Proteomics Investigation of Breast Cancer Biomarkers in Urine and Blood for Diagnosis and Monitoring ______

PROTEOMICS INVESTIGATION OF BREAST CANCER BIOMARKERS IN URINE AND BLOOD FOR DIAGNOSIS AND MONITORING ______

Julia Beretov BSc. Dip Med Sc., Dip Edu

A thesis submitted in fulfilment of the requirements for the degree of Doctor of Philosophy

Faculty of Medicine St George and Sutherland Clinical School

March 2016

ORIGINALITY STATEMENT

I hereby declare that this submission is my own work and to the best of my knowledge it contains no materials previously published or written by another person, or substantial proportions of material which have been accepted for the award of any other degree or diploma at UNSW or any other educational institution, except where due acknowledgement is made in the thesis. Any contribution made to the research by others, with whom I have worked at UNSW or elsewhere, is explicitly acknowledged in the thesis. I also declare that the intellectual content of this thesis is the product of my own work, except to the extent that assistance from others in the project's design and conception or in style, presentation and linguistic expression is acknowledged.

Signed ……………………………………………...... Date ……………………………………………......

‘I hereby grant the University of New South Wales or its agents the right to archive and to make available my thesis or dissertation in whole or part in the University libraries in all forms of media, now or here after known, subject to the provisions of the Copyright Act 1968. I retain all proprietary rights, such as patent rights. I also retain the right to use in future works (such as articles or books) all or part of this thesis or dissertation. I also authorise University Microfilms to use the 350 word abstract of my thesis in Dissertation Abstract International (this is applicable to doctoral theses only). I have either used no substantial portions of copyright material in my thesis or I have obtained permission to use copyright material; where permission has not been granted I have applied/will apply for a partial restriction of the digital copy of my thesis or dissertation.'

Signed ……………………………………………......

Date ……………………………………………......

AUTHENTICITY STATEMENT

‘I certify that the Library deposit digital copy is a direct equivalent of the final officially approved version of my thesis. No emendation of content has occurred and if there are any minor variations in formatting, they are the result of the conversion to digital format.’

Signed ……………………………………………......

Date ……………………………………………...... ABSTRACT

Breast cancer is a major world-wide health problem and is the most commonly diagnosed cancer amongst women in Australia and the world. Although the survival of patients has increased over the last two decades, due to improved screening programs and postoperative adjuvant systemic therapies (hormone therapy and chemotherapy), patients undergo aggressive treatment and many die from metastatic relapse. Whilst screening for early breast cancer detection using mammography and ultrasound is successful, it may provide a false negative result because of the density or architecture of breast tissue. Additionally, it cannot make a clear distinction of benign breast disease from malignancy. Furthermore, if breast cancer is detected at an early stage, when treatment is likely to be more effective, then more lives will be saved.

Breast cancer is a heterogeneous disease, composed of distinct molecular subtypes associated with different clinical outcomes, and would benefit from the development of new disease-specific biomarkers. Current clinical and pathological parameters are not able to monitor breast cancer progression or accurately predict its prognosis. Presently, there are no biomarkers that can be used for early diagnosis and therefore there is an urgent need to identify novel breast cancer biomarkers to improve the early detection and monitor progression. The application of proteomics and high- throughput methods based on mass spectrometry (MS) of peptide or protein mixtures provides a large number of individual proteins. Advances in proteomics technology have allowed us to dig deeper into the human proteome, provide a new insight into cancer biology and allow for the discovery of novel biomarkers.

Therefore, the main research objectives of this thesis were to conduct scientific investigation to 1) analyse biological samples including urine and blood, from breast cancer patients and control subjects using proteomics; 2) identify novel proteins or a panel of proteins which are associated with the presence of disease; 3) validate the identified potential biomarkers from urine and blood in breast cancer cell lines and primary breast cancer tissues for diagnosis. iii

After successfully developing a standardised method for urine protein extraction and precipitation using liquid chromatography-tandem mass spectrometry (LC-MS/MS) analysis (detailed in Chapter 3), this method was applied to additional urine samples from breast cancer patients (Chapter 4). Proteomic analysis of urine to identify breast cancer biomarker candidates using a label-free LC-MS/MS approach, revealed 59 urinary proteins that were significantly different in breast cancer patients compared to the normal healthy control subjects (p<0.05, fold change >3). Thirty-six urinary proteins were exclusively found in specific breast cancer stages, with 24 increasing and 12 decreasing in their abundance. Preliminary validation of 3 potential markers ECM1, MAST4 and filaggrin was performed in breast cancer cell lines by Western blotting. One potential marker, MAST4, was further validated in human breast cancer tissues as well as human breast cancer urine samples with immunohistochemistry (IHC) and Western blotting (WB), respectively.

The importance of the process of blood collection itself is critical to the accuracy and reproducibility of quantitative biomarker analysis (detailed in Chapter 5). Applying a label-free LC-MS/MS approach to analyse the blood protein abundances in four different blood tubes (2 serum and 2 plasma collection tubes) revealed that there was up to 70% variation between the tubes. The variation in proteins and individual protein abundances detected across the different tubes identified that the richest source was found using the serum gold BD Vacutainer® SST™ gel tube and plasma purple BD Vacutainer ® EDTA tubes. Additionally, applying fractionation it was apparent that the most informative data was obtained from the low to medium (3- 50kDa) fraction.

Comparative profiling of the LC-MC/MC data (Chapter 6) identified 50 proteins modulated in the different breast cancer stages (p<0.05 and fold change ⩾3.0) and amongst these, a panel of 17 unique up-regulated stage specific breast cancer proteins were identified. Using this information, five candidates were identified: Clusterin, Insulin-like growth factor-binding protein 3 (IGFBP3), Leucine-rich alpha-2- glycoprotein (LRG-1), S100-A6 and Vitronectin. These putative markers were validated with WB and IHC in breast cancer cell lines, along with serum, plasma and iv tumour samples from patients and healthy subjects. Serum and plasma samples were mined for new biomarkers using a proteomics approach, which highlighted the proteins which are significantly associated with breast cancer. This study has demonstrated that the expression of these selected proteins could be further developed into clinically relevant diagnostic biomarkers, capable of discriminating patients with breast cancer from healthy individuals.

The results from my research have identified several important proteins in urine and blood for the detection, monitoring invasive progression, prediction of progression and therapeutic management of breast cancer. This could have a major impact on the prognosis of breast cancer for the tens of thousands of women who succumb to the disease each year.

ACKNOWLEDGMENTS

“Happiness is a way of life, not a destiny”. “We are the master of our destiny. We are responsible for our own happiness” William Ernest Henley.

This thesis represents the culmination of seven years of part-time doctoral study yet only contains a fraction of the amount of work performed during this time and certainly reflects only a very small portion of my personal growth during this period. The time has come to officially thank all the people who have helped me through the whole PhD journey, which was only possible due to the work and support of my family, friends and colleagues that I have had the opportunity to thank here. Everybody surrounding me was constantly reminding me to see the “big picture” and tackle the real issue: let’s see what we can find in breast cancer patients to help identify the disease early.

Firstly, I would like to thank all my supervisors: Associate Professor Yong Li, Dr Valerie Wasinger and Associate Professor Peter Graham. Thank you for giving me this opportunity to take on this journey, supporting me all the way and reviewing my thesis. I would like to express my utmost gratitude to my supervisor Associate Professor Yong Li, for giving me the opportunity to embark on this journey. Whose support, encouragement and help has been invaluable for the thesis and for my development as a researcher at large. By the time I came under Yong Li’s supervision, he was already recognised for his skills and patience and he never failed to show these great qualities. He reviewed all my papers and this thesis, with enormous dedication, effort and patience and worked hard to ensure that my work was progressing at every vi stage. Thank you for calming me down when things got too much and never losing faith.

The same sincere acknowledgement goes to my co-supervisor Associate Professor Peter Graham, for always being available for discussion and committed to the cause. You are an exceptional man with great knowledge and have helped me all the way. Thank you for the financial support and providing valuable advice and guidance at key points in the project on experimental design and clinical associations for my studies.

I thank Dr Valerie Wasinger, an expert on proteomics. You shared all your knowledge from LC-MS to statistical analysis, brained stormed about experiments and data and were always there when I needed help and you de-stressed me with your cheerful attitude. I have learnt so much from you, and I am truly grateful that you guided me and we became friends. To the kind and great scientists from Bioanalytical Mass Spectrometry Facility, UNSW, A/Prof Mark Raftery and the team, thank you for your kindness and full support and for allowing me to work with your team.

Also, I wish to acknowledge the funding support from the St George Cancer Care Centre Research Trust Fund for this study. A huge thanks also goes to my colleagues in the Cancer Research Lab, Cancer Care Centre, and St George Hospital, Sydney; Jie Ni (Kevin), Jingli Hao (Helen), Lei Chang (Nancy), Junli Deng (Cyndi) and Duojia Wu for travelling the journey with me through the good times and bad. I cannot express enough my appreciation for your technical assistance and supporting me with knowledge and cups of tea.

We also appreciate the support from Dr Peter Schwartz at St George Private Hospital Sydney, Australia in providing us with patients for urine and blood collection for the biomarkers patients. Thank you to all the volunteers who provided the healthy control samples. St George and Sutherland Clinical School staff, especially Ms Val Reid and Dr Ashish Diwan thank you for your support through the good times and the bad. To the scientists and Pathologists in the Department of Anatomical Pathology St

vii

George Hospital (SEALS) Kogarah, who are more like friends than colleagues: your care, love and friendship were all a source of strength for me to finish this journey. Thank you.

Last but not least, I would like to thank my beloved family who made it all possible. To my parents Slobodanka and George, my role models, for you are both generous, hardworking and strong willed and always make me feel that anything is possible. To my beautiful sister Mary and nephew Dylan, my inspiration. This journey has been hard in more ways than just my PhD. Thank you so much for accompanying me all the while, laughing and crying together, helping me finalising my thesis and cheering me up. My very special and heartfelt thanks to my three beautiful children David, William and Georgia who have no idea what a normal mum is. They fill my heart with joy every day of my life and made it possible for me to enjoy doing my research. I watched them grow up from toddlers to teenagers and adults, they truly made me feel like I could have my cake and eat it too. I thank them for their patience, humour and endless love while I enjoyed the privilege of this journey.

A very wise man would often say to me success is 10% inspiration and 90% perspiration and he was right. To my dearest husband, Ewan, you are truly the most amazing person I have ever met. Thank you for your love and care, being my rock with consistent support and encouragement and for being the best thing to ever happen to me. Because of your understanding and support, I was able to concentrate on my study for all these years and you have helped me all the way. I thank you and love you very much. This PhD was not just my sole work.

This PhD was a group effort and it showed me how important and how much people can achieve with the support, love and care from all around. It has become one of my most valuable memories. Thank you all very much. I dedicate this work to all the women and their families who have suffered with breast cancer. “I too have travelled this path,” my heart is with you all always.

viii

TABLE OF CONTENTS

Originality Statement...... ii

Abstract ...... iii

Acknowledgments ...... vi

Table of contents ...... ix

List of Figures ...... xviii

List of Tables ...... xxii

List of Abbreviations ...... xxiv

List of Publications and presentations ...... xxvii

Journal articles associated with my PhD Study ...... xxvii

Conference oral presentations ...... xxvii

Conference poster presentations ...... xxviii

Other publications as a co-author ...... xxix

Other Conference poster presentations as a co-author ...... xxx

Hypothesis and Aims ...... 1

1 Introduction ...... 3

Introduction to breast cancer ...... 3 Epidemiology...... 4 Worldwide Incidence ...... 4 National Incidence ...... 6 Aetiology ...... 7

Pathology of Breast cancer ...... 11 Cell of origin ...... 11

Morphology of normal breast ...... 17 Histological morphology ...... 18 Normal Breast ...... 18

Benign Breast Disease and Precursor Lesions ...... 19 Breast Cancer ...... 21

Molecular classification of breast cancer & current biomarkers...... 24

Clinical Aspects ...... 28 Clinical presentation ...... 28 Population Screening ...... 28 Mammography ...... 29 Ultrasound ...... 29 Clinical Staging ...... 31 Primary treatment ...... 33 Conservative Surgery ...... 33 Mastectomy ...... 33 Adjuvant and Neoadjuvant Therapy for Breast Cancer ...... 33 Radiotherapy ...... 34 Hormone Therapy ...... 35 Chemotherapy ...... 35 Targeted Therapy for HER2 positive cancer...... 36 Summary of Clinical Aspects ...... 36

Current proteomics studies for breast cancer biomarker ...... 37 What is a Biomarker?...... 37 What is proteomics? ...... 38 Proteomics applications in breast cancer ...... 40

Key proteomic techniques in biomarker research ...... 41 Gel-based techniques ...... 42 One dimensional gel electrophoresis (1DGE) ...... 44 Two-dimensional gel electrophoresis (2DGE) ...... 44 Two-dimensional Difference Gel Electrophoresis ...... 46 Mass spectrometry based techniques in urine biomarker research ...... 47 Label free MS quantitation ...... 48 MALDI-TOF-MS ...... 49

SELDI-TOF ...... 50 Capillary electrophoresis–mass spectrometry...... 52 Liquid chromatography mass spectrometry ...... 53 MS/MS Data to Identified Proteins ...... 56 Global identification and quantification of proteins ...... 57 Peptide analysis using MS technology ...... 60 Multiple reaction monitoring ...... 60 Label-based MS quantitation ...... 62 Isotope Coded Affinity Tag ...... 63 Isobaric tag (iTRAQ) ...... 64 SILAC ...... 65

Proteomics techniques and urinary biomarkers in breast cancer ...... 65 Urine as a potential source for biomarkers ...... 66 Advantages of urine as breast cancer biomarker ...... 67 Urine collection, storage and sample preparation ...... 68 Factors affecting urine collection and handling...... 68 Summary of urine collection & handling ...... 70 Pooling urine samples for protein analysis ...... 71 De-salting urine samples and protein purification ...... 71 Precipitation and concentration of urine ...... 72 Challenges in urine biomarker research ...... 74 Potential breast cancer biomarkers in urine ...... 75

Biomarkers in Breast Cancer ...... 79 Blood biomarkers for breast cancer ...... 81 Serum & Plasma proteomics ...... 84 Plasma Vs Serum ...... 85 Serum and Plasma collection ...... 87 Proteomics information in plasma and serum...... 88 Immunodepletion ...... 89

Mass spectrometry and breast cancer ...... 91 Epigenetics and breast cancer...... 91 Biomarkers in tissue, nipple aspirate, saliva and tears ...... 92 xi

Blood Profiling ...... 93 Protein Expression Patterns in Blood ...... 93 Blood proteomics data bases and internet sites ...... 105

Summary of literature review ...... 106

Thesis Aims ...... 107

2 General materials, method and equipment ...... 109

Ethics Approval ...... 109

Materials...... 110 Solutions and supplier ...... 110 Preparation of buffers and reagents ...... 115 Breast cancer cell lines ...... 118 Primary Antibodies ...... 119

Sample collection ...... 122 Urine samples collection ...... 122 Plasma and serum sample collection ...... 123

Proteomics Methods ...... 125 Urine preparation for proteomics analysis ...... 125 Acetone precipitation of proteins ...... 125 Tri-chloroacetic acid (TCA) protein precipitation ...... 127 Ultra- filtration method ...... 127 Glyco amino glycan (GAG) precipitation ...... 127 Sonication-“cell shearing” ...... 128 Solubilising the protein pellet ...... 129 Protein quantitation (2D Quant) ...... 129 Tris/glycine SDS-PAGE ...... 130 Tris/glycine SDS running buffer ...... 130 Silver staining ...... 130 Coomassie Blue R250 staining ...... 131 Protein clean up and digestion ...... 131 In-solution digestion...... 132 C18 STAGE tip purification for LC-MS/MS analysis ...... 133 xii

Blood preparation for proteomics analysis ...... 134 Fractionation ...... 134 Acetone precipitation of serum and plasma ...... 136 LC-MS/MS analysis on protein samples ...... 136 Progenesis LC-MS/MS statistical analysis ...... 137 Reference run ...... 138 Peak picking parameters...... 139 Peptide statistics ...... 140 Peptide identification ...... 141 Database Searching and Validation...... 142 Ingenuity pathways analysis ...... 144

Validation Methods ...... 145 Cell culture ...... 145 Breast cancer cell lines and cell culture ...... 145 Dissociation and subculturing of cells ...... 146 Cell viability using Trypan Blue ...... 146 Cell preservation/cell thawing ...... 147 Frozen cell thawing ...... 147 Protein extraction ...... 147 Protein quantification (BCA assay) ...... 148 Western blot analysis ...... 149 Protein extraction and quantification...... 149 Western blot analysis of protein extracts ...... 149 Immunohistochemistry ...... 151 Paraffin sections ...... 151 Immuno-staining procedure ...... 152 Assessment of immunostaining results ...... 153 Tissue microarrays ...... 153 Statistical analysis ...... 154

3 Standardised urine preparation protocol for lc-ms/ms ...... 156

Introduction to urinary proteome analysis ...... 156

xiii

Urine preparation for protein analysis Method ...... 157 Urine collection protocol ...... 157 Protein extraction and precipitation techniques ...... 158 Protein separation and examination ...... 162 SDS PAGE ...... 162 Protein desalting ...... 162 Proteomics LC-MS/MS analysis ...... 163

Results of urine precipitation study ...... 163 Urinary proteins identified with LC-MS/MS ...... 163 Combination approach with acetone, then TCA at HSC ...... 167 Urinary proteins in breast cancer and control samples at high speed centrifugation ...... 169 Precipitation with organic solvent and acid alone ...... 169 Other precipitation methods ...... 169

Discussion ...... 170

4 Urine breast cancer biomarker candidates ...... 174

Introduction ...... 174

Materials and Methods ...... 176 Urine sample collection and processing ...... 176 Urine sample collection ...... 176 Urine proteins precipitation ...... 178 Urine sample protein clean-up and digestion ...... 178 Trypsin digestion ...... 178 Protein sample desalting and purification ...... 179 LC-MS/MS analysis of urine sample ...... 179 Label-free LC-MS quantitative profiling ...... 180 Generation of the heat map ...... 180 Characteristics of breast cancer cell lines ...... 181 Western blot analysis ...... 181 Immunohistochemistry staining ...... 183 Assessment of immunostaining ...... 184

xiv

Results and Discussion ...... 184 Circulating urinary markers in breast cancer ...... 184 Classification of identified urine proteins ...... 187 Urine protein distribution in breast cancer patients ...... 187 Urine protein distribution in benign disease patients ...... 205 Interaction networks of human urine proteins ...... 206 Validation of the identified potential urine markers in breast cancer cell lines ...... 210 Preliminary validation of identified potential markers in human primary breast cancer tissues...... 211 Validation of potential marker MAST4 in individual human breast cancer urine samples...... 211

Conclusions ...... 216

5 Evaluation of blood collection tubes using lc-ms/ms: serum vs plasma for breast cancer proteomic biomarker studies ...... 220

Introduction ...... 220

Materials and methods ...... 223 Patient population ...... 223 Blood collection ...... 223 Fractionation ...... 224 Blood fraction C18 clean-up and Trypsin digestion ...... 225 LC-MS/MS analysis ...... 225 Data processing and analysis ...... 226

Results and discussion ...... 226 Prevalence of varied protein expression ...... 226 Percentage variation between tubes ...... 229 Direct protein comparisons in 3-50kDa fraction ...... 234

Blood tubes results 0-3kDa ...... 247

Conclusion ...... 250

6 A panel of novel serum and plasma protein revealed by lc-ms/ms in breast cancer...... 254

Introduction ...... 254

Materials and methods ...... 256 Study design and ethics ...... 256 Sample blood collection protocol ...... 257 Blood samples purification, concentration and digestion ...... 257 Purification of protein extract ...... 258 Protein identification by LC-MS/MS analysis ...... 259 Label-free LC-MS serum-plasma profiling and data analysis ... 259 Cell lines and cell culture...... 259 Immunological confirmation of serum protein markers by Western blotting ...... 260 Immunohistochemistry staining and analysis ...... 261 Assessment of immunostaining ...... 261

Results ...... 262 Blood tube comparison fold change ratio ...... 262 Progenesis PCA analysis of the blood samples ...... 263 Normalised abundances in 3-50kDa serum...... 267 Heat map of diseases and bio functions ...... 272 Canonical pathway analysis and heat map ...... 274 Heat map of disease and bio function ...... 275 Validation of up-regulated proteins in blood and tissues ...... 278 Western blot in breast cancer cell lines ...... 278 Western blot in serum and plasma ...... 280 Immunohistochemistry analysis of human breast carcinomas and normal breast tissue ...... 282 Prognostic value of the serum and plasma biomarker panel ... 284

Discussion ...... 286

Conclusion ...... 290

7 General discussion and future directions ...... 292 xvi

Thesis summary ...... 292

Conclusions and future perspective ...... 294

8 References ...... 297

xvii

LIST OF FIGURES

Figure 1-1. Estimated new cancer cases and deaths worldwide...... 4

Figure 1-2. Bar chart of 5-year prevalence of breast cancer...... 5

Figure 1-3. Breast Cancer Susceptibility Loci and Genes...... 10

Figure 1-4. Hypothetical models of cell division...... 12

Figure 1-5. Normal cellular structure of breast tissue...... 14

Figure 1-6. Hypothetical models explaining breast tumour subtypes...... 15

Figure 1-7. Hypothetical model of breast tumour progression...... 16

Figure 1-8. Diagram of normal breast tissue...... 17

Figure 1-9. Morphology images of normal breast tissue...... 18

Figure 1-10. Microscopic images of benign breast disease...... 19

Figure 1-11. Morphology of ductal carcinoma in situ...... 22

Figure 1-12. Images of some histological types of breast cancer...... 23

Figure 1-13. IHC staining of invasive carcinoma...... 25

Figure 1-14. Illustration of a mammogram examination...... 30

Figure 1-15. Schematic diagram of protein structure...... 39

Figure 1-16. Summary of current proteomics technologies...... 43

Figure 1-17. Schematic diagram of shotgun LC-MS/MS analysis...... 54

Figure 1-18. Peptide sequencing by MS/MS for protein identification...... 58

xviii

Figure 1-19. Strategies for global protein identification and quantification...... 59

Figure 1-20. Schematic diagram of the composition of whole blood...... 86

Figure 2-1. Urine sample collection and handling protocol for proteomics analysis...... 123

Figure 2-2. A flow chart for serum and plasma sample collection...... 124

Figure 2-3. Serum and plasma samples tubes used for blood collection ...... 124

Figure 2-4. Protein purification and precipitation protocol with acetone...... 126

Figure 2-5. Mini-bead beater: Biospec product ...... 128

Figure 2-6. Sample preparation workflow for LC-MS/MS...... 132

Figure 2-7. Microcon fractionation procedure...... 135

Figure 2-8. LTQ Velos- Orbitrap at BMSF in UNSW ...... 136

Figure 2-9. Chromatogram alignment to reference run...... 139

Figure 2-10. Principal component analysis...... 140

Figure 2-11. Correlation analysis for blood collection tubes data...... 141

Figure 2-12. Mascot search results detailed for DCIS 3-50kDa serum sample ...... 144

Figure 2-13. Western blot gel run ...... 150

Figure 3-1. The comparison of urinary precipitation methods on SDS-PAGE...... 164

Figure 3-2. Venn diagram comparison of proteins identified by LC-MS/MS ...... 168

Figure 4-1. Heat map analysis of urine proteins from breast cancer patients and control subjects...... 186

Figure 4-2. Sub-cellular locations of the 59 significant urinary proteins...... 192 xix

Figure 4-3. Ingenuity cell growth and proliferation analysis of the urine proteins detected in breast cancer...... 207

Figure 4-4. In silico identification of interactive networks...... 208

Figure 4-5. Ingenuity pathway analysis showing the top enriched canonical pathways...... 209

Figure 4-6. Validation of urine protein markers using WB and IHC...... 212

Figure 4-7. Expression of MAST4 in the individual urine samples from breast cancer patients and controls...... 214

Figure 4-8. Kaplan-Meier plot of recurrence free survival by MAST4 mRNA expression in breast cancer...... 215

Figure 4-9. Workflow diagram of the urine analysis for novel protein markers...... 216

Figure 5-1. Scatter plot of distribution of proteins in serum and plasma ...... 228

Figure 5-2. Percentage variation of protein abundances in 0-3 kDa serum and plasma fraction...... 230

Figure 5-3. Percentage variation of protein abundances in 3-50kDa serum and plasma fraction...... 233

Figure 6-1. Summary of workflow showing the steps for blood fractionation and concentration...... 258

Figure 6-2. Work flow showing the steps required for LC-MS/MS of blood for biomarker discovery...... 258

Figure 6-3. Principal component analysis of all identified proteins from LC-MS/MS...... 267

Figure 6-4. Venn diagram for the distribution of proteins identified by LC-MS/MS in breast cancer...... 272 xx

Figure 6-5. Cluster analysis of the diseases and bio-functions in breast cancer and BBD are summarised in a heat-map form...... 273

Figure 6-6. IPA showing the top related canonical pathways in breast cancer...... 274

Figure 6-7. IPA analysis of associated disease and functional networks with identified potential breast cancer markers...... 276

Figure 6-8. IPA analysis of associated disease and functional networks showing association with fatty acid metabolism and cellular infiltration...... 277

Figure 6-9. Validation of serum proteins in breast cancer cell lines by WB...... 279

Figure 6-10. Validation of identified serum proteins in serum and plasma by WB. 281

Figure 6-11. IHC of selected protein markers in human breast cancer tissue...... 283

Figure 6-12. Kaplan-Meier survival analysis for identified four potential proteins.285

xxi

LIST OF TABLES Table 1-1.The five most commonly diagnosed cancers in females in Australia...... 6

Table 1-2. Breast carcinoma subtypes: histopathological, molecular and clinical features...... 26

Table 1-3. Clinical staging based on the combined TNM system...... 32

Table 1-4. Potential breast cancer urine biomarkers identified by proteomic technologies...... 76

Table 1-5. Prognostic markers currently applied to the assessment of breast tumour...... 80

Table 1-6. Serum biomarkers used for breast cancer diagnosis...... 83

Table 1-7. Breast cancer blood (serum or plasma) biomarkers in the literature discovered using proteomic techniques...... 95

Table 1-8. Additional serum biomarkers identified using MS analysis...... 104

Table 2-1. Summary of breast cancer cell lines...... 119

Table 2-2. Antibodies used for WB and IHC...... 120

Table 2-3. Strong cation exchange solutions...... 134

Table 3-1. Summary of all the urine precipitation techniques applied...... 160

Table 3-2. Summary of number of proteins identified with LC-MS/MS and total protein extracted with the different urine protein precipitation methods...... 165

Table 4-1. Histopathology characteristics and parameters, of the patients in this study...... 177

Table 4-2 Characteristics of breast cancer cell lines...... 182

xxii

Table 4-3. A list of urinary proteins identified by LC-MS/MS, uniquely associated with specific stages of breast cancer...... 188

Table 4-4. A list of proteins with decreased expression in urine, uniquely associated with certain stages of breast cancer(DCIS, IBC and MBC)...... 194

Table 4-5. A list of differentially expressed urinary proteins in breast cancer and benign breast disease...... 197

Table 4-6. A list of urine proteins up-and-down regulated in benign breast disease...... 202

Table 5-1. Percentage variation of proteins in BC, BBD patients and control subjects for 0-3kDa serum and plasma fraction...... 229

Table 5-2. Percentage variation of proteins in breast cancer, BBD patients and control subjectsfor 3-50kDa serum and plasma fraction...... 232

Table 5-3. Differential expression of proteins in serum and plasma in 3-50 kDa fraction...... 236

Table 5-4. Differential expression of proteins in serum and plasma in 0-3kDa fraction...... 248

Table 6-1. Comparison of protein abundance across 3-50kDa serum and plasma fraction in breast cancer and BBD...... 264

Table 6-2. A list of significant human breast cancer proteins identified by LC–MS/MS in 3-50kDa serum fraction...... 268

xxiii

LIST OF ABBREVIATIONS

1DE One-dimensional electrophoresis 2D LC-MS Two-dimensional liquid chromatography-tandem mass spectroscopy 2D Two dimensional 2D-DIGE Differential gel electrophoresis 2DE Two-Dimensional Electrophoresis 2DGE Two-dimensional gel electrophoresis ATCC American Type Culture Collection ATM Ataxia telangiectasia mutated BC Breast cancer BSA Bovine serum albumin CE Capillary electrophoresis CEA Carcino-embryonic antigen CPC Cetyl pyridium chloride CTL Control DAB Diaminobenzidine DCIS Ductal carcinoma in situ dH2O deionised and distilled water DMSO Dimethyl sulfoxide DNA Deoxyribonucleic acid DPBS Dulbecco’s phosphate buffered saline DTT Dithiothreitol EDTA Ethylene diamine tetra-acetic acid (disodium) EGFR Epidermal growth factor receptor EM Oestrogen metabolites ER Oestrogen receptor ESI Electrospray ionization FBS Foetal bovine serum FCS Foetal calf serum FDA Food and drug administration FDR False discovery rate

xxiv

FTICR Fourier transform ion cyclotron resonance GAG Glyco-amino glycan GC-MS Gas chromatography-mass spectrometry GE Gel electrophoresis H&E Haematoxylin and eosin HCl Hydrochloric acid HER2 Human epidermal growth factor receptor 2 hr Hour HPLC High-performance liquid chromatography HSC High speed centrifugation ICAT Cleavable isotope-coded affinity tags IDC Invasive ductal carcinoma IEF Isoelectric focusing Ig Immuno globulin IGFBP Insulin-like growth factor-binding protein IHC Immunohistochemistry IPA Ingenuity Pathways Analysis iTRAQ Isobaric tags for relative and absolute quantitation LC Liquid chromatography LCM Laser capture micro-dissection LC-MS Liquid chromatography-mass spectrometry LC-MS/MS Liquid chromatography- tandem mass spectrometry LRG Leucine-rich glycoprotein LN Lymph node m/z Mass/charge ratio MAb Monoclonal antibody MALDI-TOF Matrix-assisted laser desorption/ionization time-of-flight Min Minutes mRNA messenger RNA

Milli-Q-H2O Milli-Q-water MS Mass spectrometry MS-MS Tandem mass spectrometry MW Molecular weight xxv

NAF Nipple aspirate fluid o/n Overnight oC Degrees Celsius PAb Monoclonal antibody PAGE Polyacrylamide gel electrophoresis PAI-1 Plasminogen activator inhibitor 1 PBS Phosphate buffered saline

RB Rehydration buffer RNA Ribonucleic acid RPM Revolutions per minute RT Room temperature SDS Sodium dodecyl sulfate

SDS-PAGE sodium dodecyl sulfate-polyacrylamide gel electrophoresis SELDI-TOF Surface- enhanced laser desorption ionization and time-of-ﬂight SILAC Stable isotope labelling by amino acids in cell culture TBS Tris-Buffer Saline TCA tri-chloro acetic acid TDLU Terminal duct lobular unit TOF Time-of-flight Tris 2-amino-2-hydroxy-(hydroxymethyl)-propane-1, diol UF Ultrafiltration UNSW University of New South Wales US Ultrasound UV Ultra violet v/v, w/v Volume per volume, weight per volume WB Western blotting x g Relative centrifugal force (x g)

L, mL, L microlitre, millilitre, litre

M, mM, M micromolar, millimolar, molar

xxvi

LIST OF PUBLICATIONS AND PRESENTATIONS

JOURNAL ARTICLES ASSOCIATED WITH MY PHD STUDY

Beretov J, Wasinger VC, Schwartz P, Graham PH, Li Y. A standardized and reproducible urine preparation protocol for cancer biomarkers discovery. Biomarkers in Cancer. 2014; 6:21-7.

Beretov J, Wasinger VC, Graham PH, Millar EK, Kearsley JH, Li Y. Proteomics for breast cancer urine biomarkers. Adv Clin Chem. 2014; 63:123-67.

Beretov J, Wasinger VC, Millar, EKA, Schwartz P, Graham PH, Li Y. Proteomic analysis of urine to identify breast cancer biomarker candidates using a label-free LC-MS/MS approach. PloS One. Published online 2015; 10(11):e0141876

Beretov J, Wasinger V. C, Millar E. K. A, Schwartz P, Graham P. H, Li Y. A panel of novel serum and plasma proteins revealed by LC-MS/MS in breast cancer. Breast cancer Research. Submitted March 2016. [Under review].

CONFERENCE ORAL PRESENTATIONS

Beretov J. Discovering Urine Protein Biomarkers in Breast Cancer using Proteomic Technology. The St George & Sutherland Medical Research Symposium, 11 Oct, 2012 Sydney, NSW, Australia. St George and Sutherland Clinical School 2012 Research Excellence Second Prize for Scientific Presentation Award, UNSW, Australia.

Beretov J, Wasinger VC, Schwartz P, Graham PH, Kearsley JH and Li Y. Proteomics Investigation of Biomarkers using Human Plasma and Serum for Early Breast Cancer diagnosis. APAF, Proteomics and Beyond Symposium. Macquarie University, 12 November 2014.

xxvii

Beretov J. Novel biomarkers in urine and blood for the early detection of breast cancer. St George and Sutherland Clinical School and UNSW, Three Minute Thesis Presentation. 11th June 2014, St George Hospital, Kogarah, Sydney, Australia. People’s Choice Award.

CONFERENCE POSTER PRESENTATIONS

Beretov J. Graham PH, Kearsley JH, Li Y. The HDAC inhibitor Panabinostat (LBH589) significantly enhances Radiotherapy effect in Breast Cancer Cell Lines. 13th Milan International Breast Cancer Conference. 22-24 June 2011, Milan, Italy.

Beretov J, Wasinger V, Schwartz P, Graham P, Kearsley J, Li Y. Discovering Urine Protein Biomarkers in Breast Cancer using Proteomics Technology. APAF, Proteomics and Beyond Symposium. 7th November 2012, Macquarie University, Sydney, Australia.

Beretov J, Wasinger V, Graham P, Kearsley J, Li Y. Proteomic Identification of Novel Urine Biomarkers- for the Early Detection of Breast Cancer. HUPO-11th Annual World Congress- 9-13 September 2012, Boston, USA.

Beretov J, Wasinger VC, Schwartz P, Graham PH, Kearsley JH, Li Y. Proteomic investigation of breast cancer biomarkers using human plasma and serum samples and comparing blood tubes. HUPO 12th Annual World Congress, 14-18 September 2013, Yokohama, Japan.

Beretov J, Wasinger VC, Schwartz P, Graham PH, Kearsley JH and Li Y. Proteomic investigation of breast cancer biomarkers using human plasma and serum samples. HUPO 13th Annual World Congress, 5-8 October 2014, Madrid, Spain. UNSW Travel grant award.

Beretov J, Wasinger VC, Schwartz P, Graham PH, Kearsley JH, Y Li, Proteomic investigation of biomarkers using human plasma and serum samples for early breast

xxviii cancer diagnosis. The St George & Sutherland Medical Research Symposium, 13 Oct 2014, Sydney, NSW, Australia.

OTHER PUBLICATIONS AS A CO-AUTHOR

Abbas Rizvi SM, Song EY, Raja C, Beretov J, Morgenstern A, Apostolidis C, Russell PJ, Kearsley JH, Abbas K, Allen BJ. Preparation and testing of bevacizumab radio- immuno-conjugates with Bismuth-213 and Bismuth-205/Bismuth-206. Cancer Biol Ther. 2008; 7 (10):1547-54.

Li Y, Song E, Abbas Rizvi SM, Power CA, Beretov J, Raja C, Cozzi PJ, Morgenstern A, Apostolidis C, Allen BJ, Russell PJ. Inhibition of micro metastatic prostate cancer cell spread in animal models by 213Bilabelled multiple targeted alpha radio- immunoconjugates. Clin Cancer Res. 2009; 15(3):865-75.

Wang L, Madigan MC, Chen H, Liu F, Patterson KI, Beretov J, O'Brien PM, Li Y. Expression of urokinase plasminogen activator and its receptor in advanced epithelial ovarian cancer patients. Gynecol Oncol. 2009; 114(2):265-72.

Millar EK, Graham PH, O'Toole SA, McNeil CM, Browne L, Morey AL, Eggleton S, Beretov J, Theocharous C, Capp A, Nasser E, Kearsley JH, Delaney G, Papadatos G, Fox C, Sutherland RL. Prediction of local recurrence, distant metastases, and death after breast-conserving therapy in early-stage invasive breast cancer using a five- biomarker panel. J Clin Oncol. 2009; 27(28):4701-8.

Hao J, Chen H, Madigan MC, Cozzi PJ, Beretov J, Xiao W, Delprado WJ, Russell PJ, Li Y. Co-expression of CD147 (EMMPRIN), CD44v3-10, MDR1 and mono-carboxylate transporters is associated with prostate cancer drug resistance and progression. Br J Cancer. 2010; 103(7):1008-18.

Chen H, Wang L, Beretov J, Hao J, Xiao W, Li Y. Co-expression of CD147/EMMPRIN with monocarboxylate transporters and multiple drug resistance proteins is

xxix associated with epithelial ovarian cancer progression. Clin Exp Metastasis. 2010; 27(8):557-69.

Wang L, Chen HM, Pourgholami MH, Beretov J, Hao J, Chao H, Perkins AC, Kearsley JH, Li Y. Anti-MUC1 monoclonal antibody (C595) and docetaxel markedly reduce tumour burden and ascites, and prolong survival in an in vivo ovarian cancer model. PLoS One. 2011; 6(9): e24405.

Hao J, Madigan MC, Khatri A, Power CA, Hung TT, Beretov J , Chang L, Xiao W, Cozzi PJ, Graham PH, Kearsley J, Li Y. In vitro and in vivo prostate cancer metastasis and chemoresistance can be modulated by expression of either CD44 or CD147. PLoS One. 2012;7(8):e40716.

Ni J, Cozzi P, Hao J, Beretov J, Chang L, Duan W, Shigdar S, Delprado W, Graham P, Bucci J, Kearsley J, Li Y. Epithelial cell adhesion molecule (EpCAM) is associated with prostate cancer metastasis and chemo/radioresistance via the PI3K/Akt/mTOR signalling pathway. Int J Biochem Cell Biol. 2013; 45(12):2736-48.

Ni J, Cozzi P, Hao J, Beretov J, Chang L, Duan W, Shigdar S, Delprado W, Graham P, Bucci J, Kearsley J, Li Y. CD44 variant 6 as a cancer stem cell-like marker is associated with prostate cancer metastasis and chemo-/radio-resistance. The Prostate. 2014; 74(6):602-17.

Krilis S, Zhang P, Weaver JC, Chen G, Beretov J, Atsumi T, Qi M, Bhindi R, Qi JC, Madigan MC, Giannakopoulos B. The Fifth Domain of Beta 2 Glycoprotein I Protects from Natural IgM Mediated Cardiac Ischaemia Reperfusion Injury. PloS One. Accepted March 2016, awaiting publication.

OTHER CONFERENCE POSTER PRESENTATIONS AS A CO-AUTHOR

Ni J, Cozzi P, Beretov J, Duan W, Delprado W, Graham PH, Kearsley J, Li Y (2012). Over-expression of EpCAM (CD326) and CD44 variants in human prostate cancer cells

xxx are potential therapeutic targets for targeted therapy. 13th Australasian Prostate Cancer Conference, 31 July-8 August 2012, Melbourne, VIC, Australia.

Ni J, Cozzi P, Hao JL, Beretov J, Chang L, Duan W, Delprado W, Peter G, Bucci J, Kearsley J, Li Y (2012). EpCAM (CD326) and CD44 variants are biomarkers associated with prostate cancer metastasis and progression. The St George & Sutherland Medical Research Symposium, 11 Oct, 2012 Sydney, NSW, Australia.

Ni J, Cozzi P, Hao JL, Beretov J, Chang L, Duan W, Delprado W, Peter G, Bucci J, Kearsley J, Li Y (2013). CD44 variant 6 is a biomarker associated with prostate cancer metastasis and progression. Lowy Cancer Symposium, 15-17 May 2013, Sydney, NSW, Australia.

Ni J, Cozzi P, Hao JL, Beretov J, Chang L, Duan W, Delprado W, Peter G, Bucci J, Kearsley J, Li Y (2013). CD44 variant 6 is a biomarker associated with prostate cancer metastasis and progression. The St George & Sutherland Medical Research Symposium, 17 Oct 2013, Sydney, NSW, Australia.

Ni J, Cozzi P, Hao J, Beretov J, Chang L, Duan W, Delprado W, Graham P, Bucci J, Kearsley J, Li Y (2014). CD44 isoform variant 6 is associated with prostate cancer progression, metastasis and chemo-/radio-resistance via PI3K/Akt/mTOR and Wnt/β-catenin signalling pathways in vitro. The American Association for Cancer Research Annual Meeting 2014, 5-9 Apr 2014, San Diego, CA, USA.

Ni J, Cozzi P, Hao JL, Beretov J, Chang L, Deng JL, Duan W, Graham P, Bucci J, Kearsley J, Li Y (2015). Epithelial cell adhesion molecule (EpCAM) is associated with prostate cancer chemo-/radio-resistance in cell lines in vitro and in animal models in vivo. The 2nd Prostate Cancer World Congress, 17-21 August 2015, Cairns, QLD, Australia.

Chang L, Wasinger V, Graham P, Hao JL, Ni J, Beretov J, Bucci J, Cozzi P, Kearsley J, Li Y (2013). Identification of novel biomarkers for prostate cancer radioresistance using the label-free LC-MS/MS approach. HUPO 12th Annual World Congress, 14-18 September 2013, Yokohama, Japan. xxxi

Hao JL, Graham P, Chang L, Ni J, Wasinger V, Beretov J, Bucci J, Cozzi P, Li Y (2015). Identification of lactate dehydrogenase A (LDHA) as a potential therapeutic target for prostate cancer radiotherapy. AACR annual meeting, 18-22 Apr 2015, Pennsylvania, PA, USA.

Hao JL, Graham P, Chang L, Ni J, Wasinger V, Beretov J, Deng JL, Bucci J, Cozzi P, Li Y (2015). Combination of LBH589 and radiotherapy overcomes radioresistance in radioresistant prostate cancer cells in vitro and animal models in vivo. The 2nd Prostate Cancer World Congress, 17-21 August 2015, Cairns, QLD, Australia.

xxxii

HYPOTHESIS AND AIMS

The central hypothesis of this thesis is to mine various biological fluids for proteins associated with breast cancer and to identify specific proteins or a panel of proteins which can identify the presence of breast cancer early and even possibly highlight progression to invasive cancer and metastatic disease (breast cancer progression and relapse).

The overall aim of this project was to identify patterns of protein expression in urine and blood for the early detection of breast cancer, which may translate to improved outcomes through early therapeutic intervention. The specific aims of the project were to: 1. Collect urine and blood samples from different stages of breast cancer patients and age matched control subjects. 2. Develop a method for precipitation and purification to analyse the samples using proteomics technology. 3. To identify the up-regulated and down-regulated proteins specific to each disease state. 4. To validate a selected group of up-regulated proteins in the patients’ tumour samples and cell lines. 5. Select the tubes and source plasma or serum which optimises the best available information. 6. Analyse the blood samples collected from breast cancer patients. 7. Identify the proteins which highlight the presence of disease and to correlate these finding and patterns with protein expression in the patients’ tumour. 8. To identify a panel of proteins that may be used as indicators of the presence of breast cancer, as a possible screening tool. Correlate patterns of abnormal expression in breast cancer with known prognostic indicators and determine their potential utility as prognostic markers of disease progression and relapse.

Chapter 1 Introduction

This chapter, will begin with an introduction of breast cancer, its morphology, and current detection and treatment approach, along with a comprehensive summary of emerging proteomics technology applied for breast cancer biomarker identification. Also, significant finding in both urine and blood will be highlighted.

Parts of this literature review have been published in: Beretov J, Wasinger VC, Graham PH, Millar EK, Kearsley JH, Li Y. Proteomics for breast cancer urine biomarkers. 2014; 63:123-67. Review.

1 INTRODUCTION

INTRODUCTION TO BREAST CANCER

Breast cancer is a commonly diagnosed cancer and the leading cause of cancer deaths in women worldwide with an estimate 1.7 million cases and 521,900 deaths in 2012 (Ferlay J, Soerjomataram I et al. 2012). The five-year relative survival rate in Australian women increased, from 72 % in 1982-1987, to 89% in 2006–2010 (AIHW 2012). Although the treatment outcomes have significantly improved, the survival rate correlates with the size of the tumour after diagnosis and can range from 98% for 0–10 mm tumours to 73 % for women with tumours 30 mm or greater.

Classified according to its histological and molecular features, breast cancer is a heterogeneous disease which grows locally but often spreads by lymphatic or haematogenous routes to lymph nodes or distal sites within the body. Important pathological features which give prognostic and predictive information at the time of diagnosis include tumour size, grade (degree of tumour cell differentiation), and hormone receptor status: estrogen receptor (ER), progesterone receptor (PR), and human epidermal growth factor receptor 2 (HER2). These features guide therapeutic decisions made by a multi-disciplinary team which includes oncologists, surgeons and pathologists (Goldhirsch, Winer et al. 2013). They discuss the relevant features of each case to personalise patient treatment, within relatively broad prognostic and treatment responsive groups, but these features are an imperfect predictor of behaviour for individual patients.

A population mammographic screening program for women (age 50-70) is in place to improve early detection, but is less effective in young women with dense breasts and in patients with rapidly growing more biologically aggressive tumours (Esserman, Shieh et al. 2011). Early detection of breast cancer is a key predictor of improved survival. The current standard of mammography is associated with radiation exposure, uncomfortable, and in some groups expensive. Hence, a new 3

approach to target the whole population should be considered and is why finding biomarkers for the early detection of breast cancer is essential.

Epidemiology

Worldwide Incidence

The World Health Organisation (WHO), International Agency for Research on Cancer (IARC), GLOBOCAN 2012 Cancer Incidence and Mortality Worldwide report (including 27 cancers from 21 world regions) estimated that 14.1 million new cancer cases and 8.2 million cancer deaths occurred in 2012 worldwide (Torre, Bray et al. 2015). The Globocan 2012, 5-year global prevalence report shows breast cancer is the most frequently diagnosed cancer and the leading cause of cancer death among females worldwide, with an estimated 1.7 million cases and 521,900 deaths in 2012 (Ferlay J, Soerjomataram I et al. 2012), shown in Figure 1-1.

Figure 1-1. Estimated new cancer cases and deaths worldwide.

Illustration of the cancer incidence and death rate for males and females worldwide (Excluding non-melanoma skin cancers). Source: GLOBOCAN 2012.

The three top ranked cancers commonly diagnosed cancers in women (excluding non-melanoma skin cancer) are depicted in Figure 1-2, which account for more than 43% of all cancers (Ferlay J, Soerjomataram I et al. 2012, Bray, Ren et al. 2013, Yip and Taib 2014).

Figure 1-2. Bar chart of 5-year prevalence of breast cancer.

Bar chart indicates the global 5-year prevalence of breast cancer in counts (millions) by cancer site (Bray, Ren et al. 2013). The stacked bars denote prevalence among patients alive at the end of 2008 who were diagnosed in 2008 (0-1 years), 2006- 2007 (2-3 years) and 2004-2005 (4-5 years). All cancer combined (except non- melanoma skin cancers) and includes both sexes, plus ages 15 years and over.

National Incidence

In Australia, breast cancer is the most the commonly diagnosed invasive cancer in women. On average, 1 in 8 Australian women will be diagnosed with breast cancer before the age of 85, and one in 39 females will die (AIHW 2013). In 2008, of the five most commonly diagnosed cancers (Table 1-1) breast cancer was the highest with, a total of 13,567 women diagnosed with up to 70% in the 40–69 age group (AIHW 2013).

Table 1-1.The five most commonly diagnosed cancers in females in Australia.

Cancer type Number of Cases Percentage of all cancers in females

Breast 13,567 28.2 Bowel 6,375 13.2 Melanoma of skin 4,581 9.5 Lung 3,944 8.2 Lymphoid cancers 3,181 6.6

Note: Breast cancer is the most commonly diagnosed cancers in females (excluding basal and squamous cell carcinoma of the skin), in Australia. Source: AIHW Australian Cancer Database 2008 (AIHW 2012).

Aetiology

Who Is at Risk? Breast cancer is by far the most common cancer in women (Ferlay, Shin et al. 2010). The World Cancer Report 2008 (IARC 2008) and National Breast and Ovarian Cancer Centre’s 2009 report (NBOCC 2009a), estimated incidence rates of breast cancer around the world, shows Australia to be amongst the top four (83 new cases per 100,000 women), more than double the average for the world incidence.

According to the National Cancer Institute 2014, not only are women at a greater risk (100 times) of developing breast cancer than men, behavioural and environmental factors also play a large part. Major risk factors include increasing age and genetic susceptibility, women who inherit gene mutations associated with breast cancer (Hulka and Moorman 2008). Other associated high risks are increased exposure to endogenous oestrogen (due to early menarche and late menopause), null parity (child bearing history), and oestrogen-progesterone hormones (Parkin, Fernandez et al. 2006). Factors that are thought to aid in reducing the risk of breast cancer include early pregnancy, compared with nulliparous women or women who give birth after age 35 years and breast-feeding women (Cancer 2002).

Women with dense breasts have an increased risk, proportionate to the degree of density (McCormack and dos Santos Silva 2006, Boyd, Martin et al. 2009). There is evidence that links lifestyle to cancer outcomes. These behavioural factors include smoking, poor quality diet, medication, excessive alcohol consumption, physical inactivity (Demark-Wahnefried, Platz et al. 2012, Ligibel 2012) and obesity (Schaub, Jones et al. 2009), particularly in postmenopausal obese women (Kwan, Kushi et al. 2010). Environmental factors associated with cancer include exposures to: ionizing radiation (such as x-rays or full body scans) especially during puberty or young adulthood, certain chemicals (pesticides), biological agents, and toxins (like second hand smoke). Though many risk factors can increase the chance of developing breast cancer, it is not exactly known how these risk factors cause cells to become cancerous. 7

Genetic alterations can either be inherited or acquired. Carcinogenesis is considered a multistage process with sequential acquisition of gene mutations (Hanahan and Weinberg 2000, Campeau, Foulkes et al. 2008, Hanahan and Weinberg 2011). Unfortunately, inherited DNA mutations can dramatically increase the risk for developing breast cancer, however only a small proportion (≤10%) of breast cancer are due to hereditary mutations (Marchina, Fontana et al. 2010). Models suggest that a larger fraction of so-called sporadic cases of breast cancer might be attributable to the action of multiple genes (Pharoah, Antoniou et al. 2002). Genes are inherited and provide the instructions for how the cells should function. Women who have first-degree relatives, with a history of the disease are at an increased risk (Beebe-Dimmer, Yee et al. 2015).

The first discovered susceptibility genes for breast cancer were human tumour suppressor genes including BRCA1 and BRCA2, both are normally expressed in the cells of breast and other tissues. They are involved in DNA repair (Venkitaraman 2002) and therefore play a role in ensuring the stability of the cell's genetic material. Most pathogenic BRCA1 or BRCA2 mutations block protein production from the mutated allele. Carriers of this mutation confer a high risk of developing breast cancer, with 10-30 times higher than other women in the general population (Antoniou, Pharoah et al. 2003, Mavaddat, Peock et al. 2013). BRCA1-related breast cancers are usually high-grade carcinomas that do not express ER, PR, or HER2 and hence are called triple negative.

They often express cytokeratin 5 and 14, vimentin, epidermal growth factor receptor (EGFR), P-cadherin (CDH3), and are classified as basal-like breast cancer (Turner and Reis-Filho 2006, Campeau, Foulkes et al. 2008) according to their molecular type. Although BRCA1 and BRCA2 mutations are rare in most populations (approximately 1 in 400 persons), they are much more common in the Ashkenazi Jewish population in which 1 of 40 persons carries one of three main disease- causing mutations (Narod and Foulkes 2004).

A second class of breast cancer susceptibility alleles which are referred to as moderate risk alleles, as they are rare in most populations, include genes, CHEK2, ATM, BRIP1, and PALB2(Foulkes 2008). Other genes implicated in the inheritance of breast cancer are TP53, PTEN, STK11, CHEK2, ATM, PALB2, BRIP1, and CASP8 (Njiaju and Olopade 2012).

Common, low-risk genes for breast cancer are SNPs in FGFR2 and TOX3, and those on chromosomes 5p and 2q that specifically increase the risk of ER positive breast cancer. Breast cancer susceptibility loci and genes implicated in the inheritance of breast cancer in relation to relative risk are shown in Figure 1-3. To date, breast cancer risk assessment is largely restricted to testing for high-penetrance mutations by genetic methods such as BRAC1 and BRCA2 (Njiaju and Olopade 2012).

In summary, cancer can be caused by a variety of factors as discussed, and may develop over a number of years. An important and well-established breast cancer risk factor is family history. Other risk factors include increasing age, mammographic breast density, reproductive history, menopausal status and exogenous hormones. Additionally, lifestyle factors related to Westernisation and affluence show a higher risk. A healthy lifestyle and preventing exposure to certain environmental risk factors may help prevent the development of cancer.

Figure 1-3. Breast Cancer Susceptibility Loci and Genes.

Genes implicated in the inheritance of breast cancer in relation to relative risk. Source: (Foulkes 2008)

PATHOLOGY OF BREAST CANCER

Cell of origin

Cancer is caused by the evolutionary accumulation of somatic mutations in the progeny of a normal cell, which leads to uncontrolled proliferation (Huang, Harvie et al. 2005) and is a multi-step, multi-hit process.

Cellular and molecular events enable the malignant transformation of cells harbouring oncogenic alterations. These events include uncontrolled proliferation through activation of oncogenes; evasion of tumour suppression; inhibition of cell death (apoptosis); creation of a favourable tumour microenvironment containing increased blood vessels, stromal and immune cells; and the acquisition of invasive and metastatic potential (Hanahan and Weinberg 2011).

Cells are the basic unit of life and the building block that make up all tissues and organs of the body, including the breast. Normal cells grow, divide and die and new cells replace them (as shown in Figure 1-4A). In cancer cells, this mechanism is faulty and leads to uncontrolled replication. This process of normal cell growth is regulated by the cell cycle (shown in Figure 1-4B). Tumour suppressor genes which act as braking signals during G1/S (first gap -part of the interphase), stop or slow down the cell cycle to detect and repair abnormalities in DNA before S phase (for DNA synthesis) and replication. DNA repair genes are active throughout the cell cycle, particularly during G2 (second gap), after DNA replication and before the chromosomes prepare for cell division (mitosis) during the M phase. Usually, DNA repair is successful and the cell cycle will complete, resulting in cell division.

Figure 1-4. Hypothetical models of cell division.

Schematic model of normal cell and cancel cells replicating: (A) Normal cell (pink) division and growth that eventually stops whereas with damaged cells (blue) the cancer cells continue to divide with uncontrolled replication; (B) The normal cell cycle is depicted in the diagram showing how the cell progresses through the cell cycle checkpoints with the three main checkpoints for: cell growth (at G1), DNA synthesis (at S), chromosomes and DNA not damaged (G2) and mitosis (at M).

Unsuccessful DNA repair usually results in programmed cell death by apoptosis. Malfunction in the system, the failure to sense DNA abnormality, unsuccessful repair or inability of the cell to die by apoptosis will lead to transmission of DNA abnormalities to progeny cells. This will eventually lead to uncontrolled cell growth and tumour formation (Hanahan and Weinberg 2000, Hanahan and Weinberg 2011).

Most cancers acquire mutations in proto-oncogenes (normal genes), involved in controlling the signals that regulate the cell's cycle for growth and division. These genes encode proteins that function as growth factors, growth factor receptors, signal-relaying molecules, and nuclear transcription factors (proteins that bind to genes to start transcription). When the proto-oncogene is mutated or overregulated, it becomes an oncogene and results in unregulated, accelerated cell growth and transformation. This means that the cells lose the ability to react to signals and die, and hence proliferate uncontrollably. Common signalling pathways related to biomarkers, progression and targeted therapy include HER signalling pathway (EGFR/HER1, HER2, HER3 and HER4), tyrosine kinases family (such as Src), insulin- like growth factor (IGF)/IGF-receptor (IGFR); agents that interfere with critical pathways, such as PI3K/AKT/mTOR inhibitors (targets PIK3CA) , RAS/MEK/ERK inhibitors; agents that promote apoptosis such as Poly ADP ribose polymerase (PARP) inhibitors and agents that target invasion and metastasis such as MMP inhibitors (Munagala, Aqil et al. 2011, Gagliato, Jardim et al. 2016).

A tumour can either be benign or malignant. A benign tumour is usually well circumscribed with little infiltration of surrounding tissue and is usually not harmful. A malignant tumour however may be life threatening as it can show destructive local invasion of tissues and organs, and can metastasise to other parts for the body. The precise cell types that give rise to breast cancers and the mechanism of tumour heterogeneity are poorly understood. The mammary gland is comprised of two main cell types, the luminal and myoepithelial cells (Figure 1-5). Stem cells can develop into any of the two cell types through a series of 13

differentiation steps (Vlahou, Laronga et al. 2003). Breast cancer may express characteristics of both basal and luminal phenotypes.

Figure 1-5. Normal cellular structure of breast tissue.

Schematic depiction of the cellular structure of the mammary duct/lobule in normal breast tissues. The mature duct features an outer layer of myoepithelium cells (red) which are ER negative and inner layer of luminal epithelial cells (blue), that are ER positive. Stem cells (orange) reside in a basal population in between.

These subtypes are conserved across ethnic groups and are already evident at the ductal carcinoma in situ (DCIS) stage (Shen, Kim et al. 2005). Comprehensive gene expression profiling studies has revealed five major molecular subtypes of breast cancer: basal-like, luminal A, luminal B, HER2+/ER–, and normal breast–like (Perou, Sorlie et al. 2000, Echan, Tang et al. 2005, Hu, Fan et al. 2006, Sorlie, Wang et al. 2006, Villar-Garea, Griese et al. 2007). These molecular profiles have distinct clinical outcomes and responses to treatment: luminal A-type tumours have the best prognosis, and the basal-like tumours have the worst prognosis (Sorlie, Perou et al. 2001). The two possible explanations for this extensive intra- and inter-tumoural heterogeneity are: distinct cell of origin of cancer stem cells (CSC) and tumour subtype–specific genetic or epigenetic events (Figure 1-6). One theory is based on the cell of origin hypothesis. The different tumour subtypes are initiated from the different cell types, presumably stem or progenitor cell. Alternatively, the cell of origin (stem cell) can be the same for the different tumour subtypes and the tumour

phenotype is primarily determined by acquired genetic and epigenetic events (Polyak 2007).

Figure 1-6. Hypothetical models explaining breast tumour subtypes.

Schematic diagram of cell of origin depicting: (A) tumour subtype–specific transforming event models from stem or progenitor cell and; (B) different tumour subtypes and tumour phenotypes event from the same cell of origin, possibly determined by acquired genetic and epigenetic events. Sourced from (Polyak 2007).

DCIS is thought to be a precursor of invasive ductal carcinoma based on molecular, epidemiological, and pathological studies (Pieper, Su et al. 2003, Zhou, Lucas et al. 2004). Therefore, the micro environmental (cellular composition) influences breast cancer initiation and progression through defined pathological and clinical stages (Figure 1-7), with subsequent evolution into in situ and invasive carcinomas, and finally into metastatic disease (Pieper, Su et al. 2003, Bjorhall, Miliotis et al. 2005, Polyak 2007). The ability to detect early-stage disease, and the factors involved would allow us to interfere with tumour progression.

Figure 1-7. Hypothetical model of breast tumour progression.

A schematic view of normal, in situ, invasive, and metastatic carcinoma progression. Normal breast ducts are composed of the basement membrane and a layer of luminal epithelial and myoepithelial cells. Stroma cell composition includes various leukocytes, fibroblasts, myofibroblasts, and endothelial cells. In in-situ carcinomas, the myoepithelial cells are epigenetically and phenotypically altered and their number decreases, potentially due to degradation of the basement membrane. At the same time, the number of stromal fibroblasts, myofibroblasts, lymphocytes, and endothelial cells increases. Loss of myoepithelial cells and basement membrane results in invasive carcinomas, in which tumour cells can invade surrounding tissues and can migrate to distant organs, eventually leading to metastases. Sources from: (Polyak 2007).

MORPHOLOGY OF NORMAL BREAST

Normal breast tissue consists of fat and glandular tissue (arranged in lobules, which can produce milk in women), lactiferous ducts (which carry milk to the nipple) and connective tissue (Figure 1-8). This structural unit is known as the terminal duct lobular unit (TDLU). Breast tissue shows physiological hormonal changes which occur during the menstrual cycle, pregnancy, and menopause.

Figure 1-8. Diagram of normal breast tissue.

Illustration of the anatomy of the human breast. Adapted from the virtual medical centre.com (http://www.myvmc.com/anatomy/breast/) and Hopkins medicine health library. http://www.hopkinsmedicine.org/healthlibrary/GetImage.aspx?ImageId=161343

Histological morphology

Normal Breast

The normal breast TDLU has a specialised supporting stroma which contains fibroblasts and immune cells. Within each lobule there are two populations of cells within the acini and ducts: Luminal cells (ER+) and myoepithelial cells (ER-, p63, actin positive). Typical breast tissue is shown in Figure 1-9.

Figure 1-9. Morphology images of normal breast tissue.

Representative microscopic images of haematoxylin and eosin (H&E) staining, showing: (A) normal terminal duct lobular unit found in normal breast tissue (magnification x100); (B) normal breast tissue at high power (x400); (C) normal breast tissue with p63 IHC staining for myoepithelial cells (x400).

Benign Breast Disease and Precursor Lesions

Benign breast disease (BBD) is a spectrum of different conditions which can present with varied clinical symptoms e.g. lump, thickening, pain, nipple discharge. This group of diseases most commonly includes fibrocystic change, a spectrum of histopathological changes which include epithelial hyperplasia of usual type, fibrosis and cyst formation. Fibroadenoma, and intraductal papilloma can also be present with or without these changes. The microscopic appearance of the various morphological types of BBD is shown in Figure 1-10.

Figure 1-10. Microscopic images of benign breast disease.

Representative H&E images of selected morphological types of BBD including: (A) fibroadenoma (magnification x100); (B) fibrocystic disease (x200) and; (C) columnar cell change (x400), illustrating the different architectural changes in specialised stroma and epithelium.

Fibroadenoma is a mobile, solid firm benign mass. Intraductal papilloma grows exophytically (outwardly projecting) in the milk duct of the breast near the nipple. It is usually microscopic but may grow to 10 mm in diameter. The predominant symptom is spontaneous discharge from one nipple. Columnar cell change is the dilatation of the lobular units with columnar morphology. These may appear cytologically atypical and if so are referred to as flat epithelial atypia.

Benign lesions are often classified into three pathologic categories: non- proliferative, proliferative without atypia, and atypical hyperplasia. It has been suggested that due to the heterogeneous nature of BBD, the possibility of developing into breast cancer is 2-fold in women with proliferative lesions without atypia or with atypical hyperplasia it conifers a 5-fold risk (Singletary 2003, Wang, Costantino et al. 2004, Worsham, Raju et al. 2007, Cheng, Qiu et al. 2008).

The borderline malignant lesions include flat epithelia atypia (columnar cell change with atypia) and atypical ductal hyperplasia which are precursor lesions for the development of low grade DCIS and invasive carcinoma (Lakhani, I.O. et al. 2012, Vuong, Simpson et al. 2014). Lobular carcinoma in-situ (LCIS) is also included in this group although labelled as “carcinoma” as its association with malignancy is relatively low (compared to DCIS). These changes, if found within an otherwise benign breast biopsy, are indicative of increased risk of subsequent development of breast cancer (relative risk 4-5 fold above population risk). Increased surveillance and close follow-up are therefore warranted for these patients. This risk increased further in those individuals with a positive family history of breast cancer.

Breast Cancer

Breast cancer can be classified simplistically as carcinoma in-situ (i.e. non-invasive) or invasive carcinoma.

1.3.1.3.1 Carcinoma In Situ

DCIS is a malignant proliferation of epithelial cells within the large ducts of the breast. The cells grow and remain within the duct space, without invading through the basement membrane into the surrounding stroma. Additionally, there is preservation of myoepithelial cells at the periphery of the involved ducts. Diagnostically, loss of myoepithelial cells is often regarded as a feature of invasive carcinoma.

Based on the architectural pattern, DCIS is comprised of several subtypes which include micropapillary, papillary, solid, cribriform, and comedo (shown in Figure 1-11). DCIS is graded according to nuclear size and pleomorphism into low grade, intermediate and high-grade. Luminal necrosis can be accompanied by coarse microcalcification making this change visible on a mammogram. DCIS is usually local excision and radiotherapy, without lymph-node dissection since axillary metastases are very rare. The presence of DCIS indicates a high-risk for the development of invasive carcinoma (10 times the population risk) in the same breast. Therefore, close follow-up is required. When DCIS is found close to the central ducts of the nipple, it can present with Paget’s disease of the nipple (an erythematous crusting change in the skin), a feature which results from the spread of single malignant cells up the large ducts and into the epidermis of the skin of the nipple.

Figure 1-11. Morphology of ductal carcinoma in situ.

Representative microscopic H&E images of the morphology of ductal carcinoma in situ. (A) DCIS with comedo necrosis and calcification (magnification x100); (B) micropapillary DCIS (x200); (C) cribriform DCIS (x400).

1.3.1.3.2 Invasive Carcinoma

Invasive breast cancer (IBC) constitutes a heterogeneous group of lesions at the phenotypic and molecular levels. Representative morphological variants and lymph node metastases are shown in Figure 1-12. There are two main histological types of breast cancer. Invasive ductal carcinoma (IDC) and invasive lobular carcinoma (ILC) are the most common type of breast cancer, which comprise up to 80% and 10% of all breast cancers respectively. Less frequently found are the so-called “special types”: mucinous, medullary, tubular, and other rare salivary gland type tumours e.g. adenoid cystic carcinoma (Lakhani, I.O. et al. 2012, Corben 2013). ILC is characterised by a distinctive single cell discohesive infiltrative growth pattern mirrored by loss of expression of the adhesion molecule E-cadherin (Lakhani, I.O. et al. 2012, Corben 2013), shown in Figure 1-12 C & D. Inflammatory breast cancer is a clinical diagnosis of a locally advanced high-grade breast cancer associated with extensive dermal lymphatic spread. It is often characterised by inflamed, red patches on the skin, which can mimic inflammation of the breast (mastitis).

Figure 1-12. Images of some histological types of breast cancer.

Representative microscopic H&E images of: (A) IDC grade 1 (magnification x200); and (B) high-power image shows IDC grade 3 at which increased mitotic rate, arrows indicating mitoses (x400); (C) ILC (x100); (D) Micropapillary carcinoma (x100); (E) Mucinous carcinoma (x100); and (F) LN metastasis (x100).

1.3.1.3.3 Histological grading

Grade is an independent prognostic factor of outcome and provides information regarding the degree of differentiation and biological behaviour. All breast cancers are graded using the Nottingham grading scheme which gives a score for each of the assessment of tubule formation (1, 2 or 3), nuclear pleomorphism (1, 2 or 3) and mitotic count (0, 1, 2 or 3). The mitotic count is assessed as the total number of mitoses in ten high power fields (x400, according to a specific field diameter). Each score is added together to give a final score which will classify a tumour as grade 1 (1-5 points), grade 2 (6 & 7 points) and grade 3 (8 & 9 points).

MOLECULAR CLASSIFICATION OF BREAST CANCER &

CURRENT BIOMARKERS

Important prognostic features are determined by the pathological examination of the tumour, which provides phenotypic and molecular characteristics which guides treatment decision making e.g. tumour grade and size, ER, PR and HER 2 status (Ludwig and Weinstein 2005, Payne, Bowen et al. 2008). In 2001, Perou and colleagues described for the first time a molecular classification (intrinsic subtype) of breast cancer, based on gene expression microarray signatures (Perou, Sorlie et al. 2000, Perou and Borresen-Dale 2011). Based on these gene signatures, four major breast cancer subtypes were initially classified: luminal A and B, HER2- enriched and basal-like subtype. Also revealed was the substantial tumour heterogeneity within and between different molecular subtypes, each with distinct biological and clinical characteristics that have moulded breast cancer classification (Sorlie, Perou et al. 2001, Sorlie, Tibshirani et al. 2003).

Translation of these complex multigene signatures into a simplified classification, is now be routinely applied to all cancers at the time of diagnosis using a panel of 4 IHC biomarkers: ER, PR, HER2 and Ki67 (Millar, Graham et al. 2011) shown in Figure 1-13 and Table 1-2.

Figure 1-13. IHC staining of invasive carcinoma.

Representative images of IHC of tissue markers applied to IDC with positive nuclear expression of: (A) ER (magnification x100); (B) PR (x100); (C) Strong cell membrane staining for HER2 (x400); and (D) E-cadherin negative ILC surrounding a normal positive duct (x200).

The assessment of ER/PR is mandatory in the selection of patients for treatment with hormone therapy, while HER-2 is essential in selecting patients for treatment with HER2 targeting agents such as Herceptin (trastuzumab), (Goldhirsch, Wood et al. 2003, Goldhirsch, Wood et al. 2007, Goldhirsch, Wood et al. 2011). Ki-67 is used as a marker of tumour proliferation for all type of breast cancer. Absence of ER and PR, and up-regulation of proliferation marker Ki-67 (Mib-1) are additional indicators of poor prognosis in breast cancer patients (Dowsett, Nielsen et al. 2011, Sheri and Dowsett 2012, Polley, Leung et al. 2013) whose only treatment option is chemotherapy. Urokinase plasminogen activator (uPA) and plasminogen activator inhibitor 1 (PAI-1) were validated as prognostic markers only used in Europe, for lymph node (LN)-negative breast cancer patients (Molina, Barak et al. 2005), as indicators of breast cancer recurrence (Borstnar, Vrhovec et al. 2002, Duffy and Duggan 2004), and may also be useful for adjuvant chemotherapy treatment planning (to exclude node-negative patients).

Table 1-2. Breast carcinoma subtypes: histopathological, molecular and clinical features.

Molecular Pre- Immuno- Histological TP53 Prognosis Consensus recommendation subtype valencea histochemistry Grade mutation for (Neo) adjuvant systemic definitionb treatmentb Luminal A 50–60% ER+ and/or PR+ 1-2 Low Good Endocrine therapy alonec HER2− Ki-67 low

Luminal B 15–20% ER+ and/or PR+ 2-3 High Moderate Endocrine therapy ± (HER2 negative) HER2− Chemotherapyc Ki-67 high

Luminal B 6% ER+ and/or PR+ 2-3 High Moderate Endocrine + cytotoxic + anti- (HER2 positive) Ki-67 &HER2+ HER2 therapy

HER2-enriched 10–15% HER2 + 3 High Poor Chemotherapy + anti-HER2 ER−,PR− therapyd Basal-like 10–20% HER2− ER− & PR− 3 High Poor Chemotherapy for triple negative BC (ductal)

Notes: a: Based on prevalence data reported (Carey, Perou et al. 2006, Blows, Driver et al. 2010, Kennecke, Yerushalmi et al. 2010, Arvold, Taghian et al. 2011); b: According to St. Gallen International Expert Consensus 2011 (Goldhirsch, Wood et al. 2011); c: Inclusion and type of chemotherapeutic agents may depend on level of endocrine receptor expression, perceived risk and patient preference; d: Patients at very low risk e.g. pT1aN0 may be observed without systemic adjuvant treatment. Molecular features are detailed in (Lam, Jimenez et al. 2014). Abbreviations: BC, breast cancer; ER, oestrogen receptor; IHC, immunohistochemistry; PR, progesterone receptor; +, positive; -, negative. Sourced from (Lam, Jimenez et al. 2014). 26

Based on gene expression patterns for breast cancer classification, correlations have been made between breast cancer subgroups and survival (Blows, Driver et al. 2010, Fountzilas, Dafni et al. 2012) and metastatic spread (Kennecke, Yerushalmi et al. 2010), response to chemotherapy (Rouzier, Perou et al. 2005, Blows, Driver et al. 2010) and relapse (Fountzilas, Dafni et al. 2012). Importantly, the presence of ER which defines luminal breast cancer indicates a more favourable prognosis and treatment with anti-oestrogen/ hormonal therapy. The presence of HER2 amplification indicates a high-grade tumour with a relatively poor prognosis without treatment. However, the recent introduction of Herceptin (Trastuzumab), and other HER2 receptor targeted therapies, has largely improved the outcome for this group of patients. Triple negative breast cancers that do not express ER, PR or HER2 make up about 10% of all breast cancers. This group of tumours grow rapidly and spread faster than most other types of breast cancer. They are found predominantly in younger patients and are associated with BRCA1/2 mutations and familial breast cancer. They have a high risk of recurrence within 5 years and show a predominance of brain and lung metastases. No specific targeted therapy currently exists and chemotherapy is required, following surgical excision.

Tumour protein p53, is a well-studied marker in breast cancer. The protein is located inside the cell and senses damaged DNA and aids in regulation of DNA repair. The p53 gene that codes for a protein that functions as a tumour suppression and regulates the cell cycle by inducing cell cycle arrest until the DNA is repaired or apoptosis when the cell damaged and cannot be repaired (Yamashita, Toyama et al. 2006). Mutations in the tumour suppressor gene p53 are present in 15%–25% of primary breast carcinomas (Alsner, Yilmaz et al. 2000, Bai, Zhang et al. 2014). Mutant p53 cannot bind to DNA and stop the signal for cell division and the marker of its status has a significant predicting prognostic role. Tumour protein p63, a member of the p53 gene family, is involved in cellular differentiation and is expressed in the nuclei of myoepithelial cells of normal breast. Loss of p63 is used to identify early invasive breast cancer (Russell, Jindal et al. 2015).

Further demonstrating the complexity of breast cancer, the advent of next generation sequencing has led to comprehensive molecular portraits of human breast tumours. The genes and mutational processes involved illustrate the degree of heterogeneity which exists in breast cancer at the molecular level, with a few key driver mutations (Stephens, Tarpey et al. 2012, The Cancer Genome Atlas Network 2012).

CLINICAL ASPECTS

Clinical presentation

Most patients with breast cancer will present with a lump or other palpable abnormality, a nipple discharge which may be blood stained, or through a breast screening abnormality. Less commonly, patients may present with a lump, skin dimpling, changes in skin colour or texture, changes in appearance of the nipple shape, inverted nipple, rash or redness on the skin or around the nipple, lump or swelling in the axilla. Patients rarely present with disseminated malignancy or advanced metastatic disease and may have symptoms of bony pain or jaundice due to metastasis.

Population Screening

Early detection is vital as prognosis is closely related to tumour size. Larger tumours which are detected later in the natural history of the disease, have a higher incidence of local and distant metastases, usually requiring more substantial adjuvant therapies (Levenson 2007). The Clinical Oncology recommendations for breast cancer detection include: self-examination, physical examination by a doctor, mammogram and ultrasound followed by a biopsy, surgery and pathological examination of the tumour (Winer, Gralow et al. 2009). Mammography is the primary imaging modality for population-based breast cancer screening and early detection (Boyd, Martin et al. 2009, Giess, Frost et al. 2012), and has contributed to 28

the reduction in breast cancer mortality over the past two decades (Wright and McGechan 2003, Gotzsche and Nielsen 2006).

Mammography

Mammography screening, shown in Figure 1-14 is currently the best available approach for early detection of breast cancer and to reduce mortality from breast cancer, especially among women aged 39 to 69 years (Nelson, Tyne et al. 2009, Kalager, Zelen et al. 2010). The Australian breast cancer screening program targets women between 50 and 70 years of age to have mammograms every 2 years. Due to its success, this has now recently been extended to include patients’ up to 75 years old. Patients whose cancers are found early and treated in a timely manner are more likely to survive than are those whose diagnosis is delayed until they become symptomatic.

The availability of screening tests to detect cancers early provides better opportunities for patients to obtain more effective treatment with fewer side effects. Mammography can detect different abnormalities which include breast cancer, benign tumours, macro-and micro-calcifications (deposits of calcium) and cysts. It may be used either for screening or to evaluate possible abnormalities discovered by palpation. The mammographic images are analysed by a radiologist, with targeted tissue biopsy and histopathological examination of the tissue to determine the exact nature of the finding.

Ultrasound

Ultrasonography (sound waves used to create an echo image of underlying tissue) is the most effective procedure to diagnose small tumours in women with dense breast tissue and to differentiate solid lesions from cystic lesions (Flobbe, Bosch et al. 2003), and are thought to be better than mammography in dense breasts for detecting invasive breast cancer (Benson, Blue et al. 2004). The combination of mammography and ultrasound does significantly improve breast cancer detection.

Figure 1-14. Illustration of a mammogram examination.

Illustration demonstrates the mammogram procedure, whereby the breast is compressed using a dedicated mammography unit. Here parallel-plate compression evens out the thickness of breast tissue to increase image quality by reducing the thickness of tissue and holds the breast still. This reduces the dose of x-rays that must penetrate the tissue, decreasing the amount of scattered radiation and preventing image blurring. Source: http://health.learninginfo.org/mammogram.htm

Breast cancer survival has improved over the last few decades, due to screening programs and postoperative systemic adjuvant therapies (chemotherapy and hormone therapy). However, not all lesions are detectible with mammography or ultrasound particularly in young women who have dense breasts (Antman and Shea 1999). Diagnostic problems with mammograms can be divided into detection errors and errors of assessment and management (Giess, Frost et al. 2012). Increased breast density accounted for 67.6% of the decrease in mammographic sensitivity (Buist, Porter et al. 2004). A BRCA1/2 mutation and high breast density contribute to false-negative mammography results (Tilanus-Linthorst, Verhoog et al. 2002). Additionally, rapidly growing and high grade tumours are missed in screening programs as they can present clinically between the 2 yearly screens (Esserman, Shieh et al. 2009, Esserman, Shieh et al. 2011, Gotzsche, Nielsen et al. 2011).

In summary, the aim of population screening is to identify the cancer early and reduce breast cancer mortality. Though mammographic screening has been associated with a reduction in the death rate from breast cancer, this only accounts for a third of the total reduction in breast cancer mortality since the 1970s (Roder, Houssami et al. 2008, Kalager, Zelen et al. 2010). It has not reduced the incidence of advanced breast cancer (Jorgensen 2012).

Clinical Staging

The American Joint Committee on Cancer (AJCC), and the International Union for Cancer Control update the tumour-node-metastasis (TNM) cancer staging system used for breast cancer clinically staging (Edge and Compton 2010), and as a predictor of clinical outcome. The TNM classification Table 1-3 incorporates the T classification on the primary tumour size, N as presence or absence of LN involvement and M indicates presence or absence of metastases. This classification helps to guide the type of surgical treatment (mastectomy vs. breast-conserving surgery) and the need for adjuvant therapy using hormonal treatment, chemotherapy and/or radiation therapy. In addition, regardless of the tumour size, other features such as the direct extension of the tumour to the chest wall and/or to 31

the skin, may upgrade any T category to T4. Moreover, in-situ disease (DCIS or LCIS, designated as Tis) and Paget’s disease are also part of this system. Based on the combined score, clinical staging I to IV is designated.

Table 1-3. Clinical staging based on the combined TNM system.

Stage Tumour Node Distal Metastasis size Involvement

0 Tis N0 M0 IA T1 N0 M0 IB T0-1 N1 micro met M0 IIA T0-1 N1 M0 T2 N0 M0 IIB T2 N1 M0 T3 N0 M0 IIIA T0-2 N2 M0 T3 N1-2 M0 IIIB T4 N0-2 M0 IIIC Any T N3 M0 IV Any T Any N M1

Notes: Sourced from the publications AJCC cancer staging classification (Singletary, Allred et al. 2002, Woodward, Strom et al. 2003, Edge and Compton 2010). Abbreviations: T: classification on the primary tumour size (T1: <20 mm; T2: 20- 50 mm, and T3: >50 mm, largest dimension); N: number of LNs involved by metastatic disease (N1: 1-3 nodes; N2: 4-9 nodes, and N3: >10 nodes); M: stage referring to the presence of distal metastasis (M0 vs. M1, respectively).

Primary treatment

Conservative Surgery

Breast conserving surgery (local excision or lumpectomy), is currently the most common operation, to remove the cancer and a small amount of the normal tissue that surrounds it. The surgeon will also perform axillary sentinel LN biopsy to identify possible early metastatic deposits which will guide staging and management decisions.

Mastectomy

If local excision with clear margins cannot be obtained, or if the tumour is larger or multi-focal, then surgery to remove the whole breast (all breast tissue) may be required. Sometimes, the chest wall pectoral muscle is also removed if involved by carcinoma. If there is sentinel node involvement, the surgeon will usually perform a complete axillary dissection to remove all LNs.

Once excision of the tumour mass is completed and a formal histopathology report is issued, the patient is then clinically staged to help assess prognosis and further management.

Adjuvant and Neoadjuvant Therapy for Breast Cancer

Adjuvant therapy for breast cancer is any treatments given after primary surgical excision to attempt eradication of undetectable micro metastatic disease, which could potentially be present in all patients with invasive breast cancer and therefore increase the chance of long-term survival. Neoadjuvant therapy is the treatment given for large locally advanced tumours before primary surgical excision. Adjuvant therapy includes radiation therapy, hormonal therapy, chemotherapy, HER2 the targeted drug or a combination of treatments.

Radiotherapy

The current standard of care for patients with early-stage breast cancer consists of breast-conserving surgery, followed by 5-6 weeks’ postoperative 50 Gy radiotherapy (Fisher, Dignam et al. 1999, Julien, Bijker et al. 2000, Houghton, George et al. 2003, Bijker, Meijnen et al. 2006, Goodwin, Parker et al. 2009, Partridge, Pagani et al. 2014). Per unit weight of the organ or tissue is called absorbed dose and is expressed in units of gray (Gy).

Local breast radiation therapy is used after surgery to destroy any possible residual breast cancer cells in the breast or chest wall associated with local LN chains such as axilla, supraclavicular fossa or internal mammary. High-energy rays are used on the excision site and adjacent tissues only. The radiotherapy can be given two ways, usually external-beam radiation therapy or brachytherapy which is less common. External-beam irradiation, delivers a dose to the whole breast and the tumour bed, from outside the body. The treatment usually takes up to 6 weeks, requiring daily treatment for 5 days per week. Brachytherapy is administered inside the breast twice daily for a week, where the radioactive material is loaded into thin tubes inserted into the breast through a small incision. The radioactive substance is only loaded for a few minutes and then removed.

Women who get radiation therapy after lumpectomy have a 50 percent lower risk of breast cancer recurrence and a 20 percent lower risk of breast cancer death compared to women who get lumpectomy alone (Darby, McGale et al. 2011). Many women who have a mastectomy do not benefit from radiation therapy. However, in some cases radiation is used after mastectomy to treat the chest wall, the axillary and supraclavicular LNs. After mastectomy and axillary dissection, radiotherapy reduces both recurrence and breast cancer mortality in women with one to three positive LNs (McGale, Taylor et al. 2014).

Hormone Therapy

Hormone therapy also called anti-oestrogen treatment is a targeted drug treatment for hormone-receptor-positive (ER+) BC. These drugs act to either block the ER (Tamoxifen) or reduce oestrogen synthesis (aromatase inhibitors e.g. Anastrazole, Letrazole). They reduce the growth proliferative effect of oestrogen on breast cancer cells. Anti-oestrogen therapy is an effective treatment for ER+ breast cancer as it reduces recurrence by 50% and death by 30%. Oophorectomy (surgical removal of the ovaries), chemical or radiation is also an option for pre-menopausal women alone or in combination with Tamoxifen. The highest risk of recurrence for breast cancer patients is during the first five years following treatment and this is also affected by whether the breast cancer had spread to local LNs.

Chemotherapy

Chemotherapy may be given before or after surgery and utilises different combinations of drugs to kill dividing cancer cells. Numerous formulations exist which have evolved over the past 20 years but currently commonly include Adriamycin, cyclophosphamide and a taxanes. Unfortunately, this means that fast- growing cancer cells and normal cells will be damaged alike, leading to significant toxicity and side effects such as inflammation of the lining of the digestive tract, decreased blood cells production and hair loss. Although typically these problems disappear after treatment, there may develop long term problems such as heart injury and secondary malignancies such as leukaemia. A major clinical problem is often deciding exactly which patients, including those that are ER positive, who will benefit from chemotherapy as there are no biomarkers to predict responsiveness. For triple negative breast cancer, chemotherapy is the only adjuvant systemic option. OncotypeDX and PAM50 diagnostic tests, that analyses the activity of a panel of genes that can affect how a cancer is likely to behave and respond to treatment identify. They can be applied as a guide to predicting the benefit of adding chemotherapy to endocrine therapy for ER positive, LN negative, early breast cancer.

Targeted Therapy for HER2 positive cancer

Breast cancers that overexpress HER2 (human EGFR2), are often high grade and poor prognosis. Approximately, 10-15 % of all breast cancers are HER2 positive. These patients may receive targeted therapy to block the HER2 receptor. These medications include Trastuzumab (Herceptin) and Pertuzumab (Perjeta®) which are monoclonal antibodies (MAbs) that binds to the HER2/neu receptor (Nahta, Hung et al. 2004, Tripathy 2004). Ado-trastuzumab emtansine (Kadcyla™), a MAb attached to a chemotherapy drug and Lapatinib (Tykerb) a tyrosine kinase inhibitor (not an antibody), given orally. It was reported that one year of treatment with Trastuzumab after adjuvant chemotherapy, significantly improved disease-free survival among women with HER2-positive breast cancer (Piccart-Gebhart, Procter et al. 2005). The side effects of this targeted therapy include nausea, diarrhoea and vomiting and serious long term effects with breathing difficulties and heart failure. Other targeted therapies in preclinical studies often benefit only a small group of patients with the right mutations. Mechanisms of resistance to anti-HER2 therapies include over activity of the PI3K/AKT/mTOR pathway, IGF1R, the MET/MAPK pathways and Src (Gagliato, Jardim et al. 2016).

Summary of Clinical Aspects

In summary, most patients will present with a lump or mammographic abnormality. Breast cancer diagnosis and prognosis is based on the histological findings and tumour stage. Once the tumour is adequately excised and dependant on biomarker expression, treatment options will include endocrine therapy, chemotherapy and radiotherapy. In many cases, high grade cancer which has spread to LNs, requires aggressive treatment. Therefore, it is important to find new biomarkers for the early detection of breast cancer.

CURRENT PROTEOMICS STUDIES FOR BREAST CANCER

BIOMARKER

As previously discussed, current prognostic markers include age at diagnosis, tumour size and grade, LN involvement and receptor status: ER,PR and HER-2, and proliferation marker-Ki-67 (Penault-Llorca, Bilous et al. 2009, Gutierrez and Schiff 2011, Millar, Graham et al. 2011, O'Toole, Selinger et al. 2011). However, these markers alone are not sufficient for precise risk-group discrimination in breast cancer (Fitzgibbons, Page et al. 2000). Additionally, biomarkers are mostly sourced from blood. The main marker currently used for breast cancer monitoring is an increase in the serum cancer antigen CA 15-3. However, CA 15-3 measurement is not helpful in diagnosis, especially in patients with early stage cancers, and is not useful in the therapeutic decision-making of patients with breast cancer (Duffy, Duggan et al. 2004, Lumachi, Basso et al. 2004).

The detection of novel markers of breast cancer is the foundation for early detection and monitoring. Proteomics driven by MS is the key to the detection of proteomics patterns in urine, blood and tissue which could further lead to the development of new specific biomarkers for the detection of breast cancer. New molecular markers could potentially be used to identify small lesions, undetectable by usual screening techniques and provide an opportunity to treat breast cancer much earlier in its evolution, before it becomes capable of invasion and metastasis.

What is a Biomarker?

The biomarker definitions working group has defined the term biomarker (biological marker) as “a characteristic that is objectively measured and evaluated as an indicator of normal biological processes, pathogenic processes, or pharmacologic responses to a therapeutic intervention”(Atkinson, Colburn et al. 2001). However, it is important to note that a biomarker can only be defined when further verification and quantitation with a large number of samples have been examined. Depending on the application, biomarkers can be defined as: a) 37

diagnostic, often used for early detection of diseases; b) predictive, to determine the probable response to a specific therapeutic agent (such as therapeutic treatment) to patients; c) metabolic, to assess the drug response and toxicity in the human body; and d) outcome biomarkers, to predict progress and recurrence of diseases (Khleif, Doroshow et al. 2010). Alternatively, biomarkers can also be grouped according to where they originate: either in tissue or human body fluid markers. Compared to tissue biomarkers, human body fluid biomarkers such as urine have the advantages of accessible collection, low invasive risk and are much better tolerated and accepted for collection by patients.

What is proteomics?

Proteomics is the large-scale study of proteins that are expressed in a cell, tissue and organism and includes the study of protein structure and function, characterization of network pathways and signaling in relation to normal and disease states (Somiari, Somiari et al. 2005). Proteomic technology was reviewed in detail in our recent publication (Beretov, Wasinger et al. 2014), and can be applied to identify and decipher protein expression and proteome changes in relation to a disease development and progression (Hondermarck, Vercoutter-Edouart et al. 2001, Hanash and Hanash 2003, Wulfkuhle, Paweletz et al. 2003).

Proteins are vital parts of living organisms, as they are the functional components of the physiological pathways of cells. Proteins have diverse functions in the cell ranging from structural support, to being involved in bodily movement or defence against microorganisms. Proteins vary in structure (as shown in Figure 1-15) and are usually described and classified according to amino acid sequence, structure or function (Hensen, Meyer et al. 2012). They are responsible for maintaining homeostasis, so any changes in composition could allow for the identification of a disease process. Tissues, cell lines, primary cell cultures and body fluids such as plasma or cerebrospinal fluid are used as a source of proteins. Proteins derived from their parent genes often exist as one of several possible isoforms and can be further modified by glycosylation and phosphorylation. Thus, whilst there 38

are approximately 20,000 coding genes in the human genome there are hundreds of thousands of protein products in the body. Proteomics encompasses protein identification, protein ontology, protein–protein interaction, signalling pathways, along with quantification and functional network analysis. Proteins are particularly rich in biological information and interaction maps are the key to understanding the complex world of biological processes inside the cell (Mosca, Pons et al. 2013).

Figure 1-15. Schematic diagram of protein structure.

Depicted are the differing structures of proteins that have been associated with breast cancer: (A) Transthyretin, (B) Complement C3 and (C) EGFR. The institute of nano and molecular medicine has shown the structure of plasma protein Transthyretin which consisting of 127 amino acids that binds retinol and thyroxine. Sourced from (http://nanomed.missouri.edu/institute/research/Carborane.html); also complement C3, innate and adaptive immunity depend on the activation of the complement system, which consists of 1,560 amino-acid residues and has 12 domains that binds various proteins and receptors to effect its functions (Janssen, Huizinga et al. 2005); along with the complex structure of the EGFR (Zhang, Gureasko et al. 2006).

Proteomic research is also capable of improving our understanding of the molecular mechanisms underlying cancer and in the identification of novel proteins that

represent the presence of disease. In doing so, proteomics has developed into a key source of information in the study of human disease and when translated to the clinic, may offer improved detection, prognostic and predictive information to further guide treatment decisions in the age of precision medicine as gene expression signatures have done (van de Vijver, He et al. 2002).

Proteomics applications in breast cancer

Proteomic approaches based on the utilization of MS technologies, can be applied to urine (Beretov, Wasinger et al. 2014, Beretov, Wasinger et al. 2015), tissue samples, plasma, serum, saliva, nipple fluid, cerebrospinal fluid and cell lines. Technological advances in MS allow for the simultaneous analysis of thousands of proteins at one time, which in turn provides a wealth of potential new protein biomarkers. MS has been shown to be a sensitive, robust and quantitative technology to identify proteomics patterns in serum for ovarian cancer (Petricoin, Ardekani et al. 2002) and prostate cancer (Grizzle, Adam et al. 2003).

Proteomics research has shown that breast cancer profiles, along with novel proteins and peptides have been identified in tumours (Zhang, Tai et al. 2005, Ricolleau, Charbonnel et al. 2006, Sanders, Dias et al. 2008, Zhang, Tai et al. 2008, Sahab, Man et al. 2010, Weitzel, Byers et al. 2010, Davalieva, Kiprijanovska et al. 2012, Chung, Shibli et al. 2013), nipple aspirate fluid (Paweletz, Trock et al. 2001, Sauter, Zhu et al. 2002, Higgins, Matloff et al. 2005, Mendrinos, Nolen et al. 2005), breast cancer cell lines (Adam, Boyd et al. 2003, Hathout, Gehrmann et al. 2004, Geiger, Madden et al. 2012, Leong, McKay et al. 2012, Leong, Nunez et al. 2012, Boyer, Collier et al. 2013, Pavlou, Dimitromanolakis et al. 2013, Shaw, Chaerkady et al. 2013, Vergara, Simeone et al. 2013), in animal models (Sun, Zhang et al. 2008, Warmoes, Jaspers et al. 2012) and in plasma (Hanash, Pitteri et al. 2008, Cohen, Wang et al. 2013) and serum (Li, Zhang et al. 2002).

In last decade, advances in proteomic techniques and their application have made great progress in breast cancer research, and are very useful and promising in breast 40

cancer biomarker identification. However, until now a biomarker that can differentiate between normal subjects or BBD and breast cancer has not been found. Presently, there are no molecular biomarker which is sufficiently powered for use in current clinical practice for breast cancer screening or early detection (Tang and Gui 2012).Therefore, searching for novel biomarkers from human breast cancer samples, especially from urine using a non-invasive approach holds a great promise for future biomarker studies. The details of all proteomic techniques in use will be discussed in the following section.

KEY PROTEOMIC TECHNIQUES IN BIOMARKER RESEARCH

Proteomic technologies provide the ability to separate thousands of proteins & peptides simultaneously. It also offers a platform for the quantification and the identification of novel biomarkers which could ultimately be used for improving early detection and accuracy of diagnosis, determining the aggressiveness of breast cancer and monitoring efficacy of treatment. Applying this tool to urine and blood, is very promising for the screening of breast cancer biomarkers.

The predominant approaches for proteomics are two-dimensional gel electrophoresis (2DGE) and MS. Protein quantification can be achieved through isobaric tags for relative and absolute quantification (iTRAQ) or stable isotope labelling by amino acids (SILAC). Although, several proteomic techniques have been used to characterise the human urinary proteome, quantitative liquid chromatography- mass spectrometry (LC-MS) has already been used to identify over 2000 unique proteins. LC-MS/MS is the combination of two mass analysers in one mass spectrometry instrument. Pisitkun et al. applied ultracentrifugation and LC-MS/MS and identified 295 highly abundant unique proteins isolated from human urinary exosomes (Pisitkun, Shen et al. 2004). They further discovered a myriad of proteins and peptides using surface-enhanced laser desorption ionization time-of- flight (SELDI-TOF) and capillary electrophoresis (CE-MS) mass spectrometry system (Pisitkun, Johnstone et al. 2006). Sun et al. used three different multidimensional LC-MS/MS, approach on the human urine proteome and identified

226 soluble proteins, predominantly of low molecular weight (MW) proteins (Sun, Li et al. 2005). Using LC-MS/MS, Adachi et al. found that the human urinary proteome contains 1543 unique proteins of which a large proportion were membrane proteins (Adachi, Kumar et al. 2006). In 2009, the number of identified urine proteins was expanded to 1132 (Gonzales, Pisitkun et al. 2009). However, the understanding of the human urinary proteome is still incomplete.

In this section, we look at key proteomic technologies used in breast cancer including gel electrophoresis, liquid chromatography and MS and discuss their potential for breast cancer biomarker research. Current proteomic techniques applied to biological fluids for biomarkers discovery are summarised in Figure 1-16.

Gel-based techniques

Electrophoresis is used to separate complex mixtures of proteins, and to purify proteins for use in downstream applications. In biomarker research, combinations of gel-based and gel-free techniques are used. The most commonly used gel-based techniques are 2DGE, which can be used to separate and visualise thousands of proteins in a poly acrylamide gel matrix. In biomarker discovery, this technique can be used to detect protein spots which are altered in abundance between disease and control samples. This is then followed by identification of protein spots of interest using MS.

Figure 1-16. Summary of current proteomics technologies.

Abbreviations: 2D-DIGE, two-dimensional difference gel electrophoresis 2DGE, two-dimensional gel electrophoresis; CE-MS, capillary electrophoresis –MS; ICAT, isotope-coded affinity tags; iTRAQ, isobaric tags for relative and absolute quantitation; LC-MS, liquid chromatography–MS; MALDI-TOF-MS, matrix assisted laser desorption/ionization; MRM, multiple reaction monitoring; SDS-PAGE, sodium dodecyl sulfate polyacrylamide gel electrophoresis; SELDI-TOF, surface-enhanced laser desorption ionization time-of-flight; SILAC, stable isotope labelling by amino acids.

One dimensional gel electrophoresis (1DGE)

1DGE, is also called sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE), is commonly used for the separation of complex protein mixtures as a fast and simple method to identify abundant proteins. SDS-PAGE is applied to biological fluids that can have a proteome of hundreds of proteins. All the proteins migrate according to their mass in a polyacrylamide matrix and appear as bands in a one–dimensional separation. The higher the acrylamide concentration the smaller the gel pore size is, which determines the migration rate of the protein (Gallagher 2001). Disadvantages are that multiple proteins appear in each band and determining which one quantitatively is contributing can be difficult. The insolubility of hydrophobic proteins and the difficulty in the detection and separation of the low-abundant proteins are additional weaknesses (Petricoin, Belluco et al. 2006). In proteomic analysis, 1DGE is widely used coupled to MS-based protein identification (Paulo, Lee et al. 2011).

1DGE plus LC/MS/MS can be used to analyse numerous separation methods in sample preparation to determine the best method for the extraction and precipitation of urine (Sun, Li et al. 2005), to fractionate proteins before LC/MS/MS (Zoidakis, Makridakis et al. 2012), to assess protein profiles in human lymph and plasma (Clement, Aphkhazava et al. 2013) and nipple aspirate patterns in breast cancer (Brunoro, Ferreira et al. 2014).

Two-dimensional gel electrophoresis (2DGE)

2DGE was first reported for protein separation in 1975 by O′Farrell (O'Farrell 1975) and Klose (Klose 1975), and is still widely used today. In 2DGE, the proteins are separated in the first dimension by isoelectric focusing (IEF) whereby the proteins migrate to their isoelectric point in an immobilised pH-gradient. Then, the proteins migrate in the second dimension, by SDS-PAGE, based on their molecular masses (Fliser, Novak et al. 2007, Orenes-Pinero, Corton et al. 2007, Candiano, Santucci et

al. 2010). 2DGE provides a powerful method for the separation of complex protein mixtures and subsequent quantitative analysis and identification of proteins.

Traditional 2DGE is the most common technique used to discover and identify cancer-associated proteins, as this technique can provide more detailed information on the composition of proteins in urine. Using this technology, cancer biomarkers have been identified from urine in both urological malignancies (Rehman, Azzouzi et al. 2004, Irmak, Tilki et al. 2005) and systemic malignancies (Tantipaiboonwong, Sinchaikul et al. 2005, Yi, Chang et al. 2009). A large number of 2DGE-based approaches have been developed in pre-clinical breast cancer models, including cancer cell lines (Smith, Welham et al. 2007, Lee 2008, Zhou, Nitschke et al. 2008) and animal models (Li, Wang et al. 2006, Sun, Zhang et al. 2008). In addition, 2DGE has been applied to biological fluids including serum (Wulfkuhle, Sgroi et al. 2002, Pieper, Gatlin et al. 2003, Rui, Jian-Guo et al. 2003, Goufman, Moshkovskii et al. 2006, Kawate, Iwaya et al. 2015) and nipple aspirate fluids (Alexander, Stegner et al. 2004), tears (Molloy, Bolis et al. 1997, Evans, Vockler et al. 2001) and tumour tissues (Stastny, Prasad et al. 1984, Deng, Zhou et al. 2007, Semaan and Sang 2011) to identify breast cancer markers. Image analysis software can be used to evaluate fold difference based on spot volumes and a digital reference or master protein profile created.

One of the main advantages of using 2DGE is that a large number of proteins can be analysed generally within mass ranging from 20-220kDa (Lopez 2007), even if only a small sample (10 µg) is loaded onto a small gel. It is regarded as the most practical and useful separation technique because of its sensitivity, allowing high-resolution separation of proteins with good reproducibility. Therefore, it is commonly used as a screening tool to visually demonstrate the complexity of the sample. Proteins of interest can be excised from the gel, subjected to in-gel proteolysis and analysed by MS. Advances in this technology, allow efficient separation of low abundance proteins. In addition, all 2DGE profiles are already incorporated and compared within existing literature and databases (Expasy, SWISS-2DPAGE, UCD-2DPAGE Database, World-2DPAGE, and Siena-2DPAGE). 45

Unfortunately, certain proteins are difficult to separate with 2DGE, including very large or small proteins (<10kDa), extremely acidic (<3) or basic proteins (>9), very hydrophobic proteins, or proteins in low-abundance. Additionally, membrane proteins are difficult to identify and separate because they are difficult to solubilise. At least, triplicates of each sample are essential to generate robust results, limiting the ability to analyse low volume samples and increasing the time and labour required (Choe and Lee 2003). This technique involves a large amount of sample handling as it requires considerable technical skill, is time-consuming and therefore remains challenging mainly because of its low sensitivity and reproducibility. Even though 2DGE profiles are incorporated in existing literature and databases, without a standardised collection and preparation protocol comparing results is difficult. Furthermore, it is not an automated process and downstream technologies require the individual attention of each spot of interest.

Two-dimensional Difference Gel Electrophoresis

Two-dimensional Difference Gel Electrophoresis (2D-DIGE) is based on the classical 2DGE where the proteins being covalently labelled with fluorescent dyes (Cy3 and Cy5) prior to electrophoresis, are excited and emit at different wavelengths, which are then separated within the same gel based on charge and mass (Timms and Cramer 2008). This allows two unrelated samples and an internal control to be separated and detected together on one gel. The key for this technique is the normalization within an experiment via the inclusion of an internal control (Cy2) in all sample sets analysed which eliminate between-gel variations. In urine biomarker studies, 2D-DIGE has been used to establish a near-standard 2D human urine proteomic map (Oh, Pyo et al. 2004). This technique was applied to analyse the proteome of breast cancer cells (Lim, Choong et al. 2009, Ambrosino, Tarallo et al. 2010, DeAngelis, Li et al. 2011, Duru, Fan et al. 2012, Leong, McKay et al. 2012), tumour samples (Somiari, Sullivan et al. 2003), body fluids such as plasma (Michlmayr, Bachleitner-Hofmann et al. 2010) and saliva (Zhang, Sun et al. 2012), to screen biomarker candidates in breast cancer serum samples (Rui, Jian-Guo et al.

2003) and was utilised to distinguish between isoforms of serum proteins (Huang, Stasyk et al. 2006).

The main advantages of 2D-DIGE are the high mass range and number of proteins that can be analysed at any one time. This technique enables the simultaneous examination and comparison of 3 different samples in one gel, requiring fewer gels per study, eliminating technical replicates and therefore reducing the gel-to-gel variability (associated with 2DGE) improving the accuracy of quantitative protein profiling (Weeks 2010). Huang et al. demonstrated how this technique allowed the analysis of multiple tumour samples on the same gel, using different dyes and then aligning multiple gels to distinguish between isoforms of proteins (Huang, Stasyk et al. 2006). Unfortunately, this technique has a lower throughput as it is time- consuming and requires a high level of laboratory skill to obtain good results. Numerous steps are involved and it takes multiple days to complete. DIGE is also not a good technique for the analysis of extremely acidic, basic or hydrophobic proteins. The comparison of several different experiments using 2D-DIGE remains challenging.

Mass spectrometry based techniques in urine biomarker research

MS-based technology is one of the most powerful tools used in the last decade for the analysis of complex protein samples, with high analytical sensitivity and in a high-throughput fashion. This technique provides molecular information that cannot be gained from gel based techniques such as analysing proteins with extreme molecular mass/pI, and is complementary to them (Wolff, Otto et al. 2006). It is an important analytical tool in clinical proteomics, primarily in the disease-specific discovery, identification and characterization of proteomic biomarkers and patterns (Palmblad, Tiss et al. 2009).

A mass spectrometer consists of three basic components including an ionization source, a mass analyser and a detector(Aebersold, Mann et al. 2003). MS determines the mass/charge ratio (m/z) and the number of ions for each m/z value of a pressurised gas phase ion mixture. The predominant ionization techniques applied in cancer proteomics are matrix-assisted laser desorption ionization spectrometry (MALDI) (Karas and Hillenkamp 1988) for the analysis of simple peptide structures, and electro-spray ionization (ESI)(Fenn, Mann et al. 1989). For more complex samples, SELDI (Hutchens and Yip 1993) and LC-MS are applied. The two established MS-based strategies which have been widely adopted include: label-free quantitative proteomics or stable isotope labels. In the label-free MS approach, all samples are processed and analysed in parallel allowing the flexibility to conduct multiple comparisons. The total number of identified peptides corresponding to a certain protein (spectral count) or peptide ion intensity is used as a measurement for the relative quantity of a particular protein.

The four most common mass analysers are time of flight (TOF), quadrupole, quadrupole/ion trap and Fourier transform ion cyclotron resonance (FTICR) (Aebersold, Mann et al. 2003). There are two types of quantitation methods using MS: label free quantitation and labelling-based quantitation.

Label free MS quantitation

Researchers are increasingly turning to label-free shotgun proteomic techniques for faster, highly reproducible and accurate quantification that is less expensive. MS- based label-free quantitative proteomics falls into two general categories: the measurements of changes in chromatographic ion intensity (peptide peak heights) and spectral counting of identified proteins (Yang, Feng et al. 2011). Regardless of which label-free quantitative proteomics method is used, the main fundamental steps include: sample preparation (protein extraction, reduction, alkylation, and digestion), sample separation and analysis (LC-MS/MS), data analysis and peptide/protein identification, quantification and statistical analysis.

Chromatographic peak intensity (either peak area or peak height) is based on the observation that the ion current increases with increasing concentration of an injected peptide (Chelius and Bondarenko 2002). The assumption is that there is no variation in factors such as: temperature and pressure, retention time, sample preparation and volume injected. The spectral counting strategy compares the number of MS/MS spectra identified from the same protein across multiple LC-MS runs. Increasing the protein abundance, results in increased protein sequence coverage, which increases the number of unique peptides identified and the number of identified total MS/MS spectra (spectral count) (Duncan, Aebersold et al. 2010).

The advantages of label free techniques are it can be applied to different sample types which include urine, blood, tissue, cell culture and organisms and there is no cost for labelling reagents though this does increase the MS analysis time. Improvements in MS-based techniques have boosted the use of protein proﬁling techniques, with the application of SELDI-TOF-MS (Gast, Schellens et al. 2009) and MALDI-TOF-MS.

MALDI-TOF-MS

Matrix assisted laser desorption/ionization (MALDI-TOF-MS) is a vaporization and ionization methods in a single step, normally applied to analyse relatively simple protein mixtures. The sample is mixed with a solution of an appropriate matrix and allowed to co-crystallise directly onto specialised sample plates (Caprioli, Farmer et al. 1997). MALDI-TOF peptide mass fingerprinting has identified unique proteins in patients urine for renal cell carcinoma (Pieper, Gatlin et al. 2004), in prostate cancer (Rehman, Azzouzi et al. 2004), bladder cancer (Lin, Tsui et al. 2006, Welton, Khanna et al. 2010, Bryan, Wei et al. 2011) and non-Hodgkin lymphoma (Yoo, Kong et al. 2010) for diagnosis and monitoring disease progression. Additionally, this technique has been applied in breast cancer biomarkers studies, to serum (Liu, Sun et al. 2014), tissue (Casadonte, Kriegsmann et al. 2014) and cell lines (Jeon, Kim et al. 2013).

Main benefits of MALDI are that it is automated, very fast with high sensitivity and low cost small and sample volumes are required. This technique is multidimensional, gives absolute mass measurements and works well with large polypeptides (>30kDa). However, there are several problems associated with MALDI analysis that include sensitivity to contaminants such as salts and finding appropriate matrices to work with. The biggest disadvantage is that sometimes ions may collisionally relax and this creates problematic data analysis, which in turn affects reproducibility.

SELDI-TOF

The surface-enhanced laser desorption/ionization-time of flight-mass spectrometry (SELDI-MS) is an ionization method whereby the sample is spotted on the SELDI surface, often a chemically modified surface on a biochip, to absorb energy and allow vaporization and ionization by the laser for further MS detection. The impurities are removed by washing with buffer and the sample is analysed with a TOF spectrometer. It is used to detect proteins in tissue, blood and urine samples. The complexity of a biological sample is reduced by selective interactions of polypeptides with different active surfaces whereby the SELDI surface is modified with a chemical (a strong anion exchanger and antibodies) to make it hydrophobic. After the interaction phase, some of the proteins in the samples will bind to this surface, depending on the pH, concentration, salt content, and the presence of interfering compounds like lipids others which are removed by washing. Ideally, this enables the generation of mass spectra (and consequently information on polypeptides) from highly complex samples by elimination of known interfering compounds.

Published data have shown that this technique has been widely used with numerous reports on biomarkers for a variety of diseases and was applied to urine for protein profiling and methodology (Schaub, Wilkins et al. 2004). SELDI-TOF-MS technology offers a unique platform for high throughput urine protein profiling (Schaub, Wilkins et al. 2004). It requires very small sample volumes (2 µL) (Zhou, Huang et

al. 2004, Mischak, Julian et al. 2007), can detect small peptides (approximately 500 Da) (Grus, Podust et al. 2005) along with low MW, truncated, and modified or fragmented proteins and peptides (Petricoin and Liotta 2004), with high sensitivity and specificity (Lebrecht, Boehm et al. 2009). By using the respective chemical properties of each array it is possible to focus the analysis to either negatively or positively charged proteins or to target specific metal binding or phosphoproteins.

SELDI-TOF-MS technique, has been used to investigate novel and improved biomarkers in breast cancer from serum (Hu, Zhang et al. 2005, Li, Zhao et al. 2005, Opstal-van Winden, Krop et al. 2011), nipple aspirate fluid (Li, Zhao et al. 2005, Pawlik, Fritsche et al. 2005, Sauter, Shan et al. 2005, He, Gornbein et al. 2007, Noble, Dua et al. 2007), tumour tissues (Carter, Douglass et al. 2002, Ricolleau, Charbonnel et al. 2006), tears (Lebrecht, Boehm et al. 2009) and cell lines (Goncalves, Charafe- Jauffret et al. 2008). SELDI-TOF analysis of serum in breast cancer studies has identified specific discriminatory pattern in invasive ductal breast cancer (Belluco, Petricoin et al. 2007), low MW fragment patterns for early breast cancer detection (Petricoin and Liotta 2004), familial protein profiles (Garrisi, Tommasi et al. 2013) and potential novel biomarkers (Chung, Moore et al. 2014). Furthermore, this technique has been applied to newly identified biomarker profiles to predict LN metastasis and recurrence-free survival in high-risk primary breast cancer (Nakagawa, Huang et al. 2006, Gast, Zapatka et al. 2011) as well as to monitor treatment efficacy in metastatic disease (Pusztai, Gregory et al. 2004).

The major drawback of SELDI-TOF-MS is that the results are biased towards peptides and smaller proteins (proteins <30kDa). It is also problematic in detecting larger MW proteins and post translation modification (PTM) identification. Other down falls of SELDI-MS approach include the low-resolution (De Bock, de Seny et al. 2010), ion suppression and quantification of individual proteins when protein composition is complex. Although the system is quite simple to use, it is prone to generate artefacts (Schaub, Wilkins et al. 2004) and due to different chip surfaces reproducibility of data sets are low making comparison difficult (Mischak, Julian et al. 2007, Najam-ul-Haq, Rainer et al. 2007). Although SELDI-TOF-MS offers a unique 51

platform for high throughput urine protein profiling, there is a lack of comparability and therefore standardization of analysis conditions which is critical for accurate data interpretation and comparison (Schaub, Wilkins et al. 2004).

Capillary electrophoresis–mass spectrometry

CE-MS is a proteomic technology combining chromatographic techniques with MS to enhance the resolving ability, and is a widely used approach for the proteomic analysis of urine. The CE separates the proteins with high resolution based on their migration through a buffer-filled capillary column in an electrical field (300-500 V/cm) in a single step. CE-MS offers several advantages, it is a useful platform for peptide analysis as it is multidimensional, fast and highly sensitive and requires only small sample volumes (Johannesson, Wetterhall et al. 2004, Theodorescu, Fliser et al. 2005). It is compatible with most buffers and analytes (Hernandez-Borges, Neususs et al. 2004, Kaiser, Kamal et al. 2004), uses inexpensive capillaries (Kolch, Neususs et al. 2005), the CE can be interfaced with almost any MS and can be reconditioned with NaOH after each run.

CE-MS technology was applied to determine all the peptides and proteins in human urine (Wittke, Fliser et al. 2003), metabolic profiling (Ramautar, Nevedomskaya et al. 2011), identifying urinary metabolites in breast cancer (Yu, Jiang et al. 2013) and in general to develop a standardised processing protocol (Thomas, Sexton et al. 2010). CE-MS enables sequence analysis via MS/MS with platform-independent sample separation which facilitates the independent entry of different sequencing platforms for peptide sequencing of CE-MS-defined biomarkers from highly complex mixtures (Zurbig, Renfrow et al. 2006). Additionally, this tool has been applied for the discovery of potential biomarkers from various biological specimens (Gamagedara and Ma 2011) in a wide range of diseases (Theodorescu, Wittke et al. 2006, Coon, Zurbig et al. 2008, Wu, Chen et al. 2010).

Unfortunately, CE-MS is not suitable for the analysis of high MW proteins (>20kDa) as they tend to precipitate in the acidic buffers that are generally used for CE-MS

analysis. Also, as only small sample volume can be loaded onto the capillary, this has shown to cause lower selectivity compared with LC-MS (Mischak, Coon et al. 2009).

Liquid chromatography mass spectrometry

Liquid chromatography mass spectrometry (LC-MS) is an analytical chemistry technique that combines the physical separation capabilities of liquid chromatography with the mass analysis capabilities of MS. It enables the direct analysis of very complex protein mixtures through the separation of small quantities of peptides and the identification of peptides at the rate of many thousands of sequences per day. This proteomic technique enhances the resolution by separating large amounts of analytes on a highly sensitive LC-column. It is multidimensional and highly sensitive, and can be used to detect large MW molecules after tryptic digestion.

Tandem mass spectrometry (or MS/MS), is the preferred method for protein identification and is performed on peptides derived by trypsin digestion of proteins. LC-MS/MS is most commonly used for proteomic analysis of complex samples, where peptide masses may overlap even with a high-resolution mass spectrometer. Liquid chromatography is used as a liquid mobile phase to separate the components of a mixture. These analytes are present in a liquid phase or dissolved in a solvent, and then forced to flow through a chromatographic column usually under high pressure referred to as high-performance liquid chromatography (HPLC) in the column, where the mixture is resolved into its components. Peptides are eluted in as small a volume as possible (as signal intensity in a mass spectrum is directly proportional to analyte concentration) which is achieved by using chromatographic columns usually 50–150μM inner diameter. Such columns can be loaded with low microgram amounts of total peptide and are run at 100–500 nl /min, which should produce peptide peaks with widths of 10–60 s (Steen and Mann 2004).

The HPLC system separates the chemicals using conventional chromatography on a column, usually by reverse phase chromatography. The protein binds to the column

by hydrophobic interactions in the presence of a hydrophilic solvent (for instance water) and is eluted off by a more hydrophobic solvent (methanol or acetonitrile). At the end of the column, it enters the mass detector, where the solvent is removed and the proteins are ionised (detector can only work with ions, not neutral molecules). Only ions fly through a vacuum, so the removal of the solvent is a vital first step. A full high-resolution spectrum is produced by the mass detector once it scans the molecules by mass as shown in Figure 1-17.

During this process, the prepared samples of interest are injected onto an HPLC column that comprises a narrow stainless steel tube (usually 150 mm length and 2 mm internal diameter) packed with fine, chemically modified silica particles. The compounds are then separated on the basis of their relative interaction with the chemical coating of these particles (stationary phase) and the solvent eluting through the column (mobile phase). Components eluting from the chromatographic column are then introduced to the mass spectrometer via a specialised interface. The two most common interfaces used for LC/MS are the electrospray ionisation and the atmospheric pressure chemical ionisation interfaces (Grebe and Singh 2011).

Figure 1-17. Schematic diagram of shotgun LC-MS/MS analysis.

Applying MS/MS analysis the sample is trypsinised, injected into the mass spectrometer, ionized and accelerated, then analysed by mass spectrometry (MS1). Ions from the MS1 spectra are then selectively fragmented and analysed by mass spectrometry (MS2) to give the spectra for the ion fragments.

Application of label-free quantitative LC-MS was applied to urine for proteome analysis (Weeks 2010, Court, Selevsek et al. 2011), to create comprehensive human urine standards for comparability and validation (Mischak, Kolch et al. 2010), for biomarker discovery and quantification (Pisitkun, Johnstone et al. 2006, Cutillas 2010, Thomas, Sexton et al. 2010) and in establishing a standardised preparation protocol (Sun, Li et al. 2005, Mischak, Kolch et al. 2010, Thomas, Sexton et al. 2010, Court, Selevsek et al. 2011). Liquid chromatography, quantitative analysis of the normal urinary proteome revealed that a high proportion of the proteins are secreted, membrane, and relatively high MW proteins (Nagaraj and Mann 2011).

To date, very few breast cancer studies have looked at the LC-MS-based approaches for the protein profiling of urine. LC -MS was used to profile breast cancer using urinary RNA metabolites (Henneges, Bullinger et al. 2009) and for urinary biomarker studies for nucleosides in breast cancer (Hsu, Lin et al. 2011). This technique had been applied to detect the modified proteins in breast cancer cells (Zhu, Kim et al. 2003), modified urinary nucleosides as tumour markers (Liebich, Muller-Hagedorn et al. 2005) and serum markers of LN metastasis (Wang, Su et al.). Its application to breast cancer markers in blood includes looking at low abundance proteins (Meng, Gormley et al. 2011, Such-Sanmartin, Bache et al. 2015), glycoproteome changes (Zeng, Hincapie et al. 2010) and putative peptides biomarker (van den Broek, Sparidans et al. 2010), as well as proteome and transcriptome profiles in the mouse model (Schoenherr, Kelly-Spratt et al. 2011).

LC-MS/MS has seen enormous growth in clinical laboratories in last 10–15 years because it offers analytical specificity superior to that of immunoassays or HPLC for low MW analytes and has higher throughput than GC-MS such that in the USA reference/referral laboratories, most steroids and biogenic amines are now assayed by LC-MS/MS (Grebe and Singh 2011). LC is a very powerful fractionation method whereby the LC-columns can separate large amounts of analytes with high resolution (Aebersold, Mann et al. 2003). Multidimensional LC systems are usually chosen to separate serum/plasma samples into different fractions, based on 55

different principles such as pI or hydrophobicity, thus decreasing the complexity of the samples with the subsequent increase in the probability of identifying biomarkers at lower concentrations (Luque-Garcia and Neubert 2007). More sophisticated mass detectors such as triple quadrupole and ion-trap instruments can be set up to carry out more detailed structure-dependent analyses on the eluent from the HPLC system. LC/MS with slow elution times (2h) nano-flow LC and high accuracy mass determination (<1ppm) makes identification based solely on fingerprint and pharmacokinetic studies have demonstrated it is both accurate and reliable in the analysis of low concentrations of proteins in blood (Zhang, Bast et al. 2004, Choi, Lee et al. 2012)

The disadvantages with the LC-based systems are that large amounts of information are obtained, so it is time consuming to analyse. It is not suited for the separation of larger molecules and analytes covering a broad range of size and hydrophobicity. Additionally, this technique can be far more sensitive towards interfering compounds (e.g., lipids or detergents, large molecules) that may precipitate and/or adsorb to the column (Liu, Chen et al. 2009). Therefore, as the chosen proteomics technology, sample preparation protocol was critical to our breast cancer biomarkers investigation.

MS/MS Data to Identified Proteins

To achieve amino acid sequence data, the peptides are first separated by MS and then peptides of one particular mass are selected for collision-induced dissociation (CID) in a collision cell where a small amount of inert gas is present. These precursor ions are fragmented with sufficient energy to break the peptide bonds which are generally the weakest bonds, and these peptide fragments are detected to produce an MS/MS spectrum. The actual sequence of the peptide can be determined as all amino acids (except leucine and isoleucine) have unique masses. Proteins identification is facilitated by using the MS/MS data to interrogate protein sequence databases to achieve matches (Brewis and Brennan 2010). The process of peptide sequencing by MS/MS that enables protein identification is shown in Figure 1-18.

Global identification and quantification of proteins

According to the Brewis and Brennan report, the process of LC-MS/MS required to achieve protein qualification from cells or biological fluids is shown in Figure 1-19. The workflow required proteins solubilisation and commences with enrichment steps: Subcellular fractionation of the cell samples and protein depletion (immunodepletion methods to largely remove up to 20 of the most abundant proteins) of biological fluids in order to analyse the lower abundance proteins. The proteins can be separated by two-dimensional electrophoresis (2DE), where gels spots are excised, trypsin digestion and resulting peptides separated by nanoscale LC before MS/MS. The LC-MS/MS data is used to search existing protein databases to achieve a match and therefore a protein identification based on amino acid sequences typically derived from multiple MS/MS spectra. The method of choice for 2DE protein quantification is 2D-DIGE, whereby samples are labelled with different fluorescence dyes. The second method for protein identification is to first separate proteins by one-dimensional electrophoresis (1DE) before subjecting individual protein bands to digestion and LC-MS (the GeLC-MS workflow). Thirdly, in gel-free LC-MS proteomics method, the sample is trypsin digest and peptides are separated by LC on the basis relative hydrophobicity and often also charge as a multidimensional separation. One of the advantages of this protein identification workflow is that it is possible to achieve quantitative data at the same time by introducing a peptide labelling step, such as iTRAQ or using labelled peptides as internal standards or to perform protein labelling, such as SILAC, without the need for additional LC-MS. Label-free MS quantification can also be performed using spectral counting or ion current measurements (Brewis and Brennan 2010).

Figure 1-18. Peptide sequencing by MS/MS for protein identification.

This diagram was taken from Brewis and Brennan 2010. As shown: (A) the peptides are initially detected by MS and the peptides of a distinct mass are selected for MS/MS (the precursor ion), (B) in this example, a 1733.799-Da peptide was subjected to collision-induced dissociation (CID) to achieve peptide fragmentation and the MS/MS spectrum, (C) a diagrammatic representation of MS/MS fragmentation of peptides due to peptide bond cleavage. Shown is a five amino acid peptide (amino acid side chains denoted by R) with the amino acids connected by peptide bonds (CO-NH) and the R5 amino acids is either arginine or lysine for trypsin digestion (Brewis and Brennan 2010).

Figure 1-19. Strategies for global protein identification and quantification.

LC-MS is at the centre of all of these workflows showing how proteins may be derived from cells or biological fluid (Brewis and Brennan 2010).

Peptide analysis using MS technology

The mass spectrometer used in the proteomics analysis is a key component to convert peptides into data about the peptide sequences present. Though a variety of MS technologies are available to analyse peptides, for peptide sequence identification, tandem mass spectrometers are mostly used. MS- technology is continually evolving to create more capability and opportunity in proteomics (Domon and Aebersold 2006). Hybrid mass spectrometers incorporating high resolution, high mass accuracy analysers such as the Fourier transform ion cyclotron resonance (FTMS) and Orbitrap (also a form of FTMS) have also become incredibly useful for proteomics (Chen and Yates 2007). MS/MS can be performed in the linear ion trap. The LTQ Orbitrap Velos is a high performance LC-MS system, combining rapid LTQ ion trap data acquisition with high mass accuracy Orbitrap mass analysis. It includes a higher-energy C-trap dissociation (HCD) and electron transfer dissociation (ETD), that are complementary to the LTQ’s collision induced dissociation (CID) option. In an Orbitrap mass analyser, ions are trapped in an electrostatic field between an inner and outer electrode (Makarov 2000). As the ions rotate around the inner electrode, they process along its axis with a frequency characteristic of their mass-to charge (m/z) ratio. Acquisition of transients and the Fourier transformation of that signal yields frequencies and their intensities. A simple relationship converts frequencies into m/z values. The LTQ-Orbitrap hybrid mass spectrometer is a high mass accuracy, high resolution mass analyser, capable of achieving resolving power > 100,000 and having mass accuracy ≤ 5 ppm (Makarov, Denisov et al. 2006). The distinct advantages are its high sensitivity, resolution and mass accuracy, coupled to a fast scan rate.

Multiple reaction monitoring

Multiple reaction monitoring (MRM) requires the MS to be set up to monitor only specific mass/charge ratio (m/z) values of interest and therefore allows the detection of low level peptides. It is performed on a triple quadrupole or linear ion- trap MS. These instruments have the ability to select a parent (MS1) peptide ion of

the target protein based on m/z, subject them to collision-induced dissociation, and monitor selected fragments (MS2) product ions, again based on m/z. The peptides are chosen based on their high abundance and repeated identification in mass spectral runs, as well as sequence specificity for the protein of interest. The control peptides are proteotypic and served as representations of the intact protein of interest (Fusaro, Mani et al. 2009, Kitteringham, Jenkins et al. 2009). Other criteria for selection of candidate peptides include peptide length and the absence of missed cleavage sites.

MS-based MRM assays show superior multiplexing detection capabilities (Anderson and Hunter 2006, Keshishian, Addona et al. 2007, Keshishian, Addona et al. 2009, Kuzyk, Smith et al. 2009). Configuration starts with the selection of “signature” peptides that act as surrogates for the quantification of protein of interest. This approach does not rely on antibodies. In addition, it can also facilitate the measurement of PTM, which is a difficult task for antibody-based systems (Unwin, Griffiths et al. 2005, Mayya, Rezual et al. 2006, Hulsmeier, Paesold-Burda et al. 2007).

MRM is capable of quantifying low molecular mass (<1kDa) peptides (Gergov, Ojanpera et al. 2003), low levels of small molecules including drugs or metabolites (Anderson and Hunter 2006, Luna, Williams et al. 2008) and multiple proteins during a single MS analysis (Anderson and Hunter 2006, Addona, Abbatiello et al. 2009, Keshishian, Addona et al. 2009). In breast cancer research, recently applied label-free MRM-MS technique has been included to identify and validate potential breast cancer biomarkers in an animal model (Whiteaker, Zhang et al. 2007), human blood (Guo, Gu et al. 2008), human tissues for diagnosis of metastasis (Metodieva, Greenwood et al. 2009, Sprung, Martinez et al. 2012), and to investigate the role of phosphorylation stoichiometry (Domanski, Murphy et al. 2010, van den Broek, Sparidans et al. 2010, van den Broek, Sparidans et al. 2010). Most urine studies with MRM are focused on preclinical studies using a combination approach with other proteomic technologies for searching biomarkers or for verification of identified biomarkers (Fang, You et al. 2010). 61

MRM is highly reproducible, and a very sensitive label-free technique for quantifying targeted protein/peptides. It offers superior multiplexing capabilities, allowing for the simultaneous quantitation of numerous proteins in parallel and is considered to be a very selective and sensitive MS scan function (Thomas, Sexton et al. 2010). It can identify and confirm the presence of precursor and product ions of the anolyte of interest and isotopically labelled internal standards allow absolute quantitation (Addona, Abbatiello et al. 2009). MRM has been considered to be one of the most effective tools available for quantitative clinical proteomics (Han and Higgs 2008, Yocum and Chinnaiyan 2009). For human breast cancer urine and blood biomarker research, this technique could be useful for the verification of identified proteins, peptides or metabolites for breast cancer early diagnosis and monitoring progression.

Unfortunately, different MRM-based quantification results have been reported on the same target proteins by different groups (Keshishian, Addona et al. 2007, Fortin, Salvador et al. 2009, Kuzyk, Smith et al. 2009), which may be due to incompetent trypsin digestion of parental ions (Kuzyk, Smith et al. 2009), different ionisation of individual peptides caused by instrument or interfering molecules (the presence of high abundances molecules in the sample) (Addona, Abbatiello et al. 2009), or peptide modifications (Addona, Abbatiello et al. 2009). Interfering substances in the mixture can affect the accuracy of MRM quantitation, therefore appropriate fraction preparation is preferable for analysing mixtures containing high abundance proteins such as plasma or serum (Addona, Abbatiello et al. 2009) and certainly a standardised method for sample preparation if data is to be compared.

Label-based MS quantitation

In the protein labelling approaches, the different protein samples are combined together and once the labelling is finished, the pooled mixtures are taken through the sample preparation step before being analysed. The labelling based quantitative profiling of proteins via MS includes isotope-coded affinity tags (ICATs) (Gygi, Rist 62

et al. 1999) and iTRAQ (Ross, Huang et al. 2004, Bouchal, Roumeliotis et al. 2009). The advantage to differential labelling is that only one sample (healthy and diseased mixture) is processed and analysed by MS/MS. Limitations include increased time and complexity of sample preparation, the high cost of the labelling reagents, incomplete labelling, and the requirement for specific quantification software.

Isotope Coded Affinity Tag

Isotope Coded Affinity Tag (ICAT) reagents are specific for cysteine residues of denatured proteins. There are two versions of the reagent, the light (or normal form) and the heavy (or deuterated form). The deuterated form uses deuterium to replace a hydrogen atom in the normal form. Because there are only two versions of the reagents, only two samples can be compared at a time. In this technique, the extracted proteins from test and control samples are labelled with either light or heavy ICAT reagents by reacting with cysteinyl thiols on the proteins. Peptides containing the labelled and unlabelled ICAT tags are recovered by avidin affinity chromatography and are then analysed by LC-MS/MS.

Protein labelling with ICAT followed by MS/MS is suitable for both large-scale analysis of complex samples (no human biological fluid studies to date) including whole proteomes and small-scale analysis of sub-proteomes, and allows quantitative analysis of proteins, including those that are difficult to analyse by gel- based proteomics technology (Pisitkun, Bieniek et al. 2006). The ICAT technique has been widely used for protein identification and quantification in breast cancer in preclinical studies (Pawlik, Hawke et al. 2006, Kang, Ahn et al. 2010). This application cannot be used to analyse proteins that do not have a cysteine residue, accounting for around 10% of the total proteome. In addition, ICAT can only identify a maximum of 300-400 proteins, far fewer than the 2DGE method, and the peptides contain large labels, which makes database searching more difficult, especially for short peptides. Until now, only one report has demonstrated that comparative urinary protein profiling was analysed using the ICAT isotopic labelling based LC-MS/MS approach in a murine model of renal injury (Braun, Li et al. 2006).

Isobaric tag (iTRAQ)

Isobaric tag for relative and absolute quantitation (iTRAQ) is a shot-gun based proteomic technique, also referred to as bottom up approach, which allows the concurrent identification and relative quantification of hundreds of proteins in up to 8 different biological samples in a single experiment (Ross, Huang et al. 2004, Pierce, Unwin et al. 2008). Digested samples are labelled with the 8-plex iTRAQ reagents (113, 114, 115, 116, 117, 119, 120. 121) and the sample is pooled and prepared for MS/MS (Ross, Huang et al. 2004).

Quantitative proteomics applying iTRAQ is a promising approach for global comparison of protein expression in relatively small amounts of samples. It provides both quantitation and multiplexing in a single reagent, and is suited for biomarker applications. Several authors have reported predictive markers from breast cancer cell lines using iTRAQ (Ho, Kong et al. 2009, Leong, Nunez et al. 2012, Coumans, Gau et al. 2014, See, Chong et al. 2014, Johansson, Sanchez et al. 2015), animal models (Choong, Lim et al. 2010), serum (Opstal-van Winden, Krop et al. 2011) and tumour tissues (Bouchal, Roumeliotis et al. 2009, Muraoka, Kume et al. 2012) for diagnosis, metastasis and drug response. Validation of biomarkers requires relative amounts of a peptide, MS/MS info and absolute quantitation.

The iTRAQ proteomic technology provides multiplexing capacity with high throughput, with high sensitivity and specificity therefore reducing the analysis time. It has potential for future breast cancer biomarker research for early diagnosis and monitoring progression. The protein sequence coverage obtained using iTRAQ reagents are similar to that obtained using other shotgun proteomic approaches. Additionally, it is useful for identifying and quantifying proteins across diverse MW and pI ranges, functional categories, cellular locations and abundances. Unfortunately, this technique is very time consuming, extremely laborious and very expensive.

SILAC

Stable Isotope Labelling by Amino Acids (SILAC) is another predominant method for protein quantification, which involves the incorporation of isotopically labelled amino acids (lysine and arginine are the two most commonly used) via growth medium, into proteins as they are synthesised by the growing mammalian cell lines. This method was originally developed by Matthias Mann and his colleagues (Ong, Blagoev et al. 2002) and allows for several comparison within one experiment using labels13C and 15N. Disadvantages are that SILAC does not work well with plant cells and some mammalian cells are harder to grow in the dialyzed serum required for SILAC, due to the loss of essential growth factors (Ong and Mann 2005, Harsha, Jimeno et al. 2008).

MS technology along with developments in protein bioinformatics, will improve our biological understanding of the protein components of biological systems and provide key insights into the composition, regulation and function of molecular complexes and pathways (Cravatt, Simon et al. 2007). In addition, the identification and quantification of proteins provide powerful information for improved biomarker discovery (Aebersold and Mann 2003, Yates, Gilchrist et al. 2005, Domon and Aebersold 2006).

PROTEOMICS TECHNIQUES AND URINARY BIOMARKERS IN

BREAST CANCER

Medicine today has greatly advanced with improvements in knowledge about disease pathophysiology (along with the detection and treatment) which has been mirrored by advances in biomarker development. Modern advances in the field of proteomic techniques hold the promise of providing new tools to find novel breast cancer biomarkers for early diagnosis and prognosis. Biomarkers are required for early diagnosis, prediction of disease progression, and response to treatment.

Urine and blood are the two most commonly used human body fluids and has been subjected to intense investigation using current trends in biomarker research. Advances in proteomics, especially in MS (Yates, Ruse et al. 2009, Pan, Chen et al. 2011) have rapidly changed our knowledge of urine proteins which have simultaneously led to the identification and quantification of thousands of unique proteins and peptides in a complex biological fluid (Thongboonkerd, McLeish et al. 2002, Adachi, Kumar et al. 2006). Proteomic studies of urine are more informative than serum and tissues, as urine has a less complex composition, and has been successfully used to discover novel markers for cancer diagnosis and surveillance (Husi, Stephens et al. 2011, Hassanein, Callison et al. 2012, Lei, Zhao et al. 2012), as well as for monitoring cancer progression (Linden, Lind et al. 2012, Zoidakis, Makridakis et al. 2012).

In this section, a review current breast cancer biomarker and proteomic research, along with a comprehensive summary of urine analysis for biomarker detection was provided. The advantages of urine as a source for biomarkers studies, along with collection and sample preparation techniques will be discussed.

Urine as a potential source for biomarkers

Urine is a waste product generated by the filtering of blood in the kidneys. It is estimated that 150-180 litres of plasma are filtered in a day, resulting in the excretion of proteins, peptides and metabolites. Therefore, urine will contain information not only from the urinary system but also from other parts of the body. Analysis of pooled urine from healthy individuals (both male and female) showed that 70% of the urine proteins originate from the urinary system and the remaining 30% represent proteins filtered by the kidney (Davis, Spahr et al. 2001, Thongboonkerd and Malasit 2005). The high percentage of non-urinary system proteins strongly suggests that urine can be a useful resource for biomarker discovery for a systemic disease such as breast cancer.

With the development of proteomic techniques, the number of proteins identified in urine have increased from 124 in 2001 (Spahr, Davis et al. 2001), to 295 in 2004 (Pisitkun, Shen et al. 2004) using a LC-MS to 1500 proteins in 2006 using an Orbitrap MS (Thermo Fisher Scientific, Waltham, MA, USA) (Adachi, Kumar et al. 2006). In addition, more than 5,000 peptides have been identified in urine using proteomics technology CE-MS (Coon, Zurbig et al. 2008).

These data indicate that novel proteomic techniques provide the potential for identifying more breast cancer biomarkers for diagnosis and prognosis. Cancer biomarkers could be members of the immune system, metabolites, and fragments of tumour-associated antigens (TAAs), oncofetal proteins, adhesion proteins, acute phase proteins and others after filtration through urine. These markers can be detected by modern proteomic techniques.

Advantages of urine as breast cancer biomarker

The difficulty in obtaining human body fluid samples is a major problem which hinders the progress of searching for biomarkers. Urine has become one of the most widely used clinical samples for biomarker discovery, due to its less complex composition (Pisitkun, Johnstone et al. 2006, Thongboonkerd and Thongboonkerd 2008). The total protein concentration in urine excreted from a normal adult is low, less than 150 mg/day, and is typically not greater than 10 mg/100 mL in a single specimen (Adachi, Kumar et al. 2006).

The composition of urine is an accurate indicator of the health or disease state of the whole body. In addition, the larger sample volume available with urine allows for repeating analysis, with the introduction of varied technology and validation studies, of the same sample that will in turn allow improvement in the accuracy of research results. Furthermore, studies have shown that the urine proteome is stable for up to 6 hours(hr) at room temperature (RT) (Schaub, Wilkins et al. 2004), and for several years at -20 °C (Decramer, Gonzalez de Peredo et al. 2008). It was reported that urine can be stored at -70 °C for >17 years and still allow high-quality 67

datausing gel-based proteomics (Thongboonkerd 2007), allowing possible follow up investigation on the original samples.

Urine collection, storage and sample preparation

The urine sample needs to be collected and handled in the correct standard manner and the sample needs to be undergone precipitation and concentration to provide a purified urine protein sample for proteomic analysis.

Factors affecting urine collection and handling

The collection and handling of a urine specimen is crucial to protein analysis as the urine sample is affected by intrinsic and extrinsic factors: Intrinsic factors affect the sample include blood concentration and direct sample collection. A random urine specimen is probably the easiest collection method, but often not the first choice, as the sample could be too dilute depending on the patients’ diet, hydration status and varied collection time which in turn affects results. Traditionally, the first morning urine collection is quite often chosen as it has an increased concentration of proteins and has shown the greatest number of total proteins using 2DGE. However, it was reported that it provided the smallest number of spots on gel analysis (Thongboonkerd, Chutipongtanate et al. 2006). Although a 24 hr sample collection would demonstrate what is happening in the body throughout the day, the first morning specimen showed the least variation when compared to the 24 hr sample (Thomas, Sexton et al. 2010). Using SELDI-TOF- MS technology, it was demonstrated that there were significant differences in protein ratio between the first void and midstream samples (Schaub, Wilkins et al. 2004). However, the majority of experts in this area have suggested that a clean, second morning void is preferred over the first morning sample and 24 hr collection (Mischak, Allmaier et al. 2010, Vaezzadeh, Briscoe et al. 2010). Therefore it is correct to conclude that a random, midstream collection, blood free, in a clean, sterile container should be the preferred method of collection. Once collected, it is essential to first remove cell debris with low speed centrifugation (2000x g) for 10 min at RT 68

(Thongboonkerd and Thongboonkerd 2007, Thomas, Sexton et al. 2010), to minimise protein release from insoluble contents.

The effects of extrinsic factors on urinary proteins have been examined to determine the best method of storage to minimise any protein degradation. Extrinsic factors include urine storage post void, the effect of temperature and pH on sample storage along with the freeze-thaw cycles. Using SELDI-TOF-MS, Schaub et al found that there were no changes in the protein profile (midstream urine samples) after three days at 4 °C (Schaub, Wilkins et al. 2004). However, Zerefos and Vlahou reported changes in data using both 2DGE and MALDI-TOF-MS even when the samples were stored at 4 °C for 24 hours and that these changes were reduced for shorter storage times (< 6 hours) at 4 °C (Zerefos and Vlahou 2008). Using gel electrophoresis, others reports have also shown the damaging effect of long term storage at -20 °C (Zhou, Yuen et al. 2006). Thongboonkerd et al. demonstrated that urine can be stored at -70 °C for >17 years and still allow high-quality datausing gel-based proteomics (Thongboonkerd 2007). The differences derived from studies may be due to: using a variety of proteomic approaches, urine sample preparation method, equipment, experimental conditions and the skills of performers. In proteomic urine research, the storage of urine samples at -20 °C for several weeks is the most common laboratory practice. Despite the numerous studies conducted, there is still no agreement on how stable urinary proteins are and how they should be stored.

Numerous groups have recommended pre-analytical treatment with a protease inhibitor (Mischak, Kolch et al. 2010), along with the adjustment of sample pH to neutral, for sample preservation (Kania, Byrnes et al. , Havanapan and Thongboonkerd 2009), although neither are proven essential. Levels of some urinary proteins change in 4-7 freeze-thaw cycles (Thongboonkerd 2007) whilst others report that repetitive freeze-thaw cycles had minimal impact at either the protein or peptide level in a manner that degrades information obtainable by MS (Schaub, Wilkins et al. 2004, Lee, Monigatti et al. 2008) even after several years at - 70 °C (Parekh, Kao et al. 2007). From this evidence, it seems that proteins in urine are quite stable and most of the proteolytic degradation by endogenous proteases is 69

completed during the storage time in the bladder. Once collected, all handling of the urine sample must be on ice and storage of samples at 4 °C should be for no more than a few hours and then frozen (Parekh, Kao et al. 2007). Further to this, vortexing the sample will ensure that all the proteins are fully resuspended before analysis (Zhou, Yuen et al. 2006).

Summary of urine collection & handling

Currently, no single standard urine collection approach is used by researchers. An accepted “Standard Protocol” for adult urine collection and storage would be a significant starting point to keep urine handling and preparation consistent for data comparison between groups. A urine collection protocol in agreement to the “standard for proteomic analysis” is under development by Human Urine and Kidney Proteome Project (HUKPP) and European Urine and Kidney Proteomics (EuroKUP) (www.eurokup.orgorwww.hukpp.org).

In the literature, there are common themes found for urine collection, cell removal, processing and storage to guide the development of a standard procedural method, though nothing has been established as a standard protocol. The take-home message is that it is essential to plan the collection and handling of the urine specimen before the commencement of the study to ensure that all sample handling and treatment is consistent. Essentially, samples should be stored within hours of collection and kept cold during handling and then frozen in smaller volume aliquots, to reduce the freezing and thawing times (Kentsis, Monigatti et al. 2009). Checking and recording the pH of the sample before freezing and after thawing, along with freezing and thawing times are helpful for proteomic analysis and comparative analyses. Urinary pH and protein tests should be determined before storage to ensure consistent reliable results. The recommendations for a standardised handling and preparation method will be discussed in chapter 3.

Pooling urine samples for protein analysis

Several factors affect the quality of urine proteomic analysis. Normally, the standard 100-200 µg of protein from urine is loaded for 2DGE and MS (Thongboonkerd, Mungdee et al. 2009, Thomas, Sexton et al. 2010). Therefore, precipitation and concentration is essential and the technique applied should be very carefully considered in the experimental design.

Analysing individual samples in proteomic experiments requires a great deal of resources. Consequently, the majority of proteomic studies apply the pooled urine sample approach. Pooling the samples reduces the amount of work and resources required which in turn reduces the cost along with the complexity and length of time for the analysis of the data (Adachi, Kumar et al. 2006, Yang, Chu et al. 2011). In pilot studies with small groups of urine samples, pooling the samples within disease and non-disease states is a good way to estimate the prevalence of disease in the population whilst at the same time saving on cost, especially with MS samples. Pooling samples also dilutes the biological variance which is a hindrance as it requires too many samples to overcome the great natural variation in the population and needs to highlight only those peptides/proteins that are altered for the majority of the population. However, the main drawback of pooling samples is the inability to detect individual change in biomarker discovery.

De-salting urine samples and protein purification

Using the appropriate method of separation will result in a sufficiently high protein yield. Therefore, a further factor to consider is that salt in urine is a major problem for proteomic studies, especially as it generates detrimental high current and heat. The presence of high-salt content, non-soluble matter, aggregates and proteoglycans and other interfering compounds makes high resolution 2DE separation of proteins difficult and hinders the detection of low abundant proteins. By de-salting the urine sample, filtering and removal of interfering components, resolution on 2D gels can be improved, allowing better quantitative recovery and 71

the detection of lower-abundance proteins (Thongboonkerd and Thongboonkerd 2007) along with improved reproducibility. For that reason, sample protein purification is essential for the efficient application to urinary proteomics and several different methods of desalting urine have been examined. The desalting of samples is essential for the success of many analysis techniques such as 2DGE, MS, capillary electrophoresis and anion exchange chromatography and various methods which are currently used. Ultra-filtration is a common procedure applied in sample preparation (Pieper, Gatlin et al. 2004), to concentrate and desalt urine samples and to remove the insoluble components with minimal protein loss (Wittke, Fliser et al. 2003).

Precipitation and concentration of urine

Protein precipitation is an important factor in proteomic analysis, which is based on the solubility of the protein and involves the use of various organic solvents at different concentrations (Thongboonkerd, McLeish et al. 2002, Sun, Li et al. 2005). Analysis using 2DGE has demonstrated that the yield of protein recovery was greater with high percentage organic compounds (Thongboonkerd, Chutipongtanate et al. 2006, Sigdel, Lau et al. 2008, Crosley, Duthie et al. 2009). Urine proteomic analysis requires the precipitation of proteins to provide a concentrated protein pellet and in most studies the precipitation is performed using organic solvent: ethanol, acetone, and methanol. Thongboonkerd et al. reported that precipitation using acetonitrile demonstrated the greatest number of spots in 2DGE urine analysis (Thongboonkerd, Chutipongtanate et al. 2006). Sun el al. used acetone to separate and precipitate the low molecular mass urinary proteins for 2DGE. In addition, numerous studies performing 2DGE with MS have used a combination of acetone precipitation and trichloroacetic acid (TCA) to yield the best quality protein extraction yield (Oh, Pyo et al. 2004, Fung, Yip et al. 2005, Tantipaiboonwong, Sinchaikul et al. 2005, Tyan, Guo et al. 2006, Zerefos and Vlahou 2008, Gonzales, Pisitkun et al. 2009, Kentsis, Monigatti et al. 2009, Magistroni, Ligabue et al. 2009). Furthermore, for LC-MS analysis, applying organic solvent precipitation followed by in-solution tryptic digestion was superior to in gel digestion and achieved rates of 72

higher peptide identification (Court, Selevsek et al. 2011) as in-solution tryptic digestion is an emerging technique.

An important step in urine proteomic analysis is protein concentration. Commonly applied techniques to concentrate the urinary proteins include; ultra-filtration and ultracentrifugation (often for 30-60 min at very high speeds, 11,000 x g), a combination of acetone precipitation and ultra-filtration, TCA and ultracentrifugation (Pisitkun, Shen et al. 2004, Fung, Yip et al. 2005). Ultrafiltration simultaneously concentrates and desalts solutes, and involves the separation of particles from fluid according to their molecular size, using a filter membrane which allows water and salts to pass through. Vaezzadeh’s group employed the one-step sample preparation method, which requires reduction and alkylation along with the use of Vivaspin 6 spin-filters (column) and anti-human serum albumin to deplete the albumin (Vaezzadeh, Briscoe et al. 2010). Some groups use reverse-phase C18 to simultaneously desalt and concentrate their samples/digests prior to analysis by LC-MS (Rappsilber, Ishihama et al. 2003). Others use reverse-phase separation columns to separate the sample and affinity column or beads for protein enrichment (Castagna, Cecconi et al. 2005, Tantipaiboonwong, Sinchaikul et al. 2005). The MicroFlow MF10 is a pre-fractionation device for low volume, low abundance complex samples that can also enrich for very specific species of proteins based on charge and/or size either in native or denaturing format, as well as desalt and deplete samples of contaminating ions or proteins demonstrated with plasma and urine proteins (Wasinger, Ly et al. 2008). It was reported that the pre-fractionation and simplification of biological samples prior to MS can show an increased depth of coverage using the MF10, electrophoretic partitioning systems (Kiernan, Tubbs et al. 2003). This device should be useful for urine sample concentration and desalting for proteomic analysis in the future.

In addition to the above factors, urinary pH may also affect the proteomic analysis result as urine pH can vary from pH 4.5 - 8.0. This variation in pH between patients could be due to diet and pathological reasons and may indicate an underlying condition which could inevitably interfere with the protein study, therefore patient 73

history is essential. Although with the one-step sample preparation using off-gel electrophoresis, adjustment to neutral pH was essential (Vaezzadeh, Briscoe et al. 2010). It was noted that the adjustment was not essential for 2DGE analysis (Thongboonkerd, Klein et al. 2004, Thongboonkerd, Mungdee et al. 2009).

Challenges in urine biomarker research

Proteomic MS involves the breakdown and examination of both molecular and cellular processes at the protein level. Although the development of proteomic techniques has accelerated the discovery of potential urine biomarkers, analysis of urine samples also presents several challenges. The major challenges include a lack of standardised procedures for sampling, verification and quantification methods. As urine proteomic profiles alter with various collection methods, it is important to use a standard operating procedure for biomarker research and comparison. However, to date there is no agreement on standard collection method for urine sampling and purification.

Biomarker validation is also important. The identification of proteins in proteomic studies depends on identification of a unique set of peptides and proteins using m/z peak lists generated from MS techniques, within protein databases. Therefore, it is important to confirm findings in an independent group of samples using different quantitative techniques such as enzyme-linked immunosorbent assay (ELISA) or MRM-MS. The second challenge is the size of the patients’ pooled urine samples needed for biomarker discovery. With a small sample size and no follow-up verification to determine specificity and sensitivity, it is almost impossible to determine the biomarker potential of these proteins. The third challenge is that as the majority of published breast cancer biomarker studies are metabolomics studies from urine (Table 1-4), very few groups have looked at proteomic protein markers in breast cancer.

Owing to the huge dynamic range of proteins present, urine specimens can show a high degree of variability in protein concentration (particularly in the case of kidney 74

damage or dysfunction) and total protein excreted from person to person and from the same individual collected at different times (Goligorsky, Addabbo et al. 2007, Crosley, Duthie et al. 2009), all of which have stalled the development of urinary biomarkers. However, with appropriate exclusion criteria and control groups as well as development of new proteomic equipment, these drawbacks can be minimised.

Potential breast cancer biomarkers in urine

Identification of novel biomarkers in urine for breast cancer diagnosis and monitoring progression is a recently developed research area. Most of the publications have used MS technologies to analyse urine metabolomic breast cancer biomarkers which could be useful to improve diagnostic sensitivity and accuracy. Although, potential urine breast cancer metabolomic biomarkers have been identified using proteomic techniques (detailed in Table 1-4), few have been independent validated. Carter et al. demonstrated a correlation between breast cancer patients’ clinical characteristics and urinary Tamoxifen metabolite profiles using non-aqueous capillary chromatography/electro spray/mass spectrometry (nCE/ESI/MS)(Carter, Li et al. 2001). Using the LC-MS method, absolute quantities of 15 endogenous oestrogens and their metabolites in urine were measured in pre- and post-menopausal women, that demonstrated patterns which were thought to influence the risk of breast cancer (Xu, Veenstra et al. 2005). Applying column- switching liquid chromatography–electro spray/tandem mass spectrometry (LC- ESI/MS/MS), oxidized nucleosides measures in urine provided further understanding to breast cancer pathogenesis (Cho, Jung et al. 2006). Using the newly developed ultra-performance (UPLC/MS-MS) methodology, Gaikwad et al. compared oestrogen metabolites to their respective conjugates and found they were significantly elevated in the high risk breast cancer patients, indicating that de- purinating oestrogen-DNA adducts are possible early biomarkers for breast cancer detection (Gaikwad, Yang et al. 2008). Further analysis of those samples and a larger

Table 1-4. Potential breast cancer urine biomarkers identified by proteomic technologies.

Markers Technique Urine sample types Purposes of study References

Allantoin and four F2- LC/MS-MS BC (n=23) Oxidative status for drug (Il'yasova, Kennedy et al. isoprostanes resistance/ toxicity 2011)

Estrogens & metabolites LC/MS-MS BC (n=17), high risk BC (n=12), Early diagnosis (Gaikwad, Yang et al. 2008, control (n=46) Faupel-Badger, Fuhrman et al. 2010)

Estrogens & metabolites UP-LC/MS-MS BC (n=40), high BC risk (n=40), Early detection of BC risk (Gaikwad, Yang et al. 2009) control (n=40)

Estrogens & metabolites LC/MS-MS Control for BC (n=362 pre- Population Study for (Faupel-Badger, Fuhrman et menapausal+168 hormonal carcinogenesis al. 2010) postmenopausal)

Estrogens metabolitesa LC-MS pre- and postmenopausal BC screening via (Xu, Veenstra et al. 2005) women (n=20) estrogen metabolites

Estrogens metabolitesb GC-MS or LC-MS BC (n=10), control (n=22) BC diagnosis and (Woo, Kim et al. 2009) stratification

Estrogens metabolitesc GE-MS BC (n=50), control (n=50) Diagnosis of BC (Nam, Chung et al. 2009)

Nucleosides LC-ESI/MS/MS BC (n=30), control (n=30) Understanding (Cho, Jung et al. 2006) pathogenesis

Nucleosides LC-IT-MS BC (n=113), control (n=99) Diagnosis and (Frickenschmidt, Frohlich et classification of BC al. 2008)

Nucleosides: cytidine, 3- HPLC/EI-MS-MS BC (n=36) BC biomarkers for early (Hsu, Lin et al. 2011) methylcytidine & inosine detection

Polyamines LC-MS/MS BC (n=30), control (n=30) Monitoring BC (Byun, Lee et al. 2008) progression

Tamoxifen & metabolites nCE-ESI-MS BC (n=47) Staging BC (Carter, Li et al. 2001)

Notes: Oestrogen metabolites; a. metabolites (2-, 4-methoxy, 2-, 4-,16 alpha-hydroxy derivatives, 2-hydroxyestrone-3-methyl ether; oestradiol& 2-, 4-methoxy & 2-, 16alpha-hydroxy derivatives, 16-epiestriol, 17-epiestriol, and 16-ketoestradiol); b. metabolites (5- hydroxymethyl-2-deoxyuridine and 8-hydroxy-2-deoxyguanosine); c. (Homovanillate, 4-hydroxyphenylacetate, 5- hydroxyindoleacetate & urea). Abbreviations: BC, breast cancer; GC-MS, gas chromatography-mass spectrometry; LC-ESI/MS/MS, liquid chromatography–electro spray/tandem mass spectrometry; LC-IT-MS, liquid chromatography ion trap mass spectrometry; nCE/ESI/MS, non-aqueous capillary chromatography/electro spray/mass spectrometry; SPE/UPLC/MS-MS, ultra-performance liquid chromatography/tandem mass spectrometry.

sample number provided quantification data useful to assess individual risk of developing breast cancer and to determine appropriate risk reducing strategies (Gaikwad, Yang et al. 2009).

Using liquid chromatography and an ion trap mass spectrometer (LC-IT-MS) technique, Frickenschmidt et al. identified urinary ribonucleosides as potential biomarkers to diagnose breast cancer (Frickenschmidt, Frohlich et al. 2008). Through the analysis of polyamine in urine and serum, Byun et al. developed a quantitative LC-MS/MS method to monitor breast cancer progression, four serum polyamines (1, 3-diaminopropane, putrescine, spermine and N-acetylspermidine) were shown to be increased in breast cancer patients (Byun, Lee et al. 2008). Using gas chromatography-mass spectrometry (GC-MS) or LC-MS, Woo et al. found that two known biomarkers (5-hydroxymethyl-2-deoxyuridine and 8-hydroxy-2- deoxyguanosine) were increased in breast cancer urine (Cho, Jung et al. 2006). These breast cancer metabolomic biomarkers identified are useful for diagnostic tools and patient stratification, but may be mapped on metabolic network to reflect disease states.

Nam et al. combined tissue transcriptomics and metabolomics for breast cancer biomarker identification using GC-MS approaches, and found nine altered metabolic pathways and identified four metabolic biomarkers (Homovanillate, 4- hydroxyphenylacetate, 5-hydroxyindoleacetate and urea) as different in breast cancer versus normal subjects (Nam, Chung et al. 2009). In a population-based case-control study of breast cancer, Faupel-Badger et al. reported that LC-MS/MS was a promising approach to test urinary oestrogen metabolites (EM) as it was more sensitive than radioimmunoassay’s (RIA) and ELISA (Faupel-Badger, Fuhrman et al. 2010). In addition, it was reported that LC-MS/MS could be used to identify breast cancer patients who are especially susceptible to drug resistance and/or drug toxicity by assessing the oxidative status of urinary levels of allantoin and four F2- isoprostanes (Il'yasova, Kennedy et al. 2011). One recent study indicated that analysis of urinary nucleosides (cytidine, 3-methylcytidine, and inosine) by high- 78

performance liquid chromatography/electrospray ionization tandem mass spectrometry (HPLC/EI-MS-MS) were found to be potential tumour markers in breast cancer (Hsu, Lin et al. 2011).

Amongst all the breast cancer urine biomarkers studies using proteomic techniques, only one study has tested for sensitivity and specificity (Frickenschmidt, Frohlich et al. 2008). None of the urine biomarkers identified in breast cancer have been verified in an independent group of patients. If identified biomarkers are not validated, there will be no clinical significance. Validation of the identified proteins in tissue and urine from breast cancer patients is paramount. This validation process determines whether breast cancer or its microenvironment are the source of the protein detected in urine and whether these markers can be used in the detection and treatment of breast cancer patients. Future studies in urine and proteomic research should be carefully assess sensitivity and specificity as well as validation of identified results in independent groups of breast cancer patients. The protocol developed for urine sample preparation and novel findings from LC-MS/MS analysis will be discussed in detail in Chapters 3 and 4, respectively.

BIOMARKERS IN BREAST CANCER

Prognostic markers currently applied to assess breast tumours using IHC, associated with targets and proposed application are shown in Table 1-5. Hormone receptor status of a tumour (ER and PR) are markers of and predict the likely response to hormone therapy in both early and advanced breast cancer. Tumours with ER, PR expression are usually associated with improved prognosis compared to ER, PR negative tumours. HER2 is used in determining prognosis and treatment with Trastuzumab (Harris, Ismaila et al. 2016).

There is emerging interest in proteomics studies for the expression of proteins in tumour tissue, urine, blood and tears, as a means to identify potential novel markers

Table 1-5. Prognostic markers currently applied to the assessment of breast tumour.

Tissue Target Updated Proposed Application a Based Marker ER and PR Protein Prognostic markers , predicting response to hormone therapy (Tamoxifen)

HER-2 Protein/DNA Prognostic markers, predicting response to Herceptin and other HER2 therapies

Protein/DNA Possibly predict response to adjuvant anthracycline -based therapy

Ki-67 Protein Prognostic markers, predicting chemotherapy response

uPA/PAI-1 Protein Possible prognostic markers and predicting response to chemotherapy

Oncotype mRNA derived Possible prognostic markers, predict the need to DX test algorithm of 21 treat with chemotherapy for ER+ LN – treated genes with Tamoxifen

Notes: Experts opinion of the updated application at aSt Gallen Conference (Goldhirsch, Wood et al. 2003, Goldhirsch, Wood et al. 2007, Goldhirsch, Wood et al. 2011). Abbreviations: BC, breast cancer; ER, Oestrogen receptor; IHC, immunohistochemistry; PR, Progesterone receptor; HER2, Human epidermal growth factor receptor 2 status; LN, Lymph node status.

for diagnostic and therapeutic targets in women with breast cancer. The goal for breast cancer proteomics research today, is to identify biomarkers for the early detection of breast cancer using easily accessible body fluids, such as urine and blood. A plasma and serum protein profile to differentiate between disease and normal states may provide unique signatures, which alongside current screening methods could greatly improve detection.

Serum and plasma are often the preferred sources in biomarker studies that are performed for breast cancer. Blood is a rich source for protein information as it is actively engaged in the response to the physiological and pathological processes of the human body. Currently, there are several novel blood biomarkers available (these will be discussed in detail in Chapter 1.10), although none to date are being used in the clinic for the early detection of breast cancer.

Blood biomarkers for breast cancer

Blood markers offer the potential for the early detection of cancer, to identify different types of cancer present and to monitor and evaluate therapeutic response. Blood is an ideal source of biomarkers due to its ease of collection, availability and the protein information it holds regarding the whole body rather than a specific compartment. Circulating proteomic biomarkers are secreted by the body tumour or non-tumour source are part of a systemic reaction into blood, which makes it a promising fluid for disease diagnosis and therapeutic monitoring.

A search of the literature in early 2015, using MEDLINE(TM) for “serum” and “plasma” and “proteomics” reveals 3700 journal papers describing various approaches for the identification of blood proteins. When this search is narrowed down to include breast cancer, only 174 papers were found. In 2002, two- dimensional (2D) gels and MS contributed to the list of 289 proteins reported in plasma (Anderson and Anderson 2002). With improvements in proteomics technology, came the ability to profile the blood peptidome which provided 81

information about cancer and potential for early-stage cancer diagnosis (Petricoin, Belluco et al. 2006).

The Food and Drug Administration (FDA) federal agency of the United States Department of Health and Human Services regulates the use of tumour markers for monitoring, screening, prognostic and staging. This equivalent agency in Australia is “The Therapeutic Goods Administration”. An updated list of serum cancer biomarkers used in primary care, are shown in Table 1-6, based on the literature(Opstal-van Winden, Rodenburg et al. 2012). The main FDA approved serum markers related to breast cancer, presently practised are mucin glycoproteins MUC-1 and cancer antigens CA15-3, CA-125 and carcinoembryonic antigen (CEA). Numerous serum studies have supported their use for breast cancer detection (Arslan, Serdar et al. 2000, Norum, Erikstein et al. 2001, Lumachi, Basso et al. 2004, Ludwig and Weinstein 2005, Molina, Barak et al. 2005, Sturgeon, Duffy et al. 2008, Kulasingam, Zheng et al. 2009, Samy, Ragab et al. 2010, Zhang, Hu et al. 2013), claiming that CA 15-3 and CA27.29 are the best characterised serum markers related to breast cancer (Duffy 2006, Mathelin, Cromer et al. 2006), while others do not recommend them for diagnostic use due to low sensitivity (Drake, Cazares et al. 2011).

Other potential blood biomarkers of breast cancer include transformation- associated protein osteopontin (OPN) in plasma (Bramwell, Doig et al. 2006, Pietrowska, Marczak et al. 2009) and serum (Fedarko, Jain et al. 2001), whereby high levels in serum and plasma have been detected in patients with metastatic disease (Wai and Kuo 2004, Cook, Tuck et al. 2005, Rodrigues, Teixeira et al. 2007, Shevde, Das et al. 2010). Cancer antigens as biomarkers include matrix metalloproteinase (MMPs), which are critical for angiogenesis and tumour invasion (Zeng, Ou et al. 2014), along with haptoglobin, prolactin, CA19-9, leptin and migration inhibitory factor (MIF). Additional biomarkers of breast cancer recurrence include levels of lysosomal cysteine proteases cathepsin B and cathepsin L, which have been positively correlated to relapse-free survival and overall survival after treatment of primary breast tumour (Levicar, Kos et al. 2002). 82

Table 1-6. Serum biomarkers used for breast cancer diagnosis.

Blood Biomarker Cancer Application References

CA-125* Monitoring advanced BC & (Norum, Erikstein et al. 2001) ovarian cancer

CA15-3* Monitoring marker of advanced (Ludwig and Weinstein 2005, BC Sturgeon, Duffy et al. 2008, Zhang, Hu et al. 2013)

CA19-9* Monitoring advanced pancreatic, (Ludwig and Weinstein 2005, Park, gastrointestinal cancer Choi et al. 2009) Increased serum levels in BC (Kim, Lee et al. 2009)

CA27-29* Monitoring breast disease

CEA* Monitoring advanced BC (Ludwig and Weinstein 2005, Sturgeon, Duffy et al. 2008) Prognostic marker of recurrence (Zhang, Hu et al. 2013)

EGF* Risk marker for BC (Pitteri, Amon et al. 2010)

Haptoglobin Increased serum levels in BC (Hamrita, Chahed et al. 2009, Tabassum, Reddy et al. 2012)

Leptin Plasma levels altered in BC (Tessitore, Vizio et al. 2000, Pietrowska, Marczak et al. 2009, A marker of BC progression Maccio, Madeddu et al. 2010) Blood marker of increased BC risk (Garofalo, Koda et al. 2006) Serum levels associated with BC (Wu, Chou et al. 2009) (Mohammadzadeh, Ghaffari et al. 2014)

Migration Increased in breast tumours (Bando, Matsumoto et al. 2002) inhibitory Increased serum levels in BC (Xu, Wang et al. 2008, Verjans, factor Noetzel et al. 2009)

Osteopontin Increased levels in BC in blood & Plasma (Pietrowska, Marczak et al. tumour samples 2009), serum (Fedarko, Jain et al. 2001) and breast tumour (Rudland, plasma markers of disease Platt-Higgins et al. 2002) progression (Bramwell, Doig et al. 2006, Tuck, Chambers et al. 2007)

Prolactin Risk marker for BC (Tworoger, Eliassen et al. 2007, Faupel-Badger, Sherman et al. 2010, Tikk, Sookthai et al. 2014)

Abbreviations: BC, breast cancer; CA15-3, cancer antigen 15-3; CA19-9, cancer antigen 19-9; CA-125, cancer antigen 125; CEA, carcinoembryonic antigen; MIF, migration inhibitory factor. OPN, osteopontin; EGF, Epidermal growth factor receptor. FDA approved*(Ludwig and Weinstein 2005)therapeutic effect in breast cancer.

However, these tumour markers are not recommended for diagnostic and therapeutic purpose, due to their low sensitivity and specificity (Mathelin, Cromer et al. 2006, Harris, Fritsche et al. 2007, Drake, Cazares et al. 2011, Opstal-van Winden, Rodenburg et al. 2012).

Additionally, serum cytokeratin fragment 21.1 (CYFRA 21.1) was suggested as a tumour marker for breast cancer (Nakata, Ogawa et al. 2000). Originally, CYFRA 21.1 assay identified a serum fragment of cytokeratin 19 (CK19), which can be used to detect lung cancer (Giovanella, Ceriani et al. 2002). Application of this assay in breast cancer demonstrated that CK19 fragments were expressed in the primary and metastatic lesions and that CK19 mRNA in peripheral blood (Giovanella, Ceriani et al. 2002), and measurement of the serum CYFRA 21-1 titre were used to monitor recurrence and evaluating the patients with advanced disease (Nakata, Ogawa et al. 2000).

Serum & Plasma proteomics

Important information is provided by blood proteins that come from the body in response to the tumour present or which are directly secreted by the tumour cells. Non-invasive biomarkers for the early detection of breast cancer are urgently needed, as currently there are no established clinical blood biomarkers. Proteomics is a rapidly developing field that can explore the heterogeneity of breast cancer and supplement the wealth of information gained from genomics. This analysis technique has been used to differentiate cancer profiles from benign profiles in samples from sera, plasma, tissue, nipple fluid, and ductal lavage. 84

Plasma Vs Serum

Blood is a major component of the human body which is used to transport the blood cells, fluid and electrolytes around the body. It is a very complex biological fluid that contains carbohydrates, lipids, amino acids, nucleic acids, hormones, vitamins, metabolites and proteins (Hanash, Pitteri et al. 2008).

Whole blood is separated into three layers which include plasma, red blood cells, buffy layer (leukocytes, and platelets) as shown in Figure 1-20. Plasma makes up 55 % of the blood volume and consists of 92% water, 8% blood plasma proteins, and trace amounts of other materials. Plasma transports various ions (Na+, Ca2+, HCO3−), glucose and amino acids to and from the cells as required, along with organic acids, hormones, cholesterol, urea and waste. Important components of plasma include serum albumin, blood-clotting factors to facilitate coagulation, immunoglobulins, lipoprotein particles, other proteins and various electrolytes such as sodium and chloride. Serum is essentially plasma from which all clotting proteins are removed by clotting of the blood, and contains all the proteins not used in blood clotting, along with antibodies, antigens, hormones, and any exogenous substances, such as drugs and/or microorganisms. Serum is a rich source of protein information (Adkins, Varnum et al. 2002, Bons, Wodzig et al. 2005). Serum proteins make up 6–8% of the blood, equally divided between serum albumin and a great variety of serum globulins.

Figure 1-20. Schematic diagram of the composition of whole blood.

The diagram depicts: (A) whole blood collected into correct blood tube and either left to clot or centrifuged immediately can be separated into; (B) serum or (C) plasma with erythrocytes (red blood cells) at the bottom of the centrifuge tube. Allowing blood to clot, the clear solution in the upper phase separates as serum (no clotting factors). Plasma separation includes a clear upper layer, thin middle layer (buffy coat) of white blood cells mixed with platelets. Plasma is 90% water and 10% solutes: Hormones, nutrients and electrolytes.

Serum and Plasma collection

Blood sample collection, handling and storage are important as they impact on the sensitivity, selectivity, and reproducibility of an analysis. This in turn affects the blood samples used in biomarker-related studies. Some studies suggest that procedure and storage conditions had relatively minor effects (Hsieh, Chen et al. 2006), although most agree that sampling and handling procedures have great effects on proteome profiling, and that standardised protocols for serum/plasma sampling, handling and storage are essential (Rai, Gelfand et al. 2005). A standard operating procedure for serum and plasma collection for early detection research is available (Tuck, Chan et al. 2009).

Plasma Handling Considerations To prevent clotting and collect a suitable plasma sample, anticoagulant plasma collection tubes contain different anticoagulants such as ethylene-diamine-tetra- acetic acid (EDTA), heparin, or sodium citrate must be used. The blood sample is centrifuged and the plasma is separated from the blood cells (collected from the top). It should be noted that the additives (anticoagulant) can impact on the protein makeup in the plasma and can influence potential uses of these samples in proteomics or genomics analysis (Rai, Gelfand et al. 2005). Therefore, careful consideration should be made to blood tube selection to ensure the best possible specimen.

Serum Handling Considerations Serum is plasma minus the clotting factors, achieved by allowing the blood samples to clot at RT for 30–60 minutes. Special serum-separating tubes, which contain a gel media, are used to facilitate coagulation and help provide a serum sample. The coagulated blood sinks to the bottom of the test tube, leaving the serum on top (minus fibrinogen, i.e. clotting factors). Serum samples that are allowed to sit less than 30 min are likely to retain cellular elements and other contaminants impacting future analysis. Samples that sit for longer than 60 min are likely to experience lysis

of cells in the clot, releasing cellular components not usually found in serum samples.

Experts in the Human Proteome Organization (HUPO) panel supported the use of plasma instead of serum, with EDTA (or citrate) for anticoagulation and recommend MS/MS technologies (Omenn, States et al. 2005). The adverse effects of long term storage conditions and freeze-thaw cycles have been discussed (Luque-Garcia and Neubert 2007), although it seems no major differences have been reported for long- term storage of plasma/serum samples at −20, −80 °C or using liquid nitrogen (Rai, Gelfand et al. 2005). Although freeze/thaw cycles have not been reported to have caused profile changes (Hogdall, Johansen et al. 2000, de Noo, Tollenaar et al. 2005), a good technical preparation practise requires small sample volumes being frozen for easy access and to avoid more than one freeze and thaw cycle.

Proteomics information in plasma and serum

Plasma and serum are both complex fluids containing a diverse range of proteins with varied concentrations from very high to very low abundances. Plasma contains thousands of proteins, which originate from a variety of cells and tissues through either active secretion or leakage from blood cells or tissues. The main plasma protein is albumin at 30g/L, with 50% of plasma proteins comprising of immunoglobulins (Igs), fibrinogens, transferrin, haptoglobin and lipoproteins in order of intensity and concentration (Thadikkaran, Siegenthaler et al. 2005). Many of the high abundance plasma proteins involved in coagulation, immune defense, small molecule transport, and protease inhibition, have been functionally characterised and are associated with disease processes. Plasma protein concentration range is very dynamic, and comprises of nine orders of magnitude, which consisting mostly of low abundance proteins (Adkins, Varnum et al. 2002).

Application of MS to protein analysis has led to the development of a variety of emerging proteomic methods applied to clinical specimens such as serum/plasma and tissues used for protein-expression profiling to distinguish cancer-specific 88

cohorts from non-cancer groups. Initially, a list of 289 plasma proteins was reported (Anderson and Anderson 2002), followed by a more precise list of over 1,000 (1175) on the human plasma proteome (Anderson, Polanski et al. 2004, Rose, Bougueleret et al. 2004). The HUPO large-scale collaborative study, involving 18 laboratories, characterised the human serum and plasma proteomes using LC-MS/MS and the International Protein Index database. They identified 9,504 proteins with one or more peptides, 3,020 proteins with two or more peptides, and provided a final a list (with a 95% confidence level) of 889 proteins (States, Omenn et al. 2006). The human serum proteome displayed nearly 3700 chromatographically separated protein spots on 2DGE and 325 distinct proteins were identified (Pieper, Gatlin et al. 2003). Using microcapillary LC-MS, 490 proteins were detected in serum, including some lower abundance (ng/ml range) serum proteins such as human growth hormone, interleukin-12, and prostate-specific antigen (PSA) (Adkins, Varnum et al. 2002). Importantly, the protein profiles obtained from plasma and serum are very different and can also be affected by sample handling (Banks, Stanley et al. 2005, Hsieh, Chen et al. 2006). Therefore, the selection of blood tubes for biomarker studies is of great interest in our research and will be detailed in Chapter 5.

Immunodepletion

Immunodepletion is a method for removing a target molecule from a mixture. The main practical challenge with mass analysis of blood samples is in the purification and detection of proteins, due to its complexity and the presence of several high- abundance proteins: albumin, IgGs, transferrin, α-1-antitrypsin, IgAs, IgMs, α-2- macroglobulin, haptoglobin, and high-density lipoproteins (HDL, mainly ApoA-I and ApoA-II), orosomucoid (α-1-acid glycoprotein) and fibrinogen.

The purification and immunodepletion technique for the removal of high abundance proteins is thought of as essential for the reliable detection of low abundance proteins (Govorukhina, Reijmers et al. 2006). Diverse procedures can be applied to remove the most abundant proteins from serum and blood samples, which include 89

ultrafiltration or the usage of diverse antibodies attached to solid supports to increase the dynamic protein concentration range (Villar-Garea, Griese et al. 2007). The application of Agilent multiple affinity removal columns in biomarkers studies, removes abundant proteins to unmask low molecular weight proteins. The potential problem for this removal is the nonspecific binding during the depletion procedures and whether low MW proteins species do actually bind to the column caused by the ability of albumin or other high-abundant proteins to bind peptides or protein fragments (Zolotarjova, Martosella et al. 2005). Alternatively, a procedure commonly applied to split a complicated mixture is fractionation combined with protein digestion, as it takes advantage of the desalting step required for MS. The techniques employed include the application of dye for albumin depletion and micro beads to remove IgG (Omenn, States et al. 2005).

Other methods of immune-affinity-based protein subtraction chromatography (IASC), involve the application of purified commercially available polyclonal antibodies (PAbs) to human plasma samples. The PAbs are applied to remove several high abundance proteins like albumin, and Igs. Unfortunately, they also bind other proteins of interest in breast cancer such as α-1-antitrypsin, IgAs, transferrin, alpha-2-HS glycoprotein and haptoglobin (Pieper, Su et al. 2003, Bjorhall, Miliotis et al. 2005, Echan, Tang et al. 2005, Zolotarjova, Martosella et al. 2005). Alternative immune depletions are also available. These include avian IgY PAbs, raised against several high-abundance human plasma proteins, which include: albumin, IgGs, transferrin, α-1-antitrypsin, IgAs, IgMs, α-2-macroglobulin, haptoglobin, and high- density lipoproteins (HDL, mainly ApoA-I and ApoA-II), orosomucoid (α-1-acid glycoprotein) and fibrinogen (Huang, Harvie et al. 2005). Several groups have demonstrated that the removal of the high-abundance proteins significantly increases the detection of the low-abundance proteins in plasma. At the same time, this removal has raised concerns about the no specific binding of the less abundant protein with the high abundance proteins being removed, and the cross reactivity of the reagent applied in depletion method (Pieper, Su et al. 2003, Bjorhall, Miliotis et al. 2005, Echan, Tang et al. 2005, Omenn, States et al. 2005, Zolotarjova, Martosella et al. 2005). 90

Therefore, it seems that fractionation of the serum or plasma samples was the preferred method as it would allow for the separation of the sample into different MW fractions, without loss of low mass proteins, whilst simultaneously desalting the sample in preparation for downstream MS analysis.

MASS SPECTROMETRY AND BREAST CANCER

In the search for diagnostic and prognostic markers, proteins with potential clinical significance have been identified in breast cancer proteome studies applied to tissue samples and biological fluids which include urine, serum, plasma and nipple aspirate (Hondermarck, Vercoutter-Edouart et al. 2001, Bertucci, Birnbaum et al. 2006, Goncalves, Esterni et al. 2006, Gast, Schellens et al. 2009). Several review papers have summarised some of the candidate biomarkers found in blood, tissues and nipple aspirate fluid from breast cancer patients (Galvao, Martins et al. 2011, Opstal-van Winden, Vermeulen et al. 2012, Tang and Gui 2012).

Epigenetics and breast cancer

Epigenetic epidemiology includes the study of variation in epigenetic traits and the risk of disease in populations, with potential to serve as excellent diagnostic and prognostic markers of cancer. The three cornerstones of epigenetics which may be involved in carcinogenesis include: DNA methylation, chromatin and histone modifications and non-coding RNAs (Barrow and Michels 2014).

Ataxia telangiectasia mutated (ATM), a serine/threonine protein kinase recruited and activated by DNA double-strand breaks, phosphorylates several key proteins that initiate activation of the DNA damage checkpoint, leading to cell cycle arrest and DNA repair or apoptosis. Numerous blood studies in breast cancer have looked at DNA methylation levels of ATM (Brennan, Garcia-Closas et al. 2012) or measure DNA methylation at 27578 CpGs (Xu, Bolick et al. 2013), as markers of breast cancer risk. Others have performed promoter methylation analysis of four cancer-related genes: RASSF1A, GSTP1, APC and RARβ2 (Brooks, Cairns et al. 2010), and miRNA 91

levels differences between women who remain cancer-free versus those who later develop cancer(Godfrey, Xu et al. 2013). Real-time quantitative PCR analysis detected elevated circulating levels of cell-free DNA (cfDNA) in malignant breast cancer plasma samples that correlated with malignant tumour size (Zhong, Ladewig et al. 2007).

Biomarkers in tissue, nipple aspirate, saliva and tears

In recent breast cancer studies a number of potential biomarkers from tissue, nipple aspirate, saliva and tears were identified and validated. Several differentially expressed proteins in breast cancer tissue included ubiquitin, protein S100-A8, -B- crystalin, HER3, cathepsin H (CATH), heat shock protein beta-1 (Hsp27), protein S100-A6, and desmoglein-3 (DSG3) (Galvao, Martins et al. 2011), numerous other S100 proteins (11 isoforms as 7 members) (Cancemi, Di Cara et al. 2010), calreticulin (Kabbage, Trimeche et al. 2013) and protein disulfide isomerase A3 (PDIA3) (Song, Moon et al. 2012). Additionally, IHC studies detected several cancer biomarkers to be specifically overexpressed in breast cancer. These markers include metastasis-associated in colon cancer-1 (MACC1), which seemed to be associated with survival (Huang, Zhang et al. 2013) and lysosome-associated protein transmembrane 4 beta (LAPTM4B) which correlated with disease progression (Xiao, Jia et al. 2013). Others biomarkers detected include γ-glutamyl hydrolase (GGH) and fatty acid amide hydrolase (FAAH) (Shubbar, Helou et al. 2013). Also, bone morphogenetic protein 6 (BMP6) (Rose, Bougueleret et al. 2004) and huntingtin-associated protein 1 (HAP1) (Zhu, Song et al. 2013), which were found to be under expressed in breast cancer tissue.

Other differentially expressed potential biomarkers include fluid, gross cystic disease fluid protein 15 (GCDFP-15), 1-acid glycoprotein (AAG) and basic fibroblast growth factor (bFGF) (Tang and Gui 2012) in nipple aspirate from breast cancer patients. Also using 2D-DIGE, carbonic anhydrase 6 (CA6) was found in saliva

(Zhang, Xiao et al. 2010) and C1Q1, ALDH3A, or TPI in tears with the application of SDS-PAGE and MALDI-TOF/TOF (Bohm, Keller et al. 2012).

Blood Profiling

MS was applied to serum profiling in breast cancer pilot studies (Becker, Cazares et al. 2004, Pusztai, Gregory et al. 2004, Chao, Ladd et al. 2013), which included the comparison of women with and without the BRCA-1 mutations (Laronga, Becker et al. 2003, Becker, Cazares et al. 2004), along with classification of pre and post- surgery breast cancer profiles (Vlahou, Laronga et al. 2003).

The first stage of the biomarker's discovery pipeline in the identification of candidate markers is MS-based profiling. The applications of MS to analyse serum and plasma proteome have been used to detect changes related to breast cancer progression and response to treatment (Li, Zhang et al. 2002, Pusztai, Gregory et al. 2004, Heike, Hosokawa et al. 2005, Goncalves, Esterni et al. 2006, Villanueva, Shaffer et al. 2006, Pietrowska, Marczak et al. 2009, Pietrowska, Polanska et al. 2010, Solassol, Rouanet et al. 2010), all showing that the identified markers could potentially be used to detect or classify breast cancer.

The Human Proteome Organization (HUPO) Plasma Proteome Project (PPP) applied a comprehensive proteomics approaches involving the input of expert research groups globally over many years, to classify several components of a biological system. These studies range from the identification of every protein in the plasma proteome (Omenn, States et al. 2005) to measuring very low abundance proteins (Mallick and Kuster 2010)

Protein Expression Patterns in Blood

Using LC-MS analysis, varying protein expression patterns have been observed in breast cancer blood samples, when compared to healthy controls, which have the

potential to serve as diagnostic biomarkers for breast cancer as well as benign breast diseases. An updated list of several breast cancer associated proteins and potential biomarkers which have already been identified using proteomics technology are detailed in the Table 1-7.

Briefly, several serum studies have identified differential expression in breast cancer of the following: isoform 1 of inter-alpha trypsin inhibitor heavy chain (ITIH4), fibronectin 1, CXCL9, apolipoproteins ApoA1, ApoA2, ApoC1, ApoC2, ApoC3, and ApoE, C3a des-arginine anaphylatoxin (C3adesArg), C3f, C4a, platelet factor 4, haemoglobin -chain and -chain, transferrin, EGFR, mammaglobin, afamin, - 2-macroglobulin, bradykinin, ceruloplasmin, fibrinogen, , fibrinopeptide A and transthyretin showed (Opstal-van Winden, Vermeulen et al. 2012, Tang and Gui 2012).

Other breast cancer associated serum proteins detected, using multiple fractionation steps (protein depletion, lectin affinity fractionation, isoelectric focusing (IEF) separation, and LC-MS analysis include: thrombospondin-5 (TSP5), serum amyloid P-component (SAP), alpha-1B-glycoprotein (A1BG), complement C3, alpha-1-antitrypsin, transferrin, along withthrombospondin-1 (TPS1) and tenascin- X (TN-X) (Zeng, Hincapie et al. 2010, Zeng, Hincapie et al. 2011). Nasim et al identified serum proteome expression profiles for breast cancer that included: APO- A1, some acute phase proteins (APP), haptoglobin (Hp), inter-alpha-trypsin inhibitor heavy chain 4 (ITIH4), serum albumin, serum amyloid A (SAA). In addition, some complement system component (C3 and C8) fractions were up-regulated, though Ig kappa C region and transthyretin (TTR) were down regulated (Nasim, Ejaz et al. 2012). Furthermore, using MS on breast cancer samples, apoptosis protein-like protein-2 (ILP-2) was determined in serum (Xiang, Zhou et al. 2012). It was reported that Biotinidase (Kang, Ahn et al. 2010), thrombospondin-1 (THBS1), bromodomain and WD repeat-containing protein 3 (BRWD3) were overexpressed in plasma (Suh, Kabir et al. 2012). Additional serum biomarkers identified using MS analysis are shown in Table 1-8.

Table 1-7. Breast cancer blood (serum or plasma) biomarkers in the literature discovered using proteomic techniques.

Protein MW Protein Function Biomarker Purpose Spec. Proteomic Sample Reference (Da) Type Method Size (BC: Normal) Lipid Metabolism

Apolipoprotein 28,078 Participates in the Protein signature Serum SELDI-TOF 48M+, (Goncalves, ApoA-I reverse transport of correlated with 33M- Esterni et al. cholesterol from metastatic relapse 2006) tissues to the liver Predictive marker Serum 2D-nanoLC- 68: 68 (Opstal-van MS/MS Winden, Krop et al. 2011) Predicts recurrence- Serum SELDI-TOF 99: 51 (Chung, Moore et free survival in al. 2014) women with ER- tumours Serum 1D-SDS PAGE & 48 BC (Nasim, Ejaz et al. Diagnostic makers LC-MS/MS 13 BBD 2012) Apolipoprotein 8707 Stabilises HDL Response to Serum SELDI-TOF 6 pre, 6 (Heike, ApoA-II structure by its docetaxel infusion post trt Hosokawa et al. association with lipids treatment 2005) and affect the HDL metabolism Apolipoprotein 43,402 Role in VLDL secretion Comparison of Serum MALDI- 21: 33 (Villanueva, ApoA-IV and catabolism different cancer types TOF/TOF Shaffer et al. 2006)

Biomarker discovery Plasma 2D-iTRAQ-LC- 12: 12 (Meng and by low abundance MS/MS Veenstra 2011) protein enrichment Apolipoprotein 6630 Modulates the Protein signature Serum SELDI-TOF 33M-, (Goncalves, ApoC-I interaction of ApoE correlated with 48M+ Esterni et al. metastatic relapse 2006) Diagnosis Serum SELDI-TOF 124: 158 (Fan, Wang et al. 2010) Predictive marker Serum SELDI-TOF 68:68 (Opstal-van Winden, Krop et 99: 51 al. 2011) (Chung, Moore et al. 2014) Apolipoprotein 34,236 Mediates binding, Pre-diagnostic serum Serum SELDI-TOF, 2D 68: 68 (Opstal-van ApoE internalization, protein pattern & LC-MS/MS Winden, Krop et catabolism of identification al. 2011) lipoprotein particles Leptin 16,026 Part of signalling Biomarker discovery Serum MALDI-TOF 92: 104 (Pietrowska, pathway that regulates cancer and controls Marczak et al. the size of the body fat comparison 2009) depot

Inflammation and immunity

Alpha 2HS- 30,221 Promotes endocytosis Chemotherapy Plasma 2D-DIGE, LC- 25 pre, (Michlmayr, glycoprotein & influences mineral response: epirubicin MS/MS 25 post Bachleitner- (AHSG) phase of bone & docetaxel MALDI-TOF trt Hofmann et al. 2010) Early BC detection Serum MALDI-MS 36: 36 (Fernandez- Grijalva, Aguilar- 96

Lemarroy et al. 2014) Markers of BC stage & Serum MALDI-TOF 150 (Schaub, Jones et obesity patients al. 2009) (525 samples) Auto-antibodies serum MALDI-TOF 81: 73 (Yi, Chang et al. biomarkers for BC 2009) screening Ceruloplasmin 120,085 Acute phase reactant, Pre-diagnostic serum Serum SELDI-TOF, 2D 68: 68 (Opstal-van copper-binding & protein pattern and LC-MS/MS Winden, Krop et carrying protein in identification al. 2011) blood Complement C3a 8939 Inflammatory process, Identification and Serum SELDI-TOF 93: 46 (Li, Zhang et al. mediates complement validation of serum 37 BBD 2002) pathway protein markers

Diagnostic value Serum SELDI-TOF 124: 158 (Fan, Wang et al. 2010) Pre-diagnostic serum Serum SELDI-TOF, 2D, 68: 68 (Opstal-van protein pattern LC-MS/MS Winden, Krop et al. 2011)

Cancer type Serum MALDI- 21: 33 (Villanueva, Comparison TOF/TOF Shaffer et al. 2006) Chemotherapy Plasma 2D-DIGE, 25 pre, (Michlmayr, Response: epirubicin MALDI-TOF 25 post Bachleitner- & docetaxel treatmen Hofmann et al. t 2010)

Potential BC Serum LC-MS/MS 60: 60 (Solassol, biomarker Rouanet et al. 2010) Diagnostic makers Serum 1D-SDS PAGE & 48 BC (Nasim, Ejaz et al. LC-MS/MS 13 BBD 2012) Predicts recurrence Serum SELDI-TOF 99: 51 (Chung, Moore et with ER- tumours al. 2014) Complement C4A 84,183 Activation of the Comparison of Serum MALDI- 21: 33 (Villanueva, classical complement different cancer types TOF/TOF Shaffer et al. system 2006)

Complement C4B 71,678 Mediates complement Biomarker discovery Plasma 2D-iTRAQLC- 12: 12 (Meng and pathway in local by low abundance MS/MS Veenstra 2011) inflammatory process protein enrichment

Chemotherapy Plasma 2D-DIGE, LC- 25 pre, (Michlmayr, response: epirubicin MS/MS 25 post Bachleitner- & docetaxel trt Hofmann et al. 2010) C-reactive protein 23,047 Associated with host Protein markers for Plasma mTRAQ-LC- 54: 30 (Suh, Kabir et al. (CRP) defence, interacts with early diagnosis MS/MS 2012) DNA, scavenges nuclear material release from damaged circulating cells Evaluation of Serum Immuno- assay 814 BC (Pierce, Ballard- inflammatory Barbash et al. biomarkers and 2009) prognostic value

Haptoglobin Acute phase response Early detection of Serum MALDI-TOF 30:30 (Liu, Sun et al. protein triple negative BC 2014) Early detection of Serum 2-DE and MS 39: 40 (Hamrita, Chahed infiltrating ductal et al. 2009) breast carcinomas Inter-α-trypsin 70,585 Involved in acute Comparison of Serum MALDI- 21: 33 (Villanueva, inhibitor heavy phase response to different cancer types TOF/TOF Shaffer et al. chain H (ITIH4) trauma 2006) Response to Plasma 2D, DIGE, 25 pre, (Michlmayr, chemotherapy of MALDI-TOF 25 post Bachleitner- epirubicin and trt Hofmann et al. docetaxel 2010)

Prediagnostic serum Serum SELDI-TOF, 2D, 68: 68 (Opstal-van protein pattern and LC-MS/MS Winden, Krop et identification al. 2011)

Markers of BC stage Serum MALDI-TOF 150 (Schaub, Jones et and obesity patients al. 2009) (525 samples) Diagnostic makers Serum 1D-SDS PAGE & 48 BC (Nasim, Ejaz et al. LC-MS/MS 13 BBD 2012)

ITIH4 fragments Novel BC markers Serum SELDI-TOF 152: 129 (Gast, Van Gils et al. 2009) Monocyte 35,159 Mediates the innate Comparison of Plasma ICAT-LC- 27: 27 (Kang, Ahn et al. differentiation immune response to plasma proteome MS/MS 2010) antigen CD14 bacterial pattern

lipopolysaccharide (LPS) Serum amyloid A- 11,682 Major acute phase Evaluation of Serum Immunological 814 BC (Pierce, Ballard- 1 (SAA-1) protein inflammatory assay Barbash et al. biomarkers and 2009) prognostic value Markers of BC stage & Serum MALDI-TOF 150 (Schaub, Jones et obesity patients al. 2009) (525 samples) Correlated with the Serum Enzyme-linked 118: 30 (Zhang, Bast et al. BC stage immune- assay BBD 21 2004)

Serum amyloid_P Serum Multi- fraction 5:5 (Zeng, Hincapie (SAP) & LC-MS et al. 2011) α-1-acid 21,651 Transport protein (Kang, Ahn et al. glycoprotein modulating acute 2010) (ORM2) phase α-1-antitrypsin 45,265 Inhibits the formation Early detection of Serum MALDI-TOF 30:30 (Liu, Sun et al. (A1AT) of the active triple negative BC 2014) angiotensin-2 Early detection of Serum 2-DE and MS 39: 40 (Hamrita, Chahed infiltrating ductal et al. 2009) breast carcinomas Blood coagulation

Bradykinin 1060 Regulation of blood Profile of disease Serum MALDI- 21: 33 (Villanueva, vessel dilation TOF/TOF Shaffer et al. 2006)

100

Coagulation Factor 79,244 Activated by thrombin Profile of disease Serum MALDI- 21: 33 (Villanueva, XIIIa and calcium ion to a TOF/TOF Shaffer et al. trans-glutaminase 2006) Fibrinogen 91,358 Polymerise into fibrin Comparison of Serum MALDI- 21: 33 (Villanueva, & cofactor in platelet different cancer types TOF/TOF Shaffer et al. aggregation 2006) Markers of BC stage Serum MALDI-TOF 150 (Schaub, Jones et and obesity patients al. 2009) (525 samples) Kininogen(KNG) 69,896 Role in blood Response to Serum SELDI-TOF 6 pre, 6 (Heike, HWHK coagulation- release of docetaxel infusion post Hosokawa et al. bradykinin treatment 2005) Markers of BC stage Serum MALDI-TOF 150 (Schaub, Jones et and obesity patients al. 2009) (525 samples) Prothrombin Acute phase response Markers of BC stage Serum MALDI-TOF 150 (Schaub, Jones et protein involved in and obesity patients al. 2009) blood coagulation (525 samples) Cell matrix and adhesion

BRWD3 4193 Regulates cell Protein markers for Plasma mTRAQ-LC- 54: 30 (Suh, Kabir et al. (bromodomain & morphology and early diagnosis MS/MS 2012) WD repeat domain cytoskeletal containing 3) organization

Fibronectin 440 Involved in cell Serum ELISA 133: 119 (Ruiz-Garcia, kDa adhesion, motility, Scott et al. 2010) 101

wound healing, osteoblast mineralization. Glutathione 23,313 Protects cells and Comparison of Plasma ICAT-LC- 27: 27 (Kang, Ahn et al. peroxidase 3 enzymes from plasma proteome MS/MS 2010) (GPX3) oxidative damage pattern

Osteopontin 33,713 Bone remodelling & Biomarker discovery Serum MALDI-TOF 92: 104 (Pietrowska, (OPN) bio-mineralization by comparison of Marczak et al. with cell-matrix cancers and healthy 2009) interaction controls

Protocadherin 477,396 Plays a role in the Biomarker discovery Plasma 2D-iTRAQ LC- 12: 12 (Meng and Fat 2 (FAT2) migration of epidermal by low abundance MS/MS Veenstra 2011) cells protein enrichment

Thrombo-spondin 43,943 Adhesive glycoprotein, Protein markers for Plasma mTRAQ-LC- 54: 30 (Suh, Kabir et al. 1 (THBS1) that mediates cell-to- early diagnosis MS/MS 2012) cell and cell-to-matrix interactions, binds heparin. Thrombo-spondin Serum Multiple 5:5 (Zeng, Hincapie 5 (THBS5) fraction & LC- et al. 2011) MS Other Alpha-enolase 47,037 Multifunctional Prediction & early Plasma NanoLC- 420: 420 (Amon, Pitteri et (ENO1) enzyme,glycolysis role, detection of ER+ MS/MS al. 2012) stimulates Ig cancer production.

102

Biotinidase (BTD) 56,771 Catalytic release of Comparison of Plasma ICAT-LC- 27: 27 (Kang, Ahn et al. biotin from biocytin plasma proteome MS/MS 2010) pattern EGFR Involved in growth Pre-diagnostic serum Serum LTQ-FT MS 105: 105 (Li 2011) factor signalling protein Fructose- 39,288 Key role in glycolysis Prediction & early Plasma Nano 420: 420 (Amon, Pitteri et bisphosphate and gluconeogenesis, a detection of ER+ LC-MS/MS al. 2012) [43] aldolase A scaffolding protein cancer (ALDOA) Glyceraldehyde-3- 35,922 Role in glycolysis Prediction & early Plasma Nano 420: 420 (Amon, Pitteri et phosphate nuclear functions, and detection of ER+ LC-MS/MS al. 2012) dehydrogenase membrane trafficking cancer (GAPDH) Huntingtin (HTT) 347,603 Microtubule-mediated Biomarker discovery Plasma 2D-iTRAQLC- 12: 12 (Meng and transport or vesicle by low abundance MS/MS Veenstra 2011) function protein enrichment Transthyretin TTR 13,761 Thyroid hormone- Comparison of Serum MALDI- 21: 33 (Villanueva, binding protein, different cancer types TOF/TOF Shaffer et al. transports thyroxine 2006) from bloodstream to the brain Predicts recurrence Serum SELDI-TOF 99: 51 (Chung, Moore et for ER- tumours al. 2014)

Notes: Protein biomarkers identified in breast cancer serum or plasma, table adapted and modified from Chung et al., 2012(Chung and Baxter 2012). Abbreviations: BBD, benign breast disease; BC, breast cancer; M+, metastasis positive; M-,metastasis negative; Pre trt, pretreatment; Post trt, post treatment; Spec, specimen. 103

Using IHC, A-Kinase Anchor Protein 4 (AKAP4), was found to be overexpressed in breast cancer tissues. ELISA assay showed that the anti-AKAP4 autoantibodies were elevated in the sera of breast cancer patients (Saini, Jagadish et al. 2013).

Table 1-8. Additional serum biomarkers identified using MS analysis.

Novel Breast Cancer Protein Breast Cancer Association Reference

Inhibitor of apoptosis Only detected in testis & (Xiang, Zhou et al. 2012) protein-like protein-2 lymphoblastoid cells (ILP-2)

Dermcidin Disease progression & (Brauer, D'Arcy et al. survival 2014)

Interleukin-6 Related to tumour size (Celis, Moreira et al. 2005) Hsp27 (Belluco, Petricoin et al. 2007)

14-3-3 sigma Patterns in breast cancer (Belluco, Petricoin et al. stage 2007)

Mammaglobin 10-kd Elevated serum levels (Belluco, Petricoin et al. protein/ lipophilin 2007) B complex (Laronga, Becker et al. 2003, Bauer, Chakravarthy et al. 2010)

Interleukin 27 (IL-27) Critical role in immune (Lu, Zhou et al. 2014) regulation of infection, autoimmunity

VEGF Decreased after radical (Lu, Zhou et al. 2014) mastectomy

With the application of antibody-based array platforms and algorithms, (188 breast cancer and 175 normal subjects) found serum concentrations of EGF, soluble CD40- ligand and pro-apolipoprotein A1 were increased in breast cancer compared (Kim, Lee et al. 2009). Additional work found HMW-Kininogen, apolipoprotein A1, soluble vascular cell adhesion molecule-1, plasminogen activator inhibitor-1, vitamin-D binding protein and Vitronectin were decreased in the cancer group (Kim, Lee et al. 2009).

104

Molecular portraits, that identified 21 serum protein signatures in breast cancer patients (240 samples from 64 BC patients), included APO-A4, complement C1-5 along with CD40 and Interleukins IL1-7. They were used to classify patients with primary breast cancer and predict the development of distant metastases, independent of the type of adjuvant therapy received (Thadikkaran, Siegenthaler et al. 2005).

Blood proteomics data bases and internet sites

Proteomics data bases, internet sites devoted to blood proteomics include: The Human Proteome Organization (HUPO), human plasma proteome project http://www.hupo.org http://www.hupo.org/initiatives/plasma-proteome-project/ Human serum proteome http://www.plasmaproteomedatabase.org/ Universal Protein Resource (UniProt)http://www.uniprot.org

The Universal Protein Resource (UniProt) is freely accessible, accurately annotated protein sequence knowledgebase, which integrates, interprets and standardises data from numerous resources to achieve the most comprehensive catalogue of protein sequences and functional annotation (UniProt 2013). The UniProt Consortium, consists of groups from the European Bioinformatics Institute (EBI), the SIB Swiss Institute of Bioinformatics (SIB) and the Protein Information Resource (PIR).

Plasma Proteome Database provides a resource for proteomics research. It was initially described in 2005 as a part of Human Proteome Organization's (HUPO's) initiative on Human Plasma Proteome Project. This database currently contains information on 10,546 proteins detected in serum/plasma to facilitate clinical and basic research by serving as a comprehensive reference of plasma proteins in humans (Nanjappa, Thomas et al. 2014).

105

SUMMARY OF LITERATURE REVIEW

Breast cancer is a common disease affecting 1 in 9 women in Australia. Survival for breast cancer patients requires early detection and optimal treatment. Several targeted therapies currently exist and are used in combination with radio therapy and surgery, which have resulted in improved survival. However there is still a group of women who present with either late stage disease or move aggressive early tumours who will not respond to current therapies.

In the last decade of cancer research, there has been a major focus on cancer biomarker discovery driven by a strong clinical need to improve the early detection and diagnosis of breast cancer, the detection of early recurrence and in the prediction of therapeutic responsiveness. Numerous biomarkers have been identified in the last 20 years but only a few that have been vigorously validated (ER, PR and HER2) used in the clinic. This in turn has created an expansion in proteomic research and has intensified efforts to mine the human proteome for these answers. As discussed in the introduction, the evolution of MS-based proteomic technologies has led to major advances in our understanding of how these technologies work and the human proteome. MS has detected thousands of proteins at nanomolar concentrations, along with markers to diagnose disease, disease progression during treatment or responsiveness to therapy, holding great promise for the detection of candidate protein biomarkers for clinical application to improve breast cancer patients’ outcome.

Proteomic analysis of human body fluids has become one of the most promising approaches to the discovery of biomarkers for human diseases. Although there has been significant progress in breast cancer research, due to its heterogeneous nature, the underlying mechanisms are still poorly understood and potential candidate protein biomarkers remain to be discovered by proteomics. Therefore, investigation of breast cancer biological samples to develop a protein profile and discover novel biomarkers for the early detection of breast cancer is urgently needed. Several blood biomarkers have been identified that are related to breast cancer, though none to

106

date have been found in urine and warrants investigation. The study of urinary proteins, as a non-invasive sample, holds the potential key to provide this information, with the ready identification of abundant and depleted proteins.

The application of the improved separation and purification techniques followed by the combination of SDS-DGE and LC-MS/MS-based proteomic methods for the analysis of urine and blood will be a powerful tool to uncover novel candidate breast cancer protein markers of clinical relevance.

THESIS AIMS

Changes occur in blood and urine protein profiles between patients with breast cancer and women with benign disease or healthy subjects. These changes are related to the different stages of breast cancer and could be used to diagnose breast cancer. Investigating the protein present in urine and blood and identifying a protein profile will help us to understand the mechanisms of breast cancer and discover of potential novel biomarkers to distinguish the different stages of breast cancer and detect breast cancer early. Thus, the studies in this thesis are aimed as follows: Aim 1: Establish a standardised protocol of urine proteins precipitation and concentration (Chapter 3). Aim 2: Identify the novel proteins expressed in urine and compare the difference found in DCIS (non-invasive), Invasive and metastatic breast cancer (Chapter 4). Aim 3: Apply several different blood tubes and determine the tube-medium (serum or plasma) which allows for the most extensive protein information technically for future analysis (Chapter 5). Aim 4: Identify the proteins differentially expressed in the serum and plasma of breast cancer patients and control subjects (Chapter 6). Aim 5: Validate the identified biomarkers in breast cancer samples for future clinical trials with the specialist groups on breast cancer tumour samples and cell lines (Chapter 4 and Chapter 6).

107

Chapter 2 Materials and Methods

Chapter 2 describes the general materials and methods used in this study. Testing, optimisation and comparison between various samples, along with preparation methods are also included and are further described in Chapters 3-6. In this thesis, the biological samples were collected in a randomised fashion and assessed according to their pathological diagnosis. In our experimental work, all samples were carefully handled, prepared and analysed to ensure consistency, reproducibility and that the results represent the correct breast cancer groups.

108

2 GENERAL MATERIALS, METHOD AND EQUIPMENT

ETHICS APPROVAL

In this study, all the female breast cancer subjects had received detailed diagnostic procedures, i.e. a physical breast examination, mammography, ultrasound and biopsy or excision with a detailed pathological report on the cancer. Ethics approval for the collection and use of human urine, blood and tissue samples was obtained from the South East Sydney Local Area Health District Research Ethics Committee, Southern Section (SEA HRCE) (#07/71Li).

The study was designed and conducted in accordance with the ethical principles and all participants signed informed consent forms prior to sample collection. The Declaration of Helsinki regarding the use of human subjects in research was adhered to. None of the subjects had received any prior treatment, either endocrine or chemotherapy. All urine and blood samples, along with clinical history were collected from consenting breast cancer patients at St. George Private Hospital Sydney (NSW, Australia). Age and sex-matched control samples were collected from volunteers at St George Hospital, Cancer Care Centre (NSW, Australia), who additionally provided medical history to confirm they were disease free and were not taking medication for any underlying illnesses.

109

MATERIALS

Solutions and supplier

Material and chemical Supplier 3.5 µm particle size (300 Å pore size) Agilent Technologies (Palo Alto, CA, USA) 96-well polypropylene V-well plate Sigma-Aldrich (Castle Hill, NSW, Greiner Australia), Cat No. M0936 96-well polypropylene clear wells Sigma-Aldrich (Castle Hill, NSW, Greiner plates, round bottom Australia), Cat No. M7935, M2186 Acetic acid Thermo Fisher Scientific (Scoresby, Vic, Australia) Acetone HPLC grade or better Thermo Fisher Scientific (Scoresby, Vic, Australia) Acetonitrile Sigma-Aldrich (Castle Hill, NSW, Australia) Ammonium persulfate Sigma-Aldrich (Castle Hill, NSW, Australia) Ampholyte(pH 3-10) Pharmalyte, GE Healthcare, Sigma- Aldrich (Castle Hill, NSW, Australia), Cat No 17-0456-01 BCA Protein Assay Kit-Pierce™ Thermo-Fisher Scientific (Scoresby, Vic, Australia), Cat No. PIE 23225 Benzonase (ultra-pure) Sigma-Aldrich (Castle Hill, NSW, Australia) Blood tubes BD Diagnostics (North Ryde, NSW, Australia) Bovine serum albumin (BSA) Sigma-Aldrich (Castle Hill, NSW, Australia), Cat No. A4503 Bromophenol Blue 0.1%(w/v) Pharmacia-LKB (Uppsala, Sweden) Cellstar® 15 and 50 mL tubes Greiner Cell star (Interpath, VIC, Australia), Cat No. 82050-276, 82050-346

110

CHAPS Sigma-Aldrich (Castle Hill, NSW, Australia) Diaminobenzidine (DAB) DAKO, Sigma -Aldrich (Castle Hill, 3,3'-diamino-benzidine substrate NSW, Australia) Dimethyl sulfoxide Sigma-Aldrich (Castle Hills, NSW, Australia), Cat No. 472310 Di-sodium hydrogen orthophosphate Ajax Chemicals (Auburn, NSW, Australia) DL-Dithiothreitol (DTT) Promega (Alexandria, NSW, Australia), Cat No. V3151 DMEM Thermo-Fisher Scientific (Scoresby, Vic, Australia), Cat No. 12800-017 Cellstar® cell culture flasks 75 cm with Greiner Bio One Supplier via Thermo Filter Cap, Sterile. Greiner bio-one Fisher (Scoresby, VIC, Australia), Cat No, 178883 and Cell star (Interpath, VIC, Australia), Cat No. 658175V Phosphate-buffered saline Dulbecco’s Life technologies (Mulgrave, NSW, Australia), Cat No. 21600-010 Ethanol absolute Sigma-Aldrich & Thermo-Fisher Scientific Ethylene-diamine-tetra-acetic acid Ajax Chemicals (Auburn, NSW, Australia) Fetal Bovine Serum (FBS) Life technologies (Mulgrave, NSW, Australia), Cat No. 26140-079 Formaldehyde Sigma-Aldrich (Castle Hill, NSW, Australia) Formic acid Sigma-Aldrich (Castle Hill, NSW, Australia) Glycerol Sigma-Aldrich (Castle Hill NSW, Australia) Goat serum Sigma-Aldrich (Castle Hill NSW, Australia), Cat No. G9023 HEPES Gibco (Grand Island, NY, USA) Hydrochloric acid 32% Sigma-Aldrich (Castle Hill NSW, Australia)

111

Immun-Blot TM PVDF Membrane Bio- BioRAd (Gladesville, NSW, Australia), Rad Cat No. 1620177 LDS Sample Buffer (4X) Life technologies (Mulgrave, NSW, Australia), Cat No. NP0007 Insulin Sigma-Aldrich (Castle Hill, NSW, Australia), Cat No. 19278 Iscove's Modified Dulbecco's Medium Life technologies, Invitrogen™ (IMDM) Thermo Fisher Scientific (Scoresby VIC, Australia), Cat No. 1471300 Microcentrifuge tubes (0.5 and 1.5 mL) Sigma-Aldrich (Castle Hill NSW, Greiner Australia), Cat No. 617873 Micron Centrifugal filter devices: YM-3 Merck Millipore Life Science, and YM-50 (Bayswater, VIC, Australia). Cat No. 42404, 42416 MOPS SDS Running Buffer (20X) Life technologies (Mulgrave, NSW, Australia), Cat No. NP0001-02 MultiCelTM trypan blue (0.5%) Sigma-Aldrich (Castle Hill, NSW, exclusion dye Trace Biosciences Australia) N’-tetramethylethyl-ethylenediamine Sigma-Aldrich (Castle Hill, NSW, Australia) NuSep 10-20% Long life gels pre-cast NuSep Ltd, (Lane Cove, NSW, polyacrylamide electrophoresis iGels Australia), Cat No. NH11-1020 NuSep sample buffer NuSep Ltd, (Lane Cove, NSW, Australia) Non-sterile 5 mL polystyrene round- Becton Dickinson (Franklin Lakes, bottom tubes (FACS tubes) New Jersey, USA) Penicillin/Streptomycin (5,000 U/mL) Life Technologies, Thermo Fisher Scientific (Scoresby, VIC, Australia), Cat No. 15070-063, 15140122 (10,000U/mL) Potassium chloride Ajax Chemicals (Auburn, NSW, Australia) Phosphate-buffered saline Dulbecco’s Life technologies and Thermofisher (Scoresby, VIC, Australia), Cat No. 21300025 Potassium-dihydrogen Ajax Chemicals (Auburn, NSW, orthophosphate Australia

112

Potassium hydrogen carbonate Ajax Chemicals (Auburn, NSW, Australia) Precision Plus ProteinTM Bio-Rad (Hercules, CA, USA) All Blue molecular weight standards Protease inhibitor cocktail Sigma-Aldrich (Castle Hill, NSW, Australia) Proxeon C18-tip Thermo Fisher Scientific (Rockford, IL, USA) RPMI-1640 medium Miltenyi Biotech, (Macquarie Park, NSW, Australia), Cat No. 130-091- 439 Sample Reducing Agent (10X) Life technologies (Mulgrave, NSW, Australia), Cat No. NP0009 SDS_Page Gels 12.5% Criterion™ Tris- Bio-Rad Laboratories Pty. Ltd HCl Gel (Gladesville, NSW, Australia), Cat No.3450014 Sequencing grade trypsin Promega (Alexandria, NSW, Australia), Cat No. V5111 Slides (glass) Livingstone International (Rosebery, NSW, Australia) Superfrost slides Thermo Fisher Scientific, (Scoresby, Vic, Australia) Sodium azide BDH Chemicals Ltd. (Poole, UK) Sodium chloride Ajax Chemicals (Auburn, NSW, Australia) Sodium dodecyl sulfate (SDS) Sigma-Aldrich (Castle Hill, NSW, Australia) Sodium fluoride Sigma-Aldrich (Castle Hill, NSW, Australia) Sodium hydroxide Ajax Chemicals (Auburn, NSW, Australia) Sodium Lowy Sulfate Thiosulphate Sigma-Aldrich (Castle Hill, NSW, Australia) Streptavidin/HRP Dako (North Sydney NSW, Australia), Cat No. P0397

113

Streptomycin 50 µg/mL Life Technologies (Bayswater, VIC, Australia), Cat No.15140122 SuperSignal West Pico Thermo-Fisher Scientific (Scoresby, Chemiluminescent Substrate Vic, Australia), Cat No. PIE34080 Tetramethylethylenediamine Thermo-Fisher Scientific (Scoresby, (TEMED) Vic, Australia), Cat No. 17919 Thiourea Sigma-Aldrich (Castle Hill, NSW, Australia) Transfer Buffer (20X) Life technologies (Mulgrave, NSW, Australia), Cat No. NP0006-1 Triethyl-ammonium bicarbonate Sigma-Aldrich (Castle Hill, NSW, Australia) Trifluoroacetic acid Sigma-Aldrich (Castle Hill, NSW, Australia) Tris(hydroxymethyl)aminomethane Sigma-Aldrich (Castle Hill, NSW, (Tris) Australia) Triton X-100 Sigma-Aldrich (Castle Hill, NSW, Australia) TruPAGE™ Precast Gels Sigma-Aldrich. Invitrogen (Castle Hill, NSW, Australia) Tru Sep SDS, sample boiling buffer NuSep Holdings Ltd (Homebush, NSW, Australia) Trypsin (0.25% w/v)-EDTA (0.05% Life technologies (Scoresby VIC, w/v) Australia), Cat No. 25200-056 (no phenol red, Cat No. 15400-054) Tween-20 Sigma-Aldrich (Castle Hill, NSW, Australia) Urea Merck Pty Ltd, (Kilsyth, VIC, Australia) Water- Milli-RO 12 Plus water (Merck-Millipore, VIC, Australia) purification system

114

Preparation of buffers and reagents

Note: All media and buffers were prepared with Milli-Q water.

Bead beating lysis buffer: 7M urea, 0.1 M CHAPS (3-[(3 cholamidopropyl) dimethylammonio]-1-propanesulfonate), 0.1 M dithioerythritol (DTT), and 35 mM Tris-base). Made up with Milli-Q- H2O, add 0.1% (w/v) Bromophenol Blue.

Cell culture media: RPMI 1640 (Invitrogen Australia Pty Ltd, Melbourne, Victoria, Australia-Miltenyi Biotec Cat#S11800L0498) and Improved Minimal Essential Medium (IMEM; GIBCO/Invitrogen, Grand Island, NY), supplemented with 10% (v/v) heat-inactivated fetal bovine serum (FBS), 50 U/mL of penicillin and 50 µg/mL of streptomycin (Invitrogen Pty Ltd, VIC, Australia). Stored at 4 °C.

Cell freezing media consisting of 40% (v/v) FBS and 20% dimethyl sulfoxide (DMSO) (Sigma-Aldrich, Pty Ltd, Castle Hills, NSW, Australia) was prepared in the appropriate media for each cell line using an aseptic technique.

Citrate buffer 0.1 M, pH 4.2: Combine 0.03% H2O2: 9.802 g citric acid, 14.4 g sodium citrate-2H20. Adjust pH to 4.1 and add 1 mL 30% H2O2. Bring volume to 1L with dH20. Stored at 4°C.

Dulbecco’s phosphate buffered saline (DPBS) for tissue culture, was prepared according to manufacturer’s instructions (Invitrogen Australia Pty Ltd, VIC, Australia), followed by filtration by Stericup®& SteritopTM vacuum-driven filtration systems (0.22µm) (Millipore Co, MA, USA), pH 7.2. Stored at 4°C for use.

Gel composition (Tris-HCl 375 mM, acrylamide/bis 8-12% (w/v), ammonium persulfate 0.008% (w/v), TEMED 0.008% (w/v)).

Immuno-precipitation cell lysis buffer: (10 mM sodium phosphate, 0.15 M sodium chloride, 2 mM EDTA, 50 mM NaF, 1% (v/v) sodium orthovanadate 115

(Na3VO4), 1% (v/v) protease inhibitor cocktail and 1% (v/v) Triton X- 100). Stored at −20 °C.

IPG strip re-hydration buffer (RB): 7M urea, 2M thiourea, 1% (w/v) CHAPS, 50 mM DTT, applied for protein assay analysis and IPG-rehydration. Stored at −20 °C. Freeze the solution in 1 mL aliquots without DTT & Ampholyte, store at-80 °C for up to several months. Before use, 0.1 % (w/v) Bromophenol Blue was added,

Laemmli sample buffer4X: 240 mM Tris-HCl, pH 6.8, 20% (v/v) glycerol, 4% (v/v)

SDS, 0.02% (w/v) Bromophenol Blue were made up in Milli-Q H2O. Solution was approximately pH 6.8 and stored at −20 °C. This solution was especially formulated for protein sample preparation to be used in SDS-PAGE system.

Laemmli sample buffer2X: 50mM/L Tris-HCl pH 6.8, 10% glycerol (v/v), 2% SDS (v/v), 100mM/L DTT, 0.01% (w/v) bromophenol blue.

Lysis buffer: (Salts50mM tris/HCL, pH 8.0, 10 mM EDTA, 100 mM NaCl, 10 mM dithiothreitol (DTT), 0.1% SDS and protease inhibitor cocktail (1tablet/50mL: Sigma). Stored at −20 °C.

Protein extraction buffer: 50 mmol/L Tris-Hydrochloric acid (HCl) at pH 8.0, 150 mmol/L NaCl, 0.1% sodium dodecyl sulfate (SDS), 10 mmol/L NaF, 1 mmol/L Na3VO4, 0.5% sodium deoxycholate and 1% Triton X-100. Stored at −20 °C.

Phosphate buffered saline 10X: 1.4 M Sodium chloride (NaCl), 27 mM Potassium chloride (KCl), 81 mM Disodium hydrogen phosphate (Na2HPO4), 14.7 mM Potassium dihydrogen phosphate (KH2PO4). Adjust to pH 7.4. Stored at 4 °C.

Sodium Citrate 0.2 M, pH 3.5.Combine: 25 g Sodium citrate-2H2O, 40 g Citric acid and 7.2 g NaCl in 800 mL dH2O. Adjust pH to 3.5 and bring volume to 1 L with dH2O.

116

Sodium dodecyl sulfate (SDS)- 1D SDS-PAGE (2x) sample buffer: 0.5M Tris-HCl pH 6.8, 87% (w/w) Glycerol,10% (w/v) SDS, 0.1% (w/v) Bromophenol Blue, 2-5%

(v/v) 2-mercaptoethanol made up in Milli-Q H2O. Stored at − 20 °C.

Sodium orthovanadate100mM: prepare 100 mL of solution, adjust to pH 9 and boil the solution until colourless, cover to minimise volume loss and cool to room temperature and check pH. Adjust pH if necessary and coil again until colourless, (bring up to 100 mL), repeat procedure until solution remain at pH9 and colourless when cooled, store in small volume aliquots at -20 °C.

Tris acetate EDTA (TAE) buffer: 20 mM Tris, 0.11% Glacial Acetic acid (CH3COOH) and 0.2% EDTA, stored at RT.

Tris Buffer Saline10X (TBS) stock was prepared as follows: 60.4 g Tris and 87.6 g

NaCl was dissolved in 1 L Milli Q H2O, and pH adjusted to 7.6. Stored at 4 °C.

Tris Buffer Saline 1X (50 mM Tris-Cl, pH 7.6; 150 mM NaCl): Combine 6.61 g Tris-

HCl, 0.97 g Tris Base, 8.77 g NaCl in 800 mL dH2O. Adjust pH to 7.4 and bring volume to 1 L. Stored at room temperature (RT).

TBS -Tween buffer (TBST): consists of 0.1% (v/v) Tween 20 in 1 x TBS. Stored at RT.

Tris/glycine SDS running buffer10X: 250 mM Tris, 1.9 M Glycine, and 1% (v/v)

SDS in Milli-Q H2O. Dissolve 30.2 g Tris-base and 144 g glycine in Milli-Q-H2O, then

1 g SDS and mix. Add Milli-Q-H2O to a final volume of 1 L.

117

Breast cancer cell lines

All breast cancer cell lines are Homo sapiens and were purchased from American Type Culture Collection (ATCC, Manassas, VA, USA). The different primary and metastatic BC cell lines are summarised in Table 2-1.

BT-474 (ATCC® HTB-20™) derived from mammary gland, breast duct; 60 year old adult Caucasian woman with ductal carcinoma and maintained in Iscove's Modified Dulbecco's Medium (IMDM) supplemented with 10% (v/v) heated-inactivated FBS, 50 U/mL of penicillin and 50 µg/mL of streptomycin.

MDA-MB-231(ATCC®HTB-26™), epithelial cell type obtained from metastatic breast carcinoma, pleural effusion of a 51 years adult Caucasian woman and maintained in RMPI-1640 medium supplemented with 10% (v/v) heat-inactivated FBS, 50 U/mL of penicillin and 50 µg/mL of streptomycin.

MCF-7 (ATCC® HTB-22™) epithelial cell type obtained from metastatic breast carcinoma, pleural effusion of a 69 years adult Caucasian woman. Maintained in RMPI-1640 medium supplemented with 10% (v/v) heat-inactivated FBS, 50 U/mL of penicillin and 50 µg/mL of streptomycin.

SK-BR-3 (ATCC® HTB-30™), obtained from metastatic breast carcinoma, pleural effusion of a 43 years adult Caucasian woman. . Maintained in RMPI-1640 medium supplemented with 10% (v/v) heat-inactivated FBS, 50 U/mL of penicillin and 50 µg/mL of streptomycin.

118

Table 2-1. Summary of breast cancer cell lines.

Breast cancer Cell type ATCC® Tissue cell lines number BT-474 epithelial HTB-20™ mammary gland; breast/duct ductal carcinoma MDA-MB-231 epithelial HTB-26™ mammary gland; breast: pleural effusion epithelial adenocarcinoma MCF7 epithelial HTB-22™ mammary gland; breast: pleural effusion epithelial adenocarcinoma SK-BR-3 epithelial HTB-30™ mammary gland; breast: pleural effusion

Primary Antibodies

The primary, secondary and control antibodies used for validation studies of novel markers detected, are detailed in Table 2-2.

119

Table 2-2. Antibodies used for WB and IHC.

Antibody Source Type WB IHC Dilution & Dilution & Incubation Incubation

Mouse anti-Secretory Component Glycoprotein (SC-05) Abcam (ab3924) MAb 1:1000, 4°C 1:500, 4°C Glycoprotein Abcam O/N O/N Ab(EPR6701) (ab126629)

Mouse anti-human Abcam (ab13413) MAb 1:200, 4°C 1:100, 4°C Vitronectin (VN58-1) O/N O/N

Mouse anti-human Abcam (ab11591) MAb 1:500, 4°C 1:500, 4°C VIT-2 O/N O/N

Rabbit anti- human Abcam PAb 1:500, 4°C 1:100, RT CLUAP1 (ab198193) O/N 2hr

Rabbit anti-human C3 Abcam (ab97462) PAb 1:1000, 4°C 1:500, 4°C O/N O/N

Rabbit anti- human Abcam (24584) PAb 1:1000 1:100 Filaggrin

Rabbit anti- human Abcam (ab76001) PAb 1:50, 4°C 1:50, RT IGFBP3 O/N 2hr

Rabbit anti- human Abcam MAb 1:1000, 4°C 1:200, RT LRG1 [EPR12362] (ab178698) O/N 2hr

Rabbit anti- human Abcam PAb 1:500, 4°C 1:100, 4°C MAST4 (ab196777) O/N O/N Abcam (ab87734) Abcam (ab67854)

Rabbit anti-S100 Abcam MAb 1:10000, 4°C 1:100, RT alpha 6 antibody (ab181975) O/N 2hr [EPR13084-69]

Goat anti-mouse IgG- Santa Cruz IgG 1:2000, 4°C HRP Biotechnology, O/N Cat No. sc-2005

Goat anti-mouse IgG- Dako Cytomation IgG 1:200, 1hr HRP Cat No. P0447

120

Goat anti-rabbit IgG- Santa Cruz IgG 1:2000, 4°C HRP Biotechnology, O/N Cat No. sc-2004

Goat anti-rabbit Ig- Dako Cytomation IgG 1:200, 1hr HRP Cat No. P0448

Mouse anti-GAPDH EDM Millipore, Cat PAb 1:2000, 4°C No. MAB374 O/N

Mouse anti-human β- Sigma-Aldrich, Cat MAb 1:2000, 4°C tubulin No. T4026 O/N

Mouse IgG1 (negative Dako, Cat No. IgG 1:1000 1:100 control) X0931

Rabbit IgG1(negative Dako, Cat No. IgG 1:1000 1:100 control) X0903

Abbreviations: HRP, horseradish peroxidase; Mab, monoclonal antibody; O/N, overnight; PAb, polyclonal antibody; RT, room temperature.

121

SAMPLE COLLECTION

Urine samples collection

A mid-stream, clean catch (no skin contamination), 40-80 mL urine sample was collected from breast cancer patients and control subjects into sterile polypropylene urine containers. After collection, the urine samples were transported on ice to the Research Building, St George Hospital within 30 minutes (min), to prevent microbe contamination and proteolysis. The individual samples were encoded with the patient’s name and medical record number. Patients’ medical records were checked to ensure no other existing diseases were present. To confirm there were no minor ailments, the physical properties of the urine sample were examined for colour, odour and cloudiness. Normal urine is clear and light yellow (no blood) and should be clear (no turbidity) with pH 4.5-8.0.

Insoluble materials and cellular debris from samples were removed by centrifugation within 30 min of collection, to prevent protein release from these artefacts. Urine samples were centrifuged at 4,000 revolutions per min (RPM) at 4 °C for 10 min, supernatants carefully collected and aliquoted into smaller labelled tubes (1-2 mL) and stored at -20 °C for use. The samples were then transferred to a –80°C freezer for long term storage. Summary of the urine sample collection and handling protocol is shown in Figure 2-1. To prevent technical and analytical variation due to handling, all samples were collected, processed and stored following the same procedural conditions, and by the same laboratory personnel until a final protocol was established (discussed in detail in Chapter 3). Protease inhibitors were not added to the urine samples in this study, as the literature has not shown any major benefits, assuming that samples are frozen immediately.

122

Figure 2-1. Urine sample collection and handling protocol for proteomics analysis.

As an additional screening test, all the urine samples were pre-tested for protein and creatinine levels. Dipstick test results were negative for protein (displaying a reading of <15 mg/dL) and negative for microscopic haematuria. As an indicator of normal healthy renal function, all the urine samples included in the study, were correlated with the patients’ serum creatinine level (SEALS, St George Hospital Biochemistry department). Normal creatinine range for healthy women is 88-128 mL/min.

Plasma and serum sample collection

General blood collection protocol: Blood samples were drawn by peripheral venepuncture and blood serum/plasma collected according to standardised operating procedures. For each patient, 2-4 mL of blood was collected into each BD Vacutainer® blood collection tubes (Becton Dickinson (BD) Biosciences, North Ryde, Sydney, Australia) according to the sample being collected. After collection, all samples were appropriately inverted, then kept upright and transferred immediately to the laboratory. Serum samples were left to stand upright at RT for 30-60 min before centrifuging for 10 min, at 1300 x g relative centrifugal force (RCF) acceleration or 3,300 RPM angular velocity. Plasma samples were centrifuged (1300 x g for 10 min), within 30 min of collection. Serum and plasma samples were carefully removed and 100 μL aliquots of each sample were frozen at -80°C, at St George Hospital Cancer Care research laboratory for future analysis. 123

Figure 2-2. A flow chart for serum and plasma sample collection.

Four different blood collection tubes were used in this study. The Becton Dickinson (BD) Vacutainer blood tubes used for each patient are shown in Figure 2-3.

Figure 2-3. Serum and plasma samples tubes used for blood collection

Plasma samples were collected in two different BD Vacutainer® PST™ Plasma

Separation Tubes. The BD Vacutainer® K2E, lavender K2EDTA (Cat No: BD

367839) are plastic tubes spray-coated with K2EDTA. Anticoagulant EDTA (Ethylene diamine tetra acetic acid) functions by binding calcium ions thus blocking the coagulation cascade. BD Vacutainer® LH PST™ II, light green cap (Cat no:

124

BD367375), lithium heparin (anticoagulant) tubes were used to prevent blood clotting. After blood collection, the tubes are immediately inverted 8 times and then centrifuged (1300 x g for 10 min).

Serum samples were performed in two different serum collection tubes. BD Vacutainer®, Serum BD Hemogard™, Red top plastic tubes (Cat No: BD367812) with clot activator, silicone coated interior. And the BD Vacutainer® Plastic SST™ II Serum Separation Tubes with BD Hemogard™ Closure. Gold cap tubes (Cat No: BD367954) contain clot activator and gel for serum separation. During centrifugation, the gel forms a physical barrier between serum and blood cells. After blood collection in serum tubes, blood tubes were inverted 5 times to ensure mixing of the clot activator with blood, then allowed 30 min clotting time, followed by centrifugation (1300 x g for 10 min). Clotting time for serum blood tubes should not exceed 1 hour as the haemoglobin and other cellular proteins from lysed blood cells may contaminate the serum sample.

PROTEOMICS METHODS

During all procedures, all equipment and samples should be covered and gloves must be worn, to ensure there is no skin, hair or other contaminants.

Urine preparation for proteomics analysis

Acetone precipitation of proteins

Following the removal of cells and debris by low-speed centrifugation (LSC), pH levels of individual samples were measured and urinary proteins were isolated by precipitation. The lyophilised protein pellet was stored -20 °C for use.

Urine pH can vary from pH 4.5- 8.0, which may affect proteomic analysis results and should be adjusted to neutral during preparation and acidified for MS analysis. The characteristics of the sample (colour, turbidity and pH) were noted to assist with

125

data interpretation later. Protein concentrations were measured to determine the correct sample volume to be pooled for each breast cancer group and ensure the same total concentration of proteins was applied for proteomics analysis.

The pooled urine supernatants from each group were subjected to total protein precipitation, by combining one part urine sample volume to eight parts 100% ice- cold acetone at -20 °C (sample-solvent ratio 1 urine: 8 acetone) in acetone- compatible tubes. The acetone used must be HPLC grade or better (from Fisher Scientific). The samples were vortexed and allowed to precipitate by incubating for 1 hour at −20 C and then centrifuged (Eppendorf centrifuge 5804R) at high speed, 11,000 x g, for 30 min at 4C. The supernatant was carefully decanted and the protein pellet was air-dried for 20-30 min at RT (whilst lightly covered with Para- film to prevent air-borne contaminants), to evaporate all traces of acetone. Acetone protein pellets were stored at -20 C till used. All bench procedures and centrifugation were carried out at 4 C. An overview of the procedures used in our investigation is illustrated in Figure 2-4.

Figure 2-4. Protein purification and precipitation protocol with acetone.

126

Tri-chloroacetic acid (TCA) protein precipitation

TCA solution was freshly prepared for each precipitation procedure. Ten grams of TCA in 10 mL Milli-Q H2O was used. The urine samples were precipitated with one part fresh TCA solution (100%, w/v) to four parts of each urine sample (4:1 sample- to-solvent ratio), vortexed and incubated at 4 °C for 1 hr. A milky white sample indicated proteins were present. The samples were centrifuged at 11,000 x g for 30 min at 4 C, and the supernatants were clear with a white protein pellet. After carefully discarding the supernatants, the protein pellets were washed with 1 mL 100% ice-cold acetone (resuspend pellet completely), vortexed well to ensure complete resuspension and allowed to stand for 15 min, then centrifuged (11,000 x g for 15 min at 4 C). The acetone wash was repeated twice, to remove all traces of TCA. Following the final wash, acetone was carefully discarded and the pellets were completely air-dried at RT to remove all acetone traces, as the presence of residual acetone will make sample re-suspension more difficult (Beretov, Wasinger et al. 2014).

Ultra- filtration method

Ultra-filtration approach involves the use of a high performance ultrafiltration centrifugal filter device, which was carried out according to the procedure provided by the manufacturer (Amicon® Ultra-15 Centrifugal Filter Units, Millipore Bedford USA) in order to concentrate the sample. Briefly, urine samples (no more than 15 mL) were centrifuged at 11 000 x g, for 45 min to reduce the initial volume of urine to 500 μL.

Glyco amino glycan (GAG) precipitation

GAG precipitation method was modified based on previously published protocols (Verdier, Dussol et al. 1992, Catterall, Rowan et al. 2006). In my study, the GAGs were precipitated using 3 mg of Cetyl pyridium chloride (CPC) for every milligram of GAG. Each protein pellet was incubated in a 5% CPC solution at RT for 30 min (CPC solution-to-protein pellet ratio 3:1), centrifuged at 11,000 x g for 30 min at 4°C. 127

To disrupt CPC-GAG complex, following CPC treatment, pellets were then washed with 1M NaCl (sodium chloride) solution for 10 min and re-pelleted by centrifugation at 11,000 x g at 4°C for 30 min. Wash step was repeated twice.

Sonication-“cell shearing”

Urine protein pellets were resuspended with 100 µL lysis buffer (7M urea, 0.1 M CHAPS (3-[(3 cholamidopropyl) dimethylammonio]-1-propanesulfonate), 0.1 M dithioerythritol (DTT), and 35 mM Tris-base) and 0.1g of zirconium beads (0.1mm diameter) were added to the suspension. Samples were sheared at 5000 RPM in a mini-bead beater (shown in Figure 2-5), for 90 seconds and then kept on ice for 5 min to limit heating. This procedure was repeated three times. Initially, the samples were centrifuged at 10,000 x g for 10 min on a mini spin bench top centrifuge to clarify, and then further centrifuged at 11,000 x g for 45 min at 4 C. The supernatant was carefully collected, to prevent disruption of the bead pellet.

Figure 2-5. Mini-bead beater: Biospec product

128

Solubilising the protein pellet

Urine protein pellets were resuspended and solubilised in 100 µL of SDS re- rehydration buffer (RB) solution before use and vortexed to ensure the pellets were completely dissolved. There were several different RBs applied in this study. The preparation methods are detailed in Chapter 2.2.2.

RB consists of chaotrophes 2 M thiourea, 7 M urea, detergent 1% (w/v) 3-[(3 cholamidopropyl) dimethylammonio]-1-propanesulfonate (CHAPS), reducing agent 50 mM DTT and 0.1% (w/v) Bromothymol Blue to follow the run of protein sample on the gel. Alternatively, for protein sample preparation the pellets were resuspended in the TruSep buffer 3:1 for urine samples and/ or 2x Laemmlii buffer for serum and plasma samples 2:1.

Protein quantitation (2D Quant)

The protein pellets were dissolved in the appropriate RB which contains detergents, reductants, chaotropes and carrier ampholytes compatible with this protein assay. Protein assays monitor protein concentration and allow control of experimental loading and yield. Protein concentrations were determined by 2-D Quant Kit (GE healthcare-Life sciences, Product code 80-6483-56), according to the manufacturer’s instructions. In this colorimetric assay, once the sample was precipitated, the copper ions specifically bound to protein. The assay was run with serial dilutions of bovine serum albumin (BSA) as a reference. Absorbance was read at 480 nm using a Kinetic Microplate Reader (Bioclone, Australia).

A standard curve was generated by plotting the absorbance of the standards against the quantity of protein. This standard curve was used to determine the protein concentration of the samples. The standard curve was also used to create a trend line and a linear equation established was used to calculate the protein values with R2> 0.99.

129

Tris/glycine SDS-PAGE

SDS polyacrylamide gel electrophoresis (SDS-PAGE) analysis was used to separate the proteins in the sample. The SDS sample buffer was used to denature proteins and make them negatively charged prior to being loaded. In this manner, each protein migrated in the electrophoretic field in a measure proportional to its length.

Tris/glycine SDS running buffer

All protein samples were diluted 3:1 with TruSep SDS buffer (NuSep) or 4x Laemmli sample buffer, mixed well and denatured by heating at 95 °C for 5 min, then kept on ice. Using a micropipette with gel loading tips, the MW marker and the total prepared protein samples were loaded onto a 12% Tris/glycine precast gel or 4- 20% gradient SDS-polyacrylamide gel (10 cm in length). The gels were run at a constant 180V, 30 mA for 1 hour in Tris/glycine SDS running buffer detailed in Chapter 2.2.2 (25 mM Tris, 192 mM Glycine, and 0.1% SDS in Milli-Q water). The run was terminated when the bromophenol blue ran to the bottom of the gel. The protein bands were visualised by staining in Silver stain or CoomassieTM Blue R-250, for 1 hour on a gentle shaker. The staining gel image was acquired by EPSON scanjet- 5100 scanner and imported into image analysis software program, PDQuest Ver 7.3.1 (Bio-Rad, Hercules, CA, USA).

Silver staining

Following electrophoresis, the gels were fixed for 30 min in a solution of 30% (v/v) ethanol, and 10% (v/v) acetic acid, rinsed in 5% (v/v) methanol for 5 min and washed 3 x 5 min in Milli-Q H2O. The gels were then sensitised in 0.8 mM sodium thiosulfate (2g/100 mL stock solution, 1:100 working solution was used) for 2 min, followed by 3x 5 min wash in Milli-Q H2O. After washing, the proteins were stained in 0.2% (w/v) silver nitrate for 30 min and the reaction was stopped using fresh Tris/ acetic acid solution (0.04% w/v Tris and 20 mL 0.02% v/v). The gels were washed 3 x 5 min with Milli-Q H2O and transferred to developing solution [3% (w/v)

130

Sodium carbonate, 0.05% (v/v) formaldehyde (37%), 20 uL of Sodium thiosulphate (2% (w/v) stock solution)], for up to 10 min on a shaker until the protein bands were visible. When an adequate degree of staining was achieved, the gel was transferred to the 5% acetic acid stop solution for at least 30 min.

Coomassie Blue R250 staining

Methanol (MeOH)/acetic acid solution was prepared for staining and de-staining:

50% (v/v) MeOH, 10% (v/v) acetic acid in Mill-Q-H2O. After electrophoresis, gel was exposed to MeOH/acetic acid solution for 1- 3 hours (at RT with gentle agitation and at least three solvent changes), to ensure adequate removal of SDS. The gel was stained with 0.1% (w/v) Coomassie Blue R250 (in 50% methanol, 10% acetic acid), for the 30-60 min to visualise the bands of interest and desired band intensity was reached. The gel was de-stained if necessary by soaking in 50% methanol, 10% acetic acid, until the background was nearly clear.

Protein clean up and digestion

Sample preparation for SDS-PAGE and MS-based analysis is a critical step in the proteomics workflow. The quality and reproducibility of sample extraction, and preparation significantly impact MS results. The aim of my study was to examine the protein present in breast cancer biological samples. Therefore, for optimal results, a customised protocol was designed to incorporate urine precipitation and concentration and blood fractionation. The protein lysates were desalted (salts and buffers were removed using C-18 tips), followed by in-solution digestion of protein fragments into peptides fragments, to prepare the sample for LC-MS/MS analysis, summary of sample work flow shown in Figure 2-6.

131

Figure 2-6. Sample preparation workflow for LC-MS/MS.

(Modified from original source: https://www.thermofisher.com/au/en/home/life- science/protein-biology/protein-biology-learning-center/protein-biology- resource-library/pierce-protein-methods/sample-preparation-mass- spectrometry.html).

In-solution digestion

The dried protein pellet was re-suspended in 20μL of solution: 8 M urea, 100mM ammonium bicarbonate (NH4HCO3=AMBIC), 3mM DTT, stirred in Eppendorf Thermomixer R (Eppendorf North America, Westbury, NY) for 1 hour at RT.

Samples for LC-MS/MS analysis were digested with Promega sequencing grade modified trypsin (12.5ng/µl sequencing grade trypsin, Sigma-Aldrich, St Louis, MO,

USA). Enzyme stock solution was diluted with 25mM NH4HCO3 AMBIC or 50mM acetic acid to obtain a 20μg/mL working solution. For each solution digest, 100 µg of urine protein pellet was resuspended in AMBIC, then digested with 2μL of 1μg/μL trypsin in an enzyme: protein ratio of 1:50 (w/w) and incubated overnight at 37 °C. The digestion was quenched by adding 10 μL of 1% (v/v) formic acid, adjusted to the final pH ~ 3. 132

In-solution digestion of plasma and serum samples: The 3 kDa, and 3-50 kDa peptide fraction was digested with trypsin to a final enzyme: protein ratio of 1:20 (w/w), and incubated overnight at 37°C. The reaction was stopped using 20 µL of 1% formic acid. The digested samples were completely dried in a Centrifugal Vacuum Concentrator (Savant SpeedVac® Plus SC210A) at low speed for 30 min.

C18 STAGE tip purification for LC-MS/MS analysis

Preparing the sample for MS/MS analysis requires clean up, concentration, and desalting prior to MS/MS analysis, as the salts ionize more efficiently than peptides and can suppress MS signals. Following trypsin digestion, the peptide samples were purified using C18 Stage Tips (Thermo Scientific, USA) following the manufacturer’s instructions.

Stage tipping, dried peptide samples were resuspended in with 20μL of 5% v/v formic acid (FA), test pH<3, then purified using C18 Stage Tips, 200μL tips (Proxeon, Odense, Denmark) as per manufacturers’ instructions. Briefly, each Stage Tip was initialised with 20μL of 50% v/v MeOH and 5% v/v FA and then re-equilibrated and washed twice with 20μL of 0.5% v/v FA. The sample was loaded into the Stage Tip and slowly passed through the tip to allow peptide binding to the C18 column. The tip was washed with 20μL of 5% v/v FA twice, then the sample was eluted into a clean Eppendorf tube with 20μL of 80% v/v ACN and 0.1% v/v FA. This procedure was performed three times. The eluate was then dried completely in the SpeedVac Centrifugal Vacuum Concentrator at low speed for 30 min, or until all ACN evaporated, without over drying. Sample was frozen at -80 °C for further use. Lyophilised samples were stored -20 °C and resuspended in 10 µL 0.1% formic acid for LC-MS/MS analysis. The strong cation exchange solutions and composition are detailed in Table 2-3.

133

Table 2-3. Strong cation exchange solutions.

Solution Composition in ddH2O (v/v)

Rehydration solution 50mM ammonium bicarbonate(v/v) pH 8.0

Priming solution 50% methanol

Equilibration solution 50% methanol, 5% formic acid

Equilibration and washing 5% formic acid

Buffer A: Wash solution 5% methanol

Buffer B: Elution solution 80% Acetonitrile: 0.1% formic acid

Blood preparation for proteomics analysis

Fractionation

Sample purification, concentration and clean-up are required for successful analysis of low-abundance proteins in serum/ plasma. Protein samples can be concentrated using concentrators of varying molecular weight cut-off (MWCO) ranges. To ensure compatibility with the membrane and reduce viscosity, the sample composition of organic solvent (acetonitrile, methanol), detergents (e.g. SDS) and glycerol should be <10% to none.

In general, the blood analysis requires pooled representative plasma or serum samples according to breast cancer type and blood collection tube. For each breast cancer stage and control group, a 100 µg protein was made up to 100 µL with DTT/Urea solution (5 mM DTT + 2M Urea). The two serum and two plasma samples for each breast cancer stage and control samples were then concentrated into

134

different MW fractions using a microcentrifuge, providing fractions < 50kDa, 3- 50kDa and< 3kDa, respectively. The fractions were vacuum centrifuge dried and stored at -20 °C.

Briefly, each ultrafiltration device was pre-rinsed with 30µL MilliQ-H2O by centrifugation (9,000 x g) at 4 °C for 10 min. The filtration device was placed into an eppendorf tube labelled with sample identification and < 50,000kDa. The 200 µL of sample was added to the 50,000kDa MW (Amicon® Ultra 50K) device and ultracentrifuged (9,000 x g) for 30 min at 4 C, shown in Figure 2-7. The recovered sample in the bottom of the tube was the less than <50,000kDa fraction. The contents of the filter, >50,000kDa fraction, was collected by inverting the device reservoir into a clean vial and spinning at 9,000 rpm for 10 min, stored at -20°C . The <50,000kDa fraction was then further separated into <3kDa using a 3,000kDa MW cut off filtration device (Amicon® Ultra 3K), centrifuged (9,000 x g) for 60 min at 4 C. The 3-50kDa was collected from the filter reservoir by spinning at 9,000 rpm on a bench top and centrifuged for 10 min.

Figure 2-7. Microcon fractionation procedure.

135

Acetone precipitation of serum and plasma

For precipitation of serum and plasma, 200 µL of each serum or plasma sample were mixed with 100% ice-cold acetone in a ratio of 1:4 (sample: acetone). Samples were incubated at -20 °C for 2 hours to allow protein precipitation and then centrifuged at 11,000 x g, 4 °C for 15 min. The supernatant was discarded and each serum/ plasma sample pellet was air dried.

LC-MS/MS analysis on protein samples

Protein identification was performed on all protein digests, using label-free LC- MS/MS quantification: Orbitrap Velos (LTQ-Orbitrap, Thermo Scientific, USA) in the Bioanalytical Mass Spectrometry Facility (BMSF), UNSW, shown in Figure 2-8. The mass spectrometer was linked to a micro auto sampler and a Nano flow HPLC pump, as per previously published method (Beretov, Wasinger et al. 2014). All urine and blood samples were run in triplicate.

Figure 2-8. LTQ Velos- Orbitrap at BMSF in UNSW

136

Digested peptides were reconstituted in 10 μL of 0.1% (v/v) formic acid and separated by nano-LC using an Ultimate 3000 HPLC and auto sampler system (Dionex Amsterdam, Netherlands). Mobile phase buffers A and B were prepared.

Buffer A includes 2% (v/v) ACN CH3CN, 0.1% (v/v) FA, 98% deionised water and

Buffer B includes 98% (v/v) CH3CH, 2% H2O, 0.1% (v/v) formic acid (80% (v/v)

CH3CN, 0.1% (v/v) formic acid, 20% (v/v) deionised water). The samples (0.6 µL, 2 µg total load) were loaded onto a micro C18 pre-column (500 μM × 2 mm, Michrom Bio-resources, Auburn, CA, USA) with Buffer A at 10 μL/min. After a 4-min wash, the pre-column was switched (Valco 10 port valve, Dionex, Houston, Texas, USA) into the line with a fritless nano column (75 μm diameter × 12 cm) containing reverse phase C18 media (3 μM, 200Å Magic, Michrom Bio-resources). Peptides were eluted using a linear gradient of Buffer A: Buffer B at a flow rate of ~250nL/min over 60 min and then washed with Buffer B for 1 min at a flow rate of ~250nL/min. High voltage (2000 V) was applied to a low volume tee (Upchurch Scientific, Oak Harbor, WA, USA) and the column tip positioned ~0.5 cm from the heated capillary (T = 280 °C) of an Orbitrap Velos (Thermo Electron, Bremen, Germany) mass spectrometer. Positive ions were generated by electrospray and the Orbitrap was operated in data-dependent acquisition (DDA) mode. A survey scan mass spectra was acquired in the Orbitrap in the 350-1750 m/z range with the resolution set to a value of 30 000 at m/z= 400 (with an accumulation target value of 1 000 000 ions), with lock-mass enabled. The 10 most intense ions (>5000 counts) with charge states +2 to +4 were sequentially isolated and fragmented within the linear ion trap using collisionally -induced dissociation with an activation q=0.25 and activation time of 30 milliseconds at a target value of 30 000 ions. The m/z ratios selected for MS/MS were dynamically excluded for 30 seconds to prevent repetitive selection of the same peptide.

Progenesis LC-MS/MS statistical analysis

Biomarker discovery was analysed using Peak Integration Progenesis ® software v4.0 (Non-linear Dynamics, Durham, UK). Progenesis is a data analysis program used to quantify proteins of interest in a label free sample based on expression and 137

characterisation. MS peak intensities were analysed using Progenesis LC-MS/MS data analysis software v4 (Nonlinear Dynamics, Newcastle upon Tyne, UK). Ion intensity maps from each run were aligned to a reference sample and ion feature matching was achieved by aligning consistent ion m/z and retention times. The peptide intensities were normalised against total intensity (sample specific log-scale abundance ratio scaling factor) and compared between groups by one-way analysis of variance (ANOVA, p≤0.05 for statistical significance) and post hoc multiple comparison procedures. Type I errors were controlled for false discovery rate (FDR) with q value significance set at 0.01. Results were reported as mean ± SD (normalised ion intensity score). The information was analysed on Progenesis and Scaffold. Proteins were considered to be significantly different at P <0.05; Q< 0.02; fold change >3.

Progenesis analysis involves performing multivariate statistical analysis on tagged and selected groups. Peptide searches were conducted to manage the export of MS/MS spectra, and the import of peptide ions from Peptide Search engines. Peptide filter parameters (p-value, fold change) were applied to manage peptide ions. Protein view provided validation and resolution of the peptide identified, and conflicts for data entered from Database Search engines. Finally, a report was generated for peptides and proteins.

Reference run

The raw LC-MS data files were imported and appeared as a 2D representation of the run. All the files were reviewed and the most appropriate file was selected to align all the other runs manually, then normalised and finally the alignment quality was reported. Vectors were added manually (shown in Figure 2-9) to refine the chromatogram alignment.

138

Figure 2-9. Chromatogram alignment to reference run.

Peak picking parameters

Progenesis CoMet finds the ions in the experiment, by examining each of the runs selected in the peak picking parameters window and defining filters for features based on retention time, m/z, maximum charge and isotope numbers.

The runs (with data files) for each experimental condition that were applied were selected. The sensitivity of the peak picking algorithm was adjusted to apply a noise estimation algorithm. The retention time was set at 0.08 min, rejecting ions eluted in shorter period. Peak picking limits for urine analysis were set as default with maximum charge of ions to be detected set at 5. The blood analysis was set at auto sensitivity between default and 4, with maximum charge set at 20.

Filtering features were applied to both urine and blood analysis that included properties and limits of interest for Progenesis analysis. In the urine data, no value was set for m/z. The experiment area inside 20-78 & 17- 49 minutes was viewed, at charge state 2+, 3+ and 4+ charges ions. In the blood data, no value was set for m/z. The experimental area inside 16-78 min was viewed, to include 2+, 3+ and 4+ charges ions. Non-matching isotopes features were deleted.

139

Additionally, sample groups’ conditions and data for comparison were defined and significant tags were applied (p<0.05, fold change >3 and q<0.02).

Peptide statistics

Principle component analysis (PCA) shows the abundance levels and grouping across the triplicate runs within the breast cancer groups along with BBD and the controls, as shown in Figure 2-10.

Figure 2-10. Principal component analysis.

Correlation analysis represented visually as a dendogram showing clusters of proteins according to how strongly correlated the proteins are. Various parts of the dendogram show significant features between the different blood tubes or the different breast cancer stages, were additionally picked and tagged. The standard expression profiles of a group with similar expression profiles which can be examined more closely, are shown in Figure 2-11 with the blood tube analysis.

140

Figure 2-11. Correlation analysis for blood collection tubes data.

Peptide identification

The parameter set during peptide identification was important for MS/MS data acquisition. The MS/MS data in the runs and features detected were matched to known peptides, using an external search engine (Mascot). Before exporting the peak lists, the query was further refined to include the most 10 most reliable MS/MS spectra (setting rank at >10), to reduce the query. The protein sequence and peptide match information, selected for homology and identity threshold, were exported (saved as xml. File). Protein Search Results.xml file, was imported into Progenesis and the features list was updated.

The behaviour of the assigned peptides at the protein level was looked at in ‘Resolve conflict’. The total number of conflicts in protein resolution was observed

141

and proteins with the greatest number of peptides were assigned. Peptides were considered to be confidently identified when matches had a high ion score >20 and were then assigned to a protein. Additionally, in Protein View tag filters for the selected features (p<0.05 protein; fold change >3 protein; q<0.02 protein) were applied to proteins of interest or significance to include in the final report. The report for the analysis based on the significant features, was exported to Excel for further statistical analysis.

Database Searching and Validation

All MS/MS spectra of differentiating peptides were searched against human nonredundant NCBInr database using the Mascot search program (Matrix Science, London, UK, www.matrixscience.com). The most intense fragment ions (top 500 for urine and 200 for blood) for each raw product ion spectrum were used for searches against the International Protein Index (IPI) human database (International Protein Index, version 3.56) using Mascot version 3.2.17. Peak proteins lists were generated from the raw MS/MS data using Mascot Daemon/extract_msn (Matrix Science, Thermo, London, UK), applying the default parameters, and submitted to Mascot 2.1 (Matrix Science London, UK).

The following criteria were set for urine protein identification: (1) species, Homo sapiens; (2) allowed one missed cleavage; (3) variable modifications, Deamination (NQ), Oxidation (M), Phospho (ST) and Phospho(Y); (4) peptide mas tolerance, ±6 ppm; (5) fragment MS/MS tolerance, ± 0.6Da; (6) default charge states peptide +2, +3 and +4; (7) maximum of one missed cleavage was allowed; and (8) enzyme specificity, none. All data were searched against a merged target and decoy database with an electrospray ionization (ESI) Ion Trap- Orbitrap Mass analyser. Additionally, all criteria for blood protein samples analysis remained except setting at (4) peptide tolerance was set at ±4 ppm; and (5) MS/MS tolerance, ± 0.4Da. The results were imported into Progenesis LC-MS/MS software and peptides were considered to be confidently identified when matches had a high ion score >20 and

142

peptides were assigned to a protein. An example of Mascot search result is shown in Figure 2-12.

For identification, the FDR was set to 0.01 on the protein and on the peptide levels. The peptide level score cut off for each of the runs was automatically adjusted to ensure a 1% FDR throughout the experiments. Peptides isoelectric points (pI) were calculated. In each fraction, peptides with calculated pI values beyond the (0.5 window around the average pI were discarded.

Summary of Mascot MS/MS Ion search parameters: Search parameters Setting applied Database Sprot Trypsin None Allowed up to 0 cleaves Taxonomy restriction set to Mammalian Variable modifications Deamination (NQ), Oxidation(M), Phospho (ST) and Phospho (Y)

Peptide Tolerance 6ppm (urine experiment); 4ppm (blood experiment) MS/MS Tolerance 0.6 Da (urine); 0.4 Da (blood) Peptide charge 2+, 3+, 4+ Data File D: LTQ-data/ Val/ Julia…… (Samba/val wasinger/Julia/orbitrap data……) Instrument ESI-Trap Report the top 500 hits (urine experiment); 200 hits (blood experiment)

143

Figure 2-12. Mascot search results detailed for DCIS 3-50kDa serum sample

Ingenuity pathways analysis

Ingenuity Pathways Analysis (IPA) (Ingenuity Systems http://www.ingenuity.com) is a web-based software application tool which is designed to extract biological information from large protein lists collated and generated through large scale LC- MS/MS experiments. This powerful analysis and search tool provides a high-level overview of the general biology, constructs possible protein networks that are associated with proteomics data and uncovers the significance of data and new targets within the context of biological systems.

144

In my study, the protein list was entered into the data base and the IPA Canonical Pathways Analysis tool was used to identify the signalling and metabolic pathways associated. The significance of the association between the dataset and the canonical pathway was measured in two ways, the fold-enrichment and the significance (p value). Pathway analysis, canonical pathway, heat maps and overlapping pathways analysis were used to determine the most significantly affected biological pathways.

VALIDATION METHODS

Cell culture

Breast cancer cell lines and cell culture

The human primary breast cancer cell line (BT-474) and metastatic cell lines (MDA- MB-231, MCF-7 and SK-BR-3) were obtained from the American Type Tissue Culture Collection (ATCC, Rockville, MD). Origin and supply of each cell line is detailed in Chapter 2.2.3. Tissue culture reagents were supplied by Invitrogen Australia Pty Ltd (Melbourne, VIC, Australia), unless otherwise stated. Other three metastatic breast cancer cell lines MDA-MB-231, MCF-7 and SKBR-3 were cultured in RPMI-1640 medium, supplemented with 10% (v/v) FBS, 50 U/mL of penicillin and 50 µg/mL of streptomycin. The primary breast cancer cell line (BT-474) was maintained in Iscove's Modified Dulbecco's Medium (IMDM) supplemented with 10% (v/v) heated-inactivated FBS, 50 U/mL of penicillin and 50 µg/mL of streptomycin. All cell lines were maintained in a humidified incubator at 37 °C and 5% CO2. All breast cancer cell lines were cultured in 75 cm2 culture flasks (Thermofisher Australia, Cat No 178883 and Cell Star supplier number 658175 Cat No T-3001-2E140711D or T- 3001-3L) with 20 mL complete medium, until the cells form a sub confluent layer, during this time complete medium is changed every few days or according to cell line requirement.

145

Dissociation and subculturing of cells

The following general procedure was used to rapidly remove adherent cell lines from the substratum while maintaining the cellular integrity. Cells with 80-90% confluence were rinsed twice with DPBS (pH7.2). Attached healthy cells were then detached from the culture flask, after adding 3 mL 0.25% trypsin/0.05% EDTA and incubating at 37 °C for 5 min. Trypsin action was terminated by the addition 2 mL of complete medium with FCS. The resulting cell suspension was collected and centrifuged at 1200 RPM (Eppendorf centrifuge 5804R) for 5 min at RT. The supernatant was discarded and the cell pellet was resuspended in fresh complete medium. Cell viability was assessed and cells were subcultured at 1 x 106 cells per 75 cm2 culture flask.

Cell viability using Trypan Blue

Cell preservation and cell thawing is critical and Trypan blue is a vital dye, used to determine cell viability. Reactivity is based on the fact that the chromophore is negatively charged and interacts with a damaged cell membrane. Based on this principle, non-viable cells with blue cytoplasm are excluded and all healthy cells with a clear cytoplasm are included in the viable cell count.

Briefly, adherent cells were trypsinised and detached from the flask, pelleted with centrifugation (1200 RPM for 5 min) and resuspended in serum free tissue culture medium. Cell viability was performed using 0.5 mL of the cell suspension and 0.1 mL of 0.4% Trypan Blue Stain, mixed well and then observed microscopically using a haemocytometer. Both live and dead cells were counted, and percent viability was calculated. Cell density was calculated (average number of cells x dilution x104cells/mL).

146

Cell preservation/cell thawing

Sub confluent cells were harvested, cell viability and cell number assessed (as detailed in section 2.5.1.3). Viable cells were resuspended in freezing medium to create a suspension of approximately 1 x 106 cells per ml. Cell freezing media consisted of 40% (v/v) FBS and 20% (v/v) DMSO (Sigma-Aldrich Pty Ltd, Castle Hills, NSW, Australia) and 40% (v/v) cell culture medium. Following pipetting, the cell suspension was aliquoted in 1 mL pre-labelled cryo-vials. The vials were then immediately transferred to the Nalgene® Cryo 1 °C Freezing Container (Thermo Scientific, MA, USA) which allows a -1 °C /min, cooling rate required for successful cell cryopreservation and recovery. The Nalgene® Cryo 1°C Freezing Container was then placed in the -80 °C freezer for at least 48 hr before the vials were transferred to the liquid nitrogen tank for long-term storage.

Frozen cell thawing

Frozen cells were removed from liquid nitrogen and immediately thawed in a water bath (Labec, NSW, Australia) at 37 C, within 1 min. The cells were added to 5 mL of completed medium (breast cancer cell lines MCF-7, MDA-231 and SKBR-3 used in this study were cultured in RPMI-1640, the BT474 in IMEM, details in Chapter2.2.3). After discarding the supernatants, cell pellets were resuspended with fresh complete media (RPMI or IMEM supplemented 10% FBS, 50 U/mL of penicillin and 50 µg/mL of streptomycin) and then transferred to an appropriate size flask. All cell lines were maintained in a humidified incubator at 37 °C and 5% CO2. Sub-confluent cells grown, were then harvested as indicated in Chapter 2.5.1.2.

Protein extraction

Protein from each cell line was extracted with lysis buffer containing: 50 mmol/L Tris-HCl (pH 8.0), 150 mmol/L Sodium chloride (NaCl), 0.1% SDS, 10 mmol/L NaF, 1 mmol/L Sodium orthovanadate (Na3VO4), 0.5% sodium deoxycholate, 1% Triton X-100, and 1/12 (v/v) protease inhibitor cocktail (Sigma-Aldrich, Pty Ltd, Castle Hill, NSW, Australia). Complete protease inhibitors were added to the cell lysis 147

buffers prior to use. All steps for protein extraction were performed at 4 °C to prevent protein degradation.

Once the cells achieved 80-90% confluence, the culture medium was discarded and cells were placed on ice. The adherent cells were washed twice with ice-cold DPBS (pH7.2). After aspirating the buffer, appropriate volumes of the protein lysis buffer (300µL per 5x106 cells/75 cm2 flask) were added to the cells. The samples were incubated for 20 min, on a gentle shaker to maintain constant agitation and ensure lysis buffer covered all of the cells. The cells were then vortexed for 30 seconds and adherent cells were scraped off the dish using a cell scraper. The cell suspension was gently transferred into a pre-cooled microcentrifuge tube. After centrifuging at maximum speed (14,000 rpm) for 10 min at 4 °C, the supernatants were aspirated and the pellets discarded. The protein lysate was placed in fresh labelled eppendorf tubes and stored at -80 °C until use for WB experiment.

Protein quantification (BCA assay)

Protein quantification was assessed using a BCA Protein assay kit (Thermo Scientific™ Pierce™ VIC, Australia), following the manufacturer’s instructions. The colorimetric detection and quantitation of total protein concentrations were performed to compare to protein albumin standard using bicinchoninic acid (BCA). The BCA standards and protein samples were loaded onto 96-well plates and BCA working reagent added. The plate was sealed and incubated at 37 °C for 30 min. The optical density was measured with a BIO-TEC micro-plate reader (BIO-RAD, Hercules, CA, USA) at A562 nm absorbance.

Unknown concentrations of protein samples were determined by comparing the absorbance readings of the samples to the standard curve generated from the BSA standards.

148

Western blot analysis

Western blot (WB) technique was used to detect the presence of specific novel protein molecules identified in the urine and blood samples from breast cancer patients. Additionally, WB was used to evaluate the MW of proteins of interest, and to assess the level of protein expression in the various breast cancer cell lines representing four different molecular subtypes.

Protein extraction and quantification

Whole cell protein lysates were extracted and isolated from each breast cancer cell line using cell protein lysis buffer as described in Chapter 2.5.1.6. Total protein concentration was measured using a BCA Protein assay kit as described in 2.5.1.7 on the cell extract, human urine protein pellet and human blood samples. Human urine protein pellets (following acetone/TCA extraction) were re-suspended in RB. This concentration was used to determine the volume required for each separation.

Western blot analysis of protein extracts

Protein expression levels were semi-quantified using WB analysis. Firstly, protein sample is resuspended with RB containing SDS detergent, to denature the proteins into unfolded linear chains and coats them with a negative charge.

Equal amounts of total protein from human breast cancer cell lines, human urine and blood samples were separated by gel electrophoresis according to their sizes by NuPAGE® 4-12% Bis-Tris gel (Invitrogen Australia Pty Ltd, VIC, Australia). Based on protein concentrations an appropriate volume of the sample was added (10 µg protein) and made up to 13 µL with lysis buffer or RB. Prior to loading the protein sample, 5µl NuPAGE® lithium dodecyl sulfate (LDS) sample buffer and 2µl reducing agent were added. The mixture was denatured by heating at 95 °C for 5 min then kept on ice. Marker (Precision Plus Dual Coloured Standard) and samples were added and gels were run at 200 Volts for 60 min in MOPS, SDS Running buffer

149

(Invitrogen Cat No. NP0001-02), as shown in Figure 2-13. Following separation, the proteins were then transferred onto a polyvinylidene fluoride membranes (PVDF) blotting membrane, in NuPAGE® transfer buffer, at 30 Volts for 2 hours. The membrane was blocked with 5% BSA (Sigma-Aldrich, Sydney, NSW, Australia) in Tris-buffer with 0.1% Tween-20 (TBS-T) for 1 hour, to prevent nonspecific reactions occurring. The membrane blots were incubated with specific primary antibody, at appropriate working dilutions (5% BSA (w/v) in TBST buffer), which specifically binds to the protein of interest, overnight at 4 °C.

Figure 2-13. Western blot gel run

Specific antibodies to proteins of interest and concentrations, are detailed in chapter 2.2.4. Biomarker analysis in urine, the following antibodies and dilutions were used: mouse anti-secretory glycoprotein (ECM1) MAb (1:500 dilution, Abcam, USA), rabbit anti-MAST4PAb (1:1000 dilution, Abcam, USA) and rabbit anti-filaggrin PAb (1:1000 dilution, Abcam, USA). Biomarkers analysis in blood, the following antibodies and dilutions were used: Rabbit anti-CLUAP1 PAb (Abcam ab198193 at 1:500 dilution), mouse anti- vitronectin MAb (Abcam ab11591 at 1:500 dilution), rabbit anti-IGFP3 PAb (Abcam ab76001 at 1:50 dilution), anti-LRG Mab (Abcam ab178698 at 1:1000) and S100-A6 rabbit MAb (Abcam ab181975 at 1:1000 dilution).

150

Following primary antibody, membranes were subsequently washed 3 x 5 min in TBST, then incubated with secondary (HRP)-conjugated antibody (Santa Cruz, CA, USA), goat anti-rabbit or goat anti-mouse secondary antibodies (1:2000 dilution), at RT for 1 hr. Membranes were again washed 3 x 5 min with TBST, immuno-reactive bands were detected using enhanced chemiluminescence (ECL) WB substrate (SuperSignal West Pico Substrate, Thermo Fisher Scientific, VIC, Australia and Pierce Chemical Co, Rockford, USA) and imaged using the ImageQuant LAS4000 system (GE Health care, USA). Semi-quantification of WB was conducted using ImageJ 1.49 (NIH). The protein bands on films were scanned and processed in Adobe Photoshop.

To confirm equal loading of protein lysates, urine and blood proteins on biological membranes, housekeeping antibodies GAPDH (MAb: EDM Millipore), β-tubulin (MAb: Sigma-Aldrich Pty Ltd, Australia) or mouse anti-β-actin MAb (1:2000 dilution, Abcam, USA) were incubated at the same time as the primary or the membranes were stripped (Restore Western Blot Stripping Buffer, Thermo Scientific, USA) and re-probed with housekeeping antibodies, then processed as above.

Immunohistochemistry

IHC was used to identify the location and distribution of target antigens tissues by staining with a specific antibody to the protein identified with MS.

Paraffin sections

Paraffin sections were used to investigate the expression of markers in human tumour samples. Paraffin blocks were prepared in Anatomical Pathology SEALS, St George Hospital Kogarah, as follows. The patients’ breast tissue was fixed in 10% neutral buffered formalin for 24-48 hours at RT. The breast tumours were sectioned and blocked by a pathologist. The blocked breast tissues were dehydrated through graded alcohols, put through three changes of absolute alcohol, three changes of xylene, followed by three changes of liquid paraffin and finally embedded in a 151

paraffin block. The human breast cancer tumour paraffin blocks, were serially sectioned at 4μm and mounted on Superfrost Plus adhesion slides (Lomb Scientific, Australia- Thermo Fisher Scientific, NSW, Australia). After drying at 56 °C for 30 min, one slide was stained with haematoxylin and eosin (H&E) to examine tissue histology for a diagnosis whilst the other slides were stored at -20 °C for IHC staining.

Immuno-staining procedure

Expression of the novel molecular markers was assessed using standard IHC procedure, to visualise proteins of interest. Briefly, paraffin sections of breast cancer tissues and normal breast tissues were de-paraffinised , followed by a graded series of alcohol (100%, 100%, 95%, and 75%), rehydrated in water and then washed in TBS (pH 7.5) for 5 mins.

To enhance antigen retrieval, tissue sections were subsequently immersed and boiled in 0.01 M citrate buffer (pH 6.0) for 20 min, prior to unmasking in Low pH Target Retrieval Solution (DAKO Carpinteria, CA, USA Thermo Fisher Pty Ltd, VIC, Australia), for 30 min,. Tissue sections were rinsed in TBS for 5 min, and then quenched with 3% hydrogen peroxidase (in methanol) for 10 min, and rinsed in TBS. After blocking in 10% (v/v) goat serum for 20 min and washing with TBS, the sections were incubated overnight (o/n) at 4 °C in primary antibodies according to molecular target (antibody applied and dilutions are detailed in Chapter 2.24).

After washing with TBS, slides were incubated with biotinylated secondary antibody goat anti-rabbit IgG (1:100 dilution), for 45 min at RT. After rinsing in TBS, immunoreactivity was developed with 3,3’ diaminobenzidine (DAB) substrate solution (Sigma-Aldrich, Pty Ltd, Castle Hills, NSW, Australia) containing 0.03% hydrogen peroxide (VWR International, QLD, Australia) as a chromogen, then counterstained with Harris Hematoxylin (Thermo Fisher Pty Ltd, VIC, Australia) for 1 min. and blued with Scott’s Bluing solution for 30 seconds. Tissue sections were washed in water, dehydrated (95%, 100%, and 100% alcohol), cleared in xylene and

152

cover slipped. Control slides were treated in an identical manner, and stained with an isotype matched non-specific immuno-globulin as a negative control. MDA-MB- 231 cell line, positive controls tissue sections of skin, breast cancer, colon cancer and lung (as recommended by the manufacturer Abcam and Genecard) was used as positive control.

Assessment of immunostaining results

IHC-staining was scored using light and confocal microscopic (Leica microscope, Nussloch, Germany), the percentage of the tumour cells with positive cytoplasmic or nuclear staining was assessed. Staining intensity was graded between 0 and 3. The criteria for assessment were as follows: 0 (negative, 0 %); 1 (weak, 10-45 %); 2 (moderate, 45-70%); 3 (strong, >70%) of the tumour tissue stained. Evaluation of tissue staining was performed independently (JB, YL) and confirmed by a pathologist at St George Hospital (EM). All specimens were scored blind and an average of scores was taken. A final percentage of positively stained cells was calculated by averaging the percent positivity across each patient’s tumour.

Tissue microarrays

The H&E stained sections of the tumour paraffin donor blocks of breast tissue were marked up for pathologist determined areas, showing various patterns of invasive ductal carcinoma, normal tissue and DCIS , as a guide to select the regions for sampling.

Tissue microarray (TMA) was created from these marked representative sites. A tissue array (Beecher Instruments, Sun Prairie, WI), was used to harvest 1 mm cores of tissue from the donor blocks prior to insertion into recipient blocks. Three cores of each patient tumour were harvested due to the limited size of the foci and the potential for missing the areas of interest with the 1 mm diameter harvesting needle. Cores of breast tissue were placed at recorded random positions on the TMAs to

153

facilitate accurate orientation under the microscope. A design of the array block with core samples identification and location was made to help the analysis process.

The TMA blocks were serially sectioned at 4 μm and mounted on Superfrost Plus adhesion slides (Lomb Scientific, Australia). Verification of the final pathology on the TMAs occurred at 2 phases of the IHC studies. Firstly, representative sections of every TMA were routinely stained with haematoxylin and eosin (H&E). The pathology of the lesions at this phase was verified by a pathologist (Assoc. Prof. E Millar). IHC for protein markers of interest was performed on the TMAs, and assessed by a breast pathologist (E.K.A.M.) blinded to protein marker and clinical outcome.

Statistical analysis

The Kaplan-Meier analysis allows estimation, or computing the probabilities of occurrence or survival even when patients drop out or are studied for different lengths of time (Bewick, Cheek et al. 2004, Rich, Neely et al. 2010). In medical research, it is often used to measure the fraction of patients living for a certain amount of time after treatment. Kaplan-Meier analyses of publically available mRNA expression data were performed on the data at http://kmplot.com/analysis/index.php?p=service&cancer=breast (Gyorffy, Lanczky et al. 2010). Genes of interest were analysed at a median cut point of mRNA expression to identify any association with outcome (recurrence free survival).

154

Chapter 3 Urine Preparation Protocol for Breast Cancer Biomarkers

As discussed in Chapter 1, MS has the potential to play an important role in the diagnosis and treatment of cancer, as a unique type of biomarker technology applied to biological specimens. Sample preparation is a critically important step, therefore appropriate and careful handling of biological samples is essential. A suitable and standardised urine protein purification technique is fundamental to maintain consistency and to allow data comparison between proteomic studies for urine biomarker discovery. Ultimately, efforts should be made to standardise urine preparation protocols. In this chapter, several purification methods are explored, to precipitate and concentrate the breast cancer urine samples for LC-MS/MS analysis.

The work in this Chapter has been published in: Beretov J, Wasinger VC, Schwartz P, Graham PH, Li Y. A standardized and reproducible urine preparation protocol for cancer biomarkers discovery. Biomarkers in Cancer. 2014 Nov 3; 6:21-7.

155

3 STANDARDISED URINE PREPARATION PROTOCOL FOR LC- MS/MS

The aim of this study was to develop an optimal analytical protocol to achieve maximal protein yield and to ensure that this method was applicable to examine urine protein patterns that distinguish disease and disease-free states. In this pilot study, seven different urine sample preparation methods to remove salts, and to precipitate and isolate urinary proteins were compared. SDS-PAGE profiles showed that the sequential preparation of urinary proteins by combining acetone and TCA alongside high speed centrifugation (HSC) provided the best separation, and retained the most urinary proteins. Therefore, this approach is the preferred method for all further urine protein analysis.

INTRODUCTION TO URINARY PROTEOME ANALYSIS

The identification of novel biomarkers for the early detection of cancer, monitoring cancer progression and assessing response to therapy holds promise for improving clinical outcomes. Currently, a major obstacle in the early detection of breast cancer is the development of methods that efficiently and accurately identify potential proteomics biomarkers. In breast cancer, most of the urinary markers identified to date are metabolomic markers (Cho, Jung et al. 2006, Gaikwad, Yang et al. 2008).

Urinary proteome analysis is attractive in clinical proteomics research as urine is relatively simple and easy to collect. It is commonly used for the diagnosis and classification of diseases (Adachi, Kumar et al. 2006). As a biological sample, human urine is abundant in proteins, which reflects the physiological and the pathological state of an individual (Said 2005). Due to its complex nature, containing compounds such as salts, peptides, oligosaccharides, and glycosaminoglycan’s can interfere with the electrophoretic migration of other proteins(Verdier, Dussol et al. 1992). Thus, precipitation and concentration steps are essential to purify and isolate the proteins of interest. The precipitation of proteins occurs in solutions of extreme ionic

156

strength, high concentrations of organic solvents and low pH. Most proteomic researchers use different protocols, devised to suit the biological specimen or analytes of interest. These protocols include ultra-filtration(Pieper 2008), ethanol precipitation (Thongboonkerd, Mungdee et al. 2009), precipitation with various concentrations of acetone(Sun, Li et al. 2005), acetone and TCA combination (Tyan, Guo et al. 2006, Kentsis, Monigatti et al. 2009), and combinations of ultra-filtration, TCA and ultracentrifugation(Pisitkun, Shen et al. 2004, Fung, Yip et al. 2005).

Although there are numerous published urine preparation protocols, when applied to the metastatic breast cancer specimens in our laboratory, these existing techniques could not efficiently desalt and concentrate the urine samples. No group of investigators to date have published standardised technical information on the preparation of urine to achieve maximal protein yield for breast cancer protein biomarkers. The validity of breast cancer biomarker discovery greatly relies on the handling of urine samples in a uniform manner, thus highlighting the importance of a standardised protocol. Here an efficient and reliable technical method for urine sample preparation is demonstrated, including the sequential preparation of urinary proteins by acetone and TCA in combination with HSC. The results establish that this approach can maintain consistency and reproducibility to allow urinary data comparison.

URINE PREPARATION FOR PROTEIN ANALYSIS METHOD

Urine collection protocol

Urine samples were collected from female patients (ranging 35-60 years) with metastatic breast cancer (n =15, mean age 46.64 ± 7.38) and age matched healthy disease free control group (n=18, age 50.24 ± 6.19). Collection and preparation of urine samples was performed as detailed in Chapter 2.3.1.

157

Protein extraction and precipitation techniques

To achieve a representative urinary proteome that portrays the group pattern sample variability(Liu, Shao et al. 2012), the frozen aliquots were completely thawed and mixed well. Accurate volumes of each donor supernatant were pooled into the breast cancer or control group according to protein concentration.

The pooled urine supernatants were then subjected to seven different urine protein extraction-precipitation methods, along with various combinations of these techniques. The main precipitation techniques included acetone, TCA, ultrafiltration (UF) and acetone plus TCA combined (these methods are labelled as methods 1-4). All precipitation procedures were performed at 4 °C. Methods 1-4 were initially centrifuged (Eppendorf Centrifuge 5804R) at a low speed centrifugation (LSC) 4,000 x g, then each procedure was repeated at a high speed centrifugation (HSC) 11,000 x g to attempt to further desalt and remove non-soluble materials. All supernatant washes for each technique were kept and analyzed for protein loss. All pellets were centrifuged at the appropriate speed as indicate, initially low then high. Briefly, the procedure for all the urine proteins precipitated methods applied (detailed in Chapter 2.4.1 and shown in Table 3-1) are as follows:

(Method 1) Acetone method: Eight parts of ice cold acetone were combined with 1 part of urine sample (1:8 urine sample-to-solvent ratio) and the mixture was stored at −20°C for 1 hour. Centrifuge at 4,000 x g, discard supernatant and air-dry pellet to remove residual acetone;

(Method 2) TCA method: One part TCA solution was added into 4 parts of urine (4:1 urine sample-to-solvent ratio) and the mixture was vortexed and then incubated for 1 hour at 4°C, then centrifuge at 4,000 x g;

(Method 3) Ultra-filtration method: Amicon® Ultra-15 Centrifugal Filter Units, Millipore, applied according to the procedure provided with the device, to concentrate the initial urine volume of urine to 500μL; 158

(Method 4) Combined Acetone/TCA method: Acetone precipitation was performed for 1 hour at −20°C (as per method 1), dried-off for 5 min at RT, followed by TCA precipitation (as per method 2), vortexed and then incubated for 1 hour at 4°C. All protein pellets were initially collected at low speed at 4,000 x g;

(Method 5) High speed centrifugation (HSC) method: This approach includes the application of centrifugation at 11,000 x g for 30 min after each precipitation method 1-4 (see Table 3-1: method 5) to concentrate the protein samples.

Additional purification steps which included either Glyco amino glycan (GAG) precipitation (Verdier, Dussol et al. 1992, Catterall, Rowan et al. 2006) or sonication- “cell shearing” were examined separately with each single precipitation method 1- 4, the combined techniques applied are listed in Table 3-1.

The details for the double precipitation methods include; (Method 6) GAG precipitation method: Following precipitation with the single methods (1-4) at high speed centrifugation (method 5), each protein pellet was incubated in a 5% Cetylpyridium chloride (CPC) solution (CPC solution-to-protein pellet ratio 3:1)at RT for 30 min, and then washed twice with 1M NaCl and (Method 7) Sonication-“cell shearing” method as described in Chapter 2.4.1.5. Briefly, the pellets were resuspended in 100 µL lysis buffer and 0.1g of zirconium beads (0.1mm diameter). Shear samples were at 5000 rpm in a mini-bead beater for 90 seconds and then kept on ice for 5 min to limit heating. Procedure was repeated three times, and samples were then centrifuged at 10,000 x g for 10 min. The supernatants were carefully remove and finally centrifuged at high speed for 45 min.

159

Table 3-1. Summary of all the urine precipitation techniques applied.

Method Precipitation Technique Sample Volume Precipitation Centrifugation Time Protein Applied Time/ Speed x g at 4C Pellet Temperature Extracted

1 Acetone sample: acetone 1 hour at -20 C Low speed centrifugation 30 min Y 1: 8 (LSC)

2 Trichloroacetic-Acid (TCA) sample: TCA 1 hour at 4 C LSC 30 Y 4: 1 3 Ultra-filtration (UF) 15 mL LSC 30 Y

4 Acetone/ TCA sample: acetone 1 hour at -20 C LSC 30 Y (1: 8) sample: TCA (4: 1) 1 hour at 4 C LSC 30 5 High speed centrifugation 11,000 30 (HSC)

1, 5 Acetone- HSC HSC 30 Y 2, 5 TCA- HSC HSC 30 Y

3, 5 UF- HSC HSC 30 Y 4, 5 Acetone/TCA-HSC HSC 30 Y 6 Glyco-amino glycan removal CPC: pellet 30 min at 26 C 11,000 30 (GAG) (3:1) 1, 5, 6 Acetone-HSC-GAG HSC 30 N

2, 5, 6 TCA- HSC-GAG HSC 30 N

160

3, 5, 6 UF- HSC-GAG HSC 30 N

4, 5, 6 Acetone/TCA- HSC-GAG HSC 30 Y

7 Cell shearing (CS) 100 µL lysis Bench top 10 buffer 11,000 45 1, 5, 7 Acetone-HSC-CS Y 2, 5, 7 TCA- HSC- CS N

3, 5, 7 UF- HSC- CS N 4, 5, 7 Acetone/TCA-CS Y

Abbreviations: CPC, cetyl-pyridium-chloride solution; Y, protein pellet was extracted; N: no protein pellet was extracted. Notes: The 7 urine preparation methods (1-7) including 16 individual approaches as shown. The main precipitation techniques are methods 1-4; HSC (5) has 4 combination approaches with methods 1-4; GAG (6) has 4 combination approaches with methods 1-4; Cell shearing (7) also has 4 combination approaches with methods 1-4.At the end of each extraction method the ability to achieve a protein pellet was shown.

161

The precipitated protein pellets from all methods were resuspended and solubilised in 100 µL of RB, 7M urea, 2M thiourea, 1% CHAPS, 50 mM DTT (methods section 2.2.2), for protein assay analysis and SDS-PAGE. The samples were subjected to protein quantitation using the 2-D Quant kit (detailed in Chapter 2.4.2), with BSA as a reference standard. Numerous protein assays were conducted and a standard curve was established for each assay. All samples were run in triplicate. The standard curve was used to create a trend line and a linear equation established was used to calculate the protein values with R2> 0.99.

Protein separation and examination

SDS PAGE

SDS-PAGE was used to determine the best method for the extraction and precipitation for LC-MS/MS. The method details are described in Chapter 2.4.4. Briefly, a volume of protein solution was taken to load 30 µg of protein, which was mixed with an equal volume of TruSep SDS sample Buffer (NuSep) and boiled for 5 min. The whole sample was run on the 12%T-Tris/glycine precast mini gels at 180 V, 50 mA/gel for 1 hour with Tris-glycine running buffer (25 mM Tris, 192 mM Glycine, and 0.1% SDS in Milli-Q H2O pH 8). Gels were either stained using a solution of Coomassie Blue R-250, 0.1% w/v in 10% Methanol (CB R-250) or silver stain as detailed in Chapter 2.4.4.3 and Chapter 2.4.4.2 respectively.

Protein desalting

Protein desalting was done with C18 Stage Tips (Proxeon) as recommended by the manufacturer and detailed in Chapter 2.4.5. Briefly, peptide fractions were digested with trypsin (12.5 ng/μL trypsin proteomic grade, Sigma-Aldrich, St. Louis, MO, USA) to a final enzyme: protein ratio of 1:100 (w/w). All the samples were digested and prepared for LC-MS/MS analysis using equivalent fixed amounts of protein starting material of 10 µg.

162

Proteomics LC-MS/MS analysis

Proteomics analysis was conducted on the protein digests by LC-MS/MS using a LTQ-Velos Orbitrap ETD (Thermo Electron, Bremen, Germany) as described previously(Coumans, Gau et al. 2014), detailed in Chapter 2.4.7. Briefly, C18-LC elution was conducted over 60 min linear gradients. The false discovery rate (FDR) was less than 2% at 95% confidence for peptides. Protein dataset (Peak lists) were generated using Mascot Daemon software (Matrix Science, London, England), and analysed using Peak Integration with Progenesis LC-MS (Non-Linear Dynamics, UK), detailed in Chapter 2.4.8.

RESULTS OF URINE PRECIPITATION STUDY

Urinary proteins identified with LC-MS/MS

The representative results for SDS-PAGE results shown in Figure 3-1 and LC- MS/MS data matching the gels samples (Table 3-2), clearly indicate that the quality and the variability of the urinary protein recovery were greatly affected by the preparation protocols used. Initially, all precipitation methods 1-4 which include acetone, TCA, ultrafiltration and acetone/TCA combination, were initially centrifuged at low speed, and no MS data was detected in these protein sample extracts (Table 3-2: A1-2, C1-3). Additionally, the excessive drag in the gel results from the metastatic breast cancer urine samples (Figure 3-1: C1-3), demonstrated a difference between the metastatic breast cancer and the control samples (Figure 3-1: A2). Therefore, additional desalting steps to clarify the sample further were required. To observe the optimal method for urine sample preparation for metastatic breast cancer and healthy control urine samples, seven different urine precipitation techniques were compared. This information was used to determine the method that could provide the highest resolution on SDS-PAGE and the greatest number of urinary proteins detected with LC-MS/MS.

163

Figure 3-1. The comparison of urinary precipitation methods on SDS-PAGE.

The effects of precipitation techniques and centrifugation on urinary proteins were examined on healthy controls (A-B, n=18) and metastatic breast cancer samples (C- D, n=15). The details of the technique employed, total protein concentrations and number of proteins identified by LC-MS/MS are summarised in Table 3-1.

164

Table 3-2. Summary of number of proteins identified with LC-MS/MS and total protein extracted with the different urine protein precipitation methods.

SDS-PAGE Total Protein N0 of Proteins ID with ID Protein Precipitation Technique Method N0 Extracted (µg) LC-MS/MS Normal control urine samples- LSC and HSC M Mass Marker A1 TCA LSC 2 215 ND A2 Acetone LSC 1 576 ND A3 TCA at HSC 2, 5 345 73 A4 Acetone at HSC 1, 5 905 113 A5 Ultra-filtration 3 187 47 A6 Acetone/TCA* at HSC 4, 5 987 149* Normal control urine samples, HSC B1 Acetone 1, 5 834 115 B2 Acetone and Cell shearing 1, 5, 7 676 70 B3 TCA 2, 5 476 79 B4 Acetone/TCA* 4, 5 1184 154* B5 Acetone/TCA - CPC 4, 5, 6 93, 55 ND B6 Acetone/TCA and Cell shearing 4, 5, 7 313 55

165

Metastatic BC urine samples- LSC C1 Acetone 1 155 ND C2 TCA 2 146 ND C3 Acetone/TCA 4 202 ND C4 Ultra-filtration at HSC 3, 5 108 ND C5 All CPC washes <30 # C6 All Acetone washes <30 # Metastatic BC urine samples- all HSC D1 Acetone 1, 5 865 117 D2 TCA 2, 5 589 ND D3 Acetone/TCA* 4, 5 1023 165, 167* D4 All TCA washes <20 # D5 All acetone/TCA washes <20 #

Abbreviations: CPC, cetyl-pyridium-chloride; HSC, high speed centrifugation; ID, identified; LSC, low speed centrifugation (x 4,000 g); N0, Number; ND, not detected; TCA, trichloroacetic acid. Notes: All urine precipitation methods are tabulated against the protein concentration, in both disease free control (shown in A-B) and breast cancer specimens (C-D). Corresponding gel images are shown in Figures 1. The majority of the techniques employed HSC centrifugation (11,000 x g) at 4C for 30 min (except A1-2, A5 and C1-3 where LCS was applied). * Highlights the precipitation technique (Acetone/ TCA at HSC), with the highest total protein extract and number of proteins. # indicates there was minimal loss of protein found in the washes. 166

A total of 20 gels were examined which included each protein extraction method, where a pellet was extracted (Table 3-1), being run at least 5 times across several different gels. The four representative gels demonstrated the protein extracts achieved following the application of the seven different precipitation techniques in various combinations (Figure 3-1). The total protein extracted and LC-MS/MS data, which corresponds to the SDS-PAGE results (Figure 3-1), are summarised in Table 3-2.

Combination approach with acetone, then TCA at HSC

Overall, the greatest number of urinary proteins and total protein yield, representing both the high and low MW proteins, as well as best gel resolution, were observed by combining 3 precipitation techniques (i.e. acetone, then TCA all using HSC, see methods 4 and 5) in both the control and breast cancer samples (Figure 3-1 and Table 3-2 in *A6, B4 and D3). The number of MS/MS spectra obtaining from the peptide and protein identified (using Mudpit approach with SCX and C18 separation prior to MS, ~13 hour run) for both labelled and unlabelled analysis (Table 3-2), showed the increased numbers of proteins identified in both the breast cancer (159, 167) and control samples (149, 154), compared to acetone alone (methods 1, 5) with only 117 and 115 proteins identified in the same samples. The efficiency of the acetone/TCA at HSC technique (methods 4 and 5) was supported by the increase in the concentration of the total protein extracted for each sample (around 1000 µg) and was also confirmed with minimal protein loss demonstrated in the washes (Figure 3-1 and Table 3-1: D5).

A Venn diagram (Figure 3-2) was used to demonstrate that an increased number of proteins were identified by LC-MS/MS using the acetone/TCA protein precipitation at HSC technique in both the control and metastatic breast cancer urine samples. Unfortunately, the applications of double precipitation with cell shearing showed no additional benefit and instead lead to the loss of proteins with each sequential extraction (Figure 3-1: B2, B6). 167

Figure 3-2. Venn diagram comparison of proteins identified by LC-MS/MS

The protein identified in the control and metastatic breast cancer samples following the application of two major precipitation techniques, acetone and acetone/TCA, were compared; (A) in the urine samples from health controls and (B) metastatic breast cancer patients. The mass spectrometry data (Table 3-2) for the protein extracts indicated that an increased number of proteins were identified in the acetone/TCA protein precipitates in both the controls (154) and breast cancer samples (167), compared to 115 and 117 respectively. The overlap represents the proteins in common.

168

Urinary proteins in breast cancer and control samples at high speed centrifugation

Comparing all the precipitation techniques (methods 1-4), the gel results clearly demonstrated that the HSC was essential to desalt and remove non-soluble materials, and isolate urinary proteins in the breast cancer samples (Figure 3-1: C1- 3 low speed against D1-3 HSC). As mentioned above, the LC-MS/MS data confirmed that urine proteins extracted using methods 1-4 with low speed centrifugation (LCS) were either not detectible or low (Table 3-2: controls A1-2 and breast cancer samples C1-4). However, the same methods at HSC could achieve a far superior extraction for both the control (Figure 3-1: A3-4, B1-4) and breast cancer samples (Figure 3-1: D1-3).

Precipitation with organic solvent and acid alone

Although the application of HSC (method 5) has been shown to be more effective in desalting and concentrating the urine samples, acetone (method 1, see Figure 3-1: A4, B1 and D1) was more successful than TCA (method 2) as shown on SDS-PAGE (Figure 3-1: A3, B3 and D2). This finding was further confirmed with the LC-MS/MS data from acetone (Table 3-2: A4, B1 and D1) and TCA (Table 3-2: A3, B3 and D2). The results also demonstrated that the combination of organic solvent and acid at HSC (method 4 and 5) provided a superior extract (Figure 3-1 and Table 3-2 A6, B4, D3).

Other precipitation methods

In addition to the above mentioned methods, ultra-filtration of urine was also found to not be an effective protein precipitation technique (Figure 3-1: A5 and C4), the protein extracts from this method had a low protein yield and were difficult to detect with MS (Table 3-2: A5 and C4).

169

To achieve an increased protein yield, a double precipitation method was applied in which urinary proteins were first precipitated by one of the methods 1-4 at HSC (5) and then the pellet was re-precipitated with the CPC method to remove the GAG.

Although this technique was attempted numerous times, it was ineffective at further purifying the samples. Instead, there was aggregation and high salt contamination on the gels (Figure 3-1: B5). Despite clean-up attempts using SCX and C18 stage tips, this method showed the least total protein extract and proteins were not detectible using MS (Table 3-2: B5), even though the washes indicated no protein loss (Figure 3-1: C5). This technique was very laborious with excessive sample handling and is not recommended.

In conclusion, the results indicate that the approach of acetone and TCA precipitation at HSC is an optimal method for increased protein yield, number of protein detected with LC-MS/MS, ease of handling and reproducibility compared to other approaches.

DISCUSSION

In the current study, my objective was to determine the best method to achieve the highest yield from urine samples for future biomarker discovery of breast cancer patients. To achieve a comprehensive comparison for the preparation of urinary proteins, seven different techniques were systematically investigated which included 16 individual precipitation approaches. To optimise the removal of proteins, the techniques applied examined the effect of organic solvent, acid, ultrafiltration. Initially the combination method of organic solvent and acid all at LSC and repeated at HSC were compared. Then, additional purification steps were analysed that included 8 combination approaches (Table 3-1). Although 16 procedures (single methods: 1-4; combination methods: 5-7) were investigated, only 11 of the techniques successfully achieved a protein pellet. The effectiveness of these different techniques was different as shown on SDS-PAGE and LC-MS/MS Table 3-2. SDS-PAGE allowed for the identification of abundant proteins. The clarity, drag,

170

aggregation and number of distinct bands are good indicators for the downstream success in the identification of proteins related to the contaminants, salts, chemical degradation, sampling degradation content within the samples.

LC-MS/MS is one of the most popular proteomic techniques and requires a very small volume of protein sample for analysis. Due to the low concentration of urinary proteins and high concentration of salt and metabolites, the purity of the urine protein pellet for proteomic analysis is critical. Therefore, my aim in this study was to consistently achieve a high quality recovery, urinary proteins extract for downstream analysis. The data presented here clearly demonstrates the importance of the precipitation and concentration method applied to urine samples. Although other literature shows a common theme for urine collection, processing and downstream analysis (Thomas, Sexton et al. 2010), my current study is only a guide to the development of a standard method. The precipitation of urine proteins occurs in solutions of extreme ionic strength, high concentrations of organic solvents and low pH. My findings indicate that the techniques applied to urine preparation described in the literature were unsuccessful when applied to our metastatic breast cancer urine samples. I found that aggregation and high salt concentration in most of the protein extracts from breast cancer urine made it difficult to obtain sufficient MS spectra. Following a detailed review of urine sample preparation, using different techniques for proteomic biomarker studies in breast cancer (Beretov, Wasinger et al. 2014), I modified the urine preparation techniques to improve protein yield for MS analysis.

In this study, the most robust and reproducible techniques were Acetone at HSC (method 1 and 5) and Acetone/TCA at HSC (methods 4 and 5). Overall, the best resolution on the gels and the highest number of protein were achieved using the precipitation with acetone/TCA at HSC in both the metastatic breast cancer and control urine samples. My results demonstrate an optimised urine sample preparation method which allows us to obtain the best urinary protein MS data. This recommended method can produce a protein rich fraction and limit protein loss. In addition, I also found that this approach is a robust and reproducible analytical 171

protocol that can separate the proteins from interfering compounds and achieve the maximum yield of protein precipitate with minimal sample handling. Furthermore, the protein fraction from this method can produce the highest resolution on SDS- PAGE and the greatest number of proteins by LC-MS/MS.

Urine sample preparation is a complicated process. There is not a standardised urine sample preparation method currently available. The experimental design for urine sample preparation should also include sample collection and storage as these are crucial to the final protein analysis. The pooled urine specimens, within disease and non-disease states, can enhance the population proteins to estimate the prevalence of disease in the population and dilute the natural biological variance for proteomic studies.

In summary, although some progress has been made, challenges still remain in the development of an optimal sample preparation method for proteomic analysis of urine for gel-based electrophoresis and label-free LC-MS/MS analysis. This study reports a protocol suitable for an efficient extraction of proteins from urine. It was demonstrated that the acetone/TCA at prolonged time of centrifugation at high speed is the most promising approach for urine preparation. And therefore, provides optimal conditions for urinary protein precipitation for our metastatic breast cancer samples. Our current recommended urine preparation approach also includes the assessment of biological characteristics to ensure that there are no major underlying conditions that could affect urine stability and its analysis. The discovery of novel proteins and the validity of biomarker discovery greatly rely on the handling of urine samples in a uniform manner and thus highlight the need for a reproducible standardised protocol.

172

Chapter 4 Breast Cancer Biomarkers in Urine

“A Doctor Examining Urine”. French painter Trophime Bigot (1579–1650).

Breast cancer is a complex heterogeneous disease and is a leading cause of death in women, as described in Chapter 1. Early diagnosis and monitoring progression of breast cancer are important for improving prognosis. Despite significant advances in proteomics research and biomarkers in breast cancer, a significant gap still exists from bench top to clinical application. Following on from the success of developing a method for urine protein extraction and precipitation for LC-MS/MS analysis, detailed in Chapter 3, this method was now applied to additional urine samples from breast cancer patients. Proteomic analysis using a label-free LC-MS/MS approach was conducted on urine samples from breast cancer patients to identify biomarker candidates.

The work in this Chapter has been published in: Beretov J, Wasinger V. C, Millar E. K. A, Schwartz P, Graham P. H, Li Y. Proteomic analysis of urine to identify breast cancer biomarker candidates using a label-free LC- MS/MS approach. PloS One. Published online 2015 Nov 6; 10(11): e0141876.

173

4 URINE BREAST CANCER BIOMARKER CANDIDATES

INTRODUCTION

Breast cancer is a major public health problem worldwide. Despite the widespread use of mammographic screening, which has contributed to reduced mortality, breast cancer is still the most common form of cancer among women. It can only be detected using mammography if there is a visible, detectible abnormality with architectural distortion or calcification, which correlates with the presence of several hundred thousand tumour cells. Once breast cancer has been biopsied and the diagnosis has been confirmed pathologically, the tumour is surgically excised. The complexity and heterogeneity of individual tumours play an important role in therapeutic decision making. Pathological examination is still the gold standard for diagnosis and assessment of prognostic indicators in breast cancer include tumour size, grade (degree of tumour cell differentiation), presence or absence of positive lymph nodes (metastases), IHC expression of key proteins such as ER, PR and HER2 (Lakhani, I.O. et al. 2012).

Although advances in breast cancer diagnosis have been made in the last decade, there are still many patients who cannot be diagnosed in the early stages of disease or monitored adequately for tumor recurrence using current techniques. To reduce morbidity and mortality from breast cancer, novel approaches must be considered for screening, early detection and prevention, as well as for monitoring cancer progression or recurrence. The early detection of DCIS or invasive breast cancer (IBC) may prevent the development of life threatening metastatic disease. Additionally, monitoring metastatic progression could identify early breast cancer recurrence and help guide therapeutic decision making.

Human urine is one of the most interesting and useful bio-fluids for clinical proteomics studies. Advances in proteomics, especially in MS (Yates, Ruse et al. 2009, Pan, Chen et al. 2011) have rapidly changed our knowledge of urine proteins 174

which have simultaneously led to the identification and quantification of thousands of unique proteins and peptides in a complex biological fluid (Thongboonkerd, McLeish et al. 2002, Adachi, Kumar et al. 2006). Proteomic studies of urine are highly informative, and have been successfully used to discover novel markers for cancer diagnosis and surveillance (Husi, Stephens et al. 2011, Hassanein, Callison et al. 2012, Lei, Zhao et al. 2012, Beretov, Wasinger et al. 2014) as well as for monitoring cancer progression (Linden, Lind et al. 2012, Zoidakis, Makridakis et al. 2012). Technological development combined with the addition of urine screening would increase the knowledge about patient status and further assist assessment and treatment in clinical practice. Proteomic analysis of urine holds the potential to apply a non-invasive method to identify novel biomarkers of breast cancer. However, investigation of urinary proteins from different stages of breast cancer patients using a LC-MS/MS proteomic approach has not been reported to date.

In this study, we used a label free LC-MS/MS technique to test the feasibility of urine as a source for breast cancer biomarkers and identify the urinary proteins for breast cancer diagnosis and monitoring progression. In the breast cancer urine samples one potential marker, extracellular matrix protein 1 (ECM1) which has been previously identified and associated with breast cancer (Lee, Nam et al. 2014), and two novel potential protein markers MAST4-microtubule associated serine/threonine kinase family member 4 and filaggrin were identified and were also validated in breast cancer cell lines. Furthermore, MAST4 was validated in a small number of primary tumour tissues and in the individual human breast cancer urine samples, demonstrating the link of these proteins with breast cancer. The proteins identified showed significant differences in abundance between the different breast cancer disease stages which provide a useful reservoir of biomarkers for the detection of early and advanced breast cancer.

175

MATERIALS AND METHODS

Ethics approval for the collection and use of human urine and tissue samples, are detailed in Chapter 2.1. The healthy disease free control group (n=20) were age matched with the breast cancer patients (range 35-70 years, mean age, 51 ± 10.5 years). Urine samples were collected prior to surgery, breast cancer tumour along with normal part of breast tissues were collected after surgery (St George Hospital, Sydney, Australia). The collected samples were evaluated and grouped in the analysis according to histopathology report, after diagnosis. The breast carcinoma typing and grading were performed by a pathologist according to the World Health Organization criteria (Lakhani, I.O. et al. 2012). Once the tissue specimens were pathologically assessed the urine samples were separated into breast cancer or benign breast disease (BBD) (n=6). The breast cancer samples were then regrouped into 3 different breast cancer stages: DCIS (n=6), early invasive breast cancer (IBC), with or without axillary LN involvement, but no distant metastases (n=8), and metastatic breast cancer (MBC) (distant metastases to viscera or bone, n=6). The histopathology characteristics and clinical features are summarised in Table 4-1.

Urine sample collection and processing

Urine sample collection

Clean catch, midstream 30-50 mL urine samples were collected in a sterile tube for all breast cancer patients and healthy volunteers. Urine samples were collected as detailed in Chapter 2.3.1, and immediately transported on ice. Briefly, urine was centrifuged at 4000 RPM, at 4 C for 10 min to remove insoluble materials and cellular debris. The supernatants were aliquoted and frozen at -20 C and then transferred to -80C for long term storage. All samples were handled by the same standard operating procedures and processed for storage within one hour of collection.

176

Table 4-1. Histopathology characteristics and parameters, of the patients in this study.

Patient Patient Tumour Tumour Histological DCIS LN Biomarkers HER-2 Group Number Size Grade Diagnosis +/- ER/PR status (mm) DCIS 6 25-48 3 DCIS

IBC 5 9-32 2 or 3 ductal or lobular present + +/+ - 3 14-22 2 or 3 ductal or lobular present - +/- -

MBC 6 10-23 2 or 3 ductal or lobular present + +/+ - 10-24 2 or 3 ductal or lobular present + +/- +

BBD 6 - - fibrocystic change, - - - - fibroadenoma

Notes: LN: Lymph Node involvement; +/-: positive involved/ negative not involved; Tumour grade 1-3 (Lakhani, I.O. et al. 2012); Abbreviations: DCIS, ductal carcinoma in-situ; IBC, invasive BC; MBC, metastatic BC; BBD, benign breast disease; ER, oestrogen receptor status; PR, progesterone receptor status; HER2, human epidermal growth factor receptor status.

177

Urine proteins precipitation

All urine samples had protein concentration and urine creatinine levels measured, and abnormal samples were excluded from the study. The appropriate volume of urine samples was then pooled within the appropriate group to ensure the same total concentration of proteins for proteomics analysis. Urine precipitation and concentration, along with protein quantitation previously published (Beretov, Wasinger et al. 2014) and described in Chapter 2.4. Briefly, pooled urine supernatants from each group were subjected to total protein precipitation by 1:8 sample-solvent ratio of ice-cold (-20 °C) acetone, mixed and stored for 1 hour at −20C, and then high speed centrifuged 11,000 x g for 30 min at 4C. The supernatants were removed and the pellets were further air-dried. To further precipitate and concentrate the proteins, the pellets were resuspended in 2 mL of fresh TCA solution at a 4:1 sample-to-solvent ratio, vortexed, incubated at 4 C for 1 hour and then centrifuged, 11,000 x g at 4 C for30 min. After carefully discarding the supernatants, protein pellets were washed twice with ice-cold acetone followed by HSC at 4 C for 15 min. All pellets were air-dried.

All protein pellets were resuspended in 100 µL of RB solution (2 M thiourea, 7 M urea, 40mM Tris-base, 1% 3-[(3 cholamidopropyl) dimethylammonio]-1- propanesulfonate (CHAPS), 50mM DTT and 0.1% Bromothymol Blue) before use, and vigorously vortexed to ensure the pellets were completely dissolved. The protein concentrations of samples were determined with 2-D Quant Kit method, described in section Chapter 2.4.3.

Urine sample protein clean-up and digestion

Trypsin digestion

Urine sample protein clean-up and digestion were performed as previously described in Chapter 2.4.5.1. Briefly, peptide fractions were enzymatically digested with trypsin. Lyophilised protein samples were reconstituted with 25 µL of 50 mM 178

Ammonium bi-carbonate (AMBIC) (pH 8). Trypsin (12.5 ng/μL trypsin proteomic grade, Sigma-Aldrich, St. Louis, MO, USA) was added to a final enzyme-to-protein ratio of 1:100 (w/w) and incubated at 37 C o/n. The reaction was stopped by acidifying the preparation to ~pH 3 using neat formic acid (FA), dried and in a vacuum centrifuge, and stored at -20 C.

Protein sample desalting and purification

Following trypsin digestion, the peptide samples were purified using Strong Cation exchange (SCX) and C18 Stage Tips (Thermo Scientific, USA) detailed in Chapter 2.4.5.2.

LC-MS/MS analysis of urine sample

LC-MS/MS proteomics analysis was performed to mine the breast cancer urine protein samples and identify the significant proteins in the different breast cancer stages. All the procedures in LC-MS/MS study are detailed in Chapter 2.4.7. Briefly, label-free LC-MS/MS quantification was performed using an Orbitrap Velos (LTQ- Orbitrap, Thermo Scientific, USA). All protein samples were run in triplicate. Trypsin-digested samples (0.6 µL, 2 µg total load) were reconstituted in 10 µL of 0.1% FA and were injected onto a nano-LC using an Ultimate 3000 HPLC with a micro C18 pre-column (500μM × 2 mm, Michrom Bio-resources, Auburn, CA, USA), then into line with a fritless nano column (75 μm diameter × 12 cm) containing reverse phase C18 media (3 μM, 200Å Magic, Michrom Bio-resources). Peptides were eluted over 60 min at a flow rate of 250nL/min and effluent was electro- sprayed into the LTQ mass spectrometer (Thermo Fisher Scientific Inc., Waltham, MA). Mass spectra was acquired in the 350-1750 m/z range, resolution set to a value of 30 000 at m/z= 400 with lock-mass enabled. The 10 most intense ions (>5000 counts) with charge states +2 to +4 were sequentially isolated and fragmented.

179

Label-free LC-MS quantitative profiling

MS peak intensities were analyzed using Progensis QI, LC-MS data analysis software (version 4.1, Nonlinear Dynamics, Newcastle upon Tyne, UK) as detailed in Chapter 2.4.8. Ion intensity maps from each run were aligned to a reference map and ion feature matching was achieved by aligning consistent ion m/z and retention times. The peptide intensities were normalised against total intensity (sample specific log- scale abundance ratio scaling factor) and compared between groups by one-way analysis of variance (ANOVA, p ≤ 0.05 for statistical significance). Type I errors were controlled by FDR with q value set at 0.02 (Storey and Tibshirani 2003, Karp, McCormick et al. 2007).

MS/MS spectra were searched and identified against the human protein database Uni-Prot database (downloaded January 2013) using the database search program MASCOT (Matrix Science, London, UK, www.matrixscience.com). Parent and fragment ions were searched with tolerances of ± 6 ppm and ± 0.6 Da, respectively. Searched peptide charge states were limited to +2 to +4. Deamination (M), Oxidation, and Phosphorylation was chosen as variable modifications. Only peptides with an ion score >25 were considered for protein identification. Proteins were considered to be significantly different at p< 0.05; fold change >3. The raw datasets have been submitted to ProteomeXchange submission PXD002524.

Generation of the heat map

A heat map was generated from the raw protein data, reflecting expression values in several conditions. The area under curve, of all MS1 peaks generated from comparisons among different stages of breast cancer in urine (Progensis data), was normalised to the mean of all AUC using TIBCo spotfire (Boston, MA, USA). The clustering method used is UPGMA and distance measure was Cosine correlation in logarithmic scale for rows. Columns were clustered using a Ward’s method with distance measured using Half Square Euclidean.

180

Characteristics of breast cancer cell lines

In this Chapter, the breast cancer cells, BT474, MDA-MB2-31, MCF-7 and SK-BR-3 were used. The details for the cell lines are listed in Chapter 2.2.3. All cell lines used were cultured following the method detailed in Chapter 2.5.1. Characteristics of the breast cancer cell lines detailing molecular subtype and receptor status are shown in Table 4-2.

Western blot analysis

Protein expression levels were determined by WB as previously described (Chapter 2.5.1). Different primary antibodies and secondary are detailed in Chapter 2.2.4.

Protein from each cell line was extracted as detailed in Chapter 2.5.1.1. Briefly, cells were rinsed twice with DPBS then lysed in cell lysis buffer for 20 min at 4 °C. The lysates were collected and centrifuged at 14,000 rpm for 10 min at 4 °C and the supernatants-protein extracts were collected and stored at -80 °C for WB experiment. Human urine protein pellets (following acetone/TCA extraction) were re-suspended in rehydration buffer (25mMol/L Tris-HCl (pH 8.0), 0.5% SDS). Protein concentration was determined by BCA assay kit and optical density measured with a BIO-TEC micro-plate reader as previous described in Chapter 2.5.1.7.

181

Table 4-2 Characteristics of breast cancer cell lines.

BC cell type Molecular Receptor Status Treatment Response subtype MCF-7 Luminal A ER+, PR+/–, HER2– endocrine responsive, Ki67 low, often chemotherapy responsive BT474 Luminal B ER+, PR+/–, HER2+ usually endocrine Ki67 high, responsive, variable to chemotherapy. HER2+ are, ZR75 trastusumab responsive. MDA-MB-468, Basal ER–, PR–, HER2– endocrine SUM190 EGFR+ and/or nonresponsive, often cytokeratin 5/6+, Ki67 chemotherapy

high, responsive MDA-MB-231 Basal ER–, PR–, HER2– Ki67, Intermediate response E cadherin, claudin3,4 to chemotherapy (Claudin low) and claudin7 low SK-BR-3, HER2 positive ER–, PR–, HER2+ Ki67 trastuzumab high, are: responsive, and MDA-MB-453 chemotherapy responsive,

Notes: The 4 breast cancer cells lines representing 4 molecular classifications of breast cancer, along with receptor status. Abbreviations: EGFR, epidermal growth factor receptor; ER, oestrogen receptor; HER2, human epidermal growth factor receptor 2; PR, progesterone receptor.

182

Protein expression of ECM1, MAST4, and filaggrin in breast cancer cells, BT474, MDA -MB-231, MCF-7 and SK-BR-3 were determined by WB as previously described in Chapter 2.5.2. Primary antibodies were obtained from Abcam as detailed in Table 2-2.

Briefly, proteins from human breast cancer cell lines and human urine samples were resolved on 4-20% Tris-Nupage gels (Invitrogen Australia Pty Ltd, Melbourne, VIC, Australia) and transferred to polyvinylidene fluoride membranes. Membranes were blocked with 5% BSA (Sigma-Aldrich, Sydney, NSW, Australia) in TBST for 1 h, and then incubated with mouse anti-secretory glycoprotein (ECM1) MAb (1:500 dilution, Abcam, USA), rabbit anti-MAST4 PAb (1:1000 dilution, Abcam, USA) and rabbit anti-filaggrin PAb (1:1000 dilution, Abcam, USA) at 4 °C o/n, followed by incubation in HRP-conjugated goat anti-rabbit or goat anti-mouse secondary antibodies (1:2000 dilution). Immunoreactivity bands were detected using ECL substrate (Pierce Chemical Co, Rockford, USA), and imaged using the ImageQuant LAS4000 system (GE Health care, USA). To confirm equal loading of protein lysates, membranes were stripped (Restore Western Blot Stripping Buffer, Pierce) and re- probed using housekeeping antibodies GAPDH (MAb: EDM Millipore), then processed as above. Images were processed in Adobe Photoshop.

Immunohistochemistry staining

Standard IHC procedures were used to visualise expression of ECM1, MAST4, and filaggrin as previously described in Chapter 2.5.3. Primary antibodies used were obtained from Abcam as detailed in Chapter 2.2.4.

Standard IHC procedures were used to visualise MAST4 expression as detailed in Chapter 2.5.2. Briefly, paraffin sections including breast cancer tissues and normal breast tissues were deparaffinised in xylene, followed by a graded series of alcohols (100%, 95%, and 75%) and re-hydrated in water followed by TBS (pH 7.5). Slides were subsequently immersed in boiling 0.1 M citrate buffer (pH 6.0) for 30 min to enhance antigen retrieval, treated with 3% hydrogen peroxide and then incubated 183

with primary rabbit anti-MAST4 PAb (1:100 dilution) overnight at 4 °C. After washing with TBS, slides were incubated with goat anti-rabbit IgG (Dako, North Sydney NSW, Australia) secondary antibody (1:100 dilution) for 45 min at RT. Sections were finally developed with DAB substrate solution (Sigma-Aldrich, Pty Ltd, Castle Hills, NSW, Australia) as a chromogen, then counterstained with hematoxylin and blued with Scott’s Bluing solution. Control slides were treated in an identical manner, and substituted with an isotype matched non-specific immunoglobulin as a negative control. The MDA-MB-231 cell line was used as a positive control.

Assessment of immunostaining

Staining intensity (0-3) was assessed using light microscopy (Leica microscope, Germany) at a x 40 objective as - (negative), + (weak), ++ (moderate), and +++ (strong) using our previously published method. Evaluation of tissue staining was done, independently, by two experienced observers. All specimens were scored blind and an average of grades was taken. If discordant results were obtained, differences were resolved by joint review and consultation with a third observer, experienced in IHC pathology.

RESULTS AND DISCUSSION

Circulating urinary markers in breast cancer

Label-free LC-MS/MS quantification was used to characterise the differential expression of urinary proteins in various human breast cancer stages. The urine samples were analysed from patients with DCIS, invasive and metastatic breast cancer, BBD and normal healthy control subjects (ANOVA p< 0.05; q<0.02). A reverse database was also searched to determine protein level FDR. Using Progenesis software to compare protein expression between all the samples, I

184

identified a total of 166 proteins with 1% FDR (for protein identification, determined by searching a reverse database).

Using the raw urine data and Tibco-spotfire software Inc. 2014, a biological heat map of clusters from different stages of breast cancer patients and normal health control subjects was produced. The representative data are shown in Figure 4-1. This analysis demonstrates datasets as clustered patterns which show an overview of the distribution of urine proteins represented according to their expression.

The data obtained with Progenesis LC-MS analysis was then applied to calculate the fold change (FC) as a normalised ratio for disease compared to healthy control subjects. This statistical analysis revealed, 59 significant urinary proteins with >3- fold change relative to the normal healthy control subjects. These protein profiles are all recorded Tables 4-3, 4-4, 4-5, and 4-6. A thorough review of the literature demonstrated that the 59 significant proteins changing in abundance have not been detected in human urine in either breast cancer or BBD. Several of these proteins identified were previously reported in blood, tissue and human breast cancer cell lines (associated references are shown in Tables 4-3, 4-4, 4-5, 4-6), supporting that these proteins detected are associated with breast cancer. Therefore, in this study the focus was on the unreported proteins that revealed changes in abundance, and their biological significance in breast cancer. In addition, several proteins typically associated with plasma (CO3, KV101, ALBU, A1AG1, FETUA, LAC2, TTHY, A1BG, CO6A1, FIBA, CERU and HAPT) which are known to be excreted in urine (Candiano, Santucci et al. 2010) were also detected. These proteins are shown in the Table 4-3 and Table 4-5, Table 4-6, with protein spots marked *. Several plasma associated proteins (except LAC2 and FIBA) were also found to be associated with breast cancer, further substantiating our findings.

185

Figure 4-1. Heat map analysis of urine proteins from breast cancer patients and control subjects.

Raw LC-MS/MS data used to create a heat map and dendogram of urinary proteins identified showing the up or down expression of urine proteins from different stages of breast cancer patients and normal healthy control subjects.

186

Classification of identified urine proteins

The 59 significant breast cancer urinary proteins (p<0.05, >3-fold) identified (see Table 4-3, 4-4, 4-5, 4-6) were classified according to their subcellular locations based on the Uni-Prot entry information available. Protein locations shown in Figure 4-2, demonstrated that 52% (32) of the proteins are secreted, 18% (11) cytoplasmic, 24% (15) membrane-associated, and 6% (4) grouped as others consisted of nuclear, mitochondrial, cell organelle or unknown sub-cellular origin. The majority of significant breast cancer related proteins detected are secreted and membrane associated in nature, either tumour or host in origin but associated with the presence of disease.

Urine protein distribution in breast cancer patients

To investigate novel urine biomarkers in breast cancer, proteins which were both increasing and decreasing in abundance needed to identify, and associated with breast cancer prognosis. Our proteomics screening data provided a list of signature proteins for breast cancer (Table 4-3 and Table 4-4) and benign disease (Table 4-6). Firstly, this signature list highlights 37 unique circulating proteins which were found to be expressed only in specific stages of breast cancer and not across all the urine samples. These breast cancer profiles included 24 up-regulated proteins (Table 4-3) and 12 down regulated proteins (Table 4-4). Additionally, 23 proteins were identified which appeared across the different breast cancer urine samples, some of which displayed similar patterns of protein expression in DCIS and IBC (Table 4-5).

187

Table 4-3. A list of urinary proteins identified by LC-MS/MS, uniquely associated with specific stages of breast cancer.

Human Uni-Prot Protein Description Pep. Score Fold SL Breast Cancer References Non Breast Cancer Access. ID ID References ID

DCIS

CO3 P01024 Complement C3 * Ř 2 92 3.9 S (Li, Orlandi et al. 2005, Goncalves, Esterni et al. 2006, Cho, Jung et al. 2010, Solassol, Rouanet et al. 2010, Zeng, Hincapie et al. 2010, Dowling, Clarke et al. 2012, Nasim, Ejaz et al. 2012, Profumo, Mangerini et al. 2013)

CYTA P01040 Cystatin-A Ř 4 227 3.7 C (Parker, Ciocca et al. 2008) (Leinonen, Pirinen et al. 2007, Chang, Wu et al. 2010) KV101 P01593 Ig kappa chain V-I region AG * 4 385 5.3 MF (Whiteside and Ferrone 2012) Ř

LRC36 Q1X8D7 Leucine-rich repeat-contain 2 73 3.7 CM protein 36Đ

MAST4 O15021 Microtubule-associated 3 73 4.2 C (Sun, Gu et al. 2006, serine/ threonine-protein Robinson, Kalyana- kinase 4 Sundaram et al. 2011)

188

ALBU P02768 Serum albumin * Ř 44 1974 4.1 S (Somiari, Sullivan et al. 2003, Kadowaki, Sangai et al. 2011)

CI131 Q5VYM1 Uncharacterized protein 3 93 4.4 U C9orf131

IBC

ANXA1 P04083 Annexin A1 Ř 6 334 3.0 C, PM (Shen, Chang et al. 2005, Cao, Li et al. 2008, Wang, Bi et al. 2010, Yom, Han et al. 2011) DYH8 Q96JB1 Dynein heavy chain 8, 3 79 3.4 C axonemal

HBA P69905 Haemoglobin subunit alpha Đ 3 89 3.1 C (Woong-Shick, Sung-Pil et al. 2005, Rho, Qin et al. 2008) IGHG2 P01859 Ig gamma-2 chain C region Ř 7 250 3.2 S (Cho, Jung et al. 2010)

ITIH4 Q14624 Inter-alpha-trypsin inhibitor 10 523 14.0 S (Mohamed, Abdul-Rahman et (Zhang, Bast et al. 2004, heavy chain H4 Ř al. 2008, Nasim, Ejaz et al. Abdullah-Soheimi, Lim 2012) et al. 2010) TRFL P02788 Lacto transferrin Ř 2 100 17.6 S (Schulz, Bollner et al. 2009) (Deng, Zhang et al. 2013) NGAL P80188 Neutrophil gelatinase-assoc. 2 93 3.3 S (Provatopoulou, Gounaris et al. lipocalin Ř 2009, Wenners, Mehta et al. 2012)

PEPA P0DJD8 Pepsin A Đ 8 362 12.8 S PEPC(Vizoso, Sanchez et al. 1995)

189

VTNC P04004 VitronectinŘ 2 58 4.0 S (Aaboe, Offersen et al. 2003, (Peng, Liu et al. 2013) Kim, Lee et al. 2009, Kadowaki, Sangai et al. 2011)

MBC

AGRIN O00468 Agrin Đ 4.5 S, CM (Klein-Scory, Kubler et al. 2010) A1AG1 P02763 Alpha-1-acid glycoprotein 4 182 4.4 S (Semaan, Wang et al. 2012) 1*Ř

FETUA P02765 Alpha-2-HS-glycoprotein * Ř 103 3.5 S (Yi, Chang et al. 2009, Opstal- van Winden, Krop et al. 2011) APOA4 P06727 Apolipoprotein A-IV Ř 2 3.6 S (Custodio, Lopez-Farre et al. 2012) CAH1 P00915 Carbonic anhydrase 1Ř 5 276 3.6 Mit (Somiari, Sullivan et al. 2003)

SULF2 Q8IWU5 Extracellular sulfatase Sulf-2 3 238 3.2 CS (Morimoto-Tomita, Uchimura Ř et al. 2005, Khurana, Jung- Beom et al. 2013) NEGR1 Q7Z3B1 Neuronal growth regulator 1 7 184 4.2 CM (Takita, Chen et al. (+2) Đ 2011)

SPRL1 Q14515 SPARC-like protein 1 Ř 3 100 3.4 S (Bellahcene and Castronovo (Yamanaka, Kanda et al. 1995) 2001, Esposito, Kayed et al. 2007, Turtoi, Musmeci et al. 2012, Vafadar-Isfahani, Ball et

190

al. 2012, Yin, Liu et al. 2013)

Notes: Accession ID, Human Accession identification; Uni-Prot ID, Protein identification based on the Protein knowledge base UniProtKB/Swiss-Prot ID (http://www.uniprot.org); Pep ID, Assigned Peptides Identified; Score, Mascot score; SL, Sub-cellular location as annotated in UniProtKB. Fold: Fold change for breast cancer samples against normal health control. The proteins of interest showing biological significance are underlined. For ease of navigation, all proteins reported in the literature associated with breast cancer are marked Ŕ or with other disease marked Đ. Plasma Proteins reported in Normal Urine*(Candiano, Santucci et al. 2010). Sub-cellular location abbreviations: C, Cytoplasm; CM, Cell membrane; CS, Cell surface; MF, Membrane fraction; Mit, Mitochondrion; PM, Plasma membrane; S, Secreted; U, Unknown.

191

Figure 4-2. Sub-cellular locations of the 59 significant urinary proteins.

Protein locations illustrate that 52% are secreted, 24% membrane-associated, 18% cytoplasmic, and 6% are either nuclear, mitochondrial, cell organelle or unknown sub-cellular origin. (p<0.05, >3-fold).

192

Our findings in the current study indicate that several significant urinary proteins, in the breast cancer samples, had a strong relationship with breast cancer stage and are potentially promising breast cancer markers. The profile lists contained several makers which have already been investigated and are known to be associated with breast cancer (Table 4-5). Therefore, the aim of this study was to highlight the novel additional proteins which are not yet reported. Three abundant proteins were found to be associated with pre-invasive breast cancer (i.e. DCIS patients) including Leucine-rich repeat-containing protein 36 (LRC36), MAST4 and a novel uncharacterised protein C9orf131 (CI131) (Table 4-3). Also in the DCIS samples, Secretogranin-1 was found to be decreased (Table 4-4).

When breast cancer invades beyond the basement membrane of the lobule or duct and into the stroma, it is treated as an invasive cancer, as it can then spread to LNs and distant organs. Finding markers to detect the cancer before it spreads would prevent life threatening metastatic disease. Several of the proteins detected in this study have been reported as markers of breast cancer in other biological samples which included ANXA1, Vitronectin, Lacto transferrin, ITIH4 and NGAL (see Table 4-3). However, there were six unreported potential markers of early IBC were identified in urine. These included Dynein heavy chain 8 (DYH8), Haemoglobin subunit alpha (HBA) and Pepsin A (PEPA >10 FC) (Table 4-3), along with uncharacterized protein C4orf14 (CD014, >200-fold), filaggrin (>30-fold) and Multimerin-2 (down >40-fold in DCIS, see Table 4-5), which are markedly elevated in the IBC samples. Protein DYH8 was previously detected in serum of normal health subjects (He, He et al. 2005). Although HBA was reported as a potential serum biomarker in ovarian (Woong-Shick, Sung-Pil et al. 2005) and colon cancer (Rho, Qin et al. 2008), only hemoglobin subunit β was reported to be elevated in the breast cancer patients (Nasim, Ejaz et al. 2012). Notably, Desmoglein-1, Kallikrein-1, Keratin, type II cytoskeleton 2 epidermal (K22E) and Poliovirus receptor (PVR) were all proteins significantly down regulated in IBC (Table 4-4).

193

Table 4-4. A list of proteins with decreased expression in urine, uniquely associated with certain stages of breast cancer(DCIS, IBC and MBC).

Human Uni-Prot Protein Description Peptides Score Fold SL Breast Cancer Non Breast Cancer Access. ID ID References References ID

DCIS ENOA P06733 Alpha-enolase Ř 3 85 7.8 C (Shih, Lai et al. 2010, Tu, (Liu, Chen et al. 2012) Chang et al. 2010, Zamani- Ahmadmahmudi, Nassiri et al. 2013) KNG1 P01042 Kininogen-1 Ř 9 469 3.6 S (Kim, Lee et al. 2009) (Liu, Liu et al. 2012, Wang, Wang et al. 2013) SCG1 P05060 Secretogranin-1 Đ 4 178 8.2 S (Yang, Wang et al. 2013)

IBC AMPN P15144 Aminopeptidase N Ř 5 228 3.1 C (Liang, Zhao et al. 2006, (Hashida, Takabayashi et Ranogajec, Jakic-Razumovic al. 2002, Surowiak, Drag et al. 2012) et al. 2006, Zoidakis, Makridakis et al. 2012) DCD P81605 DermcidinŘ 2 117 4.9 S (Porter, Weremowicz et al. (Stewart, Skipworth et 2003) al. 2008)

194

DSG1 Q02413 Desmoglein-1 Đ 2 55 3.0 S (Myklebust, Fluge et al. 2012) KLK1 P06870 Kallikrein-1 Đ 1 71 3.2 S HKK2-3 (Rittenhouse, Finlay et al. 1998, Black and Diamandis 2000) KLK5 (Avgeris, Papachristopoulou et al. 2011) KLK12-13 (Yousef, Chang et al. 2000, Yousef, Magklara et al. 2000) K22E P35908 Keratin, type II 27 1619 5.0 C cytoskeletal 2 epidermal PVR P15151 Poliovirus receptor Đ 2 86 3.3 CM & S PVRL4 (Pavlova, Pallasch et (Sloan, Eustace et al. al. 2013), (Fabre-Lafay, 2004) Monville et al. 2007)

MBC TRFE P02787 SerotransferrinŘ 12 511 3.8 S (Somiari, Sullivan et al. (+1) 2003, Zeng, Hincapie et al. 2010) VASN Q6EMK4 VasorinĐ 2 80 5.5 CM (Loftheim, Midtvedt et al. 2012) VMO1 Q7Z5L0 Vitelline membrane 3 110 5.4 S outer layer protein 1 homolog

195

Notes: Accession ID, Human accession identification; Uni-Prot ID, Protein identification based on the Protein knowledge base UniProtKB/Swiss-Prot ID (http://www.uniprot.org); Peptides ID, Assigned peptides identified; Score, Mascot score; SL, Sub-cellular location as annotated in UniProtKB. Fold: Fold change for breast cancer samples against control. The proteins of interest showing biological significance are underlined. For ease of navigation, all proteins reported in the literature in association with breast cancer were marked Ŕ or with other disease marked Đ. Plasma Proteins detected in Normal Urine*(Candiano, Santucci et al. 2010). Sub-cellular location abbreviations: C, Cytoplasm; CM, Cell membrane; S, Secreted. Abbreviations: DCIS, ductal carcinoma in-situ; IBC, invasive breast cancer; MBC, metastatic breast cancer.

196

Table 4-5. A list of differentially expressed urinary proteins in breast cancer and benign breast disease.

Human Uni-Prot Protein Description SL Pept Score DCIS IBC MBC BBD Breast Cancer Non Access. ID ID FC FC FC FC References Breast Cancer ID References

Up-regulated in DCIS APOA1 P02647 Apolipoprotein A-IŘ S 2 103 ↑5.5 ↓4 (Goncalves, Esterni (Zhang, Bast et al. et al. 2006, Kim, Lee 2004, Chen, Chen et et al. 2009, Cho, Jung al. 2010, Loftheim, et al. 2010, Hamrita, Midtvedt et al. 2012, Ben Nasr et al. 2011, Chen, Lin et al. 2013, Meng, Gormley et al. Lei, Zhao et al. 2011) 2013). ECM1 Q16610 Extracellular matrix S 3 69 ↑13 ↑30 (Xiong, Zhang et al. protein 1 Ř 2012, Nutter, Holen et al. 2014)

KV118 P01610 Ig kappa chain V-I MF 5 392 ↑8 ↑10 region WEA LAC2 P0CG05 Ig lambda-2 chain C MF 5 246 ↑5 ↑4 ↑11 regions* TTHY P02766 Transthyretin*Ř S, C 3 142 ↑97 ↓38 (Goncalves, Esterni (Zhang, Bast et al. et al. 2006, Nasim, 2004, Dekker, Ejaz et al. 2012, Boogerd et al. 2005, 197

Majidzadeh and Rompp, Dekker et Gharechahi 2013) al. 2007, Sigdel, Lau et al. 2008)

Up-regulated in IBC A1BG P04217 Alpha-1B- S 2 104 ↓8 ↑14 ↑10 (Zeng, Hincapie et (Kreunin, Zhao et al. glycoprotein* Ř al. 2010) 2007, Soggiu, Piras et al. 2012) CATA P04040 Catalase Ř CO 1 57 ↑4 ↑5 (Yeghiazaryan, Mamlouk et al. 2007, Glorieux, Dejeans et al. 2011, Panis, Victorino et al. 2012) CO6A1 P12109 Collagen alpha-1(VI) S 2 114 ↑4.5 ↑7 (Abba, Drake et al. (Fan, Gao et al. 2012, chain*Ř 2004) CO6A2 (Yi, Chaudhary, Peng et al. 2013) Bhatnagar et al. 2013)

FILA P20930 Filaggrin Đ N 1 60 ↑32 ↓11 (Pellerin, Henry et al. 2013, Scharenberg, Eckardt et al. 2013) HV303 P01764 Ig heavy chain V-III MF 2 80 ↑41 ↑6 region VH26

198

MMRN2 Q9H8L6 Multimerin-2 Đ S 1 34 ↓43 ↑6 (Soltermann, Ossola et al. 2008, Shield- Artin, Bailey et al. 2012) PI16 Q6UXB8 Peptidase inhibitor 16 M 2 92 ↓4 ↑9.5 ↑15 (Reeves, Dulude et Đ al. 2006, Freue, Sasaki et al. 2010) AMBP P02760 Protein AMBP Ř S 11 664 ↑11 ↑4 (Cho, Jung et al. (Braoudaki, 2010, Cohen, Wang Lambrou et al. et al. 2013) 2013) CD014 Q8NC60 Uncharacterized M 3 78 ↑228 ↓12 protein C4orf14

Up-regulated in MBC FIBA P02671 Fibrinogen alpha S 3 248 ↓4 ↓5 ↑3 ND FIBG (Dirix, Salgado chain* Đ et al. 2002) K1C10 P13645 Keratin, type I C 25 1475 ND ↓5 ↑4 ↓3 K1C 16,18- 19 (Chen, Cheng et al. cytoskeletal 10 Đ (Somiari, Sullivan et 2006) al. 2003, Rezaul, Thumar et al. 2010, Rower, Koy et al. 2011, Yi, Peng et al. 2013) Up-regulated in BBD CADH1 P12830 Cadherin-1Ř CM 2 83 ↓5 ↑4 (Wendt, Taylor et al. 2011)

199

E-cadherin (Qureshi, Linden et al. 2006) KV122 P04430 Ig kappa chain V-I MF 1 64 ↓21 ↑3 region BAN

Down-regulated in DCIS

FILA2 Q5D862 Filaggrin-2 Đ U 2 88 ↓3 ↓5 (Scharenberg, Eckardt et al. 2013) HORN Q86YZ3 Hornerin Đ C 1 51 ↓8 ↓15 (Fleming, Ginsburg et al. 2012, Pellerin, Henry et al. 2013) IGHG4 P01861 Ig gamma-4 chain C S 4 142 ↓12 ↓12 region NUCB1 Q02818 Nucleobindin-1 Đ C, M 1 32 ↓11 ↓200 (Zhu, Chen et al. 2013) TLR4 O00206 Toll-like receptor 4 Ř M 3 83 ↓43 ↓7 ↓4 (Gonzalez-Reyes, Marin et al. 2010, Yang, Zhou et al. 2010, Basith, Manavalan et al. 2012, Theodoropoulos,

200

Saridakis et al. 2012)

Down-regulated in IBC & BBD

PLAK P14923 Junction plakoglobin Ř C, 2 77 ↓7 ↓5 (Mukhina, Mertani M et al. 2004, Holen, Whitworth et al. 2012) MASP2 O00187 Mannan-binding lectin S 2 86 ↓10 ↓32 (Ytting, Christensen serine protease 2Đ et al. 2005, Fisch, Zehnder et al. 2011, Alves, Pereira et al. 2013)

Notes: The expression patterns of the various proteins either up-regulated (↑) or down regulated (↓), demonstrate an important relationship between the different stages of breast cancer. Accession ID, Human accession identification; Uni-Prot ID, Protein identification based on the Protein knowledge base UniProtKB/Swiss-Prot ID (http://www.uniprot.org); Pept ID, Assigned Peptides Identified; Score, Mascot score; SL, Sub-cellular location as annotated in UniProtKB. FC: Fold change for breast cancer samples against control. The proteins of interest showing biological significance are underlined. All proteins reported in the literature Ř indicate that these proteins have been reported to be associated with breast cancer or associated with other cancers Đ. Plasma Proteins detected in Normal Urine*(Candiano, Santucci et al. 2010). Abbreviations: C, Cytoplasm; CM, Cell membrane; CO, Cell organelle; DCIS, ductal carcinoma in-situ; IBC, invasive breast cancer; M, Membrane; MF, Membrane fraction; MBC, metastatic breast cancer; N, Nucleus; S, Secreted.

201

Table 4-6. A list of urine proteins up-and-down regulated in benign breast disease.

Human Uni-Prot Protein Description Peptides Score Fold SL Breast Cancer Reference Non Breast Cancer Accession ID ID Reference ID

BBD up-regulated

KV309 P04433 Ig kappa chain V-III region 1 35 1070 MF VG (Fragment) IGLL5 B9A064 Immunoglobulin lambda- 4 157 3.6 S like polypeptide 5 F184A Q8NB25 Protein FAM184A Đ 3 75 3.6 C

E2F8 A0AVK6 Transcription factor E2F8 3 102 3.9 N E2F-1 (Zacharatos, (Deng, Wang et al. 2010, Đ Kotsinas et al. 2004), E2F- Xanthoulis and Tiniakos 4 (Rakha, Pinder et al. 2013) 2004)

BBD down regulated ADML P35318 ADM (Adrenomedullin) Ř 1 46 10. 6 S (Martinez, Vos et al. 2002, (Nikitenko, Leek et al. Oehler, Fischer et al. 2003) 2013) A2ML1 A8K2U0 Alpha-2-macroglobulin- 2 84 16.0 S (Kadowaki, Sangai et al. like protein 1Ř 2011, Opstal-van Winden, Krop et al. 2011) 202

CERU P00450 Ceruloplasmin* Đ 9 382 4.8 S (Peng, Liu et al. 2013) DMKN Q6E0U4 DermokineĐ 1 56 4.1 S (Tagi, Matsui et al. 2010, Szabo, Tihanyi et al. 2012, Watanabe, Oochiai et al. 2012) HPT P00738 Haptoglobin*Đ 1 65 3.4 S (Bharti, Ma et al. 2004, Kwak, Ma et al. 2004) KV203 P01616 Ig kappa chain V-II region 1 75 37.9 MF MIL K1C9 P35527 Keratin, type I cyto - 27 1660 3.2 C K1C9 (Yi, Peng et al. 2013) skeletal 9 K1C18-19(Somiari, Sullivan et al. 2003, Rezaul, Thumar et al. 2010, Rower, Koy et al. 2011) SPRR3 Q9UBC9 Small proline-rich protein 13 450 4.0 C (Kim, Yu et al. 2012) (Cho, Jo et al. 2010) -3 Ř THBG P05543 Thyroxine-binding 3 154 5.6 S globulin

Notes: Accession ID, Human Accession identification; Uni-Prot ID, Protein identification based on the Protein knowledge base UniProtKB/Swiss-Prot ID (http://www.uniprot.org); Pep ID, Assigned Peptides Identified; Score, Mascot score; SL, Sub-cellular location as annotated in UniProtKB. Fold: Fold change for BC samples against control. The proteins of interest showing biological significance are underlined. For ease of navigation, all proteins reported in the literature in association with BC are marked Ŕ or with other disease marked Đ. Plasma Proteins detected in Normal Urine*(Candiano, Santucci et al. 2010). Sub-cellular location abbreviations: C, Cytoplasm; MF, Membrane fraction; N, Nucleus. Abbreviations: BBD, benign breast disease. 203

Desmoglein-1 was reported as a prognostic marker in anal carcinoma (Myklebust, Fluge et al. 2012). However, only Desmoglein-3 was found to be associated with breast cancer cells (Caruso and Stemmer 2011). Kallikrein-1 has not been assigned for a specific biological function, even though Kallikrein-2 and 3 are serum and tissue markers for diagnosis and monitoring of prostate cancer and breast cancer (Rittenhouse, Finlay et al. 1998, Black and Diamandis 2000). Poliovirus receptor (also called CD155/PVR) which was down regulated is known to have a key role in motility during cancer cell invasion, migration (Sloan, Eustace et al. 2004)and cell adhesion in breast cancer cell lines (Pavlova, Pallasch et al. 2013).

Breast cancer which has spread to distant sites within the body is classified as metastatic. A detection pattern for MBC could provide the opportunity for early therapeutic intervention. Our analysis results provide a list of proteins with numerous proteins already linked to breast cancer including A1AG1, FETUA, APOA4, CAH1, SULF2 and SPRL1 (see Table 4-3). In metastatic samples, the novel proteins detected include AGRIN, and Neural growth regulator 1 (NEGR1) (Table 4-3), as well as Fibrinogen alpha chain (FIBA) and Keratin type 1 cytoskeleton 10 (KIC10) (Table 4-5) which were exclusively elevated in the MBC samples. Only nerve growth factor 1 (NEGF1) was previously reported in relation to the survival and proliferation of breast cancer cells (Descamps, Lebourhis et al. 1998, Dolle, Adriaenssens et al. 2004). Additionally, in this group of patients, Vasorin and Vitelline membrane outer layer protein 1 homolog (VMO1) were both down regulated (Table 4-4).

Analysis of the differential expression patterns of urine samples across different stages of breast cancer highlights potential protein markers that can identify some similarities or differences between DCIS and IBC. DCIS is a non-invasive malignant process which can progress to IBC. If a biomarker or a panel of biomarkers could identify in DCIS, then early action taken may prevent IBC development. Therefore, the identified potential breast cancer markers have potential clinical significance.

204

The 23 proteins listed in Table 4-5 include 3 potential progression markers of breast cancer: immune response proteins immunoglobulin (Ig) kappa chain V-I region WEA (KV118), lambda -2 chain C region (LAC2) and ECM1.Protein ECM1 was elevated in breast cancer (Table 4-5). ECM1 is a secreted glycoprotein, previously reported to be associated with breast cancer metastatic bone homing (Nutter, Holen et al. 2014) and plays an important role in breast cancer progression (Lee, Nam et al. 2014). High levels are detected in aggressive tumorigenic cancer cell lines MDA435 (Han, Ni et al. 2001), in ductal breast carcinomas (Wang, Yu et al. 2003) and its expression is also correlated with poor prognosis (Lal, Hashimi et al. 2009) and metastatic potential in cancer (Wu, She et al. 2012). In the current study, this protein was detected in IBC urine samples and validated in primary and metastatic breast cancer cell lines, further confirming its link with breast cancer patients. This urine marker ECM1 demonstrates the potential for breast cancer diagnosis and monitoring.

Urine protein distribution in benign disease patients

Although some BBDs are associated with an increased risk of subsequent breast cancer, this condition is generally considered as a noncancerous disorder. In such cases, markers at this stage are important for early detection in high risk individuals and understanding of disease progression. In this study, several unreported proteins of interest in association with BBD (Table 4-5 and Table 4-6) were also identified. The 6 up-regulated proteins uniquely found only in BBD included FAM184A, transcription factor E2F8 and Cadherin-1 along with Ig-response proteins kappa chain V-III region VG (KV309, >1000-FC), lambda like polypeptide 5 (IGLL5) (Table 4-5), and Ig kappa chain V-I region BAN (KV122) (Table 4-6). Numerous down regulated proteins were also identified in the benign breast patients (Table 4-6). The most significant benign breast protein identified was Nucleobindin-1, which has a 200-fold change (Table 4-5). Nucleobindin-1 is a major intracellular calcium-binding protein previously detected in colorectal cancer cells after treatment with anti-tumour compounds (Zhu, Chen et al. 2013).

205

Interaction networks of human urine proteins

Ingenuity Pathway Analysis (IPA) Software (IPA; Ingenuity® Systems, www.ingenuity.com Release date: 05-02-2013, Ingenuity Systems, Redwood City, CA, USA) was used for canonical pathway enrichment analysis and the derivation of mechanistic networks. The breast cancer urine proteins detected were analysed with Ingenuity, to identify the major biologic pathways involved in breast cancer, The lists of proteins were uploaded directly into IPA for analysis, and functional pathways or networks with the highest confidence scores were then determined. Cell growth and proliferation analysis are shown in Table 4-3. The proteins detected from Table 4-3, 4-4 and 4-5 were found to be associated with tumour growth and progression (Figure 4-3A), suggesting that these proteins are involved in the inhibition or proliferation of various cells in breast cancer patients (Figure 4-3B). The enriched pathways associated with breast cancer urine are shown in Table 4-4, while the best scored networks selected are shown in Figure 4-4. Highly interconnected networks are likely to represent significant biological functions associated with breast cancer progression.

Proteomics results from this study demonstrated that the proteins detected in breast cancer urine are involved in the Liver X receptor LXR/RXR activation and acute-phase response pathways, which are active during inflammation and/or as a contribution of the immune response to cancer. Other pathways also enriched include production of nitric oxide and reactive oxygen species (ROS) in macrophages and IL12 signalling and production in macrophages Figure 4-5. Ingenuity pathway analysis highlighted cholesterol metabolism as significant in our breast cancer samples. Cholesterol is an essential structural component of the cell membrane and proliferating cells. Cancer cells are believed to have increased requirements for cholesterol. Cancer cells can increase lipid biosynthesis and uptake cholesterol from the bloodstream (Antalis, Uchida et al. 2011). It seems that

206

Figure 4-3. Ingenuity cell growth and proliferation analysis of the urine proteins detected in breast cancer.

IPA analysis showing: (A) cell growth and proliferation analysis showing the urine proteins detected in breast cancer patients are associated with tumour growth and progression and (B) results demonstrating that the strong biological involvement and the direct effect of these identified proteins have their roles in inhibition and proliferation of various cells in the body.

207

LDL-cholesterol enriched systemic environment promotes breast cancer progression by activating key signalling pathways and modulating cell behaviour.

Figure 4-4. In silico identification of interactive networks.

In silico identification of interactive networks using breast cancer urinary proteins analysed with LC-MS/MS.

208

Figure 4-5. Ingenuity pathway analysis showing the top enriched canonical pathways.

The enrichment of canonical pathways in the significant up-regulated urine proteins associated with breast cancer.

209

LDL-cholesterol signalling was shown to induce breast cancer proliferation and invasion (dos Santos, Domingues et al. 2014). Proteomics analyses demonstrated that the proteins involved in vitamin D/E binding, heme/iron binding and transportation, and lipid/steroid transportation/metabolic systems were down- regulated in colon cancer and the same set of proteins were down-regulated in the LXR/RXR activation and acute-phase response pathways, revealing a plausible mechanistic connection between vitamin D deficiency, iron homeostasis, and colon cancer.

These canonical pathway results show (as might be predicted) that multiple pathways and networks are involved in the systemic response to breast cancer and that intrinsic and endocytosis signalling pathways play a role, along with communication between the innate and the adaptive immune system. Increasing evidence indicates that the immune response plays an important role in breast cancer disease (Coffelt and de Visser 2015, Coffelt, Kersten et al. 2015). Therefore, these immune response proteins in urine have potential to be used as breast cancer biomarkers for diagnosis and monitoring. The interaction between these identified pathways in patients’ urine and breast cancer disease is complex and biologically significant. Further study of the roles and functions of these signalling pathways in breast cancer will be performed in the future.

Validation of the identified potential urine markers in breast cancer cell lines

Verification of the novel proteins identified was performed using WB and IHC studies. To find an association of identified potential urine protein with human breast cancer, one existing marker-ECM1 and another two selected novel protein markers MAST4, and filaggrin were evaluated in the human primary breast cancer cell line (BT474) and metastatic breast cancer cell lines (MDA-MB-231, MCF-7 and SK-BR-3) by WB. As shown in Figure 4-6A, ECM1 and MAST4 were positive in all 4 breast cancer cell lines and filaggrin was positive in the 3 metastatic breast cancer cell lines, suggesting the identified potential urine

210

markers from breast cancer patients, are closely associated with human breast cancer.

Preliminary validation of identified potential markers in human primary breast cancer tissues

To further investigate the clinical significance of our findings, preliminary validation with IHC of one novel protein marker MAST4 was conducted, using a small number of representative human primary breast cancer tissue samples, including DCIS and IBC patients and normal breast tissues. My results indicate that MAST4 was positive in 80% (4/5) of IBC and in 60% (3/5) of DCIS, respectively and no positive staining was seen in normal breast tissues (5/5). The typical staining results are shown in Figure 4-6B. These findings further strengthen the link of the identified potential urine marker with breast disease. Due to the limitation of patients’ tissue samples, a lager sample size is required in a subsequent study.

Validation of potential marker MAST4 in individual human breast cancer urine samples

In order to confirm MAST4 overexpression in the individual DCIS patients, the remaining breast cancer urine samples available were re-examined using WB. The results clearly indicate that high levels of MAST4 expression were found in the individual DCIS urine samples and low levels in invasive and metastatic breast cancer urine samples, while no expression was seen in the samples from BBD patients and normal health control subjects (see Figure 4-7), further confirming that a strong link exists between MAST4 protein and the DCIS urine samples identified with the LC-MS/MS.

211

Figure 4-6. Validation of urine protein markers using WB and IHC.

Validation results of identified potential urine proteins ECM1, MAST4 and filaggrin from breast cancer patients in human breast cancer cell lines and protein MAST 4 in primary breast cancer tissues. (A) High level of MAST4 was found in the primary breast cancer cell line (BT474) and medium levels of MAST4 were found in the metastatic breast cancer cell lines (MDA-MB231, MCF-7 and SKBR-3). High levels of ECM1 and filaggrin were found in the metastatic breast cancer cell lines (MDA-MB231, MCF-7 and SKBR-3) while low level of ECM1was seen in the primary breast cancer cell line (BT474) and no filaggrin expression was detected in the primary breast cancer cell line (BT474). GAPDH was used as

212

a loading control. (B) Illustration of the positive expression of MAST4 in breast cancer using IHC. Moderate cytoplasmic expression of MAST4 (++) was seen in the primary IBC (B1) and DCIS (B4) tumours (n=5 for each stage of breast cancer); there was no expression of MAST4 in the negative controls for either IBC (B2) or DCIS (B5). No staining was seen for MAST4 in normal breast tissues (B3 and B6), (n=5). Magnifications x 400 in B1, B2, B4 and B5; magnifications x 200 in B3; magnifications x 100 in B6. Brown indicates positive staining and blue indicates nuclei. All results were from 3 independent experiments (n=3).

MAST4 is a protein-coding gene and its roles and functions in cancers have not been reported so far. Differential expression of MAST4 was observed between the four different cell lines, with high MAST4 expression found in BT474 cell line which is an aggressive luminal B subtype, suggesting this protein could be used as a therapeutic target of interest for future studies of endocrine resistance. Though proteomics screening detected MAST4 as a significantly increased marker in DCIS in the urine samples, the results from breast cancer cell lines and human breast cancer tissues suggest that MAST4 may also be involved in breast cancer progression. This is also in line with the positive expression in the individual urine samples from IBC and MBC patients. Additionally a preliminary analysis of publically available mRNA expression profiling data (Kaplan-Meier Plotter, an online survival analysis tool (Gyorffy, Lanczky et al. 2010) http://kmplot.com/analysis/index.php?p=service&cancer=breast), demonstrated that MAST4 expression is highly significantly correlated with survival in a cohort of over 3,000 breast cancer patients. Therefore, we plan to further investigate the protein expression in a large independent cohort of breast cancer patients as part of a subsequent study.

213

Figure 4-7. Expression of MAST4 in the individual urine samples from breast cancer patients and controls.

WB was performed on the remaining urine samples from breast cancer patients and control subjects to confirm the expression of MAST4. Results show (A) overexpression of MAST4 was shown in the individual DCIS urine samples (DCIS 1-5), weak expression in IBC1 urine sample and no expression in control urine samples and (B) no MAST4 expression found in the BBD urine sample but medium expression in IBC (IBC 2-5) and MBC urine samples (MBC 1-3). ß- tubulin was chosen as a loading control.

214

Figure 4-8. Kaplan-Meier plot of recurrence free survival by MAST4 mRNA expression in breast cancer.

215

CONCLUSIONS

In this chapter the aim was to identify protein biomarkers in urine for early screening detection and monitoring of invasive breast cancer progression. We performed a comparative proteomic analysis using ion count relative quantification label free LC-MS/MS analysis of urine from breast cancer patients (n=20) and healthy control women (n=20). Workflow diagram representing a model of the urine breast cancer study (Figure 4-9), shows how with the application of MS to the protein extracts from breast cancer urine samples, a unique pattern was derived that enabled the different stages of breast cancer to be identified.

Figure 4-9. Workflow diagram of the urine analysis for novel protein markers.

216

To assess the biological significance, the LC-MS/MS-based proteomics was used to provide a profile of abundant proteins in the biological system of breast cancer patients. Amongst the 59 significant urinary proteins identified, a list of 13 novel up-regulated proteins were revealed that may be used to detect breast cancer. These include stage specific markers associated with pre-IBC in the DCIS samples (Leucine LRC36, MAST4 and Uncharacterised protein CI131), early IBC (DYH8, HBA, PEPA, uncharacterized protein C4orf14 (CD014), filaggrin and MMRN2) and MBC (AGRIN, NEGR1, FIBA and Keratin KIC10). Expression of ECM1, MAST4 and filaggrin (as potential markers) were found to be positive in breast cancer cell lines and MAST 4 was expressed in human breast cancer tissues and positive in human breast cancer urine samples.

Identification of new biomarkers in early and advanced breast cancer is important in the prevention and monitoring of disease progression and is a recently developed research area. LC-MS is a very powerful technique for the comprehensive analysis of proteins in urine. As a fluid which is relatively easy and non-invasive to collect, therefore urine is the ideal biological sample for clinical management of patients and to search for biomarkers.

Within breast cancer publications, most studies have used MS technologies to analyse urine metabolomic biomarkers. In this study, LC-MS/MS analysis was performed on urine samples from breast cancer patients, benign patients and healthy control subjects.

The results of this comparative discovery study provide a panel of novel significantly altered urinary proteins that are abundant in pre-invasive and IBC, that have not yet been previously detected in urine or other biological samples. These breast cancer proteins identified provide further insight into the complex signalling pathway interactions occurring during the progression of breast cancer. This study indicates that the majority of the abundant breast cancer urinary proteins detected are secreted proteins. Additionally, the list of novel up- regulated proteins detected (Figure 4-3), provides information which could be

217

used to create a panel of targets which could form part of urine screening “dipstick” test for the detection of non-invasive and invasive breast cancer.

Breast cancer cell lines are preclinical models that represent different breast tumour subtypes. To link the potential of urine markers identified with breast cancer disease, we validated one existing marker and two novel biomarkers in human breast cancer cell lines by WB analysis. Significantly elevated expression of three interesting markers-ECM1, MAST4 and filaggrin were demonstrated in a panel of human breast cancer cell lines and one marker MAST4 in a small group of clinical breast cancer tissue samples (Figure 4-6), indicating their promise for further investigation.

Urinary proteins can potentially provide a preliminary indication of the presence of breast cancer during screening and could assist with direct examination and pathology testing for final diagnosis. The development of a non-invasive test of breast cancer risk has been a major goal for more than 20 years. In the current study, potential breast cancer stage related protein biomarkers were identified, that could be used for early diagnosis and monitoring cancer progression in urine. These novel protein markers in urine require to be further evaluated in breast cancer tissues and an independent group of breast cancer urine samples, to test their specificity and sensitivity for breast cancer early diagnosis and lead to potential applications in cancer surveillance and prevention.

218

Chapter 5 Serum versus plasma for breast cancer biomarkers

The hypothesis is that anticoagulants and additives provide differential protein profiles that affect the total number of proteins and peptides identified. In addition, this will consequently have an effect on the proteins available in the detection of novel biomarkers and would influence the perceived pathophysiology. My objectives were to identify the effect of a variety of anticoagulants (2 plasma tubes) and additives (serum tubes) by applying a label- free LC-MS/MS approach, compare the protein abundances in the four different blood collection tubes and use this information to evaluate the impact on the number and type of proteins detected in serum versus plasma. Additionally, I wanted to determine if these blood tubes had an effect on the low mass to high mass proteins of samples in biomarker discovery.

219

5 EVALUATION OF BLOOD COLLECTION TUBES USING LC- MS/MS: SERUM VS PLASMA FOR BREAST CANCER PROTEOMIC BIOMARKER STUDIES

INTRODUCTION

Serum and plasma proteins are used as indicators of health status and treatment response. Examples include the tumour markers carcino-embryonic antigen (CEA), CA 19-9, CA 27.29, CA 125, alpha-fetoprotein (AFP), prostate-specific antigen (PSA) (Perkins, Slater et al. 2003), and the inflammatory response markers serum amyloid A (SAA) and C-reactive protein (CRP) (Zakynthinos and Pappa 2009), as described in Chapter 1.6 and 1.9.

Recent proteomic studies of serum and plasma have revealed proteome patterns that may be used to distinguish breast cancer patients from healthy individuals, as discussed in Chapter 1.8. The quantitation of proteins in serum and plasma has widespread utility for the evaluation of disease aetiology and progression. The initial discovery of breast cancer biomarkers is determined by quantitative analysis to establish the abundance of the protein and its association with the clinical stage. For proteomic studies where biomarker discovery and validation are important themes, the selection of an appropriate blood sample collection strategy is a necessary analytical consideration. Serum and plasma are frequently used samples for proteomic analysis, but the preparation of the samples varies. Blood collection tubes and preparation protocol should aim to provide a diversity of proteins, representative of the specimen and deliver a wealthy source of difference that could be explored during the biomarker discovery pipeline. Clinical proteomics can provide molecular pathway information to facilitate treatment and reveal drug resistance. Those proteins should be preserved (not modified as a consequence of collection), be stable (not destroyed during the process of extraction), and be recoverable in a form compatible with downstream analysis.

220

The most critical issue in proteomic analysis of blood is sample collection and preparation. The blood sample is collected via venepuncture directly into a blood collection tube in the presence or absence of an anticoagulant. The major differences between plasma and serum are the removal of fibrinogen and associated proteins such as HMW von Willebrand factor and the cellular secreted products extracted as a result of the coagulation process (detailed in Chapter 1.9). In proteomics, plasma appears to be the favoured sample, a protocol preventing coagulation limits endogenous proteolysis. Comparative studies by The Human Proteome Organization (HUPO), regarding optimal handling conditions for proteomic investigations (Rai, Gelfand et al. 2005), have suggested that EDTA (ethylenediaminetetra acetic acid) or citrate plasma are the preferred analytical matrix rather than serum (Omenn, States et al. 2005). Anticoagulants such as EDTA, sodium citrate, heparin, and warfarin are commonly used for plasma preparation (Drake, Bowen et al. 2004, Aebersold, Anderson et al. 2005, Anderson 2005, Ludwig and J. N. Weinstein 2005, Anderson and Hunter 2006, Drake, Cazares et al. 2007, McKay, Sherman et al. 2007). The EDTA and citrate function as anticoagulants by chelating calcium ions (Sadagopan, Li et al. 2003) while heparin “activates” antithrombin, thus inhibiting the coagulation pathway and preventing clot formation. Factors influencing plasma quality include anticoagulant selection, processing, and storage. Serum quality is influenced by collection container and clot retraction time.

In clinical medicine, the differences between serum and plasma are well known and blood tube selection is based on its application and test set. The EDTA plasma separation tubes are used for haematological testing that include full blood count and platelet counts. The serum tubes are applied for clinical chemistry testing including electrolytes and immunoglobulin studies. While the dynamic range of protein concentration in plasma or serum cannot be precisely stated, it has been previously described that the protein concentration of serum was less than that of plasma with an average value of 7.45 gm/dL for heparinised plasma, 7.21 gm/dL for serum (Lum and Gambino 1974) versus a similar difference in mean protein concentration between serum (7.29 gm/dL) and plasma (7.58 gm/dL) 221

(Ladenson, Tsai et al. 1974). The authors suggested the differences were largely due to the removal of fibrinogen and platelets during clot formation and also observed an increase of albumin in serum. During the process of whole blood coagulation, the cellular elements such as platelets can secrete a variety of components into the serum (Lindemann, Tolley et al. 2001). Comparison of serum and plasma concentrations of vascular endothelial growth factor (sVEGF and pVEGF), demonstrated a difference in both normal individuals (serum concentration of 250 pg/mL vs a plasma concentration of 30 pg/mL) and breast cancer patients (median VEGF concentration of 833 pg/mL in serum compared to 249 pg/mL in plasma)(Benoy, Salgado et al. 2002), which suggested that platelets contributed to VEGF levels in both plasma and serum but more markedly so in serum. The immediate separation of plasma or serum from cellular elements provided optimal analyte stability at RT and if prolonged contact with cells is unavoidable, the use of serum is recommended because of the higher instability of plasma analytes (Boyanton and Blick 2002).

Low-abundance proteins in the nano-grams/mL concentrations range (Zhao, Chang et al. 2016) have a lower chance of being detected as they tend to be obscured by high-abundance proteins (Ray, Reddy et al. 2011). However, the low to medium abundance proteins are promising biomarker candidates, as they include proteins that leak into the blood from tissues. It is well established that the identification of biomarkers in the blood proteome is hindered by the broad concentration range of blood proteins and their complex composition (Qian, Jacobs et al. 2006). Due to the complex nature and presence of high abundance proteins (detailed in Chapter 1.9.4), various methods are applied to serum and plasma which include affinity columns for depletion of high abundance proteins and/or fractionate the sample (Whiteaker, Zhang et al. 2007, Tu, Rudnick et al. 2010, Millioni, Tolin et al. 2011).

To perform this analysis, the blood samples were fractionated with 3kDa and 50kDa MW filters and the remaining proteins were digested with trypsin. Fractionation allowed for fragments and small peptides to be analysed, alleviating the challenge of highly abundant proteins competing with less 222

abundant proteins for ionisation and hence detection. The remaining peptides were further separated by C18 chromatography prior to analysis to concentrate and remove salts and contaminants. All <3kDa and 3-50kDa fractions were analysed by LC-MS/MS. The Progenesis statistical analysis was used and additional filters were applied (p<0.05, q, 0.02 and fold change >3). With the increase in number of proteins identified, we also showed a large variation in protein abundance in the four blood tubes. Therefore, the blood collection tube used for biomarker analysis may well determine what proteins are detected for their abundance.

MATERIALS AND METHODS

Patient population

Ethics approval for the collection and usage of blood and tissue samples was detailed in Chapter 2.1. The study cohort consisted of 20 breast cancer patients including women with ductal carcinoma in situ (DCIS, n=6), invasive breast cancer (IBC, n=8), metastatic breast cancer (MBC, n=6), benign breast disease (BBD, n=6), and 20 healthy controls (CTL). The blood specimens were collected from each patient and volunteer in four different commonly used blood tubes (2 serum/2 plasma). As part of an ongoing clinical laboratory quality program, all samples were collected by a trained phlebotomist. All blood samples were handled according to the manufacturer’s instructions and inverted 5 times immediately after collection.

Blood collection

In this study, LC-MS/MS analysis was used to detect serum and plasma proteins in breast cancer patients and healthy individuals. This enabled a list of proteins to be created and allowed a comparison of the impact of the four different blood collection tubes on the protein levels across all the samples. The method and tubes applied are detailed in Chapter 2.3.2. 223

A total 184 venous blood samples were acquired from female breast cancer patients (n=20) and BBD (n=6) prior to surgery and cancer-free female volunteers (n=20). Blood samples were collected in four different BD Vacutainer blood collection tubes (2 serum/2 plasma) for each patient. The breast cancer group consisted of 40 serum and 40 plasma samples from different breast cancer stages. The same procedure was applied for blood collection from healthy controls (CTL). All samples were collected by a phlebotomist prior to surgery, processed for serum and plasma isolation.

Briefly, serum samples were collected in serum gold (SG) top (BD Vacutainer® SST™ Gold top Tubes. Ref Cat No 367986: with clot activator and gel) and serum red top (SR) Clot activator, Silicone coated tubes (BD Vacutainer® Serum Red Top. Ref Cat No 367812: coated with silicone and micronised silica particles). Plasma samples were collected in EDTA (PP) tubes (BD Vacutainer® K2E. Ref Cat No 367839: Lavender Tubes spray-coated with K2EDTA) and lithium heparin (PG) tubes (BD Plastic Vacutainer® PST II -Lithium heparin. Ref Cat No 367375: Light green gel tube). Samples were collected and processed following a standardised protocol, inverted on collection to ensure mixing of clot activator with blood. The plasma samples were processed for plasma within 30 min and serum samples were allow to clot at RT for 30 min. Serum and plasma were collected by spinning at 1300 x g for 10 min and stored at -80 °C, until needed. The protein concentrations of the samples were determined by the BCA method (2D Quant kit). Summary of blood collection and handling is shown in Figure 2-2.

Fractionation

The proteins in the serum and plasma were partitioned into three fractions based on protein MW. Method applied is detailed in Chapter 2.4.6.

Briefly, to examine the low molecular mass fraction in blood and separate highly abundant proteins, the blood samples were separated into fractions of <3kDa, 3- 50kDa and >50kDa MW for analysis. Equal volumes of blood samples from the breast cancer stages, BBD and healthy subjects were pooled respectively into 224

each cohort, based on tube type. Briefly, 100 µg pooled protein sample was made up to 100 µL with DTT/Urea solution (10mM DTT + 2M Urea) and concentrated using Microcon® ultra Centrifugal filters, 3,000kDa MW and 50,000kDa MW cut off filters (YM3 & YM50). All samples were centrifuged and concentrated at 9,000 x g for 30 min at 4 °C. The samples were vacuum centrifuge dried and frozen at - 20 °C. Protein concentration was determined by the 2D-Quant method (GE healthcare-Life sciences).

Blood fraction C18 clean-up and Trypsin digestion

Fractions were desalted using C18 Stage Tips and was digested with trypsin. The method applied is detailed in Chapter 2.4.5.

Briefly, samples were further cleaned-up and purified for LC-MS/MS analysis with C18 STAGE tip (Thermo Scientific, USA), following the manufacturer’s instructions. In-solution digestion was carried out with Promega sequencing grade modified trypsin (12.5 ng/µL sequencing grade trypsin, Sigma-Aldrich, St Louis, MO, USA) in an enzyme: protein ratio of 1:20 (w/w), overnight at 37 °C. The digested samples were completely dried in a Centrifugal Vacuum Concentrator (Savant SpeedVac® Plus SC210A) at a low speed for 30 min and stored at -20C.

LC-MS/MS analysis

LC-MS/MS analysis was performed to compare the differences in protein abundances in serum and plasma between breast cancer and healthy controls. All the procedures in this LC/MS-MS study are detailed in Chapter 2.4.7. Briefly, protein identification was performed on all protein digests, using label-free LC- MS/MS quantification (LTQ-Orbitrap, Thermo Scientific, USA). The mass spectrometer was linked to a micro auto sampler and a Nano flow HPLC pump, as per previously published method (Beretov, Wasinger et al. 2014).

225

Data processing and analysis

Peak lists were generated using ‘Mascot Daemon/extract_msn’ (Matrix Science), default parameters and then submitted to the protein database search program Mascot (Matrix Science, Boston, USA). Database searching and validation are detailed in Chapter 2.4.9.

Briefly, the raw data files were converted into *. mfg files and then searched against UniProt human data base using Mascot v3.2.17 local server. Precursor and product ion spectra were searched with a mass tolerance of ± 4ppm and ± 0.04 Da, respectively. Variable modifications Deamination (NQ), Oxidation (M), Phospho (ST) and Phospho(Y) were selected. All data were searched against a target and decoy database. The target decoy strategy was applied to control both peptide and protein level false discovery rate (FDR) at lower than 1%.

MS peak intensities were analyzed using Progenesis QI, LC-MS data analysis software (version 4.1, Nonlinear Dynamics, Newcastle upon Tyne, UK) as detailed in Chapter 2.4.8. Data was tabulated using Microsoft Excel (Microsoft Corporation, Redmond, WA, USA) and analysed using SPSS for Microsoft Windows, version 13.0 (SPSS, Chicago, IL, USA). Statistical comparisons and unpaired t-test were performed to compare all numerical data significance using Excel and Prism 6 package (GraphPad, CA, USA). Multivariate analysis by binary logistical regression and multi-linear regression were performed.

RESULTS AND DISCUSSION

Prevalence of varied protein expression

Four commonly used blood collection tubes were examined by LC-MS/MS for all blood samples collected from breast cancer and BBD patients and healthy control volunteers. The LC-MS/MS data provided detailed lists of the proteins detected in the four blood collection tubes. Hundreds of differentially abundant proteins

226

were identified in each breast cancer stage, BBD and control samples. In the 0- 3kDa fraction, the number of proteins identified ranged from 56 in DCIS to 114 in MBC. A larger number of proteins were detected in the 3-50kDa fraction, which ranged from 86 proteins identified in BBD to 140 in MBC. Overall, a greater number of proteins were identified in the 3-50kDa fraction collected from serum Gold-top SSTII tubes and the two plasma separator tubes (purple EDTA and green lithium heparin). In contrast, in the serum red top findings within both fractions, 20% less proteins were detected.

A scatter chart (Figure 5-1) was used to display and compare the correlation between the protein abundance values for all the proteins detected. The chart shows the distribution of the protein values for each blood tube in the control (A) and breast cancer samples (B) and highlights the large deviation from the unity of tube related abundances, especially compared to SG abundances.

To analyse the wide range of data and determine any natural variation in protein abundances in serum and plasma, strategies were designed to compare the impact of the tube and the relative difference in protein expression between serum SG: SR, plasma PP: PG and serum: plasma. The normalised abundances for individual proteins were compared across the four blood tubes and recorded as a ratio of serum gold tube SST (SG): Serum red top (SR): Plasma EDTA purple (PP): Plasma Heparin green (PG). This allowed for serum to be compared with serum, plasma with plasma and serum: plasma. This ratio was formulated and the degree of difference was observed as a fold change between collection tube types, whereby >5 fold was noted as a significant variation in abundance. Comparison following these criteria was conducted for all the blood samples: DCIS, IBC, MBC, BBD and control samples in both the 0-3kDa and 3-50kDa fractions.

Fractionation of the samples made it possible to detect the low abundance proteins. The > 50kDa fraction was technically a difficult sample to analyse with both SDS-PAGE and LC-MS/MS and was therefore not used.

227

Figure 5-1. Scatter plot of distribution of proteins in serum and plasma

Scatter plot of the protein abundances in four blood tubes for: (A) healthy control and (B) breast cancer blood samples.

228

Percentage variation between tubes

The initial analysis and fold change ratio calculated, relating the normalised proteins abundances value for the 0-3kDa and 3-50kDa fractions, was observed across all the proteins detected in all four blood tubes and the variation was calculated as a percentage (%). The p-value was <0.05.

Within the 0-3kDa fraction, the percentage variations for the proteins were detected in all the blood tubes for breast cancer, BBD and control samples. On average, the percentage was greater than 70% as shown in Table 5-1. A graphical representation for the comparison of percentage variation in the 0-3kDa fraction was used to highlight the significant difference across the four blood tubes for all sample types shown in Figure 5-2.

Table 5-1. Percentage variation of proteins in BC, BBD patients and control subjects for 0-3kDa serum and plasma fraction.

Sign Total Blood proteins proteins Percentage (%) variation across blood tubes sample detected compared SG:PP SG: PG SG:SR PP: PG PP: SR PG:SR EDTA Li Hep DCIS 56 33 85 64 64 73 61 52 IBC 85 49 59 88 73 78 78 59 MBC 114 62 79 85 74 66 68 82 BBD 90 71 73 63 81 76 90 58 CTL 80 46 87 93 76 65 59 76

Average 77 79 74 72 71 65

Abbreviations: BBD, benign breast disease; CTL, control sample; DCIS, ductal carcinoma in-situ; IBC, invasive breast cancer; MBC, metastatic breast cancer; PP, purple plasma EDTA tube; PG, green plasma lithium heparin tube; SG, serum gold top tube; SR, serum red top tube.

229

Figure 5-2. Percentage variation of protein abundances in 0-3 kDa serum and plasma fraction.

A comparison of percentage variation in breast cancer, BBD patients and control subjects’ blood samples in the 0-3kDa fraction, across the four different blood tubes.

230

The 3-50kDa fraction was a very rich source of protein data. The percentage variation in protein abundance across the four blood tubes was also high at 69- 78%, shown in Table 5-2. An interesting finding in the 3-50kDa fraction was that the percentage variation was lower in DCIS (55%) and IBC (32%) between serum gold: plasma purple EDTA. This difference indicated that the large number of the proteins detected in these two breast cancer samples (DCIS and IBC) had similar abundances in the serum and plasma samples, depending on the individual proteins compared. This information is important for biomarker detection when mining a patient’s blood sample for novel proteins, as it is essential to use a collection tube that will provide the most information as to what is happening in the patient’s sample/body. Overall, the average difference across all the patients’ samples was still high at 64%. The data was graphically presented to highlight the differences in protein abundances in the 3-50kDa fraction (in the four blood tubes), shown in Figure 5-3.

This evidence indicates that if the same group of proteins are compared, a large number of the protein abundances detected in the four tubes are significantly different. This variation was seen even when comparing serum (SG) to serum (SR) and plasma (PP EDTA) to plasma (PG Heparin). Any reduction in protein detection, along with reduced abundance warrants concern as this would indicate that certain proteins may not be identified, therefore resulting in loss of potential biomarkers of breast cancer. Therefore, the blood tube selection for patient sampling is important for both detection and future application as a population biomarker, if assessing a low abundance protein.

231

Table 5-2. Percentage variation of proteins in breast cancer, BBD patients and control subjectsfor 3-50kDa serum and plasma fraction.

Blood Sign Total sample proteins proteins Percentage (%) variation across blood tubes type detected compared SG:PP SG: PG SG:SR PP: PG PP: SR PG:SR EDTA Li Hep DCIS 122 80 55 85 70 75 88 75 IBC 106 44 32 77 77 80 84 84 MBC 140 70 82 80 77 46 76 77 BBD 84 43 74 91 67 72 79 74 CTL 140 95 78 59 90 73 43 79

Average 64 78 76 69 74 78

232

Figure 5-3. Percentage variation of protein abundances in 3-50kDa serum and plasma fraction.

A comparison of percentage variation in breast cancer, BBD patients’ and control subjects’ blood samples in the 3-50kDa fraction, across the four different blood tubes.

Abbreviations: BBD; benign breast disease; CTL; control sample; DCIS, ductal carcinoma in-situ; IBC; invasive breast cancer; MBC; metastatic breast cancer; PP, purple plasma-EDTA tube; PG, green plasma-lithium heparin tube; SG, serum gold top tube; SR; serum red top tube.

233

Direct protein comparisons in 3-50kDa fraction

The 3-50kDa data represented the large variation in protein numbers and abundance detected across the different tubes. Therefore, the natural variation of circulating levels in a panel of 15 proteins (some previously discussed in Chapter 1-9 and 1-10, blood biomarkers) was assessed in a cohort of breast cancer, BBD and healthy individuals. Using the protein information in the 3- 50kDa fractions shown in Table 5-3, we examined the influence of the four different blood collection tubes on serum and plasma protein levels, by comparing abundance ratio and normalised fold change for each tube type. Anova (p)* value indicates that these proteins are significantly different across the tubes.

In the 0-3kDa and the 3-50kDa fraction the abundances across the tubes demonstrated several factors: protein levels varied across the different blood tubes and that specific proteins were only detected in certain tubes. Additionally, abundant plasma proteins previously reported as having concentrations in the order of > 10 in magnitude (detailed in Chapter 1.9.4) such as: albumin, Igs, alpha-1-antitrypsin, fibrinogen, α1-antitrypsin, α2-macroglobulin, ceruloplasmin, complement C1q, C3, C4, plasminogen, haptoglobin and blood coagulation factor VII, were surveyed. Our findings showed that several of these proteins were consistently identified in all the samples, although their abundances changed depending on the collection tube used. More importantly, these proteins did not dominate the total proteins detected in the samples. Furthermore, serum albumin was present in equal amounts in most of the tubes and fibrinogen was absent or low in the serum SST (SG) samples as expected.

The fifteen proteins selected for closer comparison included: Alpha-2-HS- glycoprotein, Clusterin, Cystatin-C, Extracellular matrix protein 1, Insulin-like growth factor-binding protein 3, 5- 6, Leucine-rich alpha-2-glycoprotein, Osteopontin, Profilin-1, Selenoprotein P, Serglycin, Transthyretin, Uncharacterized protein C1orf56, Vitamin D-binding protein, Vitronectin and Zyxin. In addition, complement proteins and Keratins (type 1 and II) abundance patterns were also 234

examined. The results revealed that 10 out of 15 of proteins were differentially expressed between the four different blood tubes. A detailed comparison of the proteins is outlined as follows:

Alpha-2-glycoprotein was detected in the control and DCIS samples. Protein abundance in serum SST: plasma EDTA in DCIS samples only varied 2 fold. The rest of the comparison showed 5-fold variation between SG: SR and PP: PG was approximately 5-14 fold and in serum alone 3-6 fold.

Clusterin was detected in all the samples. In the CTL blood sample, serum: plasma abundance of Clusterin varied 1-4 fold. In the breast cancer samples (DCIS and IBC), there was a small difference between protein expressions in serum SG: SR and serum SG: PP EDTA tubes at approximately 2-fold change. In the MBC and BBD samples there were marked differences in abundance between tubes, ranging from 3-39 fold (SG: SR difference at 15-21 fold, plasma PP: PG at 8 fold and serum: plasma variation at 3-39 fold).

Cystatin-C in the CTL, breast cancer and BBD samples SG: PP EDTA was 1-6 fold. Therefore, this protein was detected in both serum SST and plasma EDTA tubes at similar abundances. In CTL and BBD serum SG: SR, protein expression was 2 fold, and average serum: plasma was 3 fold. The variation was slightly higher in the breast cancer samples for SG: PP EDTA which were similar at 1-7 fold, though serum: serum and plasma: plasma ranging from 5- 142 fold (SG: SR at 48-142 and plasma tubes PP: PG at 5-35 fold change variation).

Extracellular matrix protein 1(ECM1) was detected in the CTL and MBC samples. In the CTL’s, there were similarities between three tubes, plasma PP: PG: SR at 1-2 fold variation and the remaining comparisons range 4-16 fold (serum SG: SR and SG: plasma) difference in protein abundance. In MBC samples, abundances overall varied 2-16 fold (serum SST: Plasma 2-8; serum: serum 6; plasma to plasma 16; serum red: plasma 1-12).

235

Table 5-3. Differential expression of proteins in serum and plasma in 3-50 kDa fraction.

Accession Peptides Score Anova Fold Protein Description Blood tube comparison ratio (p)* Ctl_3-50kDa SG:PP SG:PG SG:SR PP:PG PP:SR PG:SR FETUA_HUMAN 5 333.5 5.1E-05 5.6 Alpha-2-glycoprotein 6 1 5 5 1 4 CERU_HUMAN 16 (11) 810.7 6.6E-05 7.0 Ceruloplasmin 7 2 7 3 1 4 CLUS_HUMAN 10 (8) 565.4 1.5E-04 4.2 Clusterin 3 1 4 3 1 4 CO3_HUMAN 34 (31) 2009.7 2.0E-03 3.6 Complement C3 3 3 4 1 1 1 CO4B_HUMAN 33 (28) 1806.0 6.9E-05 14.1 Complement C4-B 6 14 10 2 2 1 CO5_HUMAN 5 200.2 1.1E-04 14.5 Complement C5 7 6 15 1 2 2 CO9_HUMAN 7 384.2 9.6E-04 39.6 Complement component C9 40 4 20 9 2 4 CFAB_HUMAN 6 (4) 319.7 6.4E-04 25.8 Complement factor B 4 5 5 19 1 26 CFAD_HUMAN 2 66.3 1.0E-02 25.3 Complement factor D 7 2 16 12 2 25 CYTC_HUMAN 3 159.3 3.7E-03 7.3 Cystatin-C 1 3 2 5 1 7 ECM1_HUMAN 2 107.8 1.6E-04 9.3 Extracellular matrix protein 1 7 9 4 1 2 2 GELS_HUMAN 5 249.9 2.0E-03 16.3 Gelsolin 11 15 16 1 1 1 IBP3_HUMAN 1 41.9 4.5E-04 46.9 Insulin-like growth factor- 5 3 14 18 3 47 binding protein 3 K1C10_HUMAN 22 (13) 1343.0 3.5E-06 33.5 Keratin, type I cytoskeletal 10 6 6 5 34 1 30 K1C14_HUMAN 21 (5) 1128.1 2.4E-03 40.3 Keratin, type I cytoskeletal 14 40 2 33 18 1 15 K1C16_HUMAN 31 (15) 1635.7 6.1E-05 110.5 Keratin, type I cytoskeletal 16 79 11 110 7 1 10 K1C17_HUMAN 18 (4) 875.8 4.4E-04 262.6 Keratin, type I cytoskeletal 17 8 1 31 10 263 26 K1C9_HUMAN 19 (15) 1118.2 1.7E-05 84.4 Keratin, type I cytoskeletal 9 21 4 6 84 4 23 K2C1_HUMAN 27 (20) 1739.6 8.2E-05 50.2 Keratin, type II cytoskeletal 1 23 2 7 50 3 16 K22E_HUMAN 16 (10) 1137.0 5.3E-05 62.2 Keratin, type II cytoskeletal 2 2 34 1 62 3 24 epidermal K2C5_HUMAN 14 (6) 676.4 1.5E-03 26.5 Keratin, type II cytoskeletal 5 10 1 26 8 3 20 236

K2C6A_HUMAN 25 (2) 1506.8 1.5E-03 291.5 Keratin, type II cytoskeletal 6A 291 25 128 12 2 5 K2C6B_HUMAN 24 (4) 1393.8 4.8E-03 35.9 Keratin, type II cytoskeletal 6B 36 6 23 6 2 4 KNG1_HUMAN 24 (23) 1334.9 4.6E-03 7.7 Kininogen-1 4 1 8 4 2 7 A2GL_HUMAN 5 270.8 1.3E-03 27.7 Leucine-rich-alpha-2-glyco 4 4 6 17 2 28 protein OSTP_HUMAN 2 97.2 1.1E-03 21.0 Osteopontin 12 21 10 2 1 2 SEPP1_HUMAN 4 209.6 2.5E-04 24.5 Selenoprotein P 4 3 24 1 7 9 SRGN_HUMAN 7 344.1 8.7E-05 15.9 Serglycin 3 1 5 3 16 5 ALBU_HUMAN 32 (21) 1802.3 2.1E-03 3.1 Serum albumin 1 1 3 1 2 3 TSP1_HUMAN 2 (1) 73.4 8.5E-03 2716.7 Thrombospondin-1 25 13 213 313 9 2708 TTHY_HUMAN 80 (77) 5899.4 1.8E-03 15.2 Transthyretin 5 2 15 2 3 7 CA056_HUMAN 1 120.9 1.3E-03 13.3 Uncharacterized protein 4 13 7 3 2 2 C1orf56 VTDB_HUMAN 7 (5) 279.2 2.2E-03 4.3 Vitamin D-binding protein 1 1 3 2 2 4 VTNC_HUMAN 4 212.7 9.2E-04 29.0 Vitronectin 29 20 18 1 2 1 ZYX_HUMAN 9 (8) 311.8 4.0E-05 103.9 Zyxin 2 104 5 58 3 22

DCIS_3-50kDa SG:PP SG:PG SG:SR PP:PG PP:SR PG:SR A1BG_HUMAN 11 (9) 556.7 9.6E-05 11 Alpha-1B-glycoprotein 2 5 4 11 8 1 FETUA_HUMAN 5 439.7 5.1E-04 24 Alpha-2-HS-glycoprotein 2 25 5 14 3 5 APOA4_HUMAN 19 928.3 4.6E-03 6 Apolipoprotein A-IV 1 4 4 6 6 1 APOE_HUMAN 8 360.5 2.5E-03 7 Apolipoprotein E 1 1 5 2 7 4 APOF_HUMAN 2 91.1 6.4E-04 22 Apolipoprotein F 11 7 2 1 22 14 APOL1_HUMAN 3 186.7 9.1E-03 4 Apolipoprotein L1 4 1 1 3 3 1 CAH2_HUMAN 2 93.1 3.2E-03 40 Carbonic anhydrase 2 2 15 19 32 40 1 CD99_HUMAN 6 266.0 3.4E-03 3 CD99 antigen 2 2 2 1 3 3 CERU_HUMAN 27 (21) 1600.0 3.1E-03 10 Ceruloplasmin 5 1 2 6 10 2

237

CLUS_HUMAN 7 352.5 4.4E-03 7 Clusterin 2 7 2 4 1 4 CFAB_HUMAN 13 (11) 655.8 8.1E-03 13 Complement factor B 2 6 5 13 10 1 CFAD_HUMAN 3 95.1 3.8E-03 389 Complement factor D 6 4 388 1 62 89 DEMA_HUMAN 1 87.4 1.0E-02 23 Dematin 12 9 2 1 23 17 DSG3_HUMAN 4 122.5 1.2E-05 193 Desmoglein-3 18 7 11 3 193 72 FIBG_HUMAN 12 663.9 4.4E-03 13 Fibrinogen gamma chain 13 3 3 4 5 1 HPT_HUMAN 20 (5) 1029.1 2.0E-03 4 Haptoglobin 2 3 1 2 2 4 IGHA1_HUMAN 7 (2) 495.6 5.6E-03 7 Ig alpha-1 chain C region 3 1 2 4 7 2 IGHG1_HUMAN 17 (2) 1111.8 5.7E-04 14 Ig gamma-1 chain C region 14 4 4 4 4 1 HV103_HUMAN 3 (2) 109.6 9.1E-03 10 Ig heavy chain V-I region V35 10 2 1 5 8 2 IGKC_HUMAN 3 345.4 2.0E-04 3 Ig kappa chain C region 3 1 1 3 3 1 KV301_HUMAN 2 (1) 196.6 2.5E-05 151 Ig kappa chain V-III region B6 22 151 15 7 1 10 KV308_HUMAN 2 (1) 114.1 7.2E-03 13 Ig kappa chain V-III region CLL 5 3 2 13 3 4 LAC2_HUMAN 3 (2) 245.8 9.4E-03 13 Ig lambda-2 chain C regions 4 2 3 7 13 2 IGHM_HUMAN 4 192.7 1.0E-03 28 Ig mu chain C region 11 28 14 3 1 2 IBP5_HUMAN 4 149.8 1.0E-02 59 Insulin-like growth factor- 1 3 50 3 59 18 binding protein 5 K1C10_HUMAN 27 (13) 1658.9 2.0E-04 26 Keratin, type I cytoskeletal 10 1 18 23 21 26 1 K1C14_HUMAN 15 (7) 803.5 4.9E-03 102 Keratin, type I cytoskeletal 14 1 102 93 72 66 1 K1C15_HUMAN 8 (2) 289.0 5.4E-04 35 Keratin, type I cytoskeletal 15 2 35 6 17 3 6 K1C9_HUMAN 24 1834.9 1.2E-03 69 Keratin, type I cytoskeletal 9 1 23 69 16 47 3 K22E_HUMAN 25 (17) 2103.1 1.9E-03 27 Keratin, type II cytoskeletal 2 1 24 27 20 23 1 epidermal K2C5_HUMAN 11 (1) 563.9 6.3E-04 95 Keratin, type II cytoskeletal 5 1 95 57 74 44 2 K2C6A_HUMAN 8 (1) 428.9 4.6E-03 169 Keratin, type II cytoskeletal 6A 4 169 14 41 3 12 KNG1_HUMAN 12 539.7 1.0E-02 21 Kininogen-1 2 21 16 13 10 1

238

A2GL_HUMAN 10 (9) 503.1 3.6E-04 11 Leucine-rich alpha-2- 4 11 2 3 2 6 glycoprotein MUC16_HUMAN 13 (7) 349.1 6.7E-05 105 Mucin-16 28 18 4 2 105 66 OSTP_HUMAN 2 136.0 2.6E-03 165 Osteopontin 3 152 1 45 4 165 PROF1_HUMAN 5 306.9 4.9E-03 292 Profilin-1 5 7 292 1 54 40 S10A6_HUMAN 1 54.1 1.0E-02 4600 Protein S100-A6 8 RET4_HUMAN 4 199.7 1.6E-03 144 Retinol-binding protein 4 2 9 84 15 144 9 SEPP1_HUMAN 3 (2) 107.1 5.8E-03 6358 Selenoprotein P 7 >100 1 >100 9 >100 ALBU_HUMAN 52 (27) 2993.0 1.2E-03 8 Serum albumin 3 2 2 2 8 4 SAA4_HUMAN 3 112.9 2.0E-04 32 Serum amyloid A-4 protein 4 8 5 32 21 2 TTHY_HUMAN 11 816.8 1.7E-03 10 Transthyretin 3 2 3 2 10 5 VTDB_HUMAN 15 721.6 3.9E-05 19 Vitamin D-binding protein 1 3 18 3 19 6 ZA2G_HUMAN 5 253.9 1.1E-03 24 Zinc-alpha-2-glycoprotein 1 3 18 4 24 5 ZYX_HUMAN 4 170.1 1.8E-03 4 Zyxin 4 1 1 4 4 1

Invasive_3-50kDa SG:PP SG:PG SG:SR PP:PG PP:SR PG:SR A1AG1_HUMAN 0 482.2 8.1E-03 52.7 Alpha-1-acid glycoprotein 1 1 53 10 37 7 5 A1AG2_HUMAN 8 (4) 418.2 9.5E-03 26.5 Alpha-1-acid glycoprotein 2 2 27 3 11 1 8 A1BG_HUMAN 8 408.1 2.0E-02 30.4 Alpha-1B-glycoprotein 2 5 7 3 12 30 APOA4_HUMAN 75 (69) 4026.9 5.2E-04 193.6 Apolipoprotein A-IV 2 65 194 42 124 3 APOB_HUMAN 47 (43) 2575.1 5.5E-04 6.6 Apolipoprotein B-100 1 6 1 7 1 7 APOE_HUMAN 20 (19) 1266.9 5.0E-02 16.8 Apolipoprotein E 1 2 12 3 17 6 CERU_HUMAN 29 (28) 1754.2 8.0E-03 10.3 Ceruloplasmin 1 3 3 3 3 10 CLUS_HUMAN 11 (10) 500.3 2.8E-03 27.1 Clusterin 1 27 17 25 16 2 CO3_HUMAN 94 (84) 5680.2 3.0E-02 7.2 Complement C3 1 3 5 4 7 2 CO4B_HUMAN 47 (45) 2720.8 2.0E-02 5.6 Complement C4-B 2 6 4 3 3 1

239

CO9_HUMAN 10 (7) 522.6 1.4E-03 74.3 Complement component C9 14 2 5 7 74 11 CFAD_HUMAN 2 85.2 4.0E-02 35.8 Complement factor D 1 10 36 9 30 3 CYTC_HUMAN 5 297.2 4.0E-02 79.9 Cystatin-C 2 80 48 35 21 2 IGHG1_HUMAN 18 (4) 958.1 2.9E-04 27.7 Ig gamma-1 chain C region 3 9 2 28 5 5 IGHG2_HUMAN 17 (9) 1090.8 6.8E-03 80.6 Ig gamma-2 chain C region 2 38 4 81 8 10 K1C10_HUMAN 20 (7) 1160.1 1.2E-03 11.2 Keratin, type I cytoskeletal 10 2 2 6 1 9 11 K1C14_HUMAN 7 (3) 304.4 3.0E-02 83.5 Keratin, type I cytoskeletal 14 1 67 19 84 24 3 K1C9_HUMAN 12 (11) 707.7 6.4E-03 15.1 Keratin, type I cytoskeletal 9 1 2 7 2 7 15 K2C1_HUMAN 22 (11) 1372.4 4.0E-02 11.5 Keratin, type II cytoskeletal 1 3 3 12 1 4 3 K22E_HUMAN 12 (7) 925.2 4.6E-03 23.4 Keratin, type II cytoskeletal 2 1 2 16 3 23 7 epidermal KNG1_HUMAN 21 1200.1 3.9E-03 76.9 Kininogen-1 1 39 56 53 77 1 PROF1_HUMAN 4 312.2 1.0E-02 61.0 Profilin-1 4 61 7 16 2 8 RET4_HUMAN 4 159.2 1.7E-03 150.2 Retinol-binding protein 4 2 2 70 5 150 29 SRGN_HUMAN 12 534.3 2.0E-02 79.5 Serglycin 6 79 2 14 2 32 ALBU_HUMAN 74 (37) 4448.4 2.0E-02 50.5 Serum albumin 1 40 5 51 6 9 TTHY_HUMAN 59 (57) 4647.0 1.3E-03 46.6 Transthyretin 2 10 22 21 46 2 VTNC_HUMAN 4 257.9 6.0E-03 14.2 Vitronectin 4 14 1 4 3 10 ZA2G_HUMAN 3 147.6 3.0E-02 12.8 Zinc-alpha-2-glycoprotein 1 3 4 3 4 13

MBC 3-50kDa SG:PP SG:PG SG:SR PP:PG PP:SR PG:SR A1BG_HUMAN 17 (16) 999. 0.01 4.9 Alpha-1B-glycoprotein 3 3 2 1 4 5 FETUA_HUMAN 15 (13) 710.8 9.42E-04 52.7 Alpha-2-HS-glycoprotein 4 3 20 10 5 53 APOA4_HUMAN 114(100) 5753.9 1.83E-03 43.7 Apolipoprotein A-IV 28 13 44 2 2 3 APOC3_HUMAN 6 646.8 3.73E-03 3.3 Apolipoprotein C-III 2 2 3 1 2 2 APOE_HUMAN 36 (25) 2200.5 7.93E-03 26.9 Apolipoprotein E 27 17 19 2 1 1

240

APOF_HUMAN 3 191.3 1.55E-03 74 Apolipoprotein F 74 3 4 24 18 1 APOL1_HUMAN 6 (5) 398.6 1.95E-03 49.8 Apolipoprotein L1 50 6 30 8 2 5 CERU_HUMAN 33 (31) 1529.3 3.66E-07 6.5 Ceruloplasmin 5 6 7 1 1 1 CLUS_HUMAN 13 678.9 5.26E-03 38.7 Clusterin 39 5 15 8 3 3 CO9_HUMAN 8 (6) 401.6 2.90E-03 112.5 Complement component C9 7 2 113 3 17 60 CFAB_HUMAN 19 (18) 805.9 8.57E-03 44.6 Complement factor B 3 6 7 2 23 45 CFAI_HUMAN 4 (3) 200.9 5.16E-03 14.0 Complement factor I 2 4 3 9 2 14 CYTC_HUMAN 5 324.6 2.46E-07 386.8 Cystatin-C 7 3 142 18 22 386 ECM1_HUMAN 4 235.0 4.95E-03 15.8 Extracellular matrix protein 1 8 2 6 16 1 12 GELS_HUMAN 14 708.2 3.81E-03 12.3 Gelsolin 2 1 7 2 12 7 IGHG1_HUMAN 18 (4) 1025.6 1.10E-04 118.7 Ig gamma-1 chain C region 119 95 28 1 4 3 IGHG2_HUMAN 12 (5) 792.8 4.83E-04 40.3 Ig gamma-2 chain C region 40 28 13 1 3 2 IGHG4_HUMAN 15 (2) 944.8 5.70E-03 16.2 Ig gamma-4 chain C region 16 12 2 1 7 5 HV103_HUMAN 4 (3) 150.7 9.46E-03 3.4 Ig heavy chain V-I region V35 3 2 1 1 3 3 K1C10_HUMAN 32 (23) 1624.9 4.98E-03 23.0 Keratin, type I cytoskeletal 10 23 6 8 4 3 1 K1C9_HUMAN 25 (22) 1480.7 5.65E-05 7.5 Keratin, type I cytoskeletal 9 8 1 4 6 2 4 K2C1_HUMAN 41 (14) 2304.8 6.61E-03 9.9 Keratin, type II cytoskeletal 1 2 6 10 4 7 2 K22E_HUMAN 26 (14) 1638.3 3.33E-03 30.1 Keratin, type II cytoskeletal 2 30 8 8 4 4 1 epidermal K2C5_HUMAN 15 (2) 672.4 1.80E-04 37.4 Keratin, type II cytoskeletal 5 37 5 8 8 5 2 KNG1_HUMAN 32 1613.6 6.77E-05 24.6 Kininogen-1 4 2 25 2 7 15 A2GL_HUMAN 11 (9) 508.7 9.56E-04 32.2 Leucine-rich alpha-2- 9 8 3 1 32 27 glycoprotein AMBP_HUMAN 5 (4) 248.7 8.93E-03 5.7 Protein AMBP 6 5 2 1 3 3 RET4_HUMAN 7 321.2 1.31E-03 611.7 Retinol-binding protein 4 1 6 106 6 110 612 SEPP1_HUMAN 5 296.3 1.11E-03 62.6 Selenoprotein P 63 9 15 7 4 2 SRGN_HUMAN 11 522.3 7.22E-04 33 Serglycin 17 8 2 2 33 16

241

ALBU_HUMAN 70 (47) 3418.3 3.22E-04 8.3 Serum albumin 7 8 1 1 6 8 TETN_HUMAN 4 250.4 9.04E-04 29.9 Tetranectin 18 10 30 2 2 3 TSP1_HUMAN 7 (6) 314.7 3.05E-04 357.8 Thrombospondin-1 11 21 358 2 32 17 TTHY_HUMAN 87 (77) 6352.2 2.99E-03 40.8 Transthyretin 17 34 41 2 2 1 CA056_HUMAN 3 (2) 151.8 8.35E-03 73.7 Uncharacterized protein 18 7 74 2 4 10 C1orf56 VTDB_HUMAN 17 784.6 2.49E-04 9.0 Vitamin D-binding protein 5 8 1 2 4 6 ZYX_HUMAN 6 280.6 2.49E-04 47.4 Zyxin 5 12 47 3 10 4 ZA2G_HUMAN 7 336 0.01 24.4 Zinc-alpha-2-glycoprotein 24 13 3 2 9 5

Benign_3-50kDa SG:PP SG:PG SG:SR PP:PG PP:SR PG:SR A1BG_HUMAN 7 399.4 8.1E-03 158.1 Alpha-1B-glycoprotein 1 63 153 66 159 2 APOA4_HUMAN 10 (9) 510.8 3.0E-02 15.4 Apolipoprotein A-IV 4 2 7 8 2 15 CD99_HUMAN 5 206.6 2.0E-02 6.3 CD99 antigen 2 4 1 2 3 6 CLUS_HUMAN 5 207.5 1.3E-03 21.1 Clusterin 8 21 21 3 3 1 CO4B_HUMAN 27 (25) 1522.5 5.8E-03 4.3 Complement C4-B 3 1 2 3 4 1 CFAD_HUMAN 3 96.0 8.0E-03 11.5 Complement factor D 6 11 7 2 1 2 CYTC_HUMAN 3 219.9 2.0E-02 37.8 Cystatin-C 4 38 28 10 7 1 DCD_HUMAN 3 146.0 2.0E-02 6.0 Dermcidin 4 6 2 2 2 3 FIBA_HUMAN 134 (122) 9200.3 9.5E-03 138.4 Fibrinogen alpha chain 25 17 6 1 138 93 FIBB_HUMAN 42 (38) 1867.9 2.8E-03 95.3 Fibrinogen beta chain 20 4 5 5 95 21 HEMO_HUMAN 20 (19) 806.4 3.0E-02 28.6 Hemopexin 2 7 19 11 29 3 IGHG1_HUMAN 11 (3) 711.0 5.4E-03 10.7 Ig gamma-1 chain C region 8 11 2 1 4 5 LAC2_HUMAN 3 258.9 7.4E-03 4.5 Ig lambda-2 chain C regions 4 3 1 1 3 2 IBP5_HUMAN 6 242.3 2.0E-02 64.2 Insulin-like growth factor- 2 12 32 25 64 3 binding protein 5 K1C10_HUMAN 25 (20) 1454.4 4.2E-03 47.2 Keratin, type I cytoskeletal 10 9 47 14 5 1 3

242

K1C16_HUMAN 14 (3) 665.0 2.0E-02 516.7 Keratin, type I cytoskeletal 16 26 6 20 146 517 4 K1C9_HUMAN 29 (27) 1762.7 2.5E-03 35.2 Keratin, type I cytoskeletal 9 20 35 12 2 2 3 K2C1_HUMAN 33 (14) 2094.5 1.1E-03 81.9 Keratin, type II cytoskeletal 1 30 82 16 3 2 5 K22E_HUMAN 22 (10) 1562.8 5.7E-03 46.9 Keratin, type II cytoskeletal 2 3 47 10 16 3 5 epidermal A2GL_HUMAN 7 339.3 5.6E-04 111.4 Leucine-rich alpha-2- 3 111 15 37 5 7 glycoprotein PCDAD_HUMAN 6 (3) 152.1 4.0E-02 392.2 Protocadherin alpha-13 344 20 1 17 392 23 SRGN_HUMAN 9 481.5 3.0E-02 29.5 Serglycin 16 30 1 2 14 27 ALBU_HUMAN 38 (26) 2223.2 8.0E-04 9.3 Serum albumin 4 3 2 1 9 8 SAA4_HUMAN 2 83.2 5.0E-02 153.4 Serum amyloid A-4 protein 1 117 64 153 84 2 TTHY_HUMAN 10 778.0 2.8E-04 65.7 Transthyretin 3 66 4 26 2 17 VTDB_HUMAN 9 431.2 2.0E-02 45.5 Vitamin D-binding protein 2 10 19 25 45 2 ZYX_HUMAN 5 229.6 5.6E-04 126.9 Zyxin 10 8 12 79 127 2

Abbreviation: PG, plasma heparin green; PP; plasma EDTA purple; SG, serum STT gold; SR serum red.

243

Insulin-like growth factor-binding protein (IGFBP) 3, 5- 6 were detected in various blood samples and appeared in serum SG: PP EDTA showing 1-5 fold abundance differences. The remaining tube comparisons showed extreme variation ranging 3-60 fold.

Leucine-rich alpha-2-glycoprotein was detected in CTL, breast cancer and BBD samples ranging 4-32 fold variation in abundances, with a few similarities in plasma PP: PG tubes in the DCIS and IBC samples (3 fold variation). The serum SG: SR showed 2-15 fold variation in abundance, plasma 1-37 fold and serum gold SST: plasma showing 3-11 fold differences (with SR: Plasma 2-32).

Mucin-16 was only detected in DCIS which showed comparable finding between both serum tubes (SG: SR) and plasma tubes (PP: PG), although there was a large difference in abundance between serum: plasma samples ranging from 18-105 fold.

Osteopontin was only detected in CTL and DCIS samples. The protein abundance was 1-10 fold in serum SG: SR, plasma PP: PG 2-45 fold and serum SG SST gold: plasma 3-21 fold difference.

Profilin-1 was detected in CTL, DCIS and IBC with large degrees of variation in abundance between the four blood tubes. Plasma tubes PP: PG and Serum SG: plasma varied 1-16 fold and the serum tubes SG: SR and SR: plasma ranged 1- 292 (SG: SR 7- 292 fold and SR: plasma ranged 1- 54).

Selenoprotein P was detected in CTL and breast cancer samples. The abundance was highly varied in all four blood tubes ranging 3-6000 fold. Only in the CTL samples, similarities were found with serum SST SG: Plasma 3-4 fold and plasma PP: PG 1 fold variation. Also, DCIS had comparable abundance of selenoprotein P in the two serum samples (SG: SR 1 fold).

Serglycin was detected across the various samples with similar abundances in serum (SG: SR 1-5 fold). Although with the plasma samples, serglycin abundance varied between the control and breast cancer samples. The findings from CTL and

244

BBD were 2-3 fold and in breast cancer there was an 8-14 fold difference. Furthermore, the expression in serum: plasma tubes ranged 5-79 fold, which clearly demonstrated that although the breast cancer samples had an effect on protein abundance, blood tube selection does impact on the detection of serglycin.

Transthyretin was detected in all blood samples with similar abundance values in the serum SG SST gold tube and plasma PP and PG (average variation 2-5 fold). The serum SG: SR ranged 3-41 fold, plasma PP: PG 2-26 fold and SR: plasma ranged 1-46 fold.

Uncharacterized protein C1orf56 was detected in the CTL and MBC sample. The abundance was comparable in serum: plasma tubes with an approximate fold change of 2-7.

Vitamin D-binding protein was detected in the CTL, BC (IBC and MBC) and BBD with similarities in serum SG SST: plasma 1- 4 and plasma PP: PG at 2-3 fold (except in BBD 25), although the variation was higher in serum SG: SR at 1-19 fold.

Vitronectin was identified in the CTL and IBC samples. The abundance was quite varied across all four blood tubes. Plasma PP: PG was similar at 1-4 fold, whilst serum SG: SR and serum: plasma varied 1-29 fold.

Zyxin expression was also varied across the 4 tubes. The most similar abundance was found in the serum SG SST: plasma ranging 1-12 fold, whilst all other tubes had varied expression 1-127 fold (SG: SR at 1-47 fold, plasma PP: PG at 1-58 fold and SR: plasma 2-127).

Additionally, comparison of some more frequently detected proteins across all the samples, which include complement proteins (C3, C4, C5, C9, B and D), keratins and apolipoproteins, revealed that there were large variations in expression of these proteins in the four different blood tubes. The complement protein abundances were markedly inconsistent.

245

The control sample comparison showed SR: PP-EDTA only 1-2 fold (except one case KC17= 263), whilst serum to serum (SG: SR) and plasma to plasma (PP: PG) varied 1-20 fold; and overall serum to plasma had 2-40 fold difference (SG: PP 3-40 fold change; SG: PG 2-15; PG: SR 1-26). Similar results were found in the breast cancer samples (DCIS, IBC and MBC) with serum SST: Plasma EDTA found to be 1-7 fold whilst other tubes when compared showed large variations (serum SG: SR 4-<100; plasma PP: PG 1-13; SR: plasma 10-89). BBD- Serum: Plasma was found to be 1-11 fold, serum SG: SR 2-7 fold and plasma PP: PG variation less than 2-3 fold. The different Keratins (type 1 and II) proteins detected, had varied abundances in the four blood tubes for all the blood samples. The serum SG: SR ranged 4-16 fold in certain samples and others with 3 to >100 fold, with the same pattern for plasma to plasma (6 to >100) and serum to plasma (2-80 fold), demonstrating that the protein amount detected in each of the blood tubes were quite different for Keratin. The apolipoproteins were mainly detected in the breast cancer samples, with similar abundances in the DCIS and IBC in serum SG: PP EDTA tube comparison at 1-2 fold (the remaining tubes ranged 1-194 fold). In the MBC samples, comparisons ranged 1-74 fold across all the blood tubes (serum: plasma 1-74 fold, SG: SR 3-44 and PP: PG 1-24).

In summary, seven proteins including alpha-2-glycoprotein, clusterin, IGPBP3, mucin-16, selenoprotein, uncharacterized protein C1or f56 and Vitamin D-binding protein were found to have similar abundances between serum SST separation tube and plasma EDTA separation tube. By comparing the remaining blood tubes (plasma PP: PG and plasma PP/PG: serum SR tube types), we found significantly different levels of greater than 5-fold change in nine proteins including: extracellular matrix protein-1, insulin growth factor binding proteins, Leucine-rich alpha-2- glycoprotein, osteopontin, profilin-1, serglycin, mucin-16 and Vitamin D-binding protein, zyxin.

The apolipoproteins detected had similar abundances in the serum SG: PP EDTA samples. The complement proteins (C3, C4, C5, C9, B and D) and keratins were significantly different in levels between all four tube types. Taken together, these observations of mean differences and ratio variations were supported statistically, 246

indicating that tube type is a significant determinant of protein level and selection of serum or plasma proteins will vary if the correct blood tube is not selected.

BLOOD TUBES RESULTS 0-3KDA

The 0-3kDa fraction (analysed in the same way statistically as the 3-50kDa fraction) protein comparison showed a large variation across the four blood tubes. This fraction was rich in novel proteins with potential for biomarker application, as shown in Table 5-4.

The different abundance patterns for several of the novel proteins detected across the four blood tubes are summarised in Table 5-4: in the control samples abundances of proteins A-kinase anchor protein 13, Dermcidin and Retinol-binding protein 4 were varying, whereas similar abundances of Hemopexin (serum: plasma 1-4 fold, PP: PG 4 fold) and Protocadherin Fat 2 (SG: SR: PP 1-2) were detected. In the DCIS samples, three proteins Ankyrin-1, Myosin-binding protein C and Protein ELYS had varied abundances, whilst protein Epiplakin abundance were similar in most tubes (1-4 fold, except SG: SR 14 fold). Several proteins were found to be similar in abundances in IBC samples over most blood tubes: Dystonin, Neuron navigator, Protein ELYS, Protein sidekick-1, Tankyrase-1, Zinc finger protein 638 and Carcinoembryonic antigen-related cell adhesion molecule 5 (1-7 fold). However, there were large differences in four proteins detected including E3 ubiquitin-protein ligase RNF2, Neuroligin-2, Ryanodine receptor 1 and uncharacterized protein C12 orf35. In MBC samples, there were five proteins with similar abundances in the SG SST: PP EDTA tubes; Dedicator of cytokinesis protein 11, Girdin, Myosin-8, Nesprin-1, usher in (SG SST: PP 1-5 fold).

In BBD, most of the proteins compared over the four tube types were similar for five proteins including Ceramide kinase, Condensin complex subunit 3, Desmoglein-1, Placental protein 13 and Titin (SG: PP: PG 1-5 fold and PP: PG 1-3 fold). One protein, Filaggrin-2, was present at comparable levels in the plasma tubes (PP: PG: SR 3 fold).

247

Table 5-4. Differential expression of proteins in serum and plasma in 0-3kDa fraction.

Accession Peptides Score Anova Fold Protein Description Blood tube comparison ratio (p)* Control 0-3kDa SG:PP SG:PG SG:SR PP:PG PP:SR PG:SR AKP13_HUMAN 1 36.2 8.1E-06 26.6 A-kinase anchor protein 13 17 27 19 2 1 1 DCD_HUMAN 1 54.5 1.3E-03 80.2 Dermcidin 80 3 14 31 6 5 HEMO_HUMAN 16 (14) 739.9 5.6E-03 15.4 Hemopexin 4 15 11 4 3 1 FAT2_HUMAN 5 (2) 114.6 4.5E-03 14.0 Protocadherin Fat 2 1 11 1 9 2 14 RET4_HUMAN 3 (2) 168.4 7.2E-03 877.6 Retinol-binding protein 4 35 641 877 18 25 1

DCIS 0-3kDa ANK1_HUMAN 2 (1) 67.5 2.3E-03 425.3 Ankyrin-1 162 397 1 2 173 425 EPIPL_HUMAN 3 (1) 97.1 2.7E-03 14.5 Epiplakin 4 3 14 1 4 5 MYPC2_HUMAN 2 (1) 61.2 7.2E-04 8.1 Myosin-binding protein C, fast-type 1 7 6 8 7 1 ELYS_HUMAN 3 107.5 2.4E-04 49.0 Protein ELYS 18 49 45 3 3 1

IBC 0-3kDa CEAM5_HUMAN 4 (3) 92.1 2.0E-03 7.1 Carcinoembryonic antigen-related 7 6 5 1 1 1 cell adhesion molecule 5 DYST_HUMAN 5 (4) 113.7 1.5E-05 6.5 Dystonin 2 3 2 7 4 2 RN213_HUMAN 3 (2) 74.1 2.5E-03 116.9 E3 ubiquitin-protein ligase RNF213 35 13 117 3 3 9 NLGN2_HUMAN 1 34.2 6.6E-04 136.5 Neuroligin-2 26 58 137 2 5 2 NAV3_HUMAN 6 (3) 132.7 2.9E-03 4.3 Neuron navigator 3 4 2 3 2 1 1 ELYS_HUMAN 3 (2) 89.8 3.3E-04 26.1 Protein ELYS 7 11 26 2 4 2 SDK1_HUMAN 3 (2) 78.9 4.9E-03 3.1 Protein sidekick-1 2 1 1 3 2 2 RYR1_HUMAN 1 50.3 2.8E-04 5586 Ryanodine receptor 1 823 241 5530 3 7 23 TNKS1_HUMAN 4 (2) 101.8 9.4E-05 8.7 Tankyrase-1 2 9 3 5 1 3

248

CL035_HUMAN 4 (2) 90.2 2.9E-03 175.5 Uncharacterized protein C12orf35 13 104 175 8 13 2 ZN638_HUMAN 4 (3) 86.2 1.7E-04 57.8 Zinc finger protein 638 2 3 58 1 32 21

MBC 0-3kDa DOC11_HUMAN 5 (1) 100.2 1.2E-03 54.3 Dedicator of cytokinesis protein 11 2 5 12 3 21 54 DCD_HUMAN 1 59.3 4.2E-04 92.4 Dermcidin 7 3 13 22 92 4 GRDN_HUMAN 4 (3) 101.1 2.2E-03 8.0 Girdin 2 8 2 3 1 4 MYH8_HUMAN 3 (1) 71.7 6.4E-04 167.8 Myosin-8 1 1 154 2 108 168 MTMRD_HUMAN 3 (1) 72.8 2.0E-03 54.8 Myotubularin-related protein 13 44 38 1 1 55 47 SYNE1_HUMAN 8 (1) 208.0 9.1E-03 6.5 Nesprin-1 2 6 5 4 3 1 S12A5_HUMAN 3 (2) 75.3 5.7E-04 4.5 Solute carrier family 12 member 5 5 1 2 4 2 2 RIF1_HUMAN 3 (2) 67.3 8.6E-03 11.9 Telomere-associated protein RIF1 6 11 1 2 7 12 USH2A_HUMAN 3 (1) 96.0 1.0E-02 27.3 Usherin 3 27 4 9 1 8

BBD 0-3kDa CERK1_HUMAN 1 23.5 1.5E-03 10.7 Ceramide kinase 3 2 4 7 11 1 CND3_HUMAN 3 (1) 82.2 2.2E-05 12.0 Condensin complex subunit 3 2 1 5 2 12 6 DSG1_HUMAN 5 (4) 261.2 2.2E-03 26.0 Desmoglein-1 2 1 17 2 26 16 PPL13_HUMAN 1 38.6 6.3E-03 8.1 Placental protein 13 1 2 6 3 8 2 TITIN_HUMAN 13 (6) 287.5 7.8E-05 8.5 Titin 5 6 1 1 7 8

Abbreviation: PG, plasma heparin green; PP; plasma EDTA purple; SG, serum STT gold; SR serum red.

249

Overall, it seems that the majority of these interesting and novel proteins in the 0- 3kDa fraction were detected in serum (SST gold blood collection tube) and plasma (purple EDTA collection tube), demonstrating that the use of these tubes would be helpful in the detection of novel low mass proteins as potential markers.

CONCLUSION

Serum and plasma proteins are currently used as routine clinical indicators of health status and treatment response e.g. biochemically a useful clinical tool for HDL (high- density lipoprotein), LDL (low-density lipoprotein) assessment of the cardiovascular system, monoclonal immunoglobulins/paraproteins as a marker of multiple myeloma (plasma cell malignancy) and C-reactive protein (CRP) to assess for the presence of an inflammatory response. Serum and plasma contain intrinsic plasma proteins including circulatory proteins and other HMW proteins released by tissues and cells as part of a pathological process (Marshall, Jankowski et al. 2004), making it a complex medium, as detailed in Chapter 1.9. Therefore, it is very important in collection and preparation of blood for proteomics analysis.

Advances in proteomics technology hold great promise in the understanding and treatment of the molecular basis of disease. Though before routine proteomic analysis can be achieved in the clinic one of the main technical obstacles is standardisation of methodologies. The blood collection process is critical to the accuracy of biomarker detection and discovery. To discover and validate biomarkers, serum and plasma proteins pose a significant and well documented analytical challenge. The process of blood collection itself is critical to the accuracy and reproducibility of quantitative biomarker assays. Due to concerns regarding whether serum or plasma is better for biomarker discovery, we investigated the use of serum and plasma samples using four different BD Vacutainer to determine which blood tube would be most suitable in the application of breast cancer biomarker studies. The handling and preparation were performed according to manufacturer’s instructions and remained constant.

250

Using a combination of MS technologies with improvements in sample preparation, we performed a proteomic analysis and found the anticoagulants and additives did provide differential protein profiles that affected the total number of proteins and peptides identified. We detected up to 130 proteins in serum and plasma by LC- MS/MS. The comparison of protein abundances across all four blood tubes were greatly varied in the breast cancer, BBD and control samples, showing a 70% variation.

Consequently, the tube applied also had an effect on the individual proteins available and this was an important factor in the detection of novel biomarkers. We selected a panel of 15 abundant proteins to analyse and further evaluate the fold changes across the different blood tubes. Although occasionally the magnitudes of the discrepancies seemed clinically substantial due to the sample analysed, we showed statistically that the differences in the levels of several of the proteins could be attributed to the blood collection tube. Additionally, certain novel proteins (Mucin- 16, S100-A6) were only detected in either serum or plasma, which would influence the perceived analysis of the altered pathophysiological state.

Therefore, we have identified that the blood tube selected for proteomic analysis has a large effect on the proteins detected and their abundances dependant on the anticoagulants and additives. Consequently, tube selection is critical to detecting biomarkers and establishing a clinically relevant detection test. Furthermore, accurate quantitation of proteins is possible when the problematic tubes are avoided. As shown in this study, the magnitude of difference when comparing the BD Vacutainer ® Serum Red Tubes (coated with micronised silica particles) to the serum BD Vacutainer® SST™ gel tube and to the plasma tubes, was so high that it is not recommended for use in serum application analysis. Although the lithium plasma tubes (BD Vacutainer ® LH PST™ II, heparin green tubes) with plasma separator gel were easy to work with, the results were often varied when compared to the serum gold and plasma EDTA tubes.

251

Additionally, the aim of this study was to determine the tube which was most suitable for biomarker studies. After comparing a number of proteins and individual protein abundance in the four different blood collection tubes the results demonstrated that both serum and plasma were useful only with the application of the serum gold BD Vacutainer® SST™ gel tube and plasma collection with BD Vacutainer ® EDTA purple tubes respectively. The most informative data was obtained from the low to medium (3-50kDa) fraction.

The process of blood collection itself is critical to the accuracy and reproducibility of quantitative biomarker analysis. Chapter 6 will detail these tubes and the biomarker findings. This study should provide useful blood sample collection information for researchers contemplating undertaking protein biomarker studies. Finally, the time between venepuncture and freezing, processing/storing containers, centrifugation speed, and the temperature of storage are the most critical variables for plasma. Critical process and variables for serum are: process/storage containers, time of clot retraction/removal of the fibrin clot with associated platelets and other cellular elements, centrifugation speed, and temperature of storage. Previous studies have shown the benefit of protease inhibition in stabilising the presence of endogenous plasma peptides when samples have been maintained at sub-optimal temperatures for extended time periods (Yi, Kim et al. 2007, Yi, Liu et al. 2008). However, for clinical biomarker validation this situation is undesirable and could be managed by strict adherence to protocols by adequate resourcing of technical staff required for sample collection and processing (Randall, McKay et al. 2010). This extensive analysis performed on serum and plasma proteins to date in relation to blood tube collection has laid the foundation for the selection and application of blood tubes in the identification of novel protein biomarkers in breast cancer. If the samples are handled correctly and standardised by most publications, then the results should provide a true comparison. This study provides practical blood sample collection information for researchers undertaking protein biomarker studies.

252

Chapter 6 Blood biomarkers in breast cancer

Improving the early diagnosis of breast cancer is a major clinical challenge due to its biological heterogeneity. Blood profiling has great potential to be developed as a simple test for clinical practise to augment current screening protocols which may translate into improved prognosis and monitoring of disease progression. Therefore, in this chapter, serum and plasma profiling, using proteomic techniques was applied to identify novel proteins. The proteins expressed were then correlated with the different stages of breast cancer. Additionally, the findings from serum and plasma were further validated in breast cancer cell lines as well as patients’ blood and tissue samples to reinforce their strength as a potential panel for breast cancer detection.

The work in this Chapter has been submitted for publication: Beretov J, Wasinger V. C, Millar E. K. A, Schwartz P, Graham P. H, Li Y. A panel of novel serum and plasma proteins revealed by LC-MS/MS in breast cancer. Breast cancer Research. Submitted March 2016. [Under review]. 253

6 A PANEL OF NOVEL SERUM AND PLASMA PROTEIN REVEALED BY LC-MS/MS IN BREAST CANCER.

INTRODUCTION

The incidence and mortality of breast cancer as the most common malignancy in women is highlighted in chapter 1.1 (Ferlay J, Soerjomataram I et al. 2012). Although prognosis is relatively favourable for breast cancer when diagnosed at an early clinical stage (10-year disease-free survival usually exceeds 80%), a high percentage of patients (usually about 20-30%) are at high risk of recurrence or metastasis. These patients require adjuvant endocrine chemo- and/or radiotherapy although only a fraction of the patients will benefit from the additional treatment.

Cancer biomarkers studies have identified a dynamic range of proteins in plasma and serum (Hanash, Pitteri et al. 2008), which provides an active representation of physiological and pathological status of individuals and holds the key to achieving accurate diagnosis and prognosis. Unfortunately, the only markers current clinically utilised are CA 15-3, CA 27.29, CEA in blood and ER, PR, HER2, uPA, PAI-1 and certain multi-parameter gene expression assays in tissue (Harris, Fritsche et al. 2007, Harris, Ismaila et al. 2016). These tests lack the sensitivity and specificity for use as screening tests for the early detection of breast cancer and are not recommended by the American Society of Clinical Oncologists (ASCO)(Duffy 2006), only being approved for use to guide treatment decision of patients with breast cancer. Presently, breast cancer biomarker discovery includes proteins and peptides which have been identified in breast cancer cell lines, nipple aspirate fluid, breast tissue, in addition to serum and plasma, as discussed in detail in Chapter 1.9, although none have been successfully applied to clinical use to date, especially as markers of early disease. Therefore, there is an urgent need to search for novel breast biomarkers to improve the early detection and accuracy of diagnosis, to determine the aggressiveness of breast cancer and monitor efficacy of treatment.

254

Blood is an easily accessible biological sample allowing the detection of various cancer related biomolecules either as the body’ systemic response to cancer or secretions directly from cancer cells, as discussed in Chapter 1.9. This makes the analysis of blood a favourable sample, which may link to the pathological changes occurring in breast cancer progression. Analysis of the serum or plasma proteome represents the most extensively studied biological matrix for cancer biomarkers (Hanash, Pitteri et al. 2008). Proteomics has been shown to be a promising method in breast cancer marker candidate for serum (Fan, Wang et al. 2010), which with the improvement in LC-MS methodologies, has come the renewed interest in proteomics analysis of blood (Martosella, Zolotarjova et al. 2005, Zolotarjova, Martosella et al. 2005).

A technical challenge in serum proteome analysis is that serum contains thousands of proteins and peptides that are present in a large dynamic concentration (Anderson and Anderson 2002). Reviews of MS-based blood analysis is an emerging method of clinical proteomics and cancer diagnostics, especially looking at the LMW component of the blood proteome in serum as a promising source of previously undiscovered biomarkers (Aebersold and Mann 2003, Liotta, Ferrari et al. 2003, Rosenblatt, Bryant-Greenwood et al. 2004, Liotta and Petricoin 2006). When mining for novel biomarkers, it was demonstrated (see Chapter 5) that depletion of abundant proteins with sample fractionation was essential when searching for low- abundance serum proteins or peptides. The data discovered in Chapter 5 was used to further assess differential protein expression, as a potential diagnostic tool in the different stages of breast cancer.

For the discovery phase, LC-MS/MS proteomics was performed on 184 blood samples employed for quantitative analysis and selected protein signatures. Serum and plasma samples were mined for biomarkers for breast cancer. After performing the quantitative proteomics, CLU, IGFBP3, LTG1, S100-A6 and VTN, were selected for validation studies, applying WB to breast cancer cell lines and serum/ plasma samples. Furthermore, IHC was performed in tumour tissues to distinguish the

255

expression of four of these proteins in breast cancer tissues from paired adjacent normal tissues.

MATERIALS AND METHODS

Study design and ethics

The study was performed at the Cancer Research laboratory, St George and Sutherland Clinical School, and was approved by the South Eastern Sydney Local Health District Human Research Ethics Committee (SEA HRCE) (#07/71Li), detailed in Chapter 2.1. All participants provided informed written consent indicating their voluntary participation. None of the subjects had received any prior treatment, either endocrine or chemotherapy.

The serum and plasma samples were obtained from breast cancer patients at St George Private Hospital, Sydney, Australia prior to surgery. The median age was 51 ± 10.5 years (range 35-70 years). The non-cancer control samples included blood samples from 6 patients with BBD and 20 healthy controls. The healthy controls were age matched with the breast cancer patients, median age was 49 ± 8 years (ranging 35-65).

All tissue specimens were removed as part of routine surgery and subjected to routine pathological examination at the Department of Anatomical Pathology (SEALS) St George Hospital. The breast carcinoma typing and grading were performed by a pathologist according to the World Health Organization criteria (Lakhani, I.O. et al. 2012). Once the tissue specimens were pathologically assessed, the blood samples were separated into the breast cancer stages or BBD (n=6). The study cohort consisted of 20 breast cancer patients with DCIS (n=6), IBC (with or without axillary lymph node involvement, but no distant metastases n=8) and MBC (distant metastases to viscera or bone, n=6). The histopathology characteristics and clinical features are summarised Table 4-1.

256

Sample blood collection protocol

The blood collection method and information regarding blood tube used, are detailed in Chapter 2.3.2. Blood samples were collected in four different BD Vacutainer blood collection tubes (2 serum/2 plasma) for each patient. The serum tubes used were BD Vacutainer® SST™ Tubes gold (SG) and BD Vacutainer® Serum Red Top (SR) tubes. Plasma samples were collected in purple EDTA tubes BD Vacutainer® K2E (PP) and light green lithium heparin tubes BD Plastic Vacutainer® PST II (PG). Serum and plasma were collected by spinning at 1300 x g for 10 min and stored at -80 °C, until needed.

Blood samples purification, concentration and digestion

The workflow for blood analysis consisted of sample collection, protein recovery with fractionation (purified was using a 3000 Da MWCO filter 5000 Da MWCO filter) and trypsin digestion followed by LC-MS/MS analysis. The serum and plasma fractionation equipment and method are detailed in Chapter 2.4.6

Briefly, two serum and two plasma samples for each breast cancer stage, BBD and control samples were concentrated into different MW fractions using microcentrifuge 50,000kDa MW and 3,000kDa MW cut off filtration devices. The samples were ultracentrifuged (9,000 x g) for 30 min at 4 C. Three fractions were collected for each sample: < 3kDa, 3-50kDa and > 50kDa. The fractions were then concentrated by speed-vac, vacuum centrifuge and stored at -20°C. A summary of the workflow is shown in Figure 6-1. The protein fractions were enzymatically digested with trypsin in an enzyme-to-protein ratio of 1:20 (w/w) and incubated at 37 C overnight, as detailed in Chapter 2.4.5.1.

257

Figure 6-1. Summary of workflow showing the steps for blood fractionation and concentration.

Purification of protein extract

After tryptic digestion, prior to LC-MS/MS analysis, the samples were desalted and concentrated using C18 StageTips. The workflow summary is shown in Figure 6-2. Protein clean-up and digestion was performed as previously described (see details in Chapter 2.4.5.2).

Figure 6-2. Work flow showing the steps required for LC-MS/MS of blood for biomarker discovery.

258

Protein identification by LC-MS/MS analysis

Label-free LC-MS/MS quantitative analysis was performed (using a LTQ-Velos Orbitrap) to detect abundant proteins in DCIS, IBC, MBC, BBD and CTL patient’s blood samples and to identify the proteins with significant differences in abundances (Beretov, Wasinger et al. 2015). The procedure for LC-MS/MS analysis of protein samples, is detailed in Chapter 2.4.7.

Label-free LC-MS serum-plasma profiling and data analysis

The 200 most intense fragment ions of each raw product ion spectrum were used for searches against the IPI human database with the following search parameters applied: default charge states of 2+, 3+, and 4+; mass tolerance of 4 ppm and fragment ion search tolerance of 0.4 Da; variable modifications permitted Deamination (NQ), Oxidation (M), Phospho (ST) and Phospho (Y). The peptide level score cut-off for each of the runs was automatically adjusted to ensure a 1% FDR throughout the experiments.

Data analysis was performed using Peak Integration with Progenesis LC-MS, as detailed in Chapter 2.4.9. Ingenuity Pathway Analysis (IPA) Software was used for canonical pathway enrichment analysis and the derivation of mechanistic networks, detailed in Chapter 2.4.10.

Cell lines and cell culture

In this Chapter, the human primary breast cancer cell line BT-474 and metastatic cell lines MDA-MB-231, MCF-7 and SK-BR-3 were used. The information for the cell lines and cell culture method were described in Chapter 2.2.3 and Chapter 2.5.1, respectively. Characteristics of the breast cancer cell lines detailing molecular subtype and receptor status are shown in Table 4-2.

259

Immunological confirmation of serum protein markers by Western blotting

Protein expression levels in breast cancer cell lines and patients’ blood samples were determined by WB, as previously described in Chapter 2.2. Details of the primary antibodies used are listed in Table 2-2.

Serum and plasma protein samples were prepared in Laemmli sample buffer and quantified using BCA protein assay, detailed in chapter 2.5.1.7. Protein expression levels of five putative protein markers were determined by WB analysis as previously described in Chapter 2.5.2 (Beretov, Wasinger et al. 2015). Briefly, proteins were resolved on 4-12% SDS-PAGE gels, transferred onto PVDF membrane and blocked for 1 hr in 5% BSA (in 0.1% TBST). The membranes were probed with primary antibody in 5% BSA (in 0.1% TBST) at the recommended dilution, details in Chapter 2.2.4, Table 2-2.

Western blotting was conducted using primary antibodies against Clusterin (anti- CLUAP1 rabbit polyclonal, Abcam ab198193) at 1:500 dilution, Vitronectin (anti- vitronectin mouse monoclonal, Abcam ab11591 ) at 1:500 dilution, IGFP3 (anti- IGFP3 rabbit polyclonal, Abcam ab76001) at 1:50 dilution, LRG (anti-LRG rabbit monoclonal, Abcam ab178698) at 1:1000 dilution and S100-A6 (rabbit monoclonal, Abcam ab181975) at 1:1000 dilution, was incubated o/n at 4 °C. This was followed by incubation in HRP-conjugated secondary antibodies; goat anti-rabbit (1:2000) or goat anti-mouse secondary antibodies (1:2000 dilution), for 1 hr at RT. Immuno- reactive protein bands were visualised using enhanced ECL substrate and imaged using the ImageQuant LAS4000 system (GE Healthcare NSW Australia). Mouse anti- β-actin MAb (1:2000 dilution, Abcam, USA) and / or mouse anti-GAPDH MAb (1:2000 dilution) were used as loading controls.

260

Immunohistochemistry staining and analysis

Standard immunoperoxidase procedures on paraffin sections were used to visualise protein expression on human tissue using the method detailed in Chapter 2.5.3. The information for all primary and secondary antibodies used is summarised in Table 2-2.

Briefly, paraffin sections from breast cancer and normal breast tissues were deparaffinised and re-hydrated in water, then washed in TBS. Antigen retrieval was performed for 30 min in 0.1 M citrate buffer (pH 6.0), then endogenous peroxidase activity was blocked with peroxide and non-specific interactions blocked with goat serum. The sections were then incubated with primary antibodies against either Clusterin (anti-CLUAP1 rabbit polyclonal, Abcam ab198193) at 1:100 dilution, Vitronectin (anti-vitronectin mouse monoclonal, Abcam ab11591) at 1:100 dilution, IGFP3 (anti-IGFP3 rabbit polyclonal, Abcam ab76001) at 1:100 dilution and LRG (anti-LRG rabbit monoclonal, Abcam ab178698) at 1:100 dilution, o/n at 4°C. The following day, sections were incubated in HRP-conjugated secondary goat anti- rabbit or anti-mouse antibodies (1:100 dilution) for 45 min at RT. Staining was visualised with DAB, then counterstained with Harris haematoxylin and Scott’s blueing. Control slides were treated in an identical manner, and stained with an isotype matched non-specific immuno-globulin (Rabbit polyclonal IgG). MDA-MB- 231 cell line was used as positive control. IHC staining was evaluated by (JB and YL) and a pathologist (EM).

Assessment of immunostaining

Staining intensity was assessed as detailed in Chapter 2.5.3.3, (Beretov, Wasinger et al. 2015). The staining intensity was scored between 0 and 3. The criteria for assessment were as follows: 0 (negative, 0 %); 1 (weak, 10-45 %); 2 (moderate, 45- 70%); 3 (strong, >70%) of the tumour cells stained. Evaluation of tissue staining was performed independently (JB, YL) and confirmed by a pathologist at St George Hospital (EM). All specimens were scored blind and an average of scores was taken.

261

RESULTS

Blood tube comparison fold change ratio

A total 184 venous blood samples were acquired from female breast cancer patients (n=20) and BBD (n=6) prior to surgery and cancer-free female volunteers (n=20). LC-MS/MS analysis was performed on fractionated samples to identify novel blood markers for early breast cancer. The comparative proteomic profiles of proteins detected in the four different blood tubes representing serum and plasma proteins in breast cancer, BBD and healthy control samples, were reported in Chapter 5. The list of 140 protein were generated, significant with p< 0.05, fold change > 3 and q< 0.02.

These results indicated that there are differences between serum and plasma protein expression patterns in breast cancer samples. Therefore, 3-50kDa fraction blood data on the breast cancer and BBD blood samples were compared to the normal control values and this fold change ratio was used to evaluate the significance of protein expression level between the different stages of breast cancer. A fold change ratio >3 fold was recorded as significant. Using this criteria, 56 proteins were observed to be significantly up-regulated or down-regulated, summarised in Table 6-1. Several of the proteins have been previously described in relation to breast cancer and some novel proteins were distinguished. Among the differential expressed proteins, there were 19 proteins of interest including Alpha- 2-HS-glycoprotein, Ceruloplasmin, Clusterin (CLU), Cystatin-C, Extracellular matrix protein 1 (ECM1), Kininogen-1, Leucine-rich alpha-2-glycoprotein, Osteopontin, Plasminogen, Protein AMBP, Profilin-1, Selenoprotein P, Serglycin, Tetranectin, Transthyretin, Uncharacterized protein C1orf56, Vitamin D-binding protein, Vitronectin (VTN) and Zyxin.

We subsequently focused on several proteins that were identified as increased in abundance (Table 6-1), including serum CLU and insulin-like growth factor-binding

262

protein 3 (IGFBP3), Leucine-rich alpha-2-glycoprotein (LRG1) and VTN, along with S100 Calcium Binding Protein A6 (S100-A6) only found in DCIS (Figure 5-3).

Progenesis PCA analysis of the blood samples

Using Progensis, the PCA plot analysis of normalised protein abundance data was used to capture any variation between samples according to the protein expression variation. The biplot shows the similarities between the triplicate runs of the same groups (same coloured dots), (Figure 6-3). The results within each group showed close congregation; DCIS (blue), IBC (beige), MBC (orange), BBD (purple) and CTL (pink), which indicates highly reproducible results amongst the three technical replicates.

Additionally, the information captures five distinct groups and demonstrates the variances between the individual breast cancer stages (DCIS, IBC, and MBC), BBD and CTL groups, indicating that there is no associated relationship.

263

Table 6-1. Comparison of protein abundance across 3-50kDa serum and plasma fraction in breast cancer and BBD.

Accession Protein Description DCIS: CTL Ratio IBC: CTL Ratio MBC: CTL Ratio BBD: CTL Ratio SG PP PG SR SG PP PG SR SG PP PG SR SG PP PP SR EDTA Li-H EDTA Li-H EDTA Li-H EDTA Li-H A1AG1_HUMAN Alpha-1-acid 8 122 31 -2 2 14 51 8 2 92 29 -2 glycoprotein- 1 A1AG2_HUMAN Alpha-1-acid 1 113 -2 - 6 872 180 38 -3 188 11 1 glycoprotein- 2 >100 A1BG_HUMAN Alpha-1B-glycoprotein 17 5 116 -2 16 7 75 2 18 -2 -53 FETUA_HUMAN Alpha-2 -glycoprotein 13 1 -2 -2 9 -2 22 -10 APOA4_HUMAN Apolipoprotein A-IV -23 -36 -188 -576 13 4 -10 -84 7 -9 -3 -35 -59 -471 -52 APOB_HUMAN Apolipoprotein B-100 9 -2 18 2 APOC3_HUMAN Apolipoprotein C-III 11 11 11 -1 APOE_HUMAN Apolipoprotein E -3 -1 -2 -65 5 11 5 -11 16 -1 2 -5 APOF_HUMAN Apolipoprotein F 3 -14 -14 -1 9 -36 -2 -3 APOL1_HUMAN Apolipoprotein L1 -1 -13 -3 -5 7 -23 -2 -15 CERU_HUMAN Ceruloplasmin 31 5 53 1 11 7 33 10 CLUS_HUMAN Clusterin 1 -4 -6 -6 17 5 -2 -4 12 -9 2 -5 -2 -51 -50 -188 CFAB_HUMAN Complement factor B 37 23 35 2 10 9 >100 -3 CO3_HUMAN Complement C3 18 7 2 -1 CO4B_HUMAN Complement C4-B 53 6 -2 1 6 -3 -2 -1 CO9_HUMAN Complement C9 7 2 3 -16 19 -13 2 -90 CFAD_HUMAN Complement factor D 155 3 57 -39 85 10 13 -7 CFAI_HUMAN Complement factor I 14 1 >100 -2

264

CYTC_HUMAN Cystatin-C 8 2 -3 -13 5 -2 43 -66 3 -2 -4 -23 ECM1_HUMAN Extracellular matrix 16 -3 3 -1 protein 1

GELS_HUMAN Gelsolin 19 3 1 -6 2 -2 -245 -60

HRG_HUMAN Histidine-rich -glycoprot 41 5 842 -1 15 1 109 -104 IGHG1_HUMAN IgG-1 chain C region 4 55 6 34 7 2 25 29 -5 26 8 15 2 17 8 10 IGHG2_HUMAN Ig G-2 chain C region 39 4 318 37 1 13 8 4 IGHG4_HUMAN Ig G-4 chain C region 8 4 36 -4 HV103_HUMAN Ig heavy chain V-I -V35 9 22 14 2 IBP6_HUMAN IGFBP-6 17 1 2 1 K1C10_HUMAN Keratin, type I- C 10 1 5 3 118 -1 3 -16 21 -5 26 -5 8 -5 10 2 14 K1C14_HUMAN Keratin, type I –C14 -85 -1 3 36 4 190 -8 7 -58 -193 -10 4 K1C9_HUMAN Keratin, type I -C9 -2 16 3 211 -1 20 -9 37 1 167 -3 26 -6 64 1 11 K2C1_HUMAN Keratin, type II –C1 -18 4 -12 5 2 25 -8 1 -60 11 -2 2 K22E_HUMAN Keratin, type II-C2 9 19 6 165 5 6 -3 51 -1 45 -6 4 2 11 3 13 epidermal K2C5_HUMAN Keratin, type II- C5 -83 -6 2 18 -59 7 -9 3 K2C6A_HUMAN Keratin, type II -C 6A -910 1 5 2 KNG1_HUMAN Kininogen-1 17 2 -1 -7 79 24 2 -5 34 2 18 -6 A2GL_HUMAN Leucine-rich alpha-2- >100 >100 >100 47 6 14 >100 -4 glycoprotein MGP_HUMAN Matrix Gla protein 6 -11 1 2 3 -1 -1 16 CD14_HUMAN Monocyte Ag CD14 141 -3 -1 -1 OSTP_HUMAN Osteopontin 65 2 -49 7

265

PLMN_HUMAN Plasminogen 35 -29 1 7 CXCL7_HUMAN Platelet basic protein 7 2 -1 -87 11 14 11 -56 6 -1 13 -16 2 1 2 -95 PZP_HUMAN Pregnancy zone protein 435 -1 5 2

AMBP_HUMAN Protein AMBP 2 5 19 -4 PROF1_HUMAN Profilin-1 25 -2 3 -134 187 4 2 2 THRB_HUMAN Prothrombin 57 43 9 33 20 6 11 40 11 2 46 33 RET4_HUMAN Retinol-binding protein 4 5 6 11 -25 2 3 16 -52 2 1 207 -84 SEPP1_HUMAN Selenoprotein P 2 -15 -80 -12 23 -10 -1 -16 2 7 4 3 SRGN_HUMAN Serglycin 21 13 -4 2 6 1 -1 3 7 1 -4 1 ALBU_HUMAN Serum albumin 6 15 11 -1 4 2 155 6 3 13 21 -1 1 4 4 -5 QSOX1_HUMAN Sulfhydryl oxidase 1 4 -10 -1 3 TETN_HUMAN Tetranectin 99 2 38 -6

TTHY_HUMAN Transthyretin -9 -15 -11 -122 16 6 -1 -21 37 -2 -2 -17 -69 -145 -2 -173 CA056_HUMAN Uncharacterized protein 21 -3 -5 -25 C1orf56 VTDB_HUMAN Vitamin D-binding 55 47 24 -1 6 23 62 2 -3 -7 -2 -17 protein VTNC_HUMAN Vitronectin 19 2 14 2 ZYX_HUMAN Zyxin 30 5 -3 7 1 3 -7 12 5 -4 -3 13

Abbreviation: antigen (Ag); cytoskeletal (C); immunoglobulin gamma (Ig G); Insulin-like growth factor-binding protein 6 (IGFBP-6); PG, plasma heparin green; PP; plasma EDTA purple; SG, serum STT gold; SR serum red.

266

Figure 6-3. Principal component analysis of all identified proteins from LC- MS/MS.

The PCA of the proteome dynamics based on the protein abundances in breast cancer (DCIS, IBC, MBC), BBD and CTL samples generated by LC-MS/MS.

Normalised abundances in 3-50kDa serum

Further systematic data analysis of the human proteins identified by LC–MS/MS in the breast cancer patients (DCIS, IBC and MBC) and BBD distinguished 50 differentially expressed proteins in serum. The proteins abundance values for breast cancer and BBD were normalised against the control samples. The strategy for normalisation was based on fold change of protein in breast cancer against control and the resulting fold change ratio was recorded. Proteins increased more than three folds were thought to be up-regulated and those decreased by three fold were thought to be down-regulated with negative (-) which indicated down regulated. A list of up and down regulated serum proteins identified are shown in Table 6-2. Details include proteins identified, mascot score and fold change ratio whereby value >3 fold is significant.

267

Table 6-2. A list of significant human breast cancer proteins identified by LC–MS/MS in 3-50kDa serum fraction.

Accession Peptides Mascot Anova Description DCIS: CTL IBC: CTL MBC: CTL BBD: CTL Score (p)* Ratio Ratio Ratio Ratio ATHL1_HUMAN 4 (2) 89.52 5.07E-05 Acid trehalase-like protein 1 3 -1 - >100 -1 A1AT_HUMAN 45 (20) 1694.25 1.41E-04 Alpha-1-antitrypsin -3 3 25 -4 A1BG_HUMAN 6 321.7 2.38E-08 Alpha-1B-glycoprotein 3 1 10 2 A2MG_HUMAN 19 (14) 863.53 1.89E-08 Alpha-2-macroglobulin 3 3 2 2 ANK3_HUMAN 2 (1) 46.06 0.02 Ankyrin-3 -5 -10 196 -8 ANXA2_HUMAN 2 (1) 48.31 0.01 Annexin A2 -1 134 2 -1 APOA4_HUMAN 63 (48) 2573.22 1.48E-03 Apolipoprotein A-IV -4 2 9 -6 APOE_HUMAN 19 (14) 956.55 4.72E-04 Apolipoprotein E -7 2 17 -3 APOF_HUMAN 2 85.94 0.05 Apolipoprotein F 3 2 4 2 APOL1_HUMAN 4 (2) 217.78 8.55E-03 Apolipoprotein L1 -3 3 -1 -10 CAH1_HUMAN 6 261.71 6.40E-06 Carbonic anhydrase 1 6 -2 -9 -9 CMGA_HUMAN 3 139.75 0.000213 Chromogranin-A 1 8 9 0 CTRL_HUMAN 2 (1) 54.26 2.75E-06 Chymotrypsin-like protease CTRL-1 4 1 -6 1 CLUS_HUMAN 11 (8) 443.22 0.03 Clusterin 2 22 13 1 CO5_HUMAN 9 (8) 301.62 5.12E-05 Complement C5 -5 20 5 -3 CO6_HUMAN 2 68.58 7.01E-03 Complement component C6 -17 1 5 -1 CO9_HUMAN 5 (3) 207.12 1.02E-04 Complement component C9 3 8 20 2 CFAD_HUMAN 2 62.52 1.42E-03 Complement factor D -14 -4 11 -22 ETAA1_HUMAN 2 57.78 1.12E-04 Ewing's tumor-associated antigen 1 -5 12 67 1 FABPL_HUMAN 1 35.92 1.82E-04 Fatty acid-binding protein, liver 0 2 3 0 268

GELS_HUMAN 6 (5) 289.92 0.01 Gelsolin -2 1 18 1 GPX3_HUMAN 7 258.31 1.07E-05 Glutathione peroxidase 3 -2 2 3 -5 GSTP1_HUMAN 1 31.73 2.92E-06 Glutathione S-transferase P 6 3 -61 -1 HDAC3_HUMAN 2 (1) 44.25 6.19E-03 Histone deacetylase 3 20 2 -14 1 IBP3_HUMAN 2 94.49 0.01 Insulin-like growth factor-binding -43 8 1 -8 protein 3 K1C10_HUMAN 12 (4) 583.95 6.81E-03 Keratin, type I cytoskeletal 10 -2 -2 -4 -2 KNG1_HUMAN 20 (19) 744.23 9.92E-04 Kininogen-1 -2 3 3 -1 A2GL_HUMAN 7 (4) 305.4 3.58E-07 Leucine-rich alpha-2-glycoprotein 7 2 2 2 MATN3_HUMAN 2 (1) 47.46 8.86E-12 Matrilin-3 3 35 -10 2 NF1_HUMAN 4 (1) 100.75 6.60E-04 Neurofibromin 0 0 179 0 PI16_HUMAN 1 44.76 0.00794 Peptidase inhibitor 16 0 0 3 0 PLMN_HUMAN 7 (5) 294.92 0.01 Plasminogen 1 1 11 1 PLLP_HUMAN 1 39.37 6.67E-08 Plasmolipin 1 1 0 4 PZP_HUMAN 10 (5) 400.03 1.29E-05 Pregnancy zone protein 1 -2 198 -4 HEG1_HUMAN 5 (3) 191.02 2.04E-03 Protein HEG homolog 1 -76 15 12 1 KIBRA_HUMAN 2 (1) 43.1 1.45E-07 Protein KIBRA 1 2 0 3 ZPI_HUMAN 1 101.35 1.98E-03 Protein Z-dependent protease inhibitor -22 24 3 -6 RGAG1_HUMAN 2 (1) 46.74 4.80E-03 Retrotransposon gag domain-containing 3 0 222 1 protein 1 GDIR2_HUMAN 2 85.09 7.70E-05 Rho GDP-dissociation inhibitor 2 8 2 -22 -1 SEPP1_HUMAN 2 108.19 1.03E-03 Selenoprotein P -18 -4 4 -48 SEMG2_HUMAN 3 (1) 68.24 6.61E-07 Semenogelin-2 >200 90 0 5

269

SRGN_HUMAN 7 (5) 249.26 9.74E-03 Serglycin 1 3 -2 1 ISK5_HUMAN 1 51.72 3.29E-06 Serine protease inhibitor Kazal-type 5 -8 -3 -2 4 ALBU_HUMAN 29 (17) 1214.57 7.01E-04 Serum albumin 6 -1 2 1 VTDB_HUMAN 12 (11) 529.18 1.79E-03 Vitamin D-binding protein 13 2 2 1 VTNC_HUMAN 2 88.97 0.02 Vitronectin 2 11 4 2 XKR6_HUMAN 2 68.88 8.01E-08 XK-related protein 6 2 2 6 2 ZN217_HUMAN 2 (1) 43.32 3.64E-03 Zinc finger protein 217 0 12 65 0 ZN609_HUMAN 3 (1) 69.3 0.02 Zinc finger protein 609 -1 -1 -8 5 ZA2G_HUMAN 4 (3) 180.19 3.28E-03 Zinc-alpha-2-glycoprotein 7 1 -5 2

270

Comparative profiling of the serum proteome from breast cancer patients revealed stage specific profiles, summarised in Table 6-2, with 17 proteins identified that were up-regulated in serum with a fold change ⩾3.0. These included 7 proteins identified in DCIS patients including; Carbonic anhydrase 1, Glutathione S- transferase P, Histone deacetylase 3, Leucine-rich alpha-2-glycoprotein, Serum albumin, Vitamin D-binding protein and Zinc-alpha-2-glycoprotein. In IBC samples, 3 upregulated proteins were identified including Annexin A2, Insulin-like growth factor-binding protein 3 and Serglycin. In the MBC patients, 7 up-regulated proteins were identified including Ankyrin-3, Gelsolin, Neurofibromin, Peptidase inhibitor 16, Plasminogen, Pregnancy zone protein and Selenoprotein P.

Several proteins were found to be up-regulated in two groups of patients' samples. In both DCIS and IBC samples, 8 up-regulated proteins were detected including Chromogranin-A, Clusterin, Complement C5, Ewing's tumour-associated antigen 1, Kininogen-1, Protein HEG homolog 1, Vitronectin and Zinc finger protein 217. Two proteins, Matrilin-3 and Semenogelin-2 were detected in both IBC and MBC patients.

Using IPA software, comparison of all the significantly abundant proteins detected in breast cancer against the controls is shown in Figure 6-4, highlighting both unique and common proteins in DCIS, IBC and MBC. The Venn diagrams illustrates that seven, three and seven unique proteins were identified in DCIS, IBC and MBC respectively. Also shown in this comparison was the number of proteins in common between the different breast cancer types whereby 27, 18 and 8 between DCIS-IBC, IBC-MBC and DCIS-MBC respectively.

271

Figure 6-4. Venn diagram for the distribution of proteins identified by LC- MS/MS in breast cancer.

IPA analysis, illustrated that unique proteins were identified in DCIS, IBC and MBC.

Heat map of diseases and bio functions

The IPA cluster analysis for diseases and bio-functions, of the protein expression levels in breast cancer and BBD is summarised in a heat-map form, shown in Figure 6-5. The details make evident that numerous proteins detected in breast cancer are associated with different disease states and cellular processes.

272

Figure 6-5. Cluster analysis of the diseases and bio-functions in breast cancer and BBD are summarised in a heat-map form.

273

We subsequently focused on the same proteins identified in the 3-50kDa serum blood tube data (Table 6-1), and studied these proteins in the normalised statistical data analysis, shown in Table 6-2. Our findings identified four up-regulated proteins, as potential markers of breast cancer stage were CLU (UniProt identification

P10909) detected in IBC and MBC (22 and 13 fold respectfully), IGFBP3 (P17936) detected in IBC (8 fold), LRG1 (P02750) detected in DCIS (7-fold ratio) and VTN (P04004) detected in IBC and MBC (11 and 4 fold respectively).

Canonical pathway analysis and heat map

Using IPA software, the complex protein profiles for the different diagnostic groups (Table 6-2) were analysed and compared. The level of protein expression changes were classified according to their biological functions and relationship to disease. The top 16 pathways were shown in Figure 6-6. We found that the canonical pathway related to acute phase response, LXR/ RXR activation extrinsic prothrombin pathway activation and production of nitric oxide and reactive oxygen species in macrophages. Down-regulated proteins were related with the complement system.

Figure 6-6. IPA showing the top related canonical pathways in breast cancer.

274

Heat map of disease and bio function

IPA core analysis, functional networks display of differentially expressed proteins detected in breast cancer and BBD are shown in Table 6-6, 6-7. This detailed analysis highlights the association of the abundant detected in breast cancer with various diseases and bio-functions. We found the significant proteins in breast cancer were associated with tumour growth and progression, suggesting that these proteins are involved in the proliferation of cells in breast cancer patients.

The heat maps display diseases and metabolic pathway, from the dataset of protein expression levels identified in breast cancer. The relevant diseases or bio-functions are ranked by statistical significance (by Fisher’s Exact Test), are displayed as shown in Figure 6-7 and Figure 6-8. The information shown in the right panel is the association of the different proteins as a small network, showing activation or inhibitory regulation control.

We then focused on the specific bio-functions identified as significant in breast cancer (dark orange colour indicates higher absolute z-score corresponding to a prediction of increase) connected to the casual network associated. The selected breast cancer proteins (CLU. IFFBP3, VITN), also showed a strong functional significance in other disorders. Based on the activity of proteins CLU, IFFBP3, VITN, we demonstrated that there was a strong correlation with their elevated levels activating the inflammatory response, cell homing and chemo-taxis (shown in Figure 6-7) along with fatty acid metabolism and cellular infiltration (shown in Figure 6-8). These findings illustrate the essential key role of our novel biomarkers on important biological processes in breast cancer.

275

Figure 6-7. IPA analysis of associated disease and functional networks with identified potential breast cancer markers.

The disease and bio-functions associated with the proteins identified in IBC and MBC showing protein involvement in pathway activation: (A) VTN in inflammatory response; (B) CLU and VTN involved in cell homing and (C) CLU, VTN and IGFPB3 involved in chemotaxis. The protein molecules in the network represented as red/pink are up regulation/elevated, relative to green that represents down regulation. The orange dotted line indicates predicted activation and blue predicted inhibition. Proteins of interest are tagged (*).

276

Figure 6-8. IPA analysis of associated disease and functional networks showing association with fatty acid metabolism and cellular infiltration.

The disease and bio-functions associated with the proteins identified in IBC and MBC are highlighted here showing: (A) CLU and VTN are involved in activating fatty acid metabolism and (B) IGFPB3 involved in inhibition of cellular infiltration. The protein molecules in the network represent as red/pink are up-regulated/elevated, relative to green that represents down regulation. The orange dotted line indicates predicted activation and blue predicted inhibition. Proteins of interest are tagged (*).

277

Validation of up-regulated proteins in blood and tissues

Numerous up-regulated proteins were identified by the LC-MS/MS in the different breast cancer stages. In order to further verify our findings, five up-regulated protein candidates including CLU, IGFBP3, LRG1, S100-A6 and VITN were validated in breast cancer cell lines, blood and tumour samples from breast cancer patients using WB and IHC, respectively.

Western blot in breast cancer cell lines

There was consistent expression of the five markers (CLU, IGFPB-3, LRG1, S100-A6 and VTN) in the human primary breast cancer cell line (BT474) and metastatic breast cancer cell lines (MDA-MB-231, MCF-7 and SK-BR-3) by WB, as shown in Figure 6-9.

IGFBP-3 was positive in all four breast cancer cell lines with different levels of expression. CLU was positive in the three metastatic cell lines (MDA-MB-231, MCF- 7 and SK-BR-3) but negative in primary breast cancer cell line (BT474). Proteins LRG1, S100-A6 and VTN were expressed in primary (BT474) and in at least two different metastatic breast cancer cell lines: LRG1 was positive in MDA-MB-231 and MCF-7; S100-A6 was positive in MDA-MB-231 (high) and SKBR3 and VTN showed low expression in MCF-7 and SKBR3. Parallel loadings of actin and GAPDH were used as protein loading quantification controls. Using the four cell lines (representing the various breast cancer subtypes), WB results confirmed that the potential serum markers from breast cancer patients, are closely associated with human breast cancer.

278

Figure 6-9. Validation of serum proteins in breast cancer cell lines by WB.

The identified protein expression for CLU, IGFPB-3, LRG1, S100-A6 and VTN, representing potential markers of various breast cancer stages, were observed in primary (BT474) and metastatic breast cancer (MDA-MB-213, MCF-7 and SKBR-3) cell lines using WB. Actin and GAPDH were used as loading controls. The results were from three independent experiments.

279

Western blot in serum and plasma

WB was used to determine if CLUS, IGFBP-3, LRG1and VIT were present in serum and plasma from breast cancer patients and normal healthy control subject samples. Our results demonstrated a strong relationship between serum findings with LC- MS/MS, shown in Figure 6-10.

Validation of four candidate proteins CLUS, IGFBP3, LRG1and VIT was performed by WB in breast cancer samples (n=13) and healthy volunteers (n=13). All thirteen breast cancer specimens demonstrated up-regulated levels of CLUS and IGFBP3 in plasma and most in serum (IGFBP3, 12 of 13) compared to the controls. LRG1and VIT had varying expression in serum and plasma: detected in 10 and 11 of 13 respectively in serum and only 6 of 13 in the plasma samples. These finding are consistent with LC-MS/MS screening of protein abundance data (Chapter 5) in Table 5-3. Therefore, we expect a baseline level of protein expression in the controls. Significantly the density of the protein expression in the disease samples is greater than the controls and consistent with LC-MS/MS normalised results (fold change ratio of disease: CTL), which were up-regulated >3 fold in breast cancer compared to the normal control samples.

The validation results of these four candidate proteins also supports the proteomics findings as they are also shown to be present in the individual patient’s blood samples.

280

Figure 6-10. Validation of identified serum proteins in serum and plasma by WB.

WB validation of 4 candidate proteins CLUS, IGFBP3, LRG1and VIT in breast cancer: (A) serum; (B) plasma and (C) healthy control serum samples. Differential expression of 4 proteins was observed in serum and plasma from breast cancer patients and health controls. Actin was used as a loading control.

281

Immunohistochemistry analysis of human breast carcinomas and normal breast tissue

The expression of CLUS, IFGBP3, LRG1 and VIT was further validated in human breast cancer tissues by IHC. Immunoreactivity was observed as areas of brown colour with staining intensity, as shown in Figure 6-11. The staining intensity was evaluated as a score from (A) low (score 1), (B) moderate (score 2) and, (C) strong expression (score 3), with positive expression defined as greater than 10% tumour cell staining.

The positive staining with CLUS, IFGBP3, LRG1 and VIT was found in the invasive component of the breast cancer cells. In breast cancer tissue, there was primarily cytoplasmic and nuclear staining of tumour cells with a granular pattern for clusterin (Figure 6-11). Overall, six of eight cases of breast cancer showed overexpression of CLU in IBC. For IGFBP-3, immunoreactivity was primarily nuclear within tumour cells with overexpression of IGFBP-3 in five of eight IBC tumour cases. LRG1 displayed cytoplasmic staining in tumour tissue, with overexpression in six of eight cases. Vitronectin was expressed in the cytoplasm and nuclear component of invasive breast cancer, five of eight cases of breast cancer showed overexpression of VTN. Most normal ducts and lobules displayed weak staining for CLUS, IFGBP3, LRG1 and VIT.

282

Figure 6-11. IHC of selected protein markers in human breast cancer tissue.

IHC staining of CLUS, IGFBP3, LRG1 and VIT expression in breast cancer tissue. This study shows cytoplasmic and/or nuclear staining for CLU, IGFBP3, LRG1 and VTN in invasive carcinoma, with mild staining in normal breast ducts (arrows showing normal duct). The varied intensities of positive staining with each protein are depicted as: (A) low expression (score 1); (B) moderate (score 2); (C) strong positive expression (score 3).

283

Prognostic value of the serum and plasma biomarker panel

The diagnostic accuracy of CLUS, IFGBP3, LRG1 and VIT was analysed in serum and plasma and compared to healthy subjects by WB and IHC.

Exploratory survival estimates were subsequently performed using Kaplan-Meier (K-M) analysis of publically available mRNA expression profiling data (Kaplan- Meier Plotter database, an online survival analysis tool http://kmplot.com/analysis/index.php?p=service&cancer=breast). This demonstrated that the mRNA expression of the selected proteins CLU, IGFBP3, LRG1 and VIT was associated with clinical outcome in a cohort of over 3,000 breast cancer patients (Gyorffy, Lanczky et al. 2010), shown in Figure 6-12. In this analysis, high expression of CLU, LRG1 and VIT was shown to be associated with improved recurrence-free survival. Therefore, we plan to further investigate the protein expression in a large independent cohort of breast cancer patients as part of a subsequent study.

In conclusion, protein mass profiling by LC-MS/MS revealed five serum proteins which, in combination, can distinguish between women with DCIS, IBC and MBC patients. We examined the association of the five-protein panel in serum and plasma along four with tumour tissue from the patients and found that overall overexpression was significantly associated with breast cancer. The results of this study showed up-regulated expression of clusterin, IGFBP3 LRG1 and vitronectin in tissue and serum/ plasma samples in nearly all breast cancer stages from the patient’s using WB and IHC validation strategies. The K-M analysis also showed that relative expression of clusterin, IGFBP3 LRG1 and vitronectin were also associated with outcome (recurrence-free survival).

284

Figure 6-12. Kaplan-Meier survival analysis for identified four potential proteins.

Kaplan-Meier survival analyses demonstrated that patients’ survival was associated with expression level of the target protein mRNA. Prognosis was improved with high expression levels: (A) CLU, (C) LRG1 and (D) VTN, but was associated with poor outcome for IGFBP3 (B).

285

DISCUSSION

The power of proteomics in cancer research has been most successfully used to date in the field of cancer research (Hanash and Hanash 2003, Tyers and Mann 2003). The focus of current proteomic research is the discovery of protein biomarkers for breast cancer diagnosis, prognosis and prediction of response. Protein profiles can serve as a powerful diagnostic tool, that can also predict treatment outcome in breast cancer. This study has provided a large scale systematic analysis of the differentially expressed proteins in breast cancer. In Chapter 5, we demonstrated that there were many variations in protein abundances detected in breast cancer patients’ samples depending on the blood tube applied. This resulted in a more detailed analysis of the serum proteins in the 3-50kDa fraction, which appeared to be a rich source of potential targets. Proteins secreted or shed by cancerous cells can therefore be captured in a peripheral blood sample for use as a diagnostic screening tool. Therefore, the aim of this study was to look for putative circulating markers of the early and advanced stages of breast cancer identified by high-resolution label- free proteomics. This was achieved by the depletion of high abundance proteins from serum and plasma with fractionation and then selecting serum as a source of the medium to low abundance proteins that provided detailed information for the breast cancer patients’ samples. Subsequent validation studies confirmed their altered expression in a panel of breast cancer cell lines, patient’s serum samples and matched patient tumour samples.

Statistical analysis revealed 50 proteins modulated in the different breast cancer stages (p<0.05) and among these we identified a panel of 17 up-regulated serum proteins. These included unique proteins for the different breast cancer stages: DCIS (7), IBC (3) and MBC (7) and additionally in both DCIS and IBC (8) plus IBC and MBC (2). Using this information, five candidate proteins were selected for validation with WB in breast cancer cell lines and four: CLU, IGFBP3, LRG-1, S100-A6 and VTN were validated in blood samples from patients and healthy subjects. Validation studies on patients’ tissue samples using IHC confirmed clusterin, IGFBP3, LRG and VIT as good diagnostic markers for breast cancer targets. CLU was detected in IBC and MBC (22

286

and 13 fold respectfully), IGFBP3 was detected in IBC (8 fold), LRG1 was detected in DCIS (7-fold ratio) and VTN was detected in IBC and MBC (11 and 4 fold respectively). Validation of the individual patients’ blood samples confirmed that the abundance of these proteins was found across all samples and not the results of skewing of the pooled samples by one very high patient sample reading.

Combining breast cancer protein profile data with IPA pathway analysis provided further insight into the biology of the disease processes which showed that CLU, IGFBP3, LRG1 and VTN, are involved with inflammatory response, cell homing, chemotaxis and cellular infiltration amongst others. Most of these overexpressed proteins were predicted to have activating effects on their respective pathways.

Clusterin (apolipoprotein J) is a 75 - 80kDa disulfide-linked heterodimeric protein associated with the clearance of cellular debris and apoptosis. It is involved in a variety of biological processes such as lipid transport, regulation of the complement cascade, sperm maturation, immune regulation, regulation of apoptosis, membrane recycling, cell adhesion and morphological transformation (Trougakos and Gonos 2002). Secretory clusterin protein is an inhibitor of apoptosis with a cytoprotective function (Zhang, Kim et al. 2005). Clusterin expression has been associated with tumorigenesis of various malignancies, including pancreatic (Xie, Motoo et al. 2002), ovarian (Xie, Lau et al. 2005), prostate (Miyake, Nelson et al. 2000) and colon cancer (Pucci, Bonanno et al. 2004), along with tumorigenesis and progression of human breast carcinomas (Redondo, Villar et al. 2000). Anti-clusterin treatment of breast cancer cells (MCF7 and MDA231) has demonstrated increased sensitivity to chemotherapy and tamoxifen and counteracts the inhibitory action of dexamethasone on chemotherapy-induced cytotoxicity (Redondo, Tellez et al. 2007). Additionally, a more recent study has shown that clusterin can sensitise breast cancer cells to kinase inhibitors (Redondo, Garcia-Aranda et al. 2015). Clusterin can also regulate breast cancer cell migration and invasion via MAPK and MMP9 (Li, Jia et al. 2012) and upregulate phosphatidylinositol 3- kinase (PI3K)/protein kinase B (AKT) pathway and insulin-like growth factor (IGF)-1 activates the PI3K/AKT pathway through upregulation of sCLU (Ammar and Closset 287

2008, Ma and Bai 2012). The expression of CLU was assessed by WB and IHC in breast cancer patients. The IHC revealed that six of eight cases showed high cytoplasmic and nuclear staining in invasive breast carcinoma tissue samples. In addition, WB analysis revealed up-regulated expression of CLU in serum and plasma.

The role of IGFBP-3 is to modulate IGF/IGF type I receptor (IGF-IR)-dependent and -independent actions in the circulation and immediate extracellular environment (Baxter 2013, Baxter 2014, Perks and Holly 2015). In addition, IGFBP-3 induces apoptosis by reducing the bioavailability of IGF-1 to the IGF-1 receptor (Jogie- Brahim, Feldman et al. 2009), inhibits cell proliferation in various cell lines (Mohanraj and Oh 2011, Baxter 2014) and was found in the tumour tissues, synthesised and secreted by various cell types in vitro (Baxter 2014). According to early epidemiological studies, low levels of IGFBP-3 were independently associated with a high risk of human malignancies, such as colorectal cancer, lung cancer, and breast cancer (Wang, Wang et al. 2013, Tas, Karabulut et al. 2014, Pankaj, Kumari et al. 2015). It was proposed that the IGFBP-3 gene could be a putative tumour suppressor gene and/or therapeutic target for human cancers (Choi, Park et al. 2013, Cao, Lindstrom et al. 2014). IGFBP3 protein expressions in human breast cancer is association with hormonal factors and obesity (Probst-Hensch, Steiner et al. 2010). In the present study, the expression of IGFBP-3 was assessed by WB and IHC in breast cancer patients. IHC revealed in five of eight cases showed nuclear IGFBP-3 staining in invasive breast carcinoma tissue samples. In addition, WB analysis revealed up-regulated expression of IGFBP-3 in serum and plasma. Interestingly, high expression level of IGFBP-3 mRNA in breast cancer was associated with poor prognosis (relapse free survival) by K-M analysis (p=0.048).

S100 proteins have multifunctional properties with a regulatory role in a variety of cellular and extracellular processes. The majority of the S100 genes are located on chromosome 1q21, which are often rearranged in cancer. Clinical interest of S100 as putative cancer biomarkers is continuously expanding as S100 proteins are a potentially promising group of biomarkers in cancer development and progression. 288

An analysis of S100 showed overexpression in breast cancer tissues although S100A2, S100A6, S100A11 and S100A13, appeared more sporadic (Cancemi, Di Cara et al. 2010). S100-A6 was detected only in DCSI samples and validated in breast cancer cell lines but required further validation in patient’s serum and tumour samples. Additionally, S100A14, acts as a modulator of HER2 signalling via ERK and PI3K/AKT (Xu, Chen et al. 2014).

Leucine-rich alpha-2-glycoprotein-1 (LRG) is a serum glycoprotein with predicted molecular weight of 34 to 36kD. The function of LRG1 remains unknown, although reports have predicted its role in cell adhesion (Takahashi, Takahashi et al. 1985, Kobe and Kajava 2001), cell survival and apoptosis (Ai, Druhan et al. 2008, Weivoda, Andersen et al. 2008). Elevated LRG1 in breast cancer is novel and has only been detected as elevated in ovarian (Andersen, Boylan et al. 2010), pancreatic (Kakisaka, Kondo et al. 2007), lung (Okano, Kondo et al. 2006), liver (Kawakami, Hoshida et al. 2005), colorectal (Wang, Shan et al. 2016, Zhang, Zhu et al. 2016) and bladder cancer (Linden, Lind et al. 2012). Our results showed increased serum LRG was observed in breast cancer patients by LC-MS/MS which was validated by WB and IHC. LRG1 displayed cytoplasmic staining in tumour tissue, with overexpression in six of eight cases.

Vitronectin (VTN) is a 65kDa cell adhesion molecule that interact with glycosaminoglycans and proteoglycans and has already been reported to be a promising marker for breast cancer in serum (Kim, Lee et al. 2009, Cho, Jung et al. 2010). Vitronectin has been shown to be altered in DCIS (Aaboe, Offersen et al. 2003, Kadowaki, Sangai et al. 2011). Its interaction with the uPAR, in tumour growth is thought to facilitate cancer cell invasion into the surrounding tissue (Aaboe, Offersen et al. 2003, Pirazzoli, Ferraris et al. 2013). The present study also demonstrated with WB that VTN was expressed at 65kDa, and that its expression was significantly increased in the serum of breast cancer patients when compared with that of normal controls. We also demonstrated increased expression of VTN in the cytoplasm and nuclei of breast cancer patient’s tumour tissue (in five of eight cases), suggesting that it could serve as a marker for the detection of breast cancer. 289

CONCLUSION

This study demonstrated that the expression of these selected proteins could be used as discriminating biomarkers in breast cancer with diagnostic relevance. Serum and plasma samples were mined for new biomarkers using a proteomics approach, which highlighted the proteins which are significant in breast cancer.

In summary, we have demonstrated that serum proteomics is a valuable approach to the discovery of protein biomarkers for breast cancer. The extensive LC-MS/MS protein information obtained from human serum and plasma revealed a list of significant proteins associated with different breast cancer stages. This study revealed candidate biomarkers such as CLU, IGFBP, LRG1, S100-A6 and VTN that may have a diagnostic value for breast cancer. These proteins could also play a role in the prediction of prognosis of breast cancers.

290

Chapter 7 General discussion

Artwork of a woman crying following mastectomy. Science photo library: by Paul Brown

291

7 GENERAL DISCUSSION AND FUTURE DIRECTIONS

THESIS SUMMARY

Breast cancer is a common complex and heterogeneous disease, the survival from which is dependent on stage at diagnosis, molecular subtype and response to therapy. Early detection offers the opportunity to modify the disease with earlier therapeutic intervention and provides the patient with a higher probability of improved prognosis before life threatening metastatic disease has supervened. Although current breast screening by mammography has reduced mortality over the last 20 years, it is far from perfect and is not capable of detecting disease in younger women with dense breasts or in those women with rapidly growing tumours which appear between two yearly screens. Thus the development of a simple blood or urine test to augment current screening protocols to improve early detection was the main initial aim of this thesis. This aim was driven by recent improvements in proteomics technologies which allow us to capture thousands of proteins from body fluids and to assess their value as a potential clinically valid marker of breast cancer.

The first part of the work in this thesis reviewed: (1) the worldwide incidence and aetiology of breast cancer; (2) general pathology of breast cancer; (3) morphology of breast cancer; (4) molecular classification of breast cancer and biomarkers available; (5) clinical aspects including presentation, screening, clinical staging and treatment; (6) current proteomics studies for breast cancer biomarker; (7) key proteomic techniques in biomarker research; (8) proteomics techniques and urinary biomarkers in breast cancer; (9) biomarkers in breast cancer and (10) mass spectrometry and breast cancer (Beretov, Wasinger et al. 2014).

General materials and methods for extraction, purification and application in proteomics analysis for urine and blood biomarkers studies, were detailed in Chapter 2. 292

A significant hurdle to the discovery of novel proteins and the validity of biomarker discovery is the handling of urine samples in a uniform manner. Therefore, a novel reproducible and standardised protocol for urine collection and for LC-MS/MS (Chapter 3) demonstrated that acetone/TCA precipitation with prolonged centrifugation time at a high speed was the most promising approach for urine preparation. This provided the optimal method for urinary protein precipitation for our breast cancer samples (Beretov, Wasinger et al. 2014). In Chapter 4, I identified novel urine candidate breast cancer biomarkers using LC-MS/MS-based proteomics. This provided a profile, detailing 59 significant urinary proteins and 13 novel up- regulated proteins, associated with the presence of breast cancer. These include stage specific markers associated with pre-invasive disease in the DCIS samples (Leucine LRC36, MAST4 and Uncharacterised protein CI131), early IBC (DYH8, HBA, PEPA, uncharacterized protein C4orf14 (CD014), filaggrin and MMRN2) and MBC (AGRIN, NEGR1, FIBA and Keratin KIC10). Expression of ECM1, MAST4 and filaggrin (as potential markers) were subsequently validated in breast cancer cell lines. MAST 4 expression was further confirmed in human breast cancer tissues and human breast cancer urine samples (Beretov, Wasinger et al. 2015).

Proteomics techniques were then used to differentiate the protein expression patterns between the different stages of breast cancer. The systematic searches for biological indicators, with the use of LC-MS/MS enabled the direct identification of proteins as putative blood biomarkers. Similarly to Chapter 3, standardisation of urine preparation performed, Chapter 5 also focused on laboratory methods: to establish a standardised collection system and an optimal blood tube selection for blood based biomarker detection. A detailed literature review demonstrated that the current focus of proteomic research in the discovery of blood-protein biomarkers showed inconsistency in blood collection medium (tubes) and preparation. This resulted in variability in biomarker identification, creating problems for data comparison. Therefore, in Chapter 5, I evaluated four different blood collection tubes using LC-MS/MS: two serum versus two plasma, additionally applying fractionation to isolate the large molecular weight proteins for breast 293

cancer proteomic biomarker studies. Therefore, serum and plasma profiling using proteomic techniques based quantitative LC-MS/MS analysis, showed altered proteomic patterns associated with different breast cancer stages. My observations described how a large variation between blood tubes in both the number of proteins detected (70% variation) and the abundances of the individual protein greatly varied across the breast cancer, BBD and control samples. Consequently, I demonstrated that both serum and plasma were useful only with the application of the serum gold top BD Vacutainer® SST™ gel tube or plasma collection with BD VACUETTE® EDTA purple tubes respectively. Also, I demonstrated how the low- medium MW (3-50kDa) component of the blood proteome was a promising source of previously undiscovered biomarkers.

Finally, in Chapter 6, the extensive LC-MS/MS protein information obtained from human serum and plasma in Chapter 5 was further analysed. From this information, five novel candidate serum-biomarkers: CLU, IGFBP, LRG1, S100-A6 and VTN were overexpressed in breast cancer and validated in a panel of breast cancer cell lines, serum, plasma and breast cancer tumour samples from patients. In addition, IPA pathway data analysis, identified roles of these molecules in the inflammatory response, chemotaxis, cell homing, activating fatty acid metabolism and cellular infiltration as well as association with common signalling pathways in breast cancer.

CONCLUSIONS AND FUTURE PERSPECTIVE

The Australian Institute of Health and Welfare, estimated that 15,270 new cases of breast cancer were diagnosed in 2014, with projections to suggest 17,210 in 2020 (AIHW 2015). This would equate to 47 females being diagnosed with breast cancer every day in 2020. Survival from breast cancer depends on two main factors – early detection and optimal treatment. Breast cancer diagnosed in the early clinical stages has a favourable prognosis (10-year disease-free survival usually exceeds 80%). Early diagnosis of breast cancer and monitoring its progression is still a major challenge.

294

The body is an extremely complex biological system of protein complexes that can be regulated by the host, signals emanating from cancer cells and their surrounding tissue microenvironment. An opportunity to provide patient care and management before the onset of the disease and can aid in the drug treatment plan and save lives. This thesis has identified novel methods and several important unique proteins associated with presence of breast cancer compared to healthy controls. The potential application of these findings are significant, as they offer the possibility for improving the early detection of breast cancer which was the preliminary aim of this project. Additionally, they may also be useful for monitoring the progression of disease and therapeutic monitoring to help detect early disease recurrence and subsequently allow earlier instigation of alternative therapies. These possibilities have the potential to have a major impact on the prognosis of breast cancer for the tens of thousands of women who succumb to the disease each year. However, the importance of a standardised and uniform approach to specimen collection and handling provided a good platform for serum and plasma protein profiling in breast cancer and was shown to be paramount to reproducible biomarker detection.

Confirmation of the clinical validity of my preliminary findings of these significant breast cancer associated biomarkers in urine and blood will require larger scale validation studies. Therefore, we plan to confirm our findings in a larger independent set of breast cancer patients (n=50), in a prospective manner, for patients with a new diagnosis of breast cancer with blood and urine collected pre- operatively. Furthermore, those markers which were over expressed in human breast cancer tissue samples (MAST4, CLU IGFBP3, LRG1 and VTN) and associated with clinical outcome on K-M analysis, will be investigated further as a potential IHC based biomarker of outcome for routine diagnostic histopathology.

In conclusion, the human proteome reflects the complex interactions of multiple biological systems in our body. The data identified with LC-MS/MS, on breast cancer patients, has highlighted this complexity whilst at the same time provided valuable information. Proteomics holds the key to revealing major signalling pathways, 295

activated in breast cancer, which can further provide useful data for application in patient treatment. The protein biomarkers discovered and validated, hold great promise as cancer biomarkers or drug targets of the future. They may substantially improve our knowledge on cancer biology and the clinical management of cancer. Despite the enormous proteomics investigations performed in the past decade on breast cancer biomarkers, none have been successfully applied to clinical practice. I hope that my findings will change this position in the near future and have potential clinical applications to improve the early detection of breast cancer.

296

8 REFERENCES

Aaboe, M., B. V. Offersen, A. Christensen and P. A. Andreasen (2003). "Vitronectin in human breast carcinomas." Biochim Biophys Acta 1638(1): 72-82.

Abba, M. C., J. A. Drake, K. A. Hawkins, Y. Hu, H. Sun, C. Notcovich, et al. (2004). "Transcriptomic changes in human breast cancer progression as determined by serial analysis of gene expression." Breast Cancer Res 6(5): R499-513.

Abdullah-Soheimi, S. S., B. K. Lim, O. H. Hashim and A. S. Shuib (2010). "Patients with ovarian carcinoma excrete different altered levels of urine CD59, kininogen-1 and fragments of inter-alpha-trypsin inhibitor heavy chain H4 and albumin." Proteome Sci 8: 58.

Adachi, J., C. Kumar, Y. Zhang, J. V. Olsen and M. Mann (2006). "The human urinary proteome contains more than 1500 proteins, including a large proportion of membrane proteins." Genome Biology 7(9): R80.

Adam, P. J., R. Boyd, K. L. Tyson, G. C. Fletcher, A. Stamps, L. Hudson, et al. (2003). "Comprehensive proteomic analysis of breast cancer cell membranes reveals unique proteins with potential roles in clinical cancer." J Biol Chem 278(8): 6482-6489.

Addona, T. A., S. E. Abbatiello, B. Schilling, S. J. Skates, D. R. Mani, D. M. Bunk, et al. (2009). "Multi-site assessment of the precision and reproducibility of multiple reaction monitoring-based measurements of proteins in plasma." Nature Biotechnology 27(7): 633-641.

Adkins, J. N., S. M. Varnum, K. J. Auberry, R. J. Moore, N. H. Angell, R. D. Smith, et al. (2002). "Toward a human blood serum proteome: analysis by multidimensional separation coupled with mass spectrometry." Mol Cell Proteomics 1(12): 947-955.

Aebersold, R., L. Anderson, R. Caprioli, B. Druker, L. Hartwell and R. Smith (2005). "Perspective: A Program to Improve Protein Biomarker Discovery for Cancer." J. Proteome Res. 4(4): 1104-1109.

297

Aebersold, R. and M. Mann (2003). "Mass spectrometry-based proteomics." Nature 422(6928): 198-207.

Aebersold, R., M. Mann, R. Aebersold and M. Mann (2003). "Mass spectrometry-based proteomics." Nature 422(6928): 198-207.

Ai, J., L. J. Druhan, M. G. Hunter, M. J. Loveland and B. R. Avalos (2008). "LRG- accelerated differentiation defines unique G-CSFR signaling pathways downstream of PU.1 and C/EBPepsilon that modulate neutrophil activation." J Leukoc Biol 83(5): 1277- 1285.

AIHW (2012). "Australian Institute of Health and Welfare & Cancer Australia 2012. Breast cancer in Australia: an overview." Cancer series no. 71. Cat no. CAN 67. Canberra.

AIHW (2013). "Cancer in Australia: Actual incidence data from 1991 to 2009 and mortality data from 1991 to 2010 with projections to 2012." Asia Pac J Clin Oncol 9(3): 199-213.

AIHW (2015). "Cancer in Australia 2014: actual incidence data from 1982 to 2011 and mortality data from 1982 to 2012 with projections to 2014." Asia Pac J Clin Oncol 11(3): 208-220.

Alexander, H., A. L. Stegner, C. Wagner-Mann, G. C. Du Bois, S. Alexander and E. R. Sauter (2004). "Proteomic analysis to identify breast cancer biomarkers in nipple aspirate fluid." Clinical Cancer Research 10(22): 7500-7510.

Alsner, J., M. Yilmaz, P. Guldberg, L. L. Hansen and J. Overgaard (2000). "Heterogeneity in the clinical phenotype of TP53 mutations in breast cancer patients." Clin Cancer Res 6(10): 3923-3931.

Alves, G., D. A. Pereira, V. Sandim, A. A. Ornellas, N. Escher, C. Melle, et al. (2013). "Urine screening by Seldi-Tof, followed by biomarker identification, in a Brazilian cohort of patients with renal cell carcinoma (RCC)." Int Braz J Urol 39(2): 228-239.

Ambrosino, C., R. Tarallo, A. Bamundo, D. Cuomo, G. Franci, G. Nassa, et al. (2010). "Identification of a hormone-regulated dynamic nuclear actin network associated with estrogen receptor alpha in human breast cancer cell nuclei." Molecular & cellular proteomics : MCP 9(6): 1352-1367.

Ammar, H. and J. L. Closset (2008). "Clusterin activates survival through the phosphatidylinositol 3-kinase/Akt pathway." J Biol Chem 283(19): 12851-12861.

298

Amon, L. M., S. J. Pitteri, C. I. Li, M. McIntosh, J. J. Ladd, M. Disis, et al. (2012). "Concordant release of glycolysis proteins into the plasma preceding a diagnosis of ER+ breast cancer." Cancer Res 72(8): 1935-1942.

Andersen, J. D., K. L. Boylan, R. Jemmerson, M. A. Geller, B. Misemer, K. M. Harrington, et al. (2010). "Leucine-rich alpha-2-glycoprotein-1 is upregulated in sera and tumors of ovarian cancer patients." J Ovarian Res 3: 21.

Anderson, L. (2005). "Candidate-based proteomics in the search for biomarkers of cardiovascular disease." J. Physiol 563(1): 23-60.

Anderson, L. and C. L. Hunter (2006). "Quantitative Mass Spectrometric Multiple Reaction Monitoring Assays for Major Plasma Proteins." Mol Cell Proteomics 5(4): 573- 588.

Anderson, L. and C. L. Hunter (2006). "Quantitative mass spectrometric multiple reaction monitoring assays for major plasma proteins." Molecular & Cellular Proteomics 5(4): 573-588.

Anderson, N. L. and N. G. Anderson (2002). "The human plasma proteome: history, character, and diagnostic prospects." Molecular & Cellular Proteomics 1(11): 845-867.

Anderson, N. L., M. Polanski, R. Pieper, T. Gatlin, R. S. Tirumalai, T. P. Conrads, et al. (2004). "The human plasma proteome: a nonredundant list developed by combination of four separate sources." Mol Cell Proteomics 3(4): 311-326.

Antalis, C. J., A. Uchida, K. K. Buhman and R. A. Siddiqui (2011). "Migration of MDA- MB-231 breast cancer cells depends on the availability of exogenous lipids and cholesterol esterification." Clin Exp Metastasis 28(8): 733-741.

Antman, K. and S. Shea (1999). "Screening mammography under age 50." JAMA 281(16): 1470-1472.

Antoniou, A., P. D. Pharoah, S. Narod, H. A. Risch, J. E. Eyfjord, J. L. Hopper, et al. (2003). "Average risks of breast and ovarian cancer associated with BRCA1 or BRCA2 mutations detected in case Series unselected for family history: a combined analysis of 22 studies." Am J Hum Genet 72(5): 1117-1130.

Arslan, N., M. Serdar, S. Deveci, B. Ozturk, Y. Narin, S. Ilgan, et al. (2000). "Use of CA15-3, CEA and prolactin for the primary diagnosis of breast cancer and correlation

299

with the prognostic factors at the time of initial diagnosis." Ann Nucl Med 14(5): 395- 399.

Arvold, N. D., A. G. Taghian, A. Niemierko, R. F. Abi Raad, M. Sreedhara, P. L. Nguyen, et al. (2011). "Age, breast cancer subtype approximation, and local recurrence after breast-conserving therapy." J Clin Oncol 29(29): 3885-3891.

Atkinson, A. J., W. A. Colburn, V. G. DeGruttola, D. L. DeMets, G. J. Downing, D. F. Hoth, et al. (2001). "Biomarkers and surrogate endpoints: Preferred definitions and conceptual framework*." Clin Pharmacol Ther 69(3): 89-95.

Avgeris, M., G. Papachristopoulou, A. Polychronis and A. Scorilas (2011). "Down- regulation of kallikrein-related peptidase 5 (KLK5) expression in breast cancer patients: a biomarker for the differential diagnosis of breast lesions." Clin Proteomics 8(1): 5.

Bai, X., E. Zhang, H. Ye, V. Nandakumar, Z. Wang, L. Chen, et al. (2014). "PIK3CA and TP53 gene mutations in human breast cancer tumors frequently detected by ion torrent DNA sequencing." PLoS One 9(6): e99306.

Bando, H., G. Matsumoto, M. Bando, M. Muta, T. Ogawa, N. Funata, et al. (2002). "Expression of macrophage migration inhibitory factor in human breast cancer: association with nodal spread." Jpn J Cancer Res 93(4): 389-396.

Banks, R. E., A. J. Stanley, D. A. Cairns, J. H. Barrett, P. Clarke, D. Thompson, et al. (2005). "Influences of blood sample processing on low-molecular-weight proteome identified by surface-enhanced laser desorption/ionization mass spectrometry." Clin Chem 51(9): 1637-1649.

Barrow, T. M. and K. B. Michels (2014). "Epigenetic epidemiology of cancer." Biochem Biophys Res Commun.

Basith, S., B. Manavalan, T. H. Yoo, S. G. Kim and S. Choi (2012). "Roles of toll-like receptors in cancer: a double-edged sword for defense and offense." Arch Pharm Res 35(8): 1297-1316.

Bauer, J. A., A. B. Chakravarthy, J. M. Rosenbluth, D. Mi, E. H. Seeley, N. De Matos Granja-Ingram, et al. (2010). "Identification of markers of taxane sensitivity using proteomic and genomic analyses of breast tumors from patients receiving neoadjuvant paclitaxel and radiation." Clin Cancer Res 16(2): 681-690.

300

Baxter, R. C. (2013). "Insulin-like growth factor binding protein-3 (IGFBP-3): Novel ligands mediate unexpected functions." J Cell Commun Signal 7(3): 179-189.

Baxter, R. C. (2014). "IGF binding proteins in cancer: mechanistic and clinical insights." Nat Rev Cancer 14(5): 329-341.

Becker, S., L. H. Cazares, P. Watson, H. Lynch, O. J. Semmes, R. R. Drake, et al. (2004). "Surfaced-enhanced laser desorption/ionization time-of-flight (SELDI-TOF) differentiation of serum protein profiles of BRCA-1 and sporadic breast cancer." Ann Surg Oncol 11(10): 907-914.

Beebe-Dimmer, J. L., C. Yee, M. L. Cote, N. Petrucelli, N. Palmer, C. Bock, et al. (2015). "Familial clustering of breast and prostate cancer and risk of postmenopausal breast cancer in the Women's Health Initiative Study." Cancer 121(8): 1265-1272.

Bellahcene, A. and V. Castronovo (1995). "Increased expression of osteonectin and osteopontin, two bone matrix proteins, in human breast cancer." Am J Pathol 146(1): 95- 100.

Belluco, C., E. F. Petricoin, E. Mammano, F. Facchiano, S. Ross-Rucker, D. Nitti, et al. (2007). "Serum proteomic analysis identifies a highly sensitive and specific discriminatory pattern in stage 1 breast cancer." Ann Surg Oncol 14(9): 2470-2476.

Benoy, I., R. Salgado, C. Colpaert, R. Weytjens, P. B. Vermeulen and L. Y. Dirix (2002). "Serum interleukin 6, plasma VEGF, serum VEGF, and VEGF platelet load in breast cancer patients." Clin Breast Cancer 2(4): 311-315.

Benson, S. R., J. Blue, K. Judd and J. E. Harman (2004). "Ultrasound is now better than mammography for the detection of invasive breast cancer." Am J Surg 188(4): 381-385.

Beretov, J., V. C. Wasinger, P. H. Graham, E. K. Millar, J. H. Kearsley and Y. Li (2014). "Proteomics for breast cancer urine biomarkers." Adv Clin Chem 63: 123-167.

Beretov, J., V. C. Wasinger, E. K. Millar, P. Schwartz, P. H. Graham and Y. Li (2015). "Proteomic Analysis of Urine to Identify Breast Cancer Biomarker Candidates Using a Label-Free LC-MS/MS Approach." PLoS One 10(11): e0141876.

Beretov, J., V. C. Wasinger, P. Schwartz, P. H. Graham and Y. Li (2014). "A standardized and reproducible urine preparation protocol for cancer biomarkers discovery." Biomark Cancer 6: 21-27.

301

Bertucci, F., D. Birnbaum and A. Goncalves (2006). "Proteomics of breast cancer: principles and potential clinical applications." Mol Cell Proteomics 5(10): 1772-1786.

Bewick, V., L. Cheek and J. Ball (2004). "Statistics review 12: survival analysis." Crit Care 8(5): 389-394.

Bharti, A., P. C. Ma, G. Maulik, R. Singh, E. Khan, A. T. Skarin, et al. (2004). "Haptoglobin alpha-subunit and hepatocyte growth factor can potentially serve as serum tumor biomarkers in small cell lung cancer." Anticancer Res 24(2c): 1031-1038.

Bijker, N., P. Meijnen, J. L. Peterse, J. Bogaerts, I. Van Hoorebeeck, J. P. Julien, et al. (2006). "Breast-conserving treatment with or without radiotherapy in ductal carcinoma- in-situ: ten-year results of European Organisation for Research and Treatment of Cancer randomized phase III trial 10853--a study by the EORTC Breast Cancer Cooperative Group and EORTC Radiotherapy Group." J Clin Oncol 24(21): 3381-3387.

Bjorhall, K., T. Miliotis and P. Davidsson (2005). "Comparison of different depletion strategies for improved resolution in proteomic analysis of human serum samples." Proteomics 5(1): 307-317.

Black, M. H. and E. P. Diamandis (2000). "The diagnostic and prognostic utility of prostate-specific antigen for diseases of the breast." Breast Cancer Res Treat 59(1): 1-14.

Blows, F. M., K. E. Driver, M. K. Schmidt, A. Broeks, F. E. van Leeuwen, J. Wesseling, et al. (2010). "Subtyping of breast cancer by immunohistochemistry to investigate a relationship between subtype and short and long term survival: a collaborative analysis of data for 10,159 cases from 12 studies." PLoS Med 7(5): e1000279.

Bohm, D., K. Keller, J. Pieter, N. Boehm, D. Wolters, W. Siggelkow, et al. (2012). "Comparison of tear protein levels in breast cancer patients and healthy controls using a de novo proteomic approach." Oncol Rep 28(2): 429-438.

Bons, J. A., W. K. Wodzig and M. P. van Dieijen-Visser (2005). "Protein profiling as a diagnostic tool in clinical chemistry: a review." Clin Chem Lab Med 43(12): 1281-1290.

Borstnar, S., I. Vrhovec, B. Svetic and T. Cufer (2002). "Prognostic value of the urokinase-type plasminogen activator, and its inhibitors and receptor in breast cancer patients." Clin Breast Cancer 3(2): 138-146.

Bouchal, P., T. Roumeliotis, R. Hrstka, R. Nenutil, B. Vojtesek and S. D. Garbis (2009). "Biomarker discovery in low-grade breast cancer using isobaric stable isotope tags and

302

two-dimensional liquid chromatography-tandem mass spectrometry (iTRAQ-2DLC- MS/MS) based quantitative proteomic analysis." Journal of Proteome Research 8(1): 362- 373.

Boyanton, B. L., Jr. and K. E. Blick (2002). "Stability studies of twenty-four analytes in human plasma and serum." Clin Chem 48(12): 2242-2247.

Boyd, N. F., L. J. Martin, J. M. Rommens, A. D. Paterson, S. Minkin, M. J. Yaffe, et al. (2009). "Mammographic density: a heritable risk factor for breast cancer." Methods Mol Biol 472: 343-360.

Boyer, A. P., T. S. Collier, I. Vidavsky and R. Bose (2013). "Quantitative proteomics with siRNA screening identifies novel mechanisms of trastuzumab resistance in HER2 amplified breast cancers." Mol Cell Proteomics 12(1): 180-193.

Bramwell, V. H., G. S. Doig, A. B. Tuck, S. M. Wilson, K. S. Tonkin, A. Tomiak, et al. (2006). "Serial plasma osteopontin levels have prognostic value in metastatic breast cancer." Clin Cancer Res 12(11 Pt 1): 3337-3343.

Braoudaki, M., G. I. Lambrou, K. Vougas, K. Karamolegou, G. T. Tsangaris and F. Tzortzatou-Stathopoulou (2013). "Protein biomarkers distinguish between high- and low- risk pediatric acute lymphoblastic leukemia in a tissue specific manner." J Hematol Oncol 6: 52.

Brauer, H. A., M. D'Arcy, T. E. Libby, H. J. Thompson, Y. Y. Yasui, N. Hamajima, et al. (2014). "Dermcidin expression is associated with disease progression and survival among breast cancer patients." Breast Cancer Res Treat 144(2): 299-306.

Braun, M. C., L. Li, B. Ke, W. P. Dubinsky, M. C. Pickering and J. Y. Chang (2006). "Proteomic profiling of urinary protein excretion in the factor H-deficient mouse." Am J Nephrol 26(2): 127-135.

Bray, F., J. S. Ren, E. Masuyer and J. Ferlay (2013). "Global estimates of cancer prevalence for 27 sites in the adult population in 2008." Int J Cancer 132(5): 1133-1145.

Brennan, K., M. Garcia-Closas, N. Orr, O. Fletcher, M. Jones, A. Ashworth, et al. (2012). "Intragenic ATM methylation in peripheral blood DNA as a biomarker of breast cancer risk." Cancer Res 72(9): 2304-2313.

Brewis, I. A. and P. Brennan (2010). "Proteomics technologies for the global identification and quantification of proteins." Adv Protein Chem Struct Biol 80: 1-44.

303

Brooks, J. D., P. Cairns, R. E. Shore, C. B. Klein, I. Wirgin, Y. Afanasyeva, et al. (2010). "DNA methylation in pre-diagnostic serum samples of breast cancer cases: results of a nested case-control study." Cancer Epidemiol 34(6): 717-723.

Brunoro, G. V., A. T. Ferreira, M. R. Trugilho, T. S. Oliveira, L. C. Amendola, J. Perales, et al. (2014). "Potential correlation between tumor aggressiveness and protein expression patterns of nipple aspirate fluid (NAF) revealed by gel-based proteomic analysis." Curr Top Med Chem 14(3): 359-368.

Bryan, R. T., W. Wei, N. J. Shimwell, S. I. Collins, S. A. Hussain, L. J. Billingham, et al. (2011). "Assessment of high-throughput high-resolution MALDI-TOF-MS of urinary peptides for the detection of muscle-invasive bladder cancer." Proteomics Clinical Applications 5(9-10): 493-503.

Buist, D. S., P. L. Porter, C. Lehman, S. H. Taplin and E. White (2004). "Factors contributing to mammography failure in women aged 40-49 years." J Natl Cancer Inst 96(19): 1432-1440.

Byun, J. A., S. H. Lee, B. H. Jung, M. H. Choi, M. H. Moon and B. C. Chung (2008). "Analysis of polyamines as carbamoyl derivatives in urine and serum by liquid chromatography-tandem mass spectrometry." Biomed Chromatogr 22(1): 73-80.

Campeau, P. M., W. D. Foulkes and M. D. Tischkowitz (2008). "Hereditary breast cancer: new genetic developments, new therapeutic avenues." Hum Genet 124(1): 31-42.

Cancemi, P., G. Di Cara, N. N. Albanese, F. Costantini, M. R. Marabeti, R. Musso, et al. (2010). "Large-scale proteomic identification of S100 proteins in breast cancer tissues." BMC Cancer 10: 476.

Cancer, C. G. o. H. F. i. B. (2002). "Breast cancer and breastfeeding: collaborative reanalysis of individual data from 47 epidemiological studies in 30 countries, including 50302 women with breast cancer and 96973 women without the disease." Lancet 360(9328): 187-195.

Candiano, G., L. Santucci, A. Petretto, M. Bruschi, V. Dimuccio, A. Urbani, et al. (2010). "2D-electrophoresis and the urine proteome map: where do we stand?" Journal of Proteomics 73(5): 829-844.

Cao, Y., Y. Li, M. Edelweiss, B. Arun, D. Rosen, E. Resetkova, et al. (2008). "Loss of annexin A1 expression in breast cancer progression." Appl Immunohistochem Mol Morphol 16(6): 530-534.

304

Cao, Y., S. Lindstrom, F. Schumacher, V. L. Stevens, D. Albanes, S. Berndt, et al. (2014). "Insulin-like growth factor pathway genetic polymorphisms, circulating IGF1 and IGFBP3, and prostate cancer survival." J Natl Cancer Inst 106(6): dju085.

Caprioli, R. M., T. B. Farmer and J. Gile (1997). "Molecular imaging of biological samples: localization of peptides and proteins using MALDI-TOF MS." Anal Chem 69(23): 4751-4760.

Carey, L. A., C. M. Perou, C. A. Livasy, L. G. Dressler, D. Cowan, K. Conway, et al. (2006). "Race, breast cancer subtypes, and survival in the Carolina Breast Cancer Study." Jama 295(21): 2492-2502.

Carter, D., J. F. Douglass, C. D. Cornellison, M. W. Retter, J. C. Johnson, A. A. Bennington, et al. (2002). "Purification and characterization of the mammaglobin/lipophilin B complex, a promising diagnostic marker for breast cancer." Biochemistry 41(21): 6714-6722.

Carter, S. J., X. F. Li, J. R. Mackey, S. Modi, J. Hanson and N. J. Dovichi (2001). "Biomonitoring of urinary tamoxifen and its metabolites from breast cancer patients using nonaqueous capillary electrophoresis with electrospray mass spectrometry." Electrophoresis 22(13): 2730-2736.

Caruso, J. A. and P. M. Stemmer (2011). "Proteomic profiling of lipid rafts in a human breast cancer model of tumorigenic progression." Clin Exp Metastasis 28(6): 529-540.

Casadonte, R., M. Kriegsmann, F. Zweynert, K. Friedrich, G. Baretton, M. Otto, et al. (2014). "Imaging mass spectrometry to discriminate breast from pancreatic cancer metastasis in formalin-fixed paraffin-embedded tissues." Proteomics 14(7-8): 956-964.

Castagna, A., D. Cecconi, L. Sennels, J. Rappsilber, L. Guerrier, F. Fortis, et al. (2005). "Exploring the hidden human urinary proteome via ligand library beads." Journal of Proteome Research 4(6): 1917-1930.

Catterall, J. B., A. D. Rowan, S. Sarsfield, J. Saklatvala, R. Wait and T. E. Cawston (2006). "Development of a novel 2D proteomics approach for the identification of proteins secreted by primary chondrocytes after stimulation by IL-1 and oncostatin M." Rheumatology 45(9): 1101-1109.

Celis, J. E., J. M. Moreira, I. Gromova, T. Cabezon, U. Ralfkiaer, P. Guldberg, et al. (2005). "Towards discovery-driven translational research in breast cancer." FEBS J 272(1): 2-15.

305

Chang, K. P., C. C. Wu, H. C. Chen, S. J. Chen, P. H. Peng, N. M. Tsang, et al. (2010). "Identification of candidate nasopharyngeal carcinoma serum biomarkers by cancer cell secretome and tissue transcriptome analysis: potential usage of cystatin A for predicting nodal stage and poor prognosis." Proteomics 10(14): 2644-2660.

Chao, T., J. J. Ladd, J. Qiu, M. M. Johnson, R. Israel, A. Chin, et al. (2013). "Proteomic profiling of the autoimmune response to breast cancer antigens uncovers a suppressive effect of hormone therapy." Proteomics Clin Appl 7(5-6): 327-336.

Chaudhary, N., S. Bhatnagar, S. Malik, D. P. Katare and S. K. Jain (2013). "Proteomic analysis of differentially expressed proteins in lung cancer in Wistar rats using NNK as an inducer." Chem Biol Interact 204(2): 125-134.

Chelius, D. and P. V. Bondarenko (2002). "Quantitative profiling of proteins in complex mixtures using liquid chromatography and mass spectrometry." J Proteome Res 1(4): 317- 323.

Chen, C. L., T. S. Lin, C. H. Tsai, C. C. Wu, T. Chung, K. Y. Chien, et al. (2013). "Identification of potential bladder cancer markers in urine by abundant-protein depletion coupled with quantitative proteomics." J Proteomics 85: 28-43.

Chen, E. I. and J. R. Yates, 3rd (2007). "Cancer proteomics by quantitative shotgun proteomics." Molecular Oncology 1(2): 144-159.

Chen, J., X. Cheng, M. Merched-Sauvage, C. Caulin, D. R. Roop and P. J. Koch (2006). "An unexpected role for keratin 10 end domains in susceptibility to skin cancer." J Cell Sci 119(Pt 24): 5067-5076.

Chen, Y.-T., C.-L. Chen, H.-W. Chen, T. Chung, C.-C. Wu, C.-D. Chen, et al. (2010). "Discovery of novel bladder cancer biomarkers by comparative urine proteomics using iTRAQ technology." Journal of Proteome Research 9(11): 5803-5815.

Cheng, J., S. Qiu, U. Raju, S. R. Wolman and M. J. Worsham (2008). "Benign breast disease heterogeneity: association with histopathology, age, and ethnicity." Breast Cancer Res Treat 111(2): 289-296.

Cho, D. H., Y. K. Jo, S. A. Roh, Y. S. Na, T. W. Kim, S. J. Jang, et al. (2010). "Upregulation of SPRR3 promotes colorectal tumorigenesis." Mol Med 16(7-8): 271- 277.

306

Cho, S. H., B. H. Jung, S. H. Lee, W. Y. Lee, G. Kong and B. C. Chung (2006). "Direct determination of nucleosides in the urine of patients with breast cancer using column- switching liquid chromatography-tandem mass spectrometry." Biomed Chromatogr 20(11): 1229-1236.

Cho, W., K. Jung and F. E. Regnier (2010). "Sialylated Lewis x antigen bearing glycoproteins in human plasma." J Proteome Res 9(11): 5960-5968.

Choe, L. H. and K. H. Lee (2003). "Quantitative and qualitative measure of intralaboratory two-dimensional protein gel reproducibility and the effects of sample preparation, sample load, and image analysis." Electrophoresis 24(19-20): 3500-3507.

Choi, C. I., H. I. Lee, J. W. Bae, Y. J. Lee, J. Y. Byeon, C. G. Jang, et al. (2012). "Determination of tamsulosin in human plasma by liquid chromatography/tandem mass spectrometry and its application to a pharmacokinetic study." J Chromatogr B Analyt Technol Biomed Life Sci 909: 65-69.

Choi, Y. J., G. M. Park, J. K. Rho, S. Y. Kim, G. S. So, H. R. Kim, et al. (2013). "Role of IGF-binding protein 3 in the resistance of EGFR mutant lung cancer cells to EGFR- tyrosine kinase inhibitors." PLoS One 8(12): e81393.

Choong, L. Y., S. Lim, P. K. Chong, C. Y. Wong, N. Shah and Y. P. Lim (2010). "Proteome-wide profiling of the MCF10AT breast cancer progression model." PLoS One 5(6): e11030.

Chung, L. and R. C. Baxter (2012). "Breast cancer biomarkers: proteomic discovery and translation to clinically relevant assays." Expert Rev Proteomics 9(6): 599-614.

Chung, L., K. Moore, L. Phillips, F. M. Boyle, D. J. Marsh and R. C. Baxter (2014). "Novel serum protein biomarker panel revealed by mass spectrometry and its prognostic value in breast cancer." Breast Cancer Res 16(3): R63.

Chung, L., S. Shibli, K. Moore, E. E. Elder, F. M. Boyle, D. J. Marsh, et al. (2013). "Tissue biomarkers of breast cancer and their association with conventional pathologic features." Br J Cancer 108(2): 351-360.

Clement, C. C., D. Aphkhazava, E. Nieves, M. Callaway, W. Olszewski, O. Rotzschke, et al. (2013). "Protein expression profiles of human lymph and plasma mapped by 2D- DIGE and 1D SDS-PAGE coupled with nanoLC-ESI-MS/MS bottom-up proteomics." J Proteomics 78: 172-187.

307

Coffelt, S. B. and K. E. de Visser (2015). "Immune-mediated mechanisms influencing the efficacy of anticancer therapies." Trends Immunol 36(4): 198-216.

Coffelt, S. B., K. Kersten, C. W. Doornebal, J. Weiden, K. Vrijland, C. S. Hau, et al. (2015). "IL-17-producing gammadelta T cells and neutrophils conspire to promote breast cancer metastasis." Nature.

Cohen, A., E. Wang, K. A. Chisholm, R. Kostyleva, M. O'Connor-McCourt and D. M. Pinto (2013). "A mass spectrometry-based plasma protein panel targeting the tumor microenvironment in patients with breast cancer." J Proteomics 81: 135-147.

Cook, A. C., A. B. Tuck, S. McCarthy, J. G. Turner, R. B. Irby, G. C. Bloom, et al. (2005). "Osteopontin induces multiple changes in gene expression that reflect the six "hallmarks of cancer" in a model of breast cancer progression." Mol Carcinog 43(4): 225-236.

Coon, J. J., P. Zurbig, M. Dakna, A. F. Dominiczak, S. Decramer, D. Fliser, et al. (2008). "CE-MS analysis of the human urinary proteome for biomarker discovery and disease diagnostics." Proteomics Clinical Applications 2(7-8): 964.

Corben, A. D. (2013). "Pathology of invasive breast disease." Surg Clin North Am 93(2): 363-392.

Coumans, J. V., D. Gau, A. Poljak, V. Wasinger, P. Roy and P. Moens (2014). "Green fluorescent protein expression triggers proteome changes in breast cancer cells." Exp Cell Res 320(1): 33-45.

Court, M., N. Selevsek, M. Matondo, Y. Allory, J. Garin, C. D. Masselon, et al. (2011). "Toward a standardized urine proteome analysis methodology." Proteomics 11(6): 1160- 1171.

Cravatt, B. F., G. M. Simon and J. R. Yates, 3rd (2007). "The biological impact of mass- spectrometry-based proteomics." Nature 450(7172): 991-1000.

Crosley, L. K., S. J. Duthie, A. C. Polley, F. G. Bouwman, C. Heim, F. Mulholland, et al. (2009). "Variation in protein levels obtained from human blood cells and biofluids for platelet, peripheral blood mononuclear cell, plasma, urine and saliva proteomics." Genes Nutr 4(2): 95-102.

Custodio, A., A. J. Lopez-Farre, J. J. Zamorano-Leon, P. J. Mateos-Caceres, C. Macaya, T. Caldes, et al. (2012). "Changes in the expression of plasma proteins associated with thrombosis in BRCA1 mutation carriers." J Cancer Res Clin Oncol 138(5): 867-875.

308

Cutillas, P. R. (2010). "Analysis of peptides in biological fluids by LC-MS/MS." Methods Mol Biol 658: 311-321.

Darby, S., P. McGale, C. Correa, C. Taylor, R. Arriagada, M. Clarke, et al. (2011). "Effect of radiotherapy after breast-conserving surgery on 10-year recurrence and 15-year breast cancer death: meta-analysis of individual patient data for 10,801 women in 17 randomised trials." Lancet 378(9804): 1707-1716.

Davalieva, K., S. Kiprijanovska, C. Broussard, G. Petrusevska and G. D. Efremov (2012). "Proteomic analysis of infiltrating ductal carcinoma tissues by coupled 2-D DIGE/MS/MS analysis." Mol Biol (Mosk) 46(3): 469-480.

Davis, M. T., C. S. Spahr, M. D. McGinley, J. H. Robinson, E. J. Bures, J. Beierle, et al. (2001). "Towards defining the urinary proteome using liquid chromatography-tandem mass spectrometry. II. Limitations of complex mixture analyses." Proteomics 1(1): 108- 117.

De Bock, M., D. de Seny, M. A. Meuwis, J. P. Chapelle, E. Louis, M. Malaise, et al. (2010). "Challenges for Biomarker Discovery in Body Fluids Using SELDI-TOF-MS." Journal of Biomedicine and Biotechnology. de Noo, M. E., R. A. Tollenaar, A. Ozalp, P. J. Kuppen, M. R. Bladergroen, P. H. Eilers, et al. (2005). "Reliability of human serum protein profiles generated with C8 magnetic beads assisted MALDI-TOF mass spectrometry." Anal Chem 77(22): 7232-7241.

DeAngelis, J. T., Y. Li, N. Mitchell, L. Wilson, H. Kim and T. O. Tollefsbol (2011). "2D difference gel electrophoresis analysis of different time points during the course of neoplastic transformation of human mammary epithelial cells." Journal of Proteome Research 10(2): 447-458.

Decramer, S., A. Gonzalez de Peredo, B. Breuil, H. Mischak, B. Monsarrat, J.-L. Bascands, et al. (2008). "Urine in clinical proteomics." Molecular & Cellular Proteomics 7(10): 1850-1862.

Dekker, L. J., W. Boogerd, G. Stockhammer, J. C. Dalebout, I. Siccama, P. Zheng, et al. (2005). "MALDI-TOF mass spectrometry analysis of cerebrospinal fluid tryptic peptide profiles to diagnose leptomeningeal metastases in patients with breast cancer." Mol Cell Proteomics 4(9): 1341-1349.

Demark-Wahnefried, W., E. A. Platz, J. A. Ligibel, C. K. Blair, K. S. Courneya, J. A. Meyerhardt, et al. (2012). "The role of obesity in cancer survival and recurrence." Cancer Epidemiol Biomarkers Prev 21(8): 1244-1259. 309

Deng, M., W. Zhang, H. Tang, Q. Ye, Q. Liao, Y. Zhou, et al. (2013). "Lactotransferrin acts as a tumor suppressor in nasopharyngeal carcinoma by repressing AKT through multiple mechanisms." Oncogene 32(36): 4273-4283.

Deng, Q., Q. Wang, W. Y. Zong, D. L. Zheng, Y. X. Wen, K. S. Wang, et al. (2010). "E2F8 contributes to human hepatocellular carcinoma via regulating cell proliferation." Cancer Res 70(2): 782-791.

Deng, S., H. Zhou, R. Xiong, Y. Lu, D. Yan, T. Xing, et al. (2007). "Over-expression of genes and proteins of ubiquitin specific peptidases (USPs) and proteasome subunits (PSs) in breast cancer tissue observed by the methods of RFDD-PCR and proteomics." Breast Cancer Research & Treatment 104(1): 21-30.

Descamps, S., X. Lebourhis, M. Delehedde, B. Boilly and H. Hondermarck (1998). "Nerve growth factor is mitogenic for cancerous but not normal human breast epithelial cells." J Biol Chem 273(27): 16659-16662.

Dirix, L. Y., R. Salgado, R. Weytjens, C. Colpaert, I. Benoy, P. Huget, et al. (2002). "Plasma fibrin D-dimer levels correlate with tumour volume, progression rate and survival in patients with metastatic breast cancer." Br J Cancer 86(3): 389-395.

Dolle, L., E. Adriaenssens, I. El Yazidi-Belkoura, X. Le Bourhis, V. Nurcombe and H. Hondermarck (2004). "Nerve growth factor receptors and signaling in breast cancer." Curr Cancer Drug Targets 4(6): 463-470.

Domanski, D., L. C. Murphy and C. H. Borchers (2010). "Assay development for the determination of phosphorylation stoichiometry using multiple reaction monitoring methods with and without phosphatase treatment: application to breast cancer signaling pathways." Anal Chem 82(13): 5610-5620.

Domon, B. and R. Aebersold (2006). "Mass spectrometry and protein analysis." Science 312(5771): 212-217. dos Santos, C. R., G. Domingues, I. Matias, J. Matos, I. Fonseca, J. M. de Almeida, et al. (2014). "LDL-cholesterol signaling induces breast cancer proliferation and invasion." Lipids Health Dis 13: 16.

Dowling, P., C. Clarke, K. Hennessy, B. Torralbo-Lopez, J. Ballot, J. Crown, et al. (2012). "Analysis of acute-phase proteins, AHSG, C3, CLI, HP and SAA, reveals distinctive expression patterns associated with breast, colorectal and lung cancer." Int J Cancer 131(4): 911-923.

310

Dowsett, M., T. O. Nielsen, R. A'Hern, J. Bartlett, R. C. Coombes, J. Cuzick, et al. (2011). "Assessment of Ki67 in breast cancer: recommendations from the International Ki67 in Breast Cancer working group." J Natl Cancer Inst 103(22): 1656-1664.

Drake, R. R., L. Cazares and O. J. Semmes (2007). "Mining the low molecular weight proteome of blood." Proteomics Clin Appl 1(8): 758-768.

Drake, R. R., L. H. Cazares, E. E. Jones, T. W. Fuller, O. J. Semmes and C. Laronga (2011). "Challenges to developing proteomic-based breast cancer diagnostics." Omics 15(5): 251-259.

Drake, S. K., R. A. R. Bowen, A. T. Remaley and G. L. Hortin (2004). "Potential Interferences from Blood Collection Tubes in Mass Spectrometric Analyses of Serum Polypeptides." Clin Chem 50(12): 2398-2401.

Duffy, M. J. (2006). "Serum tumor markers in breast cancer: are they of clinical value?" Clin Chem 52(3): 345-351.

Duffy, M. J. and C. Duggan (2004). "The urokinase plasminogen activator system: a rich source of tumour markers for the individualised management of patients with cancer." Clin Biochem 37(7): 541-548.

Duffy, M. J., C. Duggan, R. Keane, A. D. K. Hill, E. McDermott, J. Crown, et al. (2004). "High preoperative CA 15-3 concentrations predict adverse outcome in node-negative and node-positive breast cancer: study of 600 patients with histologically confirmed breast cancer.[Erratum appears in Clin Chem. 2004 Jun;50(6):1111]." Clinical Chemistry 50(3): 559-563.

Duncan, M. W., R. Aebersold and R. M. Caprioli (2010). "The pros and cons of peptide- centric proteomics." Nat Biotechnol 28(7): 659-664.

Duru, N., M. Fan, D. Candas, C. Menaa, H. C. Liu, D. Nantajit, et al. (2012). "HER2- associated radioresistance of breast cancer stem cells isolated from HER2-negative breast cancer cells." Clin Cancer Res 18(24): 6634-6647.

Echan, L. A., H. Y. Tang, N. Ali-Khan, K. Lee and D. W. Speicher (2005). "Depletion of multiple high-abundance proteins improves protein profiling capacities of human serum and plasma." Proteomics 5(13): 3292-3303.

311

Edge, S. B. and C. C. Compton (2010). "The American Joint Committee on Cancer: the 7th edition of the AJCC cancer staging manual and the future of TNM." Ann Surg Oncol 17(6): 1471-1474.

Esposito, I., H. Kayed, S. Keleg, T. Giese, E. H. Sage, P. Schirmacher, et al. (2007). "Tumor-suppressor function of SPARC-like protein 1/Hevin in pancreatic cancer." Neoplasia 9(1): 8-17.

Esserman, L., Y. Shieh and I. Thompson (2009). "Rethinking screening for breast cancer and prostate cancer." JAMA 302(15): 1685-1692.

Esserman, L. J., Y. Shieh, E. J. Rutgers, M. Knauer, V. P. Retel, S. Mook, et al. (2011). "Impact of mammographic screening on the detection of good and poor prognosis breast cancers." Breast Cancer Res Treat 130(3): 725-734.

Evans, V., C. Vockler, M. Friedlander, B. Walsh and M. D. Willcox (2001). "Lacryglobin in human tears, a potential marker for cancer." Clinical & Experimental Ophthalmology 29(3): 161-163.

Fabre-Lafay, S., F. Monville, S. Garrido-Urbani, C. Berruyer-Pouyet, C. Ginestier, N. Reymond, et al. (2007). "Nectin-4 is a new histological and serological tumor associated marker for breast cancer." BMC Cancer 7: 73.

Fan, N. J., C. F. Gao, C. S. Wang, G. Zhao, J. J. Lv, X. L. Wang, et al. (2012). "Identification of the up-regulation of TP-alpha, collagen alpha-1(VI) chain, and S100A9 in esophageal squamous cell carcinoma by a proteomic method." J Proteomics 75(13): 3977-3986.

Fan, Y., J. Wang, Y. Yang, Q. Liu, Y. Fan, J. Yu, et al. (2010). "Detection and identification of potential biomarkers of breast cancer." J Cancer Res Clin Oncol 136(8): 1243-1254.

Fang, Z. G., B. G. You, Y. G. Chen, J. K. Zhang, Y. Q. Liu, X. N. Zhang, et al. (2010). "Analysis of cyclosporine A and its metabolites in rat urine and feces by liquid chromatography-tandem mass spectrometry." Journal of chromatography. B, Analytical technologies in the biomedical and life sciences 878(15-16): 1153-1162.

Faupel-Badger, J. M., B. J. Fuhrman, X. Xu, R. T. Falk, L. K. Keefer, T. D. Veenstra, et al. (2010). "Comparison of liquid chromatography-tandem mass spectrometry, RIA, and ELISA methods for measurement of urinary estrogens." Cancer Epidemiology, Biomarkers & Prevention 19(1): 292-300.

312

Faupel-Badger, J. M., M. E. Sherman, M. Garcia-Closas, M. M. Gaudet, R. T. Falk, A. Andaya, et al. (2010). "Prolactin serum levels and breast cancer: relationships with risk factors and tumour characteristics among pre- and postmenopausal women in a population-based case-control study from Poland." Br J Cancer 103(7): 1097-1102.

Fedarko, N. S., A. Jain, A. Karadag, M. R. Van Eman and L. W. Fisher (2001). "Elevated serum bone sialoprotein and osteopontin in colon, breast, prostate, and lung cancer." Clin Cancer Res 7(12): 4060-4066.

Fenn, J. B., M. Mann, C. K. Meng, S. F. Wong and C. M. Whitehouse (1989). "Electrospray ionization for mass spectrometry of large biomolecules." Science 246(4926): 64-71.

Ferlay J, Soerjomataram I, Ervik M, Dikshit R, Eser S, Mathers C, et al. (2012). "GLOBOCAN 2012 v1.0, Cancer Incidence and Mortality Worldwide: IARC CancerBase No. 11 [Internet]. Lyon, France: International Agency for Research on Cancer; 2013. Available from: http://globocan.iarc.fr, accessed on 13/12/2013.".

Ferlay, J., H. R. Shin, F. Bray, D. Forman, C. Mathers and D. M. Parkin (2010). "Estimates of worldwide burden of cancer in 2008: GLOBOCAN 2008." Int J Cancer 127(12): 2893-2917.

Fernandez-Grijalva, A. L., A. Aguilar-Lemarroy, L. F. Jave-Suarez, A. Gutierrez-Ortega, P. A. Godinez-Melgoza, S. E. Herrera-Rodriguez, et al. (2014). "Alpha 2HS- glycoprotein, a tumor-associated antigen (TAA) detected in Mexican patients with early- stage breast cancer." J Proteomics.

Fisch, U., A. Zehnder, A. Hirt, F. Niggli, A. Simon, H. Ozsahin, et al. (2011). "Mannan- binding lectin (MBL) and MBL-associated serine protease-2 in children with cancer." Swiss Med Wkly 141: w13191.

Fisher, B., J. Dignam, N. Wolmark, D. L. Wickerham, E. R. Fisher, E. Mamounas, et al. (1999). "Tamoxifen in treatment of intraductal breast cancer: National Surgical Adjuvant Breast and Bowel Project B-24 randomised controlled trial." Lancet 353(9169): 1993- 2000.

Fitzgibbons, P. L., D. L. Page, D. Weaver, A. D. Thor, D. C. Allred, G. M. Clark, et al. (2000). "Prognostic factors in breast cancer. College of American Pathologists Consensus Statement 1999." Archives of Pathology & Laboratory Medicine 124(7): 966-978.

313

Fleming, J. M., E. Ginsburg, S. D. Oliver, P. Goldsmith and B. K. Vonderhaar (2012). "Hornerin, an S100 family protein, is functional in breast cells and aberrantly expressed in breast cancer." BMC Cancer 12: 266.

Fliser, D., J. Novak, V. Thongboonkerd, A. Argiles, V. Jankowski, M. A. Girolami, et al. (2007). "Advances in urinary proteome analysis and biomarker discovery." Journal of the American Society of Nephrology 18(4): 1057-1071.

Flobbe, K., A. M. Bosch, A. G. Kessels, G. L. Beets, P. J. Nelemans, M. F. von Meyenfeldt, et al. (2003). "The additional diagnostic value of ultrasonography in the diagnosis of breast cancer." Arch Intern Med 163(10): 1194-1199.

Fortin, T., A. Salvador, J. P. Charrier, C. Lenz, F. Bettsworth, X. Lacoux, et al. (2009). "Multiple reaction monitoring cubed for protein quantification at the low nanogram/milliliter level in nondepleted human serum." Anal Chem 81(22): 9343-9352.

Foulkes, W. D. (2008). "Inherited susceptibility to common cancers." N Engl J Med 359(20): 2143-2153.

Fountzilas, G., U. Dafni, M. Bobos, A. Batistatou, V. Kotoula, H. Trihia, et al. (2012). "Differential response of immunohistochemically defined breast cancer subtypes to anthracycline-based adjuvant chemotherapy with or without paclitaxel." PLoS One 7(6): e37946.

Freue, G. V., M. Sasaki, A. Meredith, O. P. Gunther, A. Bergman, M. Takhar, et al. (2010). "Proteomic signatures in plasma during early acute renal allograft rejection." Mol Cell Proteomics 9(9): 1954-1967.

Frickenschmidt, A., H. Frohlich, D. Bullinger, A. Zell, S. Laufer, C. H. Gleiter, et al. (2008). "Metabonomics in cancer diagnosis: mass spectrometry-based profiling of urinary nucleosides from breast cancer patients." Biomarkers 13(4): 435-449.

Fung, E. T., T. T. Yip, L. Lomas, Z. Wang, C. Yip, X. Y. Meng, et al. (2005). "Classification of cancer types by measuring variants of host response proteins using SELDI serum assays." International Journal of Cancer 115(5): 783-789.

Fusaro, V. A., D. R. Mani, J. P. Mesirov and S. A. Carr (2009). "Prediction of high- responding peptides for targeted protein assays by mass spectrometry." Nature Biotechnology 27(2): 190-198.

314

Gagliato, D. M., D. L. Jardim, M. S. Marchesi and G. N. Hortobagyi (2016). "Mechanisms of resistance and sensitivity to anti-HER2 therapies in HER2+ breast cancer." Oncotarget.

Gaikwad, N. W., L. Yang, P. Muti, J. L. Meza, S. Pruthi, J. N. Ingle, et al. (2008). "The molecular etiology of breast cancer: evidence from biomarkers of risk." International Journal of Cancer 122(9): 1949-1957.

Gaikwad, N. W., L. Yang, S. Pruthi, J. N. Ingle, N. Sandhu, E. G. Rogan, et al. (2009). "Urine biomarkers of risk in the molecular etiology of breast cancer." Breast Cancer (Auckl) 3: 1-8.

Gallagher, S. R. (2001). "One-dimensional SDS gel electrophoresis of proteins." Current Protocols in Protein Science Chapter 10: Unit 10.11.

Galvao, E. R., L. M. Martins, J. O. Ibiapina, H. M. Andrade and S. J. Monte (2011). "Breast cancer proteomics: a review for clinicians." J Cancer Res Clin Oncol 137(6): 915- 925.

Gamagedara, S. and Y. Ma (2011). "Biomarker analysis for prostate cancer diagnosis using LC-MS and CE-MS." Bioanalysis 3(18): 2129-2142.

Garofalo, C., M. Koda, S. Cascio, M. Sulkowska, L. Kanczuga-Koda, J. Golaszewska, et al. (2006). "Increased expression of leptin and the leptin receptor as a marker of breast cancer progression: possible role of obesity-related stimuli." Clin Cancer Res 12(5): 1447-1453.

Garrisi, V. M., S. Tommasi, A. Facchiano, I. Bongarzone, M. De Bortoli, M. Cremona, et al. (2013). "Proteomic profile in familial breast cancer patients." Clin Biochem 46(3): 259-265.

Gast, M. C., J. H. Schellens and J. H. Beijnen (2009). "Clinical proteomics in breast cancer: a review." Breast Cancer Res Treat 116(1): 17-29.

Gast, M. C., J. H. Schellens, J. H. Beijnen, M.-C. W. Gast, J. H. M. Schellens and J. H. Beijnen (2009). "Clinical proteomics in breast cancer: a review." Breast Cancer Research & Treatment 116(1): 17-29.

Gast, M. C., C. H. Van Gils, L. F. Wessels, N. Harris, J. M. Bonfrer, E. J. Rutgers, et al. (2009). "Serum protein profiling for diagnosis of breast cancer using SELDI-TOF MS." Oncol Rep 22(1): 205-213.

315

Gast, M. C., M. Zapatka, H. van Tinteren, M. Bontenbal, P. N. Span, V. C. Tjan-Heijnen, et al. (2011). "Postoperative serum proteomic profiles may predict recurrence-free survival in high-risk primary breast cancer." Journal of Cancer Research & Clinical Oncology 137(12): 1773-1783.

Geiger, T., S. F. Madden, W. M. Gallagher, J. Cox and M. Mann (2012). "Proteomic portrait of human breast cancer progression identifies novel prognostic markers." Cancer Research 72(9): 2428-2439.

Gergov, M., I. Ojanpera and E. Vuori (2003). "Simultaneous screening for 238 drugs in blood by liquid chromatography-ion spray tandem mass spectrometry with multiple- reaction monitoring." Journal of Chromatography B: Analytical Technologies in the Biomedical & Life Sciences 795(1): 41-53.

Giess, C. S., E. P. Frost and R. L. Birdwell (2012). "Difficulties and errors in diagnosis of breast neoplasms." Semin Ultrasound CT MR 33(4): 288-299.

Giovanella, L., L. Ceriani, G. Giardina, D. Bardelli, F. Tanzi and S. Garancini (2002). "Serum cytokeratin fragment 21.1 (CYFRA 21.1) as tumour marker for breast cancer: comparison with carbohydrate antigen 15.3 (CA 15.3) and carcinoembryonic antigen (CEA)." Clin Chem Lab Med 40(3): 298-303.

Glorieux, C., N. Dejeans, B. Sid, R. Beck, P. B. Calderon and J. Verrax (2011). "Catalase overexpression in mammary cancer cells leads to a less aggressive phenotype and an altered response to chemotherapy." Biochem Pharmacol 82(10): 1384-1390.

Godfrey, A. C., Z. Xu, C. R. Weinberg, R. C. Getts, P. A. Wade, L. A. DeRoo, et al. (2013). "Serum microRNA expression as an early marker for breast cancer risk in prospectively collected samples from the Sister Study cohort." Breast Cancer Res 15(3): R42.

Goldhirsch, A., E. P. Winer, A. S. Coates, R. D. Gelber, M. Piccart-Gebhart, B. Thurlimann, et al. (2013). "Personalizing the treatment of women with early breast cancer: highlights of the St Gallen International Expert Consensus on the Primary Therapy of Early Breast Cancer 2013." Ann Oncol 24(9): 2206-2223.

Goldhirsch, A., W. C. Wood, A. S. Coates, R. D. Gelber, B. Thurlimann and H. J. Senn (2011). "Strategies for subtypes--dealing with the diversity of breast cancer: highlights of the St. Gallen International Expert Consensus on the Primary Therapy of Early Breast Cancer 2011." Ann Oncol 22(8): 1736-1747.

316

Goldhirsch, A., W. C. Wood, R. D. Gelber, A. S. Coates, B. Thurlimann and H. J. Senn (2003). "Meeting highlights: updated international expert consensus on the primary therapy of early breast cancer." J Clin Oncol 21(17): 3357-3365.

Goldhirsch, A., W. C. Wood, R. D. Gelber, A. S. Coates, B. Thurlimann and H. J. Senn (2007). "Progress and promise: highlights of the international expert consensus on the primary therapy of early breast cancer 2007." Ann Oncol 18(7): 1133-1144.

Goligorsky, M. S., F. Addabbo and E. O'Riordan (2007). "Diagnostic potential of urine proteome: a broken mirror of renal diseases." Journal of the American Society of Nephrology 18(8): 2233-2239.

Goncalves, A., E. Charafe-Jauffret, F. Bertucci, S. Audebert, Y. Toiron, B. Esterni, et al. (2008). "Protein profiling of human breast tumor cells identifies novel biomarkers associated with molecular subtypes." Molecular & Cellular Proteomics 7(8): 1420-1433.

Goncalves, A., B. Esterni, F. Bertucci, R. Sauvan, C. Chabannon, M. Cubizolles, et al. (2006). "Postoperative serum proteomic profiles may predict metastatic relapse in high- risk primary breast cancer patients receiving adjuvant chemotherapy." Oncogene 25(7): 981-989.

Gonzales, P. A., T. Pisitkun, J. D. Hoffert, D. Tchapyjnikov, R. A. Star, R. Kleta, et al. (2009). "Large-scale proteomics and phosphoproteomics of urinary exosomes." Journal of the American Society of Nephrology 20(2): 363-379.

Gonzalez-Reyes, S., L. Marin, L. Gonzalez, L. O. Gonzalez, J. M. del Casar, M. L. Lamelas, et al. (2010). "Study of TLR3, TLR4 and TLR9 in breast carcinomas and their association with metastasis." BMC Cancer 10: 665.

Goodwin, A., S. Parker, D. Ghersi and N. Wilcken (2009). "Post-operative radiotherapy for ductal carcinoma in situ of the breast." Cochrane Database Syst Rev(3): Cd000563.

Gotzsche, P. C. and M. Nielsen (2006). "Screening for breast cancer with mammography." Cochrane Database Syst Rev(4): CD001877.

Gotzsche, P. C., M. Nielsen, P. C. Gotzsche and M. Nielsen (2011). "Screening for breast cancer with mammography." Cochrane Database of Systematic Reviews(1): CD001877.

Goufman, E. I., S. A. Moshkovskii, O. V. Tikhonova, P. G. Lokhov, V. G. Zgoda, M. V. Serebryakova, et al. (2006). "Two-dimensional electrophoretic proteome study of serum

317

thermostable fraction from patients with various tumor conditions." Biochemistry. Biokhimiia 71(4): 354-360.

Govorukhina, N. I., T. H. Reijmers, S. O. Nyangoma, A. G. van der Zee, R. C. Jansen and R. Bischoff (2006). "Analysis of human serum by liquid chromatography-mass spectrometry: improved sample preparation and data analysis." J Chromatogr A 1120(1- 2): 142-150.

Grebe, S. K. and R. J. Singh (2011). "LC-MS/MS in the Clinical Laboratory - Where to From Here?" The Clinical biochemist. Reviews / Australian Association of Clinical Biochemists 32(1): 5-31.

Grizzle, W. E., B. L. Adam, W. L. Bigbee, T. P. Conrads, C. Carroll, Z. Feng, et al. (2003). "Serum protein expression profiling for cancer detection: validation of a SELDI- based approach for prostate cancer." Dis Markers 19(4-5): 185-195.

Grus, F. H., V. N. Podust, K. Bruns, K. Lackner, S. Fu, E. A. Dalmasso, et al. (2005). "SELDI-TOF-MS ProteinChip Array Profiling of Tears from Patients with Dry Eye." Investigative Ophthalmology & Visual Science 46(3): 863-876.

Guo, T., J. Gu, O. P. Soldin, R. J. Singh and S. J. Soldin (2008). "Rapid measurement of estrogens and their metabolites in human serum by liquid chromatography-tandem mass spectrometry without derivatization." Clinical Biochemistry 41(9): 736-741.

Gutierrez, C. and R. Schiff (2011). "HER2: biology, detection, and clinical implications." Archives of Pathology & Laboratory Medicine 135(1): 55-62.

Gygi, S. P., B. Rist, S. A. Gerber, F. Turecek, M. H. Gelb and R. Aebersold (1999). "Quantitative analysis of complex protein mixtures using isotope-coded affinity tags." Nature Biotechnology 17(10): 994-999.

Gyorffy, B., A. Lanczky, A. C. Eklund, C. Denkert, J. Budczies, Q. Li, et al. (2010). "An online survival analysis tool to rapidly assess the effect of 22,277 genes on breast cancer prognosis using microarray data of 1,809 patients." Breast Cancer Res Treat 123(3): 725- 731.

Hamrita, B., H. Ben Nasr, S. Gabbouj, N. Bouaouina, L. Chouchane and K. Chahed (2011). "Apolipoprotein A1 -75 G/A and +83 C/T polymorphisms: susceptibility and prognostic implications in breast cancer." Mol Biol Rep 38(3): 1637-1643.

318

Hamrita, B., K. Chahed, M. Trimeche, C. L. Guillier, P. Hammann, A. Chaieb, et al. (2009). "Proteomics-based identification of alpha1-antitrypsin and haptoglobin precursors as novel serum markers in infiltrating ductal breast carcinomas." Clin Chim Acta 404(2): 111-118.

Han, B. and R. E. Higgs (2008). "Proteomics: from hypothesis to quantitative assay on a single platform. Guidelines for developing MRM assays using ion trap mass spectrometers." Briefings in Functional Genomics & Proteomics 7(5): 340-354.

Han, Z., J. Ni, P. Smits, C. B. Underhill, B. Xie, Y. Chen, et al. (2001). "Extracellular matrix protein 1 (ECM1) has angiogenic properties and is expressed by breast tumor cells." Faseb j 15(6): 988-994.

Hanahan, D. and R. A. Weinberg (2000). "The hallmarks of cancer." Cell 100(1): 57-70.

Hanahan, D. and R. A. Weinberg (2011). "Hallmarks of cancer: the next generation." Cell 144(5): 646-674.

Hanash, S. and S. Hanash (2003). "Disease proteomics." Nature 422(6928): 226-232.

Hanash, S. M., S. J. Pitteri and V. M. Faca (2008). "Mining the plasma proteome for cancer biomarkers." Nature 452(7187): 571-579.

Harris, L., H. Fritsche, R. Mennel, L. Norton, P. Ravdin, S. Taube, et al. (2007). "American Society of Clinical Oncology 2007 update of recommendations for the use of tumor markers in breast cancer." J Clin Oncol 25(33): 5287-5312.

Harris, L. N., N. Ismaila, L. M. McShane, F. Andre, D. E. Collyar, A. M. Gonzalez- Angulo, et al. (2016). "Use of Biomarkers to Guide Decisions on Adjuvant Systemic Therapy for Women With Early-Stage Invasive Breast Cancer: American Society of Clinical Oncology Clinical Practice Guideline." J Clin Oncol 34(10): 1134-1150.

Harsha, H. C., A. Jimeno, H. Molina, A. B. Mihalas, M. G. Goggins, R. H. Hruban, et al. (2008). "Activated epidermal growth factor receptor as a novel target in pancreatic cancer therapy." J Proteome Res 7(11): 4651-4658.

Hashida, H., A. Takabayashi, M. Kanai, M. Adachi, K. Kondo, N. Kohno, et al. (2002). "Aminopeptidase N is involved in cell motility and angiogenesis: its clinical significance in human colon cancer." Gastroenterology 122(2): 376-386.

319

Hassanein, M., J. C. Callison, C. Callaway-Lane, M. C. Aldrich, E. L. Grogan and P. P. Massion (2012). "The state of molecular biomarkers for the early detection of lung cancer." Cancer prevention research (Philadelphia, Pa.) 5(8): 992-1006.

Hathout, Y., M. L. Gehrmann, A. Chertov and C. Fenselau (2004). "Proteomic phenotyping: metastatic and invasive breast cancer." Cancer Lett 210(2): 245-253.

Havanapan, P.-O. and V. Thongboonkerd (2009). "Are protease inhibitors required for gel-based proteomics of kidney and urine?" Journal of Proteome Research 8(6): 3109- 3117.

He, J., J. Gornbein, D. Shen, M. Lu, L. E. Rovai, H. Shau, et al. (2007). "Detection of breast cancer biomarkers in nipple aspirate fluid by SELDI-TOF and their identification by combined liquid chromatography-tandem mass spectrometry." International Journal of Oncology 30(1): 145-154.

He, P., H. Z. He, J. Dai, Y. Wang, Q. H. Sheng, L. P. Zhou, et al. (2005). "The human plasma proteome: analysis of Chinese serum using shotgun strategy." Proteomics 5(13): 3442-3453.

Heike, Y., M. Hosokawa, S. Osumi, D. Fujii, K. Aogi, N. Takigawa, et al. (2005). "Identification of serum proteins related to adverse effects induced by docetaxel infusion from protein expression profiles of serum using SELDI ProteinChip system." Anticancer Res 25(2b): 1197-1203.

Henneges, C., D. Bullinger, R. Fux, N. Friese, H. Seeger, H. Neubauer, et al. (2009). "Prediction of breast cancer by profiling of urinary RNA metabolites using Support Vector Machine-based feature selection." BMC Cancer 9: 104.

Hensen, U., T. Meyer, J. Haas, R. Rex, G. Vriend and H. Grubmuller (2012). "Exploring protein dynamics space: the dynasome as the missing link between protein structure and function." PLoS One 7(5): e33931.

Hernandez-Borges, J., C. Neususs, A. Cifuentes and M. Pelzing (2004). "On-line capillary electrophoresis-mass spectrometry for the analysis of biomolecules." Electrophoresis 25(14): 2257-2281.

Higgins, S. A., E. T. Matloff, D. L. Rimm, J. Dziura, B. G. Haffty and B. L. King (2005). "Patterns of reduced nipple aspirate fluid production and ductal lavage cellularity in women at high risk for breast cancer." Breast Cancer Res 7(6): R1017-1022.

320

Ho, J., J. W. Kong, L. Y. Choong, M. C. Loh, W. Toy, P. K. Chong, et al. (2009). "Novel breast cancer metastasis-associated proteins." Journal of Proteome Research 8(2): 583- 594.

Hogdall, E. V., J. S. Johansen, S. K. Kjaer, P. A. Price, J. Blaakjaer and C. K. Hogdall (2000). "Stability of YKL-40 concentration in blood samples." Scand J Clin Lab Invest 60(4): 247-251.

Holen, I., J. Whitworth, F. Nutter, A. Evans, H. K. Brown, D. V. Lefley, et al. (2012). "Loss of plakoglobin promotes decreased cell-cell contact, increased invasion, and breast cancer cell dissemination in vivo." Breast Cancer Res 14(3): R86.

Hondermarck, H., A. S. Vercoutter-Edouart, F. Revillion, J. Lemoine, I. el-Yazidi- Belkoura, V. Nurcombe, et al. (2001). "Proteomics of breast cancer for marker discovery and signal pathway profiling." Proteomics 1(10): 1216-1232.

Houghton, J., W. D. George, J. Cuzick, C. Duggan, I. S. Fentiman and M. Spittle (2003). "Radiotherapy and tamoxifen in women with completely excised ductal carcinoma in situ of the breast in the UK, Australia, and New Zealand: randomised controlled trial." Lancet 362(9378): 95-102.

Hsieh, S. Y., R. K. Chen, Y. H. Pan and H. L. Lee (2006). "Systematical evaluation of the effects of sample collection procedures on low-molecular-weight serum/plasma proteome profiling." Proteomics 6(10): 3189-3198.

Hsu, W. Y., W. D. Lin, Y. Tsai, C. T. Lin, H. C. Wang, L. B. Jeng, et al. (2011). "Analysis of urinary nucleosides as potential tumor markers in human breast cancer by high performance liquid chromatography/electrospray ionization tandem mass spectrometry." Clin Chim Acta 412(19-20): 1861-1866.

Hu, Y., S. Zhang, J. Yu, J. Liu and S. Zheng (2005). "SELDI-TOF-MS: the proteomics and bioinformatics approaches in the diagnosis of breast cancer." Breast 14(4): 250-255.

Hu, Z., C. Fan, D. S. Oh, J. S. Marron, X. He, B. F. Qaqish, et al. (2006). "The molecular portraits of breast tumors are conserved across microarray platforms." BMC Genomics 7: 96.

321

Huang, H. L., T. Stasyk, S. Morandell, H. Dieplinger, G. Falkensammer, A. Griesmacher, et al. (2006). "Biomarker discovery in breast cancer serum using 2-D differential gel electrophoresis/ MALDI-TOF/TOF and data validation by routine clinical assays." Electrophoresis 27(8): 1641-1650.

Huang, L., G. Harvie, J. S. Feitelson, K. Gramatikoff, D. A. Herold, D. L. Allen, et al. (2005). "Immunoaffinity separation of plasma proteins by IgY microbeads: meeting the needs of proteomic sample preparation and analysis." Proteomics 5(13): 3314-3328.

Huang, Y., H. Zhang, J. Cai, L. Fang, J. Wu, C. Ye, et al. (2013). "Overexpression of MACC1 and Its significance in human Breast Cancer Progression." Cell Biosci 3(1): 16.

Hulka, B. S. and P. G. Moorman (2008). "Breast cancer: hormones and other risk factors." Maturitas 61(1-2): 203-213; discussion 213.

Hulsmeier, A. J., P. Paesold-Burda and T. Hennet (2007). "N-glycosylation site occupancy in serum glycoproteins using multiple reaction monitoring liquid chromatography-mass spectrometry." Molecular & Cellular Proteomics 6(12): 2132- 2138.

Husi, H., N. Stephens, A. Cronshaw, A. MacDonald, I. Gallagher, C. Greig, et al. (2011). "Proteomic analysis of urinary upper gastrointestinal cancer markers." Proteomics Clinical Applications 5(5-6): 289-299.

Hutchens, T. and T. Yip (1993). "New desorption strategies for the mass spectrometric analysis of macromolecules." Rapid Cmmunications in Mass Spectrometry 7: 576-580.

Il'yasova, D., K. Kennedy, I. Spasojevic, F. Wang, A. A. Tolun, K. Base, et al. (2011). "Individual responses to chemotherapy-induced oxidative stress." Breast Cancer Res Treat 125(2): 583-589.

Irmak, S., D. Tilki, J. Heukeshoven, L. Oliveira-Ferrer, M. Friedrich, H. Huland, et al. (2005). "Stage-dependent increase of orosomucoid and zinc-alpha2-glycoprotein in urinary bladder cancer." Proteomics 5(16): 4296-4304.

Janssen, B. J., E. G. Huizinga, H. C. Raaijmakers, A. Roos, M. R. Daha, K. Nilsson- Ekdahl, et al. (2005). "Structures of complement component C3 provide insights into the function and evolution of immunity." Nature 437(7058): 505-511.

322

Jeon, Y. R., S. Y. Kim, E. J. Lee, Y. N. Kim, D. Y. Noh, S. Y. Park, et al. (2013). "Identification of annexin II as a novel secretory biomarker for breast cancer." Proteomics 13(21): 3145-3156.

Jogie-Brahim, S., D. Feldman and Y. Oh (2009). "Unraveling insulin-like growth factor binding protein-3 actions in human disease." Endocr Rev 30(5): 417-437.

Johannesson, N., M. Wetterhall, K. E. Markides and J. Bergquist (2004). "Monomer surface modifications for rapid peptide analysis by capillary electrophoresis and capillary electrochromatography coupled to electrospray ionization-mass spectrometry." Electrophoresis 25(6): 809-816.

Johansson, H. J., B. C. Sanchez, J. Forshed, O. Stal, H. Fohlin, R. Lewensohn, et al. (2015). "Proteomics profiling identify CAPS as a potential predictive marker of tamoxifen resistance in estrogen receptor positive breast cancer." Clin Proteomics 12(1): 8.

Jorgensen, K. (2012). "Is the tide turning against breast screening?" Breast Cancer Res 14(4): 107.

Julien, J. P., N. Bijker, I. S. Fentiman, J. L. Peterse, V. Delledonne, P. Rouanet, et al. (2000). "Radiotherapy in breast-conserving treatment for ductal carcinoma in situ: first results of the EORTC randomised phase III trial 10853. EORTC Breast Cancer Cooperative Group and EORTC Radiotherapy Group." Lancet 355(9203): 528-533.

Kabbage, M., M. Trimeche, S. Bergaoui, P. Hammann, L. Kuhn, B. Hamrita, et al. (2013). "Calreticulin expression in infiltrating ductal breast carcinomas: relationships with disease progression and humoral immune responses." Tumour Biol 34(2): 1177-1188.

Kadowaki, M., T. Sangai, T. Nagashima, M. Sakakibara, H. Yoshitomi, S. Takano, et al. (2011). "Identification of vitronectin as a novel serum marker for early breast cancer detection using a new proteomic approach." J Cancer Res Clin Oncol 137(7): 1105-1115.

Kaiser, T., H. Kamal, A. Rank, H. J. Kolb, E. Holler, A. Ganser, et al. (2004). "Proteomics applied to the clinical follow-up of patients after allogeneic hematopoietic stem cell transplantation." Blood 104(2): 340-349.

Kakisaka, T., T. Kondo, T. Okano, K. Fujii, K. Honda, M. Endo, et al. (2007). "Plasma proteomics of pancreatic cancer patients by multi-dimensional liquid chromatography and two-dimensional difference gel electrophoresis (2D-DIGE): up-regulation of leucine- rich alpha-2-glycoprotein in pancreatic cancer." J Chromatogr B Analyt Technol Biomed Life Sci 852(1-2): 257-267. 323

Kalager, M., M. Zelen, F. Langmark and H. O. Adami (2010). "Effect of screening mammography on breast-cancer mortality in Norway." N Engl J Med 363(13): 1203- 1210.

Kang, U. B., Y. Ahn, J. W. Lee, Y. H. Kim, J. Kim, M. H. Yu, et al. (2010). "Differential profiling of breast cancer plasma proteome by isotope-coded affinity tagging method reveals biotinidase as a breast cancer biomarker." BMC Cancer 10: 114.

Kania, K., E. A. Byrnes, J. P. Beilby, S. A. Webb, K. J. Strong, K. Kania, et al. "Urinary proteases degrade albumin: implications for measurement of albuminuria in stored samples." Annals of Clinical Biochemistry 47(Pt 2): 151-157.

Karas, M. and F. Hillenkamp (1988). "Laser desorption ionization of proteins with molecular masses exceeding 10,000 daltons." Analytical Chemistry 60(20): 2299-2301.

Karp, N. A., P. S. McCormick, M. R. Russell and K. S. Lilley (2007). "Experimental and statistical considerations to avoid false conclusions in proteomics studies using differential in-gel electrophoresis." Mol Cell Proteomics 6(8): 1354-1364.

Kawakami, T., Y. Hoshida, F. Kanai, Y. Tanaka, K. Tateishi, T. Ikenoue, et al. (2005). "Proteomic analysis of sera from hepatocellular carcinoma patients after radiofrequency ablation treatment." Proteomics 5(16): 4287-4295.

Kawate, T., K. Iwaya, K. Koshikawa, T. Moriya, T. Yamasaki, S. Hasegawa, et al. (2015). "High levels of DJ-1 protein and isoelectric point 6.3 isoform in the serum of breast cancer patients." Cancer Sci.

Kennecke, H., R. Yerushalmi, R. Woods, M. C. Cheang, D. Voduc, C. H. Speers, et al. (2010). "Metastatic behavior of breast cancer subtypes." J Clin Oncol 28(20): 3271-3277.

Kentsis, A., F. Monigatti, K. Dorff, F. Campagne, R. Bachur and H. Steen (2009). "Urine proteomics for profiling of human disease using high accuracy mass spectrometry." Proteomics Clinical Applications 3(9): 1052-1061.

Keshishian, H., T. Addona, M. Burgess, E. Kuhn and S. A. Carr (2007). "Quantitative, multiplexed assays for low abundance proteins in plasma by targeted mass spectrometry and stable isotope dilution." Molecular & Cellular Proteomics 6(12): 2212-2229.

Keshishian, H., T. Addona, M. Burgess, D. R. Mani, X. Shi, E. Kuhn, et al. (2009). "Quantification of cardiovascular biomarkers in patient plasma by targeted mass

324

spectrometry and stable isotope dilution." Molecular & Cellular Proteomics 8(10): 2339- 2349.

Khleif, S. N., J. H. Doroshow and W. N. Hait (2010). "AACR-FDA-NCI Cancer Biomarkers Collaborative Consensus Report: Advancing the Use of Biomarkers in Cancer Drug Development." Clinical Cancer Research 16(13): 3299-3318.

Khurana, A., D. Jung-Beom, X. He, S. H. Kim, R. C. Busby, L. Lorenzon, et al. (2013). "Matrix detachment and proteasomal inhibitors diminish Sulf-2 expression in breast cancer cell lines and mouse xenografts." Clin Exp Metastasis 30(4): 407-415.

Kiernan, U. A., K. A. Tubbs, D. Nedelkov, E. E. Niederkofler, E. McConnell and R. W. Nelson (2003). "Comparative urine protein phenotyping using mass spectrometric immunoassay." Journal of Proteome Research 2(2): 191-197.

Kim, B. K., J. W. Lee, P. J. Park, Y. S. Shin, W. Y. Lee, K. A. Lee, et al. (2009). "The multiplex bead array approach to identifying serum biomarkers associated with breast cancer." Breast Cancer Res 11(2): R22.

Kim, J. C., J. H. Yu, Y. K. Cho, C. S. Jung, S. H. Ahn, G. Gong, et al. (2012). "Expression of SPRR3 is associated with tumor cell proliferation in less advanced stages of breast cancer." Breast Cancer Res Treat 133(3): 909-916.

Kitteringham, N. R., R. E. Jenkins, C. S. Lane, V. L. Elliott and B. K. Park (2009). "Multiple reaction monitoring for quantitative biomarker analysis in proteomics and metabolomics." Journal of Chromatography B: Analytical Technologies in the Biomedical & Life Sciences 877(13): 1229-1239.

Klein-Scory, S., S. Kubler, H. Diehl, C. Eilert-Micus, A. Reinacher-Schick, K. Stuhler, et al. (2010). "Immunoscreening of the extracellular proteome of colorectal cancer cells." BMC Cancer 10: 70.

Klose, J. (1975). "Protein mapping by combined isoelectric focusing and electrophoresis of mouse tissues. A novel approach to testing for induced point mutations in mammals." Humangenetik 26(3): 231-243.

Kobe, B. and A. V. Kajava (2001). "The leucine-rich repeat as a protein recognition motif." Curr Opin Struct Biol 11(6): 725-732.

325

Kolch, W., C. Neususs, M. Pelzing and H. Mischak (2005). "Capillary electrophoresis- mass spectrometry as a powerful tool in clinical diagnosis and biomarker discovery." Mass Spectrom Rev 24(6): 959-977.

Kreunin, P., J. Zhao, C. Rosser, V. Urquidi, D. M. Lubman and S. Goodison (2007). "Bladder cancer associated glycoprotein signatures revealed by urinary proteomic profiling." J Proteome Res 6(7): 2631-2639.

Kulasingam, V., Y. Zheng, A. Soosaipillai, A. E. Leon, M. Gion and E. P. Diamandis (2009). "Activated leukocyte cell adhesion molecule: a novel biomarker for breast cancer." Int J Cancer 125(1): 9-14.

Kuzyk, M. A., D. Smith, J. Yang, T. J. Cross, A. M. Jackson, D. B. Hardie, et al. (2009). "Multiple reaction monitoring-based, multiplexed, absolute quantitation of 45 proteins in human plasma." Molecular & Cellular Proteomics 8(8): 1860-1877.

Kuzyk, M. A., D. Smith, J. C. Yang, T. J. Cross, A. M. Jackson, D. B. Hardie, et al. (2009). "Multiple Reaction Monitoring-based, Multiplexed, Absolute Quantitation of 45 Proteins in Human Plasma." Molecular & Cellular Proteomics 8(8): 1860-1877.

Kwak, J. Y., T. Z. Ma, M. J. Yoo, B. H. Choi, H. G. Kim, S. R. Kim, et al. (2004). "The comparative analysis of serum proteomes for the discovery of biomarkers for acute myeloid leukemia." Exp Hematol 32(9): 836-842.

Kwan, M. L., L. H. Kushi, E. Weltzien, E. K. Tam, A. Castillo, C. Sweeney, et al. (2010). "Alcohol consumption and breast cancer recurrence and survival among women with early-stage breast cancer: the life after cancer epidemiology study." J Clin Oncol 28(29): 4410-4416.

Ladenson, J. H., L. M. Tsai, J. M. Michael, G. Kessler and J. H. Joist (1974). "Serum versus heparinized plasma for eighteen common chemistry tests: is serum the appropriate specimen?" Am J Clin Pathol 62(4): 545-552.

Lakhani, S. R., E. I.O., S. J. Schnitt, P. H. Tan and M. J. van de Vijver, Eds. (2012). WHO Classification of Tumours of the Breast, Fourth Edition. Lyon, IARC WHO Classification of Tumours.

Lal, G., S. Hashimi, B. J. Smith, C. F. Lynch, L. Zhang, R. A. Robinson, et al. (2009). "Extracellular matrix 1 (ECM1) expression is a novel prognostic marker for poor long- term survival in breast cancer: a Hospital-based Cohort Study in Iowa." Ann Surg Oncol 16(8): 2280-2287.

326

Lam, S. W., C. R. Jimenez and E. Boven (2014). "Breast cancer classification by proteomic technologies: current state of knowledge." Cancer Treat Rev 40(1): 129-138.

Laronga, C., S. Becker, P. Watson, B. Gregory, L. Cazares, H. Lynch, et al. (2003). "SELDI-TOF serum profiling for prognostic and diagnostic classification of breast cancers." Dis Markers 19(4-5): 229-238.

Lebrecht, A., D. Boehm, M. Schmidt, H. Koelbl and F. H. Grus (2009). "Surface- enhanced Laser Desorption/Ionisation Time-of-flight Mass Spectrometry to Detect Breast Cancer Markers in Tears and Serum." Cancer Genomics Proteomics 6(2): 75-83.

Lebrecht, A., D. Boehm, M. Schmidt, H. Koelbl, R. L. Schwirz, F. H. Grus, et al. (2009). "Diagnosis of breast cancer by tear proteomic pattern." Cancer Genomics & Proteomics 6(3): 177-182.

Lee, K. (2008). "Evaluation of an effective sample prefractionation method for the proteome analysis of breast cancer tissue using narrow range two-dimensional gel electrophoresis." Bioscience, Biotechnology & Biochemistry 72(6): 1464-1474.

Lee, K. M., K. Nam, S. Oh, J. Lim, Y. P. Kim, J. W. Lee, et al. (2014). "Extracellular matrix protein 1 regulates cell proliferation and trastuzumab resistance through activation of epidermal growth factor signaling." Breast Cancer Res 16(6): 479.

Lee, R. S., F. Monigatti, A. C. Briscoe, Z. Waldon, M. R. Freeman and H. Steen (2008). "Optimizing sample handling for urinary proteomics." Journal of Proteome Research 7(9): 4022-4030.

Lei, T., X. Zhao, S. Jin, Q. Meng, H. Zhou and M. Zhang (2012). "Discovery of Potential Bladder Cancer Biomarkers by Comparative Urine Proteomics and Analysis." Clinical genitourinary cancer.

Lei, T., X. Zhao, S. Jin, Q. Meng, H. Zhou and M. Zhang (2013). "Discovery of potential bladder cancer biomarkers by comparative urine proteomics and analysis." Clin Genitourin Cancer 11(1): 56-62.

Leinonen, T., R. Pirinen, J. Bohm, R. Johansson, A. Rinne, E. Weber, et al. (2007). "Biological and prognostic role of acid cysteine proteinase inhibitor (ACPI, cystatin A) in non-small-cell lung cancer." J Clin Pathol 60(5): 515-519.

327

Leong, S., M. J. McKay, R. I. Christopherson and R. C. Baxter (2012). "Biomarkers of breast cancer apoptosis induced by chemotherapy and TRAIL." J Proteome Res 11(2): 1240-1250.

Leong, S., A. C. Nunez, M. Z. Lin, B. Crossett, R. I. Christopherson and R. C. Baxter (2012). "iTRAQ-based proteomic profiling of breast cancer cell response to doxorubicin and TRAIL." J Proteome Res 11(7): 3561-3572.

Levenson, V. V. (2007). "Biomarkers for early detection of breast cancer: what, when, and where?" Biochim Biophys Acta 1770(6): 847-856.

Levicar, N., J. Kos, A. Blejec, R. Golouh, I. Vrhovec, S. Frkovic-Grazio, et al. (2002). "Comparison of potential biological markers cathepsin B, cathepsin L, stefin A and stefin B with urokinase and plasminogen activator inhibitor-1 and clinicopathological data of breast carcinoma patients." Cancer Detect Prev 26(1): 42-49.

Li, C. I. (2011). "Discovery and validation of breast cancer early detection biomarkers in preclinical samples." Horm Cancer 2(2): 125-131.

Li, D.-Q., L. Wang, F. Fei, Y.-F. Hou, J.-M. Luo, C. Wei, et al. (2006). "Identification of breast cancer metastasis-associated proteins in an isogenic tumor metastasis model using two-dimensional gel electrophoresis and liquid chromatography-ion trap-mass spectrometry." Proteomics 6(11): 3352-3368.

Li, J., L. Jia, P. Zhao, Y. Jiang, S. Zhong and D. Chen (2012). "Stable knockdown of clusterin by vectorbased RNA interference in a human breast cancer cell line inhibits tumour cell invasion and metastasis." J Int Med Res 40(2): 545-555.

Li, J., R. Orlandi, C. N. White, J. Rosenzweig, J. Zhao, E. Seregni, et al. (2005). "Independent validation of candidate breast cancer serum biomarkers identified by mass spectrometry." Clin Chem 51(12): 2229-2235.

Li, J., Z. Zhang, J. Rosenzweig, Y. Y. Wang and D. W. Chan (2002). "Proteomics and bioinformatics approaches for identification of serum biomarkers to detect breast cancer." Clinical Chemistry 48(8): 1296-1304.

328

Li, J., J. Zhao, X. Yu, J. Lange, H. Kuerer, S. Krishnamurthy, et al. (2005). "Identification of biomarkers for breast cancer in nipple aspiration and ductal lavage fluid." Clinical Cancer Research 11(23): 8312-8320.

Liang, X., J. Zhao, M. Hajivandi, R. Wu, J. Tao, J. W. Amshey, et al. (2006). "Quantification of membrane and membrane-bound proteins in normal and malignant breast cancer cells isolated from the same patient with primary breast carcinoma." J Proteome Res 5(10): 2632-2641.

Liebich, H. M., S. Muller-Hagedorn, F. Klaus, K. Meziane, K. R. Kim, A. Frickenschmidt, et al. (2005). "Chromatographic, capillary electrophoretic and matrix- assisted laser desorption ionization time-of-flight mass spectrometry analysis of urinary modified nucleosides as tumor markers." J Chromatogr A 1071(1-2): 271-275.

Ligibel, J. (2012). "Lifestyle factors in cancer survivorship." J Clin Oncol 30(30): 3697- 3704.

Lim, S., L. Y. Choong, C. P. Kuan, C. Yunhao and Y. P. Lim (2009). "Regulation of macrophage inhibitory factor (MIF) by epidermal growth factor receptor (EGFR) in the MCF10AT model of breast cancer progression." Journal of Proteome Research 8(8): 4062-4076.

Lin, C. Y., K. H. Tsui, C. C. Yu, C. W. Yeh, P. L. Chang and B. Y. Yung (2006). "Searching cell-secreted proteomes for potential urinary bladder tumor markers." Proteomics 6(15): 4381-4389.

Lindemann, S., N. D. Tolley, D. A. Dixon, T. M. McIntyre, S. M. Prescott, G. A. Zimmerman, et al. (2001). "Activated platelets mediate inflammatory signaling by regulated interleukin 1beta synthesis." J Cell Biol 154(3): 485-490.

Linden, M., S. B. Lind, C. Mayrhofer, U. Segersten, K. Wester, Y. Lyutvinskiy, et al. (2012). "Proteomic analysis of urinary biomarker candidates for nonmuscle invasive bladder cancer." Proteomics 12(1): 135-144.

Liotta, L. A., M. Ferrari and E. Petricoin (2003). "Clinical proteomics: written in blood." Nature 425(6961): 905.

Liotta, L. A. and E. F. Petricoin (2006). "Serum peptidome for cancer detection: spinning biologic trash into diagnostic gold." J Clin Invest 116(1): 26-30.

329

Liu, A. N., P. Sun, J. N. Liu, C. Y. Yu, H. J. Qu, A. H. Jiao, et al. (2014). "Analysis of the differences of serum protein mass spectrometry in patients with triple negative breast cancer and non-triple negative breast cancer." Tumour Biol.

Liu, J., C. F. Chen, C. W. Tsao, C. C. Chang, C. C. Chu and D. L. DeVoe (2009). "Polymer microchips integrating solid-phase extraction and high-performance liquid chromatography using reversed-phase polymethacrylate monoliths." Anal Chem 81(7): 2545-2554.

Liu, W., B. Liu, Q. Cai, J. Li, X. Chen and Z. Zhu (2012). "Proteomic identification of serum biomarkers for gastric cancer using multi-dimensional liquid chromatography and 2D differential gel electrophoresis." Clin Chim Acta 413(13-14): 1098-1106.

Liu, X., C. Shao, L. Wei, J. Duan, S. Wu, X. Li, et al. (2012). "An individual urinary proteome analysis in normal human beings to define the minimal sample number to represent the normal urinary proteome." Proteome Sci 10(1): 70.

Liu, Z., C. Chen, H. Yang, Y. Zhang, J. Long, X. Long, et al. (2012). "Proteomic features of potential tumor suppressor NESG1 in nasopharyngeal carcinoma." Proteomics 12(22): 3416-3425.

Loftheim, H., K. Midtvedt, A. Hartmann, A. V. Reisaeter, P. Falck, H. Holdaas, et al. (2012). "Urinary proteomic shotgun approach for identification of potential acute rejection biomarkers in renal transplant recipients." Transplant Res 1(1): 9.

Lopez, J. L. (2007). "Two-dimensional electrophoresis in proteome expression analysis." Journal of chromatography. B, Analytical technologies in the biomedical and life sciences 849(1-2): 190-202.

Lu, D. P., X. Y. Zhou, L. T. Yao, C. G. Liu, W. Ma, F. Jin, et al. (2014). "Serum soluble ST2 is associated with ER-positive breast cancer." BMC Cancer 14: 198.

Ludwig, J. A. and J. N. Weinstein (2005). "Biomarkers in Cancer Staging, Prognosis and Treatment Selection." Nat. Rev. Cancer 11: 845-856.

Ludwig, J. A. and J. N. Weinstein (2005). "Biomarkers in cancer staging, prognosis and treatment selection." Nat Rev Cancer 5(11): 845-856.

330

Ludwig, J. A. and J. N. Weinstein (2005). "Biomarkers in cancer staging, prognosis and treatment selection." Nature Reviews. Cancer 5(11): 845-856.

Lum, G. and S. R. Gambino (1974). "A comparison of serum versus heparinized plasma for routine chemistry tests." Am J Clin Pathol 61(1): 108-113.

Lumachi, F., S. M. Basso, A. A. Brandes, D. Pagano and M. Ermani (2004). "Relationship between tumor markers CEA and CA 15-3, TNM staging, estrogen receptor rate and MIB-1 index in patients with pT1-2 breast cancer." Anticancer Research 24(5B): 3221- 3224.

Luna, L. G., T. L. Williams, J. L. Pirkle and J. R. Barr (2008). "Ultra performance liquid chromatography isotope dilution tandem mass spectrometry for the absolute quantification of proteins and peptides.[Erratum appears in Anal Chem. 2008 Oct 1;80(19):7649]." Analytical Chemistry 80(8): 2688-2693.

Luque-Garcia, J. L. and T. A. Neubert (2007). "Sample preparation for serum/plasma profiling and biomarker identification by mass spectrometry." J Chromatogr A 1153(1- 2): 259-276.

Ma, X. and Y. Bai (2012). "IGF-1 activates the P13K/AKT signaling pathway via upregulation of secretory clusterin." Mol Med Rep 6(6): 1433-1437.

Maccio, A., C. Madeddu, G. Gramignano, C. Mulas, C. Floris, D. Massa, et al. (2010). "Correlation of body mass index and leptin with tumor size and stage of disease in hormone-dependent postmenopausal breast cancer: preliminary results and therapeutic implications." J Mol Med (Berl) 88(7): 677-686.

Magistroni, R., G. Ligabue, V. Lupo, L. Furci, M. Leonelli, L. Manganelli, et al. (2009). "Proteomic analysis of urine from proteinuric patients shows a proteolitic activity directed against albumin." Nephrology Dialysis Transplantation 24(5): 1672-1681.

Majidzadeh, A. K. and J. Gharechahi (2013). "Plasma proteomics analysis of tamoxifen resistance in breast cancer." Med Oncol 30(4): 753.

Makarov, A. (2000). "Electrostatic axially harmonic orbital trapping: a high-performance technique of mass analysis." Anal Chem 72(6): 1156-1162.

331

Makarov, A., E. Denisov, A. Kholomeev, W. Balschun, O. Lange, K. Strupat, et al. (2006). "Performance evaluation of a hybrid linear ion trap/orbitrap mass spectrometer." Anal Chem 78(7): 2113-2120.

Mallick, P. and B. Kuster (2010). "Proteomics: a pragmatic perspective." Nature Biotechnology 28(7): 695-709.

Marchina, E., M. G. Fontana, M. Speziani, A. Salvi, G. Ricca, D. Di Lorenzo, et al. (2010). "BRCA1 and BRCA2 genetic test in high risk patients and families: counselling and management." Oncol Rep 24(6): 1661-1667.

Marshall, J., A. Jankowski, S. Furesz, I. Kireeva, L. Barker, M. Dombrovsky, et al. (2004). "Human serum proteins preseparated by electrophoresis or chromatography followed by tandem mass spectrometry." J Proteome Res 3(3): 364-382.

Martinez, A., M. Vos, L. Guedez, G. Kaur, Z. Chen, M. Garayoa, et al. (2002). "The effects of adrenomedullin overexpression in breast tumor cells." J Natl Cancer Inst 94(16): 1226-1237.

Martosella, J., N. Zolotarjova, H. Liu, G. Nicol and B. E. Boyes (2005). "Reversed-phase high-performance liquid chromatographic prefractionation of immunodepleted human serum proteins to enhance mass spectrometry identification of lower-abundant proteins." J Proteome Res 4(5): 1522-1537.

Mathelin, C., A. Cromer, C. Wendling, C. Tomasetto and M. C. Rio (2006). "Serum biomarkers for detection of breast cancers: A prospective study." Breast Cancer Res Treat 96(1): 83-90.

Mavaddat, N., S. Peock, D. Frost, S. Ellis, R. Platte, E. Fineberg, et al. (2013). "Cancer risks for BRCA1 and BRCA2 mutation carriers: results from prospective analysis of EMBRACE." J Natl Cancer Inst 105(11): 812-822.

Mayya, V., K. Rezual, L. Wu, M. B. Fong and D. K. Han (2006). "Absolute quantification of multisite phosphorylation by selective reaction monitoring mass spectrometry: determination of inhibitory phosphorylation status of cyclin-dependent kinases." Molecular & Cellular Proteomics 5(6): 1146-1157.

McCormack, V. A. and I. dos Santos Silva (2006). "Breast density and parenchymal patterns as markers of breast cancer risk: a meta-analysis." Cancer Epidemiol Biomarkers Prev 15(6): 1159-1169.

332

McGale, P., C. Taylor, C. Correa, D. Cutter, F. Duane, M. Ewertz, et al. (2014). "Effect of radiotherapy after mastectomy and axillary surgery on 10-year recurrence and 20-year breast cancer mortality: meta-analysis of individual patient data for 8135 women in 22 randomised trials." Lancet 383(9935): 2127-2135.

McKay, M. J., J. Sherman, M. T. Laver, M. S. Baker, S. J. Clarke and M. P. Molloy (2007). "The development of multiple reaction monitoring assays for liver-derived plasma proteins." Proteomics Clin Appl 1(12): 1570-1581.

Mendrinos, S., J. D. Nolen, T. Styblo, G. Carlson, J. Pohl, M. Lewis, et al. (2005). "Cytologic findings and protein expression profiles associated with ductal carcinoma of the breast in ductal lavage specimens using surface-enhanced laser desorption and ionization-time of flight mass spectrometry." Cancer 105(3): 178-183.

Meng, R., M. Gormley, V. B. Bhat, A. Rosenberg and A. A. Quong (2011). "Low abundance protein enrichment for discovery of candidate plasma protein biomarkers for early detection of breast cancer." J Proteomics 75(2): 366-374.

Meng, Z. and T. D. Veenstra (2011). "Targeted mass spectrometry approaches for protein biomarker verification." J Proteomics 74(12): 2650-2659.

Metodieva, G., C. Greenwood, L. Alldridge, P. Sauven and M. Metodiev (2009). "A peptide-centric approach to breast cancer biomarker discovery utilizing label-free multiple reaction monitoring mass spectrometry." Proteomics Clin Appl 3(1): 78-82.

Michlmayr, A., T. Bachleitner-Hofmann, S. Baumann, M. Marchetti-Deschmann, I. Rech-Weichselbraun, C. Burghuber, et al. (2010). "Modulation of plasma complement by the initial dose of epirubicin/docetaxel therapy in breast cancer and its predictive value." British Journal of Cancer 103(8): 1201-1208.

Millar, E. K. A., P. H. Graham, C. M. McNeil, L. Browne, S. A. O'Toole, A. Boulghourjian, et al. (2011). "Prediction of outcome of early ER+ breast cancer is improved using a biomarker panel, which includes Ki-67 and p53." British Journal of Cancer 105(2): 272-280.

Millioni, R., S. Tolin, L. Puricelli, S. Sbrignadello, G. P. Fadini, P. Tessari, et al. (2011). "High abundance proteins depletion vs low abundance proteins enrichment: comparison of methods to reduce the plasma proteome complexity." PLoS One 6(5): e19603.

Mischak, H., G. Allmaier, R. Apweiler, T. Attwood, M. Baumann, A. Benigni, et al. (2010). "Recommendations for biomarker identification and qualification in clinical proteomics." Science Translational Medicine 2(46): 46ps42. 333

Mischak, H., J. J. Coon, J. Novak, E. M. Weissinger, J. P. Schanstra and A. F. Dominiczak (2009). "Capillary electrophoresis-mass spectrometry as a powerful tool in biomarker discovery and clinical diagnosis: an update of recent developments." Mass Spectrometry Reviews 28(5): 703-724.

Mischak, H., B. A. Julian and J. Novak (2007). "High-resolution proteome/peptidome analysis of peptides and low-molecular-weight proteins in urine." Proteomics Clin Appl 1(8): 792.

Mischak, H., W. Kolch, M. Aivaliotis, D. Bouyssie, M. Court, H. Dihazi, et al. (2010). "Comprehensive human urine standards for comparability and standardization in clinical proteome analysis." Proteomics Clinical Applications 4(4): 464-478.

Miyake, H., C. Nelson, P. S. Rennie and M. E. Gleave (2000). "Testosterone-repressed prostate message-2 is an antiapoptotic gene involved in progression to androgen independence in prostate cancer." Cancer Res 60(1): 170-176.

Mohamed, E., P. S. Abdul-Rahman, S. R. Doustjalali, Y. Chen, B. K. Lim, S. Z. Omar, et al. (2008). "Lectin-based electrophoretic analysis of the expression of the 35 kDa inter- alpha-trypsin inhibitor heavy chain H4 fragment in sera of patients with five different malignancies." Electrophoresis 29(12): 2645-2650.

Mohammadzadeh, G., M. A. Ghaffari, A. Bafandeh and S. M. Hosseini (2014). "Association of serum soluble leptin receptor and leptin levels with breast cancer." J Res Med Sci 19(5): 433-438.

Mohanraj, L. and Y. Oh (2011). "Targeting IGF-I, IGFBPs and IGF-I receptor system in cancer: the current and future in breast cancer therapy." Recent Pat Anticancer Drug Discov 6(2): 166-177.

Molina, R., V. Barak, A. van Dalen, M. J. Duffy, R. Einarsson, M. Gion, et al. (2005). "Tumor markers in breast cancer- European Group on Tumor Markers recommendations." Tumour Biol 26(6): 281-293.

Molloy, M. P., S. Bolis, B. R. Herbert, K. Ou, M. I. Tyler, D. D. van Dyk, et al. (1997). "Establishment of the human reflex tear two-dimensional polyacrylamide gel electrophoresis reference map: new proteins of potential diagnostic value." Electrophoresis 18(15): 2811-2815.

Morimoto-Tomita, M., K. Uchimura, A. Bistrup, D. H. Lum, M. Egeblad, N. Boudreau, et al. (2005). "Sulf-2, a proangiogenic heparan sulfate endosulfatase, is upregulated in breast cancer." Neoplasia 7(11): 1001-1010. 334

Mosca, R., T. Pons, A. Ceol, A. Valencia and P. Aloy (2013). "Towards a detailed atlas of protein-protein interactions." Curr Opin Struct Biol 23(6): 929-940.

Mukhina, S., H. C. Mertani, K. Guo, K. O. Lee, P. D. Gluckman and P. E. Lobie (2004). "Phenotypic conversion of human mammary carcinoma cells by autocrine human growth hormone." Proc Natl Acad Sci U S A 101(42): 15166-15171.

Munagala, R., F. Aqil and R. C. Gupta (2011). "Promising molecular targeted therapies in breast cancer." Indian J Pharmacol 43(3): 236-245.

Muraoka, S., H. Kume, S. Watanabe, J. Adachi, M. Kuwano, M. Sato, et al. (2012). "Strategy for SRM-based Verification of Biomarker Candidates Discovered by iTRAQ Method in Limited Breast Cancer Tissue Samples." J Proteome Res 11(8): 4201-4210.

Myklebust, M. P., O. Fluge, H. Immervoll, A. Skarstein, L. Balteskard, O. Bruland, et al. (2012). "Expression of DSG1 and DSC1 are prognostic markers in anal carcinoma patients." Br J Cancer 106(4): 756-762.

Nagaraj, N. and M. Mann (2011). "Quantitative analysis of the intra- and inter-individual variability of the normal urinary proteome." J Proteome Res 10(2): 637-645.

Nahta, R., M. C. Hung and F. J. Esteva (2004). "The HER-2-targeting antibodies trastuzumab and pertuzumab synergistically inhibit the survival of breast cancer cells." Cancer Res 64(7): 2343-2346.

Najam-ul-Haq, M., M. Rainer, L. Trojer, I. Feuerstein, R. M. Vallant, C. W. Huck, et al. (2007). "Alternative profiling platform based on MELDI and its applicability in clinical proteomics." Expert Rev Proteomics 4(4): 447-452.

Nakagawa, T., S. K. Huang, S. R. Martinez, A. N. Tran, D. Elashoff, X. Ye, et al. (2006). "Proteomic profiling of primary breast cancer predicts axillary lymph node metastasis." Cancer Research 66(24): 11825-11830.

Nakata, B., Y. Ogawa, T. Ishikawa, K. Ikeda, Y. Kato, H. Nishino, et al. (2000). "Serum CYFRA 21-1 is one of the most reliable tumor markers for breast carcinoma." Cancer 89(6): 1285-1290.

Nam, H., B. C. Chung, Y. Kim, K. Lee and D. Lee (2009). "Combining tissue transcriptomics and urine metabolomics for breast cancer biomarker identification." Bioinformatics 25(23): 3151-3157.

335

Nam, H., B. C. Chung, Y. Kim, K. Lee, D. Lee, H. Nam, et al. (2009). "Combining tissue transcriptomics and urine metabolomics for breast cancer biomarker identification." Bioinformatics 25(23): 3151-3157.

Nanjappa, V., J. K. Thomas, A. Marimuthu, B. Muthusamy, A. Radhakrishnan, R. Sharma, et al. (2014). "Plasma Proteome Database as a resource for proteomics research: 2014 update." Nucleic Acids Res 42(Database issue): D959-965.

Narod, S. A. and W. D. Foulkes (2004). "BRCA1 and BRCA2: 1994 and beyond." Nat Rev Cancer 4(9): 665-676.

Nasim, F. U., S. Ejaz, M. Ashraf, A. R. Asif, M. Oellerich, G. Ahmad, et al. (2012). "Potential biomarkers in the sera of breast cancer patients from bahawalpur, pakistan." Biomark Cancer 4: 19-34.

Nelson, H. D., K. Tyne, A. Naik, C. Bougatsos, B. K. Chan and L. Humphrey (2009). "Screening for breast cancer: an update for the U.S. Preventive Services Task Force." Ann Intern Med 151(10): 727-737, w237-742.

Nikitenko, L. L., R. Leek, S. Henderson, N. Pillay, H. Turley, D. Generali, et al. (2013). "The G-protein-coupled receptor CLR is upregulated in an autocrine loop with adrenomedullin in clear cell renal cell carcinoma and associated with poor prognosis." Clin Cancer Res 19(20): 5740-5748.

Njiaju, U. O. and O. I. Olopade (2012). "Genetic determinants of breast cancer risk: a review of current literature and issues pertaining to clinical application." Breast J 18(5): 436-442.

Noble, J. L., R. S. Dua, G. R. Coulton, C. M. Isacke and G. P. Gui (2007). "A comparative proteinomic analysis of nipple aspiration fluid from healthy women and women with breast cancer." European Journal of Cancer 43(16): 2315-2320.

Norum, L. F., B. Erikstein and K. Nustad (2001). "Elevated CA125 in breast cancer--A sign of advanced disease." Tumour Biol 22(4): 223-228.

Nutter, F., I. Holen, H. Brown, S. Cross, A. Evans, M. Walker, et al. (2014). "Different molecular profiles are associated with breast cancer bone homing compared to colonisation." Endocr Relat Cancer.

O'Farrell, P. H. (1975). "High resolution two-dimensional electrophoresis of proteins." The Journal of biological chemistry 250(10): 4007-4021.

336

O'Toole, S. A., C. I. Selinger, E. K. A. Millar, T. Lum and J. M. Beith (2011). "Molecular assays in breast cancer pathology." Pathology 43(2): 116-127.

Oehler, M. K., D. C. Fischer, M. Orlowska-Volk, F. Herrle, D. G. Kieback, M. C. Rees, et al. (2003). "Tissue and plasma expression of the angiogenic peptide adrenomedullin in breast cancer." Br J Cancer 89(10): 1927-1933.

Oh, J., J. H. Pyo, E. H. Jo, S. I. Hwang, S. C. Kang, J. H. Jung, et al. (2004). "Establishment of a near-standard two-dimensional human urine proteomic map." Proteomics 4(11): 3485-3497.

Okano, T., T. Kondo, T. Kakisaka, K. Fujii, M. Yamada, H. Kato, et al. (2006). "Plasma proteomics of lung cancer by a linkage of multi-dimensional liquid chromatography and two-dimensional difference gel electrophoresis." Proteomics 6(13): 3938-3948.

Omenn, G. S., D. J. States, M. Adamski, T. W. Blackwell, R. Menon, H. Hermjakob, et al. (2005). "Overview of the HUPO Plasma Proteome Project: results from the pilot phase with 35 collaborating laboratories and multiple analytical groups, generating a core dataset of 3020 proteins and a publicly-available database." Proteomics 5(13): 3226- 3245.

Ong, S. E., B. Blagoev, I. Kratchmarova, D. B. Kristensen, H. Steen, A. Pandey, et al. (2002). "Stable isotope labeling by amino acids in cell culture, SILAC, as a simple and accurate approach to expression proteomics." Molecular & Cellular Proteomics 1(5): 376- 386.

Ong, S. E. and M. Mann (2005). "Mass spectrometry-based proteomics turns quantitative." Nat Chem Biol 1(5): 252-262.

Opstal-van Winden, A. W., E. J. Krop, M. H. Karedal, M. C. Gast, C. H. Lindh, M. C. Jeppsson, et al. (2011). "Searching for early breast cancer biomarkers by serum protein profiling of pre-diagnostic serum; a nested case-control study." BMC Cancer 11: 381. 337

Opstal-van Winden, A. W., W. Rodenburg, J. L. Pennings, C. T. van Oostrom, J. H. Beijnen, P. H. Peeters, et al. (2012). "A bead-based multiplexed immunoassay to evaluate breast cancer biomarkers for early detection in pre-diagnostic serum." Int J Mol Sci 13(10): 13587-13604.

Opstal-van Winden, A. W., R. C. Vermeulen, P. H. Peeters, J. H. Beijnen and C. H. van Gils (2012). "Early diagnostic protein biomarkers for breast cancer: how far have we come?" Breast Cancer Res Treat 134(1): 1-12.

Orenes-Pinero, E., M. Corton, P. Gonzalez-Peramato, F. Algaba, I. Casal, A. Serrano, et al. (2007). "Searching urinary tumor markers for bladder cancer using a two-dimensional differential gel electrophoresis (2D-DIGE) approach." Journal of Proteome Research 6(11): 4440-4448.

Palmblad, M., A. Tiss and R. Cramer (2009). "Mass spectrometry in clinical proteomics - from the present to the future." Proteomics Clin Appl 3(1): 6-17.

Pan, S., R. Chen, R. Aebersold and T. A. Brentnall (2011). "Mass spectrometry based glycoproteomics--from a proteomics perspective." Molecular & Cellular Proteomics 10(1): R110.003251.

Panis, C., V. J. Victorino, A. C. Herrera, L. F. Freitas, T. De Rossi, F. C. Campos, et al. (2012). "Differential oxidative status and immune characterization of the early and advanced stages of human breast cancer." Breast Cancer Res Treat 133(3): 881-888.

Pankaj, J., J. R. Kumari, W. Kim and S. A. Lee (2015). "Insulin-like Growth Factor-1, IGF-binding Protein-3, C-peptide and Colorectal Cancer: a Case-control Study." Asian Pac J Cancer Prev 16(9): 3735-3740.

Parekh, R. S., W. H. Kao, L. A. Meoni, E. Ipp, P. L. Kimmel, J. La Page, et al. (2007). "Reliability of urinary albumin, total protein, and creatinine assays after prolonged storage: the Family Investigation of Nephropathy and Diabetes." Clin J Am Soc Nephrol 2(6): 1156-1162.

Park, I. J., G. S. Choi and S. H. Jun (2009). "Prognostic value of serum tumor antigen CA19-9 after curative resection of colorectal cancer." Anticancer Res 29(10): 4303-4308.

Parker, B. S., D. R. Ciocca, B. N. Bidwell, F. E. Gago, M. A. Fanelli, J. George, et al. (2008). "Primary tumour expression of the cysteine cathepsin inhibitor Stefin A inhibits distant metastasis in breast cancer." J Pathol 214(3): 337-346.

338

Parkin, D. M., L. M. Fernandez, D. M. Parkin and L. M. G. Fernandez (2006). "Use of statistics to assess the global burden of breast cancer." Breast Journal 12 Suppl 1: S70- 80.

Partridge, A. H., O. Pagani, O. Abulkhair, S. Aebi, F. Amant, H. A. Azim, Jr., et al. (2014). "First international consensus guidelines for breast cancer in young women (BCY1)." Breast 23(3): 209-220.

Paulo, J. A., L. S. Lee, P. A. Banks, H. Steen and D. L. Conwell (2011). "Difference gel electrophoresis identifies differentially expressed proteins in endoscopically collected pancreatic fluid." Electrophoresis 32(15): 1939-1951.

Pavlou, M. P., A. Dimitromanolakis and E. P. Diamandis (2013). "Coupling proteomics and transcriptomics in the quest of subtype-specific proteins in breast cancer." Proteomics 13(7): 1083-1095.

Pavlova, N. N., C. Pallasch, A. E. Elia, C. J. Braun, T. F. Westbrook, M. Hemann, et al. (2013). "A role for PVRL4-driven cell-cell interactions in tumorigenesis." Elife 2: e00358.

Paweletz, C. P., B. Trock, M. Pennanen, T. Tsangaris, C. Magnant, L. A. Liotta, et al. (2001). "Proteomic patterns of nipple aspirate fluids obtained by SELDI-TOF: potential for new biomarkers to aid in the diagnosis of breast cancer." Dis Markers 17(4): 301-307.

Pawlik, T. M., H. Fritsche, K. R. Coombes, L. Xiao, S. Krishnamurthy, K. K. Hunt, et al. (2005). "Significant differences in nipple aspirate fluid protein expression between healthy women and those with breast cancer demonstrated by time-of-flight mass spectrometry." Breast Cancer Research & Treatment 89(2): 149-157.

Pawlik, T. M., D. H. Hawke, Y. Liu, S. Krishnamurthy, H. Fritsche, K. K. Hunt, et al. (2006). "Proteomic analysis of nipple aspirate fluid from women with early-stage breast cancer using isotope-coded affinity tags and tandem mass spectrometry reveals differential expression of vitamin D binding protein." BMC Cancer 6: 68.

Payne, S. J., R. L. Bowen, J. L. Jones and C. A. Wells (2008). "Predictive markers in breast cancer--the present." Histopathology 52(1): 82-90.

Pellerin, L., J. Henry, C. Y. Hsu, S. Balica, C. Jean-Decoster, M. C. Mechin, et al. (2013). "Defects of filaggrin-like proteins in both lesional and nonlesional atopic skin." J Allergy Clin Immunol 131(4): 1094-1102.

339

Penault-Llorca, F., M. Bilous, M. Dowsett, W. Hanna, R. Y. Osamura, J. Ruschoff, et al. (2009). "Emerging technologies for assessing HER2 amplification." American Journal of Clinical Pathology 132(4): 539-548.

Peng, L., J. Liu, Y. M. Li, Z. L. Huang, P. P. Wang, Y. R. Gu, et al. (2013). "Serum proteomics analysis and comparisons using iTRAQ in the progression of hepatitis B." Exp Ther Med 6(5): 1169-1176.

Perkins, G. L., E. D. Slater, G. K. Sanders and J. G. Prichard (2003). "Serum tumor markers." Am Fam Physician 68(6): 1075-1082.

Perks, C. M. and J. M. Holly (2015). "Epigenetic regulation of insulin-like growth factor binding protein-3 (IGFBP-3) in cancer." J Cell Commun Signal 9(2): 159-166.

Perou, C. M. and A. L. Borresen-Dale (2011). "Systems biology and genomics of breast cancer." Cold Spring Harb Perspect Biol 3(2).

Perou, C. M., T. Sorlie, M. B. Eisen, M. van de Rijn, S. S. Jeffrey, C. A. Rees, et al. (2000). "Molecular portraits of human breast tumours." Nature 406(6797): 747-752.

Petricoin, E. F., A. M. Ardekani, B. A. Hitt, P. J. Levine, V. A. Fusaro, S. M. Steinberg, et al. (2002). "Use of proteomic patterns in serum to identify ovarian cancer." Lancet 359(9306): 572-577.

Petricoin, E. F., C. Belluco, R. P. Araujo and L. A. Liotta (2006). "The blood peptidome: a higher dimension of information content for cancer biomarker discovery." Nat Rev Cancer 6(12): 961-967.

Petricoin, E. F. and L. A. Liotta (2004). "SELDI-TOF-based serum proteomic pattern diagnostics for early detection of cancer." Current Opinion in Biotechnology 15(1): 24- 30.

Pharoah, P. D., A. Antoniou, M. Bobrow, R. L. Zimmern, D. F. Easton and B. A. Ponder (2002). "Polygenic susceptibility to breast cancer and implications for prevention." Nat Genet 31(1): 33-36.

Piccart-Gebhart, M. J., M. Procter, B. Leyland-Jones, A. Goldhirsch, M. Untch, I. Smith, et al. (2005). "Trastuzumab after adjuvant chemotherapy in HER2-positive breast cancer." N Engl J Med 353(16): 1659-1672.

340

Pieper, R. (2008). "Preparation of urine samples for proteomic analysis." Methods in Molecular Biology 425: 89-99.

Pieper, R., C. L. Gatlin, A. J. Makusky, P. S. Russo, C. R. Schatz, S. S. Miller, et al. (2003). "The human serum proteome: display of nearly 3700 chromatographically separated protein spots on two-dimensional electrophoresis gels and identification of 325 distinct proteins." Proteomics 3(7): 1345-1364.

Pieper, R., C. L. Gatlin, A. M. McGrath, A. J. Makusky, M. Mondal, M. Seonarain, et al. (2004). "Characterization of the human urinary proteome: a method for high-resolution display of urinary proteins on two-dimensional electrophoresis gels with a yield of nearly 1400 distinct protein spots." Proteomics 4(4): 1159-1174.

Pieper, R., Q. Su, C. L. Gatlin, S. T. Huang, N. L. Anderson and S. Steiner (2003). "Multi- component immunoaffinity subtraction chromatography: an innovative step towards a comprehensive survey of the human plasma proteome." Proteomics 3(4): 422-432.

Pierce, A., R. D. Unwin, C. A. Evans, S. Griffiths, L. Carney, L. Zhang, et al. (2008). "Eight-channel iTRAQ enables comparison of the activity of six leukemogenic tyrosine kinases." Molecular & Cellular Proteomics 7(5): 853-863.

Pierce, B. L., R. Ballard-Barbash, L. Bernstein, R. N. Baumgartner, M. L. Neuhouser, M. H. Wener, et al. (2009). "Elevated biomarkers of inflammation are associated with reduced survival among breast cancer patients." J Clin Oncol 27(21): 3437-3444.

Pietrowska, M., L. Marczak, J. Polanska, K. Behrendt, E. Nowicka, A. Walaszczyk, et al. (2009). "Mass spectrometry-based serum proteome pattern analysis in molecular diagnostics of early stage breast cancer." J Transl Med 7: 60.

Pietrowska, M., J. Polanska, L. Marczak, K. Behrendt, E. Nowicka, M. Stobiecki, et al. (2010). "Mass spectrometry-based analysis of therapy-related changes in serum proteome patterns of patients with early-stage breast cancer." J Transl Med 8: 66.

Pirazzoli, V., G. M. Ferraris and N. Sidenius (2013). "Direct evidence of the importance of vitronectin and its interaction with the urokinase receptor in tumor growth." Blood 121(12): 2316-2323.

341

Pisitkun, T., J. Bieniek, D. Tchapyjnikov, G. Wang, W. W. Wu, R. F. Shen, et al. (2006). "High-throughput identification of IMCD proteins using LC-MS/MS." Physiol Genomics 25(2): 263-276.

Pisitkun, T., R. Johnstone, M. A. Knepper, T. Pisitkun, R. Johnstone and M. A. Knepper (2006). "Discovery of urinary biomarkers." Molecular & Cellular Proteomics 5(10): 1760-1771.

Pisitkun, T., R. F. Shen and M. A. Knepper (2004). "Identification and proteomic profiling of exosomes in human urine." Proc Natl Acad Sci U S A 101(36): 13368-13373.

Pitteri, S. J., L. M. Amon, T. Busald Buson, Y. Zhang, M. M. Johnson, A. Chin, et al. (2010). "Detection of elevated plasma levels of epidermal growth factor receptor before breast cancer diagnosis among hormone therapy users." Cancer Res 70(21): 8598-8606.

Polley, M. Y., S. C. Leung, L. M. McShane, D. Gao, J. C. Hugh, M. G. Mastropasqua, et al. (2013). "An international Ki67 reproducibility study." J Natl Cancer Inst 105(24): 1897-1906.

Polyak, K. (2007). "Breast cancer: origins and evolution." J Clin Invest 117(11): 3155- 3163.

Porter, D., S. Weremowicz, K. Chin, P. Seth, A. Keshaviah, J. Lahti-Domenici, et al. (2003). "A neural survival factor is a candidate oncogene in breast cancer." Proc Natl Acad Sci U S A 100(19): 10931-10936.

Probst-Hensch, N. M., J. H. Steiner, P. Schraml, Z. Varga, U. Zurrer-Hardi, M. Storz, et al. (2010). "IGFBP2 and IGFBP3 protein expressions in human breast cancer: association with hormonal factors and obesity." Clin Cancer Res 16(3): 1025-1032.

Profumo, A., R. Mangerini, A. Rubagotti, P. Romano, G. Damonte, P. Guglielmini, et al. (2013). "Complement C3f serum levels may predict breast cancer risk in women with gross cystic disease of the breast." J Proteomics 85: 44-52.

Provatopoulou, X., A. Gounaris, E. Kalogera, F. Zagouri, I. Flessas, E. Goussetis, et al. (2009). "Circulating levels of matrix metalloproteinase-9 (MMP-9), neutrophil gelatinase-associated lipocalin (NGAL) and their complex MMP-9/NGAL in breast cancer disease." BMC Cancer 9: 390.

342

Pucci, S., E. Bonanno, F. Pichiorri, C. Angeloni and L. G. Spagnoli (2004). "Modulation of different clusterin isoforms in human colon tumorigenesis." Oncogene 23(13): 2298- 2304.

Pusztai, L., B. W. Gregory, K. A. Baggerly, B. Peng, J. Koomen, H. M. Kuerer, et al. (2004). "Pharmacoproteomic analysis of prechemotherapy and postchemotherapy plasma samples from patients receiving neoadjuvant or adjuvant chemotherapy for breast carcinoma." Cancer 100(9): 1814-1822.

Qian, W. J., J. M. Jacobs, T. Liu, D. G. Camp, 2nd and R. D. Smith (2006). "Advances and challenges in liquid chromatography-mass spectrometry-based proteomics profiling for clinical applications." Mol Cell Proteomics 5(10): 1727-1744.

Qureshi, H. S., M. D. Linden, G. Divine and U. B. Raju (2006). "E-cadherin status in breast cancer correlates with histologic type but does not correlate with established prognostic parameters." Am J Clin Pathol 125(3): 377-385.

Rai, A. J., C. A. Gelfand, B. C. Haywood, D. J. Warunek, J. Yi, M. D. Schuchard, et al. (2005). "HUPO Plasma Proteome Project specimen collection and handling: towards the standardization of parameters for plasma proteome samples." Proteomics 5(13): 3262- 3277.

Rakha, E. A., S. E. Pinder, E. C. Paish, J. F. Robertson and I. O. Ellis (2004). "Expression of E2F-4 in invasive breast carcinomas is associated with poor prognosis." J Pathol 203(3): 754-761.

Ramautar, R., E. Nevedomskaya, O. A. Mayboroda, A. M. Deelder, I. D. Wilson, H. G. Gika, et al. (2011). "Metabolic profiling of human urine by CE-MS using a positively charged capillary coating and comparison with UPLC-MS." Molecular BioSystems 7(1): 194-199.

Randall, S. A., M. J. McKay, M. S. Baker and M. P. Molloy (2010). "Evaluation of blood collection tubes using selected reaction monitoring MS: implications for proteomic biomarker studies.[Erratum appears in Proteomics. 2011 Dec;11(23):4593 Note: Baker, Mark S [added]]." Proteomics 10(10): 2050-2056.

Ranogajec, I., J. Jakic-Razumovic, V. Puzovic and J. Gabrilovac (2012). "Prognostic value of matrix metalloproteinase-2 (MMP-2), matrix metalloproteinase-9 (MMP-9) and aminopeptidase N/CD13 in breast cancer patients." Med Oncol 29(2): 561-569.

343

Rappsilber, J., Y. Ishihama and M. Mann (2003). "Stop and go extraction tips for matrix- assisted laser desorption/ionization, nanoelectrospray, and LC/MS sample pretreatment in proteomics." Analytical Chemistry 75(3): 663-670.

Ray, S., P. J. Reddy, R. Jain, K. Gollapalli, A. Moiyadi and S. Srivastava (2011). "Proteomic technologies for the identification of disease biomarkers in serum: advances and challenges ahead." Proteomics 11(11): 2139-2161.

Redondo, M., M. Garcia-Aranda, M. J. Roldan, G. Callejon, A. Serrano, E. Jimenez, et al. (2015). "Downregulation of clusterin mediates sensitivity to protein kinase inhibitors in breast cancer cells." Anticancer Drugs 26(1): 85-89.

Redondo, M., T. Tellez, M. J. Roldan, A. Serrano, M. Garcia-Aranda, M. E. Gleave, et al. (2007). "Anticlusterin treatment of breast cancer cells increases the sensitivities of chemotherapy and tamoxifen and counteracts the inhibitory action of dexamethasone on chemotherapy-induced cytotoxicity." Breast Cancer Res 9(6): R86.

Redondo, M., E. Villar, J. Torres-Munoz, T. Tellez, M. Morell and C. K. Petito (2000). "Overexpression of clusterin in human breast carcinoma." Am J Pathol 157(2): 393-399.

Reeves, J. R., H. Dulude, C. Panchal, L. Daigneault and D. M. Ramnani (2006). "Prognostic value of prostate secretory protein of 94 amino acids and its binding protein after radical prostatectomy." Clin Cancer Res 12(20 Pt 1): 6018-6022.

Rehman, I., A. R. Azzouzi, J. W. Catto, S. Allen, S. S. Cross, K. Feeley, et al. (2004). "Proteomic analysis of voided urine after prostatic massage from patients with prostate cancer: a pilot study." Urology 64(6): 1238-1243.

Rezaul, K., J. K. Thumar, D. H. Lundgren, J. K. Eng, K. P. Claffey, L. Wilson, et al. (2010). "Differential protein expression profiles in estrogen receptor-positive and - negative breast cancer tissues using label-free quantitative proteomics." Genes Cancer 1(3): 251-271.

Rho, J. H., S. Qin, J. Y. Wang and M. H. Roehrl (2008). "Proteomic expression analysis of surgical human colorectal cancer tissues: up-regulation of PSB7, PRDX1, and SRP9 and hypoxic adaptation in cancer." J Proteome Res 7(7): 2959-2972.

Rich, J. T., J. G. Neely, R. C. Paniello, C. C. Voelker, B. Nussenbaum and E. W. Wang (2010). "A practical guide to understanding Kaplan-Meier curves." Otolaryngol Head Neck Surg 143(3): 331-336.

344

Ricolleau, G., C. Charbonnel, L. Lode, D. Loussouarn, M. P. Joalland, R. Bogumil, et al. (2006). "Surface-enhanced laser desorption/ionization time of flight mass spectrometry protein profiling identifies ubiquitin and ferritin light chain as prognostic biomarkers in node-negative breast cancer tumors." Proteomics 6(6): 1963-1975.

Rittenhouse, H. G., J. A. Finlay, S. D. Mikolajczyk and A. W. Partin (1998). "Human Kallikrein 2 (hK2) and prostate-specific antigen (PSA): two closely related, but distinct, kallikreins in the prostate." Crit Rev Clin Lab Sci 35(4): 275-368.

Robinson, D. R., S. Kalyana-Sundaram, Y. M. Wu, S. Shankar, X. Cao, B. Ateeq, et al. (2011). "Functionally recurrent rearrangements of the MAST kinase and Notch gene families in breast cancer." Nat Med 17(12): 1646-1651.

Roder, D., N. Houssami, G. Farshid, G. Gill, C. Luke, P. Downey, et al. (2008). "Population screening and intensity of screening are associated with reduced breast cancer mortality: evidence of efficacy of mammography screening in Australia." Breast Cancer Res Treat 108(3): 409-416.

Rodrigues, L. R., J. A. Teixeira, F. L. Schmitt, M. Paulsson and H. Lindmark-Mansson (2007). "The role of osteopontin in tumor progression and metastasis in breast cancer." Cancer Epidemiol Biomarkers Prev 16(6): 1087-1097.

Rompp, A., L. Dekker, I. Taban, G. Jenster, W. Boogerd, H. Bonfrer, et al. (2007). "Identification of leptomeningeal metastasis-related proteins in cerebrospinal fluid of patients with breast cancer by a combination of MALDI-TOF, MALDI-FTICR and nanoLC-FTICR MS." Proteomics 7(3): 474-481.

Rose, K., L. Bougueleret, T. Baussant, G. Bohm, P. Botti, J. Colinge, et al. (2004). "Industrial-scale proteomics: from liters of plasma to chemically synthesized proteins." Proteomics 4(7): 2125-2150.

Rosenblatt, K. P., P. Bryant-Greenwood, J. K. Killian, A. Mehta, D. Geho, V. Espina, et al. (2004). "Serum proteomics in cancer diagnosis and management." Annu Rev Med 55: 97-112.

Ross, P. L., Y. N. Huang, J. N. Marchese, B. Williamson, K. Parker, S. Hattan, et al. (2004). "Multiplexed protein quantitation in Saccharomyces cerevisiae using amine- reactive isobaric tagging reagents." Molecular & Cellular Proteomics 3(12): 1154-1169.

Rouzier, R., C. M. Perou, W. F. Symmans, N. Ibrahim, M. Cristofanilli, K. Anderson, et al. (2005). "Breast cancer molecular subtypes respond differently to preoperative chemotherapy." Clin Cancer Res 11(16): 5678-5685. 345

Rower, C., C. Koy, M. Hecker, T. Reimer, B. Gerber, H. J. Thiesen, et al. (2011). "Mass spectrometric characterization of protein structure details refines the proteome signature for invasive ductal breast carcinoma." Journal of the American Society for Mass Spectrometry 22(3): 440-456.

Rudland, P. S., A. Platt-Higgins, M. El-Tanani, S. De Silva Rudland, R. Barraclough, J. H. Winstanley, et al. (2002). "Prognostic significance of the metastasis-associated protein osteopontin in human breast cancer." Cancer Res 62(12): 3417-3427.

Rui, Z., J. Jian-Guo, T. Yuan-Peng, P. Hai and R. Bing-Gen (2003). "Use of serological proteomic methods to find biomarkers associated with breast cancer." Proteomics 3(4): 433-439.

Ruiz-Garcia, E., V. Scott, C. Machavoine, J. M. Bidart, L. Lacroix, S. Delaloge, et al. (2010). "Gene expression profiling identifies Fibronectin 1 and CXCL9 as candidate biomarkers for breast cancer screening." Br J Cancer 102(3): 462-468.

Russell, T. D., S. Jindal, S. Agunbiade, D. Gao, M. Troxell, V. F. Borges, et al. (2015). "Myoepithelial Cell Differentiation Markers in Ductal Carcinoma in Situ Progression." Am J Pathol.

Sadagopan, N. P., W. Li, J. A. Cook, B. Galvan, D. L. Weller, S. T. Fountain, et al. (2003). "Investigation of EDTA anticoagulant in plasma to improve the throughput of liquid chromatography/tandem mass spectrometric assays." Rapid Commun Mass Spectrom 17(10): 1065-1070.

Sahab, Z. J., Y. G. Man, S. M. Semaan, R. G. Newcomer, S. W. Byers and Q. X. Sang (2010). "Alteration in protein expression in estrogen receptor alpha-negative human breast cancer tissues indicates a malignant and metastatic phenotype." Clin Exp Metastasis 27(7): 493-503.

Said, J. (2005). "Biomarker discovery in urogenital cancer." Biomarkers 10 Suppl 1: S83- 86.

Saini, S., N. Jagadish, A. Gupta, A. Bhatnagar and A. Suri (2013). "A novel cancer testis antigen, A-kinase anchor protein 4 (AKAP4) is a potential biomarker for breast cancer." PLoS One 8(2): e57095.

Samy, N., H. M. Ragab, N. A. El Maksoud and M. Shaalan (2010). "Prognostic significance of serum Her2/neu, BCL2, CA15-3 and CEA in breast cancer patients: a short follow-up." Cancer Biomark 6(2): 63-72.

346

Sanders, M. E., E. C. Dias, B. J. Xu, J. A. Mobley, D. Billheimer, H. Roder, et al. (2008). "Differentiating proteomic biomarkers in breast cancer by laser capture microdissection and MALDI MS." J Proteome Res 7(4): 1500-1507.

Sauter, E. R., S. Shan, J. E. Hewett, P. Speckman and G. C. Du Bois (2005). "Proteomic analysis of nipple aspirate fluid using SELDI-TOF-MS." International Journal of Cancer 114(5): 791-796.

Sauter, E. R., W. Zhu, X. J. Fan, R. P. Wassell, I. Chervoneva and G. C. Du Bois (2002). "Proteomic analysis of nipple aspirate fluid to detect biologic markers of breast cancer." Br J Cancer 86(9): 1440-1443.

Scharenberg, C., A. Eckardt, C. Tiede, H. Kreipe and K. Hussein (2013). "Expression of caspase 14 and filaggrin in oral squamous carcinoma." Head Neck Pathol 7(4): 327-333.

Schaub, N. P., K. J. Jones, J. O. Nyalwidhe, L. H. Cazares, I. D. Karbassi, O. J. Semmes, et al. (2009). "Serum proteomic biomarker discovery reflective of stage and obesity in breast cancer patients." J Am Coll Surg 208(5): 970-978; discussion 978-980.

Schaub, S., J. Wilkins, T. Weiler, K. Sangster, D. Rush and P. Nickerson (2004). "Urine protein profiling with surface-enhanced laser-desorption/ionization time-of-flight mass spectrometry." Kidney International 65(1): 323-332.

Schoenherr, R. M., K. S. Kelly-Spratt, C. Lin, J. R. Whiteaker, T. Liu, T. Holzman, et al. (2011). "Proteome and transcriptome profiles of a Her2/Neu-driven mouse model of breast cancer." Proteomics Clin Appl 5(3-4): 179-188.

Schulz, D. M., C. Bollner, G. Thomas, M. Atkinson, I. Esposito, H. Hofler, et al. (2009). "Identification of differentially expressed proteins in triple-negative breast carcinomas using DIGE and mass spectrometry." J Proteome Res 8(7): 3430-3438.

347

See, A. L., P. K. Chong, S. Y. Lu and Y. P. Lim (2014). "CXCL3 is a potential target for breast cancer metastasis." Curr Cancer Drug Targets 14(3): 294-309.

Semaan, S. M. and Q. X. Sang (2011). "Prefractionation enhances loading capacity and identification of basic proteins from human breast cancer tissues." Analytical Biochemistry 411(1): 80-87.

Semaan, S. M., X. Wang, A. G. Marshall and Q. X. Sang (2012). "Identification of Potential Glycoprotein Biomarkers in Estrogen Receptor Positive (ER+) and Negative (ER-) Human Breast Cancer Tissues by LC-LTQ/FT-ICR Mass Spectrometry." J Cancer 3: 269-284.

Shaw, P. G., R. Chaerkady, T. Wang, S. Vasilatos, Y. Huang, B. Van Houten, et al. (2013). "Integrated proteomic and metabolic analysis of breast cancer progression." PLoS One 8(9): e76220.

Shen, D., H. R. Chang, Z. Chen, J. He, V. Lonsberry, Y. Elshimali, et al. (2005). "Loss of annexin A1 expression in human breast cancer detected by multiple high-throughput analyses." Biochem Biophys Res Commun 326(1): 218-227.

Shen, Y., J. Kim, E. F. Strittmatter, J. M. Jacobs, D. G. Camp, 2nd, R. Fang, et al. (2005). "Characterization of the human blood plasma proteome." Proteomics 5(15): 4034-4045.

Sheri, A. and M. Dowsett (2012). "Developments in Ki67 and other biomarkers for treatment decision making in breast cancer." Ann Oncol 23 Suppl 10: x219-227.

Shevde, L. A., S. Das, D. W. Clark and R. S. Samant (2010). "Osteopontin: an effector and an effect of tumor metastasis." Curr Mol Med 10(1): 71-81.

Shield-Artin, K. L., M. J. Bailey, K. Oliva, A. K. Liovic, G. Barker, N. L. Dellios, et al. (2012). "Identification of ovarian cancer-associated proteins in symptomatic women: A novel method for semi-quantitative plasma proteomics." Proteomics Clin Appl 6(3-4): 170-181.

Shih, N. Y., H. L. Lai, G. C. Chang, H. C. Lin, Y. C. Wu, J. M. Liu, et al. (2010). "Anti- alpha-enolase autoantibodies are down-regulated in advanced cancer patients." Jpn J Clin Oncol 40(7): 663-669.

Shubbar, E., K. Helou, A. Kovacs, S. Nemes, S. Hajizadeh, C. Enerback, et al. (2013). "High levels of gamma-glutamyl hydrolase (GGH) are associated with poor prognosis and unfavorable clinical outcomes in invasive breast cancer." BMC Cancer 13: 47.

348

Sigdel, T. K., K. Lau, J. Schilling and M. Sarwal (2008). "Optimizing protein recovery for urinary proteomics, a tool to monitor renal transplantation." Clinical Transplantation 22(5): 617-623.

Singletary, S. E. (2003). "Rating the risk factors for breast cancer." Ann Surg 237(4): 474-482.

Singletary, S. E., C. Allred, P. Ashley, L. W. Bassett, D. Berry, K. I. Bland, et al. (2002). "Revision of the American Joint Committee on Cancer staging system for breast cancer." J Clin Oncol 20(17): 3628-3636.

Sloan, K. E., B. K. Eustace, J. K. Stewart, C. Zehetmeier, C. Torella, M. Simeone, et al. (2004). "CD155/PVR plays a key role in cell motility during tumor cell invasion and migration." BMC Cancer 4: 73.

Smith, L., K. J. Welham, M. B. Watson, P. J. Drew, M. J. Lind and L. Cawkwell (2007). "The proteomic analysis of cisplatin resistance in breast cancer cells." Oncology Research 16(11): 497-506.

Soggiu, A., C. Piras, L. Bonizzi, H. A. Hussein, S. Pisanu and P. Roncada (2012). "A discovery-phase urine proteomics investigation in type 1 diabetes." Acta Diabetol 49(6): 453-464.

Solassol, J., P. Rouanet, P. J. Lamy, C. Allal, G. Favre, T. Maudelonde, et al. (2010). "Serum protein signature may improve detection of ductal carcinoma in situ of the breast." Oncogene 29(4): 550-560.

Soltermann, A., R. Ossola, S. Kilgus-Hawelski, A. von Eckardstein, T. Suter, R. Aebersold, et al. (2008). "N-glycoprotein profiling of lung adenocarcinoma pleural effusions by shotgun proteomics." Cancer 114(2): 124-133.

Somiari, R. I., S. Somiari, S. Russell and C. D. Shriver (2005). "Proteomics of breast carcinoma." Journal of Chromatography B 815(1–2): 215-225.

Somiari, R. I., A. Sullivan, S. Russell, S. Somiari, H. Hu, R. Jordan, et al. (2003). "High- throughput proteomic analysis of human infiltrating ductal carcinoma of the breast." Proteomics 3(10): 1863-1873.

Song, M. N., P. G. Moon, J. E. Lee, M. Na, W. Kang, Y. S. Chae, et al. (2012). "Proteomic analysis of breast cancer tissues to identify biomarker candidates by gel-assisted digestion

349

and label-free quantification methods using LC-MS/MS." Arch Pharm Res 35(10): 1839- 1847.

Sorlie, T., C. M. Perou, R. Tibshirani, T. Aas, S. Geisler, H. Johnsen, et al. (2001). "Gene expression patterns of breast carcinomas distinguish tumor subclasses with clinical implications." Proc Natl Acad Sci U S A 98(19): 10869-10874.

Sorlie, T., R. Tibshirani, J. Parker, T. Hastie, J. S. Marron, A. Nobel, et al. (2003). "Repeated observation of breast tumor subtypes in independent gene expression data sets." Proc Natl Acad Sci U S A 100(14): 8418-8423.

Sorlie, T., Y. Wang, C. Xiao, H. Johnsen, B. Naume, R. R. Samaha, et al. (2006). "Distinct molecular mechanisms underlying clinically relevant subtypes of breast cancer: gene expression analyses across three different platforms." BMC Genomics 7: 127.

Spahr, C. S., M. T. Davis, M. D. McGinley, J. H. Robinson, E. J. Bures, J. Beierle, et al. (2001). "Towards defining the urinary proteome using liquid chromatography-tandem mass spectrometry. I. Profiling an unfractionated tryptic digest." Proteomics 1(1): 93-107.

Sprung, R. W., M. A. Martinez, K. L. Carpenter, A. J. Ham, M. K. Washington, C. L. Arteaga, et al. (2012). "Precision of Multiple Reaction Monitoring Mass Spectrometry Analysis of Formalin-Fixed, Paraffin-Embedded Tissue." J Proteome Res.

Stastny, J., R. Prasad and E. Fosslien (1984). "Tissue proteins in breast cancer, as studied by use of two-dimensional electrophoresis." Clinical Chemistry 30(12 Pt 1): 1914-1918.

States, D. J., G. S. Omenn, T. W. Blackwell, D. Fermin, J. Eng, D. W. Speicher, et al. (2006). "Challenges in deriving high-confidence protein identifications from data gathered by a HUPO plasma proteome collaborative study." Nature Biotechnology 24(3): 333-338.

Steen, H. and M. Mann (2004). "The ABC's (and XYZ's) of peptide sequencing." Nat Rev Mol Cell Biol 5(9): 699-711.

Stephens, P. J., P. S. Tarpey, H. Davies, P. Van Loo, C. Greenman, D. C. Wedge, et al. (2012). "The landscape of cancer genes and mutational processes in breast cancer." Nature 486(7403): 400-404.

Stewart, G. D., R. J. Skipworth, C. J. Pennington, A. G. Lowrie, D. A. Deans, D. R. Edwards, et al. (2008). "Variation in dermcidin expression in a range of primary human

350

tumours and in hypoxic/oxidatively stressed human cell lines." Br J Cancer 99(1): 126- 132.

Storey, J. D. and R. Tibshirani (2003). "Statistical significance for genomewide studies." Proc Natl Acad Sci U S A 100(16): 9440-9445.

Sturgeon, C. M., M. J. Duffy, U. H. Stenman, H. Lilja, N. Brunner, D. W. Chan, et al. (2008). "National Academy of Clinical Biochemistry laboratory medicine practice guidelines for use of tumor markers in testicular, prostate, colorectal, breast, and ovarian cancers." Clin Chem 54(12): e11-79.

Such-Sanmartin, G., N. Bache, A. K. Callesen, A. Rogowska-Wrzesinska and O. N. Jensen (2015). "Targeted mass spectrometry analysis of the proteins IGF1, IGF2, IBP2, IBP3 and A2GL by blood protein precipitation." J Proteomics 113: 29-37.

Suh, E. J., M. H. Kabir, U. B. Kang, J. W. Lee, J. Yu, D. Y. Noh, et al. (2012). "Comparative profiling of plasma proteome from breast cancer patients reveals thrombospondin-1 and BRWD3 as serological biomarkers." Exp Mol Med 44(1): 36-44.

Sun, B., S. Zhang, D. Zhang, Y. Li, X. Zhao, Y. Luo, et al. (2008). "Identification of metastasis-related proteins and their clinical relevance to triple-negative human breast cancer." Clinical Cancer Research 14(21): 7050-7059.

Sun, L., S. Gu, X. Li, Y. Sun, D. Zheng, K. Yu, et al. (2006). "[Identification of a novel human MAST4 gene, a new member of the microtubule associated serine-threonine kinase family]." Mol Biol (Mosk) 40(5): 808-815.

Sun, W., F. Li, S. Wu, X. Wang, D. Zheng, J. Wang, et al. (2005). "Human urine proteome analysis by three separation approaches." Proteomics 5(18): 4994-5001.

Surowiak, P., M. Drag, V. Materna, S. Suchocki, R. Grzywa, M. Spaczynski, et al. (2006). "Expression of aminopeptidase N/CD13 in human ovarian cancers." Int J Gynecol Cancer 16(5): 1783-1788.

Szabo, G. T., R. Tihanyi, F. Csulak, E. Jambor, A. Bona, G. Szabo, et al. (2012). "Comparative salivary proteomics of cleft palate patients." Cleft Palate Craniofac J 49(5): 519-523.

351

Tabassum, U., O. Reddy and G. Mukherjee (2012). "Elevated serum haptoglobin is associated with clinical outcome in triple-negative breast cancer patients." Asian Pac J Cancer Prev 13(9): 4541-4544.

Tagi, T., T. Matsui, S. Kikuchi, S. Hoshi, T. Ochiai, Y. Kokuba, et al. (2010). "Dermokine as a novel biomarker for early-stage colorectal cancer." J Gastroenterol 45(12): 1201- 1211.

Takahashi, N., Y. Takahashi and F. W. Putnam (1985). "Periodicity of leucine and tandem repetition of a 24-amino acid segment in the primary structure of leucine-rich alpha 2- glycoprotein of human serum." Proc Natl Acad Sci U S A 82(7): 1906-1910.

Takita, J., Y. Chen, J. Okubo, M. Sanada, M. Adachi, K. Ohki, et al. (2011). "Aberrations of NEGR1 on 1p31 and MYEOV on 11q13 in neuroblastoma." Cancer Sci 102(9): 1645- 1650.

Tang, S. S. and G. P. Gui (2012). "Biomarkers in the diagnosis of primary and recurrent breast cancer." Biomark Med 6(5): 567-585.

Tantipaiboonwong, P., S. Sinchaikul, S. Sriyam, S. Phutrakul, S. T. Chen, P. Tantipaiboonwong, et al. (2005). "Different techniques for urinary protein analysis of normal and lung cancer patients." Proteomics 5(4): 1140-1149.

Tas, F., S. Karabulut, E. Bilgin, D. Tastekin and D. Duranyildiz (2014). "Clinical significance of serum insulin-like growth factor-1 (IGF-1) and insulin-like growth factor binding protein-3 (IGFBP-3) in patients with breast cancer." Tumour Biol 35(9): 9303- 9309.

Tessitore, L., B. Vizio, O. Jenkins, I. De Stefano, C. Ritossa, J. M. Argiles, et al. (2000). "Leptin expression in colorectal and breast cancer patients." Int J Mol Med 5(4): 421-426.

Thadikkaran, L., M. A. Siegenthaler, D. Crettaz, P. A. Queloz, P. Schneider and J. D. Tissot (2005). "Recent advances in blood-related proteomics." Proteomics 5(12): 3019- 3034.

The Cancer Genome Atlas Network (2012). "Comprehensive molecular portraits of human breast tumours." Nature 490(7418): 61-70.

Theodorescu, D., D. Fliser, S. Wittke, H. Mischak, R. Krebs, M. Walden, et al. (2005). "Pilot study of capillary electrophoresis coupled to mass spectrometry as a tool to define potential prostate cancer biomarkers in urine." Electrophoresis 26(14): 2797-2808.

352

Theodorescu, D., S. Wittke, M. M. Ross, M. Walden, M. Conaway, I. Just, et al. (2006). "Discovery and validation of new protein biomarkers for urothelial cancer: a prospective analysis." Lancet Oncol 7(3): 230-240.

Theodoropoulos, G. E., V. Saridakis, T. Karantanos, N. V. Michalopoulos, F. Zagouri, P. Kontogianni, et al. (2012). "Toll-like receptors gene polymorphisms may confer increased susceptibility to breast cancer development." Breast 21(4): 534-538.

Thomas, C. E., W. Sexton, K. Benson, R. Sutphen and J. Koomen (2010). "Urine collection and processing for protein biomarker discovery and quantification. [Review] [67 refs]." Cancer Epidemiology, Biomarkers & Prevention 19(4): 953-959.

Thongboonkerd, V. (2007). "Practical points in urinary proteomics." Journal of Proteome Research 6(10): 3881-3890.

Thongboonkerd, V., S. Chutipongtanate, R. Kanlaya, V. Thongboonkerd, S. Chutipongtanate and R. Kanlaya (2006). "Systematic evaluation of sample preparation methods for gel-based human urinary proteomics: quantity, quality, and variability." Journal of Proteome Research 5(1): 183-191.

Thongboonkerd, V., J. B. Klein, A. W. Jevans and K. R. McLeish (2004). "Urinary proteomics and biomarker discovery for glomerular diseases." Contributions to Nephrology 141: 292-307.

Thongboonkerd, V. and P. Malasit (2005). "Renal and urinary proteomics: current applications and challenges." Proteomics 5(4): 1033-1042.

Thongboonkerd, V., K. R. McLeish, J. M. Arthur and J. B. Klein (2002). "Proteomic analysis of normal human urinary proteins isolated by acetone precipitation or ultracentrifugation." Kidney International 62(4): 1461-1469.

Thongboonkerd, V., S. Mungdee and W. Chiangjong (2009). "Should urine pH be adjusted prior to gel-based proteome analysis?" Journal of Proteome Research 8(6): 3206- 3211.

Thongboonkerd, V. and V. Thongboonkerd (2007). "Practical points in urinary proteomics." Journal of Proteome Research 6(10): 3881-3890.

Thongboonkerd, V. and V. Thongboonkerd (2008). "Urinary proteomics: towards biomarker discovery, diagnostics and prognostics." Molecular Biosystems 4(8): 810-815.

353

Tikk, K., D. Sookthai, T. Johnson, S. Rinaldi, I. Romieu, A. Tjonneland, et al. (2014). "Circulating prolactin and breast cancer risk among pre- and postmenopausal women in the EPIC cohort." Ann Oncol 25(7): 1422-1428.

Tilanus-Linthorst, M., L. Verhoog, I. M. Obdeijn, K. Bartels, M. Menke-Pluymers, A. Eggermont, et al. (2002). "A BRCA1/2 mutation, high breast density and prominent pushing margins of a tumor independently contribute to a frequent false-negative mammography." Int J Cancer 102(1): 91-95.

Timms, J. F. and R. Cramer (2008). "Difference gel electrophoresis." Proteomics 8(23- 24): 4886-4897.

Torre, L. A., F. Bray, R. L. Siegel, J. Ferlay, J. Lortet-Tieulent and A. Jemal (2015). "Global cancer statistics, 2012." CA Cancer J Clin 65(2): 87-108.

Tripathy, D. (2004). "Systemic therapy for advanced breast cancer." Breast J 10 Suppl 1: S26-27.

Trougakos, I. P. and E. S. Gonos (2002). "Clusterin/apolipoprotein J in human aging and cancer." Int J Biochem Cell Biol 34(11): 1430-1448.

Tu, C., P. A. Rudnick, M. Y. Martinez, K. L. Cheek, S. E. Stein, R. J. Slebos, et al. (2010). "Depletion of abundant plasma proteins and limitations of plasma proteomics." J Proteome Res 9(10): 4982-4991.

Tu, S. H., C. C. Chang, C. S. Chen, K. W. Tam, Y. J. Wang, C. H. Lee, et al. (2010). "Increased expression of enolase alpha in human breast cancer confers tamoxifen resistance in human breast cancer cells." Breast Cancer Res Treat 121(3): 539-553.

Tuck, A. B., A. F. Chambers and A. L. Allan (2007). "Osteopontin overexpression in breast cancer: knowledge gained and possible implications for clinical management." J Cell Biochem 102(4): 859-868.

Tuck, M. K., D. W. Chan, D. Chia, A. K. Godwin, W. E. Grizzle, K. E. Krueger, et al. (2009). "Standard operating procedures for serum and plasma collection: early detection research network consensus statement standard operating procedure integration working group." J Proteome Res 8(1): 113-117.

Turner, N. C. and J. S. Reis-Filho (2006). "Basal-like breast cancer and the BRCA1 phenotype." Oncogene 25(43): 5846-5853.

354

Turtoi, A., D. Musmeci, A. G. Naccarato, C. Scatena, V. Ortenzi, R. Kiss, et al. (2012). "Sparc-like protein 1 is a new marker of human glioma progression." J Proteome Res 11(10): 5011-5021.

Tworoger, S. S., A. H. Eliassen, P. Sluss and S. E. Hankinson (2007). "A prospective study of plasma prolactin concentrations and risk of premenopausal and postmenopausal breast cancer." J Clin Oncol 25(12): 1482-1488.

Tyan, Y. C., H. R. Guo, C. Y. Liu and P. C. Liao (2006). "Proteomic profiling of human urinary proteome using nano-high performance liquid chromatography/electrospray ionization tandem mass spectrometry." Anal Chim Acta 579(2): 158-176.

Tyers, M. and M. Mann (2003). "From genomics to proteomics." Nature 422(6928): 193- 197.

UniProt, C. (2013). "Update on activities at the Universal Protein Resource (UniProt) in 2013." Nucleic Acids Res 41(Database issue): D43-47.

Unwin, R. D., J. R. Griffiths, M. K. Leverentz, A. Grallert, I. M. Hagan and A. D. Whetton (2005). "Multiple reaction monitoring to identify sites of protein phosphorylation with high sensitivity." Molecular & Cellular Proteomics 4(8): 1134-1144.

Vaezzadeh, A. R., A. C. Briscoe, H. Steen and R. S. Lee (2010). "One-step sample concentration, purification, and albumin depletion method for urinary proteomics." Journal of Proteome Research 9(11): 6082-6089.

Vafadar-Isfahani, B., G. Ball, C. Coveney, C. Lemetre, D. Boocock, L. Minthon, et al. (2012). "Identification of SPARC-like 1 protein as part of a biomarker panel for Alzheimer's disease in cerebrospinal fluid." J Alzheimers Dis 28(3): 625-636. van de Vijver, M. J., Y. D. He, L. J. van't Veer, H. Dai, A. A. Hart, D. W. Voskuil, et al. (2002). "A gene-expression signature as a predictor of survival in breast cancer." N Engl J Med 347(25): 1999-2009. van den Broek, I., R. W. Sparidans, J. H. Schellens and J. H. Beijnen (2010). "Quantitative assay for six potential breast cancer biomarker peptides in human serum by liquid chromatography coupled to tandem mass spectrometry." Journal of chromatography. B, Analytical technologies in the biomedical and life sciences 878(5-6): 590-602. van den Broek, I., R. W. Sparidans, J. H. Schellens and J. H. Beijnen (2010). "Sensitive liquid chromatography/tandem mass spectrometry assay for absolute quantification of

355

ITIH4-derived putative biomarker peptides in clinical serum samples." Rapid Commun Mass Spectrom 24(13): 1842-1850. van den Broek, I., R. W. Sparidans, J. H. Schellens and J. H. Beijnen (2010). "Sensitive liquid chromatography/tandem mass spectrometry assay for absolute quantification of ITIH4-derived putative biomarker peptides in clinical serum samples." Rapid communications in mass spectrometry : RCM 24(13): 1842-1850.

Venkitaraman, A. R. (2002). "Cancer susceptibility and the functions of BRCA1 and BRCA2." Cell 108(2): 171-182.

Verdier, J. M., B. Dussol, P. Dupuy, Y. Berland and J. C. Dagorn (1992). "Preliminary treatment of urinary proteins improves electrophoretic analysis and immunodetection." Clin Chem 38(6): 860-863.

Vergara, D., P. Simeone, P. del Boccio, C. Toto, D. Pieragostino, A. Tinelli, et al. (2013). "Comparative proteome profiling of breast tumor cell lines by gel electrophoresis and mass spectrometry reveals an epithelial mesenchymal transition associated protein signature." Mol Biosyst 9(6): 1127-1138.

Verjans, E., E. Noetzel, N. Bektas, A. K. Schutz, H. Lue, B. Lennartz, et al. (2009). "Dual role of macrophage migration inhibitory factor (MIF) in human breast cancer." BMC Cancer 9: 230.

Villanueva, J., D. R. Shaffer, J. Philip, C. A. Chaparro, H. Erdjument-Bromage, A. B. Olshen, et al. (2006). "Differential exoprotease activities confer tumor-specific serum peptidome patterns." J Clin Invest 116(1): 271-284.

Villar-Garea, A., M. Griese and A. Imhof (2007). "Biomarker discovery from body fluids using mass spectrometry." J Chromatogr B Analyt Technol Biomed Life Sci 849(1-2): 105-114.

Vizoso, F., L. M. Sanchez, I. Diez-Itza, A. M. Merino and C. Lopez-Otin (1995). "Pepsinogen C is a new prognostic marker in primary breast cancer." J Clin Oncol 13(1): 54-61.

Vlahou, A., C. Laronga, L. Wilson, B. Gregory, K. Fournier, D. McGaughey, et al. (2003). "A novel approach toward development of a rapid blood test for breast cancer." Clin Breast Cancer 4(3): 203-209.

356

Vuong, D., P. T. Simpson, B. Green, M. C. Cummings and S. R. Lakhani (2014). "Molecular classification of breast cancer." Virchows Arch 465(1): 1-14.

Wai, P. Y. and P. C. Kuo (2004). "The role of Osteopontin in tumor metastasis." J Surg Res 121(2): 228-241.

Wang, J., J. P. Costantino, E. Tan-Chiu, D. L. Wickerham, S. Paik and N. Wolmark (2004). "Lower-category benign breast disease and the risk of invasive breast cancer." J Natl Cancer Inst 96(8): 616-620.

Wang, J., X. Wang, S. Lin, C. Chen, C. Wang, Q. Ma, et al. (2013). "Identification of kininogen-1 as a serum biomarker for the early detection of advanced colorectal adenoma and colorectal cancer." PLoS One 8(7): e70519.

Wang, L., D. Su, H. J. Yan, J. H. Xu, Z. G. Zheng, Y. J. Hu, et al. "Primary study of lymph node metastasis-related serum biomarkers in breast cancer." Anatomical Record (Hoboken, N J : 2007) 294(11): 1818-1824.

Wang, L., J. Yu, J. Ni, X. M. Xu, J. Wang, H. Ning, et al. (2003). "Extracellular matrix protein 1 (ECM1) is over-expressed in malignant epithelial tumors." Cancer Lett 200(1): 57-67.

Wang, L. P., J. Bi, C. Yao, X. D. Xu, X. X. Li, S. M. Wang, et al. (2010). "Annexin A1 expression and its prognostic significance in human breast cancer." Neoplasma 57(3): 253-259.

Wang, Y., Q. Shan, G. Hou, J. Zhang, J. Bai, X. Lv, et al. (2016). "Discovery of potential colorectal cancer serum biomarkers through quantitative proteomics on the colonic tissue interstitial fluids from the AOM-DSS mouse model." J Proteomics 132: 31-40.

Wang, Z., Z. Wang, Z. Liang, J. Liu, W. Shi, P. Bai, et al. (2013). "Expression and clinical significance of IGF-1, IGFBP-3, and IGFBP-7 in serum and lung cancer tissues from patients with non-small cell lung cancer." Onco Targets Ther 6: 1437-1444.

Warmoes, M., J. E. Jaspers, T. V. Pham, S. R. Piersma, G. Oudgenoeg, M. P. Massink, et al. (2012). "Proteomics of mouse BRCA1-deficient mammary tumors identifies DNA repair proteins with potential diagnostic and prognostic value in human breast cancer." Mol Cell Proteomics 11(7): M111.013334.

357

Wasinger, V., L. Ly, A. Fitzgerald, B. Walsh, V. Wasinger, L. Ly, et al. (2008). "Prefractionation, enrichment, desalting and depleting of low volume and low abundance proteins and peptides using the MF10." Methods in Molecular Biology 424: 257-275.

Watanabe, K., T. Oochiai, S. Kikuchi, T. Kumano, T. Matsui, K. Morimoto, et al. (2012). "Dermokine expression in intraductal papillary-mucinous neoplasm and invasive pancreatic carcinoma." Anticancer Res 32(10): 4405-4412.

Weeks, M. E. (2010). "Urinary proteome profiling using 2D-DIGE and LC-MS/MS." Methods in Molecular Biology 658: 293-309.

Weeks, M. E. (2010). "Urinary proteome profiling using 2D-DIGE and LC-MS/MS." Methods Mol Biol 658: 293-309.

Weitzel, L. R., T. Byers, J. Allen, C. Finlayson, S. M. Helmke, J. E. Hokanson, et al. (2010). "Discovery and verification of protein differences between Er positive/Her2/neu negative breast tumor tissue and matched adjacent normal breast tissue." Breast Cancer Res Treat 124(2): 297-305.

Weivoda, S., J. D. Andersen, A. Skogen, P. M. Schlievert, D. Fontana, T. Schacker, et al. (2008). "ELISA for human serum leucine-rich alpha-2-glycoprotein-1 employing cytochrome c as the capturing ligand." J Immunol Methods 336(1): 22-29.

Welton, J. L., S. Khanna, P. J. Giles, P. Brennan, I. A. Brewis, J. Staffurth, et al. (2010). "Proteomics analysis of bladder cancer exosomes (cell lines)." Molecular & Cellular Proteomics 9(6): 1324-1338.

Wendt, M. K., M. A. Taylor, B. J. Schiemann and W. P. Schiemann (2011). "Down- regulation of epithelial cadherin is required to initiate metastatic outgrowth of breast cancer." Mol Biol Cell 22(14): 2423-2435.

Wenners, A. S., K. Mehta, S. Loibl, H. Park, B. Mueller, N. Arnold, et al. (2012). "Neutrophil gelatinase-associated lipocalin (NGAL) predicts response to neoadjuvant chemotherapy and clinical outcome in primary human breast cancer." PLoS One 7(10): e45826.

Whiteaker, J. R., H. Zhang, J. K. Eng, R. Fang, B. D. Piening, L. C. Feng, et al. (2007). "Head-to-head comparison of serum fractionation techniques." J Proteome Res 6(2): 828- 836.

358

Whiteaker, J. R., H. Zhang, L. Zhao, P. Wang, K. S. Kelly-Spratt, R. G. Ivey, et al. (2007). "Integrated pipeline for mass spectrometry-based discovery and confirmation of biomarkers demonstrated in a mouse model of breast cancer." Journal of Proteome Research 6(10): 3962-3975.

Whiteside, T. L. and S. Ferrone (2012). "For breast cancer prognosis, immunoglobulin kappa chain surfaces to the top." Clin Cancer Res 18(9): 2417-2419.

Winer, E., J. Gralow, L. Diller, B. Karlan, P. Loehrer, L. Pierce, et al. (2009). "Clinical cancer advances 2008: major research advances in cancer treatment, prevention, and screening--a report from the American Society of Clinical Oncology." J Clin Oncol 27(5): 812-826.

Wittke, S., D. Fliser, M. Haubitz, S. Bartel, R. Krebs, F. Hausadel, et al. (2003). "Determination of peptides and proteins in human urine with capillary electrophoresis- mass spectrometry, a suitable tool for the establishment of new diagnostic markers." J Chromatogr A 1013(1-2): 173-181.

Wolff, S., A. Otto, D. Albrecht, J. S. Zeng, K. Büttner, M. Glückmann, et al. (2006). "Gel- free and Gel-based Proteomics in Bacillus subtilis." Molecular & Cellular Proteomics 5(7): 1183-1192.

Woo, H. M., K. M. Kim, M. H. Choi, B. H. Jung, J. Lee, G. Kong, et al. (2009). "Mass spectrometry based metabolomic approaches in urinary biomarker study of women's cancers." Clinica Chimica Acta 400(1-2): 63-69.

Woodward, W. A., E. A. Strom, S. L. Tucker, M. D. McNeese, G. H. Perkins, N. R. Schechter, et al. (2003). "Changes in the 2003 American Joint Committee on Cancer staging for breast cancer dramatically affect stage-specific survival." J Clin Oncol 21(17): 3244-3248.

Woong-Shick, A., P. Sung-Pil, B. Su-Mi, L. Joon-Mo, N. Sung-Eun, N. Gye-Hyun, et al. (2005). "Identification of hemoglobin-alpha and -beta subunits as potential serum biomarkers for the diagnosis and prognosis of ovarian cancer." Cancer Sci 96(3): 197- 201.

Worsham, M. J., U. Raju, M. Lu, A. Kapke, J. Cheng and S. R. Wolman (2007). "Multiplicity of benign breast lesions is a risk factor for progression to breast cancer." Clin Cancer Res 13(18 Pt 1): 5474-5479.

Wright, T. and A. McGechan (2003). "Breast cancer: new technologies for risk assessment and diagnosis." Mol Diagn 7(1): 49-55. 359

Wu, J., Y. D. Chen and W. Gu (2010). "Urinary proteomics as a novel tool for biomarker discovery in kidney diseases." J Zhejiang Univ Sci B 11(4): 227-237.

Wu, M. H., Y. C. Chou, W. Y. Chou, G. C. Hsu, C. H. Chu, C. P. Yu, et al. (2009). "Circulating levels of leptin, adiposity and breast cancer risk." Br J Cancer 100(4): 578- 582.

Wu, Q. W., H. Q. She, J. Liang, Y. F. Huang, Q. M. Yang, Q. L. Yang, et al. (2012). "Expression and clinical significance of extracellular matrix protein 1 and vascular endothelial growth factor-C in lymphatic metastasis of human breast cancer." BMC Cancer 12: 47.

Wulfkuhle, J. D., C. P. Paweletz, P. S. Steeg, E. F. Petricoin, 3rd, L. Liotta, J. D. Wulfkuhle, et al. (2003). "Proteomic approaches to the diagnosis, treatment, and monitoring of cancer." Advances in Experimental Medicine & Biology 532: 59-68.

Wulfkuhle, J. D., D. C. Sgroi, H. Krutzsch, K. McLean, K. McGarvey, M. Knowlton, et al. (2002). "Proteomics of human breast ductal carcinoma in situ." Cancer Res 62(22): 6740-6749.

Xanthoulis, A. and D. G. Tiniakos (2013). "E2F transcription factors and digestive system malignancies: how much do we know?" World J Gastroenterol 19(21): 3189-3198.

Xiang, M., W. Zhou, D. Gao, X. Fang and Q. Liu (2012). "Inhibitor of apoptosis protein- like protein-2 as a novel serological biomarker for breast cancer." Int J Mol Sci 13(12): 16737-16750.

Xiao, M., S. Jia, H. Wang, J. Wang, Y. Huang and Z. Li (2013). "Overexpression of LAPTM4B: an independent prognostic marker in breast cancer." J Cancer Res Clin Oncol 139(4): 661-667.

Xie, D., S. H. Lau, J. S. Sham, Q. L. Wu, Y. Fang, L. Z. Liang, et al. (2005). "Up-regulated expression of cytoplasmic clusterin in human ovarian carcinoma." Cancer 103(2): 277- 283.

Xie, M. J., Y. Motoo, S. B. Su, H. Mouri, K. Ohtsubo, F. Matsubara, et al. (2002). "Expression of clusterin in human pancreatic cancer." Pancreas 25(3): 234-238.

Xiong, G. P., J. X. Zhang, S. P. Gu, Y. B. Wu and J. F. Liu (2012). "Overexpression of ECM1 contributes to migration and invasion in cholangiocarcinoma cell." Neoplasma 59(4): 409-415.

360

Xu, C., H. Chen, X. Wang, J. Gao, Y. Che, Y. Li, et al. (2014). "S100A14, a member of the EF-hand calcium-binding proteins, is overexpressed in breast cancer and acts as a modulator of HER2 signaling." J Biol Chem 289(2): 827-837.

Xu, X., T. D. Veenstra, S. D. Fox, J. M. Roman, H. J. Issaq, R. Falk, et al. (2005). "Measuring fifteen endogenous estrogens simultaneously in human urine by high- performance liquid chromatography-mass spectrometry." Anal Chem 77(20): 6646-6654.

Xu, X., B. Wang, C. Ye, C. Yao, Y. Lin, X. Huang, et al. (2008). "Overexpression of macrophage migration inhibitory factor induces angiogenesis in human breast cancer." Cancer Lett 261(2): 147-157.

Xu, Z., S. C. Bolick, L. A. DeRoo, C. R. Weinberg, D. P. Sandler and J. A. Taylor (2013). "Epigenome-wide association study of breast cancer using prospectively collected sister study samples." J Natl Cancer Inst 105(10): 694-700.

Yamanaka, M., K. Kanda, N. C. Li, T. Fukumori, N. Oka, H. O. Kanayama, et al. (2001). "Analysis of the gene expression of SPARC and its prognostic value for bladder cancer." J Urol 166(6): 2495-2499.

Yamashita, H., T. Toyama, M. Nishio, Y. Ando, M. Hamaguchi, Z. Zhang, et al. (2006). "p53 protein accumulation predicts resistance to endocrine therapy and decreased post- relapse survival in metastatic breast cancer." Breast Cancer Res 8(4): R48.

Yang, H., H. Zhou, P. Feng, X. Zhou, H. Wen, X. Xie, et al. (2010). "Reduced expression of Toll-like receptor 4 inhibits human breast cancer cells proliferation and inflammatory cytokines secretion." J Exp Clin Cancer Res 29: 92.

Yang, M.-H., P.-Y. Chu, S. C.-J. Chen, T.-W. Chung, W.-C. Chen, L.-B. Tan, et al. (2011). "Characterization of ADAM28 as a biomarker of bladder transitional cell carcinomas by urinary proteome analysis." Biochemical & Biophysical Research Communications 411(4): 714-720.

Yang, M. S., H. S. Wang, B. S. Wang, W. H. Li, Z. F. Pang, B. K. Zou, et al. (2013). "A comparative proteomic study identified calreticulin and prohibitin up-regulated in adrenocortical carcinomas." Diagn Pathol 8: 58.

Yang, N., S. Feng, K. Shedden, X. Xie, Y. Liu, C. J. Rosser, et al. (2011). "Urinary glycoprotein biomarker discovery for bladder cancer detection using LC/MS-MS and label-free quantification." Clinical Cancer Research 17(10): 3349-3359.

361

Yates, J. R., 3rd, A. Gilchrist, K. E. Howell and J. J. Bergeron (2005). "Proteomics of organelles and large cellular structures." Nat Rev Mol Cell Biol 6(9): 702-714.

Yates, J. R., C. I. Ruse and A. Nakorchevsky (2009). "Proteomics by mass spectrometry: approaches, advances, and applications." Annual Review of Biomedical Engineering 11: 49-79.

Yeghiazaryan, K., S. Mamlouk, D. Trog, H. Moenkemann, M. Braun, W. Kuhn, et al. (2007). "Irradiated breast cancer patients demonstrate subgroup-specific regularities in protein expression patterns of circulating leukocytes." Cancer Genomics Proteomics 4(6): 411-418.

Yi, J., C. Kim and C. A. Gelfand (2007). "Inhibition of Intrinsic Proteolytic Activities Moderates Preanalytical Variability and Instability of Human Plasma." J. Proteome Res. 6(5): 1768-1781.

Yi, J., Z. Liu, D. Craft, P. Oâ€™Mullan, G. Ju and C. A. Gelfand (2008). "Intrinsic Peptidase Activity Causes a Sequential Multi-Step Reaction (SMSR) in Digestion of Human Plasma Peptides." J. Proteome Res 7(12): 5112-5118.

Yi, J. K., J. W. Chang, W. Han, J. W. Lee, E. Ko, D. H. Kim, et al. (2009). "Autoantibody to tumor antigen, alpha 2-HS glycoprotein: a novel biomarker of breast cancer screening and diagnosis." Cancer Epidemiology, Biomarkers & Prevention 18(5): 1357-1364.

Yi, W., J. Peng, Y. Zhang, F. Fu, Q. Zou and Y. Tang (2013). "[Differential protein expressions in breast cancer between drug sensitive tissues and drug resistant tissues]." Zhong Nan Da Xue Xue Bao Yi Xue Ban 38(2): 148-154.

Yin, F., X. Liu, D. Li, Q. Wang, W. Zhang and L. Li (2013). "Bioinformatic analysis of chemokine (C-C motif) ligand 21 and SPARC-like protein 1 revealing their associations with drug resistance in ovarian cancer." Int J Oncol 42(4): 1305-1316.

Yip, C. H. and N. A. Taib (2014). "Breast health in developing countries." Climacteric: 1-6.

362

Yocum, A. K. and A. M. Chinnaiyan (2009). "Current affairs in quantitative targeted proteomics: multiple reaction monitoring-mass spectrometry." Briefings in functional genomics & proteomics 8(2): 145-157.

Yom, C. K., W. Han, S. W. Kim, H. S. Kim, H. C. Shin, J. N. Chang, et al. (2011). "Clinical significance of annexin A1 expression in breast cancer." J Breast Cancer 14(4): 262-268.

Yoo, B. C., S. Y. Kong, S. G. Jang, K. H. Kim, S. A. Ahn, W. S. Park, et al. (2010). "Identification of hypoxanthine as a urine marker for non-Hodgkin lymphoma by low- mass-ion profiling." BMC Cancer 10: 55.

Yousef, G. M., A. Chang and E. P. Diamandis (2000). "Identification and characterization of KLK-L4, a new kallikrein-like gene that appears to be down-regulated in breast cancer tissues." J Biol Chem 275(16): 11891-11898.

Yousef, G. M., A. Magklara and E. P. Diamandis (2000). "KLK12 is a novel serine protease and a new member of the human kallikrein gene family-differential expression in breast cancer." Genomics 69(3): 331-341.

Ytting, H., I. J. Christensen, S. Thiel, J. C. Jensenius and H. J. Nielsen (2005). "Serum mannan-binding lectin-associated serine protease 2 levels in colorectal cancer: relation to recurrence and mortality." Clin Cancer Res 11(4): 1441-1446.

Yu, L., C. Jiang, S. Huang, X. Gong, S. Wang and P. Shen (2013). "Analysis of urinary metabolites for breast cancer patients receiving chemotherapy by CE-MS coupled with on-line concentration." Clin Biochem 46(12): 1065-1073.

Zacharatos, P., A. Kotsinas, K. Evangelou, P. Karakaidos, L. V. Vassiliou, N. Rezaei, et al. (2004). "Distinct expression patterns of the transcription factor E2F-1 in relation to tumour growth parameters in common human carcinomas." J Pathol 203(3): 744-753.

Zakynthinos, E. and N. Pappa (2009). "Inflammatory biomarkers in coronary artery disease." J Cardiol 53(3): 317-333.

Zamani-Ahmadmahmudi, M., S. M. Nassiri and R. Rahbarghazi (2013). "Serological proteome analysis of dogs with breast cancer unveils common serum biomarkers with human counterparts." Electrophoresis.

Zeng, X. H., Z. L. Ou, K. D. Yu, L. Y. Feng, W. J. Yin, J. Li, et al. (2014). "Absence of multiple atypical chemokine binders (ACBs) and the presence of VEGF and MMP-9

363

predict axillary lymph node metastasis in early breast carcinomas." Med Oncol 31(9): 145.

Zeng, Z., M. Hincapie, B. B. Haab, S. Hanash, S. J. Pitteri, S. Kluck, et al. (2010). "The development of an integrated platform to identify breast cancer glycoproteome changes in human serum." J Chromatogr A 1217(19): 3307-3315.

Zeng, Z., M. Hincapie, S. J. Pitteri, S. Hanash, J. Schalkwijk, J. M. Hogan, et al. (2011). "A proteomics platform combining depletion, multi-lectin affinity chromatography (M- LAC), and isoelectric focusing to study the breast cancer proteome." Anal Chem 83(12): 4845-4854.

Zerefos, P. G. and A. Vlahou (2008). "Urine sample preparation and protein profiling by two-dimensional electrophoresis and matrix-assisted laser desorption ionization time of flight mass spectroscopy." Methods in Molecular Biology 428: 141-157.

Zhang, A., H. Sun and X. Wang (2012). "Saliva Metabolomics Opens Door to Biomarker Discovery, Disease Diagnosis, and Treatment." Applied biochemistry and biotechnology.

Zhang, D., L. K. Tai, L. L. Wong, L. L. Chiu, S. K. Sethi and E. S. Koay (2005). "Proteomic study reveals that proteins involved in metabolic and detoxification pathways are highly expressed in HER-2/neu-positive breast cancer." Mol Cell Proteomics 4(11): 1686-1696.

Zhang, D., L. K. Tai, L. L. Wong, T. C. Putti, S. K. Sethi, M. Teh, et al. (2008). "Proteomic characterization of differentially expressed proteins in breast cancer: Expression of hnRNP H1, RKIP and GRP78 is strongly associated with HER-2/neu status." Proteomics Clin Appl 2(1): 99-107.

Zhang, H., J. K. Kim, C. A. Edwards, Z. Xu, R. Taichman and C. Y. Wang (2005). "Clusterin inhibits apoptosis by interacting with activated Bax." Nat Cell Biol 7(9): 909- 915.

Zhang, J., L. Zhu, J. Fang, Z. Ge and X. Li (2016). "LRG1 modulates epithelial- mesenchymal transition and angiogenesis in colorectal cancer via HIF-1alpha activation." J Exp Clin Cancer Res 35(1): 29.

Zhang, L., H. Xiao, S. Karlan, H. Zhou, J. Gross, D. Elashoff, et al. (2010). "Discovery and preclinical validation of salivary transcriptomic and proteomic biomarkers for the non-invasive detection of breast cancer." PLoS One 5(12): e15573.

364

Zhang, S. J., Y. Hu, H. L. Qian, S. C. Jiao, Z. F. Liu, H. T. Tao, et al. (2013). "Expression and significance of ER, PR, VEGF, CA15-3, CA125 and CEA in judging the prognosis of breast cancer." Asian Pac J Cancer Prev 14(6): 3937-3940.

Zhang, X., J. Gureasko, K. Shen, P. A. Cole and J. Kuriyan (2006). "An allosteric mechanism for activation of the kinase domain of epidermal growth factor receptor." Cell 125(6): 1137-1149.

Zhang, Z., R. C. Bast, Jr., Y. Yu, J. Li, L. J. Sokoll, A. J. Rai, et al. (2004). "Three biomarkers identified from serum proteomic analysis for the detection of early stage ovarian cancer." Cancer Res 64(16): 5882-5890.

Zhao, Y., C. Chang, P. Qin, Q. Cao, F. Tian, J. Jiang, et al. (2016). "Mining the human plasma proteome with three-dimensional strategies by high-resolution Quadrupole Orbitrap Mass Spectrometry." Anal Chim Acta 904: 65-75.

Zhong, X. Y., A. Ladewig, S. Schmid, E. Wight, S. Hahn and W. Holzgreve (2007). "Elevated level of cell-free plasma DNA is associated with breast cancer." Arch Gynecol Obstet 276(4): 327-331.

Zhou, C., A. M. Nitschke, W. Xiong, Q. Zhang, Y. Tang, M. Bloch, et al. (2008). "Proteomic analysis of tumor necrosis factor-alpha resistant human breast cancer cells reveals a MEK5/Erk5-mediated epithelial-mesenchymal transition phenotype." Breast Cancer Research 10(6): R105.

Zhou, H., P. S. Yuen, T. Pisitkun, P. A. Gonzales, H. Yasuda, J. W. Dear, et al. (2006). "Collection, storage, preservation, and normalization of human urinary exosomes for biomarker discovery." Kidney International 69(8): 1471-1476.

Zhou, L., L. Q. Huang, R. W. Beuerman, M. E. Grigg, S. F. Y. Li, F. T. Chew, et al. (2004). "Proteomic Analysis of Human Tears: Defensin Expression after Ocular Surface Surgery." Journal of Proteome Research 3(3): 410-416.

Zhou, M., D. A. Lucas, K. C. Chan, H. J. Issaq, E. F. Petricoin, 3rd, L. A. Liotta, et al. (2004). "An investigation into the human serum "interactome"." Electrophoresis 25(9): 1289-1298.

Zhu, D. J., X. W. Chen, J. Z. Wang, Y. L. Ju, M. Z. Ou Yang and W. J. Zhang (2013). "Proteomic analysis identifies proteins associated with curcumin-enhancing efficacy of irinotecan-induced apoptosis of colorectal cancer LOVO cell." Int J Clin Exp Pathol 7(1): 1-15.

365

Zhu, K., J. Kim, C. Yoo, F. R. Miller and D. M. Lubman (2003). "High sequence coverage of proteins isolated from liquid separations of breast cancer cells using capillary electrophoresis-time-of-flight MS and MALDI-TOF MS mapping." Anal Chem 75(22): 6209-6217.

Zhu, L., X. Song, J. Tang, J. Wu, R. Ma, H. Cao, et al. (2013). "Huntingtin-associated protein 1: a potential biomarker of breast cancer." Oncol Rep 29(5): 1881-1887.

Zoidakis, J., M. Makridakis, P. G. Zerefos, V. Bitsika, S. Esteban, M. Frantzi, et al. (2012). "Profilin 1 is a potential biomarker for bladder cancer aggressiveness." Molecular & cellular proteomics : MCP 11(4): M111 009449.

Zoidakis, J., M. Makridakis, P. G. Zerefos, V. Bitsika, S. Esteban, M. Frantzi, et al. (2012). "Profilin 1 is a potential biomarker for bladder cancer aggressiveness." Mol Cell Proteomics 11(4): M111 009449.

Zolotarjova, N., J. Martosella, G. Nicol, J. Bailey, B. E. Boyes and W. C. Barrett (2005). "Differences among techniques for high-abundant protein depletion." Proteomics 5(13): 3304-3313.

Zurbig, P., M. B. Renfrow, E. Schiffer, J. Novak, M. Walden, S. Wittke, et al. (2006). "Biomarker discovery by CE-MS enables sequence analysis via MS/MS with platform- independent separation." Electrophoresis 27(11): 2111-2125.

366