Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

PPP2R1A mutations in gynaecologic cancers: functional characterization and use in the genomic classification… McConechy, Melissa 2015

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

Item Metadata

Download

Media
24-ubc_2015_may_mcconechy_melissa.pdf [ 19.01MB ]
Metadata
JSON: 24-1.0166242.json
JSON-LD: 24-1.0166242-ld.json
RDF/XML (Pretty): 24-1.0166242-rdf.xml
RDF/JSON: 24-1.0166242-rdf.json
Turtle: 24-1.0166242-turtle.txt
N-Triples: 24-1.0166242-rdf-ntriples.txt
Original Record: 24-1.0166242-source.json
Full Text
24-1.0166242-fulltext.txt
Citation
24-1.0166242.ris

Full Text

 PPP2R1A Mutations in Gynaecologic Cancers: Functional Characterization and use in the Genomic Classification of Tumours by Melissa McConechy B.Sc., Simon Fraser University, 2003A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF Doctor of Philosophy in THE FACULTY OF GRADUATE AND POSTDOCTORAL STUDIES (Pathology and Laboratory Medicine) THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver) April 2015 © Melissa McConechy, 2015 ii Abstract Endometrial carcinoma is the most common gynaecological cancer in developed countries. The current pathologic classification system for endometrial cancer lacks reproducibility, which has hampered the development of new treatment approaches for these cancers. The PP2A phosphatase complexes are responsible for regulating many cellular pathways, and may play a large role in the deregulation of endometrial cancer-associated pathways. Objectives: To determine the role of somatic PPP2R1A mutations in subtype specific classification of gynaecological tumours. In addition, mutational profiles from multiple genes will be used to improve the classification of the subtypes of endometrial carcinomas. Lastly, the functional effect of mutant PPP2R1A on PP2A subunit protein interactions will be determined, in the context of endometrial cancer cell lines. Methods: Next-generation sequencing and Sanger sequencing was used to determine the presence of mutations in endometrial and ovarian carcinomas. PPP2R1A isogenic endometrial specific cell lines were generated using somatic cell gene knockout by homologous recombination. Co-immunoprecipitation coupled to mass spectrometry was used to determine the effects of the PPP2R1A W257L mutation on the ability to interact with PP2A subunits. Results: Subtype-specific somatic PPP2R1A mutations were identified in endometrial serous carcinomas. Low-grade endometrial endometrioid carcinomas were defined by mutations in the genes: ARID1A, PTEN, PIK3CA, CTNNB1, and KRAS, whereas high-grade endometrioid also harbor TP53 mutations. Endometrial serous carcinomas harbor mutations in PPP2R1A, FBXW7, PIK3CA and TP53. Consequently, the molecular profiles proved useful in assisting classification of seven tumours with overlapping morphological features that cause irreproducibility in diagnoses. Proteomic analysis of the isogenic cell lines determined that the PPP2R1A W257L mutation disrupts the interaction with PPP2R5C and PPP2R5D B subunits. In addition, PPP2R1A mutated protein caused an increased interaction with the endogenous PP2A inhibitor SET/I2PP2A. Conclusions: The integration of mutational profiles and other genomic features will be used to improve clinical and pathological classification in endometrial tumours that are difficult to diagnose. PPP2R1A mutations are likely playing an important role in the transformation of gynaecological carcinoma, by disrupting PP2A subunit interactions with tumour suppressor functions. Increased interaction of mutant PPP2R1A with SET/I2PP2A adds another layer of iii complexity to the tumour suppressive role of PP2A. In the future, targeting the PP2A complex with novel therapeutics could provide an alternative method for treating these gyneacological cancers with poor outcomes.   iv Preface  Chapter 2 is based on a published manuscript [1].  McConechy MK, Anglesio MS, Kalloger SE, Yang W, Senz J, Chow C, Heravi-Moussavi A, Morin GB, Mes-Masson AM, Australian Ovarian Cancer Study Group, Carey MS, McAlpine JN, Kwon JS, Prentice LM, Boyd N, Shah SP, Gilks CB, Huntsman DG, "Subtype-specific mutation of PPP2R1A in endometrial and ovarian carcinomas", The Journal of Pathology, 223(5), 567-573 © [2011] Pathological Society of Great Britain and Ireland, first published by John Wiley & Sons Ltd.  I was responsible for all experiments, writing, editing and coordination of the manuscript.  MS Anglesio was a co-first author who contributed in interpretation, writing and editing of the manuscript. CB Gilks, DG Huntsman all assisted in experiment interpretation and manuscript writing. Dr. CB Gilks was also the primary pathologist to review all cases in this study. W Yang, J Senz, A Heravi-Moussavi, C Chow assisted with technical experimentation. SE Kalloger, AOCSG, AM Mes-Masson assisted with obtaining and collection of tissue. All other authors assisted in the interpretation and editing of the manuscript.  All immunohistochemistry was performed at GPEC (Genetic Pathology Evalutation Centre) at the Jack Bell Research Centre, and scored with Dr. CB Gilks.   The Ethics and certificates involved in this study are the following:  BC Cancer Agency Research Ethics Board - H05-60119 The Gynaecological Cancer Tissue Bank BC Cancer Agency Research Ethics Board - H02-61375 Immunohistochemical and Fluorescent In-Situ Hybridization (FISH) Studies of Cancer BC Cancer Agency Research Ethics Board - H08-01411 NGS in Tumours BC Cancer Agency Research Ethics Board - H09-02153 Validation Cohort Study UBC Biosafety Committee – B14-0070 Huntsman Lab Biosafety   v  Chapter 3 is mostly based on a published manuscript [2]. There are portions of this chapter that are unpublished data.  McConechy MK, Ding J, Cheang MC, Wiegand KC, Senz J, Tone AA, Yang W, Prentice LM, Tse K, Zeng T, McDonald H, Schmidt AP, Mutch DG, McAlpine JN, Hirst M, Shah SP, Lee CH, Goodfellow PJ, Gilks CB, Huntsman DG, "Use of mutation profiles to refine the classification of endometrial carcinomas", The Journal of Pathology, 228(1), 20-30 © [2012] Pathological Society of Great Britain and Ireland, first published by John Wiley & Sons Ltd.  I was responsible for all aspects of the study including sample preparation for exon capture sequencing, all validation sequencing, data analysis, writing, editing, development of all manuscript figures, and coordination of the manuscript.  J Ding was a co-first author who was responsible for the exon-capture sequencing analysis, writing and editing of the manuscript. MC Cheang performed the statistical analysis. KC Wiegand, J Senz, AA Tone, W Yang, LM Prentice, K Tse, T Zeng, H McDonald, all assisted in technical experimentation or study coordination. K Tse, T Zeng, H McDonald designed the exon capture probes and performed the capture and library sequencing.  AP Schmidt, DG Mutch, JN McAlpine, PJ Goodfellow assisted with collection of samples. SP Shah, M Hirst, CH Lee, PJ Goodfellow, CB Gilks, DG Huntsman all assisted in experimental interpretation and manuscript editing. Dr. CB Gilks and Dr. CH Lee were the primary pathologists to review all cases in this study. All immunohistochemistry was performed at GPEC (Genetic Pathology Evaluation Centre) at the Jack Bell Research Centre, and scored with Dr. CB Gilks. Dr. H Horlings assisted with the immunohistochemistry scoring for C-Myc, and Samuel Leung calculated H-scores for the C-Myc TMA analysis. I also performed the technical experiments and analysis of the additional unpublished sequencing in this chapter.  The ethics and certificates involved in this study are the following:  BC Cancer Agency Research Ethics Board - H05-60119 The Gynaecological Cancer Tissue Bank BC Cancer Agency Research Ethics Board - H02-61375 Immunohistochemical and Fluorescent vi In-Situ Hybridization (FISH) Studies of Cancer BC Cancer Agency Research Ethics Board - H08-01411 NGS in Tumours BC Cancer Agency Research Ethics Board - H09-02153 Validation Cohort Study UBC Biosafety Committee – B14-0070 Huntsman Lab Biosafety  Chapter 4 is based on a published manuscript [3].   M.K. McConechy, J Ding, J Senz, W Yang, N Melnyk, A.A. Tone, L.M. Prentice, K Wiegand, J.N. McAlpine, S.P. Shah, C-H Lee, P.J. Goodfellow, C.B. Gilks, D.G. Huntsman, "Ovarian and endometrial endometrioid carcinomas have distinct CTNNB1 and PTEN mutation profiles", Modern Pathology (2014) (27), 128-134.  Chapter 4 is an extension of the Chapter 3 study. I was responsible for all aspects of the manuscript including sample preparation for exon capture sequencing, all validation sequencing, data analysis, writing, editing development of all manuscript figures, and coordination of the manuscript.  J Ding, J Senz, W Yang, N Melnyk, A.A. Tone, L.M. Prentice, K Wiegand, performed technical experimentation or assisted with study coordination. J.N. McAlpine, S.P. Shah, C-H Lee, P.J. Goodfellow, C.B. Gilks, D.G. Huntsman assisted in experimental interpretation and manuscript editing. Dr. CB Gilks was also the primary pathologist to review all cases in this study.  The ethics and certificates involved in this study are the following:  BC Cancer Agency Research Ethics Board - H05-60119 The Gynaecological Cancer Tissue Bank BC Cancer Agency Research Ethics Board - H08-01411 NGS in Tumours BC Cancer Agency Research Ethics Board - H09-02153 Validation Cohort Study UBC Biosafety Committee – B14-0070 Huntsman Lab Biosafety     vii Chapter 5 is unpublished material, and was performed in collaboration with Dr. James Brenton’s laboratory at the University of Cambridge (Cambridge, UK). I designed and generated the PPP2R1A targeting vectors in Vancouver, and then traveled to Dr. James Brenton’s lab in Cambridge to undergo the transduction and generation of the Hec1A isogenic knockout cell line. Dr. Jian Xian provided assistance and standard protocols for the somatic cell knockout technique in Dr. James Brenton’s laboratory. The cell lines were then shipped back to Vancouver for analysis and further experiments. Natalyia Melnyk performed all FISH experiments, otherwise I designed and performed all other technical experiments and interpretation described in this Chapter.   The certificates involved in this study are the following: UBC Biosafety Committee – B14-0070 Huntsman Lab Biosafety UBC Biosafety Committee – B12-0017 PP2A protein serine/threonine phosphatase mutations in cancer. Alternative splicing and CLK inhibitors.   Chapter 6 is unpublished material, performed in collaborations with Dr. Gregg Morin’s laboratory in the Genome Sciences Centre at the BC Cancer Research Centre. I performed all technical experiments and interpretation. The mass spectrometry experiments were ongoing optimization experiments over the course of my PhD studies with assistance from Dr. Annie Moradian, Dr. Vincent Chen and Dr. Christopher Hughes. The final mass spectrometry experiments and analysis presented in this thesis were performed in collaboration with Dr. Christopher Hughes. Statistical consultation was performed with Dr. Aline Talhouk.  The certificates involved in this study are the following: UBC Biosafety Committee – B14-0070 Huntsman Lab Biosafety UBC Biosafety Committee – B12-0017 PP2A protein serine/threonine phosphatase mutations in cancer. Alternative splicing and CLK inhibitors.    viii Chapter 7 is unpublished material and is written solely by myself.   Appendix D is based on published material that is currently available as an early view online version on the website for The Journal of Pathology: Clinical Research.  McConechy MK, Hoang LN, Chi MH, Senz J, Yang W, Rozenberg N, Mackenzie R, McAlpine JN, Huntsman DG, Clarke BA, Gilks, CB, Lee CH. “In-depth molecular profiling of the biphasic components of uterine carcinosarcomas”, The Journal of Pathology: Clinical Research (2015). doi: 10.1002/cjp2.18.  Appendix D is an extension of the Chapter 2 carcinosarcoma study. I was responsible for all aspects of the manuscript including sample preparation for targeted gene sequencing, all validation sequencing, data analysis, writing, editing development of all manuscript figures, and coordination of the manuscript. L.N. Hoang is a co-first author responsible for all pathologic review of samples, coordination of the study, preparation of figures, editing and writing of the manuscript. M.H. Chi, J Senz, W Yang, N Rozenberg, R MacKenzie performed technical experimentation or assisted with study coordination. J.N. McAlpine, D.G Huntsman, B.A. Clarke C.B. Gilks, CH Lee D.G. assisted with review, editing and interpretation of data for the manuscript. CH Lee assisted with experimental design, supervision of the study, interpretation and final preparations of the manuscript including writing and editing.  The ethics and certificates involved in this study are the following:  BC Cancer Agency Research Ethics Board - H05-60119 The Gynaecological Cancer Tissue Bank BC Cancer Agency Research Ethics Board - H08-01411 NGS in Tumours BC Cancer Agency Research Ethics Board - H09-02153 Validation Cohort Study UBC Biosafety Committee – B14-0070 Huntsman Lab Biosafety    ix Table of Contents  Abstract .......................................................................................................................................... ii	  Preface ........................................................................................................................................... iv	  Table of Contents ......................................................................................................................... ix	  List of Tables ................................................................................................................................ xv	  List of Figures ............................................................................................................................ xvii	  List of Symbols ........................................................................................................................... xix	  List of Abbreviations ................................................................................................................... xx	  Glossary .................................................................................................................................... xxiv	  Acknowledgements .................................................................................................................... xxv	  Dedication ................................................................................................................................ xxvii	  Chapter 1: Introduction ................................................................................................................1	  1.1	   Endometrial Carcinoma ..................................................................................................... 1	  1.1.1	   Histopathology of Endometrial Carcinoma Subtypes ................................................. 2	  1.1.2	   Molecular Genetics of Endometrial Carcinoma ......................................................... 4	  1.1.2.1	   Endometrial Endometrioid Carcinoma (EEC) ..................................................... 4	  1.1.2.2	   High-Grade Endometrial Carcinoma: Serous, Clear Cell, Carcinosarcoma ........ 5	  1.2	   Controversies in High-Grade Endometrial Classification ................................................. 6	  1.3	   Ovarian Carcinoma ............................................................................................................ 8	  1.3.1	   Histopathology of Ovarian Carcinoma ....................................................................... 9	  1.3.2	   Molecular Genetics of Ovarian Carcinoma .............................................................. 10	  1.3.2.1	   Ovarian Endometrioid (OEC) and Clear Cell Carcinoma (OCCC) ................... 10	  1.3.2.2	   Ovarian High Grade Serous Carcinoma (HGSC) .............................................. 11	  x 1.4	   Protein Phosphatase 2A (PP2A) ...................................................................................... 12	  1.4.1	   The Role of PP2A in Cancer ..................................................................................... 16	  1.4.1.1	   The Role of the PP2A A Subunits (PPP2R1A and PPP2R1B) in Cancer ......... 18	  1.5	   Omics Technology – Advances in DNA and Protein Sequencing .................................. 20	  1.5.1	   Next-Generation Sequencing Technology ................................................................ 20	  1.5.2	   Proteomics Technology ............................................................................................ 21	  1.6	   Hypotheses ....................................................................................................................... 23	  Chapter 2: Identification of Subtype-Specific PPP2R1A Mutations in Endometrial and Ovarian Carcinomas ....................................................................................................................24	  2.1	   Introduction ...................................................................................................................... 24	  2.2	   Materials and Methods ..................................................................................................... 24	  2.2.1	   Patient Samples ......................................................................................................... 24	  2.2.2	   Whole Transcriptome Sequencing Analysis ............................................................. 25	  2.2.3	   DNA Extraction ........................................................................................................ 25	  2.2.4	   PCR and Sanger Sequencing .................................................................................... 25	  2.2.5	   Immunohistochemistry Staining ............................................................................... 26	  2.3	   Results .............................................................................................................................. 26	  2.4	   Discussion ........................................................................................................................ 30	  Chapter 3: Mutational Classification of Endometrial Carcinoma Subtypes .........................33	  3.1	   Introduction ...................................................................................................................... 33	  3.2	   Methods............................................................................................................................ 33	  3.2.1	   Patient Samples ......................................................................................................... 33	  3.2.2	   DNA Isolation ........................................................................................................... 34	  xi 3.2.3	   Exon Sequencing ...................................................................................................... 34	  3.2.4	   Bioinformatics Analysis ............................................................................................ 34	  3.2.5	   DNA Validations ...................................................................................................... 34	  3.2.6	   Targeted Fluidigm-MiSeq Sequencing of PPP2R1A, FBXW7, TP53 ...................... 35	  3.2.7	   Identifying Outlier Cases .......................................................................................... 36	  3.2.8	   TMA and Immunohistochemistry ............................................................................. 36	  3.2.9	   Microsatellite Instability (MSI) Assay ...................................................................... 37	  3.2.10	   Statistical Analysis .................................................................................................. 37	  3.3	   Results .............................................................................................................................. 38	  3.3.1	   High-Grade and Low-Grade Endometrioid Carcinomas have Similar Mutation Profiles but Differ in Frequencies of TP53 Mutations .......................................................... 38	  3.3.2	   Endometrial Serous Carcinomas Show a Distinct Mutation Profile ......................... 40	  3.3.3	   Cases with Discordant Morphological Diagnosis and Mutational Profiles .............. 42	  3.3.4	   Carcinosarcomas Show Either an Endometrioid or Serous Mutation Profile ........... 45	  3.3.5	   Mutations Involving Signalling Pathways in Endometrial Carcinomas ................... 46	  3.3.6	   Microsatellite Instability ........................................................................................... 46	  3.3.7	   Analysis of PPP2R1A and FBXW7 Mutations in Endometrial Serous Carcinomas 46	  3.4	   Discussion ........................................................................................................................ 48	  Chapter 4: Differences in the Mutation Profiles of Ovarian Endometrioid and Endometrial Endometrioid Carcinomas ..........................................................................................................55	  4.1	   Introduction ...................................................................................................................... 55	  4.2	   Materials and Methods ..................................................................................................... 55	  4.2.1	   Patient Samples ......................................................................................................... 55	  xii 4.2.2	   Mutation Analysis ..................................................................................................... 56	  4.2.3	   Bioinformatics Analysis ............................................................................................ 56	  4.2.4	   Statistical Analysis .................................................................................................... 56	  4.3	   Results .............................................................................................................................. 57	  4.4	   Discussion ........................................................................................................................ 60	  Chapter 5: Construction of Endometrial Hec1A Isogenic PPP2R1A Cell Lines ...................65	  5.1	   Introduction ...................................................................................................................... 65	  5.2	   Materials and Methods ..................................................................................................... 67	  5.2.1	   Cell Culture Maintenance ......................................................................................... 67	  5.2.2	   Creation of Somatic Cell Knockout Hec1A Cell Lines ............................................ 67	  5.2.2.1	   Cloning PPP2R1A Exon 3 and Exon 4 Regions into the Targeting Vectors pSEPT and pAAV-MCS ................................................................................................... 67	  5.2.2.2	   Generating PPP2R1A Exon 3 and 4 Hec1A KO Cell Lines .............................. 69	  5.2.2.3	   Sanger Sequencing of Hec1A Single Cell Knockout Colonies ......................... 70	  5.2.2.4	   Ion Torrent Sequencing of PPP2R1A Exon 6 in Hec1A Cell Line ................... 71	  5.2.3	   M-FISH and FISH .................................................................................................... 71	  5.2.4	   Cell Proliferation Assays .......................................................................................... 72	  5.2.5	   Migration (Scratch-Wound) Assay ........................................................................... 72	  5.3	   Results .............................................................................................................................. 73	  5.3.1	   Isolation of PPP2R1A Exon 3 and 4 Knockout Isogenic Hec1A Cell Lines ............ 73	  5.3.2	   Growth Characteristics of Isogenic Cell Lines ......................................................... 77	  5.4	   Discussion ........................................................................................................................ 83	  Chapter 6: Proteomics Analysis of PPP2R1A Mutations in Model Cell Lines ......................89	  xiii 6.1	   Introduction ...................................................................................................................... 89	  6.2	   Methods............................................................................................................................ 91	  6.2.1	   Co-Immunoprecipitation ........................................................................................... 91	  6.2.2	   Western Blots ............................................................................................................ 92	  6.2.3	   Sample Digestion for Mass Spectrometry Analysis ................................................. 92	  6.2.4	   Mass Spectrometry Data Acquisition ....................................................................... 93	  6.2.5	   Mass Spectrometry Data Analysis ............................................................................ 93	  6.3	   Results .............................................................................................................................. 96	  6.4	   Discussion ...................................................................................................................... 101	  Chapter 7: Conclusion ...............................................................................................................107	  7.1	   Overall Significance of My Thesis Research ................................................................. 107	  7.1.1	   PPP2R1A Mutations in Endometrial Carcinomas .................................................. 107	  7.1.2	   Mutational Profiling of Endometrial Carcinomas ................................................... 107	  7.1.3	   PPP2R1A Isogenic Cell Line Models ..................................................................... 108	  7.1.4	   Impact of PPP2R1A Mutations on PP2A Holoenzyme Composition .................... 109	  7.2	   Limitations of Study Designs ........................................................................................ 111	  7.3	   Challenges in Studying Protein Complexes ................................................................... 113	  7.4	   Future Directions ........................................................................................................... 114	  7.4.1	   Classification of Endometrial Carcinomas ............................................................. 114	  7.4.2	   PP2A Aberrations in Gynaecological and Breast Carcinomas ............................... 114	  7.4.3	   Proteomic Studies of Endometrial Cancer Patient Samples ................................... 119	  7.4.4	   Novel Therapeutic Options for Targeting PP2A Altered Cancers .......................... 119	  Bibliography ...............................................................................................................................122	  xiv Appendices ..................................................................................................................................144	  Appendix A Chapter 2 Supplemental Tables .......................................................................... 144	  Appendix B Chapter 3 Supplemental Materials and Methods, Tables and Figures ............... 153	  	   Materials and Methods .............................................................................................. 153	  B.1	   Supplemental Tables ................................................................................................. 157	  B.2Table B.4 All endometrial carcinoma mutation data .............................................................. 200	  Table B.5 Endometrial carcinoma mutation data with immunohistochemistry scores .......... 204	  	   Supplemental Figures ................................................................................................ 209	  B.3Figure B.1 A histogram of the probability distribution of the predicted SNV positions. ....... 209	  Figure B.2 Boxplots of the mean-coverage of each gene in the cases with and without mutations ................................................................................................................................. 210	  Appendix C Chapter 4 Supplemental Table ........................................................................... 212	  Table C.7 Ovarian endometrioid carcinoma mutation data .................................................... 215	  Appendix D Molecular Profiling of Endometrial Carcinosarcomas ....................................... 216	  	   Abstract ..................................................................................................................... 216	  D.1	   Materials and Methods .............................................................................................. 216	  D.2	   Carcinosarcoma Mutation Profiles Figure ................................................................ 219	  D.3Appendix E POLE Mutations in Endometrial Carcinomas .................................................... 220	  	   Materials and Methods .............................................................................................. 220	  E.1	   Results ....................................................................................................................... 221	  E.2Appendix F PPP2R1A-IP Mass Spectrometry Peptide Identifications .................................. 222	   xv List of Tables  Table 1.1 Heterotrimeric PP2A subunit components ................................................................... 14	  Table 2.1 PPP2R1A mutation results with frequency, coding changes, and predicted amino acid changes .......................................................................................................................................... 29	  Table 3.1 Summary of all endometrial carcinoma subtypes ......................................................... 39	  Table 3.2 The frequency of mutations within all endometrial subtypes ....................................... 39	  Table 3.3 Univariate Fisher exact test (p-values) to show significant differences between mutation profiles of each endometrial carcinoma subtypes. ......................................................... 40	  Table 3.4 Multivariable logistic regression analysis of gene mutations between endometrial carcinoma subtypes. ...................................................................................................................... 42	  Table 3.5 Outlier cases with pathological review, IHC and mutation profile. ............................. 43	  Table 4.1 Comparison of mutation frequencies in low-grade OECs and low-grade EECs .......... 58	  Table 4.2 The mutation frequencies of high-grade OECs ............................................................ 58	  Table 5.1 Homology arm primers for cloning PPP2R1A exon 3 and 4 into pSEPT .................... 68	  Table 5.2 Primer sets for amplifying PPP2R1A cDNA for Ion Torrent sequencing. ................... 71	  Table 6.1 Hec1A isogenic cell line IP TMT pools for mass spectrometry analysis ..................... 93	  Table 6.2 Mass spectrometry data at the protein level after normalization ................................ 100	  Table A.1 PPP2R1A mutations by sample ................................................................................. 144	  Table A.2 Endometrial and ovarian carcinomas dataset information ......................................... 152	  Table B.3 Exon capture probe design with genomic locations ................................................... 160	  Table B.4 All endometrial carcinoma mutation data .................................................................. 200	  Table B.5 Endometrial carcinoma mutation data with immunohistochemistry scores .............. 204	  Table B.6 All mutation table for endometrial serous targeted gene sequencing ........................ 208	  xvi Table C.7 Ovarian endometrioid carcinoma mutation data ........................................................ 215	  Table E.8 The histology distribution of POLE mutated endometrial carcinoma ....................... 221	  Table F.9 Peptide list used for PPP2R1A immunoprecipitation mass spectrometry quantitation..................................................................................................................................................... 257	   xvii List of Figures  Figure 1.1 Histopathology of endometrial carcinomas ................................................................... 3	  Figure 1.2 Histopathology of ovarian carcinomas ........................................................................ 10	  Figure 1.3 3D Structure of Protein Phosphatase 2A (PP2A) ........................................................ 13	  Figure 2.1 Schematic of mutations in PPP2R1A .......................................................................... 28	  Figure 3.1 Mutation profiles of endometrial subtypes .................................................................. 41	  Figure 3.2 A case originally diagnosed as serous carcinoma, but with an ARID1A mutation and no TP53 mutation, is a mixed low-grade endometrioid and serous carcinoma (case #1120). ..... 44	  Figure 3.3 Mutation profiles of endometrial serous carcinomas .................................................. 47	  Figure 3.4 An intermediate type of high-grade endometrial carcinomas is not encompassed in the Type 1 and Type 2 model. Problematic morphological diagnoses of high-grade endometrial carcinomas are an intermediate subtype and may be further subclassified by mutational profiles........................................................................................................................................................ 50	  Figure 3.5 Mutational analysis may be an effective tool to classify morphologically problematic cases into biologically relevant treatment groups ......................................................................... 51	  Figure 4.1 Low-grade ovarian and endometrial endometrioid mutation profiles. ........................ 59	  Figure 4.2 PI3K/AKT and WNT signaling pathways. .................................................................. 63	  Figure 5.1 pAAV-PPP2R1A exon 3 targeting vector ................................................................... 69	  Figure 5.2  pAAV-PPP2R1A exon 3 targeted Hec1A single cell colony ..................................... 70	  Figure 5.3  Sanger Sequencing Traces for Hec1A KO ................................................................. 74	  Figure 5.4  Ion torrent sequencing of the parental Hec1A cell line for W257L mutation ............ 75	  Figure 5.5  M-FISH and FISH for Hec1A cells ............................................................................ 76	  Figure 5.6 Resulting Hec1A PPP2R1A isogenic knockout cell lines ........................................... 77	  xviii Figure 5.7 Cell proliferation of Hec1A mutant/mutant expressing cell lines using the ............... 78	  Figure 5.8 Cell proliferation of Hec1A PPP2R1A isogenic cell lines with differing expression . 79	  Figure 5.9 Effect of different FBS concentration on the proliferation of Hec1A PPP2R1A ........ 80	  Figure 5.10 Images of the scratch (wound) assay over time ......................................................... 82	  Figure 5.11 Relative wound density over time of Hec1A PPP2R1A isogenic cell lines .............. 83	  Figure 6.1 Pre and post normalization of all peptides identified in the IP-MS pools. .................. 95	  Figure 6.2 PPP2R1A-IP samples with immunoblot (IB) for PPP2R1A ....................................... 96	  Figure 6.3 Quantitative analysis of PPP2R1A-IP TMT labeled pools using mass spectrometry. 99	  Figure 7.1 TCGA endometrial copy number high (serous-like) group (n=60) .......................... 116	  Figure 7.2 TCGA ovarian serous carcinoma (n=316) ................................................................ 117	  Figure 7.3 TCGA breast carcinomas (all subtypes n=466) ........................................................ 118	  Figure B.1 A histogram of the probability distribution of the predicted SNV positions. ........... 209	  Figure B.2 Boxplots of the mean-coverage of each gene in the cases with and without mutations..................................................................................................................................................... 210	  Figure B.3 Endometrial unsupervised hierarchical mutation clustering analysis aids in visualizing mutational outliers. ..................................................................................................................... 211	  Figure D.4 Uterine carcinosarcoma mutation profiles ................................................................ 219	   xix List of Symbols α alpha β beta γ gamma δ delta ε epsilon xx List of Abbreviations Ab  Antibody AKT  V-AKT Murine Thymoma Viral Oncogene Homolog (Protein Kinase B Alpha) ARID1A AT Rich Interactive Domain 1A (SWI-like) ATM  Ataxia Telangiectasia BAC  Bacterial Artificial Chromosome BRCA1 Breast Cancer 1, Early Onset BRCA2  Breast Cancer 2, Early Onset BRAF  V-Raf Muring Sarcoma Viral Oncogene Homolog B BSO  Bilateral Salpingo -Oophorectomy cDNA  copy DNA CCNE1 Cyclin E CDK12 Cyclin-dependent kinase 12 CDKN2A Cyclin-dependent kinase inhibitor 2A CHD4  Chromodomain Helicase DNA Binding Protein 4 C-MYC V-Myc Avian Myelocytomatosis Viral Oncogene Homolog CNA  Copy Number Analysis   COSMIC Catalogue of Somatic Mutations in Cancer CTNNB1 Catenin (cadherin-associated protein), Beta 1 Co-IP  Co-Immunoprecipitation DNA  Deoxyribonucleic Acid DTT   Dithiothreitol EDM  Exonuclease Domain Mutations EEC  Endometrial Endometrioid Carcinoma EEC-3  High-grade (grade 3) Endometrial Endometrioid Carcinoma EP300  E1A Binding Protein p300 ER  Estrogen Receptor ESC  Endometrial Serous Carcinoma FBS  Fetal Bovine Serum FBXW7 F-box and WD repeat domain containing 7, E3 Ubiquitin Protein Ligase FFPE  Formalin Fixed Paraffin Embedded xxi FIGO  International Federation of Gynecology and Obstetrics FISH  Fluorescence In-Situ Hybridization HEAT  Huntington Elongation-A subunit TOR HEPES 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid HGSC  High-Grade Serous Carcinoma- Ovarian Cancer HPLC  High Performance Liquid Chromatography IAA  Iodoacetamide IGBP1  Immunoglobulin (CD79A) Binding Protein 1, Protein Alpha-4 IP  Immunoprecipitation iTRAQ Isobaric tags for relative and absolute quantitation KO  Knockout KRAS  Kirsten Rat Sarcoma Viral Oncogene Homolog LOH  Loss of Heterozygosity LS  Lynch Syndrome M-FISH Metaphase Fluorescence In-Situ Hybridization MLH1  MutL Homolog 1 MMMT Malignant Mixed Müllerian Tumour MRM-MS Multiple Reaction Monitoring-Mass Spectrometry MSH2  MutS Homolog 2 MSH6  MutS Homolog 6 MSI  Microsatellite Instability OCCC  Ovarian Clear Cell Carcinoma OEC  Ovarian Endometrioid Carcinoma PARP  Poly ADP Ribose Polymerase PBS  Phosphate Buffered Saline PCR  Polymerase Chain Reaction PI3K  Phosphatidylinositol-4,5-bisphosphate 3-kinase PIK3CA Phosphatidylinositol-4,5-bisphosphate 3-kinase, Catalytic Subunit alpha PIK3R1 Phosphoinositide-3-Kinase, Regulatroy Subunit 1 alpha PIK3R2 Phosphoinositide-3-Kinase, Regulatory Subunit 2 beta PMS2  PMS2 Post Meiotic Segregation Increased 2 (S. Cerevisiae)  xxii POLE  Polymerase (DNA), epsilon, catalytic subunit PP2A  Protein Phosphatase 2A PPP2R1A Protein Phosphatase 2, Regulatory Subunit A, alpha  PPP2R1B Protein Phosphatase 2, Regulatory Subunit A, beta PPP2CA Protein Phosphatase 2, Catalytic Subunit C, alpha  PPP2CB Protein Phosphatase 2, Catalytic Subunit C, beta PPP2R2A Protein Phosphatase 2, Regulatory Subunit B, alpha PPP2R2B Protein Phosphatase 2, Regulatory Subunit B, beta PPP2R5A Protein Phosphatase 2, Regulatory Subunit B’, alpha PPP2R5B Protein Phosphatase 2, Regulatory Subunit B’, beta PPP2R5C Protein Phosphatase 2, Regulatory Subunit B’, gamma PPP2R5D Protein Phosphatase 2, Regulatory Subunit B’, delta PPP2R5E Protein Phosphatase 2, Regulatory Subunit B’, epsilon PPP2R4 Protein Phosphatase 2A Activator, Regulatory Subunit 4 (PTPA) PPME1 Protein Phosphatase Methyltransferase 1 PR  Progesterone Receptor PRM  Parallel Reaction Monitoring  pSEPT  plasmid Synthetic Exon Promoter Trap PTEN  Phosphatase and Tensin Homolog RNA  Ribonucleic Acid SCNA  Somatic Copy Number Analysis SET  SET Nuclear Oncogene SILAC  Stable Isotope Labeling of Amino Acids in Cell Culture SNP  Single Nucleotide Polymorphism SNV  Single Nucleotide Variant SP3   Single-Pot Solid-Phase-enhanced Sample Preparation SPOP  Speckle-Type POZ Protein TAH  Total Abdominal Hysterectomy TCGA  The Cancer Genome Atlas TLH  Total Laparoscopic Hysterectomy TMA  Tissue Microarray xxiii TMT  Tandem Mass Tags TP53  Tumour Protein p53 UCSC  University of California Santa Cruz WHO  World Health Organization xxiv Glossary Allelic Fraction  - sequencing term used to describe the variant read number over the total depth of that specific position. Also referred to as allele ratios. Bilateral Salpingo –Oophorectomy – removal of cervix, both fallopian tubes, and ovaries Carcinoma – cancer that arises from epithelial cells  Mesenchymal – cells that develop into lymphatic, circulatory and connective tissues such as bone and cartilage. Myometrium – the uterine smooth muscle layer that is found between the epithelium and the serosa (outer layer) of the uterus. Sarcoma – a tumour that arises from mesenchymal cells (bone, fat, cartilage, vascular cells) Simple Hysterectomy – also referred to as total hysterectomy, which includes removal of the uterus and cervix. Squamous - flat looking cells that make up an epithelium Villoglandular – papillary pattern of tumour cells  xxv Acknowledgements      I would first like to thank my supervisor Dr. David Huntsman for his utmost support throughout my PhD studies. From the start of my program, David believed in my technical abilities and allowed me to be a part of many different projects. This has allowed me to experience a wide range of research topics in science and medical fields. I feel very lucky to be a part of Dr. Huntsman’s lab and the team of OvCaRe researchers. Thank you David for allowing me to have a work/life balance, where you were so supportive of my love for volleyball, snowboarding and family life. All of our meetings over the years have been invaluable, and I appreciate your always insightful advice and support regarding my future career as a scientist.       To my second mentor, Dr. Blake Gilks, thank you very much for always taking the time from your busy schedule to talk and teach from behind the microscope. You have given me so much valuable advice over the years with respect to career and life, which I appreciate a great deal. Thank you very much for all the assistance with my manuscripts, I could not have finished these projects with out your expertise.      Thank you to all my committee members, Dr. Gregg Morin, Dr. Sohrab Shah, and Dr. Mark Carey for your valuable input on my thesis research. Your different areas of expertise have allowed me to acquire a wide range of exposure to cancer genomics, proteomics and medicine. Also thank you to Dr. Haydn Pritchard for agreeing to be my committee chair, and your respected advice as a graduate advisor. I would also like to thank Dr. Kevin Bennewith who came to my rescue to act as my committee chair on a few different occasions. I am very grateful for your help.      I would like to give additional thanks to Dr. Gregg Morin for your support and assistance in the proteomics work of my thesis. It was a bumpy road for the both of us, but overall it was an important learning experience for my science career. Thank you to all of the GSC staff that has helped me along the way: Dr. Grace Cheng, Dr. Christopher Hughes, Dr. Vincent Chen, and Dr. Annie Moradian.       The collaborations I have had over the past 5 years have meant a great deal to me. Thank you to Dr. Jessica McAlpine, Dr. Cheng-Han Lee, and Dr. Lien Hoang for allowing me to be a part of all our collaborations. I have learned so much from our interactions over the years, and I appreciate your patience and time in meetings to finish up all of our exciting projects. Also, thank you to Dr. James Brenton for allowing me to visit your laboratory, and have an amazing xxvi learning experience that was extremely important for the advancement of my science career.  Your guidance along with Dr. Jian Xian’s expertise was priceless in constructing the endometrial isogenic cell lines. Lastly, thank you to Dr. Paul Goodfellow who provided endometrial samples and guidance in this research.      Special thanks to all the members of the past and present Huntsman laboratory (Winnie, Janine, Alicia, Clara, Sarah, Samantha, Niki, Dawn, Michelle, Natalyia, Genny, Leah, Rob, Tony, Hugo, Mike, Kim, Satoshi), CTAG laboratory (Peggy, Julie, Ying, Amy, Julie, Sarah, Sylvia), GPEC laboratory (Christine, Sherman, Angela), the OvCaRe team (Jessica, Anna, Margaret, Karen), the Shah lab (Ali, Jiarui, Karey, Yikan) and the Aparicio lab (Adrian, Cherry). Thank you to all of your for assistance in all aspect of my research over the last five years. Special thanks to Dr. Alicia Tone, Winnie Yang, Janine Senz, Natalyia Meylnk, and Samantha Hansford for all the laughs, talks and shoulder to cry on. Also, thank you to Jiarui Ding who was instrumental for the completion of the exon capture sequencing analysis. I could not have done any of it without the support of so many great people.     I have been very fortunate over the period of my PhD research to be funded by many scholarships, travel funds and awards. I would like to thank the following organizations for their support of my PhD career: UBC, CIHR Masters and CIHR PhD Fellowships, Canadian Cancer Society, Ovarian Cancer Canada, and the Evelyn Martin Memorial Fund.    Last but not least, I would like to thank my family and friends for their utmost support in my Ph.D. endeavor. A very special thanks to my parents who have always supported me in my independence and academics. Dad and Mom, thank you for doing whatever you did to make me love school and want to further my education. Dad, thank you for attending my defence that meant a great deal to me. Thank you to my sisters and brothers for your support, especially to Kirstin who has been my number one supporter since the beginning of the crazy thought of starting a PhD. Finally, thank you to Thibault who has been my rock in keeping me grounded over the last few years. Thank you for your love and patience with my long workdays and studying, and especially for keeping me company at the lab during a few late-night experiments. You mean the world to me, and hope we can continue to support each other in our life adventures. I am very grateful for everyone’s support. xxvii Dedication I dedicate this to my family, with special dedication to my sister Kirstin, and my supportive partner Thibault.1 Chapter 1: Introduction   1.1 Endometrial Carcinoma      Endometrial cancer, or cancer of the uterus, is the most common gyneacological malignancy in developed countries, and is the sixth most commonly diagnosed cancer in women worldwide [4]. In 2011, an estimated 143,000 new cases of endometrial cancer were reported, causing death in about 33,000 women in developed countries, and this incidence is rising [4, 5]. This is likely due to a variety of factors including increased worldwide obesity, the growing aging population, and prior tamoxifen use [6]. The majority of endometrial cancers occur in post-menopausal women; 90% being sporadic and 10% with a hereditary nature [7]. Endometrial tumours are diagnosed as multiple histopathologic subtypes: endometrioid, serous, carcinosarcoma, clear cell and mucinous carcinoma. In 1983, Bokhman proposed the classical dualistic histopathological model separating this cancer into two broad types; estrogen-dependent endometrioid endometrial carcinoma (type 1), and non-endometrioid endometrial carcinoma (type II) [8]. Approximately 70-80% of endometrial tumours are diagnosed as low-grade, low-stage endometrial endometrioid carcinomas (EEC), which are generally confined to the body of the uterus, and can be cured by a simple hysterectomy [9]. In contrast, non-endometrioid carcinomas constitute 10-20% of the histopathological subtypes serous, clear cell and carcinosarcomas. These cancers are responsible for a disproportional majority of endometrial cancer deaths, as they are generally high-grade, aggressive, and present in older post menopausal women [10]. Endometrial serous carcinomas (ESC) also referred to as high-grade endometrial serous carcinoma, uterine papillary serous carcinoma, or uterine serous carcinoma, are aggressive and invasive, and tend to relapse with metastatic disease [11-13]. The 5-year overall survival for women with serous carcinoma was 45.9%, and those with stage III and stage IV disease were 37.3% and 19.9% respectively [14, 15]. Carcinosarcomas, also referred to as MMMT, and clear cell carcinomas are extremely rare high-grade endometrial subtypes, accounting for less than 5% of all endometrial cancers that tend to recur locally and can metastasize to distant areas [16]. For the purpose of this body of research, I will focus mainly on the histological subtypes endometrioid and serous carcinomas with brief mentions of carcinosarcomas.  2      In the current clinical setting, low-stage, low-grade EEC patients undergo hysterectomy by total laparoscopic hysterectomy (TLH) or total abdominal hysterectomy (TAH) with bilateral salpingo-oophorectomy (BSO). The hysterectomy and BSO involves removal of the uterus, cervix, fallopian tubes and ovaries. Histopathological diagnosis is extremely important in clinical management, as adjuvant radiation and chemotherapy regimes will be dependent on cell type (endometrioid or non-endometrioid), grade and stage. EECs with greater than stage 1B and grade 3 nuclei will generally undergo adjuvant radiotherapy and chemotherapy, and is always administered to women diagnosed with serous, carcinosarcomas and clear cell carcinomas with myometrial invasion. This will include cycles of chemotherapeutic agents carboplatin and paclitaxel, as this treatment has been shown to be efficacious with low toxicity in advanced or recurrent endometrial carcinoma [17, 18].   1.1.1 Histopathology of Endometrial Carcinoma Subtypes       Endometrial carcinoma originates from the epithelial lining of the uterus. In premenopausal women, endometrial glandular cells undergo a monthly cycle (average of 28 days) consisting of the menstrual phase, followed by proliferative and secretory phases [19]. The histopathologic definitions of endometrial tumours are guided by the international WHO (World Health Organization) classification of female reproductive organs [20] and FIGO (International Federation of Gynecology and Obstetrics) [21]. The morphology of low-grade EEC is distinct from that of high-grade serous carcinoma [22]. The epithelial tumour cells of grade 1 EEC are well differentiated with apparent glandular or villoglandular structure with less that 5% solid structure [20] (Figure 1.1). Generally, grade 2 EECs range from 6-50% solid growth patterns with moderate differentiation patterns. Tumour grade increases to grade 3 with the presence of more than 50% solid growth and cellular atypia, including atypical pleomorphic nuclei with poor differentiation.       Endometrial serous tumours are all grade 3 neoplasms with complex papillary and glandular structures, often with solid growth, and nuclear pleomorphism [20]. Serous carcinomas are characterized by the presence of many mitotic figures per field of view, and large prominent nucleoli (Figure 1.1). Epithelial cells in the myometrium indicate invasion, however the presence 3 of metastatic cells to sites outside the uterus may be evident even with lack of myometrial invasion [20].        High-grade endometrial adenocarcinomas can have overlapping morphological features, consequently making a pathologist’s diagnoses challenging [22, 23]. Some serous tumours exhibit well-differentiated glandular structures that are characteristic of endometrioid    Figure 1.1 Histopathology of endometrial carcinomas A. H&E (Hematoxylin and Eosin) staining of a low-grade endometrial endometrioid carcinoma with glandular structures. B. H&E staining of an endometrial serous carcinoma with solid pattern, pleomorphic nuclei and many mitotic figures (arrow).   carcinomas, and some high-grade endometrioid tumours have serous-like solid papillary growth patterns with a high mitotic index [22]. The apparent ambiguity of differentiation associated histologic features can make the process of differentiating serous from endometrioid cancers extremely challenging, which can ultimately affect the clinical management of patients.  This is particularly difficult in cases with mixed histologic features (see Chapter 3 Figure 3.2 and Table 3.5).        Carcinosarcomas are biphasic high-grade tumours with carcinomatous (epithelial) and sarcomatous (mesenchymal) components, which can be found as tumours with separate distinct entities or with mixed cell-types [24]. Histologically, the epithelial component can appear with endometrioid or serous morphology, and the sarcomatous can include homologous and heterologous elements. Homologous elements include cell types that are derived from uterine   A B 4 tissues, and heterologous elements include cell types derived outside the uterus (cartilage, skeletal muscle) such as rhabdomyosarcoma, chondrosarcoma, and osteosarcoma [25].   1.1.2 Molecular Genetics of Endometrial Carcinoma  1.1.2.1 Endometrial Endometrioid Carcinoma (EEC)      Endometrial endometrioid carcinomas have been molecularly characterized using immunohistochemistry and DNA sequencing, often on a small scale, by observing only one to two proteins or genes at a time. However, advances in next-generation sequencing technology have allowed investigation of whole genomes and targeted gene panel sequencing. EECs are characterized by microsatellite instability (MSI) (17-26%) [26], PI3K pathway alterations (80%) [27, 28], ARID1A (40%), KRAS (8-40%), FGFR2 (5-16%) and CTNNB1 (2-45%) mutations [2, 29, 30]. PTEN is the most frequently altered gene in EEC (26-80%); mostly due to mutations, rare copy number alteration events, deleted by promoter methylation, or protein degradation to ultimately cause activation of the PI3K pathway [31]. In the TCGA data set, PTEN gene abnormalities were seen in 66% of cases (153/232); of those 147/153 (96%) harbor mutations, 4/153 (3%) with homozygous deletions, 2/153 (1.3%) with homozygous deletions and concurrent mutations, 1/153 with an amplification alteration [32, 33]. Overall, EECs also have a high frequency (80%) of activating mutations in additional PI3K pathway genes; PIK3CA (30-50%), PIK3R1 (20-40%), PIK3R2 (5%) [27]. Mutations in PIK3CA and PIK3R1 are generally mutually exclusive, however are often found concurrently with PTEN alterations [27, 28]. ARID1A mutations, a component of the SWI/SNF chromatin-remodeling complex, result in BAF250a protein loss, and are frequently found in low-grade and high-grade EEC [34, 35]. Mutations in TP53 are infrequent in low-grade EECs, however high-grade EECs do harbor a higher frequency of TP53 alterations.        The autosomal dominant hereditary condition, Lynch syndrome (LS), predisposes about 3% of women for EEC, and about 3% of all colorectal cancers [36, 37]. Women with LS have a 40-60% lifetime risk of acquiring endometrial cancer, and 10-12% risk for ovarian cancer [38], which is caused by germline mutations in mismatch repair (MMR) genes (MLH1, MSH2, MSH6, and PMS2) [39]. This genetic defect results in tumour microsatellite instability (MSI), and can be 5 detected by immunohistochemistry or DNA-based MSI analysis of tumours and germline tissue [39]. Clinical testing is performed on patients diagnosed with EEC, and can have benefits for both affected individuals, and at-risk family members with potential MMR germline mutations [38].  1.1.2.2 High-Grade Endometrial Carcinoma: Serous, Clear Cell, Carcinosarcoma      The endometrial high-grade subtypes characteristically do not harbor the same high frequency of mutations found in EECs, and are molecularly different than EECs [40]. Endometrial serous carcinomas are often aneuploid, with fewer mutational aberrations than EECs [41].  TP53 (50-90%) mutations are the most frequent molecular alteration, with PPP2R1A (20-40%) and FBXW7 (20%) mutations also present in a large proportion of serous carcinomas [1, 2, 29, 42]. The first exome sequencing of serous carcinomas identified mutations harbored in chromatin remodeling genes (ARID1A, EP300, CHD4) and ubiquitin ligase complex genes (FBXW7, SPOP) [43]. Most endometrial serous carcinomas exhibit high protein expression of p16, encoded by the gene CDKN2A, which is distinct from EEC [23].  This is also a distinct feature compared to ovarian high-grade serous carcinoma, as CDKN2A is frequently downregulated [44]. Conversely, Cyclin E (CCNE1) is overexpressed in a subset of poorly differentiated endometrial carcinomas, as measured by IHC [45], as well as exhibiting copy number amplifications [46], which is also a similar feature of ovarian high-grade serous carcinomas.       Endometrial clear cell carcinomas (CCC) are extremely rare tumours; therefore few molecular studies have been completed. Most of what is known about these rare tumours comes from observations of ovarian clear cell carcinomas (OCCC). Similar to OCCC, ARID1A and PIK3CA are frequently mutated [47, 48]. However in contrast to OCCC, TP53 (33%) mutations can be identified in endometrial CCC. During my PhD studies, I assisted with a published study wherein 14 prototypical endometrial CCC tumours were sequenced with a targeted endometrial specific gene panel. Mutations were detected in genes involved in chromatin remodeling and transcriptional regulation (ARID1A, ZFHX3, TSPYL2), and ubiquitin-mediated proteolysis (SPOP and FBXW7) [49]. Furthermore, mutations in TP53 (29%) and PPP2R1A (20%) were discovered, indicating these tumours have a mutational profile closely related to endometrial serous carcinomas.  6       Few comprehensive multi-gene studies have been performed to detect molecular alterations in carcinosarcomas, however low frequency mutations have been identified in KRAS and PIK3CA [50]. Defective MMR has been identified in 20% of carcinosarcomas, and a high frequency of TP53 aberrations in both the carcinoma (68%) and sarcoma (64%) components [51]. Based on molecular alterations, Taylor et al. and others, has provided evidence to suggest that carcinosarcomas are of monoclonal origin; therefore proposing that the sarcomatous component is derived from the carcinomatous component through trans-differentiation [50]. This theory is still yet to be fully established, as more comprehensive molecular studies are needed.       I also performed targeted sequencing of 27 genes in 30 cases of endometrial carcinosarcomas [52] (Appendix D) as an extension to the work presented in Chapter 3. This involved sequencing the carcinoma and sarcoma components as separate entities, as well as metastatic sites when available. Sequencing mutational analysis illustrates TP53 as the most frequently mutated gene in 24/30 (80%) carcinosarcomas, yet genes in the PIK3CA pathway (PTEN, PIK3CA, PIK3R1, and PIK3R2) were mutated in 20 of 30 (67%) of cases. The patterns of mutations can be categorized as serous-like mutation profiles, and endometrioid-like mutation profiles (Appendix D Figure D.4), which supports the observations in Chapter 3 of this thesis [2]. These mutation patterns also provides additional evidence to support the monoclonal theory of carcinosarcomas; the theory of trans-differentiation of the carcinoma to give rise to the sarcoma component of the tumour.  1.2 Controversies in High-Grade Endometrial Classification      Current endometrial classification is based solely on histological subtypes [53]. Although this classification works well for low-grade EECs, for high-grade subtypes (grade 3 EECs and serous carcinomas) there is poor pathological interobserver agreement rendering it of dubious utility [54]. The current grading system is also irreproducible, and there are inconsistencies in demonstrating its prognostic relevance. Serous carcinomas can often harbor endometrioid-like features and vise versa, which leads to some of this discordance. Due to the aggressive nature of serous carcinomas, misdiagnoses could change clinical management of the patient; therefore improved classification is much needed. The data presented in Chapter 3 suggests that mutational 7 profiles may provide useful information for improving the classification of difficult to diagnose endometrial carcinomas. Furthermore, an extension of my work presented in Chapter 3 attempted to correlate histotype and genotype in a subset of these problematic diagnoses [55]. Based purely on morphological cell type, the interobserver concordance of the diagnoses was very poor (Κ value =0.55). However, when immunostaining and genotype were considered, diagnostic reproducibility improved to Κ=0.68. This study shows that in the future, molecular and histological features will be needed for improving the diagnoses of difficult endometrial cases.        Many of the core driver mutations in endometrial cancers have been identified; The Cancer Genome Atlas (TCGA) project acquired 373 endometrial tumours to improve the definitions of endometrial genomes. Based on full exome sequencing and SNP arrays, mutations and copy number alterations (CNAs) were identified, leading to the classification of 4 categories of tumours: POLE ultra-mutated (232  ×  10−6 mutations per Mb), MSI hypermutated (18  ×  10−6 mutations per Mb) (most with MLH1 promoter methylation), copy-number low, and copy-number high [42]. Low-grade EECs were categorized into POLE ultra-mutated, MSI hypermutated, and copy-number low groups. Pathologically defined high-grade EECs were classified mostly into the POLE ultra-mutated and copy-number high categories, whereas serous carcinomas were placed primarily in the copy-number high group. Although this analysis did not result in the discovery of novel recurrently mutated genes, the POLE mutated tumours, which have a unique clinical phenotype is of great interest. The distribution of other known mutations in the TCGA defines subtypes of endometrial cancer [42] along with my own work of this disease [2], which predated TCGA is discussed in Chapter 4.       POLE encodes for the catalytic domain of DNA polymerase ε, with mutations identified in the protein exonuclease domain [42, 56, 57]. DNA polymerase ε is involved in DNA replication and repair; therefore exonuclease domain mutations (EDM) may cause inefficient proofreading for DNA replication and repair, thus resulting in the ultra-mutated phenotype. Additional studies and TCGA have identified the presence of POLE EDM mutations in 7-15% of EECs [42, 57, 58] with a frequency of 7-11%. Interestingly, in the TCGA cohort, patients with POLE mutations were shown to have an improved progression-free survival [42].  In a recent study by Meng et al., POLE mutations were identified in 15% of grade 3 EECs, and when data was analyzed 8 together with TCGA data, these patients were also significantly associated with improved progression-free survival [57]. Conversely, a separate study of 544 tumours, exhibited only 5.6% of EECs with POLE mutations and no improvement in progression-free survival [56]. To validate these observations in a large independent cohort of endometrial carcinomas, I have sequenced 406 tumours to identify 39 (9.6%) POLE somatic EDM mutations (Table E.8 (Appendix E)). Future analysis of this targeted sequencing study will be used to associate POLE mutations with progression-free survival, in order to verify the prognostic use of POLE mutation status and aid in classifying this group of endometrial patients. In addition, we will determine if POLE mutations are markers for overall good response to conventional treatment, or if this group of patients does well without the need for adjuvant treatment. This distinction is key to determining the clinical relevance of this discovery.  1.3 Ovarian Carcinoma      Epithelial ovarian carcinoma is the leading cause of gynaecological cancer deaths, and is the sixth leading cause of death of women in developed countries [4]. Similar to endometrial carcinomas, ovarian carcinomas have been histologically typed into two main groups: Type 1 consists of low-grade serous, endometrioid, clear cell, and Type II consists of high-grade serous carcinomas, and carcinosarcomas [59]. In contrast to uterine carcinomas, high-grade serous ovarian carcinoma (HGSC) is the most common ovarian subtype. HGSC is aggressive with metastatic spread to the peritoneum, and is often diagnosed late stage, thus 70% of patients will die of their disease. Original studies described ovarian cancer as a single disease, however today the view is that ovarian cancer encompasses many diseases, and may not originate from ovarian cells [60]. The cell of origin for HGSC was thought to be the ovarian surface epithelium; however there is a growing body of evidence to suggest the cell of origin is the fallopian tube epithelium [61]. In this hypothesis, the secretory epithelial cells of the distal fallopian tube (the fimbria) acquire TP53 mutations then transform into serous tubal intraepithelial carcinoma (STICs), which shed onto the surface of the ovary and into the peritoneum [62]. These transformed cells attach to the peritoneal walls, the omentum, the fallopian tubes, and the surface of the ovary; therefore this disease is also referred to as high-grade pelvic serous carcinoma [63]. These tumours are usually treated with aggressive surgery to optimally de-bulk the patient of any residual tumour, and then followed by platinum-taxane chemotherapy. Unfortunately, platinum 9 resistance will be acquired in about 25% of all patients within six months, and in most other patients after subsequent courses of treatment [64]. New treatment options for HGSC, such as PARP inhibitors, are being tested in clinical trials.        Type I endometrioid and clear cell carcinomas account for only 15-20% of all ovarian cancers. Ovarian clear cell carcinomas (OCCC) diagnosed at a high stage have an extremely poor prognosis with the second leading cause of death from ovarian cancer [65], whereas ovarian endometrioid tumour (OEC) patients generally do quite well with conventional treatment [66]. There is now considerable evidence to suggest these tumours develop from an endometriotic cyst and not from ovarian surface epithelial cells [60]. The most accepted theory is that retrograde menstruation accounts for endometriosis, which is deposited in and on the ovary; therefore these cells would originate from the uterine epithelium [67]. This theory is supported by the morphological and molecular similarities of the tumours found in the uterus and the ovary.  1.3.1 Histopathology of Ovarian Carcinoma       Similar to endometrial tumours, ovarian tumours are also classified by the WHO and FIGO systems [20]. Ovarian endometrioid and clear cell carcinomas are morphologically similar to their endometrial counterparts, due to a common endometrial cell of origin, and are also similarly graded. These tumours harbor arrangements of glandular structures with villoglandular patterns and squamous differentiation in about 30-50% of cases (Figure 1.2) [68].  OCCC is histopathologically characterized by the presence of clear cells and “hobnail” cells [20].       High-grade serous carcinoma (HGSC), the most common type of epithelial ovarian carcinoma, is composed of solid papillary and glandular structures with often large, unusual-looking nuclei with prominent nucleoli (Figure 1.2). Mitotic and atypical mitotic figures are commonly found in large numbers within a field of view.   10  Figure 1.2 Histopathology of ovarian carcinomas A. H&E of ovarian clear cell carcinoma B. H&E of ovarian endometrioid carcinoma. C. Ovarian high-grade serous carcinoma       The histological classification of ovarian cancer is now generally reproducible between pathologists, unlike high-grade endometrial carcinomas. Nevertheless, research and clinical management have historically lumped this disease into one uniform ovarian disease classification. The development of prognostic biomarkers (immunohistochemistry markers or mutations) has been unsuccessful; therefore stratification of the ovarian subtypes in research studies and clinical trials is needed for improvement in clinical treatment and care [69].  1.3.2 Molecular Genetics of Ovarian Carcinoma  1.3.2.1 Ovarian Endometrioid (OEC) and Clear Cell Carcinoma (OCCC)      Two landmark studies using whole exome and transcriptome sequencing discovered inactivating mutations in ARID1A in about 50% of OCCCs [70, 71]. About 7% of OCCCs also harbor mutations in PPP2R1A, which was the first study to report these mutations in ovarian cancer. The PI3K pathway is mutated at a high frequency, with mutations found in PIK3CA (50%) and deletion of PTEN (20%) [72, 73]. The PI3K pathway is also mutated in OECs (20%), along with the Wnt pathway [74], where the later is rare in OCCCs. OECs harbor mutations in CTNNB1 (40%), which encodes for the protein β-catenin, and are associated with squamous differentiation, low tumour grade and favorable outcome [75].       Early molecular and histopathology studies, gave evidence to suggest an association of endometriosis with OEC and OCCC, by exhibiting LOH in the same chromosomal regions of the tumour and adjacent endometriosis [76]. Additionally, the discovery of ARID1A mutations in 11 clear cell carcinoma cells and adjacent atypical endometriosis, thus provides further evidence that endometriosis may be the precursor of OCCC [71]. Although OEC and OCCC may be derived from similar endometriotic precursors, these tumours are molecularly different, suggesting they evolve by different mechanisms. Genomic studies of these two cancers and their potential cells of origin are currently underway in our laboratory. The precursor of endometriosis is the endometrial epithelium; therefore it has been hypothesized that OEC and EEC have similar molecular features [29, 77]. In Chapter 4, I show the results of sequencing a panel of genes in endometrial endometrioid carcinoma and ovarian endometrioid carcinoma to determine if there are differences in mutation profiles.   1.3.2.2 Ovarian High Grade Serous Carcinoma (HGSC)      The mutation spectrum of HGSC is sparse, however these tumours are genomically unstable due to a high number of somatic copy number alterations (SCNA) [44]. It is striking that TP53 mutations (>95%) are ubiquitous in HGSC, and harbors the highest frequency of TP53 mutations found in any type of cancer [63]. Inactivation of BRCA1 and BRCA2 by mutation or hypermethylation of the promoter region has been demonstrated in about 30-50% of sporadic HGSC [44, 60], however is a characterizing feature of hereditary HGSC [78].        The TCGA analysis of HGSC identified previously described CNAs in CCNE1, MYC, MAPK1 and KRAS. Low frequency recurrent somatic mutations were also demonstrated in the genes NF1, BRCA1, BRCA2, RB1 and CDK12 [44]. Pathway analysis identified alterations in RB (67%), PI3K/RAS (45%), NOTCH (22%), FOXM1 (84%) signaling, and alterations in homologous recombination (HR) proteins (51%) [44].  These deregulated pathways could potentially be targeted for improved treatment options for HGSC patients; hence is a focus of current research and clinical trials. An example of this research is the treatment of HGSC patients that are altered by HR defects, such as inactivation of BRCA1 and BRCA2 with PARP inhibitors in combination with DNA alkylators, which is considered a synthetic lethal treatment approach [79]. PARP inhibitors target DNA repair mechanisms, thus the cells that harbor homologous recombination defects have an inability to repair double stranded DNA breaks, and the cells undergo cell death [80]. Unfortunately, this targeted treatment can eventually drive 12 genetic reversion in BRCA-associated ovarian cancer causing drug resistance and tumour recurrence by [81].   1.4 Protein Phosphatase 2A (PP2A)      The heterotrimeric protein phosphatase 2A (PP2A) complex is the most abundant serine/threonine protein phosphatase complex, making up 1% of all cellular proteins [82]. It is a promiscuous protein complex involved in numerous cellular processes such as differentiation, development and growth [83, 84]. The core function of PP2A is to remove phosphate groups on signaling protein substrates, that may be activating or deactivating, which is a fundamental regulatory biological mechanism [85]. The core PP2A enzyme is composed of a 65kDa scaffolding A subunit, and a 36kDa catalytic C subunit. The third regulatory B subunit interacts with the core enzyme, to construct the fully functional heterotrimeric PP2A holoenzyme complex, and is the key component for substrate-specificity, cellular functions, and localization (Figure 1.3) [86]. The scaffolding A subunit is encoded by two isoforms: protein phosphatase regulatory Aα (PPP2R1A) or Aβ (PPP2R1B) genes which are 86% identical [87]. Studies have established that PPP2R1A is ubiquitously expressed, is 40 times more abundant than PPP2R1B [87], and represents about 0.1% of total protein in the cell [88]. The catalytic C subunit is also encoded by two isoforms: protein phosphatase catalytic 2A or 2B (PPP2CA and PPP2CB) [89]. The regulatory B subunits include about 15 different family members with multiple isoforms (Table 1.1) that could potentially compose around 200 different PP2A enzyme complexes [82, 90].  13 Figure 1.3 3D Structure of Protein Phosphatase 2A (PP2A) A. The blue ribbon is the scaffold A subunit (PPP2R1A), the green ribbon is the catalyticC subunit (PPP2CA), and the red ribbon structure is the regulatory B55α subunit (PPP2R2A), as assessed from crystal structures. The orientation and colours were adjusted from the original PDB (Protein Data Bank) file (3DW8.pdb) [91]. B. Cartoon of PP2A subunits with choices of  four different B subunit family members. The ability of each B subunit (including Striatin) to bind to the A subunit will compose a unique PP2A holoenzyme.  A"C"B"B"B’"B’’"Stria+n"A"B"14       Table 1.1 Heterotrimeric PP2A subunit components       The structure of PPP2R1A is composed of a curved, hook-like helical structure with 15 HEAT repeat motifs (Huntingtin-Elongation-A subunit-TOR; Huntingtin elongation factor 3, A subunit of protein phosphatase 2A, Target of Rapamycin 1) [92]. Each repeat is defined by two alpha helices that are connected by an intra-repeat loop [93]. Crystal structure studies (Figure 1.3) demonstrate that the HEAT repeats 2-8 (N-terminus) directly interacts with regulatory B subunits through hydrogen bonds [86, 94]. The A subunit HEAT repeats 11-15 (C-terminus) interacts with PP2A catalytic C subunits through hydrogen bonds and hydrophobic interactions Gene Names Additional Name Family Protein Size (kDa) Scaffold A Subunit    PPP2R1A, alpha isoform PR65a, Aα A 65 PPP2R1B, beta isoform PR65b, Aβ A 65     Catalytic C Subunit    PPP2CA, alpha isoform PP2Acα, Cα C 36 PPP2CB, beta isoform PP2Acβ, Cβ C 36     Regulatory B Subunit    PPP2R2A, B55 alpha isoform PR55α, B55α, Bα B 55 PPP2R2B, B55 beta isoform PR55β, B55β, Bβ B 55 PPP2R2C, B55 gamma isoform PR55γ, B55γ, Bγ B 55 PPP2R2D, B55 delta isoform PR55δ, B55δ, Bδ B 55     PPP2R5A B56 alpha isoform PR61α, B56α, B'α B' 56 PPP2R5B B56 beta isoform PR61β, B55β, B'β B' 56 PPP2R5C B56 gamma isoform PR61γ, B56γ, B'γ B' 56 PPP2R5D B56 delta isoform PR61δ, B56δ, B'δ B' 56 PPP2R5E B56 epsilon isoform PR61ε, B56ε, B'ε B' 56     PPP2R3A PR130, B''α1 B'' 130 PPP2R3A PR72, B''α1 B'' 72 PPP2R3B PR70/48, B''β1 B'' 70/48 PPP2R3B PR70, B''β2 B'' 70 PPP2R3C G5PR, B''γ B''      PPP2R4 PR53, PTPA  53 STRN PR110, Striatin  110 STRN3 PR93, SG2NA  93 15 [86, 94]. Each regulatory B subunit has a unique structure, and may interact with the A and C subunits differently, enabling regulation of the high complexity of PP2A functions. Through a large mass spectrometry analysis of PP2A interactions, researchers discovered that additional proteins can replace B subunits, and C subunits can form heterodimers with proteins other than with the A subunit [95].       PP2A also plays a major role in mitosis and cell cycle regulation. The PP2A-B56δ complex has demonstrated an ability to dephosphorylate CDC25C, another major phosphatase regulator of mitosis, which prevents 14-3-3 release to ultimately trigger cells to enter mitosis [96]. At mitotic exit, the PP2A-B56δ complex is also responsible for the dephosphorylation of CDC25C to repress its phosphatase activity on CDK1, thus causing exit from mitosis. If B56δ is repressed, then CDC25C is hyperphosphorylated causing activation and dephosphorylation of CDK1 leading to a delay in mitotic exit [97]. In addition, the protein kinase Gwl (Greatwall) will associate and inhibit PP2A phosphatase activity in order to promote entry into mitosis [98]. Furthermore, PP2A-B56γ3 exhibits localization to the nucleus to dephosphorylate the p27 protein, which regulates cell cycle entry by delaying G1 to S phase [99]. Overexpression of B56γ3 increases its nuclear localization and delays G1 to S, thus decreasing cell proliferation. Conversely, knockdown of B56γ3 increases the G1 to S transition and increases cell proliferation. An elegant in vitro study using live-cell imaging demonstrated that PP2A-B55α is also important in regulating mitotic cell exit. Knockdown of B55α with siRNA significantly slowed mitotic cell cycle exit, delayed disassembly of spindle-pole-associated microtubules, and delayed post-mitotic chromosome decondensation [100]. These observations may be linked to the mitotic regulator kinase MASTL, which negatively regulates PP2A-B55α to delay mitotic exit. However, when CDK1 and MASTL are inhibited, PP2A-B55α and PP2A-B55δ will be activated to dephosphorylate substrates causing transition into mitotic exit [101]. PP2A is also implicated in the tight regulation of the cohesion complex which is important at centromeres for sister chromatid cohesion during mitosis and meiosis. PP2A was found to co-localize with the shugoshin complex at centromeres to ultimately de-phosphorylate cohesion which helps protect the protein complex [102]. This is important as phosphorylation of cohesion promotes dissociation from chromosomes, and shugoshin protects the centromere until the kinetochores 16 are ready to be captured by spindles in tight spatial and temporal regulation. Overall, distinctive PP2A complexes are crucial for the different stages of cell cycle regulation.  1.4.1 The Role of PP2A in Cancer      PP2A was first described as a tumour suppressor using the selective inhibitor okadaic acid, which leads to the formation of tumours in mice [103, 104]. However, other studies have found that PP2A is required for growth and survival, thus it is not a typical tumour suppressor protein [105]. Considering the PP2A holoenzyme is a multi-protein complex encoded by hundreds of PP2A molecules, this may not be a simple prototypical classification of a tumour suppressor or oncogene protein. Tissue specificity, cellular localization and substrate specificity may contribute to PP2A complex’s functional role as a tumour suppressor or oncogene in different cellular contexts.       Tumour promoting virus proteins are able to alter PP2A activity in multiple ways; by mimicking B subunits and enabling binding to the A-C dimer complex, and altering PP2A substrate specificity by interacting directly with B subunits in the PP2A complex. It has been established that the polyoma small T (pyST), polyoma middle T (pyMT) antigen, simian virus SV40 small T (SV40ST), and E4orf4 antigens can form complexes with PP2A [106, 107]. Complete inhibition of PP2A activity leads to cellular death; therefore viral antigens control PP2A by inhibiting interaction with substrates, and altering targets for dephosphorylation. SV40ST binds PP2A to inhibit its phosphatase activity, thereby causing cellular transformation by activating specific cellular pathways [108]. Specifically, SV40ST was shown to interact with B56γ PP2A complexes, which drives cellular transformation and tumour formation [109]. In a later study, suppression of the B subunits B56α (PPP2R5A), B56γ (PPP2R5C), PR72/PR130 (PPP2R3A), and PTPA recapitulated the SV40ST cell transformation phenotype [110]. Deregulation of phosphatase activity induced by down regulating some of these PP2A B subunits, caused signaling changes in a subset of the Wnt, c-Myc and PI3K/Akt pathway to ultimately assist in transformation. An interesting study deciphered the crystal structure of B55α, and was subsequently able to demonstrate that E4orf4 binds to the substrate groove of B55α (PPP2R2A) to inhibit a specific interaction with p107 (RBL1) causing a dephosphorylation blockade. When high expression levels of E4orf4 infection are present, this causes an inability of 17 cell cycle progression substrates to be dephosphorylated by PP2A-B55α, thus causing cell death. Albeit at lower levels, E4orf4 recruits the substrates ASF/SF2/SRSF1 to enhance adenoviral replication [111].       PP2A B55 complexes have been implicated as partaking in both activating and inhibiting effects on the MAPK pathway, which is cell-type specific [82]. Dephosphorylation of the genes ERK, RAF1, KSR1, and c-SRC causes signaling in the MAPK pathway [112, 113], which is often deregulated in many cancers. Recently, the PP2A-B55α (PPP2R2A) complex was shown to specifically dephosphorylate the c-Jun T239 residue, which promotes binding with the AP-1 transcription complex to regulate tumour migration and invasion [114]. Additionally, in AML (Acute Myeloid Leukemia) patients, the protein levels of B55α were decreased compared to normal cells, and correlate with an increase in AKT T308 phosphorylation levels. The decreased B55α levels and increased AKT signaling was an adverse prognostic factor for these patients [115]. PP2A-B55β (PPP2R2B) targets cyclin E for dephosphorylation, which stabilizes the protein and protects it from ubiquitin-mediated degradation [116]. Thus, decreased levels of B55β cause overexpression of the cyclin E protein [116], which is correlated with aggressive tumours and poor patient outcome [117].        Functional studies of PP2A suggest that loss of the B56 (PR61) family is the most relevant event for tumourigenesis, however this is likely context dependent [82]. PP2A is important in controlling the accumulation of c-Myc, which is dysregulated in many cancers [118]. PP2A dephosphorylates the c-Myc S62 residue, resulting in protein instability and subsequent ubiquitination by the F-box E3-ubiquitin ligase protein, FBXW7, followed by 26S proteasome degradation [119, 120]. In a study by Arnold and Sears, knockdown of B56α (PPP2R5A) resulted in c-myc overexpression, increased levels of phospho-S62 and stability of the c-Myc protein, implicating the PP2A-B56α complex as an important regulator of c-Myc function [121]. Additionally, B56α has also been show to interact with the anti-apoptotic BCL2 protein at the mitochondrial membrane, causing dephosphorylation and inactivation of BCL2 [122]. Furthermore, the B56α-PP2A complex plays an indirect role in the regulation of p53, with opposing studies suggesting PP2A as a negative and a positive regulator of p53 signaling [82]. 18 The expression of B56γ (PPP2R5C) is decreased in primary melanoma compared to non-transformed cells, conversely higher expression of PPP2R5C was found in melanoma cell lines compared to normal melanocytes [123]. In a separate study, PPP2R5C was overexpressed in lung cancer cell lines that lacked the endogenous protein expression, resulting in partial reversal of cell line tumourgenicity and suppressed cell proliferation, implicating levels of PP2A B56γ may be important in transformation behavior of human cancer cell lines [109]. Moreover, PP2A-B56γ is also important in Wnt signaling by inhibiting formation of the APC-Axin complex, which destabilizes β-catenin and inhibits transcription of Wnt pathway genes [124]. Lastly, a recent report has linked PP2A-B56delta (PPP2R5D) with selective dephosphorylation of BCL2 Ser70. BCL2 is overexpressed in many hematopoietic tumours and hyperphosphorylation of Ser70 causes anti-apoptotic activity [125].       Specific gene inhibitors of PP2A have also been implicated to play a role in cancer. The gene CIP2A (Cancerous Inhibitor of PP2A) inhibits PP2A induced de-phosphorylation of c-Myc S62 by direct interaction, thereby preventing c-Myc degradation [126]. A second endogenous PP2A inhibitor protein, the oncoprotein SET (I2PP2), can form an inhibitory protein complex with PP2A, thus causing changes in phosphatase regulation which ultimately contributes to tumourigenesis [127]. Targeting these PP2A inhibitory proteins by novel therapeutics has shown to increase PP2A activity by affecting the stability of c-Myc in breast and prostate cancer cell lines [128, 129]. Overexpression of CIP2A is present in head and neck squamous cell carcinoma (HNSCC) and colon cancer; therefore targeting overexpressed inhibitors of PP2A could be a novel way to treat cancers with this disease hallmark.  1.4.1.1 The Role of the PP2A A Subunits (PPP2R1A and PPP2R1B) in Cancer      The first somatic mutations in PPP2R1A were identified in multiple human tumour types albeit at low frequencies [130]. Calin et al. identified four different mutations in one case of lung carcinoma (E64D), one case of malignant melanoma (R418W), and two cases of breast carcinoma (E64G and a frameshift mutation). Additional studies using site-directed mutagenesis demonstrated that the few somatic mutations of PPP2R1A disrupt the binding of B and C subunits [90]. Specifically, the mutations were shown to disrupt binding with Bα, B’α and B’’/PR72 B subunit family members. The introduction of these Aα (PPP2R1A) mutations into 19 non-transformed cell lines displayed defects in B subunit binding (B55α, all B56 family members), in particular for B56γ (PPP2R5C), which is a critical interaction in human cell transformation [131]. PPP2R1A is an essential gene, thus complete loss induces cell death [132]. Several studies have concluded that PPP2R1A mutations contribute to cell transformation by haploinsufficiency, and that the level of functional PPP2R1A is essential for the balance of cell death and transformation [131, 133]. The haploinsufficient levels of PPP2R1A act to create competition between B subunit binding for the formation of an active PP2A complex. In addition, Chen et al., also found that suppression of Aα decreased phosphatase activity, and activated the AKT pathway. Furthermore, it has been discovered that 43% of human glioma harbor decreased expression of PPP2R1A [134].       Frequent loss of heterozygosity (LOH) on 11q22-24 was reported in ovarian carcinoma, which coincides with the PPP2R1B loci [135]. Upon closer inspection, the region of LOH involved the entire 11q arm, therefore was not only restricted to the PPP2R1B gene. Additionally, no somatic mutations of PPP2R1B were identified in ovarian carcinomas [135], resulting in the conclusion that PPP2R1B is not likely involved in ovarian tumourigenesis. Mutations in PPP2R1B (Aβ) were first identified in 15% of primary lung tumours, 6% of lung cell lines, and 15% of primary colon tumours [136]. Subsequent functional analysis established that some of the mutations, but not all, altered binding to the B”/PR72 and C subunits [137]. Interestingly, wild-type Aβ did not bind to the B-subunits Bα, B’α1 when compared to Aα binding. The B”/PR72 and C subunits bound at higher levels to Aβ, 30% and 15% respectively, although still significantly less binding compared to the Aα subunit. This may be due to differences in expression levels of the two A subunits, as the Aα subunit is 40X more abundant than Aβ [87]. One study has shown that Aβ, but not Aα, is responsible for the dephosphorylation of a small GTPase RalA, which is important in many cell processes [138]. In a lung cancer cell line that harbors an Aβ mutation there is constitutive phosphorylation of RalA promoting cell transformation.  Taken together these two subunits likely have distinctive roles in growth control during tumourigenesis, and mutations in the two subunits can affect the binding of B subunits family members in the context of different diseases and tissues.   20      The majority of my thesis work has focused on assessing PPP2R1A mutations in all subtypes of endometrial and ovarian carcinomas. The discovery of PPP2R1A mutations is described in Chapter 1, which led to the work in Chapter 3 discussing full exon sequencing of PPP2R1A and eight other genes for the molecular classification of endometrial carcinomas. The prevalence of these PPP2R1A mutations in gynaecological cancers, also led to the studies presented in Chapter 5 and 6, showing how a specific endogenous mutation affects interactions between PP2A A and B subunit proteins to form the PP2A holoenzyme.  1.5 Omics Technology – Advances in DNA and Protein Sequencing 1.5.1 Next-Generation Sequencing Technology      Advances in next-generation sequencing (NGS) technology, also referred to as second-generation sequencing, have revolutionized our ability to sequence cancer genomes. Large international collaborations such as TCGA (The Cancer Genome Atlas) and the International Cancer Genome Consortium (ICGC) are completing sequencing efforts for almost every cancer type. NGS technology has dramatically decreased the cost of sequencing a human genome, although this is still expensive. The total time for sequencing has decreased, which subsequently increased throughput and the amount of data output. The mass amount of data generated has caused a computational challenge for analysis as well as for suitable data storage. The Illumina sequencers (HiSeq, MiSeq) use massively parallel sequencing by synthesis (SBS) on a chip, called a flowcell, and has become the leading NGS sequencing technology used for whole genome RNA-sequencing (RNA-seq), exomes, epigenetic, and targeting panel sequencing. Other sequencing platforms such as the Ion Torrent and Roche 454 sequencing use emulsion PCR, and are now mostly used for targeted sequencing. The newest Illumina HiSeq2500 can produce a whole genome in less than a week, and a whole exome in just a few days. New technology is constantly being released to enable higher throughput and lower sequencing costs. This is revolutionary considering the first human genome took 10 years to produce and cost billions of dollars. The first Illumina personal sequencer released in 2011, called the MiSeq, is now leading the field with targeted gene sequencing and “hotspot” sequencing for clinical applications and molecular pathology, although new sequencing chemistry can allow for exome and RNA sequencing. Sanger sequencing is still used as a gold standard in validation experiments, however the power of NGS comes from the ability to barcode DNA from one patient, then pool 21 multiple patient DNA into one large sequencing run. This also allows the ability to perform bidirectional sequencing of multiple regions and multiple patients in one run, whereas a single Sanger sequencing reaction is limited by one direction, one amplicon and one sample.       Analysis of NGS data can identify heritable SNPs, somatic mutations, copy number variations (CNVs), and structural aberrations. The enormous amounts of sequencing output data from whole genome sequencing, exomes, targeted sequencing, etc., are being used to discover cancer driver and passenger mutations, identify somatic mutation profiles, clonal evolution and cancer subclone populations [139-141]. The characterization of these cancer hallmarks is being used to progress and aid the fast growing area of personalized cancer genomics. For example, patient tumour and normal samples are sequenced with NGS technology to identify somatic mutations and alterations that may be used for clinical diagnosis, targets for novel therapeutics or response to specific treatments. Clinical trials and therapeutics are being designed based on the presence of specific mutations in the tumour genome; for example the EGFR mutation in lung cancer for treatment with gefitinib, and BRCA1 and BRCA2 mutated tumours with PARP inhibitors [142]. In this thesis work, I have focused mainly on using targeted gene sequencing with Illumina NGS technology, and Sanger sequencing to identify mutational profiles that will aid in molecular pathology and the classification of endometrial tumours.  1.5.2 Proteomics Technology      With the plethora of research and technology advancement in genomics, there is now a large need to decipher how these genetic aberrations affect the function and structure of proteins in cancer cells. The field of proteomics has emerged to address the analysis of global protein expression analysis in a complex biological sample, and to complement the massive genomic datasets. There is accumulating evidence to suggest there is a low correlation of mRNA expression change with the changes in protein counterpart expression, thus underscoring the importance of post-transcriptional modifications and the complexity of the biological system. However, proteomics technologies are plagued by a number of technical challenges, making this analysis very challenging [143]. In the case of genomic sequencing, we have the luxury of being able to amplify DNA and RNA using PCR, however there are no such amplification methods for proteins that are extracted from tissue or cells. Low abundance proteins are analyzed in their 22 physiological state, unless a large amount of starting material can be prepared. This may be feasible for some in vitro cell line experiments, but when you are working with small patient samples this can become problematic.        To overcome these challenges and to keep up with the incredible pace of advancing NGS technology, there was a sizeable need for sensitive, high-resolution mass spectrometer technology enabling the analysis of complex samples. The Orbitrap mass spectrometer (MS) was invented by Makarov in 1999-2000 based on a technology from the 1920’s called the Kingdon trap [144]. The commercial Orbitrap MS was the first new MS instrument introduced in the last 20 years and has slowly revolutionized the world of proteomics [145]. In conjunction with HPLC (High Performance Liquid Chromatography), the Orbitrap MS has enabled the analysis of complex samples, and overcome some of the old issues of sensitivity, accuracy, and a need for high quantities of sample. In traditional “bottom-up proteomics”, proteins are digested into peptides, ionized and fragmented in the MS, and lastly identified using known peptide databases. With the advent of the Orbitrap, ionizing intact proteins can now be performed using “top-down proteomics”, which can allow for full protein characterization [146, 147]. This enables improvements in the study of post-translational modifications of proteins, which is not always supported using bottom-up proteomics due to inconsistent coverage of the many peptides that make up an individual protein. However, this type of analysis is still in the very early days of research, and is limited by hurdles in sample processing, instrumentation and analysis. Similar to NGS, proteins or processed peptides can be labeled (barcoded) using chemical labels after cell lysis (TMT, dimethyl or iTRAQ), or by in vitro labeling (SILAC) to enable pooling of samples and quantitation of protein levels. TMT (Tandem Mass Tags) and iTRAQ (Isobaric Tags for Relative and Absolute Quantitation) labeling methods allows peptides from different samples (up to 8 samples) to be simultaneously identified on the MS instrument [148, 149]. This allows for quantitation of the relative abundances from MS/MS analysis, resulting in the ability to perform differential protein expression in dynamic cellular systems. An advantage of this system is the ability to label and quantitate fixed and frozen patient proteomes. For analyzing cell lines, SILAC (Stable Isotope Labeling by Amino Acids in Cell Culture) uses “heavy” and “light” amino acids in the growth media which are incorporated into all of the endogenous proteins. Different cell lines or differing treatment of cell lines (i.e. drug treatments, or cell cycles) can 23 then be pooled for identification and quantitation by LC-MS/MS [150].  These proteomics analyses are slowly bringing genomics and proteomics together to enable mapping of the global changes in the cancer genome. In Chapter 6, I have also utilized challenging proteomics mass spectrometry technology, using TMT peptide labeling, in order to determine the effects of endometrial PPP2R1A mutations (identified in Chapter 2 and 3) on PP2A B subunit interactions.   1.6 Hypotheses A) Hypothesis A (Chapter 2) Endometrial and ovarian carcinomas will harbor mutations in PPP2R1A, which will aid in defining the molecular profiles of the histological subtypes. B) Hypothesis B (Chapter 3) The histopathological subtypes of endometrial tumours need assessment of specific gene mutation profiles to help classify and diagnose tumours.  C) Hypothesis C (Chapter 4) Mutation frequencies in specific genes and co-occurrence patterns, will differ in histologically similar tumours: ovarian endometrioid and endometrial endometrioid carcinomas. D) Hypothesis D (Chapter 5 and 6) The PPP2R1A W257L mutation leads to changes in interactions with specific PP2A B subunits and additional PP2A binding proteins.  24 Chapter 2: Identification of Subtype-Specific PPP2R1A Mutations in Endometrial and Ovarian Carcinomas  2.1 Introduction      Low frequency PPP2R1A mutations were first described in 3/42 (7%) of clear cell carcinomas of the ovary [70]. Mutations in PPP2R1A have been described at very low frequency in other types of cancers [130], but not in any histological subtype of endometrial cancer or ovarian tumours. There are striking clinical and pathological similarities between endometrial and ovarian histopathological subtypes, therefore it is expected that there are molecular similarities. This is evident between high-grade serous carcinoma of the endometrium and ovary, and between endometrioid carcinoma of endometrium and ovary, however it is less clear whether tumours of comparable cell type arising in different organs are also similar with respect to underlying molecular abnormalities. This is an important consideration if treatments are based on tumour cell types rather than if the organ of origin is also being contemplated. Recently, there has been an effort to identify new markers to define patient groups that would benefit from alternative or aggressive treatments for both ovarian and endometrial cancers [69]. To accomplish this, molecular and mutational characterization is needed to better understand the carcinoma subtypes and their relationship within a spectrum of gynaecological malignancies.       The goal of this study was to sequence exon 5 and 6 of PPP2R1A in endometrial and ovarian high-grade serous and endometrioid tumours, to test the hypothesis that PPP2R1A mutations are present in other subtypes of ovarian and endometrial carcinomas. Herein, I present the novel finding that PPP2R1A is mutated in endometrioid-type endometrial and ovarian carcinomas and in a significantly higher percentage of high-grade serous endometrial carcinomas, but not high-grade serous carcinomas of the ovary.  2.2 Materials and Methods 2.2.1 Patient Samples      The tumour specimens analyzed in this study for DNA and RNA sequencing were collected via several tumour banks and tissue repositories at the BC Cancer Agency and Vancouver General Hospital (via OvCaRe), the Australian Ovarian Cancer Study (AOCS), and Montreal 25 tumour bank (Banque de Tissus et de Données, of the Réseau de Recherche sur le Cancer of the Fonds de la Recherche en Santé du Québec, affiliated with the Canadian Tumour Repository Network), see also Supplemental Table 2 (Appendix A). Patients were approached for written informed consent, before undergoing surgery; to donate tissue surplus to diagnostic requirements plus a blood sample, for use in a research ethics board (REB) approved research protocol. All patients were informed at the time of consent about the potential risks of loss of confidentiality arising from use of their samples in research and that none of the research study data would ever form part of the clinical record or be reported back to the care physicians. See also Wiegand et al [71].   2.2.2 Whole Transcriptome Sequencing Analysis      Whole transcriptome data from Wiegand et al. (European Genome–Phenome Archive accession number, EGAS00000000075) was analyzed as previously described [71, 151].   2.2.3 DNA Extraction      DNA and RNA were extracted using standard methodologies, as previously described [71, 151].  2.2.4 PCR and Sanger Sequencing      PCR primer sets were designed to amplify exons 5 and 6 of PPP2R1A. Priming sites for -21M13 forward and -27M13 reverse were added to 5′ ends to allow direct Sanger sequencing of amplicons. PPP2R1A exons 5 and 6 forward primer 5’-ACAGAGAGGGGGTCATCACTT-3’, reverse 5’-GCCTAATGGAAACCTCAGCTC-3’. FFPE samples were amplified using exon 5 forward 5’-AAAACCTGGACCCACACAAC-3’, reverse 5’-TTGGAGAACATGGGGATGAT-3’, and exon 6 using forward 5’-CTCTCCTCTCCCTAGGACTCG-3’, reverse 5’-GTGTCAGTGTCCCCACCAGT-3’. For PCR reactions Platinum Taq DNA Polymerase High Fidelity (Invitrogen, Carlsbad, CA) reagents were used. After denaturation at 94ºC for 3 min, DNA was amplified 35-45 cycles (94ºC 45 sec, 64ºC 30sec, 72ºC 30 sec), final extension was at 72ºC for 5 min using a MJ Research Tetrad (Ramsey, MN). PCR products were purified using ExoSAP-IT® (USB® Products Affymetrix, Inc., Cleveland, OH) and amplified in both forward and reverse directions using M13 oligos (M13 forward 5’-TGTAAAACGACGGCCAGT-3’, 26 M13 reverse 5’-CAGGAAACAGCTATGAC-3’) and the ABI BigDye terminator v3.1 cycle sequencing kit (Applied Biosystems, Foster City, CA). Amplified products were then sequenced using an ABI Prism 3130xl Genetic Analyzer (Applied Biosystems, Foster City, CA). All capillary traces were visually inspected to confirm their presence in tumour and absence from germline traces by using Finch TV (Geospiza, Seattle, WA) and Mutation Surveyor (SoftGenetics LLC). All mutations were confirmed by duplicate PCR reactions.  2.2.5 Immunohistochemistry Staining      4 µm sections were processed using a Ventana Discovery XT automated system (Ventana Medical Systems, Tucson, AZ, USA) as per manufacturer’s protocol with proprietary reagents. After baking at 60°C for 1 h, they were deparaffinized on the automated system with EZ Prep solution (Ventana). Heat induced antigen retrieval method was used in Cell Conditioning solution (CC1-Tris based EDTA buffer, pH 8.0, Ventana). The rabbit monoclonal ER (Estrogen Receptor), clone SP1 (cat#: RM-9101) antibody was obtained from Thermo Fisher Scientific (Ottawa, Ont, Canada), the mouse monoclonal p53, clone DO-7 (cat#: M7001) antibody, and mouse monoclonal WT-1 (Wilms Tumour 1), clone 6F-H2 (cat#: M3561) antibody both obtained from Dako (Burlington, Ont, Canada).  All primary incubations were performed for 60 min with heat at a 1:25 dilution for ER, 1:400 dilution for p53, and 1:50 dilution for WT-1 in Ventana antibody diluents. The Ventana Universal Secondary Antibody was used for 32 min at 37°C. The detection system used was the Ventana DABMap kit, and slides were then counterstained with Hematoxylin and treated with a proprietary bluing agent (p53 and WT-1 used Ventana, ER used Tissue Tek Prisma). All washes were conducted with the Ventana Reaction Buffer, and dehydration steps and coverslip procedure were completed as per manufacturer’s recommendations. Categorical scoring for WT-1 and ER was assessed as 0=no expression, 1=expression. Scores for p53 were as follows; 0=loss of expression, equivalent to deletion or nonsense mutation, 1=normal expression, 2=overexpression equivalent to missense mutation (Supplemental Table A.2 (Appendix A).  2.3 Results      From our whole transcriptome sequencing data [71] I identified and validated one mutation within exon 5 of PPP2R1A, corresponding to amino acid residue R183, in an ovarian clear cell 27 carcinoma (OCCC1). To determine if such mutations are present in other ovarian or endometrial carcinomas I used Sanger sequencing to analyze exons 5 and 6 of PPP2R1A in DNA extracted from 271 tumours (see Methods) (Table 2.1). The somatic status of PPP2R1A mutated cases was also assessed using germline DNA. Although mutations of exon 6 have not been previously described, it was included due to known interactions of amino acids P179, R183, R182 (coded within exon 5) and W257 (coded within exon 6) with the PP2A regulatory B subunits [86, 90, 152, 153]. Recurrent somatic mutations, not previously described within COSMIC [154], were identified in both ovarian and endometrial carcinomas. Mutations appeared to be subtype enriched (Fisher’s Exact p <0.0001): 40.8% (20 of 49 cases) of high-grade serous endometrial carcinomas compared to only 5.0% (3 of 60 cases) of endometrial endometrioid carcinomas and 12.2% (5 of 41 cases) within ovarian endometrioid carcinomas (Figure 2.1A, Table 2.1). Mutations were also observed at a lower frequency of 4.1% (2 of 49 cases) in ovarian clear cell carcinomas (Table 2.1; Supplemental Table A.1 (Appendix A)), including two distinct missense mutations observed in neighboring nucleotides resulting in two apparent amino acid changes in a single case of ovarian clear cell carcinoma (Table 2.1; Supplemental Table A.1 (Appendix A)). No other cases had more than a single mutation, and all mutations appeared heterozygous upon review of electropherograms. In 28 of the 31 cases, which tested positive for mutations in PPP2R1A, the normal germline DNA tested negative for PPP2R1A mutations, demonstrating that all mutations were somatic (Supplemental Table A.1 and A.2 (Appendix A)). Also to further assess the sensitivity of mutation detection by sequencing, the tumour cellularity was determined for the majority of endometrial and ovarian tumours (Supplemental Table A.2 (Appendix A)). I confirmed there were no significant cellularity differences between endometrial high-grade serous carcinomas with and without mutations (p-value >0.3, range [20%-90%]), giving good indication that it is possible to detect PPP2R1A mutations in tumours with low cellularity. 28      Figure 2.1 Schematic of mutations in PPP2R1A (A) The lower panels represent the coding sequence and protein domain structure of PPP2R1A, grey arrows denote positions (corresponding to amino acids R182 and R183) of mutations described by Jones et al. [70], while the black arrow corresponds to the position of a single variant detected in whole transcriptome sequencing data of ovarian clear cell carcinomas from Wiegand et al. [71]. Upper panel denotes the position (amino acid residue), frequency, and subtype of mutations found across the cohort of ovarian and endometrial malignancies as a histogram. (B) Three dimensional protein structure of the PPP2R1A scaffolding Aα subunit of PP2A, with 15 repeat HEAT motifs (Model generated by SWISS-MODEL [155], and manipulated using PyMOL1.1). The HEAT 5 and 7 motifs coded by exons 5 and 6 are highlighted and amino acids affected by radical changes are annotated. Amino acid residues P179, R183 and W257 are know to interact with the regulatory B subunit of PP2A.   A"B"0 2 4 6 8 10 exon5 exon6 P179 R182 R183 R249 S256 W257 R258 # of samples mutated position of mutation CCC1"(R183W)"! E-HGSC ! E-mixed ! O-EC ! O-CCC ! E-EC 173 ATG TGA 1 1770bp coding exons 269 1 589aa HEAT"HEAT"HEAT"HEAT"HEAT"HEAT"HEAT"HEAT"HEAT"HEAT"HEAT"HEAT"HEAT"HEAT"HEAT"P179"R182"R183"S256"W257"R258""C-terminus N-terminus 29  Frequency (%) Coding sequence changes (mutation occurrence) Predicted amino acid changes Endometrial Carcinoma High-Grade Serous 20/49 (40.8%) c.536C>G (6) c.536C>T (3) c.544C>T (1) c.547C>T (2) c.767C>T (6) c.767C>A (1) c.769T>G (1) p.Pro179Arg p.Pro179Leu p.Arg182Trp p.Arg183Trp p.Ser256Phe p.Ser256Tyr p.Trp257Gly Clear Cell 0/1 (0%) NA NA Endometrioid 3/60 (5.0%) c.548G>A (2) c.746G>A (1) p.Arg183Gln pArg249His Mixed 1/3 (33.3%) c.544C>T (1) p.Arg182Trp Ovarian Carcinoma High-Grade Serous 0/50 (0%) NA NA Low-Grade Serous 0/12 (0%) NA NA Clear Cell 2/49 (4.1%) c.547C>T (1) c.771G>T§ (1) c.772C>T§ (1) p.Arg183Trp p.Trp257Cys§ p.Arg258Cys§ Endometrioid 5/41 (12.2%) c.536C>G (1) c.547C>T (2) c.767C>A (1) c.769T>G (1) p.Pro179Arg p.Arg183Trp p.Ser256Tyr p.Trp257Gly       Table 2.1 PPP2R1A mutation results with frequency, coding changes, and predicted amino acid changes Sequence co-ordinates are given relative to the coding sequence, or translation, within the PPP2R1A transcript variant 1 (NM_014225.5), unique mutations observed for each subtype are listed (see also Figure 2.1A and Supplemental Table 1(Appendix A)).    §Two somatic mutations were found in side-by-side nucleotides resulting in two apparent codon      changes in ovarian clear cell carcinoma sample OCCC2.        With the exception of mutations observed at R182 and R183, all other PPP2R1A mutations described in this study have not been found in other human cancers [154]. All mutations appear to be confined to intra-repeat regions of the PPP2R1A HEAT motifs, with mutations in exon 5 and 6 corresponding to HEAT 5 and 7 motifs, respectively (Figure 2.1B). These regions are known to loosely interact with the various isoforms of the regulatory B subunit of the PP2A holoenzyme [86, 94]. There was no correlation between the intra-repeat loop affected, carcinoma type or subtype. All except one of the observed mutations resulted in a predicted radical amino acid change (Supplemental Table A.1 (Appendix A)), and none of the mutations are predicted to result in premature termination or frameshifts.   30      To give some insight into the possible morphological differences within endometrial high-grade serous carcinomas with and without mutations, in conjunction with a histopathology expert, we assessed expression of p53, WT-1 (Wilms Tumour 1), and ER (Estrogen Receptor) by immunohistochemistry (IHC) (Supplemental Table A.2 (Appendix A)). The frequency of IHC staining of abnormal p53 (nuclear overexpression immunostaining or complete loss of protein expression) and ER immunoreactivity (IHC positive) was 87.8% (43 of 49 cases), and 79.6% (39 of 49 cases) of endometrial high-grade serous carcinomas, respectively. WT-1 positive IHC staining was also present in 28.6% (14 of 49 cases), which is consistent with the typical immunohistochemical profile of endometrial serous carcinomas [156]. There were no significant differences of expression (Pearson Chi-Square probability >0.15) of these molecules within the endometrial high-grade serous carcinomas that were mutation positive and mutation negative for PPP2R1A. Therefore, PPP2R1A mutations are independent from these already known molecular markers of endometrial high-grade serous carcinoma.  2.4 Discussion      As indicated in Chapter 1, endometrial serous carcinoma is a well-recognized aggressive variant of endometrial cancer, with a high propensity for extra-uterine metastasis [11, 12, 157].  I have identified a total of 31 missense mutations within exons 5 and 6 of PPP2R1A, with novel mutations most frequently (40.8%) found in endometrial serous carcinomas.  No mutations were detected in either high-grade or low-grade serous ovarian carcinomas; however only a small number of the latter (n=12) were studied. Previous PPP2R1A mutational analysis had suggested that the mutated amino acids P179, R182, R183 and W257 are involved in the interaction with different B subunit family members [86, 137, 152, 153]. Mutation of residues 179, 182, and 183 to alanine results in variable loss (50-90%) of binding to B subunit family members, and of particular importance a W257A mutation completely abolishes binding of B subunit family members [153]. The A subunit residues P179, R183 and W257 are known to interact with B subunit residues at N206, E214 and L107 of the B’ regulatory subunit B56γ1 (PPP2R5Cγ1), respectively [86, 94], suggesting that all of these residues may influence interactions with the regulatory B subunits. Additional in silico computation, by MutationAssessor [158], of these protein mutations results in a medium functional impact score adding further evidence to the importance of these amino acid residues. Furthermore, multiple studies have found polyomavirus 31 middle and small tumour (T) antigens and SV40 small t antigens can displace the regulatory B subunits [106, 107] resulting in increased phosphorylation of PP2A substrates and increasing cellular transformation [108]. Taken together this may indicate that mutations of the above noted regions of PPP2R1A could have a dominant-negative effect, modifying or eliminating proper interaction within the holoenzyme of regulatory B subunit family members. This may result in a change of the spectrum of PP2A scaffold, catalytic and regulatory subunit combinations in cells, and possible destabilization of the PP2A complex. In Chapter 6, the effect of a specific PPP2R1A mutation on B subunit interactions is determined using immunoprecipitation coupled with mass spectrometry.   Jones et al. suggested that since PPP2R1A mutations clustered in a small region, similar to the pattern of mutations seen in many well-described oncogenes, the mutations might be pro-oncogenic [70]. However, as previously described in the Introduction of this thesis (Chapter 1) the PP2A holoenzyme has mostly been described as a putative tumour suppresser and acts to regulate multiple cellular pathways [159-161]. The PP2A B subunits have been associated with negative regulation of multiple cancer-causing pathways including c-MYC, BCL2, p53, ERBB2, and AKT (reviewed in [82]). PP2A plays a role in the regulation of MAP kinase signalling and WNT pathways, possibly by stabilizing β-catenin, which can lead to proliferation and tumourigenesis [82, 162]. PP2A also plays a role in the G2-M cell cycle transition through direct interaction and dephosphorylation of CHK2 leading to cell cycle arrest [161, 163]. Additionally, the dysfunction of PP2A has been hypothesized to play a role in progression of serous borderline tumour cell lines [164]. Overall, the numerous studies of PP2A in the literature reflects the importance of this phosphatase complex for normal cellular functions and tumourigenesis.       Previous studies have highlighted the morphological similarities between high-grade serous carcinoma of the endometrium and ovary, starting with the first detailed description of serous carcinoma of the endometrium [11], however this study demonstrates an important molecular difference between them. Mutations in PPP2R1A were present in 40.8% of endometrial serous carcinomas, but were not seen in any high-grade serous carcinomas of the ovary. Although, this does not preclude the occurrence of low frequency PPP2R1A mutation in ovarian high-grade serous carcinomas, it does make them distinct. I have also shown IHC staining for the expression 32 of p53, WT-1 and ER, which is consistent with previous reports of molecular profiling of endometrial serous carcinomas [165]. These results also confirm distinct expression differences of WT-1 between endometrial and ovarian high-grade serous carcinomas, further strengthening the argument to caution against the common practice of regarding these as equivalent tumours.        Certainly the role of PPP2R1A in cancer has not been fully resolved; this study and future studies of full PPP2R1A sequencing may reveal further evidence supporting a role for PPP2R1A in endometrial and ovarian cancers. The identification of these subtype-specific mutations in endometrial serous carcinoma may have potentially significant treatment implications. These patients almost invariably receive adjuvant therapy, and remain at high risk for relapse and death, which underscores the need for further research. I hypothesize that mutations of PPP2R1A and thus the dysfunction of the PP2A holoenzyme may contribute to disease pathogenesis in a subtype specific manner, serve as a molecular marker and, upon further study, yield molecular druggable targets for high-grade serous carcinoma of the endometrium.   33 Chapter 3: Mutational Classification of Endometrial Carcinoma Subtypes   3.1  Introduction      Before the TCGA (The Cancer Genome Atlas) endometrial genomics characterization study [42], large-scale mutational profiling had been mostly confined to single gene or hotspot screening studies using Sanger-based sequencing. As described in the Introduction, advances in next-generation sequencing technologies in the last few years has allowed sequencing of multiple genes and samples simultaneously [166], making large mutational studies achievable.        Unlike ovarian subtype classification [69], no single gene is a sensitive or specific marker for endometrial carcinoma subtypes; therefore it is likely that the analysis of gene panels will be needed to guide subclassification. Ovarian subtype classification is considered reproducible, however recent reports have shown the current pathological classification and grading system of high-grade endometrial carcinomas is limited in both reproducibility and prognostic ability [23, 167-169]. My discovery of subtype-specific PPP2R1A mutations in endometrial serous tumours, described in Chapter 2, gives significant evidence to support mutational profiles aiding in stratifying endometrial subtypes. Therefore, herein I sought to determine the mutation profiles of a large series of endometrial carcinomas, using next-generation exon capture sequencing of 9 genes known to be important in carcinogenesis (ARID1A, PTEN, PIK3CA, KRAS, CTNNB1, PPP2R1A, BRAF, TP53, and PPP2R5C) in an attempt to improve the classification of endometrial carcinomas.   3.2 Methods 3.2.1 Patient Samples      I obtained 152 endometrial tumours, and 90 corresponding buffy coat specimens originating from the BC Cancer Agency and Vancouver General Hospital via the OvCaRe Tissue Biobank repository, Vancouver, BC, Canada. An additional 76 cases of endometrial serous carcinomas, corresponding to two TMAs were also obtained from the same institutions. Patients were informed for written consent, and research ethics approved as previously described [1]. An additional 260 endometrial tumour DNA samples were obtained from Washington University, St. Louis, Missouri. The endometrial subtype, grade and microsatellite instability data was 34 previously determined in these cases. All samples from both centers’ have undergone review by gynaecological pathologists.  3.2.2 DNA Isolation      Genomic DNA, from the Vancouver cohort, was extracted from flash frozen tumours using the Ambion DNA extraction kit as per manufacturer's protocol (Ambion, Life Technologies). Genomic DNA, from the St. Louis cohort, was isolated using Proteinase K and phenol extraction or the DNeasy Tissue kit (Qiagen Inc.). For the additional endometrial serous cohort, fixed formalin paraffin embedded (FFPE) blocks were identified and extracted for DNA using the Qiagen FFPE kit as per manufacturers protocols.  3.2.3 Exon Sequencing      Genomic DNA (500ng) was used for indexed Illumina library construction [71], then underwent targeted enrichment using biotinylated RNA capture probes generated from cDNA clones or PCR amplicons [170] representing all exons of ARID1A, PTEN, PIK3CA, KRAS, CTNNB1, PPP2R1A, BRAF, TP53, and PPP2R5C (probe genomic locations can be found in Supplemental Table B.3 (Appendix B) and sequenced using Illumina (GAIIx).   3.2.4 Bioinformatics Analysis      Short reads were aligned to the human genome (hg18) using the BWA aligner v0.5.9 [171]. A Random Forest classifier (MutationSeq) trained on validated SNVs was used to remove false-positive calls [172]. SNVs in the Catalogue of Somatic Mutations in Cancer (COSMIC) [173] were considered to be true positives, so a 99% cutoff threshold was selected with a probability threshold cutoff of 0.2588 (Supplemental Figure B.1 (Appendix B)). Mean coverage was plotted for cases with and without mutations (Supplemental Figure B.2 (Appendix B)). Details found in Supplemental materials and methods (Appendix B).  3.2.5 DNA Validations      Select predicted SNVs were validated using Sanger sequencing as previously described [1]. Sanger sequencing validated a subset of the SNVs and indels called by bioinformatics’ analysis. I validated 340 predicted positions that were present in all 9 genes tested. 330/340 (97%) 35 positions were validated as true positive, and 10/340 (2.9%) were considered false positive. The majority of these false positives have low allelic fraction reads (<10%), therefore Sanger sequencing may not be sensitive enough to identify these mutations if tumour cellularity is low. Validation was performed on a range of called mutations with low and high frequency and probability. All others were considered true positive unless otherwise indicated.  3.2.6 Targeted Fluidigm-MiSeq Sequencing of PPP2R1A, FBXW7, TP53      Targeted primers were designed to cover the full exon regions of the genes (PPP2R1A, FBXW7, TP53) and synthesized by IDT Technologies. All primers were tested and re-synthesized if no amplification product was present. In brief, primer sets were designed using Primer 3 to amplify the specific gene regions, and tagged with CS1 (5’-ACACTGACGACATGGTTCTACA-3’) and CS2 (5’-TACGGTAGCAGAGACTTGGTCT-3’) sequencing tags. PCR products (150-200bp) were amplified using the Fluidigm 48X48 Access Arrays, as per manufacturers protocol, with input of 100ng for FFPE derived DNA, and 50ng for high-quality DNA from buffy coat or frozen tumour DNA. DNA was sequenced originating from 89 endometrial serous tumours with 40 from frozen tumours, and 49 from FFPE tissue. DNA barcodes (10bp) with Illumina cluster-generating adapters were added to the libraries post-Fluidigm harvest as previously described [174], and cleaned-up using Agencourt AMpure XP beads (Beckman Coulter). Barcoded PCR product pools were then quantified using the high sensitivity DNA assay and Qubit fluorometer (Life Technologies) and pooled to one total library by normalizing to equal amounts of PCR product. In total, 96 samples were pooled, denatured according to Illumina standard protocols, and sequenced using a MiSeq 300 cycle V2 kit on the Illumina MiSeq for ultra-deep validations. Uni-directional barcode sequencing was performed. All bam and VCF files were generated using Illumina MiSeq reporter. Analysis was performed using the VCF files generated by the somatic variant caller 3.2.3.0, then filtered based on reads passing filter, non-synonymous, and >5% variant allele frequency.  All potential mutations were then manually interrogated and filtered using the Integrated Genome Viewer (IGV). For validations, repeat Fluidigm-MiSeq sequencing was performed along with select Sanger sequencing, however CS1 and CS2 primers were used as a universal sequencing primers on the ABI 3130xl Genetic Analyzer (Applied Biosystems) and analyzed as previously described [1]. Normal DNA was also sequenced to check somatic status. 36  3.2.7 Identifying Outlier Cases       Outliers were identified by observing mutation profiles that did not fit the original diagnosed histological subtype; defined as ESC with PTEN and/or ARID1A mutations, and low-grade EECs with only TP53 and/or PPP2R1A mutations. With the goal of comparing mutational outliers with immuno-profiles, formalin-fixed embedded paraffin blocks were only available for 147/156 Vancouver cases, for the construction of a Tissue Microarray (TMA). These cases were used for the characterization of mutational outliers, by correlating with morphology and immunohistochemistry (IHC), and retrospectively reviewed by two independent pathologists, using the full hysterectomy case, without knowledge of mutation or IHC data.  3.2.8 TMA and Immunohistochemistry      To construct the endometrial TMA, tumours were annotated by a pathologist (BG) and two 0.6mm cores per case were arrayed. TMAs were cut at 4µm thickness onto Superfrost+ glass slides, and were processed using the Ventana Discovery XT, and the Ventana Benchmark XT automated system (Ventana Medical Systems) as per manufacturer’s protocol with proprietary reagents. After slides were baked at 60°C for 1 h, they were deparaffinized on the automated system with EZ Prep solution (Ventana). Heat induced antigen retrieval method was used in Cell Conditioning solution (CC1-Tris based EDTA buffer, pH 8.0, Ventana). The rabbit monoclonal ER, clone SP1 (cat#: RM-9101) antibody was obtained from Thermo Fisher Scientific, the mouse monoclonal p53, clone DO-7 (cat#: M7001) antibody obtained from Dako, the rabbit monoclonal PR, clone 1E2 (cat#790-2223) antibody obtained from Ventana, the rabbit monoclonal PTEN, clone 138G6 (cat#9559) antibody obtained from Cell Signaling, the mouse monoclonal p16, clone E6H4 (cat#9517) antibody obtained from mtm Laboratories, the rabbit monoclonal c-myc, clone Y69 (cat# ab3072) antibody obtained from Abcam. Primary incubations for ER were performed for 60 min with heat at a 1:25 dilution, 32 min with heat at 1:800 dilution for p53, 8 min with heat and neat for PR, 60 min without heat at 1:25 dilution for PTEN, and 32 min with heat at 1:3 dilution for p16, all using Ventana antibody diluents. The Ventana Universal Secondary Antibody was used for 32 min at 37°C. The detection system used was the Ventana DABMap kit (ER, PR), Ventana UltraMap DAB kit (PTEN), Ventana OptiView DAB kit (p16, p53), and slides were then counterstained with Hematoxylin and treated 37 with a proprietary bluing agent (Ventana). All washes were conducted with the Ventana Reaction Buffer, and dehydration steps and coverslip procedure were completed as per manufacturer’s recommendations. Categorical scoring for ER, PR, PTEN, p16 was assessed as 0=no expression (loss of expression for PTEN), 1=expression. Scores for p53 were as follows: 0=loss of expression, equivalent to deletion or nonsense mutation, 1=normal expression, 2=overexpression equivalent to missense mutation (Supplemental Table B.5 (Appendix B)). The c-myc IHC was scored as intensity =0/1/2, and nuclear expression 0-100%.  The final scoring resulted from the maximum H-score between duplicate TMA cores calculated as intensity multiplied by percent nuclear positivity.  3.2.9 Microsatellite Instability (MSI) Assay       The analysis of tumour MSI was performed as previously described [175]. I used the Bethesda Markers; Bat 25, Bat 26, DI7S250, D5S346, and D2S123 to assess the MSI status. MSI-high is reported if 2 or more markers have MSI, one or less markers with MSI were considered MSI-low, and MSS (Microsatellite Stable) indicates no markers were positive. Cases reported as N/A, did not have sufficient normal available to assess the MSI status of these tumours (Supplemental Table B.4 (Appendix B)).  3.2.10 Statistical Analysis       Fisher exact tests and multivariable logistic regression analysis were used to test the significance of associations between mutations within subtypes. All tests were two-tailed and p-value < 0.05 were considered significant. Fisher exact tests were not adjusted for multiple comparisons. The multivariable logistic regression model used step-wise selection based on the likelihood ratio test, with all genes included. The Hosmer-Lemeshow test was used to assess the goodness-of-fit of the estimated logistic regression models.     38 3.3 Results      To determine the mutation frequencies in various subtypes of endometrial carcinomas, exon capture sequencing was performed for ARID1A, PTEN, PIK3CA, KRAS, CTNNB1, PPP2R1A, BRAF, TP53, and PPP2R5C. This resulted in the detection of somatic nonsynonymous missense, truncating, indels (insertions/deletions), and splice site mutations in 90.1% (353/392) of cases. The characteristics of the endometrial carcinomas, with histology subtypes and grade, are summarized in Table 3.1. These carcinomas were stratified into low-grade (grade 1 and 2) EECs, EEC-3, ESC, carcinosarcoma, mixed, and undifferentiated, based on routine histopathological assessment, to determine the differences in mutational profiles. All mutational data are summarized in Supplemental Table B.4 (Appendix B). The mutation frequencies of ARID1A, PTEN, PIK3CA, PPP2R1A, TP53, and CTNNB1 are significantly different across four subtypes of endometrial carcinomas (Table 3.2).   3.3.1 High-Grade and Low-Grade Endometrioid Carcinomas have Similar Mutation Profiles but Differ in Frequencies of TP53 Mutations       Low-grade EECs have mutations in PTEN (67%), ARID1A (47%), PIK3CA (38%), and CTNNB1 (24%) (Table 3.2), with higher frequencies of mutations in PTEN (90%), ARID1A (60%), PIK3CA (57%), KRAS (27%), PPP2R1A (10%) and TP53 (30%) seen in EEC-3s (Table 3.2). The comparison of mutations in low-grade EEC and EEC-3 showed that PTEN (p=0.0111) and TP53 (p=0.0046) mutation frequencies are significantly different (Table 3.3). Multivariable logistic regression model fitting suggested PTEN (p=0.007) and TP53 (p<0.001) mutations as disguising features (Table 3.4).          39   All Subtypes Endometrioid 306 Grade 1 169 Grade 2 107 Grade 3 30 Serous 37 Mixed* 4 Undifferentiated 3 Carcinosarcoma 42 Total 392 Table 3.1 Summary of all endometrial carcinoma subtypes * Includes one cases as mixed serous and endometrioid carcinoma, one  case mixed G2 and G3 endometrioid and clear cell carcinoma, and two  cases as mixed serous and clear cell carcinoma.    Low-Grade Endometrioid    (G1 and 2) (n=276) High-Grade Endometrioid (G3) (n=30) Serous           (n=37) Carcinosarcoma (n=42) p-value            across all subtypes (chi-squared test) PTEN 185 (67.0%) 27 (90.0%) 1 (2.7%) 14 (33.3%) 4.63E-17 PIK3CA 105 (38.0%) 17 (56.7%) 10 (27.0%) 12 (28.6%) 0.0480 ARID1A 129 (46.7%) 18 (60.0%) 4 (10.8%) 10 (23.8%) 5.77E-06 KRAS 46 (16.6%) 8 (26.7%) 3 (8.1%) 7 (16.7%) 0.2434 CTNNB1 66 (23.8%) 6 (20.0%) 1 (2.7%) 2 (4.8%) 1.19E-03 PPP2R1A 19 (6.9%) 3 (10.0%) 16 (43.2%) 9 (21.4%) 1.50E-09 TP53 28 (10.1%) 9 (30.0%) 25 (67.6%) 27 (64.3%) 2.79E-23 BRAF 8 (2.9%) 2 (6.7%) 2 (5.4%) 1 (2.4%) 0.6186 PPP2R5C 1 (0.4%) 2 (6.7%) 0 (0%) 0 (0%) 0.002 Table 3.2 The frequency of mutations within all endometrial subtypes Bold indicates significant p-values <0.05       40  Low-Grade Endometrioid vs. High-Grade Endometrioid Low-Grade Endometrioid vs. Serous High-Grade Endometrioid vs. Serous High-Grade  Endometrioid  vs. Carcinosarcoma Serous vs. Carcinosarcoma PTEN 0.0111 6.58E-15 2.57E-14 1.09E-06 4.30E-04 PIK3CA 0.0522 0.2091 0.0235 0.0276 1.0000 ARID1A 0.1826 1.38E-05 2.42E-05 0.0030 0.1522 KRAS 0.2000 0.2328 0.0525 0.3814 0.3215 CTNNB1 0.8211 1.23E-03 0.0394 0.0602 1.0000 PPP2R1A 0.4630 4.96E-08 2.95E-03 0.3365 0.0526 TP53 4.62E-03 8.56E-14 3.17E-03 0.0080 0.8151 BRAF 0.2555 0.3352 1.0000 0.5669 0.5972 PPP2R5C 0.0263 1.0000 0.1967 0.1702 NA Table 3.3 Univariate Fisher exact test (p-values) to show significant differences between mutation profiles of each endometrial carcinoma subtypes.  3.3.2 Endometrial Serous Carcinomas Show a Distinct Mutation Profile      Of 37 ESCs, high frequencies of mutations were found in TP53, PPP2R1A, and PIK3CA (Table 3.2). TP53 and/or PPP2R1A mutations were found in 28/37 (75.7%) of ESCs, accounting for the majority of aberrations in this subtype (Figure 3.1). The comparison of EEC-3 to ESC revealed significantly different mutation frequencies for ARID1A, PTEN, PIK3CA, CTNNB1, PPP2R1A, and TP53 (p <0.05) (Table 3.3). Low frequencies to zero mutation events were noted for some genes common in both ESCs and EEC-3. In an attempt to keep all the multivariable analyses consistent across the subtype comparisons, the statistician and I included the same list of genes in the logistic regression model building between EEC-3 and ESC. As a result, there was no one reliable multivariable logistic regression model built, based on the mutation markers, to distinguish between these two subtypes (Table 3.4). As expected, the mutational profiles of low-grade EEC and ESC are significantly different (Table 3.3). Multivariable logistic regression shows, PTEN (p <0.001) with a trend of ARID1A (p=0.08) mutations associated with low-grade EEC, whereas PPP2R1A and TP53 (p <0.001) are associated with ESC (Table 3.4). Additional testing in a separate cohort would need to be performed to validate these results.41  Figure 3.1 Mutation profiles of endometrial subtypes A. Low-grade endometrioid carcinoma, including grade 1 and 2 tumours (n=276); B. High-grade endometrioid carcinoma, grade 3 tumours (n=30); C. Serous carcinoma (n=37); D. Carcinosarcoma (n=42), (+) indicates carcinosarcomas with heterologous differentiation elements; E. Undifferentiated and mixed histology subtypes, (a) undifferentiated carcinomas, (b) mixed low-grade EEC with serous carcinoma, (c) mixed endometrioid and clear cell carcinoma, (d) mixed serous and clear cell carcinoma. Rows indicate genes, columns represent tumour cases. Coloured bars indicate mutations’ including; missense, truncating, indels and splice site mutations. Grey bars indicate no mutations were detected. (*) indicates serous carcinoma outliers with ARID1A mutations; (#) indicates low-grade EECs and EEC-3s mutation outliers with serous-type mutations (TP53 or PPP2R1A). 42  Gene (Marker) Low-grade Endometrioid  (n=276) vs. High-grade Endometrioid (n=30) Low-grade Endometrioid (n=276) vs. Serous (n=37) High-grade Endometrioid (n=30) vs.                  Serous (n=37) High-grade Endometrioid (n=30) vs.  Carcinosarcoma (n=42) Serous (n=37) vs. Carcinosarcoma (n=42)  p-value Odds ratio* to high-grade endometrioid p-value Odds ratio* to  serous p-value Odds ratio* to  serous p-value Odds ratio* to carcinosarcoma p-value Odds ratio* to carcinosarcoma PTEN 0.007 5.61 (1.6-19.7) 1.89E-04 0.02 (0.002-0.14)   3.75E-05 0.05 (0.01-0.22) 6.24E-03 19.41 (2.3-162.6) PIK3CA           ARID1A   0.080 0.3 (0.08-1.2)       KRAS           CTNNB1           PPP2R1A   2.7E-04 13.28 (3.3-53.4)   0.0736 5.12 (0.86-30.7) 0.0446 0.32 (0.1-0.97) TP53 7.04E-04 4.95 (2.0-12.5) 7.64E-05 7.64 (2.8-20.9)       BRAF   1.40E-02 18.9 (1.8-196.7)       PPP2R5C           Table 3.4 Multivariable logistic regression analysis of gene mutations between endometrial carcinoma subtypes.  Reported values are only the most significant genes selected by the step-wise selection method based on the Likelihood ratio test. *Odds ratio (95% CI).  3.3.3 Cases with Discordant Morphological Diagnosis and Mutational Profiles       As discussed, ESCs were found to have a high frequency of mutations in TP53 and PPP2R1A (Figure 3.1). From the mutation profiles, I identified three histology-defined ESC cases with ARID1A and PTEN mutations and lacked TP53 mutations, a profile more indicative of EECs (Figure 3.1). Other studies have not found ARID1A or PTEN mutations in ESCs, however there have been limited studies testing for ARID1A mutations in endometrial carcinomas [31, 34, 176].  On independent histopathological review of these three cases, all were mixed tumours consisting predominantly of ESC, but with minor components of low-grade EEC in two cases, and EEC-3 with clear cell carcinoma in one case (Table 3.5).  For the two mixed ESC and low-grade EEC cases, I confirmed the section of tumour sample used for DNA extraction and subsequent sequencing exclusively contained the ESC component (Figure 3.2); however it 43 ID 841 1120 220 895 511 1034 611 Original Subtype Serous carcinoma Serous carcinoma Serous carcinoma Low-grade endometrioid carcinoma Low-grade endometrioid carcinoma Low-grade endometrioid carcinoma Low-grade endometrioid carcinoma Review mixed serous (80%) and low-grade endometrioid carcinoma, with adjacent endometrium showing focal complex atypical hyperplasia mixed serous (60%) and low-grade endometrioid carcinoma, with adjacent endometrium showing complex atypical hyperplasia Grade 3 endometrioid with clear cell changes Grade 2 endometrioid (extensively myometrial-invasive and LVI) Grade 3 endometrioid Mixed low-grade (G2) endometrioid and serous carcinoma Serous carcinoma p53-IHCa 1 1 1 2 2 2 2 ER-IHCb 0 1 0 1 1 1 1 P16-IHCb 0 0 0 0 0 1 1 PR-IHCb 0 1 0 1 1 1 NA PTEN-IHCc 1 1 0 0 0 0 0 ARID1A p.Q420*, p.R1335* p.Q2176fs p.Q548fs, p.G1847fs     PTEN   p.L265fs   splice site acceptor  PIK3CA p.G106V, p.V344M p.Q546K, p.H1047Y Y1021C     KRAS p.G13D p.G12A      PPP2R1A p.R182W    p.P179L   TP53    p.R282W p.H193L p.R248Q p.S241F CTNNB1        BRAF   p.A526V, p.P403fs     PPP2R5C        Table 3.5 Outlier cases with pathological review, IHC and mutation profile. aScoring; 0= loss of expression, 1= normal expression, 2= over-expression bScoring; 0= no expression, 1= over-expression, cScoring; 1=normal expression, 0= loss of expression 44   Figure 3.2 A case originally diagnosed as serous carcinoma, but with an ARID1A mutation and no TP53 mutation, is a mixed low-grade endometrioid and serous carcinoma (case #1120).  A. A mix of a grade 1 endometrioid (left half) and high-grade serous (right half) carcinoma, 40X magnification; B. High power (100X) image of histologically distinct low-grade endometrioid carcinoma; C. High power (100X) images of serous carcinoma component, of which the sampling of tumour was used for mutation sequencing; D. Atypical complex hyperplasia in the background endometrium (40X magnification) harbored mutations with an endometrioid profile. Immunostaining is recommended for use in diagnostically problematic cases [22], although not universally used. These three cases showed a non-serous IHC profile; p53 normal expression and p16 negative expression, while one expressed ER and PR (Table 3.5).   harbored mutations with an endometrioid profile. Immunostaining is recommended for use in diagnostically problematic cases [22], although not universally used. These three cases showed a non-serous IHC profile; p53 normal expression and p16 negative expression, while one expressed ER and PR (Table 3.5). 45      I also identified four outlier low-grade EECs that contained TP53 mutations and lacked PTEN mutations, which were also diagnostically challenging cases. Upon review, two cases showed morphological features of serous, and one case was re-classified from low-grade EEC to EEC-3. One outlier remained classified as low-grade EEC, however it was noted that this case showed extensive myometrial invasion and widespread lymphovascular invasion.  By IHC, abnormal p53 expression was confirmed in all cases. All were, however, ER-positive with PTEN loss of expression, features found primarily in EECs. In two of these cases, p16 was strongly expressed (Table 3.5). In summary, these seven outlier cases showed features intermediate between ESC and EEC in morphological, IHC and genetic analysis (Table 3.5, Table B.5, Figure B.3 (Appendix B)).      I also performed unsupervised hierarchical clustering analysis on the 147 cases with IHC and mutational status (Figure B.3, Table B.5, Supplemental methods (Appendix B)). This shows most low-grade EEC and EEC-3 subtypes cluster together, while the remaining EEC-3, serous and mixed cases are scattered. The mutational outliers with the diagnosed subtype are indicated, as well as the new classification.   3.3.4 Carcinosarcomas Show Either an Endometrioid or Serous Mutation Profile      Endometrial carcinosarcomas are relatively rare, and their classification as an endometrial carcinoma subtype or as a distinct entity is under debate [177].  In my analysis of carcinosarcomas I found mutations in TP53, PTEN, PIK3CA, ARID1A, and PPP2R1A (Table 3.2). Two subgroups of carcinosarcomas were identified; one group characterized by mutations in PTEN and ARID1A (endometrioid-type), and a second group with TP53 and PPP2R1A mutations more similar to ESC (Figure 3.1). Heterologous differentiation of the sarcomatous component was observed in a subset of tumors from both groups. Histopathological reviews of cases were not available; therefore it was not possible to correlate morphological features and mutational profiles of endometrioid-like or serous-like in the epithelial components of these tumours.  46 3.3.5 Mutations Involving Signalling Pathways in Endometrial Carcinomas      By mutational analysis of multiple genes, it is possible to identify different mutations involving a single signalling pathway that may be functionally equivalent, and to examine the relationship between mutations involving different genes/pathways. Mutations in the PI3K and MAPK signalling pathways are known to be important in EECs, therefore I further examined the prevalence of mutations in PTEN, PIK3CA, KRAS, ARID1A and CTNNB1. I found 211/276 (76.5%) low-grade EECs have PTEN and/or PIK3CA mutations (Figure 3.1). Co-existent PTEN and PIK3CA mutations were identified in 79/276 (28.6%) low-grade EECs, and 16/30 (53.3%) EEC-3s (p=0.0112). AR1D1A mutations have recently been identified in low-grade EECs; however the relationship of these mutations with other pathways such as PI3K and WNT has not been examined [34]. Of the low-grade EECs with ARID1A mutations, 112/129 (86.8%) have mutations within PTEN and/or PIK3CA (p=0.0002). EEC-3s with ARID1A mutations (n=18) all have PTEN mutations, and 13/18 (72.2%) also have PIK3CA mutations. Thus there is a significant association between ARID1A and PTEN/PIK3CA mutations.   3.3.6 Microsatellite Instability      MSI is a feature of the endometrioid subtype, therefore I determined the MSI status of 241/276 low-grade EECs and 13/30 EEC-3s. I found 97/241 (40.2%) of the low-grade EECs were MSI positive, compared to 8/13 (61.5%) of EEC-3 (Supplemental Table B.4 (Appendix B)). Considering all 110 cases of EECs with MSI, only 39/110 (35%) harbor concurrent PTEN and PIK3CA mutations, and 25/110 (23%) have co-occurring mutation in PTEN, PIK3CA and ARID1A.   3.3.7 Analysis of PPP2R1A and FBXW7 Mutations in Endometrial Serous Carcinomas      The endometrial TCGA analysis in cBioPortal (www.cbioportal.org) [32, 33], revealed similar frequencies of PPP2R1A (23%) and FBXW7 (22%) mutations in the copy-number high (serous-like n=60) group of tumours (Figure 3.3A). Of interest, these two genes with the highest frequency of mutations, not including TP53 alterations, showed a trend towards mutual exclusivity, although not statistically significant. This led to the investigation of PPP2R1A and FBXW7 mutation mutual exclusivity in a validation cohort of endometrial serous carcinoma cases (89 cases). Targeted exon sequencing (involving all exons) revealed mutations in 33/89 47 (36%) PPP2R1A, 15/89 (17%) FBXW7, 57/89 (64%) TP53 (Figure 3.3B, Table B.6 (Appendix B)). However, 6/89 (7%) cases harbored mutations in both PPP2R1A and FBXW7, which did not support statistically significant mutation mutual exclusivity in these cases. Further analysis of this cohort with c-Myc immunohistochemistry revealed that there is no significant association of FBXW7 or PPP2R1A mutations with high c-Myc expression. However, there is a significant association of high c-Myc expression (H-score >60) when both FBXW7 and PPP2R1A are mutated together (Fisher exact test, p<0.0011) (Table B.6 (Appendix B)).   Figure 3.3 Mutation profiles of endometrial serous carcinomas A. TCGA endometrial carcinoma OncoPrint copy number high (serous-like) (n=60) mutations and copy number alterations (CNA) data from cBioPortal. B. Vancouver endometrial serous carcinoma validation samples (n=89) mutations and c-Myc IHC data. Rows indicate altered genes, and columns indicate patient tumours. Grey bars indicate no mutation was detected. See Appendix B Table 8.6 for mutation and IHC details.  PPP2R1A&36%&FBXW7&17%&TP53&64%&C%Myc)IHC)Missense)Frameshi5)or)Inser7on)Nonsense)H%score)>60)H%score)<60)A)B)PPP2R1A&23%&FBXW7&22%&TP53&92%&C%Myc)CNA)Amplifica7on) Muta7on)48 3.4 Discussion      Endometrial carcinoma is a heterogeneous disease, comprised of multiple subtypes with differing risk factors, precursor lesions, and outcomes. Lack of reproducibility in histopathological diagnosis of endometrial carcinoma subtypes has hindered progress. For example, while some studies have found that EEC-3 and ESC have different outcomes [178], other studies have not [23]. This difference may reflect inclusion of different cases, based on subtly different diagnostic criteria, within these cohorts. Robust and reproducible diagnostic categories are an important first step in moving towards subtype-specific treatment, which is proceeding in ovarian carcinoma [179, 180]. However in the case of endometrial carcinoma, it is likely that molecular markers will be needed to improve the suboptimal performance of conventional histopathological assessment [181]. With the advent of next-generation sequencing technologies, the molecular profiles of many tumour cell types are being extensively characterized. These mutation profiles can potentially be used diagnostically for subclassification, and to identify relevant targets for the development/deployment of targeted therapeutics. In this study, exon capture sequencing of nine genes was performed in two large cohorts of endometrial carcinomas, revealing differing mutational landscapes for endometrial carcinoma subtypes.        As demonstrated in previous studies, I identified high frequencies of mutations within PTEN, PIK3CA, ARID1A, KRAS and CTNNB1, and lack of TP53 mutations in low-grade EECs. EEC-3s demonstrate a similar pattern of mutations, but with a significantly increased frequency of TP53 mutations. High frequencies of PTEN mutations in EECs confirm this is an early driver event in tumour progression. My results show that the frequency of MSI cases is similar in low-grade EEC and EEC-3, which supports the view that the majority of EEC-3s have progressed from low-grade EEC [23].        Recent studies identified a high frequency of concurrent PTEN and PIK3CA mutations in endometrial carcinomas [27, 182], but not in any other tumour type investigated to date [27]. In this study, this phenomenon was also observed in low-grade EECs and EEC-3s, but not in ESC or carcinosarcoma. I have determined that in low-grade EECs and EEC-3s, ARID1A mutations 49 are significantly associated with concurrent mutations in PTEN and PIK3CA, a novel finding suggesting a cooperative role of these pathways in EEC tumourigenesis.        ESCs have frequent TP53 and PPP2R1A mutations, and lack mutations in PTEN, ARID1A and CTNNB1, a mutational profile distinct from that of EECs. While it was not possible to classify tumours solely based on this nine gene mutation panel, I was able to use the mutation profile as a diagnostic adjunct for morphological subclassification in individual cases. This is an attractive prospect given the significant problems in distinguishing EEC-3 and ESC highlighted in recent studies [22, 53, 168, 169, 183, 184]. There were mutational outliers where the original diagnosis did not fit the mutation profile, specifically ESC cases with ARID1A mutations, and low-grade EECs with only TP53 mutations. In the seven outlier cases, retrospective review by two independent pathologists resulted in reclassification, agreeing with the subtype-specific mutation patterns rather than the original diagnosis.   It has previously been proposed that ESC may arise through two different tumourigenic pathways, i.e. from progression through hyperplasia and low-grade EEC, or arising via high-grade endometrial intraepithelial carcinoma, in an estrogen-independent pathway [185]. In this study, I observed two tumours initially diagnosed as ESC that showed an endometrioid mutation profile. On retrospective review the diagnosis for both was changed to mixed serous and endometrioid. This observation is not novel but does give further support to ESCs arising in some cases by an alternative molecular pathway, rather than the classical Type 2 pathway (Figure 3.4, Figure 3.5) [186]. This further suggests that the classification of endometrial carcinomas cannot be encompassed by a simple dualistic model. In particular, the high-grade subtypes show considerable heterogeneity not reflected adequately in a Type 1 versus Type 2 model. Future studies will be required to address the following issues: 1. How reproducible is molecularly supported subtype diagnoses? 2. If diagnoses can be made reproducibly, do subtypes show significant differences in stage at diagnosis, pattern of spread, prognosis or response to treatment? Only after those questions are addressed can subtype-specific management move forward, and mutation-based treatment decisions can be made for challenging diagnoses.   50     Figure 3.4 An intermediate type of high-grade endometrial carcinomas is not encompassed in the Type 1 and Type 2 model. Problematic morphological diagnoses of high-grade endometrial carcinomas are an intermediate subtype and may be further subclassified by mutational profiles.                    Type%1%% Type%2%%Low+Grade%EEC(G1%and%2)%PTEN%muta*ons%Addi7on%of%%TP53%muta7ons%Addi7onal%muta7ons%(PIK3CA,%ARID1A,%CTNNB1,%%KRAS,%MSI)%Addi7onal%muta7ons%(PIK3CA,%ARID1A,%KRAS)%EEC+3% EEC/ESC%mixed% Carcinosarcoma%Less%Aggressive%Tumours%Carcinosarcoma% ESC%Problema7c%morphological%diagnoses:%high+grade%subtypes%%Endometrioid+like%differen7a7on%Serous+like%%differen7a7on%Aggressive%Tumours%TP53%muta7ons%Addi7onal%muta7ons%(PPP2R1A,%etc)%51   Figure 3.5 Mutational analysis may be an effective tool to classify morphologically problematic cases into biologically relevant treatment groups Intermediate high-grade cell types tend to be diagnostically challenging cases, often with multiple morphological features of endometrioid and/or serous carcinomas. The addition of mutation profiles can lead to reproducible diagnosis and the future of mutation-based treatment decisions for targeted therapeutics. Blue and red colours indicate distinct mutation profiles for low-grade EEC and serous carcinomas. Yellow indicates the cases were the mutational profiles will aid in separating out the appropriate histological subtype and dictate appropriate treatment options for patients.        Type%1%Stable%Histological%%Diagnosis%Less%Aggressive%disease%Predictable%Muta<on%Profile%Problema<c%Histological%Diagnoses%Aggressive%disease%Muta<on%Profile%May%Define%Treatment%Op<ons%%Type%2%Stable%Histological%Diagnosis%%Very%Aggressive%Disease%Predictable%Muta<on%Profile%Low$Grade*Endometrioid*Carcinoma**(PTEN,&PIK3CA,&ARID1A,&KRAS,&CTNNB1&muta4ons)*Serous*Carcinoma**(TP53,&PPP2R1A,&&other*muta4ons)*Mixed*Serous*and*Endometrioid*Carcinoma*High$Grade*Endometrioid*Carcinoma*Carcinosarcomas*Endometrioid$like*muta4on*profile*Serous$like*muta4on*profile*+&TP53&muta4ons*+&TP53&muta4ons*Muta<onBbased%treatment%decisions%Muta4ons*Will*E hanc *Extant*Classifica4on*Systems**52       I also investigated the molecular profiles of carcinosarcomas. These tumours are generally rare with poor prognosis [187], and are composed of a mixture of carcinomatous and sarcomatous elements [188]. While previous studies have not identified a high number of mutations in this subtype [189], I have shown a moderate frequency of mutations in the majority of genes sequenced. This discrepancy may be due to limited exon sequencing in previous studies; in the current study all exons of these genes were interrogated. Two patterns of mutations were observed; an endometrioid-type mutation profile (ARID1A, PTEN, PIK3CA, KRAS) or a serous mutation profile (TP53, PPP2R1A). This suggests a dualistic molecular evolution of carcinosarcomas with an endometrioid-like or serous-like mutation pattern, however this analysis is limited by the 9-gene panel and may underestimate the mutation patterns (Figure 3.4).  Larger gene panels, whole exome, or whole genome sequencing would be useful to validate these endometrial mutation patterns in carcinosarcomas. Further validation studies will also be necessary to determine if these molecular profiles are associated with different morphological features in the carcinomatous or sarcomatous components, or are associated with outcome differences.        I acknowledge that there are limitations of this study; the pathologist and I were unable to perform full histopathological reviews of many cases, including all carcinosarcomas. There were also limited numbers of cases of EEC-3 and ESC in this study, therefore independent validation studies, linked with outcome [30], will be needed in these tumour types. There is also uncertainty about the sensitivity of the exon capture method, and false negatives are likely to be present in this data set. The TCGA endometrial sequencing effort will prove to be useful in validating the observations of this study.        After the publication of this body of work, I completed an in-depth sequencing analysis of PPP2R1A, FBXW7 and TP53 mutations in a large number (n=89) of endometrial serous cases. The mutation frequency of PPP2R1A was 36% in this cohort as compared to the TCGA data (23%), although my results aligned with Chapter 2 results (40% of cases with PPP2R1A mutations). Mutations in TP53 were slightly lower than what was observed by TCGA. This could be due to differences in sequencing methods and the population of the cohort. Although, this difference is likely due to sequencing FFPE material, which caused a significant amount 53 sequencing noise due to deamination (G to A and C to T) events. This resulted in areas that were difficult to confirm the presence of mutations. In addition, a second limitation of the TP53 sequencing was that exon 5 and 11 primers did not reliably amplify in some samples; therefore all TP53 exons were not always sequenced efficiently. To resolve this problem, the use of immunohistochemistry could be used to determine if aberrations are present in the tumours.        This sequencing experiment was based on the analysis of TCGA data, where there appeared to be a near mutual exclusive trend of mutations in PPP2R1A and FBXW7, independent of TP53 mutations. It has been previously shown that the PP2A pathway acts by dephosphorylating c-Myc at S62 [121], and PP2A-B56δ acts to dephosphorylate GSK3β to enable phosphorylation of c-Myc [190]. In addition, FBXW7 acts to negatively regulate c-Myc by ubiquitination and proteasome-dependent degradation [119], therefore it is possible that the mutual exclusive pattern of mutations could lead to two different pathways of activation, and converge on c-Myc. Deregulation of PP2A phosphatase activity induced by a PPP2R1A mutation could potentially increase c-Myc activity by increasing the phosphorylation levels of c-Myc, thus stabilizing the protein. Alternatively, a mutation in FBXW7, a E3 ubiquitin ligase, could disrupt its ability to ubiquitinate c-Myc, resulting in protein stabilization and lack of protein degradation. This has been demonstrated in a mouse model of leukemia where FBXW7 R465C mutation is incapable of c-Myc ubiquitination and leads to increased c-Myc protein levels [191]. However, in this cohort of 89 serous carcinomas there was no significant associations of PPP2R1A or FBXW7 mutation with the accumulation of c-Myc as assessed by IHC. There was however a significant association of high c-Myc expression when both genes, PPP2R1A and FBXW7 are mutated, although the number of cases are small. It is possible that in this specific system, deregulation of both genes is needed for the stabilization of c-Myc protein levels. Previous studies have shown that PP2A B56α (PPP2R5A) specifically regulates the phosphorylation level of c-Myc at S62, which leads to stabilization and protein accumulation [121]. This was tested in the cell line HEK-293 model, therefore it is possible that PPP2R1A mutations in endometrial carcinomas do not specifically affect the B56α interaction, or there are cell context specific differences. A second study using HEK-293 cells has also shown that a feedback loop wherein c-Myc can induce transcriptional upregulation of PPP2R5D (B56δ), which can then participate in the dephosphorylation of GSK3β–S9. This dephosphorylation causes activation of GSK3β enabling the phosphorylation 54 of c-Myc to enable degradation [190]. Additional studies of these proteins with PPP2R1A mutations, B subunit alterations and B subunit interactions are needed in context-specific disease models. It is likely that the overall biological system may differ in each disease type. Overall, the mutational analysis of these genes in endometrial serous carcinomas are important, as they define the disease to aid in classification, and may be used to design patient targeted therapies.   55 Chapter 4: Differences in the Mutation Profiles of Ovarian Endometrioid and Endometrial Endometrioid Carcinomas  4.1 Introduction      Ovarian endometrioid carcinomas are histologically similar to endometrial endometrioid carcinomas. There is accumulating evidence that OECs arise from transformed endometriosis [192], which would link a common cell of origin; endometrial epithelial cells that line the uterine cavity. The molecular features of OECs and EECs have been characterized in many studies using immunohistochemistry markers and mutational analysis by DNA sequencing (reviewed in [29, 60, 77, 193]). There have been studies that infer differences in mutation frequencies in OECs and EECs [74], however to the best of my knowledge, no prior studies have directly compared endometrial and ovarian endometrioid mutation frequencies with a uniform technical approach in a large cohort of tumors. In the previous Chapter, the mutation profiles from a panel of nine genes was investigated in the various subtypes of endometrial carcinoma [2]. Herein, I have also performed the same exon capture sequencing method and applied this to a cohort of ovarian endometrioid carcinomas. The goal of this study was to directly compare the mutation frequencies of seven specific genes in low-grade EECs and low-grade OECs to determine differences in molecular features. These features may be useful in ovarian and endometrioid classification and targets for targeted therapeutics.  4.2 Materials and Methods 4.2.1 Patient Samples      I obtained frozen tumor tissue from 129 EECs and 33 OECs, for exon capture sequencing, originating from the OvCaRe Tissue Biobank repository, Vancouver, BC, Canada. Patients provided informed consent, and research ethics was approved, and DNA extracted as previously described [1]. An additional 178 EEC tumor DNA samples, with grade information available, were obtained from Washington University, St. Louis, Missouri. All samples from both centers have undergone histological review by gynaecological pathologists. All subtype and mutational data for all endometrial carcinomas have been previously reported [2]. An additional 20 OEC cases (18 frozen tumors, 2 Formalin-Fixed Paraffin Embedded were obtained from the OvCaRe Tissue Biobank repository to test CTNNB1 mutation status, so that a total of 53 OECs were 56 tested for mutation in this gene only, and 33 OECs were tested for the other genes in the panel. Normal germline DNA, when available, was used for testing somatic status from mutation positive cases only from the OvCaRe Tissue Biobank.  4.2.2 Mutation Analysis      Genomic DNA (500ng) was used for indexed Illumina library construction, then underwent targeted enrichment using biotinylated RNA capture probes generated from cDNA clones or PCR amplicons representing exons of ARID1A, PTEN, PIK3CA, KRAS, CTNNB1, PPP2R1A, TP53, BRAF and PPP2R5C and sequenced using the Illumina (GAIIx). All sequencing validation methods and primers used are previously described [2]. The two genes BRAF and PPP2R5C were excluded from subsequent analysis due to only one mutation found in BRAF and no mutations in PPP2R5C in the OEC cases. To validate the differences in CTNNB1 mutation frequencies, I re-sequenced the hotspot exon 3 of CTNNB1 by Sanger sequencing from all 178 low-grade endometrial endometrioid DNA obtained from Washington University. Additional Sanger sequencing validations for the hotspot exon 3 of CTNNB1 were also performed on 20 low-grade OEC cases that were not included in the original select exon capture sequencing set.  4.2.3 Bioinformatics Analysis      Short reads were aligned to the human genome (hg18) using the BWA aligner v0.5.9 [171]. A Random Forest classifier trained on validated single nucleotide variants was used to remove false-positive calls [172]. Single nucleotide variants in the Catalogue of Somatic Mutations in Cancer (COSMIC) [173] were considered to be true positives, with a probability threshold cutoff for selecting positive SNVs of 0.2588 (Figure B.1 (Appendix B)). All analysis was performed as previously described [2], and can be found in Chapter 3 materials and methods and Appendix B Supplemental methods.   4.2.4 Statistical Analysis       Fisher exact tests were used to test the significance of associations between mutations within subtypes. All tests were two-tailed and p-value < 0.05 were considered significant. The Benjamini-Hochberg [194] method was used to adjust p-values to account for multiple comparisons (R stats package). 57  4.3 Results      To determine the differences in somatic mutation frequencies between endometrial and ovarian endometrioid carcinomas, I used select gene exon capture sequencing of ARID1A, PTEN, PIK3CA, KRAS, CTNNB1, PPP2R1A, and TP53. This was performed using 33 cases of ovarian endometrioid and 307 cases of EECs (Supplemental Table C.7 (Appendix C)). Comparison of the mutational frequencies of low-grade (grade 1 and 2) EECs to low-grade (grade 1 and 2) OECs showed a significant difference for PTEN (adjusted p<0.0007) and CTNNB1 (adjusted p=0.02) mutations (Table 4.1, Figure 4.1). PTEN mutations were found in 67% of low-grade EECs, while 17% of low-grade OECs harbor PTEN mutations. Mutations of CTNNB1 were identified in 53% of OECs, and only 28% of EECs. This frequency of CTNNB1 mutations in EECs is slightly higher than previously published [2] due to additional validations of mutations using Sanger sequencing and select exon capture sequencing. To further verify the high CTNNB1 mutation frequency found in OECs, I acquired an additional 20 low-grade OECs cases and tested for hotspot CTNNB1 mutations using direct Sanger sequencing. I found that 45% (9 of 20) of these cases contained hotspot CTNNB1 mutations, bringing the overall frequency to 50% (25 of 50) CTNNB1 mutations in low-grade OECs (Supplemental Table C.7 (Appendix C)). The mutation frequencies of PIK3CA, ARID1A, PPP2R1A, and TP53 are not significantly different between low-grade EECs and OECs. There is a trend towards more KRAS mutations in low-grade OECs (33%) compared to low-grade EECs (18%), however this is not significant after multiple comparison adjustments (Table 4.1).        58  Low-Grade Ovarian Endometrioid           (Grade 1 and 2) (n=30) Low-Grade Endometrial  Endometrioid                 (Grade 1 and 2) (n=276) Fisher Exact Test (p-value) Adjusted       p-value* PTEN 5 (16.6%) 185 (67.0%) 1.08e-07 0.001 PIK3CA 12 (40.0%) 107 (38.7%)+ 1 1 ARID1A 9 (30.0%) 129 (46.7%) 0.086 0.120 KRAS 10 (33.3%) 50 (18.1%)+ 0.055 0.120 CTNNB1 16 (53.3%) 76 (27.5%)+ 0.006 0.020 PPP2R1A 5 (16.6%) 19 (6.9%) 0.071 0.120 TP53 2 (6.6%) 28 (10.1%) 1 1          Table 4.1 Comparison of mutation frequencies in low-grade OECs and low-grade EECs        *p-values are adjusted using the B-H method        + additional mutations have been verified by Sanger sequencing post-original publication.    The comparisons of high-grade (grade 3) EECs with high-grade OECs cannot be statistically compared, as this is limited by the rarity of high-grade OECs (n=3). Each high-grade OECs case harbors different mutations in different genes. Similarly to low-grade EECs, high-grade EECs also have a high frequency of PTEN mutations (87%), and a lower frequency of CTNNB1 mutations (19%) (Table 4.2).   High-Grade Ovarian  Endometrioid (Grade 3) (n=3) High-Grade Endometrial Endometrioid  (Grade 3) (n=31) PTEN 1 (33.3%) 27 (87.1%) PIK3CA 1 (33.3%) 17 (54.8%) ARID1A 0 (0%) 18 (58.1%) KRAS 1 (33.3%) 8 (25.8%) CTNNB1 1 (33.3%) 6 (19.4%) PPP2R1A 0 (0%) 4 (12.9%) TP53 1 (33.3%) 10 (32.2%)                                  Table 4.2 The mutation frequencies of high-grade OECs                                     and high-grade EECs. 59  Figure 4.1 Low-grade ovarian and endometrial endometrioid mutation profiles.  A. Low-grade endometrial endometrioid carcinomas (EEC) including grade 1 and 2 (n=276).  B. Low-grade ovarian endometrioid carcinomas (OEC) including grade 1 and 2 (n=30). Individual columns designate one tumor case, and rows indicate genes. All colored boxes specify a genetic alteration such as missense, truncating, indels, splice site mutations and combinations of these mutations. Grey boxes indicate no mutations were identified by sequencing. These colors are specifically shown in the color legend.  60            Most CTNNB1 mutations found in both OECs and EECs involve known phospho-acceptor sites. In OECs, 16 of 25 (64%) contain CTNNB1 mutations that affect serine (S33 and S37) or threonine (T41) amino acid residues, which are phosphorylation targets for glycogen synthase kinase 3-beta (GSK3β). Similarly in low-grade EECs, 43 of 76 (57%) of the CTNNB1 mutations are located at serine (S33, S37, S45) and threonine (T41) residues (Appendix C Table C.7). Our analysis of all CTNNB1 coding sequences also revealed additional somatic mutations outside the hotspot serine/threonine residues.  4.4 Discussion      The distinct molecular abnormalities of endometrial and ovarian carcinoma subtypes will be the basis for subtype specific treatment, and may become an essential component of stratified management strategies. Standard treatment options for ovarian and endometrial carcinomas have not yet changed, but a shift towards subtype-specific clinical trials highlights the need to better understand the molecular abnormalities and potential therapeutic targets in the different subtypes [195]. The same subtypes at different sites in the gynaecological tract, endometrial endometrioid carcinomas and ovarian endometrioid carcinomas, have indistinguishable morphology, clinical similarities, and share at least some genetic abnormalities. I have directly compared the mutation frequencies in the same gene set, using the same technology, in low-grade ovarian and endometrial endometrioid carcinomas. This has shown that there are similar mutations patterns but there are also two distinct differences.        Previous literature often refers to endometrial and ovarian endometrioid carcinomas as having similar molecular alterations [196, 197]. The same genes are often mutated, however some suggest that OECs and EECs have similar frequencies of CTNNB1 [74] and PTEN mutations [198], but do not directly compare the two tumor types in the same study. In this study, PTEN mutations are found at higher frequencies in low-grade EECs compared to low-grade OECs, and CTNNB1 mutations at higher frequencies in OECs compared to EECs. One limitation of this study is the small number of low-grade OECs (n=30) OECs. Previous studies have reported low-grade OECs with CTNNB1 mutation frequencies between 16-54% [75, 199-204], with an average frequency of 40-50%. This range of CTNNB1 mutation frequency is due to varying 61 methods of detection and limited exon sequencing. PTEN mutations are reported in only about 20% of ovarian endometrioid carcinomas [73, 205]. All these studies also assessed small cohorts of OECs, and reported similar mutation frequencies for both CTNNB1 and PTEN.        Recently, our group [2] and others [27, 28, 30] have reported mutation frequencies in a large number of EECs. Byron et al. report a 19% CTNNB1 mutation frequency from 466 endometrial endometrioid tumors [30]. In this study I report a slightly higher CTNNB1 mutation frequency of 28% from 276 low-grade EECs. This difference is most probably due to differences in sequencing (hotspot vs. all exons) and analysis methods used. The current study included 25 EECs tumor samples also analyzed by Byron et al.; there were 66 CTNNB1 and 46 KRAS mutations reported in the original paper [2], whereas I have now identified 76 CTNNB1 and 50 KRAS mutations in these same tumours, all validated by Sanger sequencing, indicating a low false negative rate in the earlier study. This does not change the conclusions of that study [2] but does indicate that there was an underestimation of mutation frequencies. In conclusion, this study confirms observations from other studies suggesting there are differences in PTEN and CTNNB1 mutation rates in OECs and EECs. Similarly, ovarian and endometrial serous carcinoma subtypes are morphologically equivalent and were often thought to have similar mutation patterns, with both showing a high frequency of TP53 mutations. However, more detailed sequencing analyses of these tumor types have revealed mutational profile differences. Mutations of PPP2R1A are found at high frequencies in endometrial serous carcinomas but are rarely found in ovarian high-grade serous carcinomas [1, 206]. More recent studies have identified high frequencies of other gene mutations (i.e. FBXW7, CHD4) [43, 46], in endometrial serous carcinomas but not in ovarian serous carcinomas [44].         The majority of OECs are believed to arise from endometriosis [60, 76]. Although ovarian and EECs may develop from the same cell type, namely the endometrial epithelial cell, these two tumor types are exposed to different microenvironments that may reflect differences in the evolution of their mutation profiles. Endometrial endometrioid carcinomas frequently occur in postmenopausal women with unopposed estrogen [207]; exposure to high estrogen and low progesterone levels have been found to increase proliferation of endometrial cells thus increasing the risk of tumourigenesis [208]. Endometriosis is thought to occur via retrograde menstruation, 62 where endometrial epithelial cells travel from the uterus through the fallopian tubes and can establish as an endometriotic cyst within the ovary [67]. This creates a unique microenvironment where menstruation-like blood and necrotic tissue is trapped within the cyst, resulting in high concentrations of iron in a confined space [209], causing oxidative stress and a hypoxic environment leading to DNA damage and mutation accumulation [210, 211]. The cells that embed in the ovary are then exposed to a proliferative microenvironment known to be an inflammatory milieu where malignant transformation can occur [212]. Chromosomal aberrations have been identified at a high frequency in ovarian endometriotic cysts compared to extra-gonadal endometriosis [213]. The different selection pressures found in the uterus compared to the ovary are likely factors in the evolution of the tumourigenic cells. The mutational analysis of a small number of endometriosis lesions has mostly been confined to a subset of genes, PTEN, CTNNB1 and KRAS with only a small number of somatic mutations found in PTEN [73], and KRAS [214]. It will be important to determine the malignant transformation pathways of endometriosis to ovarian endometrioid carcinomas, and in so doing, identify the genetic differences between the precursors of endometrial and ovarian endometrioid carcinomas and from endometriosis with low potential for transformation. In addition to considering the microenvironment the impact that these mutations have on different cell context must be considered. The use of endometriosis derived cell lines as opposed to endometrial cell lines may be of critical importance for studying the origins of these cancers [215].       Based on the mutation frequencies found in this study, CTNNB1 mutations in the ovary and PTEN mutations in the endometrium are characteristic features of these diseases. Mutations in CTNNB1 and deregulation of the Wnt pathway are a well-established pathway in cancer signaling first characterized in colorectal cancers [216, 217]. As seen in our study, the majority of CTNNB1 mutations change the phospho-serine/threonine sites, which affects the ability of GSK3β to phosphorylate β-catenin to signal degradation. Lack of β-catenin phosphorylation results in nuclear accumulation causing expression of cell proliferation and inflammatory genes (reviewed in [218]) (Figure 4.2). Ovarian carcinomas with an accumulation of nuclear β-catenin,   63             Figure 4.2 PI3K/AKT and WNT signaling pathways.            These two signaling pathways show convergence on GSK3β and β-catenin. Genetic alterations caused by mutations in both pathways             can result in the transcription of cell growth and proliferation genes. Mutations are indicated by black stars and are found in both ovarian            and endometrial endometrioid tumor 64 either due to CTNNB1 mutations or deregulation of other Wnt family member like APC or Axin, is an indicator of good prognosis [75]. Similar to the Wnt pathway, PI3K signaling is also one of the most commonly altered cancer pathways [219]. PTEN acts as a negative regulator of the PI3K pathway by dephosphorylating the signaling lipid molecule PIP3 to PIP2; where PIK3CA (p110α) together with PIK3R1 (p85α) acts to phosphorylate PIP2 to PIP3, thus allowing signaling to proceed through AKT and mTOR. Mutations in both PTEN and PIK3CA can act to maintain constitutively activated PI3K signaling (reviewed in [220]). This activation leads to the degradation of GSK3β, and thus allows β-catenin nuclear translocation [221]. The PI3K and Wnt pathways do not occur linearly, and are interactive within their own signal transduction networks as well as with other pathways, which is shown by convergence on GSK3β (Figure 4.2). There are however many other gene regulation events resulting through the activation of PI3K/AKT/mTOR pathway that are not activated through the Wnt pathway. Therefore, in OECs that do not respond to standard treatments, it may be beneficial to target both the Wnt pathway, to inhibit β-catenin and the PI3K pathway when PI3KCA is mutated. In the case of EECs, PTEN and PIK3CA are both frequently mutated, likely leading to the up-regulation of PI3K signaling, thus targeting the PI3K pathway may be of benefit. Additionally, there is no mutual exclusivity of CTNNB1, PTEN or PIK3CA mutations in EECs and OECs (Figure 4.1), indicating that they are not functionally equivalent, therefore when both pathways are mutated; simultaneously targeting the PI3K and Wnt pathways may be appropriate. Careful consideration will be needed when deciding which molecules to target in one or multiple pathways, as well as the specific cellular context.        Ovarian and endometrial endometrioid carcinomas share obvious histogenic connections and are morphologically similar, however there are genomic differences, as shown by CTNNB1 and PTEN mutation frequencies. The occurrence of these mutations may reflect different environmental niches during oncogenesis, and ultimately point toward different routes of distinct targeted therapeutics. 65 Chapter 5: Construction of Endometrial Hec1A Isogenic PPP2R1A Cell Lines   5.1 Introduction      The most definitive way to assess gene function is by targeting the endogenous genome [222]. Gene overexpression and knockdown approaches of genes in cell lines or mouse models have traditionally been the approach of molecular functional studies. Overexpressing genes has given insight into how mutations or genes may affect the biology of the cell, however this is often an artificial system, and is frequently performed in cell lines or models that do not have any relevance to disease or cell type. Knockdown approaches have mostly been accomplished using small interfering RNA (siRNA) technologies, however this technique has limitations with high non-specific effects and incomplete inactivation of the gene of interest [223, 224].  These approaches, although have been useful, cannot compare genetic variants and do not recapitulate the naturally occurring genetic variants found in specific cell types or diseases [225].       Knockout (KO) and knockin technology in mammalian cells was traditionally very difficult and inefficient. Nevertheless, Fred Bunz and Bert Vogelstein at Johns Hopkins University utilized the recombinant adeno-associated virus (rAAV) to enable efficient somatic cell knockout by homologous recombination into mammalian cells [222, 226]. This rAAV infection approach was found to be 25-fold more efficient than transfection of plasmids for homologous recombination into the same exon. The rAAV technology exploits endogenous homologous recombination to insert a specific promoter-trap targeting construct into the endogenous gene. Thus allowing the targeted insertion of mutations or deletions into one allele or the knockout of a specific allele. The promoter-trap requires integration of a promoterless construct into the proximity of an active gene promoter to drive expression, and has been used to successfully target human loci [227]. The two diploid colorectal cancer cell lines (HCT116, DLD1) have been mostly used for this technology as they are mismatch repair (MMR) deficient thus allowing for homologous recombination to occur. Most diploid or near diploid cell lines are MMR-deficient with stable chromosomes [228]. The inactivation of MMR genes leads to the accumulation of many mutations in cancer genes, called a mutator phenotype, and the development of cancer [229]. The rAAV method has been used to target many genes including CTNNB1 [226] and PIK3CA [230] loci. 66       Previous molecular studies of PPP2R1A mutations have been based on a few very low frequency cancer mutations and experiments were performed in unrelated disease cell types [90, 131, 153]. In Chapters 2 and 3, I have presented data where I discovered a high frequency of PPP2R1A mutations in multiple types of endometrial carcinomas (serous, endometrioid, carcinosarcoma) [1, 2]. It is therefore important to study these mutations in the context of endometrial or ovarian specific cell types. To overcome the limitations of overexpressing genes, in this Chapter, I describe the generation of isogenic endometrial-derived cell lines using rAAV somatic cell knockout technology to further study an endogenous PPP2R1A mutation. As part of the endometrial exon capture analysis utilized in Chapter 3, four endometrial derived cell lines (Hec1A, KLE, ECC1, Hec50) were also sequenced (data not shown). There were two endometrial cell lines that harbor PPP2R1A mutations: the Hec1A cell line with a heterozygous W257L mutation, and the Hec50 cell line with a homozygous R183W mutation. The Hec1A cell lines originates from a human endometrial adenocarcinoma [231], is mostly diploid and MMR-deficient with a PMS2 nonsense mutation [232], which allows for homologous recombination utilized by the rAAV somatic cell knockout technique. Since this cell line is MMR-deficient it harbors a hyper-mutator phenotype, with mutations in almost every gene, and could be classified with the MSI group according the TCGA classification [42]. Using the Hec1A cell line, PPP2R1A mutant and wild-type alleles can be targeted using the rAAV somatic cell knockout system, and be used to determine the effects of PPP2R1A mutations on cell proliferation and migration. The Hec50 cell line with a homozygous R183W mutation is not MMR deficient, therefore was not a good candidate cell line for somatic cell gene knockout or knockin.  The generation of the Hec1A isogenic PPP2R1A cell line clones is described in this Chapter 5, and was crudely characterized for cell proliferation and migration characteristics. The main purpose of generating these isogenic cell lines was not to characterize a new cell line, but to use as a cell-context specific model for PPP2R1A interactions studies presented in Chapter 6. This includes performing immunoprecipitation coupled with mass spectrometry experiments to determine how the PPP2R1A mutation affects binding of PP2A B subunits and other known PP2A interactors.   67 5.2 Materials and Methods  5.2.1 Cell Culture Maintenance      Endometrial Hec1A cells [231] were grown in McCoy’s 5A media (Hyclone) with 10% fetal bovine serum (FBS). All cell lines were maintained at 370C with 5% CO2.   5.2.2 Creation of Somatic Cell Knockout Hec1A Cell Lines 5.2.2.1 Cloning PPP2R1A Exon 3 and Exon 4 Regions into the Targeting Vectors pSEPT and pAAV-MCS      The endometrial cell line Hec1A containing an endogenous missense W257L (c.770 G>T) mutation (present in exon 6) was chosen for somatic cell knockout, to isolate isogeneic cell lines that express only mutant or wild-type PPP2R1A. The construct design is based on using recombinant adeno-associated viral vector (rAAV) system described in two publications [222, 226]. The synthetic exon promoter trap pSEPT (plasmid synthetic exon promoter trap) vector was obtained from Dr. Fred Bunz’s lab at Johns Hopkins Medicine, USA [226]. Two separate pSEPT constructs were designed to target exon 3 and exon 4 of PPP2R1A. This will enable incorporation of the knockout construct upstream of the mutation that is present in exon 6, causing loss of expression of the targeted allele. To clone the PPP2R1A regions of interest into the pSEPT vector, primers were designed to incorporate PPP2R1A homology regions (HA) and restriction enzyme sites. Using the Clontech Online In-Fusion primer design tools (http://www.clontech.ca), two homology arms (left and right) were designed for each of PPP2R1A exon 3 and 4.           68 Primer Name Primer Sequence RE RE sequence PPP2R1Ax3_LHA1_F CCGCGGTGGCGGCCGCAAGGAAGAGGCAGAGATACTAACC NotI GCGGCCGC PPP2R1Ax3_LHA1_R ATCCACTAGTTCTAGACAGCACTCAGTTCTTCCATCC XbaI TCTAGA PPP2R1Ax3_RHA1_F GCATATGTATGAATTCGGAACCTTCACTACCCTGGTG EcoR1 GAATTC PPP2R1Ax3_RHA1_R GCTTGATATCGAATTCGCGGCCGCTCAAATCCCAAGATCCCAAC EcoRI, Not1 GCGGCCGC PPP2R1Ax4_LHA_F ACCGCGGTGGCGGCCGCCCTGTCAGCCCAAGTTGAAT  NotI GCGGCCGC PPP2R1Ax4_LHA_R ATCCACTAGTTCTAGACCTATTACCCATCCCGACCT  XbaI TCTAGA PPP2R1Ax4_RHA_F GCATATGTATGAATTCGGACAAGGCAGTGGAGTCCT  EcoR1 GAATTC PPP2R1Ax4_RHA_R GCTTGATATCGAATTCGCGGCCGCAGGCAGGTCTAGAGCCACAG EcoRI, Not1 GCGGCCGC Table 5.1 Homology arm primers for cloning PPP2R1A exon 3 and 4 into pSEPT The primers for LHA: left homology arm, RHA: right homology arm with restriction enzyme sites (RE) and RE sequence. The bold primer sequence contains pSEPT vector-specific sequences with restriction enzyme sequences.        Each HA was PCR amplified from Hec1A DNA using iPROOF High-fidelity polymerase (BioRAD) and purified using the PCR purification kit (Qiagen). Each HA was cloned stepwise into the pSEPT vector (Figure 5.1), using the Clontech In-Fusion Cloning kit as per manufacturer’s protocol. The targeting cassette was sequenced, using Sanger sequencing, off the pSEPT backbone to make sure the cloning was efficient and to check for possible errors in the sequence. The HA targeting PPP2R1A cassette was then excised from the pSEPT vector backbone using Not1, then purified using the QIAquick gel purification kit (Qiagen). Finally, the pAAV-MCS vector was also linearized with Not1 and purified to allow cloning of the HA targeting PPP2R1A cassette into the vector backbone using T4 ligase. The final vector with the targeting cassette was sequenced using Sanger sequencing. 69                                      Figure 5.1 pAAV-PPP2R1A exon 3 targeting vector         The SEPT cassette includes a splice acceptor (SA), Internal Ribosomal Entry Site      (IRES), neomycin selection (neo), polyadenylation site (pA). The red arrows indicate      LoxP sites. The pAAV-PPP2R1A targeting vector was utilized to make adenovirus in      AAV-293 cells, then harvested to transduce into the Hec1A parental cell line to target     the PPP2R1A exon 3 or exon 4 locus.  5.2.2.2 Generating PPP2R1A Exon 3 and 4 Hec1A KO Cell Lines      The final pAAV-PPP2R1A exon 3 and 4 targeting vector (Figure 5.1) was then transfected into AAV-293 cells to package the recombinant virus using the AAV-Helper-Free System (Agilent) as per manufacturer’s protocol. The virus was harvested and transduced into the Hec1A cell lines according to Rago et al [225]. For single cell selection, the cells were harvested, diluted to low concentrations and seeded into 10 X 96 well plates with 1200ug/mL G418 selection media (Geneticin, Life Technologies). Once the colonies had grown to cover the majority of a single well of a 96 well plate, (3-5 weeks) (Figure 5.2), the cells were transferred to larger tissue culture vessels, and tested for integration of the knockout allele.   1kb$ 1kb$ITR$ PPP2R1A$Right$HA$ ITR$SEPT$ IRES$ neo$=$SA$ pA$2 4$ 6"5Not1$ Xba1$ EcoR1$ EcoR1,$Not1$LoxP$LoxP$2 4$ 6"53$Wild$type$allele$Mutant$allele$PPP2R1A$LeG$HA$pAAVIPPP2R1A$targeKng$vector$SEPT$SEPT$70                               Figure 5.2  pAAV-PPP2R1A exon 3 targeted Hec1A single cell colony                              Bright field image of a single cell colony after 6 days on selection, and                               8 days on selection medium. The black cells are dead cells that are found      mostly surrounding the growing colony.  5.2.2.3 Sanger Sequencing of Hec1A Single Cell Knockout Colonies      To determine which PPP2R1A allele was targeted in the single cell Hec1A clones, RNA was extracted from cell pellets using the QIAamp mini RNA extraction kit (Qiagen) as per manufacturer’s protocol. The cDNA was synthesized using First Strand cDNA synthesis kit (Invitrogen), and proceeded as per manufacturers protocols. The PPP2R1A exon 6 region surrounding the W257L (c.796-771) mutation was amplified using cDNA specific primers (PPP2R1A_c381F CTTTGTGCCGCTAGTGAAGC, PPP2R1A_c889R CGGCCTCACAGTCTTTCATC), then sequenced using the same primers on the ABI 3130xl capillary sequencer in the forward and reverse directions. Sequences were analyzed using the Applied Biosystems Sequencing Analysis 5.2 software, BLAST, and visualized using FinchTV  (Geospiza, Inc). Once positively targeted clones were identified, Cre recombinase was utilized to excise the neomycin cassette (pSEPT) flanked by LoxP sites (see Figure 5.1). Ad-CMV-Cre (Cre Recombinase Adenovirus) (Vector Biolabs) was used for excision of the pSEPT construct as per Rago et al [225]. Phenotypic screening using with the same concentration of Geneticin in which they were first selected, resulted in choosing positive clones that had undergone correct pSEPT excision.  71 5.2.2.4 Ion Torrent Sequencing of PPP2R1A Exon 6 in Hec1A Cell Line      To determine how many alleles of mutant PPP2R1A W257L (c.769-771) were expressed in the parental Hec1A cells, RNA was extracted and cDNA was synthesized as described in the Sanger sequencing methods. Targeting primers were designed using Primer 3, and Ion Torrent adapter sequences were added to the target sequence and synthesized by IDT (Integrated DNA Technologies). For bidirectional sequencing, two sets of primers were synthesized and used for amplification (Table 5.2). PCR was performed using the BioRAD iPROOF Hi-fidelity enzyme, as per manufacturer’s protocol. The amplified products were checked on a 1% agarose gel, cleaned with magnetic Ampure beads, and then spiked into an existing Ion Torrent library at 1/100 dilution. The final library was loaded onto an Ion 316TM chip and sequenced on the Personal Genome MachineTM (PGM) Ion Torrent sequencer (Life Technologies).  Primer Name Adapter Full Primer Sequence PPP2R1A_ex6_F_A CCATCTCATCCCTGCGTGTCTCCGACTCAG CCATCTCATCCCTGCGTGTCTCCGACTCAGTCTGCCCCAGGAGGATCT PPP2R1A_ex6_R_trP1 CCTCTCTATGGGCAGTCGGTGAT CCTCTCTATGGGCAGTCGGTGATCACCAGTGGGACAATGTCAA       PPP2R1A_ex6_F_trP1 CCTCTCTATGGGCAGTCGGTGAT CCTCTCTATGGGCAGTCGGTGATTCTGCCCCAGGAGGATCT PPP2R1A_ex6_R_A CCATCTCATCCCTGCGTGTCTCCGACTCAG CCATCTCATCCCTGCGTGTCTCCGACTCAGCACCAGTGGGACAATGTCAA Table 5.2 Primer sets for amplifying PPP2R1A cDNA for Ion Torrent sequencing.  5.2.3 M-FISH and FISH      Metaphase cells for M-FISH and FISH were prepared by using colcemid (Sigma) treatment for 10 min, then fixed with methanol/acetic acid and stored at 4oC. To label endometrial Hec1A cells for M-FISH analysis, the MetaSystem multi-colour probe kit – 24XCyte and the MetaSystem B-tect kit was used as per manufacturer’s protocol. Imaging and analysis was performed using the MetaSystem analysis software on a Zeiss Axioplan epifluorescent microscope. For Hec1A FISH analysis, the probes were generated using BAC (bacterial artificial chromosomes) clones selected from the UCSC Human Genome Browser. BAC DNA was isolated using the rapid alkaline lysis miniprep method, then labeled with a nick translation kit as per manufacturer’s protocol (Abbott Molecular, Illinois, USA) using SpectrumOrange-11-dUTP or SpectrumGreen-11-dUTP. The specific probes to label chromosome 19 centromere and 72 PPP2R1A loci are as follows respectively: RP11-317K24 and RP11-13M19 at 19p12 (SpectrumGreen, Centromere region), RP11-890J19 at 19q13.33 (SpectrumOrange, PPP2R1A).  5.2.4 Cell Proliferation Assays      To assess cell proliferation by using the crystal violet assay, cells were seeded at various densities 10,000 cells/well, 20,000 cells/well, 40,000 cells/well in 6 replicates (number of wells =6) for each cell line into 12-well plates. Cells were plated for assessment from 0-9 days. At each time point, the cells were washed three times with PBS, fixed with 4% formalin for 10 min, then washed another three times with PBS. Plates were stored at 4oC, and all plates were simultaneously fixed and stained with 0.1% crystal violet solution for 30 min at room temperature. The crystal violet solution was washed off with distilled water and allowed to dry at room temperature. To extract dye from the cells, 2mL of 30% methanol and 10% acetic acid was added to each well and shaken for uniform colour. For each well, 150uL of solution was transferred to a 96-well plate and the absorbance was read at 590nm on a UV spectrometer plate reader. The absorbance readings were averaged with standard deviations reported. The automated IncucyteTM (Essen Biosciences) was also used to determine cell proliferation by monitoring cell confluence over a defined time period. Cells were seeded at different densities (5,000 cells/well and 10,000 cells/well) in a 96-well tissue culture plate (number of wells=4). Each well was then monitored, by taking an image with a 10X lens, every 6 hours for 4-5 days. Final analysis was performed and images viewed using the IncucyteTM ZOOM software. At each time point, the average cell confluence for each cell line was reported with standard errors reported.  5.2.5 Migration (Scratch-Wound) Assay       Each of the cell line clones (mut/mut clones: 9-14, 9-15, 33-1, 33-4; mut/wt clones: 10-23-45-3, 10-23-45-6, 10-23-60-17, 10-23-45-24) were plated at 100% confluence (~20,000cells/well) (number of wells=6) into a 96-well Essen ImageLock microplate, and allowed to grow and attach overnight. The next day, a sterile Essen 96-well WoundMakerTM was used to make a consistent wound through the middle of each well. All media was aspirated off, and carefully washed twice with culture media to get rid of any loose cells. Cell media (100uL) was added back to each well, and wells were then monitored for 4 days, every 4 hours, or until the wounds were healed using the IncucyteTM (Essen Biosciences). This protocol was adapted 73 from Essen Bioscience’s CellPlayerTM 96-well cell migration assay protocol. Final analysis, images and videos were performed using the IncucyteTM ZOOM software.   5.3 Results 5.3.1 Isolation of PPP2R1A Exon 3 and 4 Knockout Isogenic Hec1A Cell Lines      To generate Hec1A isogenic PPP2R1A cell lines, somatic cell knockout of PPP2R1A exon 3 and 4 was performed. Using this somatic cell technique, two different cell lines clones could potentially be isolated that express only the PPP2R1A L257 mutation or only the wild-type protein. The first round of transduction with the PPP2R1A targeting exon 3 in Hec1A cells resulted in 53 single colonies (Figure 5.2). The targeting of exon 4 was less efficient and resulted in 30 colonies. To test the efficiency of the knockout construct, the cDNA of all colonies was sequenced to determine which PPP2R1A allele was targeted. Three single cell colonies were isolated from the exon 3 knockout construct that expressed only the PPP2R1A (-) mutant L257 allele (clone #9, 33 and 37) (Figure 5.3). No mutants were isolated from the exon 4 knockout construct. There were also no single cell colonies that expressed only the wild-type allele. To try to isolate the PPP2R1A (wt) wild-type (W257) only expressing Hec1A cell line, two more rounds of transducing the PPP2R1A exon 3 targeting construct was completed. This resulted in an additional 95 single colonies being tested for the knockout of the PPP2R1A mutant or wild-type allele. No wild-type only colonies were identified, however 5 additional mutant only expressing cell colonies were isolated.   74 Figure 5.3  Sanger Sequencing Traces for Hec1A KO A. Sanger sequencing trace for the parental Hec1A cell line cDNA with an endogenous W257L  (G to T) mutation. B. Sanger sequencing trace for the cDNA of the mutant clone #9 with only the T nucleotide present (L257).        It was possible that the Hec1A cells are not actually diploid at the PPP2R1A locus (chr19: 52,190,039-52,229,533); therefore to determine a digital count of the allele frequency of the mutation, Hec1A cDNA was sequenced at the PPP2R1A locus using the PGMTM Ion Torrent next generation sequencer. This may give insight into the lack of isolation of the wild-type only expressing cell lines. This analysis revealed a digital count of 62% T mutant reads and 38% G reference reads giving a 2:1 mutant:wild-type allele expression of the PPP2R1A (G to T) W257L mutation in the Hec1A cell lines (Figure 5.4).  To determine if this was due to chromosome ploidy changes, M-FISH and FISH for the PPP2R1A gene region was performed using the Hec1A parental cell line (Figure 5.5). The M-FISH assay showed the parental Hec1A cell line as diploid at chromosome 19, however there was also a small piece of chromosome 19 that had been translocated to chromosome 17, as depicted by the purple on the aqua chromosome 17 (Figure 5.5A). FISH specific for the PPP2R1A region of chromosome 19 shows one normal chromosome and one chromosome with 2 regions of PPP2R1A similar to an isochromosome (Figure 5.5B). Therefore, together these results show that Hec1A parental cell line actually Only%mutant%allele%L257%Clone%#9%Mutant%Parental% W257L%A%B%75 contains 3 PPP2R1A alleles; two PPP2R1A (T) L257 mutated alleles, and one PPP2R1A (G) W257 wild-type allele, designated as (mut/mut/wt).  The knockout of one wild-type allele generates a cell with 2 PPP2R1A (mut/mut) mutant alleles, and the targeted knockout of one mutant allele generates a heterozygous PPP2R1A (mut/wt) cell.                    Figure 5.4  Ion torrent sequencing of the parental Hec1A cell line for W257L mutation                Integrated Genome Viewer (IGV) panel depicting the total reads for PPP2R1A cDNA exon 6.                 Each row indicates one sequencing read, with the total reads shown for each variant. 22571%Total%Reads%38%%G%62%%T%76     Figure 5.5  M-FISH and FISH for Hec1A cells A. M-FISH results of the Hec1A parental cell line. B. PPP2R1A specific FISH in metaphase Hec1A cells. The red probe is specific for the PPP2R1A locus (RP11-890J19 at 19q13.33), the green probe is specific for the chromosome 19 region (RP11-317K24, RP11-13M19 at 19p12).       In light of these results, the isolation of a wild-type only cell line could potentially be generated from the heterozygous single colonies isolated from the first round of targeted selection. A single Hec1A cell line (clone 10-23) that had undergone the Cre excision of the exon 3 targeting pSEPT vector, was used for a second round of targeting. A total of about 46 colonies were screened, and no 100% wild-type colonies were identified. All the cell clones screened had detectable levels of the PPP2R1A mutant L257 allele. The resulting cell line clones have equal to less mutant allele being expressed, pre-targeting, however are still mostly equivalent to a heterozygous PPP2R1A (mut/wt) W257L mutation. The resulting Hec1A PPP2R1A isogenic cell lines are shown in Figure 5.6. My inability to produce a pure wild-type isogenic line from Hec1A cells may be due to an inability of the wild-type cell clones to survive after selection.    Chr19&Chr17&A& B&77  Figure 5.6 Resulting Hec1A PPP2R1A isogenic knockout cell lines Blue boxes depict the wild-type exons, and the green exon 6 boxes encode a mutation (G>T). The red cross indicates which allele is being targeted by the pSEPT targeting construct, and thus not expressed in the cell line.  5.3.2 Growth Characteristics of Isogenic Cell Lines      To determine if the PPP2R1A mutation affects cell growth, two different methods were used to assess the cell proliferation of the isogenic cell lines. First, cell nuclei were stained using crystal violet dye, which is proportional to the number of live cells present. After 9 days of cell growth, starting with 10,000 cells/well, I discovered that the parental line proliferates at a faster rate than three clones of the mutant only expressing clones (#9, 33 and 37) (Figure 5.7A). This was replicated when the cells were seeded at a higher confluence of 40,000 cells/well, however the mutant clone #33 appears to proliferate at a faster rate than the other two mutant clones (Figure 5.7B). The heterozygous cell lines were not tested with this method because they were still in the process of being isolated and screened. 4 65#Wild#type#Heterozygous,(,express,mut/wt),Mutant,(express,only,mutant,allele,(mut/mut),Parental,(3,copies,mut/mut/wt),2 3 4 65#Mutant#2 3 4 65#Wild#type#2 3 4 65#Mutant#2 3 4 65#Wild#type#2 3 4 65#Mutant#2 3 4 65#Mutant#22 3 4 65#Mutant#2 3 4 65#Mutant#78       Figure 5.7 Cell proliferation of Hec1A mutant/mutant expressing cell lines using the      crystal violet assay      A. A total of 10,000 cells were plated (n=6) on Day1 and assessed every day for 9 days. B. A total of   40,000 cells were also plated (n=6) on Day 1 and assayed every day 9 days. The absorbance readings from n=6 for each cell line were averaged, with the error bars indicating standard deviation.       A second method of analyzing cell proliferation, based on cell confluence as measured by the IncucyteTM, also shows that after 4.5 days of cell growth, all of the mutant cells line clones proliferate slower than the parental and the heterozygous cell lines (Figure 5.8A). A closer look at only three of the cell lines (Figure 5.8B) reveals that one of the heterozygous cell clones may proliferate faster than the parental line. However, this is confounded by increased confluence of the heterozygous cell line, compared to the other two cell lines, at the beginning of the assay (time=0). When the changes in cell growth rates are compared; the parental and heterozygous cell lines show almost identical changes in growth rate (Figure 5.8C). Conversely, the mutated cell line demonstrations a slower rate of change over time compared to the other two cell lines. !0.5%0%0.5%1%1.5%2%2.5%3%0% 2% 4% 6% 8% 10%Absorbance*(590nm)**Rela3ve*cell*number*Days*Hec1A%parental%Hec1A%mut/mut%#9%Hec1A%mut/mut%#33%Hec1A%mut/mut%#37%!0.5%0%0.5%1%1.5%2%2.5%3%3.5%4%0% 2% 4% 6% 8% 10%Absorbance*(590nm)**Rela3ve*cell*number*Days*Hec1A%parental%Hec1AKO%#9%Hec1A%KO%#33%Hec1A%KO%#37%A%B%79        Figure 5.8 Cell proliferation of Hec1A PPP2R1A isogenic cell lines with differing expression      of mutant PPP2R1A. A. Cell proliferation, represented by phase cell confluence, of all cell lines (10,000cells/well) as monitored by the IncucyteTM over 120 hours. B. Cell proliferation of only one each of the parental, mutant and heterozygous cell line clones over time is shown. C. The rates of change of phase confluence as measured by the IncucyteTM to show the change in rate of proliferation of the trio of isogenic cell lines. Error bars indicate standard error.  A"B"C"Mutant"(mut/mut)"Clone"9215"Parental"(mut/mut/wt)"Heterozygous"(mut/wt)"Clone"1022324526"80      To determine the effects of growth factors on the proliferation of the different isogenic cell lines in triplicate, I assayed cells with varying concentrations of FBS (Fetal Bovine Serum). The concentration of FBS did not seem to affect the growth characteristics of the parental and heterozygous (clone 10-23-45-6) cell lines, however there were differences in growth patterns for the mutant cells (clone 9-14) (Figure 5.9).  At low levels (1% and 0.1%) of FBS, the rate of proliferation of the mutant cell line (clone 9-14) was different than the regular growth conditions (10% FBS).   Figure 5.9 Effect of different FBS concentration on the proliferation of Hec1A PPP2R1A isogenic cell lines.  Panels A, B and C depict cell proliferation over time for each of the each of the Hec1A PPP2R1A isogenic cell line clones in triplicate with varying levels of FBS. Error bars indicate standard error.  B"A"C"Parental"Heterozygous"Clone"1052354556"Mutant"Clone"9514"10%"FBS"5%"FBS"1%"FBS"0.1%"FBS"81      The migration characteristics of the Hec1A PPP2R1A isogenic cell lines were also assayed using the scratch, or wound, assay. Over a 4-day period, the different cells lines clones (parental, mut/mut clones: 9-14, 9-15, 33-1, 33-4, mut/wt clones: 10-23-45-3, 10-23-45-6, 10-23-60-17, 10-23-60-24) were monitored to determine the rate of movement into the scratched area, however only one of each clone is shown (Figure 5.10 and 5.11). The results are consistent with the previously observed cell proliferation rates, where the parental and the heterozygous cells grow and migrate into the scratch at similar rates. In addition, the mutant cells are significantly impeded from migrating into the scratch area (Figure 5.10 and 5.11). After 72 hours, the mutant cells are not able to fully close the wound, whereas both the parental and heterozygous cell lines are able to migrate to close the full scratch area. The mutant cells appear to spread into the scratch area by growing on top of the cells at the scratch boundary instead of moving into the unoccupied areas, which is how the parental and heterozygous cells seem to grow.                      82               Figure 5.10 Images of the scratch (wound) assay over time             The three rows show time-lapse images of the migration rates of the different isogenic cell lines.             The yellow lines indicate the starting scratched area boundary (wound outline). As the cells migrate into the cleared wound area, the blue line indicates the new boundary of the cells. For the parental and heterozygous cell lines, the blue line disappears after 48 hours indicating the wound area has been closed from migrating cells. For the mutant cell line, the blue lines indicate that the cells were not able to move into the scratch area to close the wound after 72 hours. The changes in the density of these areas are shown in Figure 5.11.   83            Figure 5.11 Relative wound density over time of Hec1A PPP2R1A isogenic cell lines         A graphical figure showing the differences in migration characteristics of the isogenic cell          lines (n=6) using a scratch or wound assay. The relative wound density is plotted over time to          show the rate of cell movement into the scratch area. Error bars indicate standard error.  5.4 Discussion      Overexpression cell line models and gene knockout by shRNA (small hairpin interfering RNA) experiments, in many different cell lines, has been previously performed by many groups [90, 110, 131]. The use of shRNA to knockdown PPP2R1A was not used in my PhD project experiments, as it has been previously shown that PPP2R1A knockout to less than 30-37% of endogenous levels leads to cellular apoptosis and death [131, 132]. Instead, I utilized the endometrial Hec1A cell line with an endogenously expressing PPP2R1A W257L mutation and pAAV somatic cell KO methodology, to establish the first context-specific disease cell line models. This model will be used to determine the effect of endogenous PPP2R1A mutations on PP2A formation and activity in endometrial carcinoma. The establishment of these endometrial Hec1A PPP2R1A isogenic cell lines was challenging. I was able to isolate a mutant only expressing PPP2R1A (mut/mut) L257 cell line, however the unanticipated presence of an extra PPP2R1A mutant allele in the Hec1A genome, made isolating an wild-type only expressing cell line unsuccessful. Therefore, the final Hec1A trio isogenic cell lines consists of the parental PPP2R1A (mut/mut/wt) W257L cells, mutant only PPP2R1A (mut/mut) L257 cells, and heterozygous PPP2R1A (mut/wt) cells.   Mutant&(mut/mut)&Clone&9015&Parental&Heterozygous&(mut/wt)&Clone&1002304506&84      After isolating the isogenic cell lines, I next determined if the PPP2R1A mutation had any effect on the growth characteristics of the cells. I therefore, utilized various methods of determining cell proliferation and cell migration. The results of the different cell proliferation assays showed consistently that the different PPP2R1A (mut/mut) cell lines clones demonstrated an in vitro decrease in cell proliferation over time, compared to both the parental  (mut/mut/wt) and heterozygous (mut/wt) cell line clones. Both the parental and heterozygous lines exhibited similar cell proliferation characteristics. In addition, there were large differences in the migration characteristics of the mutant (mut/mut) cells compared to the parental or heterozygous cell lines. The time-lapse scratch assay dramatically shows the extent of migration inhibition in the mutant expressing cell lines. Taken together, these results demonstrate that the one allele expressing PPP2R1A L257 mutation has inhibitory effects on cell growth and migration. However, both the W257L mutated cell lines (parental and heterozygous) proliferate quickly and have an ability to migrate. This is consistent with patient endometrial tumours, as somatic PPP2R1A mutations are usually found as heterozygous missense mutations, and rarely found as homozygous mutations [1, 206].       Each of the isogenic mutant and heterozygous and cell lines were single cell cloned after the initial PPP2R1A pSEPT targeting vector transduction. Once the mutant and heterozygous lines were identified an additional Cre-recombinase step was utilized to excise out the SEPT cassette which resulting clones were also single cell cloned. Therefore, each mutant and heterozygous cell clone could potentially have different genetic backgrounds. The multiple single cell cloning is one explanation for the differences in cell proliferation for each mutant and heterozygous cell line clones. It was important to assay multiple clones for each genotype as biological replicates, however each of the different clones had a similar trend of proliferation. A second explanation for the differences in cell proliferation is that the Hec1A is an MMR deficient cell with a hyper-mutator phenotype. It is entirely possible that there is heterogeneity within the original Hec1A cell line; therefore each single cell clone may have a slightly different mutational background. Furthermore, over time the cell lines may acquire additional mutations contributing to the different isogenic backgrounds.  85      PPP2R1A has been described as a haploinsufficient and a dominant negative tumour suppressor [105, 131, 160]. The functional level of PPP2R1A (PP2A Aα) has been described to be an important modulator of cell transformation [133]. The haploinsufficiency model, and the levels of PPP2R1A to sustain cell growth are consistent with the growth characteristics of the isogenic cell lines. The balance of PP2A levels are extremely important for keeping the cells in a transformed or normal state. The Hec1A (mut/wt) cells with one mutant PPP2R1A allele and one wild-type PPP2R1A are viable due to the typical haploinsufficient phenotype, and proliferate similar to the parental PPP2R1A (mut/mut/wt) W257L cells (Figure 5.8). The Hec1A (mut/mut) mutant only expressing cell line is likely viable because it provides enough semi-functional PP2A to allow the cells to keep proliferating, however at a slower rate. The expression of the mutant PPP2R1A has likely changed the ability of the PP2A core complex to interact with B subunits, and also alters B subunit substrate interactions, resulting in changes in functional PP2A to keep the cells in a transformed state. This is supported by established literature wherein SV40ST can bind to PPP2R1A, thus inactivating PP2A by blocking B subunit binding, to induce cellular transformation [107]. Transformation can also occur when specific B subunits are knocked down [110], which may mimic PPP2R1A mutations that disrupt binding of the B subunits to form functional PP2A phosphatases. Interestingly, the knockdown of PP2A-B56γ (PPP2R5C), in conjunction with sh-p53 and mutant CDK4 was used to transform human fallopian tube secretory epithelial cells (FTSEC) that mimic ovarian high-grade serous carcinoma [233]. This is an important observation, as high-grade serous of the ovary and endometrium are histologically similar, however differ in some landscapes of somatic genotypes. PPP2R1A mutations are an example of one of these differences, however it is possible that PP2A is being alternatively disrupted in the ovary compared to the endometrium.       There are a few hypotheses that may explain why the isolation of a wild-type only expressing cell line was unsuccessful. 1) This may be due to unspecific targeting of the pSEPT vector into non-PPP2R1A sites, which would allow for selection of drug resistant clones, but would not result in the knockout of PPP2R1A exon 3. 2) The haploinsufficiency model, and the ability of PP2A to cause cellular transformation may also explain the lack of isolation of the wild-type only cell line. It is possible that the wild-type only expressing PPP2R1A cell clones are not able to survive and thus die during selection. This may be a phenomenon of the balance of a PPP2R1A 86 mutation providing a survival advantage, to keep the cells growing in a transformed state in culture. If this transformed state is disrupted, the cells can no longer survive. A previous study has shown that the overexpression of one allele of wild-type functional PPP2R1A into cells that lack less than 50% of endogenous PPP2R1A, inhibited cell proliferation and reversed the ability of cells to form tumours [131]. In the case of the potential Hec1A wild-type only expressing cell line, it would express only one copy of the wild-type allele, which may be at a threshold of functional PP2A needed to overcome the transformed state and revive the tumour suppressor function, thus causing cell death. This may also be supported by a study assessing mutational inactivation of the BAX frameshift mutation in the MMR-deficient human colon cell line HCT-116 [234]. Ionov et al., found that single cell cloning of HCT-116 cells isolated two populations of cells: heterozygous mutated BAX (+/-), and homozygous mutated BAX (-/-). These populations of cells were inoculated in nude mice and the resulting tumours were tested to find the BAX (+/-) population composed of subclonal populations of BAX (-/-) and (+/-), but never BAX (+/+) cells. The BAX (-/-) population never produced tumours with wild-type BAX expression. The authors show that wild-type BAX alleles can acquire a mutation, however the reversion of a mutation to wild-type was not observed. These results support evidence that BAX inactivation provides a survival advantage for tumour progression. There is also evidence to support this theory in the studies of the re-expression of wild-type p53. In this well-studied tumour suppressor, which is mutated in many different cancers including endometrial and ovarian serous carcinoma, studies have shown that the reversion of mutated p53 or the introduction of wild-type p53 can cause tumour cell death [235].       The production of these Hec1A PPP2R1A isogenic cell lines using the rAAV method was an intensive, time-consuming procedure. While the rAAV somatic cell knockout eliminates the limitations of an overexpression system, there are certainly limitations to this method. Of note, the neomycin (neo) resistance coding sequence found in the targeting vector, can have unintended consequences on the target gene and surrounding genes [236, 237]. The neomycin resistance gene has cryptic splice acceptors, which could be used by target genes, therefore it is recommended to remove the neo gene by Cre-recombination if long-term culture is required. Geneticin (G418) selection of the single cell clones containing the neomycin targeting cassette takes a few weeks to a month to enable clones to grow to a significant size for long-term storage, 87 passage and further cre-exision. It is indeed possible that these aberrant effects of the neomycin gene described, could have enough time to cause unintentional consequences.       To overcome some of these limitations, in recent years there have been advances in other genome-editing technologies to allow the generation of endogenous gene targeting systems in mammalian cells. The zinc finger nucleases (ZFN) [238, 239], TALENs (transcription activator-like effector nucleases) [240] and most recently the CRISPR (clustered regularly interspaced short palindromic repeats) [241, 242] systems have enabled increased success for creating somatic in vitro cell knockout and knockin models. The ZFNs and TALENs work by inducing loci-specific double strand breaks that are repaired by error-prone non-homologous end joining (NHEJ) and homology-directed repair (HDR) [243, 244]. This produces small insertions and deletions at break spots, which disrupt the target gene ultimately causing endogenous gene knockout. These genetic engineering technologies are groundbreaking and are replacing methods that rely on inefficient homologous recombination like the rAAV somatic cell knockout method. The ZFN and TALEN systems are however limited by their genetic delivery methods that rely on inefficient plasmid DNA, viral vectors or in vitro transcribed mRNA. Gaj et al, has shown that purified ZFN proteins can cross cell membranes and induce endogenous gene disruption [245]. This may overcome some of the limitations of gene-based delivery systems and reduce off target effects of viral and plasmid vectors.        CRISPR has emerged as a more efficient alternative to ZFNs and TALENs for genome editing, as it utilizes the bacterial adaptive immune system to perform RNA-guided DNA cleavage by Cas9 nucleases [241, 246]. This system has now been applied as a whole-genome scale knockout screening method in mammalian cells [247], and been successful in many studies and cell types [248, 249]. The field of CRISPR gene editing has exploded in a very short amount of time with a plethora of online resources and protocols for using the CRISPR technique [250], and is thought to replace ZFN and TALENs as the method of choice for genome editing [251]. For future studies, I would recommend utilizing the CRISPR system to generate improved and different isogenic cell lines to study the role of PPP2R1A mutations in gynaecological cancers. The endometrial cell lines that could not be utilized for the rAAV pSEPT somatic cell knockout, due to MSI stability, would be ideal candidates to introduce or knockout PPP2R1A mutations 88 into the endogenous genome using CRISPR. Our in-house collection of cell line candidates are the endometrial Hec50 cell line derived from the peritoneal fluid of a high-grade endometrioid carcinoma with a homozygous PPP2R1A mutation, and the KLE cell line derived from a poorly differentiated endometrial carcinoma (no histology defined) that does not harbor endogenous PPP2R1A mutations. The importance of using disease specific model cell lines is extremely important, therefore one could argue that the Hec1A, Hec50 or KLE cell lines are not perfect models for investigating the function of PPP2R1A mutations due to the lower frequency of PPP2R1A mutations in the cell line derived histological subtypes. Endometrial serous carcinoma harbors the highest frequency of PPP2R1A mutations as shown in Chapter 2 and 3, therefore an in vitro endometrial serous cell line would be the most ideal in vitro model. However, these cell lines are rare and our own attempts at making these cell lines from primary endometrial serous tumour tissue has failed. I have acquired only one endometrial serous carcinoma derived cell line SPAC-1-L from Japan [231], however my own sequencing of the PPP2R1A gene revealed no endogenous mutation. In the future, this cell line would be an ideal cell line for the introduction of a PPP2R1A mutation(s) to derive isogenic cell lines by CRISPR.        In conclusion, the establishment of the endometrial Hec1A trio isogenic PPP2R1A cell lines is an important contribution to the endometrial model resources. These model cell lines will be used to determine the how this PPP2R1A mutation affects the binding of PP2A B-subunits to form a functional PP2A holoenzyme. The majority of previous PP2A studies have been performed in non-relevant cell lines and disease models, therefore these cell lines will be one of the first to take into account PPP2R1A mutations in a context-specific disease model.  89 Chapter 6: Proteomics Analysis of PPP2R1A Mutations in Model Cell Lines  6.1 Introduction      PP2A and PPP2R1A have been extensively studied over many years, however the complexity of these proteins and involvement in many protein interactions makes this a very difficult protein complex to study. To determine interactions of each of the B subunits to the core enzyme (A and C subunits), in vitro overexpressing constructs with single mutations and or deletions in PPP2R1A have been utilized [90, 153]. Ruediger et al., was the first to determine how PPP2R1A mutations (E64D, E64G and R418W) identified in various cancers disrupt the formation of PP2A. The authors employed EE-tagged (EEEEYMPME) mutant PPP2R1A constructs with in vitro synthesized radiolabeled B subunits, supplemented unlabeled PPP2CA, and finally coupled immunoprecipitation to determine if a PP2A complex was formed [90]. The study concluded that E64 mutant constructs affected the binding of the B’α subunit (PPP2R5A) only. In an earlier study, the same authors found that mutating the amino acid position P179A caused disruption in the binding of the B55α (PPP2R2A) and B’α (PPP2R5A) B subunits [153], and hypothesized that this amino acid would likely be mutated in some cancers. It was not until my study of endometrial serous carcinomas, where I discovered that this specific amino acid change (P179R) can be found in nature and is a PPP2R1A cancer mutation hotspot [1]. In addition, a separate study has utilized transformed HEK (human embryonic kidney) cells expressing large T, hTERT and H-RAS (HEK TER) with mutant PPP2R1A constructs to determine the effects of the E64D, E64G and R418W mutations on B subunit binding [131]. Similar to past studies, these mutations were found to disrupt binding of B55α and some members of the B’ subunit to the Aα subunit [131].        Many of the first functional experiments that indicate PPP2R1A mutations disrupt binding of particular PP2A B-subunits were performed using IP-Western experiments [90, 131]. This method is limited by a need for multiple antibodies for each of the PP2A subunits (>15), and many of these subunits have many alternative transcript products. This results in many non-specific bands detected by a western blot, and is often a guess to determine which band is the specific band of interest. There is also high similarity between the B-subunit family members making it difficult for specific antibody detection and production. Previous studies have 90 generated their own antibodies [109, 110], which is time consuming and costly. Testing multiple antibodies for IP-western experiments is also costly, especially when investigating multi-protein complexes involved in many protein-protein interactions. To avoid this problem, researchers have overexpressed constructs with tags (FLAG, HA), which eliminates the need for multiple B subunit antibodies [114, 252, 253]. However, the process of quantifying differences on a western blot can be subjective and inaccurate, and has also been the subject of manuscript retractions due to manipulated western blot images [254, 255]. In my own experience (data not shown), the antibodies for the specific B subunits did not show consistent results, and often the bands detected on a western blot were not the same as the predicted molecular weight. This becomes problematic to try to detect small differences in the binding of B subunits to the PP2A core protein complex.       Proteomics experiments using mass spectrometry can avoid these limitations of antibodies and western blots [256], and have been used to determine the PP2A holoenzyme complexes and PP2A interaction networks [95, 257]. These studies are imperative for understanding the complexity of the PP2A protein network, however they are limited by cell context. The cell lines that were primarily used are easy to grow and easily transfected; HEK293 (transformed human embryonic kidney cells) and HeLa (cervical carcinoma) cells. Although these cell lines are useful for initial biochemical studies, the overexpressed mutated constructs transfected into these types of cells do not take into account the disease of which the mutations are found. PP2A subunit expression and biological role may be different in tissue and cell types; therefore this needs to be taken into account for disease-specific studies. It is also apparent from previous work, and Chapter 5, that the balance of PP2A levels are important.       Overall, studies have shown that a few PPP2R1A mutations (E64A, E64G, R418W, and 171-589 deletion) identified in one case of lung carcinoma, one case of melanoma and breast carcinoma, can disrupt the binding of specific B subunit proteins to form a functional PP2A phosphatase holoenzyme complex [90, 131]. However, there have been no studies to date on how the specific PPP2R1A mutations (P179R, R183W, R182W, S256F/W, W257G/L/C, R258C), discovered in endometrial and ovarian carcinomas [1, 2, 70, 206], can disrupt the binding of the B subunits in the context of gynaecological cancer. The last goal for my thesis research was to 91 utilize the endometrial carcinoma derived Hec1A isogenic PPP2R1A cell lines, as described in Chapter 5, to study the effect of the PPP2R1A mutation at endogenous levels and in a context-specific model. I hypothesized that the Hec1A endogenous PPP2R1A W257L mutation affects the interaction of PP2A B subunit family members; this was tested by isolating the PP2A complex using co-IP (co-immunoprecipitation) then coupled with proteomics technology and analysis. State of the art proteomics technology was utilized to allow insight into the relative differences in B subunit protein interactions in the Hec1A isogenic cell lines.   6.2 Methods  6.2.1 Co-Immunoprecipitation      For all co-immunoprecipitation (co-IP) and western blots, the rat monoclonal PPP2R1A 6F9 (Covance) was used. To covalently link the PPP2R1A antibody (Ab) to agarose beads, the Pierce co-IP kit was used as per manufacturers protocols (Thermo Scientific, Pierce, IL, USA). A control normal rat IgG antibody (Santa Cruz Biotechnology) was also covalently attached to agarose beads using the same protocol. In brief, 100ug of PPP2R1A Ab or normal rat IgG Ab was covalently bound to the AminoLink Plus Coupling Resin (aldehyde-activated beaded agarose). Hec1A cells from 6 X 15cm plates were cross-linked on plastic using 1% paraformaldehyde in serum-free media for 7 min, then quenched with 250nM glycine for at least one minute. The cells were washed twice with approximately 5-8mL cold PBS, then lysed and scraped off the tissue culture dishes using 10mL of cold Pierce IP lysis buffer with 1 tablet of cOmplete Protease Inhibitor Cocktail tablets (Roche). The cells were scraped on ice, then centrifuged at 16,000Xg for 15min to clear the lysate. The Pierce BCA assay was used as per manufacturer’s protocol to quantitate the total protein concentration. In total 20mg of cell lysate was used for each co-IP reaction. The bound Ab-bead mixture was added to the cell lysate and incubated at 4oC with rocking for about 2 hours. The protein complexes and beads were spun down at 1000Xg for 5min, then applied to a small spin column to allow washing and eluting. Once all the beads were applied to the column, the beads were washed 2X with cold PBS, then eluted with 60uL of Pierce Elution Buffer. The Elution buffer was allowed to incubate on the beads for 5min at room temperature, then eluted at 1000Xg for 1min. This elution was repeated for a total of 120uL of eluted IP proteins. The antibody-bead resin was then regenerated using 92 Pierce 1X Coupling Buffer, and added back to the cell lysate to repeat the co-IP. This co-IP was repeated for an additional two times, for a total of 3 sequential co-IP elutions. After all elutions were complete, the samples were pooled to a total of 320uL, and stored at -80oC.  6.2.2 Western Blots      To determine if the PPP2R1A co-IP was successful, 10uL of each co-IP reaction was subjected to analysis on a 10% SDS-PAGE gel. Proteins were transferred to a nitrocellulose membrane and then washed in TBST (0.1% Tween-20). The blots were blocked using 5% skim milk with TBST for either 1hr at room temp or overnight at 4oC. The primary PPP2R1A Ab (Covance) was incubated for 1-2hrs at room temp, or overnight at 4oC, using 1/4000 dilution in 5% skim milk + TBST solution. The secondary anti-rat HRP-labeled antibody was used at 1/20,000 dilution in 5% skim milk +TBST. The chemiluminescence kit (Millipore) was used for detection, and exposed at various times and developed using x-ray film.  6.2.3 Sample Digestion for Mass Spectrometry Analysis      Each of the IP samples was split into 3 aliquots. The first 100uL samples were prepared with no heating for de-crosslinking, the second 100uL aliquots were heated for de-crosslinking at 68oC overnight, and the last 300uL were de-fixed at 95oC for 10min. To then digest the IP samples, 100uL of the each pooled IP sample was used for reduction and alkylation. The samples were first reduced by using 5uL per 100uL volume of 200mM DTT in 50mM HEPES pH8.5, and then incubated at 45oC for 30min. This was followed by the addition of 10uL per 100uL volume of 400mM IAA in 50mM HEPES pH8.5 then incubated at 30min at 24oC in the dark. To quench the reaction 10uL per 100uL volume of 200mM DTT in 50mM HEPES pH8.5 was added and mixed. To proceed with protein cleanup and trypsin digestion, the paramagnetic SP3 beads were utilized as per protocols cited in Hughes et al. [258]. Samples were digested overnight with 80ng sequencing grade Trypsin (Promega) at 37oC with the presence of SP3 beads. At the peptide level, 4 TMT labels (TMT6-126, TMT6-127, TMT6-128, TMT6-129) (Thermo Scientific) were used to pool samples (Table 6.1), as per Hughes et al. [258]  and manufacturer’s protocol. In brief, 20ug of each TMT label was added to the peptide-bead mix and incubated at room temp for 30min. This step was repeated and incubated for a further 30min. The samples were quenched for 5min at room temp then washed with acetonitrile. To pool samples, 10uL of 2% DMSO was 93 used to resuspend one sample, then transferred to the next sample for incorporation into the pool, and repeated until all four samples were pooled in 10uL total. The pools (Table 6.1) were then acidified to final 0.1% formic acid, and run directly on the LC coupled to the mass spectrometer.   PPP2R1A-IP Pool TMT6-126 (mut/mut/wt) TMT6-127 (mut/mut) TMT6-128 (mut/wt) TMT6-129 (mut/mut/wt ctl) Pool 1 Parental P145 9-15 mutant 10-23-45-6 het Parental P145 IgG-IP Pool 2 Parental P146 9-14 mutant 10-23-60-17 het Parental P146 IgG-IP Pool 3 Parental P147 33-1 mutant 10-23-60-24 het Parental P147 IgG-IP Table 6.1 Hec1A isogenic cell line IP TMT pools for mass spectrometry analysis P145, P146, P147 indicates passage number. For the mutant and heterozygous cell, the number indicates the clone ID.  6.2.4 Mass Spectrometry Data Acquisition      The pools of IP TMT labeled samples were analyzed on a Fusion Orbitrap system (Thermo Scientific) coupled with an EASY-nLC 1000 (Thermo Scientific) liquid chromatography system. A total of 4uL for each pooled peptide sample was loaded onto the LC with 10uL/min flow into a reverse phase C18 column. The samples were eluted from the column with a gradient flow of 250nL/min, for a total of 120 min. The gradient parameters are as follows: 1min at 2% Buffer B, 91min at 22% Buffer B, 14min at 30% Buffer B, 14min 80% Buffer B (Buffer A: 0.5% formic acid in water, Buffer B 0.5% formic acid in acetonitrile). Data acquisition on the Fusion Orbitrap was programed using the Thermo Scientific software and workflow template. The first MS scan was used for precursor selection, then MS2 fragmentation in the CID (collisional induced fragmentation), with a cycle time of 3sec. At this point the top 10 fragments were selected for MS3 in the Orbitrap HCD, to allow fragmentation for TMT label quantitation. The survey scans were used at a mass range of 380-1200(m/z), with the Orbitrap resolution set at 120,000.  6.2.5 Mass Spectrometry Data Analysis      To perform peptide and protein identification, as well as TMT quantitation, the Thermo Proteome Discoverer 1.4.1.14 software (Thermo Scientific) was utilized. All MS runs for each pool were searched together using the protein databases Mascot and Sequest HT. The analysis workflow was set up to search and group peptides into proteins, which also assigned a FDR 94 (0.05). A medium peptide confidence filter was utilized, and the raw data was extracted for subsequent analysis in the stats package RStudio Version 0.98.1074. To first filter the raw peptide data in each individual pools, peptides were removed if duplicate peptides were present, and if missing values were present in any of the TMT126, TMT127 and TMT128 labeled samples. The resulting peptides from each of the 3 pools were then merged, to allow quantile normalization [259]. Normalization of the cell lines is shown before and after normalization (Figure 6.1). The Negative control Ig-G IP values were not used in the downstream analysis. Each of the pools was then collapsed into proteins (based on accession groups), which takes the mean of the peptides for each protein and collapses into the protein group. An accession group list of PP2A subunits and PP2A interacting proteins found in the raw data (PPP2R1A, PPP2CA, PPP2CB, PPP2R2A, PPP2R5A, PPP2R5C, PPP2R5D, PPP2R5E, PPP2R4, SET, IGBP1, PPME1) were used to extract proteins of interest for additional statistical analysis. The peptide list and intensities associated with these proteins can be found in Table F.9 (Appendix F).  Proteins were included in additional analysis only if they were identified in 2 of the 3 pools. This removed only one of the B subunits (PPP2R5A) from the analysis, as it was only detected in one of the 3 IP pools. To determine significant changes for these proteins of interest, the fold changes for each individual replicate were calculated. To calculate unadjusted p-values, the Limma R package [260] was used for each cell line comparison (n=3) within each protein group. The adjusted p-values were then calculated based on all the tests performed using the B-H method [194]. 95 Figure 6.1 Pre and post normalization of all peptides identified in the IP-MS pools. Histograms for pre and post normalization by quantile normalization analysis are shown. The histograms depict peptide density (y-axis) by the log of the peptide intensity (x-axis). The first row of panels indicate pre-normalization, the middle row panels show log peptide intensities with normalization as the histogram and the smoothed lines indicate pre-normalization levels for each of the 3 pools. Red lines indicate pool 1, green as pool 2, and blue as pool 3. The post normalization histogram is the finalized normalized data for the parental, mutant and heterozygous samples. Parental(Density(Pre(Normaliza1on(Pre/Post((Normaliza1on(Post((Normaliza1on(Density(Density(Log(Pep1de(Intensity)((Mutant( Heterozygous(Log(Pep1de(Intensity)((96 6.3 Results      To determine the changes in endogenous PPP2R1A interactions in the trio of Hec1A isogenic cell lines harboring the PPP2R1A mutation, immunoprecipitation were employed for analysis by western blot and mass spectrometry. A previous study had determined that the PPP2R1A 6F9 antibody enabled the isolation of the endogenous PP2A heterotrimeric complex using sequential immunoprecipitation [261]. Therefore, by performing sequential PPP2R1A co-IPs of the Hec1A parental and isogenic cell lines, I was able to isolate the endogenous PP2A complex in each IP (Figure 6.2).                          Figure 6.2 PPP2R1A-IP samples with immunoblot (IB) for PPP2R1A                       Sequential PPP2R1A-IPs from Hec1A lysates and PPP2R1A-IB shown by Western                       blot. L=total lysate, FT=Flow-Through, E= Elution. The IgG negative control shows                        that there was no PPP2R1A pull down through immunoprecipitation.                       The monomeric PPP2R1A protein or the core subunit bound to the catalytic subunit is very abundant in the cell, however, I found that the B subunits were in limiting amounts, and difficult to detect by IP-western analysis. I also found it difficult to detect B subunit protein interaction with PPP2R1A without chemical cross-linking. I therefore utilized chemical crosslinking to allow isolation of an intact PP2A holoenzyme, and sequential IPs from each cell line were needed in order to maximize the number of PPP2R1A protein molecules and PP2A complexes binding to the PPP2R1A-IP antibody. Sequential IP’s were performed by using the unbound fraction to carry into the next IP using the agarose-conjugated PPP2R1A antibody. Even after three sequential co-IPs, the levels of PPP2R1A were not 100% depleted from the cell lysate, as PPP2R1A&PPP2R1A&&&L&&&&&&&&FT1&&&&E1&&&&&FT2&&&&E2&&&&FT3&&&&&E3&&&&&&L&&&&&&FT1&&&&E1&&&&&FT2&&&&&E2&&&&&FT3&&&&&&E3&&&&L&&&&&&&&FT1&&&&E1&&&&&FT2&&&&&E2&&&&FT3&&&&&E3&&&&&&L&&&&&&FT1&&&&E1&&&&&FT2&&&&&E2&&&&&FT3&&&&&&E3&&HEC1A&P145&PPP2R1A0IP& HEC1A&P145&IgG0IP&HEC1A&9014&Mutant&PPP2R1A0IP& HEC1A&9015&Mutant&PPP2R1A0IP&97 PPP2R1A was still present in the flow-through after the third immunoprecipitation reaction (Figure 6.2). However, there was sufficient PPP2R1A and interacting proteins in the elutions to perform mass spectrometry.            To acquire a complete analysis of the possible PPP2R1A-IP interactions, I used the sensitive Fusion Orbitrap mass spectrometer to analyze the three PPP2R1A-IP pools of isogenic cell lines containing triplicate biological replicates (three different mutant clones, heterozygous clones and different passages of the parental cell line). For the purposes of this study, I was interested in only the PP2A subunits and known interactors with PP2A, therefore the analysis was directed at only these specific proteins. In all of the samples, I was able to detect both C subunits (PPP2CA, PPP2CB), five of the B subunit family members (PPP2R2A, PPP2R5A, PPP2R5C, PPP2R5D, PPP2R5E), and four additional known PP2A interactors (PPP2R4, PPME1, IGBP1 (alpha4), and SET) (Table 6.2). Peptides from PPP2R1B, along with the other B55, B56’ and B’’ subunit family members (including Striatin (STRN)), and an additional PP2A inhibitor CIP2A were not detected in any of the IP samples.       By comparing the IP pools of Hec1A PPP2R1A mutant (mut/mut) clones with Hec1A PPP2R1A heterozygous (mut/wt) cell lines clones, there were significant differential interactions of the B subunits PPP2R5C, PPP2R5D and the PP2A inhibitor SET (p-value 0.035, 0.043, 0.043 respectively) (Table 6.2, Figure 6.3). PPP2R5C and PPP2R5D peptides were found at lower intensities in the mutant cells compared to the heterozygous cell lines. Conversely, the endogenous PP2A inhibitor, SET, was found at higher intensities in the mutant cell lines. This implies that there was an increased interaction of SET with PPP2R1A in the mutant cells compared to the heterozygous cells. There was a trend for decreased interaction of PPP2R2A and PPP2R5E with mutant PPP2R1A, however this was not significant after multiple p-value corrections (Table 6.2).        The comparison of the parental (mut/mut/wt) with the mutant cell line clones also shows significant differential interaction with PPP2R5C and PPP2R5D (Figure 6.3).  However, PPP2R1A levels were also found to be significantly different in the parental cells compared to the mutant and the heterozygous cell lines. This may be explained by the difference in the 98 number of PPP2R1A expressing alleles in the isogenic cell lines; the parental cell line harbors three expressed PPP2R1A alleles compared to only two alleles in the mutant and heterozygous cell lines. Therefore, assessing the differences in interactions between the mutant cells and the heterozygous cells was a more suitable comparison to account for the number of alleles expressing PPP2R1A. There were no other significant differences between the PPP2R1A-IP samples from the parental and heterozygous cell lines (Figure 6.3). There were also no significant effects of the PPP2R1A mutation on the catalytic C subunit  (PPP2CA or PPP2CB) binding in any of the cell lines. In addition, the PPP2R1A mutation did not seem to show an interaction difference with the PP2A interacting proteins PPP2R4, IGBP1 or PPME1.                 99   Figure 6.3 Quantitative analysis of PPP2R1A-IP TMT labeled pools using mass spectrometry. Each panel shows the comparison between the isogenic cell lines. Box plots reveal the variation of the detection of each protein in the three IP pools. Black stars indicate significant adjusted p-values.    PPP2R1A&PPP2CA&PPP2CB&PPP2R2A&PPP2R5C&PPP2R5D&PPP2R5E&PPP2R4&SET&IGBP1&PPME1&PPP2R1A&PPP2CA&PPP2CB&PPP2R2A&PPP2R5C&PPP2R5D&PPP2R5E&PPP2R4&SET&IGBP1&PPME1&PPP2R1A&PPP2CA&PPP2CB&PPP2R2A&PPP2R5C&PPP2R5D&PPP2R5E&PPP2R4&SET&IGBP1&PPME1&Log2(Fold&Change)&A.&Mutant&vs.&Heterozygous& B.&Parent&vs.&Heterozygous& C.&Parent&vs.&Mutant&100 Protein Name PPP2R1A PPP2CA PPP2CB PPP2R2A PPP2R5C PPP2R5D PPP2R5E PPP2R4 SET IGBP1 PPME1 Protein Groups P30153 P67775 P62714 E5RFR9 H0YJU0 Q14738 Q16537 A6PVN9 Q01105 P78318 Q9Y570 Parental-t126a 433000 384000 169000 2980 12300 46300 32600 20200 173000 16100 28000 Parental-t126b 327000 456000 167000 NA 15300 100000 6710 13200 94000 9760 24500 Parental-t126c 262000 352000 106000 16100 5950 48000 17220 NA 50100 9190 21600 Mutant-t127a 76400 179000 55100 2090 1090 8510 6860 22800 156000 9420 21200 Mutant-t127b 116000 212000 68000 NA 2250 32600 5400 11400 142000 8270 27400 Mutant-t127c 170000 493000 107000 12400 3770 23500 9250 NA 217000 5550 21200 Het-t128a 139000 468000 144000 14800 7540 57300 50800 25900 54500 20500 47300 Het-t128b 82900 297000 99600 NA 5960 61200 9980 15000 63300 12400 33600 Het-t128c 112000 387000 99100 19300 7220 52300 11900 NA 88200 9200 30400 MutVHet a -0.86 -1.39 -1.39 -2.82 -2.79 -2.75 -2.89 -0.18 1.52 -1.12 -1.16 MutVHet b 0.49 -0.49 -0.55 NA -1.40 -0.91 -0.89 -0.39 1.17 -0.59 -0.29 MutVHet c 0.61 0.35 0.11 -0.64 -0.94 -1.15 -0.36 NA 1.30 -0.73 -0.52 ParVHet a 1.64 -0.28 0.22 -2.31 0.71 -0.31 -0.64 -0.36 1.66 -0.35 -0.76 ParVHet b 1.98 0.62 0.74 NA 1.36 0.71 -0.57 -0.19 0.57 -0.35 -0.46 ParVHet c 1.23 -0.14 0.09 -0.26 -0.28 -0.12 0.54 NA -0.82 0.00 -0.49 ParVMut a 2.50 1.10 1.61 0.51 3.49 2.44 2.25 -0.17 0.14 0.77 0.40 ParVMut b 1.49 1.11 1.30 NA 2.77 1.62 0.31 0.21 -0.60 0.24 -0.17 ParVMut c 0.62 -0.49 -0.01 0.38 0.66 1.03 0.90 NA -2.12 0.73 0.03 MutVHet pvalue 0.87 0.30 0.21 0.03 0.01 0.01 0.03 0.60 0.01 0.08 0.15 ParVHet pvalue 0.00 0.84 0.28 0.05 0.15 0.79 0.53 0.50 0.33 0.46 0.09 ParVMut pvalue 0.00 0.23 0.05 0.44 0.0001 0.001 0.022 0.98 0.08 0.22 0.85 MutVHet pAdj 0.90 0.45 0.38 0.09 0.04 0.04 0.09 0.71 0.04 0.19 0.30 ParVHet pAdj 0.01 0.90 0.44 0.14 0.30 0.89 0.65 0.63 0.48 0.60 0.19 ParVMut pAdj 0.03 0.38 0.14 0.60 0.003 0.02 0.09 0.98 0.19 0.38 0.90 Table 6.2 Mass spectrometry data at the protein level after normalization The TMT values (126-128) in the first nine rows show the normalized protein intensities for each pool. The comparison rows (MutVHet, ParVHet, ParVMut) show the ratio of the log intensities for each protein. The unadjusted and adjusted p-value is also shown for each comparison. 101  6.4 Discussion      This is the first study to determine the effects of an endometrial endogenous PPP2R1A mutation on the interactions with PP2A B subunits. In an attempt to perform protein interaction experiments without the use of multiple antibodies, I utilized sequential co-immunoprecipitation of endogenous PPP2R1A in isogenic Hec1A cell lines, coupled with proteomic mass spectrometry analysis. As shown in Chapter 5, each of the mutant and heterozygous cell line clones had slightly different cell proliferation profiles, however all proliferate similarly within the same isogenic background. For example, the mutant clonal cell lines all proliferate slower than the parental, and the heterozygous cell clones proliferate similarly to the parental cell line. To potentially mitigate these slight variations of one particular clone, three different mutant and heterozygous clones were utilized for the IPs and combined for the overall peptide data analysis. By comparing the different PPP2R1A isogenic cell lines, I was able to show that there are significant differences in some PP2A B’ subunit interactions with mutant PPP2R1A.        Of particular interest, the Hec1A PPP2R1A mutant (mut/mut) cell clones demonstrate a significant decrease in interaction with the B56 family members PPP2R5C and PPP2R5D when compared to the heterozygous and parental cell lines. Due to the unbalanced PPP2R1A allele levels of the parental (mut/mut/wt) compared to the mutant (mut/mut) and heterozygous (mut/wt) cell lines, there was a significant difference in PPP2R1A levels detected by mass spectrometry analysis. However, it was acceptable to compare the pooled mutant and heterozygous cell lines clones (both express two PPP2R1A alleles) to determine if the mutation had an affect on B subunit binding. The intensity of the PPP2R1A peptides detected in the mutant and heterozygous cell line IP samples were not significantly different. The IP-MS analysis demonstrated the novel discovery that the PPP2R1A L257 mutation affected the interaction of both PPP2R5C and PPP2R5D. These two proteins have been previously reported as being important regulators of tumourigenesis and cellular transformation.       The crystal structure studies of PP2A revealed that the specific B’γ subunit (PPP2R5C) interacts with the A subunit (PPP2R1A) via HEAT domains. These interactions are over a large surface area, however the overall interaction is low involving only a few key amino acid 102 residues. Xu et al., showed that a small interaction area involves the B subunit HEAT repeat 2 with the A subunit HEAT repeat 7 and 8, where W257 directly hydrogen bonds with the L107 residue on the B subunit [86, 94]. In addition, amino acid residue L107 and the surrounding residues are highly conserved amino acids in all family members of the B’ family [86]. This provides further evidence that the W257 mutation found in PPP2R1A can cause disruption of the hydrogen bond that links the whole B’ family to the A subunit of PP2A. This structure interaction also provides additional validity to my observation of decreased protein interactions involving PPP2R5C and PPP2R5D with PPP2R1A that harbors the W257L mutation.       As previously described, the suppression of PPP2R5C has induced cellular transformation [109, 133]; therefore it is not surprising to find that this particular Hec1A PPP2R1A mutation (W257L) can affect the binding of PPP2R5C to potentially induce endometrial transformation. The knockdown of PPP2R5C, to replace SV40ST, has exhibited an ability to fully transform cell lines that express SV40LT, hTERT, and Ras-V12 [262]. The capability of PPP2R5C knockout to transform cells, explains the use of this knockout construct for the construction of the human fallopian tube secretory epithelial cell model in mice [233], also giving evidence to show this protein is extremely important in tumourigenesis. My observed results were also consistent with the report by Chen et al., where the authors show that PPP2R5C (B56γ) interaction is lost when a PPP2R1A E64D/G mutation is overexpressed. Loss of PPP2R5C interaction with PPP2R1A was also observed when PPP2R1A was downregulated with the use of shRNA [131]. PPP2R5C and also all B56 family members have been reported to directly inhibit the formation of the APC-Axin complex, which leads to the destabilization of β-catenin [263]. Moreover, functional PP2A complexes containing PPP2R5C can dephosphorylate p53 at Thr55, thus preventing its degradation and inhibiting cell proliferation [264]. In the tumour cell, if PP2A-PPP2R5C complexes cannot effectively dephosphorylate p53, hyperphosphorylation of p53 occurs leading to it’s degradation by the proteasome and thus kept at low levels to allow cell proliferation and tumourigenesis. In the context of endometrial serous carcinomas TP53 is almost ubiquitously mutated [42], however it is possible that PPP2R1A mutations that cause disruption of PPP2R5C interaction may provide an alternative “safe keeping” suppression of p53 by lack of Thr55 dephosphorylation. One PPP2R5C mutation (F395C) described in lung cancer has been previously found to disrupt the ability of PP2A to dephosphorylate Thr55 on p53 [265]. In 103 Chapter 3, I also sequenced PPP2R5C in a large cohort of endometrial carcinomas, and found this gene to be rarely mutated (n=3), therefore it is unlikely that PPP2R5C DNA mutations are playing a large role in tumourigenesis. It is however possible that perturbed binding of PPP2R5C to PPP2R1A or alternative post-translational modifications causes disruption of PP2A-PPP2R5C specific substrate dephosphorylation to promote endometrial tumourigenesis.       My discovery of the PPP2R1A Hec1A mutation disrupting interaction with PPP2R5D is interesting in the context of endometrial carcinoma. Firulli et al., demonstrated that B56δ-containing PP2A complexes interacts with the bHLH factors HAND1 and HAND2 [266] to reduce levels of HAND1 phosphorylation. The HAND transcription factors are required for heart, vascular, embryonic and placental development [267], and phosphorylation is required for nucleolar release, dimerization and biological function [268]. A recent publication presented evidence associating hypermethylation of HAND2 in >90% of endometrial endometrioid carcinomas, which is also correlated with the downregulation of mRNA expression [269]. The PPP2R5D gene was also noted to be hypermethylated in these endometrial carcinomas. HAND2 is expressed in the normal endometrial stroma, is regulated by progesterone and aids in progesterone suppression of estrogen induced pathways [270]. In mice that lack HAND2, epithelial proliferation is maintained by induction of fibroblast growth factors and stimulation of estrogen pathways [270]. This is fitting as the majority of endometrial endometrioid carcinomas are estrogen-dependent tumours, and the loss of HAND2 and PPP2R5D protein expression by gene hypermethylation could be an important contributor to endometrial carcinogenesis. Taken together, the deregulation of PPP2R5D in endometrial carcinoma seems to be an important mechanism of tumourigenesis and warrants further investigation. It would be interesting to determine if PPP2R1A mutations and hypermethylation of PPP2R5D and HAND2 are leading to the same pathways for promotion of tumourigenesis and cell proliferation. In addition, PP2A-PPP2R5D complexes have been found to be an important regulator of CDC25C in mitotic exit [97]. If PPP2R5D cannot dephosphorylate CDC25C in mitosis, this will lead to prolonged activation, and subsequent activation of CDK1 to delay exit from mitosis [97].       The mutated PPP2R1A cell lines also demonstrated a statistically significant increased interaction with the PP2A inhibitor SET (I2PP2A). At this time, it is unclear if SET can interact 104 more efficiently with the mutant-containing PPP2R1A-PP2A holoenzymes, or that the disruption of B subunit interactions by the PPP2R1A mutation has caused an unbalance of available free PP2A, which can then bind to SET.  Earlier studies have shown that SET/I2PP2A likely interacts with the PP2A catalytic subunit [127, 271].  With respect to the structure of PP2A, if mutated PPP2R1A causes loss of hydrogen bonds that act to facilitate A and B subunit interactions, it is possible that the pool of A-C dimers (without B subunit interaction) may be more available to interact with SET through the catalytic domain. In other words, the SET protein may be able to mimic the B subunit and interact with the A-C core PP2A protein. It is also possible that mutant PPP2R1A could cause “loose” interactions of B subunits resulting in a conformational change in PP2A, which could then allow for efficient SET interaction with the catalytic subunit. These theories are purely speculative, as these types of structure studies need to be performed in the future to determine how mutant PPP2R1A affects SET interaction.       The SET oncoprotein has been identified as overexpressed in many cancer types including B-cell chronic lymphocytic leukemia (CLL), non-Hodgkin lymphoma [272], and breast cancers [273]. In addition SET overexpression was correlated with decreased PP2A activity in chronic myelogenous leukemia (CML) cells [252]. In CML, BCR/ABL induces expression of SET to inactivate the dephosphorylation activity of PP2A, however re-activation of PP2A causes growth suppression and apoptosis. Interestingly, the inactivation of SET or CIP2A has been the subject of a few recent studies, these show that a SET inhibitor (OP449) can significantly reduce the tumourigenic potential in pancreatic cell lines and breast cancer cell lines [128, 129]. Furthermore, ceramides have exhibited antiproliferative properties, with an ability to increase PP2A activity by directly binding to SET/I2PP2A in lung cancer models [274, 275]. An FDA approved spingolipid analogue drug, FTY720, has recently been shown to directly target SET to activate PP2A, causing cell death and lung tumour suppression [268]. The importance of SET in gynaecological cancers is yet to be determined, although in light of these results this warrants further research and drug testing in endometrial cancer models.       It was interesting to note that the PPP2R1A mutation does not cause complete loss of binding of any of the observed B subunits. This may be due to the importance of the fine balance of PP2A activity for the cell’s growth, survival and death. If there is complete loss of PP2A activity 105 this will lead to cell death, therefore for the tumour cell to survive, it must not eliminate all functional PP2A activity. As described previously, mutations in PPP2R1A seem to function by disrupting B subunit binding which likely partakes in cellular transformation and tumourigenesis. From these results, I propose that the W257L PPP2R1A mutation is affecting the PP2A holoenzyme in three ways: 1) to disrupt efficient binding of particular B subunits, 2) disrupting B subunit substrate interaction for effective phosphatase activity 3) allowing the PP2A inhibitor SET to interact with PP2A (containing mutated PPP2R1A) to thus inhibit PP2A activity.         The use of mass spectrometry analysis enabled the simultaneous identification of a number of protein interactions in the PPP2R1A-IP cell line lysates. Currently, these interaction results have not been validated, and traditionally the “gold standard” would be to perform a number of western blots. As stated previously, western blots are not quantitative, and are dependent on reliable antibodies for each protein of interest. Aebersold et al., recently advocated that targeted proteomics techniques should be used to perform quantitative protein analysis and for validating mass spectrometry results, instead of the western blot [256]. These techniques, such as SRM/MRM (selected reaction monitoring/multiple reaction monitoring) or PRM (parallel reaction monitoring) [276], are extremely accurate as they provide quantitative information from multiple peptides per protein, multiple transitions per peptide with multiple measures of each signal. This level of analysis would be equivalent to performing hundreds to thousands of western blots to potentially validate the interactions of PP2A. For future studies, I will likely utilize the PRM methodology to validate the results from this study.       In conclusion, the immunoprecipitation of endogenously mutated PPP2R1A and subsequent analysis by sensitive mass spectrometry identified many subtle differences in the interaction of regulatory B subunits to PPP2R1A. I have presented a novel discovery wherein the endogenous PP2A inhibitor, SET oncoprotein, can interact with mutated PPP2R1A. Additional studies will be needed to determine if this increased interaction contributes to the suppression of PP2A activity. Overall, these findings suggest that the PPP2R1A mutation is likely playing a large role in cellular tumourigenesis by disrupting the balance of B subunit interactions. It will be interesting to study additional cell lines and patient samples for changes in PP2A subunits to determine if all hotspot PPP2R1A mutations act similarly. Future studies should also include 106 assessing the changes in B subunit substrate dephosphorylation events. Finally, targeting PP2A for targeted therapeutics may be an efficient way to inhibit tumourigenesis in gyneacological carcinomas. 107 Chapter 7: Conclusion  7.1 Overall Significance of My Thesis Research  7.1.1 PPP2R1A Mutations in Endometrial Carcinomas      My publication on the discovery of PPP2R1A mutations in endometrial carcinomas, described in Chapter 2, has lead to a number of other researchers validating and publishing their sequencing results of PPP2R1A mutations. The validation of these results by separate international groups is of immense importance that reflects the validity of the finding. It was especially important to validate the hypothesis and discovery of PPP2R1A mutations specific to endometrial serous carcinoma, and not ovarian serous carcinomas [206]. As this suggests that these cancers are molecularly distinct and thus should not be bundled as a single entity in clinical trials. Multiple groups including the TCGA consortium have validated this observation, there are, of course exceptions and PPP2R1A mutations have now been described in ovarian serous carcinoma, albeit at very low frequencies (1%) [33]. PPP2R1A mutations are not only found in endometrial serous carcinomas, as I also identified varying mutation frequencies in endometrial and ovarian endometrioid carcinomas. These two subtypes are thought to originate from the same endometrial epithelial precursor cell, however it is apparent from their differences in the frequency of PPP2R1A mutations that they evolve by different tumourigenic pathways.   7.1.2 Mutational Profiling of Endometrial Carcinomas     My work presented in Chapter 3, identified distinct molecular profiles that may aid in improving endometrial carcinoma classification leading to increased reproducibility of diagnoses. This publication was an important stepping-stone for the TCGA endometrial carcinoma study. The TCGA study used whole exome sequencing and other genomic technologies; therefore they were able to perform a more comprehensive analysis for classifying the genetic landscape of these tumours. Although my study did not have larger genomic view of the tumour, I was still able to demonstrate that there was a molecular difference between the histologically defined endometrioid and serous cases. In addition, I was able to pinpoint the misclassified cases using mutation profiles. This is important, as the subset of cases that can appear as intermediate subtypes (Figure 3.4), mixed endometrioid and serous, are in particular 108 need of subclassification, and mutational profiling seems to distinguish these challenging cases. In addition, the TCGA study was also able to show misclassified cases based on their genomic data [42]. Although endometrial carcinoma subtype diagnoses and grade are currently used in guiding patient management, mutational analysis is emerging as a realistic option in clinical practice. In the future, I predict that the mutational classification of endometrial carcinomas will become an important tool in diagnosis, and guiding mutation-based targeted treatment decisions. Mutation profiles are already being applied in other cancers for selecting targeted therapeutics, for example BRAF inhibitors in malignant melanoma [277] and BRAF and EGFR targeting in colorectal cancers [278, 279]. Determination of the role of mutational analysis in assessment of endometrial carcinomas will require additional study, with careful comparison of molecular versus conventional subclassification.       In Chapter 3, I also proposed that the traditional dualistic model of endometrial cancer that divides this cancer into type I and type II should be avoided, as it is too simple and does not capture the important distinctions between the subtypes of endometrial cancer. The very first small epidemiological study of endometrial cancers did show that risk factors were different between endometrioid and serous histologies [280], however a recent large case control study indicates that the risk factors may be similar for both types [281]. This large study also supports the notion that the type I and type II terminology should not be used, and should not be grouped separately based on histology alone, as there may be a common etiologic pathway. A recent review by Murali et al., also recommended the elimination of the dualistic type I and type II model as it does not encompass the heterogeneity of the endometrial tumours [282]. In the future, this terminology will be filtered out of mainstream publications due to the recommendations from these new studies.  7.1.3 PPP2R1A Isogenic Cell Line Models      The generation of the PPP2R1A isogenic endometrial cell lines, described in Chapter 5, provides an important model for the PP2A research community. This is the first endogenously expressing PPP2R1A mutant only cell line in a relevant disease specific model. These cell lines could be used in future studies of drug screens to assess the ability of drugs to selectively target tumours with PPP2R1A aberrations. To determine the downstream effects of the PPP2R1A 109 mutation and disruption of PP2A B subunit binding, it will be interesting to use these cell lines in additional assays, for example, phospho-proteomic, global proteomics and cell cycle regulation and mitotic checkpoint assays. This may aid in the discovery of novel targetable proteins or pathways that could be used for endometrial-specific therapeutics and diagnostic tools. These cell lines may also provide a model for the assessment of drugs that target PPP2R1A and/or the SET protein. Overall, my research has generated the first set of isogenic PPP2R1A endometrial specific cell lines for one particular mutation, therefore it will be important to generate additional cell lines to determine the effect of the other hotspot mutations identified in endometrial and ovarian cancers.  7.1.4 Impact of PPP2R1A Mutations on PP2A Holoenzyme Composition      The work presented in Chapter 6 is the first study to perform the assessment of an endogenous PPP2R1A mutation in a disease-specific cell line model system. I was able to show using sensitive mass spectrometry analysis, that the PPP2R1A W257L mutation causes changes in the interactions of PP2A B subunits PPP2R5C and PPP2R5D. There were also subtle differences in the interaction of PPP2R2A and PPP2R5E although these did not reach statistical significance. Many additional studies will be needed to determine the effects of the interaction disruption, however previous studies have shown that these particular interactions are crucial for cellular transformation. The experiments I performed in Chapter 6 only indicate changes in B subunit interactions with the PPP2R1A mutant. At this point in time I can only speculate on how these changes can affect tumourigenesis, although, literature suggests that both PPP2R5C and PPP2R5D play important roles in cancer [82, 131]. In 2011, Ruediger et al., performed an elegant knockout and knockin mouse model study [283] to determine the effects of a low frequency PPP2R1A E64G/D mutation previously found in one breast and one lung carcinoma [130]. The authors show that mutant heterozygous mice have an increased incidence of lung tumours, and there was a partial reduction of B’ holoenzymes. These results suggest that the reduction of B’ holoenzymes leads to increased cancer incidence, and PP2A acts as haploinsufficient tumour suppressor in this mouse model. Surprisingly, the mouse model with both altered PPP2R1A alleles (∆5-6/E64D) was viable, even with no expression of wild-type PPP2R1A, and gave rise to mice presenting with multiple tumours [283]. This also provides 110 support for the viability of the mutant only expressing PPP2R1A L257 Hec1A cell lines I generated for this thesis research presented in Chapter 5 and 6.       It is interesting to note that in a normal cell PPP2R5C-induces dephosphorylation of p53 Thr55, and this is important to inhibit p53 degradation and promotes its tumour suppressor function. TP53 is considered a gatekeeper of the cell to inhibit tumour formation, however if PPP2R5C is deregulated and cannot efficiently dephosphorylate p53, this may promote p53 degradation to promote tumourigenesis. There is evidence to suggest that the PP2A-PPP2R5C complex and TP53 could be working together to suppress tumour promotion, therefore when one protein or the other is de-regulated this could lead to tumour progression. In light of this, I also hypothesize that when both p53 and PPP2R1A are mutated, for example in the specific case of endometrial serous and some high-grade endometrioid carcinomas, the tumour cells have evolved an extra mechanism to suppress both PP2A and TP53 tumour suppressor function. This may contribute to the level of aggressiveness to aid tumour cell growth and survival.       There has been limited evidence to support the role of PPP2R5D in cancer, however recent evidence suggests that a PPP2R5D substrate, HAND2, is frequently hypermethylated in endometrial cancer. The study also found that PPP2R5D is frequently hypermethylated in these endometrial endometrioid tumours [269]. The downregulation of both PPP2R5D and HAND2 proteins by epigenetic regulation may lead to endometrial tumourigenesis. This study did not investigate the epigenetic regulation of the aggressive serous carcinomas, however, it is possible that PPP2R1A mutations causing disruption in the interaction of PPP2R5D may also function similarly to DNA hypermethylation to cause downregulation of gene expression.       The discovery of the PPP2R1A mutation allowing for increased interaction with the PP2A inhibitor SET is interesting and novel. This is the first report to my knowledge, which describes this phenomenon of the ability of SET to interact with mutant PPP2R1A. However, as mentioned in the discussion of Chapter 6, SET is reported to interact with the catalytic subunit and not the A regulatory subunit [127, 271]. This could indicate that SET interacts with the A-C core protein complex, without B subunits present due to the inability to interact efficiently with the mutant A subunit. Thus, the ability to immunoprecipitate SET with PPP2R1A was due to the PPP2CA/B 111 interaction. A second possibility is that the B subunits may loosely interact with the A-C complex, which can allow SET to interact with the C subunit. This may be reflected in the observation that B subunit interaction was not completely abolished by the presence of the mutation, likely attributing to the need for some PP2A protein function in the tumour. Future studies would be needed to determine this exact mechanism.        Targeting endogenous PP2A inhibitors, like SET, have been recently described to be effective in cancers with SET overexpression [128, 284]. These new exogenous SET inhibitors have been found to be effective in breast, prostate, CML, and CLL cancers, which has resulted in a plethora of recent publications [128, 129, 272, 284-286]. This may indicate new significance for targeting this protein for novel therapeutics, and may provide an alternative way to target endometrial tumours with PPP2R1A mutations to increase PP2A tumour suppressor activity.   7.2 Limitations of Study Designs      The genomic studies completed in Chapters 2, 3 and 4 were limited by targeted gene sequencing experiments without the use of whole genome or exome sequencing. However, full exon sequencing of PPP2R1A in Chapter 3 did not result in the discovery of additional hotspot mutations outside of exon 5 and 6. In Chapter 3, I was able to show that endometrial carcinomas can be classified on the basis of mutational profile, however this does not improve clinical practice or have any benefit to the patients. This work was a stepping-stone for additional molecular studies to dive deeper into endometrial molecular classifications that may stratify patients for predictive markers or targeted therapeutics. The important study by the endometrial TCGA group [42], allowed an improved subclassification of endometrial tumours, however the genomic techniques used to display these classifications are also limited by the ease of use in standard cancer care. The use of targeted gene panels in clinical diagnosis and care is still in early stages however will likely become standard practice in the future.       The functional proteomic studies of PPP2R1A mutations identified in gynaecological malignancies were also limited by the analysis of a single cell line with one PPP2R1A mutation. Unfortunately, the isolation of the isogenic cell lines and subsequent proteomics experiments were time consuming, which limited the proteomics analysis of different PPP2R1A mutations. 112 Although, I was only able to provide evidence that one PPP2R1A mutation causes disruption of a few B subunit interactions, I can only speculate that the lack of these particular B subunits interacting with PPP2R1A can cause tumourigenesis. However, literature evidence suggests a very strong link. The use of additional cell line models will be needed to validate these results, and determine if other hotspot mutations cause similar disruptions.        The use of the Hec1A MMR deficient cell line was a limitation for using the somatic cell knockout technique. This technique had previously been validated to work well in cell lines with MMR (ref), however it was uncertain when I started these experiments if other cell lines could be utilized. Therefore the Hec1A cells with the endogenous hotspot PPP2R1A W257L mutation seemed like the most reasonable cell line to start assessing a specific PPP2R1A endogenous mutation. As discussed throughout Chapter 5 and 6, the cell and disease context of the cell line models should be taken into consideration. The Hec1A cells originated from a primary endometrial endometrioid cancer with MMR deficiency and a hyper-mutator phenotype. PPP2R1A mutations are found in all subtypes of endometrial tumours, however most frequently in endometrial serous carcinomas as identified in Chapter 2 and 3. Of course, the ideal model for assessing PPP2R1A mutations would be in the context of endometrial serous carcinoma, however cell lines originating from primary endometrial serous tumours are not easily obtained. Through our laboratory’s in-house practice of establishing new cell lines from endometrial and ovarian primary tissue, we have been unsuccessful with establishing an endometrial serous cell line. I have now obtained a cell line, SPAC-1-L originating from Japan [287] that is a reported to have originated from a primary endometrial serous adenocarcinoma, however this cell line did not harbor a PPP2R1A mutation as assessed by Sanger sequencing (data not shown). In the future, researchers and future graduate students in our laboratory could use the CRISPR technique to introduce different PPP2R1A mutations into different endometrial and ovarian cell lines. These will be important future experiments as mutations could be assessed in a stable (non mutator phenotype) background, and a wild-type comparison is already available as the parental cell line. The SPAC-1-L cell line, targeted with CRISPR to make PPP2R1A isogenic cell lines, would be a superior context-specific disease model to assess how the hotspot PPP2R1A mutations affect PP2A binding and substrate dephosphorylation, compared to the current Hec1A isogenic model. Furthermore there are a few endometrial and ovarian cell lines; Hec50 113 (endometrial endometrioid cell line with a homozygous R183W), RMG2 (ovarian clear cell carcinoma, heterozygous R183W) that harbor PPP2R1A mutations that could be targeted with CRISPR to generate additional model cell lines for these diseases. These new genomic engineering approaches and use of different cell lines to assess the function of PPP2R1A mutations in gynaecological cancers pose a new set of experimental challenges and limitations.  7.3 Challenges in Studying Protein Complexes      A phosphatase review in 2009 stated that a major challenge for the future would be to find ways to identify the complex phosphatase network. The traditional approach for identifying a phospho-protein and its phosphatase is slow, incomplete and often relies heavily on serendipity [288]. The majority of cellular signaling through phosphorylation acts on serine/threonine residues, of which involves about 428 Ser/Thr kinase genes, and only about 40 phosphatase genes, a disproportion that is perhaps staggering [289]. The difference in the number of genes for kinases compared to phosphatases, underscores the complexity of how phosphatases use a low number of proteins to build multi-protein complexes for specific processes. The complex nature of PP2A emphasizes this point, and has been the source of major challenges throughout this thesis work to identify how PPP2R1A mutations identified in endometrial cancers affects the PP2A composition and thus phosphorylation levels of cellular proteins. The complexity of PP2A does not lie solely in the formation of the heterotrimeric holoenzyme with the A/B/C complexes, but also in the ability of the A and C subunits to bind to multiple other proteins. This adds another layer of finite phosphatase regulation. These noncanonical PP2A complexes involve binding of the C subunit to α4 (IGBP1), PTPA, PME1, which are competing with the A subunit to bind and activate the catalytic subunit [290]. PP2A composition is still widely unknown in different cell types, and yeast studies to understand the stoichiometry have provided conflicting results [291, 292].       The increased use and advances in proteomic technology, especially with the use of Orbitrap technology, will provide large–scale global phospho-protein and protein interaction analysis. The use of AP-MS (affinity purification-mass spectrometry), which is essentially IP-MS, have advanced the ability of proteomics to acquire high-throughput interaction data [293]. These techniques will enable application to in vitro cell models, and importantly, directly from tumours 114 and other human disease tissue. Large-scale PP2A interaction studies have been performed in vitro and suggest that the PP2A interaction network is large [95]. A chemical cross-link study of PP2A interactions identified 176 interprotein and 570 intraprotein crosslinks linking PP2A complexes, giving additional evidence to the complexity of the PP2A network [257].   7.4 Future Directions  7.4.1 Classification of Endometrial Carcinomas      Many studies have shown that the classifications of endometrial carcinomas are in need of a major overhaul. The morphological heterogeneity of the tumours can cause misclassification and irreproducibility in diagnosis [23, 53, 54]. My study along with the TCGA study used molecular markers to try to define and classify endometrial tumours, however there is still much work to be completed. The intermediate, high-risk groups of tumours are very heterogeneous and should be classified with clinical markers to aid in diagnosis and clinical management. The use of POLE mutations for future endometrial classification could be profoundly useful, as many of these tumours with POLE mutations tend to be high-grade endometrioid cases (Table E.8 (Appendix E). The patients with these types of tumours are often thought to do just as poorly as serous carcinomas, however these women tend to have near absolute favorable outcomes [42, 57]. These studies could indicate POLE as a prognostic marker for good outcome in a heterogeneous population of endometrial tumours. TCGA has also used copy number as a classification scheme for endometrial cancers, however this type of analysis is very expensive and needs technical expertise. In a clinical pathology setting, it is unlikely that this type of analysis can be performed on a day-to-day basis; therefore a different method or marker would need to be utilized as a surrogate for copy number changes. These types of research are currently underway in our laboratory and others. Lastly, the integration of histological, molecular, and clinical features of endometrial carcinomas has been proposed to improve the future classifications of endometrial cancers [282].  7.4.2 PP2A Aberrations in Gynaecological and Breast Carcinomas       The copy number high (serous-like) endometrial TCGA cBioPortal data suggests most of the PP2A B subunits are not frequently mutated, however PPP2R2A homozygous deletion occurs in 115 about 10% of copy-number high (serous-like) cases, with a trend towards mutual exclusivity with other B subunits (Figure 7.1) [32, 33]. In Chapter 6, there was a trend towards loss of PPP2R2A interaction with the W257L mutation, however this was not statistically significant in this assay. This phosphatase complex could be a very important part of endometrial serous carcinoma tumour biology, and it could be possible that other PPP2R1A mutations could cause significant disruption of PPP2R2A interaction. This would be very interesting, as it would mean that PPP2R1A mutations and deletion of the PPP2R2A could be assisting in driving aggressive endometrial tumours. Although other B-subunit aberrations are not present in the TCGA data, there is the possibility of epigenetic regulation or post-translational modifications. Therefore further work should be performed to determine the validity of this observation, which could potentially be assessed in the cohort of 89 endometrial serous carcinomas where the mutational status is known for PPP2R1A and FBXW7 (Chapter3, Figure 3.3). It would also be interesting to determine if PPP2R2A deletion is responsible for causing stability of c-Myc in this tissue type. In previous studies, PPP2R5A was responsible for the activation of GSK3β, which de-phosphorylates the S9 residue to allow phosphorylation of c-Myc and ultimately c-Myc degradation [121]. This could be explained by the importance of cell-type context for specific B subunit function, however one study has shown that the deletion of PPP2R2A is not involved in affecting c-Myc levels as assessed by an RNAi screen [121]. Conversely, another study has shown evidence that PPP2R5D is responsible for dephosphorylating GSK3β-S9 which subsequently can phosphorylate c-Myc to cause degradation [190]. Knockdown of PPP2R5D resulted in the accumulation of c-Myc. However, in my study of PPP2R1A mutations with c-Myc IHC in Chapter 3, this did not show any significant association of PPP2R1A mutations with up-regulated c-Myc protein expression. It is possible that not all PPP2R1A mutations cause disruption of PPP2R2A or PPP2R5D equally, which could explain the lack of association of PPP2R1A mutations with high c-Myc expression. 116                Figure 7.1 TCGA endometrial copy number high (serous-like) group (n=60) The cBioPortal TCGA OncoPrint shows each column indicates an individual patient.  Coloured lines indicate a DNA aberration. The grey bars indicate no DNA aberration identified.       Analysis of the TCGA ovarian serous carcinoma data set (n=316), PPP2R1A alterations (including CNA amplification, deletions, mutations) are rare (3%), but are mutually exclusive to 5% of PPP2R2A alterations (mostly consisting of homozygous deletions, and the rare amplification and mutation) (Figure 7.2). Other PP2A regulatory B subunits are also altered in this disease including: 4% CNA amplifications and mutations in PPP2R3A, 2% PPP2R2C, 3% PPP2R5C, 5% PPP2R5D, and 2% deletions in PPP2R5E (Figure 7.2). There is a trend towards mutual exclusivity, however the numbers of cases that are altered are small. Interestingly, these genomic aberrations also show a trend towards mutually exclusivity with 12% of BRCA1 and BRCA2 alterations. Endometrial serous and ovarian serous carcinomas harbor different altered genetic profiles, however histological similarities warrant investigation of PP2A alterations in this cancer type. It has also been established that about 4-7% of ovarian clear carcinoma and 12% of ovarian endometrioid carcinomas harbor PPP2R1A mutations [1, 2, 70]. 117              Figure 7.2 TCGA ovarian serous carcinoma (n=316) The cBioPortal TCGA OncoPrint shows each column indicates an individual patient.  Coloured lines indicate a DNA aberration. The grey bars indicate no DNA aberration identified.        Additionally, the cBioPortal TCGA breast data (n=466) also shows the same trend of mutual exclusivity between 2% PPP2R1A and 2% PPP2R2A alterations (Figure 7.3). The difference lies in that PPP2R1A is not mutated but putatively amplified. Additional evidence that PP2A alterations may play an important role in breast carcinomas is the discovery of ER-positive Luminal B-type breast cancers with PPP2R2A deletions [294]. The investigation of these deletions and how this plays are role in substrate dephosphorylation in the cancer cell is needed.  118            Figure 7.3 TCGA breast carcinomas (all subtypes n=466) The cBioPortal TCGA OncoPrint shows each column indicates an individual patient.  Coloured lines indicate a DNA aberration. The grey bars indicate no DNA aberration identified.        From the TCGA ovarian serous and all breast subtype mutation profiles, there appears to be a trend towards mutual exclusivity of BRCA1 and BRCA2 mutations and PP2A subunit aberrations. BRCA1 and BRCA2 are important genes for homologous recombination, and the TCGA ovarian data suggests that 50% of all ovarian serous carcinomas have defects in the homologous recombination pathway [44]. I can speculate that the mutual exclusive pattern of BRCA1/2 mutations with PP2A genes may indicate that PP2A genes are also affecting double strand break repair via homologous recombination. Kalev et al., has reported on the PP2A genes PPP2R2A, PPP2R2D, PPP2R5A and PPP2R3C involvement in double strand break repair mechanisms. Loss of function of these genes inhibited homologous recombination by increasing phosphorylation levels of ATM, thus sensitizing cells to PARP inhibition [295]. Additional analysis of the TCGA ovarian data also found that a recurrently mutated tumour suppressor gene, CDK12 (9 of 316 (3%) of tumours) [44],  trended towards mutually exclusivity (7 of 9 tumours) with BRCA1 and BRCA2 mutations [296]. Similar to PP2A B subunits loss of function, 119 knockdown of CDK12 lead to suppression of DNA repair mechanisms via homologous recombination, reduced expression of BRCA1, and also sensitized cells to PARP inhibition [296]. Future studies of the effect of PP2A alterations on DNA repair by homologous recombination in gyneacological and breast carcinomas are needed to define the relationship of these proteins. It is possible that these proteins together or separate are biomarkers for patients that may benefit from PARP inhibitors.   7.4.3 Proteomic Studies of Endometrial Cancer Patient Samples      The proteomics finding that PPP2R1A mutations causes disruption of B subunit interactions implies that the specific dephosphorylation events induced by these specific PP2A holoenzyme complex will be decreased. Additional studies of the Hec1A isogenic cell lines could be used as a model to discover the differences in phosphorylation events caused by the PPP2R1A mutation. These studies could potentially lead to alternative deregulated pathways that may be targeted by novel therapeutics. Additional MS experiments using total IP-MS and PRM for specific PP2A interacting proteins would be useful in additional gynaecological cell lines and patient samples. The integration of proteomics analysis in endometrial tumour patient samples would provide insight into the effect of PPP2R1A mutations on the PP2A complex in human tumours. However, the tools for these types of proteomics studies are slowly catching up with the field of genomics to enable proteomics profiling of patient tumours. The first draft of the human proteome was only published in May 2014 [297], which is more than 10 years behind the release of the first human genome. The integration of clinical, genomics and proteomics will be an exciting new field for cancer research.  7.4.4 Novel Therapeutic Options for Targeting PP2A Altered Cancers       Endometrial serous carcinomas are characterized almost exclusively with mutations in the tumour suppressor TP53, and a subset of these with PPP2R1A mutations. Since TP53 is also frequently mutated in many other types of cancers, there have been concentrated efforts to target p53 with novel therapeutics. Earlier studies have shown that restoring p53 alone is sufficient to cause regression of lymphomas, sarcomas, and liver carcinomas in mouse models [235, 298]. Therefore restoring the function of p53 could be an attractive therapeutic strategy for many tumours. Viruses, small molecule inhibitors, and protein chaperones are now being used in 120 clinical trials in an attempt to restore the wild-type function of p53 in tumour cells [299]. A novel potential synthetic lethal approach for targeting endometrial and perhaps ovarian serous carcinoma, includes therapeutics to restore wild-type p53 and wild-type PPP2R1A or B subunits to restore the function of PP2A.  The inhibition of PP2A can kill cancerous cells, however since this is an essential gene, inhibition would also kill normal cells. Overexpression of wild-type PPP2R1A has been shown to decrease cell proliferation and tumour growth [131]. In addition, multiple PP2A inhibitors (okadaic acid, calyculin A) are potent tumour promoters, and viruses that bind to PP2A to increase cellular transformation [300]. From my own experiments in Chapter 5, I have also shown that my attempts of isolating Hec1A isogenic cell lines with one copy of wild-type PPP2R1A were unsuccessful. This was likely due to slow growth with no growth advantage; therefore the wild-type cells are not selected. A synthetic lethal approach to target two major tumour suppressor genes by restoring wild-type function could be an “out of the box” solution for targeting the tumour cells and leaving the normal cells alone.  This would essentially restore the function of p53 and PP2A to bring the tumour cells back to a “normal” like state, which would not allow abnormal growth and proliferation.       Currently, there is a FDA approved drug for use in multiple sclerosis, FTY720, a sphingosine analogue drug that is an activator of PP2A [300], and has shown to be well tolerated in humans [301, 302]. This drug demonstrates an antitumour effect in leukemia and mouse xenographs with no toxicity [303]. Studies also suggest that prostate cancer cells are more sensitive to the drug than normal stromal cells [304]. It is interesting to note that FTY720 acts to bind and target the PP2A inhibitor SET, to induce reactivation of PP2A [284]. The use of an additional SET peptide inhibitor (OP449) significantly reduces proliferation and cell survival signaling [128, 129]. OP449 acts to antagonize SET inhibition of PP2A, and has been reported to be efficacious in vitro and in vivo in chronic myeloid leukemia (CML) and acute myeloid leukemia (AML) when used in combination with tyrosine kinase inhibitors [286]. Furthermore, the mass spectrometry results, presented in Chapter 6, illustrate an increased affinity of the PPP2R1A L257 mutation with the PP2A inhibitor SET. It would be useful to test the efficacy of these drugs on endometrial cell lines and patient derived mouse xenographs with known PP2A aberrations (PPP2R1A mutations or PPP2R2A deletions), to determine if this could be a useful targeted 121 therapy. In conclusion, future studies of targeting SET to reactivate PP2A could provide an additional way to treat PP2A disrupted cancers for novel therapeutics.      122 Bibliography 1. McConechy, M.K., Anglesio, M.S., Kalloger, S.E., Yang, W., Senz, J., Chow, C., Heravi-Moussavi, A., Morin, G.B., Mes-Masson, A.M., Carey, M.S., et al. (2011). Subtype-specific mutation of PPP2R1A in endometrial and ovarian carcinomas. The Journal of pathology 223, 567-573. 2. McConechy, M.K., Ding, J., Cheang, M.C., Wiegand, K.C., Senz, J., Tone, A.A., Yang, W., Prentice, L.M., Tse, K., Zeng, T., et al. (2012). Use of mutation profiles to refine the classification of endometrial carcinomas. The Journal of pathology 228, 20-30. 3. McConechy, M.K., Ding, J., Senz, J., Yang, W., Melnyk, N., Tone, A.A., Prentice, L.M., Wiegand, K.C., McAlpine, J.N., Shah, S.P., et al. (2014). Ovarian and endometrial endometrioid carcinomas have distinct CTNNB1 and PTEN mutation profiles. Modern Pathol 27, 128-134. 4. Jemal, A., Bray, F., Center, M.M., Ferlay, J., Ward, E., and Forman, D. (2011). Global cancer statistics. CA Cancer J Clin 61, 69-90. 5. Ferlay, J., Shin, H.R., Bray, F., Forman, D., Mathers, C., and Parkin, D.M. (2010). GLOBOCAN 2008 v1.2, Cancer Incidence and Mortality Worldwide: IARC CancerBase No. 10 Lyon, France, http://globocan.iarc.fr. In International Agency for Research on Cancer. 6. Evans, T., Sany, O., Pearmain, P., Ganesan, R., Blann, A., and Sundar, S. (2011). Differential trends in the rising incidence of endometrial cancer by type: data from a UK population-based registry from 1994 to 2006. Br J Cancer 104, 1505-1510. 7. Okuda, T., Sekizawa, A., Purwosunu, Y., Nagatsuka, M., Morioka, M., Hayashi, M., and Okai, T. (2010). Genetics of endometrial cancers. Obstet Gynecol Int 2010, 984013. 8. Bokhman, J.V. (1983). Two pathogenetic types of endometrial carcinoma. Gynecol Oncol 15, 10-17. 9. Eltabbakh, G.H., Piver, M.S., Hempling, R.E., and Shin, K.H. (1997). Excellent long-term survival and absence of vaginal recurrences in 332 patients with low-risk stage I endometrial adenocarcinoma treated with hysterectomy and vaginal brachytherapy without formal staging lymph node sampling: report of a prospective trial. International journal of radiation oncology, biology, physics 38, 373-380. 10. Hamilton, C.A., Cheung, M.K., Osann, K., Chen, L., Teng, N.N., Longacre, T.A., Powell, M.A., Hendrickson, M.R., Kapp, D.S., and Chan, J.K. (2006). Uterine papillary serous and clear cell carcinomas predict for poorer survival compared to grade 3 endometrioid corpus cancers. Br J Cancer 94, 642-646. 11. Hendrickson, M., Ross, J., Eifel, P., Martinez, A., and Kempson, R. (1982). Uterine papillary serous carcinoma: a highly malignant form of endometrial adenocarcinoma. Am J Surg Pathol 6, 93-108. 12. Chan, J.K., Loizzi, V., Youssef, M., Osann, K., Rutgers, J., Vasilev, S.A., and Berman, M.L. (2003). Significance of comprehensive surgical staging in noninvasive papillary serous carcinoma of the endometrium. Gynecol Oncol 90, 181-185. 13. Goff, B.A., Kato, D., Schmidt, R.A., Ek, M., Ferry, J.A., Muntz, H.G., Cain, J.M., Tamimi, H.K., Figge, D.C., and Greer, B.E. (1994). Uterine papillary serous carcinoma: patterns of metastatic spread. Gynecol Oncol 54, 264-268. 123 14. Kato, D.T., Ferry, J.A., Goodman, A., Sullinger, J., Scully, R.E., Goff, B.A., Fuller, A.F., Jr., and Rice, L.W. (1995). Uterine papillary serous carcinoma (UPSC): a clinicopathologic study of 30 cases. Gynecol Oncol 59, 384-389. 15. Slomovitz, B.M., Burke, T.W., Eifel, P.J., Ramondetta, L.M., Silva, E.G., Jhingran, A., Oh, J.C., Atkinson, E.N., Broaddus, R.R., Gershenson, D.M., et al. (2003). Uterine papillary serous carcinoma (UPSC): a single institution review of 129 cases. Gynecol Oncol 91, 463-469. 16. Jonson, A.L., Bliss, R.L., Truskinovsky, A., Judson, P., Argenta, P., Carson, L., Dusenbery, K., and Downs, L.S., Jr. (2006). Clinical features and outcomes of uterine and ovarian carcinosarcoma. Gynecol Oncol 100, 561-564. 17. Hoskins, P.J., Swenerton, K.D., Pike, J.A., Wong, F., Lim, P., Acquino-Parsons, C., and Lee, N. (2001). Paclitaxel and carboplatin, alone or with irradiation, in advanced or recurrent endometrial cancer: a phase II study. Journal of clinical oncology : official journal of the American Society of Clinical Oncology 19, 4048-4053. 18. Carey, M.S., Gawlik, C., Fung-Kee-Fung, M., Chambers, A., Oliver, T., and Cancer Care Ontario Practice Guidelines Initiative Gynecology Cancer Disease Site, G. (2006). Systematic review of systemic therapy for advanced or recurrent endometrial cancer. Gynecol Oncol 101, 158-167. 19. Widmaier, E.P., Raff, H., Strang, K.T., and Vander, A.J. (2014). Vander's human physiology : the mechanisms of body function, Thirteenth edition / Edition, (New York: McGraw-Hill). 20. Kurman, R.J., International Agency for Research on Cancer., and World Health Organization. (2014). WHO classification of tumours of female reproductive organs, 4th Edition, (Lyon: International Agency for Research on Cancer). 21. Benedet, J.L., Bender, H., Jones, H., 3rd, Ngan, H.Y., and Pecorelli, S. (2000). FIGO staging classifications and clinical practice guidelines in the management of gynecologic cancers. FIGO Committee on Gynecologic Oncology. International journal of gynaecology and obstetrics: the official organ of the International Federation of Gynaecology and Obstetrics 70, 209-262. 22. Darvishian, F., Hummer, A.J., Thaler, H.T., Bhargava, R., Linkov, I., Asher, M., and Soslow, R.A. (2004). Serous endometrial cancers that mimic endometrioid adenocarcinomas: a clinicopathologic and immunohistochemical study of a group of problematic cases. Am J Surg Pathol 28, 1568-1578. 23. Alkushi, A., Kobel, M., Kalloger, S.E., and Gilks, C.B. (2010). High-grade endometrial carcinoma: serous and grade 3 endometrioid carcinomas have different immunophenotypes and outcomes. Int J Gynecol Pathol 29, 343-350. 24. Norris, H.J., and Taylor, H.B. (1966). Mesenchymal tumors of the uterus. 3. A clinical and pathologic study of 31 carcinosarcomas. Cancer 19, 1459-1465. 25. McCluggage, W.G. (2002). Uterine carcinosarcomas (malignant mixed Mullerian tumors) are metaplastic carcinomas. Int J Gynecol Cancer 12, 687-690. 26. Risinger, J.I., Berchuck, A., Kohler, M.F., Watson, P., Lynch, H.T., and Boyd, J. (1993). Genetic instability of microsatellites in endometrial carcinoma. Cancer research 53, 5100-5103. 27. Cheung, L.W., Hennessy, B.T., Li, J., Yu, S., Myers, A.P., Djordjevic, B., Lu, Y., Stemke-Hale, K., Zhang, F., Ju, Z., et al. (2011). High Frequency of PIK3R1 and PIK3R2 124 Mutations in Endometrial Cancer Elucidates a Novel Mechanism for Regulation of PTEN Protein Stability. Cancer discovery 1, 170-185. 28. Urick, M.E., Rudd, M.L., Godwin, A.K., Sgroi, D., Merino, M., and Bell, D.W. (2011). PIK3R1 (p85{alpha}) Is Somatically Mutated at High Frequency in Primary Endometrial Cancer. Cancer research 71, 4061-4067. 29. O'Hara, A.J., and Bell, D.W. (2012). The genomics and genetics of endometrial cancer. Advances in Genomics and Genetics 2, 33-47. 30. Byron, S.A., Gartside, M., Powell, M.A., Wellens, C.L., Gao, F., Mutch, D.G., Goodfellow, P.J., and Pollock, P.M. (2012). FGFR2 Point Mutations in 466 Endometrioid Endometrial Tumors: Relationship with MSI, KRAS, PIK3CA, CTNNB1 Mutations and Clinicopathological Features. PloS one 7, e30801. 31. Sun, H., Enomoto, T., Fujita, M., Wada, H., Yoshino, K., Ozaki, K., Nakamura, T., and Murata, Y. (2001). Mutational analysis of the PTEN gene in endometrial carcinoma and hyperplasia. Am J Clin Pathol 115, 32-38. 32. Cerami, E., Gao, J., Dogrusoz, U., Gross, B.E., Sumer, S.O., Aksoy, B.A., Jacobsen, A., Byrne, C.J., Heuer, M.L., Larsson, E., et al. (2012). The cBio Cancer Genomics Portal: An Open Platform for Exploring Multidimensional Cancer Genomics Data. Cancer discovery 2, 401-404. 33. Gao, J., Aksoy, B.A., Dogrusoz, U., Dresdner, G., Gross, B., Sumer, S.O., Sun, Y., Jacobsen, A., Sinha, R., Larsson, E., et al. (2013). Integrative Analysis of Complex Cancer Genomics and Clinical Profiles Using the cBioPortal. Science Signaling 6, pl1-pl1. 34. Guan, B., Mao, T.L., Panuganti, P.K., Kuhn, E., Kurman, R.J., Maeda, D., Chen, E., Jeng, Y.M., Wang, T.L., and Shih Ie, M. (2011). Mutation and loss of expression of ARID1A in uterine low-grade endometrioid carcinoma. Am J Surg Pathol 35, 625-632. 35. Wiegand, K.C., Lee, A.F., Al-Agha, O.M., Chow, C., Kalloger, S.E., Scott, D.W., Steidl, C., Wiseman, S.M., Gascoyne, R.D., Gilks, B., et al. (2011). Loss of BAF250a (ARID1A) is frequent in high-grade endometrial carcinomas. The Journal of pathology 224, 328-333. 36. Hampel, H., Frankel, W., Panescu, J., Lockman, J., Sotamaa, K., Fix, D., Comeras, I., La Jeunesse, J., Nakagawa, H., Westman, J.A., et al. (2006). Screening for Lynch syndrome (hereditary nonpolyposis colorectal cancer) among endometrial cancer patients. Cancer research 66, 7810-7817. 37. Lynch, H.T., Lynch, P.M., and Harris, R.E. (1978). Minimal genetic findings and their cancer control implications. A family with the cancer family syndrome. Jama 240, 535-538. 38. Schmeler, K.M., Lynch, H.T., Chen, L.-m., Munsell, M.F., Soliman, P.T., Clark, M.B., Daniels, M.S., White, K.G., Boyd-Rogers, S.G., Conrad, P.G., et al. (2006). Prophylactic Surgery to Reduce the Risk of Gynecologic Cancers in the Lynch Syndrome. New England Journal of Medicine 354, 261-269. 39. Geurts-Giele, W.R., Leenen, C.H., Dubbink, H.J., Meijssen, I.C., Post, E., Sleddens, H.F., Kuipers, E.J., Goverde, A., van den Ouweland, A.M., van Lier, M.G., et al. (2014). Somatic aberrations of mismatch repair genes as a cause of microsatellite-unstable cancers. J Pathol. 125 40. Lax, S.F., Kendall, B., Tashiro, H., Slebos, R.J., and Hedrick, L. (2000). The frequency of p53, K-ras mutations, and microsatellite instability differs in uterine endometrioid and serous carcinoma: evidence of distinct molecular genetic pathways. Cancer 88, 814-824. 41. Pradhan, M., Abeler, V.M., Danielsen, H.E., Trope, C.G., and Risberg, B.A. (2006). Image cytometry DNA ploidy correlates with histological subtypes in endometrial carcinomas. Mod Pathol 19, 1227-1235. 42. Kandoth, C., Schultz, N., Cherniack, A.D., Akbani, R., Liu, Y., Shen, H., Robertson, A.G., Pashtan, I., Shen, R., Benz, C.C., et al. (2013). Integrated genomic characterization of endometrial carcinoma. Nature 497, 67-73. 43. Gallo, M.L., O'Hara, A.J., Rudd, M.L., Urick, M.E., Hansen, N.F., O'Neil, N.J., Price, J.C., Zhang, S., England, B.M., Godwin, A.K., et al. (2012). Exome sequencing of serous endometrial tumors identifies recurrent somatic mutations in chromatin-remodeling and ubiquitin ligase complex genes. Nature genetics. 44. Cancer Genome Atlas Research, N. (2011). Integrated genomic analyses of ovarian carcinoma. Nature 474, 609-615. 45. Cassia, R., Moreno-Bueno, G., Rodriguez-Perales, S., Hardisson, D., Cigudosa, J.C., and Palacios, J. (2003). Cyclin E gene (CCNE) amplification and hCDC4 mutations in endometrial carcinoma. J Pathol 201, 589-595. 46. Kuhn, E., Wu, R.C., Guan, B., Wu, G., Zhang, J., Wang, Y., Song, L., Yuan, X., Wei, L., Roden, R.B., et al. (2012). Identification of Molecular Pathway Aberrations in Uterine Serous Carcinoma by Genome-wide Analyses. J Natl Cancer Inst. 47. Fadare, O., Gwin, K., Desouki, M.M., Crispens, M.A., Jones, H.W., 3rd, Khabele, D., Liang, S.X., Zheng, W., Mohammed, K., Hecht, J.L., et al. (2013). The clinicopathologic significance of p53 and BAF-250a (ARID1A) expression in clear cell carcinoma of the endometrium. Mod Pathol 26, 1101-1110. 48. Kuo, K.T., Mao, T.L., Jones, S., Veras, E., Ayhan, A., Wang, T.L., Glas, R., Slamon, D., Velculescu, V.E., Kuman, R.J., et al. (2009). Frequent activating mutations of PIK3CA in ovarian clear cell carcinoma. The American journal of pathology 174, 1597-1601. 49. Hoang, L.N., Han, G., McConechy, M., Lau, S., Chow, C., Gilks, C.B., Huntsman, D.G., Kobel, M., and Lee, C.H. (2013). Immunohistochemical characterization of prototypical endometrial clear cell carcinoma-diagnostic utility of HNF-1beta and oestrogen receptor. Histopathology. 50. Kobayashi, H., Uekuri, C., Akasaka, J., Ito, F., Shigemitsu, A., Koike, N., and Shigetomi, H. (2013). The biology of uterine sarcomas: A review and update. Molecular and clinical oncology 1, 599-609. 51. Taylor, N.P., Zighelboim, I., Huettner, P.C., Powell, M.A., Gibb, R.K., Rader, J.S., Mutch, D.G., Edmonston, T.B., and Goodfellow, P.J. (2006). DNA mismatch repair and TP53 defects are early events in uterine carcinosarcoma tumorigenesis. Mod Pathol 19, 1333-1338. 52. McConechy, M.K., Hoang, L.N., Chui, M.H., Senz, J., Yang, W., Rozenberg, N., Mackenzie, R., McAlpine, J.N., Huntsman, D.G., Clarke, B.A., et al. (2015). In-depth molecular profiling of the biphasic components of uterine carcinosarcomas. The Journal of Pathology: Clinical Research, n/a-n/a. 53. Clarke, B.A., and Gilks, C.B. (2010). Endometrial carcinoma: controversies in histopathological assessment of grade and tumour cell type. J Clin Pathol 63, 410-415. 126 54. Gilks, C.B., Oliva, E., and Soslow, R. (2011). Poor Interobserver Reproducibility in the Diagnosis of High-Grade Endometrial Carcinoma. Modern Pathol 24, 248A-248A. 55. Hoang, L.N., McConechy, M.K., Kobel, M., Han, G., Rouzbahman, M., Davidson, B., Irving, J., Ali, R.H., Leung, S., McAlpine, J.N., et al. (2013). Histotype-genotype correlation in 36 high-grade endometrial carcinomas. Am J Surg Pathol 37, 1421-1432. 56. Billingsley, C.C., Cohn, D.E., Mutch, D.G., Stephens, J.A., Suarez, A.A., and Goodfellow, P.J. (2014). Polymerase varepsilon (POLE) mutations in endometrial cancer: Clinical outcomes and implications for Lynch syndrome testing. Cancer. 57. Meng, B., Hoang, L.N., McIntyre, J.B., Duggan, M.A., Nelson, G.S., Lee, C.H., and Kobel, M. (2014). POLE exonuclease domain mutation predicts long progression-free survival in grade 3 endometrioid carcinoma of the endometrium. Gynecol Oncol 134, 15-19. 58. Church, D.N., Briggs, S.E., Palles, C., Domingo, E., Kearsey, S.J., Grimes, J.M., Gorman, M., Martin, L., Howarth, K.M., Hodgson, S.V., et al. (2013). DNA polymerase epsilon and delta exonuclease domain mutations in endometrial cancer. Human molecular genetics 22, 2820-2828. 59. Kurman, R.J., and Shih Ie, M. (2010). The origin and pathogenesis of epithelial ovarian cancer: a proposed unifying theory. Am J Surg Pathol 34, 433-443. 60. Kurman, R.J., and Shih Ie, M. (2011). Molecular pathogenesis and extraovarian origin of epithelial ovarian cancer--shifting the paradigm. Human pathology 42, 918-931. 61. Lee, Y., Miron, A., Drapkin, R., Nucci, M.R., Medeiros, F., Saleemuddin, A., Garber, J., Birch, C., Mou, H., Gordon, R.W., et al. (2007). A candidate precursor to serous carcinoma that originates in the distal fallopian tube. J Pathol 211, 26-35. 62. Crum, C.P., Drapkin, R., Kindelberger, D., Medeiros, F., Miron, A., and Lee, Y. (2007). Lessons from BRCA: the tubal fimbria emerges as an origin for pelvic serous cancer. Clinical medicine & research 5, 35-44. 63. Ahmed, A.A., Etemadmoghadam, D., Temple, J., Lynch, A.G., Riad, M., Sharma, R., Stewart, C., Fereday, S., Caldas, C., Defazio, A., et al. (2010). Driver mutations in TP53 are ubiquitous in high grade serous carcinoma of the ovary. The Journal of pathology 221, 49-56. 64. Miller, D.S., Blessing, J.A., Krasner, C.N., Mannel, R.S., Hanjani, P., Pearl, M.L., Waggoner, S.E., and Boardman, C.H. (2009). Phase II evaluation of pemetrexed in the treatment of recurrent or persistent platinum-resistant ovarian or primary peritoneal carcinoma: a study of the Gynecologic Oncology Group. Journal of clinical oncology : official journal of the American Society of Clinical Oncology 27, 2686-2691. 65. Kobel, M., Kalloger, S.E., Huntsman, D.G., Santos, J.L., Swenerton, K.D., Seidman, J.D., Gilks, C.B., and Cheryl Brown Ovarian Cancer Outcomes Unit of the British Columbia Cancer Agency, V.B.C. (2010). Differences in tumor type in low-stage versus high-stage ovarian carcinomas. Int J Gynecol Pathol 29, 203-211. 66. Kumar, A., Le, N., Tinker, A.V., Santos, J.L., Parsons, C., and Hoskins, P.J. (2014). Early-stage endometrioid ovarian carcinoma: population-based outcomes in British Columbia. Int J Gynecol Cancer 24, 1401-1405. 67. Giudice, L.C. (2010). Endometriosis. New England Journal of Medicine 362, 2389-2398. 68. Brescia, R.J., Dubin, N., and Demopoulos, R.I. (1989). Endometrioid and clear cell carcinoma of the ovary. Factors affecting survival. Int J Gynecol Pathol 8, 132-138. 127 69. Kobel, M., Kalloger, S.E., Boyd, N., McKinney, S., Mehl, E., Palmer, C., Leung, S., Bowen, N.J., Ionescu, D.N., Rajput, A., et al. (2008). Ovarian carcinoma subtypes are different diseases: implications for biomarker studies. PLoS Med 5, e232. 70. Jones, S., Wang, T.L., Shih Ie, M., Mao, T.L., Nakayama, K., Roden, R., Glas, R., Slamon, D., Diaz, L.A., Jr., Vogelstein, B., et al. (2010). Frequent mutations of chromatin remodeling gene ARID1A in ovarian clear cell carcinoma. Science 330, 228-231. 71. Wiegand, K.C., Shah, S.P., Al-Agha, O.M., Zhao, Y., Tse, K., Zeng, T., Senz, J., McConechy, M.K., Anglesio, M.S., Kalloger, S.E., et al. (2010). ARID1A mutations in endometriosis-associated ovarian carcinomas. The New England journal of medicine 363, 1532-1543. 72. Campbell, I.G., Russell, S.E., Choong, D.Y., Montgomery, K.G., Ciavarella, M.L., Hooi, C.S., Cristiano, B.E., Pearson, R.B., and Phillips, W.A. (2004). Mutation of the PIK3CA gene in ovarian and breast cancer. Cancer research 64, 7678-7681. 73. Sato, N., Tsunoda, H., Nishida, M., Morishita, Y., Takimoto, Y., Kubo, T., and Noguchi, M. (2000). Loss of Heterozygosity on 10q23.3 and Mutation of the Tumor Suppressor Gene PTEN in Benign Endometrial Cyst of the Ovary: Possible Sequence Progression from Benign Endometrial Cyst to Endometrioid Carcinoma and Clear Cell Carcinoma of the Ovary. Cancer research 60, 7052-7056. 74. Catasus, L., Bussaglia, E., Rodrguez, I., Gallardo, A., Pons, C., Irving, J.A., and Prat, J. (2004). Molecular genetic alterations in endometrioid carcinomas of the ovary: similar frequency of beta-catenin abnormalities but lower rate of microsatellite instability and PTEN alterations than in uterine endometrioid carcinomas. Human pathology 35, 1360-1368. 75. Gamallo, C., Palacios, J., Moreno, G., Calvo de Mora, J., Suarez, A., and Armas, A. (1999). beta-catenin expression pattern in stage I and II ovarian carcinomas : relationship with beta-catenin gene mutations, clinicopathological features, and clinical outcome. The American journal of pathology 155, 527-536. 76. Jiang, X., Morland, S.J., Hitchcock, A., Thomas, E.J., and Campbell, I.G. (1998). Allelotyping of endometriosis with adjacent ovarian carcinoma reveals evidence of a common lineage. Cancer research 58, 1707-1712. 77. Bell, D.A. (2005). Origins and molecular pathology of ovarian cancer. Modern pathology : an official journal of the United States and Canadian Academy of Pathology, Inc 18 Suppl 2, S19-32. 78. Stratton, J.F., Gayther, S.A., Russell, P., Dearden, J., Gore, M., Blake, P., Easton, D., and Ponder, B.A.J. (1997). Contribution of BRCA1 Mutations to Ovarian Cancer. New England Journal of Medicine 336, 1125-1130. 79. Audeh, M.W., Carmichael, J., Penson, R.T., Friedlander, M., Powell, B., Bell-McGuinn, K.M., Scott, C., Weitzel, J.N., Oaknin, A., Loman, N., et al. (2010). Oral poly(ADP-ribose) polymerase inhibitor olaparib in patients with BRCA1 or BRCA2 mutations and recurrent ovarian cancer: a proof-of-concept trial. The Lancet 376, 245-251. 80. Fong, P.C., Yap, T.A., Boss, D.S., Carden, C.P., Mergui-Roelvink, M., Gourley, C., De Greve, J., Lubinski, J., Shanley, S., Messiou, C., et al. (2010). Poly(ADP)-ribose polymerase inhibition: frequent durable responses in BRCA carrier ovarian cancer correlating with platinum-free interval. Journal of clinical oncology : official journal of the American Society of Clinical Oncology 28, 2512-2519. 128 81. Bouwman, P., and Jonkers, J. (2012). The effects of deregulated DNA damage signalling on cancer chemotherapy response and resistance. Nat Rev Cancer 12, 587-598. 82. Eichhorn, P.J., Creyghton, M.P., and Bernards, R. (2009). Protein phosphatase 2A regulatory subunits and cancer. Biochimica et biophysica acta 1795, 1-15. 83. Zhou, J., Pham, H.T., and Walter, G. (2003). The formation and activity of PP2A holoenzymes do not depend on the isoform of the catalytic subunit. The Journal of biological chemistry 278, 8617-8622. 84. Virshup, D.M. (2000). Protein phosphatase 2A: a panoply of enzymes. Current opinion in cell biology 12, 180-185. 85. Hunter, T. (1995). Protein kinases and phosphatases: the yin and yang of protein phosphorylation and signaling. Cell 80, 225-236. 86. Xu, Y., Xing, Y., Chen, Y., Chao, Y., Lin, Z., Fan, E., Yu, J.W., Strack, S., Jeffrey, P.D., and Shi, Y. (2006). Structure of the protein phosphatase 2A holoenzyme. Cell 127, 1239-1251. 87. Hemmings, B.A., Adams-Pearson, C., Maurer, F., Muller, P., Goris, J., Merlevede, W., Hofsteenge, J., and Stone, S.R. (1990). alpha- and beta-forms of the 65-kDa subunit of protein phosphatase 2A have a similar 39 amino acid repeating structure. Biochemistry 29, 3166-3173. 88. Ruediger, R., Van Wart Hood, J.E., Mumby, M., and Walter, G. (1991). Constant expression and activity of protein phosphatase 2A in synchronized cells. Molecular and cellular biology 11, 4282-4285. 89. Arino, J., Woon, C.W., Brautigan, D.L., Miller, T.B., Jr., and Johnson, G.L. (1988). Human liver phosphatase 2A: cDNA and amino acid sequence of two catalytic subunit isotypes. Proc Natl Acad Sci U S A 85, 4252-4256. 90. Ruediger, R., Pham, H.T., and Walter, G. (2001). Disruption of protein phosphatase 2A subunit interaction in human cancers with mutations in the A alpha subunit gene. Oncogene 20, 10-15. 91. Xu, Y., Chen, Y., Zhang, P., Jeffrey, P.D., and Shi, Y. (2008). Structure of a protein phosphatase 2A holoenzyme: insights into B55-mediated Tau dephosphorylation. Molecular cell 31, 873-885. 92. Groves, M.R., Hanlon, N., Turowski, P., Hemmings, B.A., and Barford, D. (1999). The structure of the protein phosphatase 2A PR65/A subunit reveals the conformation of its 15 tandemly repeated HEAT motifs. Cell 96, 99-110. 93. Ruediger, R., Roeckel, D., Fait, J., Bergqvist, A., Magnusson, G., and Walter, G. (1992). Identification of binding sites on the regulatory A subunit of protein phosphatase 2A for the catalytic C subunit and for tumor antigens of simian virus 40 and polyomavirus. Molecular and cellular biology 12, 4872-4882. 94. Cho, U.S., and Xu, W. (2007). Crystal structure of a protein phosphatase 2A heterotrimeric holoenzyme. Nature 445, 53-57. 95. Goudreault, M., D'Ambrosio, L.M., Kean, M.J., Mullin, M.J., Larsen, B.G., Sanchez, A., Chaudhry, S., Chen, G.I., Sicheri, F., Nesvizhskii, A.I., et al. (2009). A PP2A phosphatase high density interaction network identifies a novel striatin-interacting phosphatase and kinase complex linked to the cerebral cavernous malformation 3 (CCM3) protein. Molecular & cellular proteomics : MCP 8, 157-171. 96. Margolis, S.S., Perry, J.A., Forester, C.M., Nutt, L.K., Guo, Y., Jardim, M.J., Thomenius, M.J., Freel, C.D., Darbandi, R., Ahn, J.H., et al. (2006). Role for the PP2A/B56delta 129 phosphatase in regulating 14-3-3 release from Cdc25 to control mitosis. Cell 127, 759-773. 97. Forester, C.M., Maddox, J., Louis, J.V., Goris, J., and Virshup, D.M. (2007). Control of mitotic exit by PP2A regulation of Cdc25C and Cdk1. Proc Natl Acad Sci U S A 104, 19867-19872. 98. Gharbi-Ayachi, A., Labbe, J.C., Burgess, A., Vigneron, S., Strub, J.M., Brioudes, E., Van-Dorsselaer, A., Castro, A., and Lorca, T. (2010). The substrate of Greatwall kinase, Arpp19, controls mitosis by inhibiting protein phosphatase 2A. Science 330, 1673-1677. 99. Lee, T.Y., Lai, T.Y., Lin, S.C., Wu, C.W., Ni, I.F., Yang, Y.S., Hung, L.Y., Law, B.K., and Chiang, C.W. (2010). The B56gamma3 regulatory subunit of protein phosphatase 2A (PP2A) regulates S phase-specific nuclear accumulation of PP2A and the G1 to S transition. The Journal of biological chemistry 285, 21567-21580. 100. Schmitz, M.H., Held, M., Janssens, V., Hutchins, J.R., Hudecz, O., Ivanova, E., Goris, J., Trinkle-Mulcahy, L., Lamond, A.I., Poser, I., et al. (2010). Live-cell imaging RNAi screen identifies PP2A-B55alpha and importin-beta1 as key mitotic exit regulators in human cells. Nature cell biology 12, 886-893. 101. Manchado, E., Guillamot, M., de Cárcer, G., Eguren, M., Trickey, M., García-Higuera, I., Moreno, S., Yamano, H., Cañamero, M., and Malumbres, M. (2010). Targeting Mitotic Exit Leads to Tumor Regression In Vivo: Modulation by Cdk1, Mastl, and the PP2A/B55α,δ Phosphatase. Cancer Cell 18, 641-654. 102. Kitajima, T.S., Sakuno, T., Ishiguro, K., Iemura, S., Natsume, T., Kawashima, S.A., and Watanabe, Y. (2006). Shugoshin collaborates with protein phosphatase 2A to protect cohesin. Nature 441, 46-52. 103. Suganuma, M., Fujiki, H., Suguri, H., Yoshizawa, S., Hirota, M., Nakayasu, M., Ojika, M., Wakamatsu, K., Yamada, K., and Sugimura, T. (1988). Okadaic acid: an additional non-phorbol-12-tetradecanoate-13-acetate-type tumor promoter. Proc Natl Acad Sci U S A 85, 1768-1771. 104. Takai, A., Bialojan, C., Troschka, M., and Ruegg, J.C. (1987). Smooth muscle myosin phosphatase inhibition and force enhancement by black sponge toxin. FEBS letters 217, 81-84. 105. Schonthal, A.H. (2001). Role of serine/threonine protein phosphatase 2A in cancer. Cancer Lett 170, 1-13. 106. Walter, G., Ruediger, R., Slaughter, C., and Mumby, M. (1990). Association of protein phosphatase 2A with polyoma virus medium tumor antigen. Proc Natl Acad Sci U S A 87, 2521-2525. 107. Pallas, D.C., Shahrik, L.K., Martin, B.L., Jaspers, S., Miller, T.B., Brautigan, D.L., and Roberts, T.M. (1990). Polyoma small and middle T antigens and SV40 small t antigen form stable complexes with protein phosphatase 2A. Cell 60, 167-176. 108. Rodriguez-Viciana, P., Collins, C., and Fried, M. (2006). Polyoma and SV40 proteins differentially regulate PP2A to activate distinct cellular signaling pathways involved in growth control. Proc Natl Acad Sci U S A 103, 19290-19295. 109. Chen, W., Possemato, R., Campbell, K.T., Plattner, C.A., Pallas, D.C., and Hahn, W.C. (2004). Identification of specific PP2A complexes involved in human cell transformation. Cancer Cell 5, 127-136. 130 110. Sablina, A.A., Hector, M., Colpaert, N., and Hahn, W.C. (2010). Identification of PP2A complexes and pathways involved in cell transformation. Cancer research 70, 10474-10484. 111. Mui, M.Z., Kucharski, M., Miron, M.J., Hur, W.S., Berghuis, A.M., Blanchette, P., and Branton, P.E. (2013). Identification of the adenovirus E4orf4 protein binding site on the B55alpha and Cdc55 regulatory subunits of PP2A: Implications for PP2A function, tumor cell killing and viral replication. PLoS pathogens 9, e1003742. 112. Eichhorn, P.J., Creyghton, M.P., Wilhelmsen, K., van Dam, H., and Bernards, R. (2007). A RNA interference screen identifies the protein phosphatase 2A subunit PR55gamma as a stress-sensitive inhibitor of c-SRC. PLoS Genet 3, e218. 113. Ory, S., Zhou, M., Conrads, T.P., Veenstra, T.D., and Morrison, D.K. (2003). Protein phosphatase 2A positively regulates Ras signaling by dephosphorylating KSR1 and Raf-1 on critical 14-3-3 binding sites. Current biology : CB 13, 1356-1364. 114. Gilan, O., Diesch, J., Amalia, M., Jastrzebski, K., Chueh, A., Verrills, N.M., Pearson, R.B., Mariadason, J.M., Tulchinsky, E., Hannan, R.D., et al. (2014). PR55alpha-containing protein phosphatase 2A complexes promote cancer cell migration and invasion through regulation of AP-1 transcriptional activity. Oncogene 0. 115. Ruvolo, P.P., Qui, Y.H., Coombes, K.R., Zhang, N., Ruvolo, V.R., Borthakur, G., Konopleva, M., Andreeff, M., and Kornblau, S.M. (2011). Low expression of PP2A regulatory subunit B55alpha is associated with T308 phosphorylation of AKT and shorter complete remission duration in acute myeloid leukemia patients. Leukemia 25, 1711-1717. 116. Tan, Y., Sun, D., Jiang, W., Klotz-Noack, K., Vashisht, A.A., Wohlschlegel, J., Widschwendter, M., and Spruck, C. (2014). PP2A-B55beta antagonizes cyclin E1 proteolysis and promotes its dysregulation in cancer. Cancer research 74, 2006-2014. 117. Keyomarsi, K., Tucker, S.L., Buchholz, T.A., Callister, M., Ding, Y., Hortobagyi, G.N., Bedrosian, I., Knickerbocker, C., Toyofuku, W., Lowe, M., et al. (2002). Cyclin E and survival in patients with breast cancer. The New England journal of medicine 347, 1566-1575. 118. Meyer, N., and Penn, L.Z. (2008). Reflecting on 25 years with MYC. Nat Rev Cancer 8, 976-990. 119. Welcker, M., Orian, A., Jin, J., Grim, J.E., Harper, J.W., Eisenman, R.N., and Clurman, B.E. (2004). The Fbw7 tumor suppressor regulates glycogen synthase kinase 3 phosphorylation-dependent c-Myc protein degradation. Proc Natl Acad Sci U S A 101, 9085-9090. 120. Yeh, E., Cunningham, M., Arnold, H., Chasse, D., Monteith, T., Ivaldi, G., Hahn, W.C., Stukenberg, P.T., Shenolikar, S., Uchida, T., et al. (2004). A signalling pathway controlling c-Myc degradation that impacts oncogenic transformation of human cells. Nature cell biology 6, 308-318. 121. Arnold, H.K., and Sears, R.C. (2006). Protein phosphatase 2A regulatory subunit B56alpha associates with c-myc and negatively regulates c-myc accumulation. Molecular and cellular biology 26, 2832-2844. 122. Ruvolo, P.P., Clark, W., Mumby, M., Gao, F., and May, W.S. (2002). A functional role for the B56 alpha-subunit of protein phosphatase 2A in ceramide-mediated regulation of Bcl2 phosphorylation status and function. The Journal of biological chemistry 277, 22847-22852. 131 123. Deichmann, M., Polychronidis, M., Wacker, J., Thome, M., and Naher, H. (2001). The protein phosphatase 2A subunit Bgamma gene is identified to be differentially expressed in malignant melanomas by subtractive suppression hybridization. Melanoma research 11, 577-585. 124. Polakis, P. (2000). Wnt signaling and cancer. Genes & development 14, 1837-1851. 125. Low, I.C., Loh, T., Huang, Y., Virshup, D.M., and Pervaiz, S. (2014). Ser70 phosphorylation of Bcl-2 by selective tyrosine nitration of PP2A-B56delta stabilizes its antiapoptotic activity. Blood 124, 2223-2234. 126. Junttila, M.R., Puustinen, P., Niemela, M., Ahola, R., Arnold, H., Bottzauw, T., Ala-aho, R., Nielsen, C., Ivaska, J., Taya, Y., et al. (2007). CIP2A inhibits PP2A in human malignancies. Cell 130, 51-62. 127. Li, M., Makkinje, A., and Damuni, Z. (1996). The myeloid leukemia-associated protein SET is a potent inhibitor of protein phosphatase 2A. The Journal of biological chemistry 271, 11059-11062. 128. Janghorban, M., Farrell, A.S., Allen-Petersen, B.L., Pelz, C., Daniel, C.J., Oddo, J., Langer, E.M., Christensen, D.J., and Sears, R.C. (2014). Targeting c-MYC by antagonizing PP2A inhibitors in breast cancer. Proceedings of the National Academy of Sciences 111, 9157-9162. 129. Farrell, A.S., Allen-Petersen, B., Daniel, C.J., Wang, X., Wang, Z., Rodriguez, S., Impey, S., Oddo, J., Vitek, M.P., Lopez, C., et al. (2014). Targeting inhibitors of the tumor suppressor PP2A for the treatment of pancreatic cancer. Molecular cancer research : MCR 12, 924-939. 130. Calin, G.A., di Iasio, M.G., Caprini, E., Vorechovsky, I., Natali, P.G., Sozzi, G., Croce, C.M., Barbanti-Brodano, G., Russo, G., and Negrini, M. (2000). Low frequency of alterations of the alpha (PPP2R1A) and beta (PPP2R1B) isoforms of the subunit A of the serine-threonine phosphatase 2A in human neoplasms. Oncogene 19, 1191-1195. 131. Chen, W., Arroyo, J.D., Timmons, J.C., Possemato, R., and Hahn, W.C. (2005). Cancer-associated PP2A Aalpha subunits induce functional haploinsufficiency and tumorigenicity. Cancer research 65, 8183-8192. 132. Strack, S., Cribbs, J.T., and Gomez, L. (2004). Critical role for protein phosphatase 2A heterotrimers in mammalian cell survival. The Journal of biological chemistry 279, 47732-47739. 133. Westermarck, J., and Hahn, W.C. (2008). Multiple pathways regulated by the tumor suppressor PP2A in transformation. Trends Mol Med 14, 152-160. 134. Colella, S., Ohgaki, H., Ruediger, R., Yang, F., Nakamura, M., Fujisawa, H., Kleihues, P., and Walter, G. (2001). Reduced expression of the Aalpha subunit of protein phosphatase 2A in human gliomas in the absence of mutations in the Aalpha and Abeta subunit genes. Int J Cancer 93, 798-804. 135. Campbell, I.G., and Manolitsas, T. (1999). Absence of PPP2R1B gene alterations in primary ovarian cancers. Oncogene 18, 6367-6369. 136. Wang, S.S., Esplin, E.D., Li, J.L., Huang, L., Gazdar, A., Minna, J., and Evans, G.A. (1998). Alterations of the PPP2R1B gene in human lung and colon cancer. Science 282, 284-287. 137. Ruediger, R., Pham, H.T., and Walter, G. (2001). Alterations in protein phosphatase 2A subunit interaction in human carcinomas of the lung and colon with mutations in the A beta subunit gene. Oncogene 20, 1892-1899. 132 138. Sablina, A.A., Chen, W., Arroyo, J.D., Corral, L., Hector, M., Bulmer, S.E., DeCaprio, J.A., and Hahn, W.C. (2007). The tumor suppressor PP2A Abeta regulates the RalA GTPase. Cell 129, 969-982. 139. Mwenifumbo, J.C., and Marra, M.A. (2013). Cancer genome-sequencing study design. Nat Rev Genet 14, 321-332. 140. Cancer Genome Atlas Research, N., Weinstein, J.N., Collisson, E.A., Mills, G.B., Shaw, K.R., Ozenberger, B.A., Ellrott, K., Shmulevich, I., Sander, C., and Stuart, J.M. (2013). The Cancer Genome Atlas Pan-Cancer analysis project. Nat Genet 45, 1113-1120. 141. Thomas, R.K., Baker, A.C., Debiasi, R.M., Winckler, W., Laframboise, T., Lin, W.M., Wang, M., Feng, W., Zander, T., MacConaill, L., et al. (2007). High-throughput oncogene mutation profiling in human cancer. Nat Genet 39, 347-351. 142. Doroshow, J.H., and Kummar, S. (2014). Translational research in oncology[mdash]10 years of progress and future prospects. Nature reviews. Clinical oncology 11, 649-662. 143. Aebersold, R., and Cravatt, B.F. (2002). Proteomics – advances, applications and the challenges that remain. Trends in biotechnology 20, s1-s2. 144. Makarov, A. (2000). Electrostatic axially harmonic orbital trapping: a high-performance technique of mass analysis. Analytical chemistry 72, 1156-1162. 145. Perry, R.H., Cooks, R.G., and Noll, R.J. (2008). Orbitrap mass spectrometry: instrumentation, ion motion and applications. Mass spectrometry reviews 27, 661-699. 146. Macek, B., Waanders, L.F., Olsen, J.V., and Mann, M. (2006). Top-down Protein Sequencing and MS3 on a Hybrid Linear Quadrupole Ion Trap-Orbitrap Mass Spectrometer. Molecular & Cellular Proteomics 5, 949-958. 147. Tran, J.C., Zamdborg, L., Ahlf, D.R., Lee, J.E., Catherman, A.D., Durbin, K.R., Tipton, J.D., Vellaichamy, A., Kellie, J.F., Li, M., et al. (2011). Mapping intact protein isoforms in discovery mode using top-down proteomics. Nature 480, 254-258. 148. Thompson, A., Schäfer, J., Kuhn, K., Kienle, S., Schwarz, J., Schmidt, G., Neumann, T., and Hamon, C. (2003). Tandem Mass Tags:   A Novel Quantification Strategy for Comparative Analysis of Complex Protein Mixtures by MS/MS. Analytical chemistry 75, 1895-1904. 149. Zieske, L.R. (2006). A perspective on the use of iTRAQ reagent technology for protein complex and profiling studies. Journal of experimental botany 57, 1501-1508. 150. Ong, S.E., Blagoev, B., Kratchmarova, I., Kristensen, D.B., Steen, H., Pandey, A., and Mann, M. (2002). Stable isotope labeling by amino acids in cell culture, SILAC, as a simple and accurate approach to expression proteomics. Molecular & cellular proteomics : MCP 1, 376-386. 151. Shah, S.P., Kobel, M., Senz, J., Morin, R.D., Clarke, B.A., Wiegand, K.C., Leung, G., Zayed, A., Mehl, E., Kalloger, S.E., et al. (2009). Mutation of FOXL2 in granulosa-cell tumors of the ovary. The New England journal of medicine 360, 2719-2729. 152. Ruediger, R., Hentz, M., Fait, J., Mumby, M., and Walter, G. (1994). Molecular model of the A subunit of protein phosphatase 2A: interaction with other subunits and tumor antigens. Journal of virology 68, 123-129. 153. Ruediger, R., Fields, K., and Walter, G. (1999). Binding specificity of protein phosphatase 2A core enzyme for regulatory B subunits and T antigens. Journal of virology 73, 839-842. 154. Forbes, S.A., Tang, G., Bindal, N., Bamford, S., Dawson, E., Cole, C., Kok, C.Y., Jia, M., Ewing, R., Menzies, A., et al. (2010). COSMIC (the Catalogue of Somatic Mutations 133 in Cancer): a resource to investigate acquired mutations in human cancer. Nucleic Acids Res 38, D652-657. 155. Arnold, K., Bordoli, L., Kopp, J., and Schwede, T. (2006). The SWISS-MODEL workspace: a web-based environment for protein structure homology modelling. Bioinformatics 22, 195-201. 156. Al-Hussaini, M., Stockman, A., Foster, H., and McCluggage, W.G. (2004). WT-1 assists in distinguishing ovarian from uterine serous carcinoma and in distinguishing between serous and endometrioid ovarian carcinoma. Histopathology 44, 109-115. 157. Abeler, V.M., and Kjorstad, K.E. (1990). Serous papillary carcinoma of the endometrium: a histopathological study of 22 cases. Gynecol Oncol 39, 266-271. 158. Reva, B., Antipin, Y., and Sander, C. (2007). Determinants of protein function revealed by combinatorial entropy optimization. Genome Biol 8, R232. 159. Janssens, V., Goris, J., and Van Hoof, C. (2005). PP2A: the expected tumor suppressor. Current opinion in genetics & development 15, 34-41. 160. Bialojan, C., and Takai, A. (1988). Inhibitory effect of a marine-sponge toxin, okadaic acid, on protein phosphatases. Specificity and kinetics. The Biochemical journal 256, 283-290. 161. Kawabe, T. (2004). G2 checkpoint abrogators as anticancer drugs. Molecular cancer therapeutics 3, 513-519. 162. Zhang, W., Yang, J., Liu, Y., Chen, X., Yu, T., Jia, J., and Liu, C. (2009). PR55 alpha, a regulatory subunit of PP2A, specifically regulates PP2A-mediated beta-catenin dephosphorylation. The Journal of biological chemistry 284, 22649-22656. 163. Dozier, C., Bonyadi, M., Baricault, L., Tonasso, L., and Darbon, J.M. (2004). Regulation of Chk2 phosphorylation by interaction with protein phosphatase 2A via its B' regulatory subunit. Biology of the cell / under the auspices of the European Cell Biology Organization 96, 509-517. 164. Woo, M.M., Salamanca, C.M., Symowicz, J., Stack, M.S., Miller, D.M., Leung, P.C., Gilks, C.B., and Auersperg, N. (2008). SV40 early genes induce neoplastic properties in serous borderline ovarian tumor cells. Gynecol Oncol 111, 125-131. 165. Hirschowitz, L., Ganesan, R., and McCluggage, W.G. (2009). WT1, p53 and hormone receptor expression in uterine serous carcinoma. Histopathology 55, 478-482. 166. Watson, I.R., Takahashi, K., Futreal, P.A., and Chin, L. (2013). Emerging patterns of somatic mutations in cancer. Nat Rev Genet 14, 703-718. 167. Soslow, R.A. (2010). Endometrial carcinomas with ambiguous features. Seminars in Diagnostic Pathology 27, 261-273. 168. Sidhu, D., Duggan, M., Clement, P., Ewanowich, C., Arseneau, J., Cesari, M., Kobel, M., and Han, G. (2012). Interobserver Agreement on High-Grade Endometrial Carcinoma and Correlation with ER, p53 and p16 Expression. Modern Pathology 25, 296A-297A. 169. Gilks, C.B., Oliva, E., and Soslow, R. (2011). Poor Interobserver Reproducibility in the Diagnosis of High-Grade Endometrial Carcinoma. In USCAP. (San Antonio, TX, USA). 170. Morin, R.D., Mendez-Lago, M., Mungall, A.J., Goya, R., Mungall, K.L., Corbett, R.D., Johnson, N.A., Severson, T.M., Chiu, R., Field, M., et al. (2011). Frequent mutation of histone-modifying genes in non-Hodgkin lymphoma. Nature 476, 298-303. 171. Li, H., and Durbin, R. (2009). Fast and accurate short read alignment with Burrows-Wheeler transform. Bioinformatics 25, 1754-1760. 134 172. Ding, J., Bashashati, A., Roth, A., Oloumi, A., Tse, K., Zeng, T., Haffari, G., Hirst, M., Marra, M.A., Condon, A., et al. (2012). Feature-based classifiers for somatic mutation detection in tumour-normal paired sequencing data. Bioinformatics 28, 167-175. 173. Forbes, S.A., Bindal, N., Bamford, S., Cole, C., Kok, C.Y., Beare, D., Jia, M., Shepherd, R., Leung, K., Menzies, A., et al. (2011). COSMIC: mining complete cancer genomes in the Catalogue of Somatic Mutations in Cancer. Nucleic Acids Res 39, D945-950. 174. Forshew, T., Murtaza, M., Parkinson, C., Gale, D., Tsui, D.W., Kaper, F., Dawson, S.J., Piskorz, A.M., Jimenez-Linan, M., Bentley, D., et al. (2012). Noninvasive identification and monitoring of cancer mutations by targeted deep sequencing of plasma DNA. Science translational medicine 4, 136ra168. 175. Goodfellow, P.J., Buttin, B.M., Herzog, T.J., Rader, J.S., Gibb, R.K., Swisher, E., Look, K., Walls, K.C., Fan, M.-Y., and Mutch, D.G. (2003). Prevalence of defective DNA mismatch repair and MSH6 mutation in an unselected series of endometrial cancers. Proceedings of the National Academy of Sciences 100, 5908-5913. 176. Tashiro, H., Blazes, M.S., Wu, R., Cho, K.R., Bose, S., Wang, S.I., Li, J., Parsons, R., and Ellenson, L.H. (1997). Mutations in PTEN are frequent in endometrial carcinoma but rare in other common gynecological malignancies. Cancer research 57, 3935-3940. 177. D’Angelo, E., and Prat, J. (2011). Pathology of mixed Müllerian tumours. Best Practice &amp; Research Clinical Obstetrics &amp; Gynaecology 25, 705-718. 178. Soslow, R.A., Bissonnette, J.P., Wilton, A., Ferguson, S.E., Alektiar, K.M., Duska, L.R., and Oliva, E. (2007). Clinicopathologic analysis of 187 high-grade endometrial carcinomas of different histologic subtypes: similar outcomes belie distinctive biologic differences. Am J Surg Pathol 31, 979-987. 179. Kobel, M., Kalloger, S.E., Baker, P.M., Ewanowich, C.A., Arseneau, J., Zherebitskiy, V., Abdulkarim, S., Leung, S., Duggan, M.A., Fontaine, D., et al. (2010). Diagnosis of ovarian carcinoma cell type is highly reproducible: a transcanadian study. Am J Surg Pathol 34, 984-993. 180. Anglesio, M.S., Carey, M.S., Kobel, M., Mackay, H., and Huntsman, D.G. (2011). Clear cell carcinoma of the ovary: a report from the first Ovarian Clear Cell Symposium, June 24th, 2010. Gynecol Oncol 121, 407-415. 181. Alkushi, A., Clarke, B.A., Akbari, M., Makretsov, N., Lim, P., Miller, D., Magliocco, A., Coldman, A., van de Rijn, M., Huntsman, D., et al. (2007). Identification of prognostically relevant and reproducible subsets of endometrial adenocarcinoma based on clustering analysis of immunostaining data. Modern Pathol 20, 1156-1165. 182. Oda, K., Stokoe, D., Taketani, Y., and McCormick, F. (2005). High frequency of coexistent mutations of PIK3CA and PTEN genes in endometrial carcinoma. Cancer research 65, 10669-10673. 183. Voss, M.A., Ganesan, R., Ludeman, L., McCarthy, K., Gornall, R., Schaller, G., Wei, W., and Sundar, S. (2011). Should grade 3 endometrioid endometrial carcinoma be considered a type 2 cancer-A clinical and pathological evaluation. Gynecologic oncology. 184. Zannoni, G., Vellone, V., Arena, V., Prisco, M., Scambia, G., Carbone, A., and Gallo, D. (2010). Does high-grade endometrioid carcinoma (grade 3 FIGO) belong to type I or type II endometrial cancer? A clinical–pathological and immunohistochemical study. Virchows Archiv 457, 27-34. 135 185. Crum, C.P., Lee, Kenneth R., Genest, David R. (2006). Adenocarcinoma, carcinosarcoma and other epithelial tumors of the endometrium. In Diagnostic gynecologic and obstetric pathology, Volume  , 1st Edition, K.R.L. Christopher P. Crum, ed. (Philadelphia, PA: Saunders/Elsevier), p. 1182 p. 186. Catasus, L., Gallardo, A., and Prat, J. (2009). Molecular genetics of endometrial carcinoma.  15, 554-563. 187. Amant, F., Cadron, I., Fuso, L., Berteloot, P., de Jonge, E., Jacomen, G., Van Robaeys, J., Neven, P., Moerman, P., and Vergote, I. (2005). Endometrial carcinosarcomas have a different prognosis and pattern of spread compared to high-risk epithelial endometrial cancer. Gynecologic oncology 98, 274-280. 188. Chiyoda, T., Tsuda, H., Tanaka, H., Kataoka, F., Nomura, H., Nishimura, S., Takano, M., Susumu, N., Saya, H., and Aoki, D. (2012). Expression profiles of carcinosarcoma of the uterine corpus—are these similar to carcinoma or sarcoma? Genes, Chromosomes and Cancer 51, 229-239. 189. Murray, S., Linardou, H., Mountzios, G., Manoloukos, M., Markaki, S., Eleutherakis-Papaiakovou, E., Dimopoulos, M.A., and Papadimitriou, C.A. (2010). Low frequency of somatic mutations in uterine sarcomas: a molecular analysis and review of the literature. Mutat Res 686, 68-73. 190. Liu, L., and Eisenman, R.N. (2012). Regulation of c-Myc Protein Abundance by a Protein Phosphatase 2A-Glycogen Synthase Kinase 3beta-Negative Feedback Pathway. Genes & cancer 3, 23-36. 191. King, B., Trimarchi, T., Reavie, L., Xu, L., Mullenders, J., Ntziachristos, P., Aranda-Orgilles, B., Perez-Garcia, A., Shi, J., Vakoc, C., et al. (2013). The ubiquitin ligase FBXW7 modulates leukemia-initiating cell activity by regulating MYC stability. Cell 153, 1552-1566. 192. Prowse, A.H., Manek, S., Varma, R., Liu, J., Godwin, A.K., Maher, E.R., Tomlinson, I.P.M., and Kennedy, S.H. (2006). Molecular genetic evidence that endometriosis is a precursor of ovarian cancer. International Journal of Cancer 119, 556-562. 193. Samarnthai, N., Hall, K., and Yeh, I.T. (2010). Molecular profiling of endometrial malignancies. Obstet Gynecol Int 2010, 162363. 194. Benjamini, Y., and Hochberg, Y. (1995). Controlling the False Discovery Rate - a Practical and Powerful Approach to Multiple Testing. J Roy Stat Soc B Met 57, 289-300. 195. Stuart, G.C.E., Kitchener, H., Bacon, M., duBois, A., Friedlander, M., Ledermann, J., Marth, C., Thigpen, T., Trimble, E., and OCCC, t.O.C.C.C. (2011). 2010 Gynecologic Cancer InterGroup (GCIG) Consensus Statement on Clinical Trials in Ovarian Cancer Report From the Fourth Ovarian Cancer Consensus Conference. Int J Gynecol Cancer 21, 750-755. 196. Geyer, J.T., Lopez-Garcia, M.A., Sanchez-Estevez, C., Sarrio, D., Moreno-Bueno, G., Franceschetti, I., Palacios, J., and Oliva, E. (2009). Pathogenetic pathways in ovarian endometrioid adenocarcinoma: a molecular study of 29 cases. Am J Surg Pathol 33, 1157-1163. 197. McCluggage, W.G. (2011). Morphological subtypes of ovarian carcinoma: a review with emphasis on new developments and pathogenesis. Pathology 43, 420-432. 198. Shedden, K.A., Kshirsagar, M.P., Schwartz, D.R., Wu, R., Yu, H., Misek, D.E., Hanash, S., Katabuchi, H., Ellenson, L.H., Fearon, E.R., et al. (2005). Histologic type, organ of origin, and Wnt pathway status: effect on gene expression in ovarian and uterine 136 carcinomas. Clinical cancer research : an official journal of the American Association for Cancer Research 11, 2123-2131. 199. Palacios, J., and Gamallo, C. (1998). Mutations in the beta-catenin gene (CTNNB1) in endometrioid ovarian carcinomas. Cancer research 58, 1344-1347. 200. Wu, R., Zhai, Y., Fearon, E.R., and Cho, K.R. (2001). Diverse mechanisms of beta-catenin deregulation in ovarian endometrioid adenocarcinomas. Cancer research 61, 8247-8255. 201. Wright, K., Wilson, P., Morland, S., Campbell, I., Walsh, M., Hurst, T., Ward, B., Cummings, M., and Chenevix-Trench, G. (1999). beta-catenin mutation and expression analysis in ovarian cancer: exon 3 mutations and nuclear translocation in 16% of endometrioid tumours. Int J Cancer 82, 625-629. 202. Saegusa, M., and Okayasu, I. (2001). Frequent nuclear beta-catenin accumulation and associated mutations in endometrioid-type endometrial and ovarian carcinomas with squamous differentiation. The Journal of pathology 194, 59-67. 203. Sagae, S., Kobayashi, K., Nishioka, Y., Sugimura, M., Ishioka, S., Nagata, M., Terasawa, K., Tokino, T., and Kudo, R. (1999). Mutational analysis of beta-catenin gene in Japanese ovarian carcinomas: frequent mutations in endometrioid carcinomas. Jpn J Cancer Res 90, 510-515. 204. Moreno-Bueno, G., Gamallo, C., Perez-Gallego, L., de Mora, J.C., Suarez, A., and Palacios, J. (2001). beta-Catenin expression pattern, beta-catenin gene mutations, and microsatellite instability in endometrioid ovarian carcinomas and synchronous endometrial carcinomas. Diagn Mol Pathol 10, 116-122. 205. Obata, K., Morland, S.J., Watson, R.H., Hitchcock, A., Chenevix-Trench, G., Thomas, E.J., and Campbell, I.G. (1998). Frequent PTEN/MMAC Mutations in Endometrioid but not Serous or Mucinous Epithelial Ovarian Tumors. Cancer research 58, 2095-2097. 206. Shih Ie, M., Panuganti, P.K., Kuo, K.T., Mao, T.L., Kuhn, E., Jones, S., Velculescu, V.E., Kurman, R.J., and Wang, T.L. (2011). Somatic mutations of PPP2R1A in ovarian and uterine carcinomas. The American journal of pathology 178, 1442-1447. 207. Siiteri, P.K., Schwarz, B.E., and MacDonald, P.C. (1974). Estrogen receptors and the estrone hypothesis in relation to endometrial and breast cancer. Gynecologic oncology 2, 228-238. 208. Key, T.J.A., and Pike, M.C. (1988). The Dose-Effect Relationship between Unopposed Estrogens and Endometrial Mitotic Rate - Its Central Role in Explaining and Predicting Endometrial Cancer Risk. Brit J Cancer 57, 205-212. 209. Yamaguchi, K., Mandai, M., Toyokuni, S., Hamanishi, J., Higuchi, T., Takakura, K., and Fujii, S. (2008). Contents of endometriotic cysts, especially the high concentration of free iron, are a possible cause of carcinogenesis in the cysts through the iron-induced persistent oxidative stress. Clinical Cancer Research 14, 32-40. 210. Van Langendonckt, A., Casanas-Roux, F., Dolmans, M.M., and Donnez, J. (2002). Potential involvement of hemoglobin and heme in the pathogenesis of peritoneal endometriosis. Fertil Steril 77, 561-570. 211. Kobayashi, H., Kajiwara, H., Kanayama, S., Yamada, Y., Furukawa, N., Noguchi, T., Haruta, S., Yoshida, S., Sakata, M., Sado, T., et al. (2009). Molecular pathogenesis of endometriosis-associated clear cell carcinoma of the ovary (Review). Oncol Rep 22, 233-240. 137 212. Fleming, J.S., Beaugie, C.R., Haviv, I., Chenevix-Trench, G., and Tan, O.L. (2006). Incessant ovulation, inflammation and epithelial ovarian carcinogenesis: revisiting old hypotheses. Mol Cell Endocrinol 247, 4-21. 213. Korner, M., Burckhardt, E., and Mazzucchelli, L. (2006). Higher frequency of chromosomal aberrations in ovarian endometriosis compared to extragonadal endometriosis: A possible link to endometrioid adenocarcinoma. Modern pathology : an official journal of the United States and Canadian Academy of Pathology, Inc 19, 1615-1623. 214. Vestergaard, A.L., Thorup, K., Knudsen, U.B., Munk, T., Rosbach, H., Poulsen, J.B., Guldberg, P., and Martensen, P.M. (2011). Oncogenic events associated with endometrial and ovarian cancers are rare in endometriosis. Molecular Human Reproduction 17, 758-761. 215. Lawrenson, K., Lee, N., Torres, H.A., Lee, J.M., Brueggmann, D., Rao, P.N., Noushmehr, H., and Gayther, S.A. (2014). Src as a novel therapeutic target for endometriosis. Gynecol Oncol 135, 100-107. 216. Su, L.K., Vogelstein, B., and Kinzler, K.W. (1993). Association of the APC tumor suppressor protein with catenins. Science 262, 1734-1737. 217. Rubinfeld, B., Souza, B., Albert, I., Muller, O., Chamberlain, S.H., Masiarz, F.R., Munemitsu, S., and Polakis, P. (1993). Association of the APC gene product with beta-catenin. Science 262, 1731-1734. 218. Klaus, A., and Birchmeier, W. (2008). Wnt signalling and its impact on development and cancer. Nature reviews. Cancer 8, 387-398. 219. Samuels, Y., Wang, Z., Bardelli, A., Silliman, N., Ptak, J., Szabo, S., Yan, H., Gazdar, A., Powell, S.M., Riggins, G.J., et al. (2004). High frequency of mutations of the PIK3CA gene in human cancers. Science 304, 554. 220. Cully, M., You, H., Levine, A.J., and Mak, T.W. (2006). Beyond PTEN mutations: the PI3K pathway as an integrator of multiple inputs during tumorigenesis. Nature reviews. Cancer 6, 184-192. 221. McCubrey, J.A., Steelman, L.S., Chappell, W.H., Abrams, S.L., Montalto, G., Cervello, M., Nicoletti, F., Fagone, P., Malaponte, G., Mazzarino, M.C., et al. (2012). Mutations and Deregulation of Ras/Raf/MEK/ERK and PI3K/PTEN/Akt/mTOR Cascades. Oncotarget 3, 954-987. 222. Kohli, M., Rago, C., Lengauer, C., Kinzler, K.W., and Vogelstein, B. (2004). Facile methods for generating human somatic cell gene knockouts using recombinant adeno-associated viruses. Nucleic Acids Res 32, e3. 223. Hannon, G.J. (2002). RNA interference. Nature 418, 244-251. 224. Bridge, A.J., Pebernard, S., Ducraux, A., Nicoulaz, A.L., and Iggo, R. (2003). Induction of an interferon response by RNAi vectors in mammalian cells. Nat Genet 34, 263-264. 225. Rago, C., Vogelstein, B., and Bunz, F. (2007). Genetic knockouts and knockins in human somatic cells. Nat Protoc 2, 2734-2746. 226. Topaloglu, O., Hurley, P.J., Yildirim, O., Civin, C.I., and Bunz, F. (2005). Improved methods for the generation of human gene knockout and knockin cell lines. Nucleic Acids Res 33, e158. 227. Hanson, K.D., and Sedivy, J.M. (1995). Analysis of biological selections for high-efficiency gene targeting. Molecular and cellular biology 15, 45-51. 138 228. Lengauer, C., Kinzler, K.W., and Vogelstein, B. (1998). Genetic instabilities in human cancers. Nature 396, 643-649. 229. Kolodner, R. (1996). Biochemistry and genetics of eukaryotic mismatch repair. Genes & development 10, 1433-1442. 230. Samuels, Y., Diaz Jr, L.A., Schmidt-Kittler, O., Cummins, J.M., DeLong, L., Cheong, I., Rago, C., Huso, D.L., Lengauer, C., Kinzler, K.W., et al. (2005). Mutant PIK3CA promotes cell growth and invasion of human cancer cells. Cancer Cell 7, 561-573. 231. Kuramoto, H., Tamura, S., and Notake, Y. (1972). Establishment of a cell line of human endometrial adenocarcinoma in vitro. American journal of obstetrics and gynecology 114, 1012-1019. 232. Risinger, J.I., Umar, A., Barrett, J.C., and Kunkel, T.A. (1995). A hPMS2 Mutant Cell Line Is Defective in Strand-specific Mismatch Repair. Journal of Biological Chemistry 270, 18183-18186. 233. Karst, A.M., Levanon, K., and Drapkin, R. (2011). Modeling high-grade serous ovarian carcinogenesis from the fallopian tube. Proc Natl Acad Sci U S A 108, 7547-7552. 234. Ionov, Y., Yamamoto, H., Krajewski, S., Reed, J.C., and Perucho, M. (2000). Mutational inactivation of the proapoptotic gene BAX confers selective advantage during tumor clonal evolution. Proc Natl Acad Sci U S A 97, 10872-10877. 235. Ventura, A., Kirsch, D.G., McLaughlin, M.E., Tuveson, D.A., Grimm, J., Lintault, L., Newman, J., Reczek, E.E., Weissleder, R., and Jacks, T. (2007). Restoration of p53 function leads to tumour regression in vivo. Nature 445, 661-665. 236. Meyers, E.N., Lewandoski, M., and Martin, G.R. (1998). An Fgf8 mutant allelic series generated by Cre- and Flp-mediated recombination. Nat Genet 18, 136-141. 237. Olson, E.N., Arnold, H.H., Rigby, P.W., and Wold, B.J. (1996). Know your neighbors: three phenotypes in null mutants of the myogenic bHLH gene MRF4. Cell 85, 1-4. 238. Porteus, M.H., and Baltimore, D. (2003). Chimeric nucleases stimulate gene targeting in human cells. Science 300, 763. 239. Porteus, M.H., and Carroll, D. (2005). Gene targeting using zinc finger nucleases. Nature biotechnology 23, 967-973. 240. Christian, M., Cermak, T., Doyle, E.L., Schmidt, C., Zhang, F., Hummel, A., Bogdanove, A.J., and Voytas, D.F. (2010). Targeting DNA double-strand breaks with TAL effector nucleases. Genetics 186, 757-761. 241. Cong, L., Ran, F.A., Cox, D., Lin, S., Barretto, R., Habib, N., Hsu, P.D., Wu, X., Jiang, W., Marraffini, L.A., et al. (2013). Multiplex Genome Engineering Using CRISPR/Cas Systems. Science 339, 819-823. 242. Mali, P., Yang, L., Esvelt, K.M., Aach, J., Guell, M., DiCarlo, J.E., Norville, J.E., and Church, G.M. (2013). RNA-Guided Human Genome Engineering via Cas9. Science 339, 823-826. 243. Gaj, T., Gersbach, C.A., and Barbas, C.F., 3rd (2013). ZFN, TALEN, and CRISPR/Cas-based methods for genome engineering. Trends in biotechnology 31, 397-405. 244. Wood, A.J., Lo, T.W., Zeitler, B., Pickle, C.S., Ralston, E.J., Lee, A.H., Amora, R., Miller, J.C., Leung, E., Meng, X., et al. (2011). Targeted genome editing across species using ZFNs and TALENs. Science 333, 307. 245. Gaj, T., Guo, J., Kato, Y., Sirk, S.J., and Barbas, C.F., 3rd (2012). Targeted gene knockout by direct delivery of zinc-finger nuclease proteins. Nature methods 9, 805-807. 139 246. Brouns, S.J.J., Jore, M.M., Lundgren, M., Westra, E.R., Slijkhuis, R.J.H., Snijders, A.P.L., Dickman, M.J., Makarova, K.S., Koonin, E.V., and van der Oost, J. (2008). Small CRISPR RNAs Guide Antiviral Defense in Prokaryotes. Science 321, 960-964. 247. Shalem, O., Sanjana, N.E., Hartenian, E., Shi, X., Scott, D.A., Mikkelsen, T.S., Heckl, D., Ebert, B.L., Root, D.E., Doench, J.G., et al. (2014). Genome-scale CRISPR-Cas9 knockout screening in human cells. Science 343, 84-87. 248. Canver, M.C., Bauer, D.E., Dass, A., Yien, Y.Y., Chung, J., Masuda, T., Maeda, T., Paw, B.H., and Orkin, S.H. (2014). Characterization of Genomic Deletion Efficiency Mediated by Clustered Regularly Interspaced Palindromic Repeats (CRISPR)/Cas9 Nuclease System in Mammalian Cells