EXPRESSION ANALYSIS AND FUNCTIONAL SIGNIFICANCE OF CHONDROITIN SULPHATE PROTEOGLYCANS AND HEPARAN SULPHATE PROTEOGLYCANS IN PROSTATE CANCER
TENG HUI FANG, YVONNE (BSc, Hons) NATIONAL UNIVERSITY OF SINGAPORE
A THESIS SUBMITTED FOR THE DEGREE OF DOCTOR OF PHILOSOPHY DEPARTMENT OF ANATOMY YONG LOO LIN SCHOOL OF MEDICINE NATIONAL UNIVERSITY OF SINGAPORE
2010 Acknowledgements
ACKNOWLEDGEMENTS
My PhD candidature would not have been easier, if not for the help, guidance
and support from many mentors, family, friends and colleagues.
My utmost gratitude goes to my mentor, Associate Professor George Yip
Wai Cheong, for his unfailing guidance, patience and trust in me. Through him, I
have learnt much, in terms of scientific skills and knowledge acquisition. In addition,
I have benefitted from his many interesting stories that have also made my candidature a very enjoyable one indeed. The most important lesson that I have learnt from Dr George Yip would be the importance of thinking critically, a skill that is not only important in the scientific arena, but also one that applies to our everyday life.
I would also like to thank my co-supervisor, Associate Professor Chia Sing
Joo, for his support and advice. His comments have always added an interesting clinical perspective to the type of scientific research that I have been working on during my candidature.
My deepest appreciation goes to Professor Bay Boon Huat for his timely advice and support. Professor Bay never fails to think for our welfare and has always shown genuine concern for us in all aspects of our life. He has taught us the importance of forming great relationships, one that supersedes any material wealth and status.
Associate Professor Tan Puay Hoon has been a source of inspiration to me.
Being one of the first female professors that I have personally met and known, she has demonstrated that with passion, hard work and determination, women can be leaders in their respective fields as well. I am grateful to her as she always takes time to help
Page | i
Acknowledgements
and support me in my work despite a busy schedule; her boundless energy is amazing
and can be infectious.
I would also like to thank Dr Aye Aye Thike for spending much time to score
the immunostained slides with me and for freely sharing her knowledge and expertise.
To Dr Chong Kian Tai, thank you for your help in liasing with the collection of the
prostate glands for the microarray study. Thank you for always trying so hard to
ensure that the prostatic chips that we collect are ‘as fresh as possible’.
This project would not have been possible without the excellent technical expertise offered by Ms Sii Lang Hiong (my immunohistochemical staining teacher),
Ms Cheok Poh Yian and Ms Tan Yen. Thank you all for helping us cut all the prostate tissue sections and for compiling the clinicopathological data.
Also to Mrs Yong Eng Siang, thank you for making the Cell and
Developmental Biology Laboratory such a safe and clean place to work in. I also deeply appreciate the nice treats you have given me and the conversations that we had which makes my candidature at the Anatomy Department so much more enjoyable and memorable. Thank you, Mrs Ng Geok Lan, for your help and troubleshooting for problems that I had encountered in the Histology Laboratory. Also to Mr Poon
Junwei, thank you for your help and expertise in the Tissue Culture Laboratory. My apologies to all of you if I had been messy or broken the rules at times!
Many thanks to all my friends and colleagues whom I have worked with, either previously or presently: to Mr Gilbert Khor, my basic cell culture teacher and the first person to help orientate me around the laboratory when I first joined; all the
Laboratory officers/Research Assistants, Ms Lee Xiao Hwee, Ms Choo Siew Hua,
Ms Grace Leong, Ms Sim Wey Cheng, Ms Wendy Sim and Mr Nick Low, for
Page | ii
Acknowledgements
helping keep things in order and in constant supply so that I can work without
worries; my seniors, Dr Guo Chunhua and Dr Koo Chuay Yeng, for their help and
advice; my fellow friends, Dr Omid Iravani, Ms Cao Shoufeng, Ms Chua Peijou,
Mr Lo Soo Ling, Ms Sen Yin Ping, Mr Matthew Tan, Mr Koo Kah Chun, Ms
Therese Kwan, Mr Alvin Low, Mr Jeffrey Lim, Mr Ong Han Kee, Mr Ho Soo
Keng and Mr Ethan Khoo– it has really been fun and enjoyable learning together and working with all of you!
I would also like to express my gratitude to all staff and students of the
Department of Anatomy, Yong Loo Lin School of Medicine, National University of Singapore, for all the help, advices and stories you have willingly shared with me.
I would also like to thank the National University of Singapore for giving me the Graduate Research Scholarship, without which, all my work would not have been made possible.
Most importantly, I would like to thank my family for their unfailing support through the years. I dedicate this work to them for their love and patience all this while.
Page | iii
Table of contents
TABLE OF CONTENTS
Acknowledgements i Table of contents iv Summary xi List of Tables xiii List of Figures xvii List of Abbreviations xx List of Publications xxiii
Chapter 1 Introduction 1 1.1 Prostate gland and prostate cancer 1 1.1.1 The anatomy of the prostate gland 1 1.1.2 Functions of the prostate gland 3 1.1.3 The histology and histopathology of the prostate gland 4 1.1.3.1 Normal prostate histology 4 1.1.3.2 Histopathology of benign prostatic hyperplasia 4 1.1.3.3 Histopathology of prostatic intraepithelial neoplasia 5 1.1.3.4 Histopathology of prostate adenocarcinoma 5 1.1.4 Clinical symptoms and diagnosis of prostate cancer 6 1.1.5 Gleason grading of prostate cancer 7 1.1.6 Prognostic markers for prostate cancer 8 1.1.7 Treatment 10 1.1.8 Current challenges in prostate cancer 12 1.1.9 Genome-wide expression profiling of prostate cancer 12 1.2 Glycosaminoglycans (GAG) and proteoglycans (PG) – an 14 introduction 1.2.1 Structure 14 1.2.2 Biosynthesis of chondroitin sulphate/dermatan sulphate 19 and heparan sulphate 1.2.3 Enzymatic remodelling of chondroitin/dermatan sulphate 25 and heparan sulphate chains
Page | iv
Table of contents
1.2.4 Functions of chondroitin/dermatan sulphate and heparan 26 sulphate and/or proteoglycans in cancer 1.2.4.1 Heparan sulphate in cancer biology 27 1.2.4.2 Chondroitin sulphate in cancer biology 31 1.3 Objectives of project 33
Chapter 2 Materials and Methods 35 2.1 Genome-wide expression profiling of prostate 35 adenocarcinoma tissues 2.1.1 Collection of prostate adenocarcinoma tissues 35 2.1.2 Haematoxylin and eosin staining of prostate tissues 35 2.1.3 RNA isolation, yield and quality from prostate tissues 35 2.1.4 Microarray analysis 36 2.1.5 Analysis of microarray data 37 2.1.6 cDNA synthesis and quantitative real-time polymerase 37 chain reaction (qPCR) 2.2 Expression analysis of chondroitin sulphate and heparan 44 sulphate in prostate adenocarcinoma tissues using immunohistochemistry 2.2.1 Clinical samples and patient data collection 44 2.2.2 Immunohistochemical staining 44 2.2.3 Immunohistochemical evaluation 46 2.2.4 Statistical analysis 46 2.3 In vitro cell culture 47 2.3.1 Cell lines 47 2.3.2 Storage of cells 47 2.3.3 Antibodies used 48 2.3.4 RNA extraction, cDNA synthesis, real-time PCR of 48 prostate tumour cell lines 2.3.5 Silencing of prostate cell lines 54 2.3.6 Western blot 55 2.3.7 GAG extraction and quantification 56 2.3.8 Proliferation assay 56
Page | v
Table of contents
2.3.9 Migration assay 57 2.3.10 Invasion assay 57 2.3.11 Adhesion assay 58 2.3.12 Cell cycle assay 59 2.3.13 Cell apoptosis assay 59 2.3.14 Flow cytometry 59 2.3.15 Microarray analysis 60
Chapter 3 Expression Profiling of Prostate Adenocarcinoma Tissues 62 3.1 Selection of tissues to be used for study 62 3.2 RNA quality of tissue samples used for Affymetrix GeneChip 62 microarrays 3.3 Quality control of microarray data 69 3.4 Differential expression of genes in prostate adenocarcinoma 71 3.5 Hierarchical clustering of microarray data 83 3.6 Principal component analysis (PCA) of microarray expression 83 data 3.7 Quantitative real-time polymerase chain reaction (qPCR) 86 verification of microarray results 3.8 Discussion 90 3.8.1 Aberrant expression of genes in prostate adenocarcinoma 90 3.8.2 Prostate cancer as a result of development gone wrong? 92 3.8.2.1 Dysregulation of Wnt signalling pathway 95 3.8.2.2 Dysregulation of homeobox genes 97 3.8.3 Prostate cancer – sustaining growth and motility while 98 evading death signals 3.8.4 Prostate cancer as a result of aberration of metabolic 98 processes? 3.8.5 Aberrant expression of GAGs and PGs genes in prostate 100 cancer 3.8.6 Conclusions 106
Page | vi
Table of contents
Chapter 4 Expression and Functional Analysis of Chondroitin 107 Sulphate in Prostate Cancer 4.1 Expression analysis of chondroitin sulphate in prostate cancer 107 4.1.1 Clinical and demographic data 107 4.1.2 Expression analysis of non-sulphated chondroitin in 111 prostate adenocarcinoma 4.1.2.1 Localization of non-sulphated chondroitin in 111 prostatic tissues 4.1.2.2 Comparison of non-sulphated chondroitin expression 111 in benign prostatic and adenocarcinoma tissues 4.1.2.3 Comparison of non-sulphated chondroitin expression 117 in different grades of prostate adenocarcinoma tissues 4.1.3 Expression analysis of chondroitin-4-sulphate in prostate 120 adenocarcinoma 4.1.3.1 Localization of chondroitin-4-sulphate in prostate 120 tissues 4.1.3.2 Comparison of chondroitin-4-sulphate expression in 123 benign prostatic and adenocarcinoma tissues 4.1.3.3 Comparison of chondroitin-4-sulphate expression in 126 different grades of prostate adenocarcinoma tissues 4.1.4 Expression analysis of chondroitin-6-sulphate in prostate 129 adenocarcinoma. 4.1.4.1 Localization of chondroitin-6-sulphate in prostate 129 tissues 4.1.4.2 Comparison of chondroitin-6-sulphate expression in 129 benign prostatic and adenocarcinoma tissues 4.1.4.3 Comparison of chondroitin-6-sulphate expression in 134 different grades of prostate adenocarcinoma tissues 4.1.5 Expression analysis of chondroitin-2,6-sulphate in 137 prostate adenocarcinoma. 4.1.5.1 Localization of chondroitin-2,6-sulphate in prostate 137 tissues
Page | vii
Table of contents
4.1.5.2 Comparison of chondroitin-2,6-sulphate expression 140 in benign prostatic and adenocarcinoma tissues 4.1.5.3 Comparison of chondroitin-2,6-sulphate expression 144 in different grades of prostate adenocarcinoma tissues 4.1.6 Comparisons of chondroitin sulphate staining amongst 147 each subtype 4.1.7 Association of chondroitin sulphate staining with 150 clinicopathological parameters 4.2 Gene transcript analysis of chondroitin-sulphate biosynthesis 157 and editing genes in prostate cancer cell lines 4.3 Functional analysis of CSGalNAcT-1 in prostate cancer cell, 161 PC-3 4.3.1 CSGalNAcT-1 is effectively silenced in PC-3 – gene 161 transcript and protein levels 4.3.2 Reduction of CS/DS chains after CSGalNAcT-1 siRNA 161 4.3.3 CSGalNAcT-1 silencing inhibited PC-3 proliferation 163 4.3.4 Apoptotic analysis of CSGalNAcT-1 silenced PC-3 cells 163 4.3.5 Cell cycle analysis of CSGalNAcT-1 silenced PC-3 cells 163 4.3.6 CSGalNAcT-1 silencing impeded PC-3 invasion 164 4.3.7 CSGalNAcT-1 silencing has no effect on PC-3 adhesion 164 to collagen type I 4.3.8 Microarray analysis of CSGalNAcT-1 silenced PC-3 cells 166 4.3.8.1 Quality control of microarray data 167 4.3.8.2 Differential expression of genes in CSGalNAcT-1 167 silenced PC-3 4.3.9 Cell cycle and apoptotic proteins were downregulated in 169 CSGalNAcT-1 silenced PC-3 cells 4.4 Discussion 177 4.4.1 Increased chondroitin sulphate expression and their potential 177 use as a biomarker of prostate cancer progression 4.4.2 Unraveling the functional significance of CSGalNAcT-1 179 upregulation in prostate cancer
Page | viii
Table of contents
Chapter 5 Expression and Functional Analysis of Heparan Sulphate 189 in Prostate Cancer 5.1 Expression analysis of heparan sulphate in prostate cancer 189 5.1.1 Localization of heparan sulphate in BPH and 189 adenocarcinoma tissues 5.1.2 Comparison of heparan sulphate expression in benign 189 prostatic and adenocarcinoma tissues 5.1.3 Comparison of heparan sulphate expression in different 190 grades of prostate adenocarcinoma tissues 5.1.4 Association of heparan sulphate staining with 197 clinicopathological parameters 5.2 Gene transcript analysis of heparan sulphate biosynthesis 206 genes in prostate cancer cell lines 5.3 Functional Analysis of HS3ST3A1 in prostate cancer 210 5.3.1 HS3ST3A1 is effectively silenced in RWPE-1 210 5.3.2 HS3ST3A1 silencing inhibited RWPE-1 proliferation 210 5.3.3 HS3ST3A1 silencing inhibited RWPE-1 migration 211 5.3.4 HS3ST3A1 silencing did not affect RWPE-1 adhesion to 211 collagen type I 5.4 Functional Analysis of HS6ST2 in prostate cancer 213 5.4.1 HS6ST2 is effectively silenced in RWPE-1 213 5.4.2 HS6ST2 silencing prevented the proliferation of RWPE-1 213 cells. 5.4.2 HS6ST2 silencing reduced the migratory ability of 214 RWPE-1 cells. 5.4.3 HS6ST2 silencing did not affect adhesion of RWPE-1 214 cells to collagen type I 5.5 Discussion 216 5.5.1 Utility of 10E4-specific heparan sulphate as a prostate 216 cancer biomarker 5.5.2 HS3ST3A1 and HS6ST2 are not tumour suppressors in 217 prostate cancer.
Page | ix
Table of contents
Chapter 6 Conclusions and Future Work 223 6.1 Identification of potential biomarkers in prostate cancer 223 6.2 Establishing the functional significance of CS/HS and their 225 biosynthesis enzymes in prostate cancer
Chapter 7 References 228 Appendix I Filtered gene list 264 Appendix II Association studies of chondroitin sulphate staining with 289 clinicopathological parameters Appendix III Reprints of publications 322
Page | x
Summary
SUMMARY
Prostate cancer is the most frequently diagnosed cancer among American men,
and is second only to lung cancer in malignancy-related deaths. In Singapore, it is
now the third commonest malignancy among all men. Current diagnostic tools are not
powerful enough to accurately predict the clinical outcome, and more biomarkers are
in need to warrant better prediction and treatment plans for each patient. In addition,
understanding the complexities of this disease at a molecular level may aid in the
identification and subsequent development of potential therapeutics.
Glycosaminoglycans and proteoglycans have been found to be important
regulators of cellular growth and metastasis. In this study, the expression of
chondroitin sulphate (CS) and heparan sulphate (HS) in prostate cancer tissues was
determined and their possible uses as potential biomarkers of prostate cancer
progression were explored. To search for plausible novel therapeutic targets, the
functional implications of CS and HS synthesis enzymes that were differently expressed in prostate cancer were identified and determined.
Genome-wide expression profiling of six prostate adenocarcinoma tissues compared to six benign hyperplasia tissues was performed. It was found that
CSGalNAcT-1, HS3ST3A1 and HS6ST2 were differentially expressed in prostate cancer. Immunohistochemistry was also performed and it was found that CS was upregulated in prostate cancer and high CS expression augurs poor prognosis.
Since CSGalNAcT-1, a gene involved in chondroitin sulphate/dermatan
sulphate synthesis, is upregulated in both prostate cancer tissues and prostate cancer
cell lines, as compared to their benign counterparts, this gene was silenced in PC-3
prostate cancer cells. The upregulation of this gene could possibly explain the
increased expression of chondroitin sulphate in prostate cancer tissues, as detected by
Page | xi
Summary
immunohistochemistry. Silencing CSGalNAcT-1 inhibited cell proliferation and
invasion, and promoted apoptosis via suppression of MAPRE-1, TPX2 and HIPK3.
These results point to a potential role of CSGalNAcT-1 and chondroitin sulphate as a novel therapeutic target.
Immunohistochemistry was also performed using monoclonal 10E4 that recognizes mixed HS domains containing N-acetylated and N-sulphated disaccharide units. 10E4-specific heparan sulphate epitopes were not found to be differentially expressed in the prostate adenocarcinoma tissues as compared to the benign hyperplastic prostate tissues. However, strong 10E4 staining in the prostate cancer cells is associated with higher clinical stage.
HS3ST3A1 and HS6ST2, both heparan sulphotransferases, were downregulated in prostate cancer and silencing either of these genes in normal prostate epithelial cell,
RWPE-1, led to growth inhibition and reduced motility. These indicated the importance of HS3ST3A1 and HS6ST2 in normal prostate cell physiology, but perhaps suggested these genes to demonstrate different roles after the acquisition of tumour cell characteristics.
On the whole, my findings reiterated the importance of CS and HS in cellular growth and behaviour, and essentially, suggested CS to be a promising biomarker of prostate cancer progression and established the role of CSGalNAcT-1 as a pro-tumour protein.
Page | xii
List of tables
LIST OF TABLES
Chapter 1 Table 1.1 Histology of prostate adenocarcinoma tissues 6 Table 1.2 Features of prostate cancer tissues with different Gleason 7 grades Table 1.3 Prostate cancer- TNM staging 9 Table 1.4 List of proteoglycans, classified according to location 18 Table 1.5 Genes coding for enzymes involved in the synthesis of 23 CSPG/DSPG and HSPG
Chapter 2 Table 2.1 Methods of normalization and criteria for determining genes 38 differentially expressed using different microarray analysis software, for microarray work involving human prostate tissues Table 2.2 Programme settings for qPCR 39 Table 2.3 Primer sequences used in real-time PCR (qPCR) for 40 microarray data verification Table 2.4 Antibodies and optimal conditions used for 45 immunohistochemistry Table 2.5 Primer sequences used in real-time PCR, for in vitro work 49 involving prostate cancer cell lines Table 2.6 ON-TARGETplus SMARTpool CSGalNAcT-1 siRNA 54 sequences Table 2.7 Ambion Silencer siRNA sequences 55 Table 2.8 Methods of normalization and criteria for determining genes 61 differentially expressed using different microarray analysis software, for microarray work involving CSGalNAcT-1 silenced PC-3 cells
Page | xiii
List of tables
Chapter 3 Table 3.1 Clinical features of prostate adenocarcinoma cases used for 66 expression profiling Table 3.2 RNA purity and quality of samples used for microarray study 69 Table 3.3 Gene Ontology (GO) annotation terms and clusters that the 865 74 filtered genes are associated with Table 3.4 Highlight of some genes that are differentially expressed in 78 prostate adenocarcinoma tissues Table 3.5 Verification of several genes from microarray list, as 87 determined by qPCR
Chapter 4 Table 4.1 Clinical features of prostate adenocarcinoma cases used for 108 immunohistochemistry Table 4.2 Expression analysis of non-sulphated chondroitin in benign 114 hyperplastic prostate tissues and prostate adenocarcinoma tissues Table 4.3 Mean values of non-sulphated chondroitin staining in prostate 116 adenocarcinoma samples versus paired BPH samples. Table 4.4 Mean values of non-sulphated chondroitin staining in different 119 grades of prostate adenocarcinoma tissues Table 4.5 Expression analysis of chondroitin-4-sulphate in benign 122 hyperplastic prostate tissues and prostate adenocarcinoma tissues Table 4.6 Mean values of chondroitin-4-sulphate staining in prostate 125 adenocarcinoma samples versus paired BPH samples Table 4.7 Mean values of chondroitin-4-sulphate staining in different 128 grades of prostate adenocarcinoma tissues Table 4.8 Expression analysis of chondroitin-6-sulphate in benign 131 hyperplastic prostate tissues and prostate adenocarcinoma tissues Table 4.9 Mean values of chondroitin-6-sulphate staining in prostate 133 adenocarcinoma samples versus paired BPH samples
Page | xiv
List of tables
Table 4.10 Mean values of chondroitin-6-sulphate staining in different 136 grades of prostate adenocarcinoma tissues Table 4.11 Expression analysis of chondroitin-2,6-sulphate in benign 139 hyperplastic prostate tissues and prostate adenocarcinoma tissues Table 4.12 Mean values of chondroitin-2,6-sulphate staining in prostate 143 adenocarcinoma samples versus paired BPH samples Table 4.13 Mean values of chondroitin-2,6-sulphate staining in different 146 grades of prostate adenocarcinoma tissues Table 4.14 Comparisons of chondroitin sulphate staining amongst each 148 subtype in the adenocarcinoma cells Table 4.15 Comparisons of chondroitin sulphate staining amongst each 149 subtype in the peritumoural stroma Table 4.16 Association studies of different chondroitin sulphate staining 152 intensities with clinicopathological parameters Table 4.17 Association studies of percentage of cells positively stained for 155 different chondroitin sulphate with clinicopathological parameters. Table 4.18 Expression fold changes of CS/DS biosynthesis and editing 159 genes and CSPG/DSPG genes in prostate cancer cell lines and tissues Table 4.19 RNA purity and quality of samples used for microarray study 166 Table 4.20 List of genes differentially expressed after CSGalNAcT-1 168 silencing and their fold changes as determined from microarray study and real-time PCR Table 4.21 Transcript data of genes possibly involved in CSGalNAcT-1- 186 mediated signaling
Chapter 5 Table 5.1 Expression profiles of 10E4 specific-heparan sulphate epitope 191 in prostate tissues Table 5.2 Mean values of 10E4-specific heparan sulphate epitope 194 staining in prostate adenocarcinoma samples and paired BPH
Page | xv
List of tables
tissues. Table 5.3 Mean values of 10E4-specific heparan sulphate staining in 196 different grades of prostate adenocarcinoma tissues Table 5.4 Association studies of 10E4-epitope specific heparan sulphate 198 staining intensity with clinicopathological parameters Table 5.5 Association of percentage of positively 10E4-stained cells with 202 clinicopathological data. Table 5.6 Relative fold changes of heparan sulphate biosynthesis genes 207 differentially expressed in prostate cancer cell lines and prostate tissues
Page | xvi
List of figures
LIST OF FIGURES
Chapter 1 Figure 1.1 Structure of proteoglycans and glycosaminoglycans 17 Figure 1.2 Structure of glycosaminoglycans 17
Chapter 3 Figure 3.1 Haematoxylin and eosin staining of BPH prostate samples 64 used for microarray study Figure 3.2 Haematoxylin and eosin staining of prostate 65 adenocarcinoma samples used for microarray study Figure 3.3 Electrophoregrams of RNA samples extracted from six 67 BPH sections and six cancer sections Figure 3.4 Electrophoregrams of fragmented cRNA samples 68 Figure 3.5 Hierarchical clustering of differentially expressed genes in 84 prostate cancer tissues Figure 3.6 Principal component analysis (PCA) 85
Chapter 4 Figure 4.1 Non-sulphated chondroitin expression in prostate tissues 113 Figure 4.2 Comparisons of non-sulphated chondroitin expression in 115 prostate adenocarcinoma samples against matched BPH samples. Figure 4.3 Comparisons of non-sulphated chondroitin expression in 118 different grades of prostate adenocarcinoma tissues Figure 4.4 Chondroitin-4-sulphate expression in BPH tissue and 121 prostate adenocarcinoma sections Figure 4.5 Comparisons of chondroitin-4-sulphate expression in 124 prostate adenocarcinoma samples against matched BPH samples. Figure 4.6 Comparisons of chondroitin-4-sulphate expression in 127 different grades of prostate adenocarcinoma tissues Figure 4.7 Chondroitin-6-sulphate expression in BPH tissue and 130
Page | xvii
List of figures
prostate adenocarcinoma sections Figure 4.8 Comparisons of chondroitin-6-sulphate expression in 132 prostate adenocarcinoma samples against paired BPH samples Figure 4.9 Comparisons of chondroitin-6-sulphate expression in 135 different grades of prostate adenocarcinoma tissues, Figure 4.10 Chondroitin-2,6-sulphate expression in BPH and prostate 138 adenocarcinoma tissues Figure 4.11 Comparisons of chondroitin-2,6-sulphate expression in 142 prostate adenocarcinoma samples against paired BPH samples. Figure 4.12 Comparisons of chondroitin-2,6-sulphate expression in 145 different grades of prostate adenocarcinoma tissues, Figure 4.13 CSGalNAcT-1 is effectively silenced and CS/DS 162 expression is reduced Figure 4.14 Effects of CSGAlNAcT-1 silencing on PC-3 cell behaviour 165 Figure 4.15 CSGalNAcT-1 silencing downregulated MAPRE-1 protein 171 – Flow cytometry Figure 4.16 CSGalNAcT-1 silencing downregulated MAPRE-1 protein 172 - Graph Figure 4.17 CSGalNAcT-1 silencing downregulated TPX2 protein – 173 Flow cytometry Figure 4.18 CSGalNAcT-1 silencing downregulated TPX2 protein - 174 Graph Figure 4.19 CSGalNAcT-1 silencing downregulated HIPK3 protein - 175 Flow cytometry Figure 4.20 CSGalNAcT-1 silencing downregulated HIPK3 protein - 176 Graph Figure 4.21 Hypothetical diagram on the mechanisms of CSGalNAcT-1 188 role in PC-3 cell behaviour
Chapter 5 Figure 5.1 10E4-specific heparan sulphate epitope staining in prostate 192
Page | xviii
List of figures
tissues Figure 5.2 Comparisons of 10E4-specific heparan sulphate epitope 193 expression in prostate adenocarcinoma samples against matched BPH samples Figure 5.3 Comparisons of 10E4-specific heparan sulphate epitope 195 expression in different grades of prostate adenocarcinoma tissues Figure 5.4 Effects of HS3ST3A1 silencing on RWPE-1 cell 212 behaviour Figure 5.5 Effects of HS6ST2 silencing on RWPE-1 cell behaviour 215
Page | xix
List of abbreviations
LIST OF ABBREVIATIONS
μg micrograms μl microlitres μm micrometres ANOVA analysis of variance AJCC American Joint Committee on Cancer ATCC American Type Cell Culture ATP adenosine triphosphate BM basement membrane BPH benign prostatic hyperplasia BSA bovine serum albumin cDNA complementary DNA cRNA complementary RNA CS chondroitin sulphate CS-E chondroitin sulphate E (also known as chondroitin-4,6- sulphate) CSPG chondroitin sulphate proteoglycan CSPG4 chondroitin sulphate 4 proteoglycan DMSO dimethylsulphoxide DNA deoxyribonucleic acid DS dermatan sulphate DSPG dermatan sulphate proteoglycan ECM extracellular matrix EGF epidermal growth factor EGFR epidermal growth factor receptor EMT epithelial-mesenchymal transition ERK extracellular signal regulated kinase FAK focal adhesion kinase FBS fetal bovine serum FGF fibroblast growth factor FGFR fibroblast growth factor receptor GAG glycosaminoglycan
Page | xx
List of abbreviations
GalNAc N-acetyl-D-galactosamine GAPDH glyceraldehyde 3-phosphate dehydrogenase GCOS GeneChip operating software Glc-6-P glucose 6 phosphate GlcA β-D-glucuronic acid GlcNAc N-acetyl-D-glucosamine GlcNS/ GlcNSO3 N-sulphate-D-glucosamine GO gene ontology GPI glycosylphosphatidylinositol HCL hydrochloric acid HGF hepatocyte growth factor HGPIN high grade PIN HS heparan sulphate HSPG heparan sulphate proteoglycan IdoA α-L-iduronic acid IRS immunoreactivity score JNK Jun N-terminal Kinase KEGG Kyoto Encyclopedia of Genome and Genes LGPIN low grade PIN MAPK mitogen-activated protein min minutes ml millilitres mm millimetres MMP-9 matrix metalloproteinase 9 mRNA messenger ribonucleic acid MTS 3-(4,5-dimethylthiazol-2-yl)-5-(3- carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H- tetrazolium NADH nicotinamide adenine dinucleotide NADPH nicotinamide adenine dinucleotide phosphate-oxidase ng nanograms nM nanomolar PAP prostate-specific acid phosphatase
Page | xxi
List of abbreviations
PAPS 3'-phosphoadenosine 5'-phosphosulphate PBS phosphate buffered saline PCR polymerase chain reaction PDGF platelet derived growth factor PG proteoglycan PI3K phosphoinositide 3-kinase PIN prostatic intraepithelial neoplasia PM/MM perfect match/mismatch PPP pentose phosphate pathway PSA prostate specific antigen PVDF polyvinylidene difluoride qPCR quantitative real-time PCR RIN RNA integrity number RMA Robust Multi-array Average rpm revolutions per minute RPP radical perineal prostatectomy RRP radical retropubic prostatectomy SDS-PAGE sodium-docedyl-sulphate polyacrylamide gel electrophoresis siRNA silencing RNA TGF transforming growth factor TRUS transrectal ultrasound guided core biopsies TURP transurethral resection of the prostate VEGF vascular endothelial growth factor w/v weight per volume WAI weighted average intensity
Page | xxii
List of publications
LIST OF PUBLICATIONS
Articles
1. Teng Hui-Fang Yvonne, Tan Puay-Hoon, Chia Sing-Joo, Nor Azhari Bin Mohd Zam, Lau Kam-On Weber, Cheng Wai-Sam Christopher, Bay Boon Huat and Yip Wai-Cheong George (2008) Increased expression of non-sulphated chondroitin correlates with adverse clinicopathological parameters in prostate cancer. Modern Pathology 21:893-901
2. Teng Hui-Fang Yvonne, Tan Puay-Hoon, Chia Sing-Joo, Aye Aye Thike, Bay Boon-Huat and Yip Wai-Cheong George. Downregulation of CSGalNAcT-1 activity in PC-3 prostate adenocarcinoma cells shows important tumour-attenuating properties: inhibition of cell proliferation, increased apoptosis and decreased invasiveness. Manuscript in preparation, 2009.
Meeting Proceedings
1. Teng Hui-Fang Yvonne and Yip Wai-Cheong George. (2006) Modulation of prostate cancer cell behaviour by sulphated proteoglycans in Proceedings of the Universiti Kebangsaan Malaysia-Microscopy Society (Singapore) Joint Meeting, 6th National Symposium on Health Sciences, Kuala Lumpur, Malaysia.
2. Teng Hui-Fang Yvonne and Yip Wai-Cheong George. (2006) Sulphated proteoglycans modulate the proliferation, migration and invasion of prostate cancer cells in Proceedings of The 16th International Microscopy Congress, Sapporo, Japan.
3. Teng Hui-Fang Yvonne, Yip Wai-Cheong George and Tan Puay Hoon. (2006) Involvement of sulphated proteoglycans in the modulation of prostate cancer cells behaviour in Proceedings of the Singapore General Hospital 15th Annual Scientific Meeting, Singapore.
4. Teng Hui-Fang Yvonne, Tan Puay-Hoon and Yip Wai-Cheong George. (2007) Expression of chondroitin in prostate tumours and their association with several clinicopathological parameters of poor prognosis in Proceedings of the Microscopy Society (Singapore) Annual General Meeting, Singapore.
Page | xxiii
List of publications
5. Teng Hui-Fang Yvonne, Chia Sing-Joo, Chong Kian-Tai, Tan Puay-Hoon, Bay Boon Huat and Yip Wai-Cheong George. (2008) Analysis of chondroitin sulphate proteoglycans in prostate cancer in Proceedings of Urology Fair 2008 in conjunction with 1st Asia-Pacific Kidney Cancer Symposium, Singapore.
6. Teng Hui-Fang Yvonne, Tan Puay-Hoon, Chia Sing-Joo, Nor Azhari Bin Mohd Zam, Lau Kam-On Weber, Cheng Wai-Sam Christopher, Bay Boon-Huat, George Yip Wai-Cheong. (2008) Expression Analysis of Differentially Sulphated Chondroitin Species in Prostate Cancer in Proceedings of Annual Conference of the Society for Glycobiology, Fort Worth, Texas, USA.
Awards
1. Travel Scholarship, Microscopy Society (Singapore) for the 16th International Microscopy Congress (IMC16), Japan, Sapporo, 2006.
2. Travel Scholarship, Microscopy Society (Singapore) for the Universiti Kebangsaan Malaysia-Microscopy Society (Singapore) Joint Meeting, 6th National Symposium on Health Sciences, Kuala Lumpur, Malaysia, 2006.
3. Travel Scholarship, Glycobiology Society (USA) for the Annual Conference of the Society for Glycobiology, Fort Worth, Texas, USA, 2008.
Page | xxiv
Chapter 1 Introduction
Chapter 1
Introduction
1.1 Prostate Gland and Prostate Cancer
1.1.1 The Anatomy of the Prostate Gland
The prostate gland is composed of a large series of independently branching ducts,
all of which open into the prostatic urethra that the prostate gland surrounds. The
somewhat conical-shaped prostate has a base, which lies against the bladder neck situated above the gland and also has an apex, which lies against the urogenital diaphragm below. Located superiorly to the prostate base are the seminal vesicles.
The seminal vesicles join the vas deferens on each side to form the ejaculatory ducts.
The two ejaculatory ducts penetrate the upper part of the posterior surface of the prostate to join the prostatic urethra at the verumontanum [Snell RS, 2008].
Traditionally, the topography of prostatic disorders is related to prostate lobes
[Lowsley O, 1912; Hammerich et al, 2009]. The prostate is incompletely divided into five lobes: anterior lobe, median lobe, posterior lobe, right lateral lobe and left lateral lobe. The non-glandular anterior lobe lies in front of the urethra while the glandular- rich median lobe is situated between the urethra and the ejaculatory ducts. The posterior lobe sits behind the urethra and below the ejaculatory ducts; glands are also present in the posterior lobe. On the sides of the urethra lie the right and left lateral lobes of the prostate, in which there are many glands present. This lobular system was used even though there was no distinct lobes observable in human. Subsequently,
McNeal established the current and most widely accepted concept of the zonal anatomy of the prostate [McNeal et al, 1980 & 1988].
The zonal anatomy of the prostate includes a peripheral zone, a central zone, a transition zone and the anterior fibromuscular stroma. The peripheral zone comprise
Page | 1
Chapter 1 Introduction
all of the prostatic glandular tissue at the apex and the tissue located posteriorly near
the capsule; the central zone is a cone-shaped area of the adult gland, extending from
the apex of the cone at the confluence of the ejaculatory ducts and the prostatic
urethra at the verumontanum. The transition zone consists of 2 equal portions of
glandular tissue lateral to the urethra in the mid gland. The anterior fibromuscular
stroma extends downward from the bladder neck over the anteromedial surface of the
prostate, narrowing to join the urethra at the prostate apex. The apical half of the anterior fibromuscular stroma consists of striated muscles that blend into the glands
and the muscles of the pelvic diaphragm while towards the base, there are more
smooth muscle cells which blend into the fibres of the bladder neck [McNeal JE,
1972].
These zones of the prostate are usually associated with certain pathology of the
prostate; prostate carcinoma and inflammation usually occurs in the peripheral region
while the transition zone is involved in benign prostatic hyperplasia (BPH) and less commonly, prostate adenocarcinoma. The central zone is resistant to both carcinoma and inflammation [Hammerich et al, 2009].
The prostatic capsule mainly consists of smooth muscle fibres and collagenous
tissues, where the relative and absolute amounts of fibrous and muscular tissue vary
from area to area; there is no clear separation between the periacinar smooth muscle
and the smooth muscle at the inner capsular border. In addition, the distance of the
terminal acini to the prostate surface is highly variable at different areas within the
same prostate. Even more so, there is inconsistency in the proportion and arrangement
of the collagenous tissue at different areas of the same prostate. As a result, the prostate capsule cannot be regarded as a well defined anatomical structure. In evaluating extraprostatic extension by prostate carcinoma, only complete penetration
Page | 2
Chapter 1 Introduction
with perforation through the capsule surface may be related to prognosis in prostate carcinoma [McNeal JE, 1990]. There is a defect in the capsule anteriorly and anterolaterally at the apex; the distal fibres of the anterior fibromuscular stroma and the striated sphincter together often mingle with the prostatic glandular tissue anterolateral to the urethra; the relative extent of these tissues vary considerably inter- individually. As a result, it is hard to determine the extent of invasion if carcinoma of the prostate apex invades anteriorly [Ayala et al, 1989].
The prostate receives its blood supply from the inferior vesical and middle rectal arteries. Veins of the prostate form the prostatic venous plexus which also receives the deep dorsal vein of the penis and numerous vesical veins. The plexus drains into the internal iliac nodes [Brooks JD, 2007].
Innervation of the prostate gland is accomplished by two neurovascular bundles located posterolaterally adjacent to the gland. The prostate receive sympathetic and parasympathetic innervations from the peripheral hypogastric ganglion, and the hypogastric and pelvic nerves respectively. Prostate stroma consists of both cholinergic and noradrenergic fibres while the smooth muscle of the capsule and the space around the blood vessels consist of cholinergic nerves which helps the secretory function of the epithelial part. During ejaculation, excitation of the sympathetic nerves closes the bladder neck and assists in the discharge of the seminal fluid into the urethra [Snell RS, 2008].
1.1.2 Functions of the prostate gland
The secretory glandular cells synthesize and secrete prostate specific antigen
(PSA), citric acid, spermine, prostate-specific acid phosphatase (PAP) into the glandular lumen, contributing 20% to 25% of the ejaculate volume [Veltri and
Rodriguez, 2007]. The seminal vesicles and Cowper’s gland are the other accessory
Page | 3
Chapter 1 Introduction
glands responsible for the secretions of the semen too. Also, this prostatic secretion is
alkaline and helps neutralize the acidity in the vagina.
1.1.3 The Histology and Histopathology of the Prostate Gland
1.1.3.1 Normal Prostate Histology
The architecture of the normal prostate consists of branched ducts embedded in a
mixture of smooth muscle and connective tissue. The glandular epithelium, resting on
a thin basement membrane, consists of two cell layers: a luminal secretory columnar
cell layer and an underlying flattened cuboidal basal cell layer. The stroma of the
prostate gland consists of fibroblasts, smooth muscles, blood vessels and nerve cells.
Corpora amylacea, composed largely of protein and nucleic acid that are formed
from desquamated epithelial cells, can be found in the lumen of the prostatic glands.
They are oftenly seen as concentric lamellae which also consists of inspissated prostatic secretions that accumulate because of partial ductal obstruction. These may
further block the ducts with dilatation and further inspissations of secretions, resulting
in inflammation and subsequent deposition of calcium salts to form
microcalcifications [Moore and Hanzel, 1936].
1.1.3.2 Histopathology of Benign Prostatic Hyperplasia (BPH)
BPH, also known as nodular hyperplasia, is the benign enlargement of the
prostate usually occurring in older men. The prostatic stromal and epithelial cells
undergo uncontrolled proliferation, leading to large nodules formation in the periurethral region of the prostate. Consequently, this may lead to compression of the urethra resulting in urinary obstruction, dysuria and increased risk of urinary retention
and infection. Absence of cytologic atypia is observed, with the basal cell layer still
intact. In fact, the ultrastructural features of the acinar epithelium of hyperplastic
Page | 4
Chapter 1 Introduction nodules and normal prostate are not significantly different [Mao et al, 1965; Fisher and Jeffrey, 1965]. Nodular expansion is more readily apparent macroscopically, where the gross appearance is that of well–defined clustered nodules with variable cystic and solid compositions [Petersen et al, 2009].
1.1.3.3 Histopathology of Prostatic Intraepithelial Neoplasia (PIN)
PIN refers to the pre-invasive stage of the uncontrolled proliferation of the prostatic ducts and acini. PIN can be classified into low grade PIN (LGPIN) and high grade PIN (HGPIN), based on the cytologic characteristics of the cell [Petersen et al,
2009]. Histologic features of LGPIN include intact basal cell layer, enlarged nuclei of variable sizes, normal or slightly increased chromatin content and presence of small or inconspicuous nucleoli. In HGPIN, there is enlargement of nuclei with uniform sizes, prominent nucleoli, increased chromatin content and disruptions in the basal cell layer. HGPIN has been linked as the pre-malignant precursor to prostate cancer in many morphologic, immunohistochemical, molecular and genetic studies. HGPIN has been found in higher frequencies in prostate with cancers than prostates without cancer. The value of HGPIN is that it serves as a strong predictive marker of adenocarcinoma, and that its presence in biopsy specimens necessitates further search for invasive carcinoma [Petersen et al, 2009].
1.1.3.4 Histopathology of Prostate Adenocarcinoma
Prostate adenocarcinoma is the malignant transformation of the epithelial cells lining the prostate glands. The microscopic features of prostate adenocarcinoma are seen both architecturally and cytologically, as shown in Table 1.1. The light microscopic features are usually sufficient for diagnosis of prostate cancer, but in certain cases, immunohistochemical markers such as prostate specific antigen (PSA),
Page | 5
Chapter 1 Introduction
high molecular weight keratin (36βE12) and p63, could prove the presence of prostate
adenocarcinoma more definitively in suspicious foci [Montironi et al, 2007]. PSA is
useful in distinguishing high grade prostate cancer from other prostate diseases like
granulomatous prostatitis and other cancers such as urothelial cancer, colon cancer
and lymphoma. It is also useful in identifying prostate tumour origin in metastatic
sites [Petersen et al, 2009]. In suspicious foci, positively stained basal cell layer with
36βE12 or p63 could confirm the non-malignancy of the foci (as adenocarcinoma cells have a lack of basal cell layer); the diagnosis of malignancy should be based also on architectural and cytological features.
Table 1.1 Histology of prostate adenocarcinoma tissues
Architectural features Cytologic features • Infiltrative pattern • Nuclear enlargement • Irregular and haphazard arrangement • Prominent nucleoli • Variation in spacing between acini • Mitotic figures • Variation in acini sizes • Nuclear hyperchromasia • Irregular contour • Absent basal cell layer
1.1.4 Clinical symptoms and diagnosis of prostate cancer
Most prostate cancer cases are asymptomatic, and are detected by digital rectal
examination. If symptoms do present, they include urinary problems such as dysuria,
urinary obstruction, erectile difficulties, hematuria and frequent pain in the lower
back, hips, or upper thighs. However, other disease conditions such as benign prostate
hyperplasia (BPH) can also be presented with such symptoms.
In addition, laboratory test like prostate specific antigen (PSA) test are available
for prostate cancer diagnosis and early prostate cancer detection. PSA is found
elevated beyond an arbitrary cut-off of 4ng/ml in majority of prostate cancer patients.
Page | 6
Chapter 1 Introduction
However, both benign prostatic hyperplasia (BPH) and prostate adenocarcinoma
conditions can result in elevated PSA levels, thus limiting its use as a good diagnostic
and prognostic marker. Nevertheless, PSA screening is still widely used as it
highlights a diseased condition that warrants further investigation, and enables early
detection of prostate cancer [Diamandis EP, 2000].
Prostate biopsies are often necessary for the diagnosis of prostate cancer. The
current standard method for detection is by transrectal ultrasound guided core biopsies
(TRUS) while transurethral resection of the prostate (TURP) can also be performed.
Microscopic features of the prostate biopsy tissue are then analyzed to confirm the
diagnosis of prostate cancer [Montironi et al, 2007].
1.1.5 Gleason grading of prostate cancer
The Gleason grading system of prostate cancer which is based on pattern of
glandular growth and differentiation is divided into 5 grades [Gleason DF, 1966]
(Table 1.2). The Gleason sum, obtained from the addition of the primary grade
(dominant pattern) and the secondary grade (second most prevalent pattern), remains
one of the most significant prognostic markers of prostate cancer [Bostwick DG,
1994].
Table 1.2 Features of prostate cancer tissues with different Gleason grades
Gleason grade Features Grade 1 single, separate, closely packed acini Grade 2 single, more loosely arranged, less uniformed acini Grade 3 single acini of variable size and separation, cribriform and papillary patterns Grade 4 irregular masses of acini and fused epithelium, presence of clear cells Grade 5 almost complete loss of glandular lumen; epithelium are in solid sheets, strands or single cells invading into stroma
Page | 7
Chapter 1 Introduction
1.1.6 Prognostic markers for prostate cancer
Gleason sum is one of the most powerful predictor of progression; patients with Gleason sum 5-6 are cured after radical prostatectomy while those with Gleason sum of 8-10 have highly aggressive tumours with advanced stage. In these patients, more aggressive treatment is required [Epstein et al, 2005; Mian et al, 2002].
The TNM staging system, as recommended by the American Joint Committee on Cancer (AJCC) and International Union Against Cancer (IUAC), assesses the cancer according to the tumour features (T), regional lymph node metastasis (N) and distant metastasis (M). Clinical classification (cTNM) is carried out during initial evaluation of patient or when pathologic classification is not possible. Pathological classification (pTNM) is based on gross and microscopic examination (Table 1.3).
Extent of local invasion includes extraprostatic extension and seminal vesicle involvement. Seminal vesicle involvement is defined as cancer invading into the muscular coat of the seminal vesicle. Extraprostatic extension is the presence of cancer cells in extraprostatic fat tissue.
Positive surgical margins are indicated when cancer cells are found at the margins of the resected prostate. Presence of cancer cells at surgical margins could indicate that not all cancer cells are surgically removed and there may be residual cancer cells left in the patient. Surgical margin positivity has been significantly associated with increased risk of progression as compared to those with negative margins [Sakr et al, 1996].
Page | 8
Chapter 1 Introduction
Table 1.3 Prostate cancer - TNM staging (AJCC/IUAC)
Primary tumour, pathologic (pT) Regional lymph nodes (N) pT2 - Organ confined NX - Regional lymph nodes cannot be assessed pT2a - Unilateral N0 - No regional lymph node metastasis pT2b - Bilateral N1 - Metastasis in regional lymph node or nodes pT3 - Extraprostatic extension Distant metastasis (M) pT3a - Extraprostatic extension MX - Distant metastasis cannot be assessed pT3b - Seminal vesicle invasion M0 - No distant metastasis pT4 - Invasion of bladder, rectum M1- Distant metastasis *No pT1 stage in pT staging
Primary tumour (T), Clinical TX - Primary tumour cannot be assessed T0 - No evidence of primary tumour T1 - Clinically inapparent tumour not palpable or visible by imaging T1a - Tumour incidental histologic finding in 5% or less of tissue resected T1b - Tumour incidental histologic finding in more than 5% of tissue resected T1c - Tumour identified by needle biopsy (e.g. because of elevated PSA) T2 - Palpable tumour confined within prostate T2a - Tumour involves one lobe T2b - Tumour involves both lobes T3 - Tumour extends through the prostatic capsule T3a - Extracapsular extension (unilateral or bilateral) T3b - Tumour invades seminal vesicle(s) T4 - Tumour is fixed or invades adjacent structures other than seminal vesicles, e.g bladder
Perineural invasion is present when prostate cancer cells are seen adjacent or
within a nerve. It has been found to correlate with extraprostatic extension and
recently, reported to be an important predictor of lymph node metastasis [Quinn et al,
2003; Vargas et al, 1999].
Lymphovascular invasion is the presence of cancer cells within endothelial-
lined spaces; vascular and lymphatic channels are difficult to differentiate and there is
also lack of reproducibility among different observers by light microscopic examination. It has been found to correlate with lymph node metastasis, biochemical
recurrence and distant metastasis [Shariat et al, 2004].
The natural history of prostate carcinoma is complex and not fully understood.
The disease can be mild in some patients but present in a more aggressive form in
Page | 9
Chapter 1 Introduction
others. Better prognostic markers that can stratify the different types of prostate
cancer will be helpful. One way is to identify molecular biomarkers that can
differentiate between quiescent cancers and aggressive ones. In fact, there is immense
research in finding new molecular markers for prostate cancer, such as several recent
studies reporting on the use of Pin1, Akt-1 and cyclin D1 as predictive factors of clinical outcome in prostate cancer [Ayala et al, 2003; Le Page et al, 2006; Drobnjak et al, 2000].
1.1.7 Treatment
Various treatment options are available for prostate cancer such as active surveillance, surgery, radiation therapy, hormone therapy and chemotherapy.
Treatment plan for each patient usually depends on various factors, such as age,
Gleason grade, cancer stage, symptoms and health status of individual.
For patients with early stage prostate cancer that appears to be slow growing,
active surveillance (watchful waiting) may appear to be a ‘treatment plan’ if the risks
and side effects of other treatments appear to outweigh the benefits; such scenarios
could include very old age or implications from other serious health problems.
Radical prostatectomy which includes open radical retropubic prostatectomy
(RRP) (entire prostate removed through a cut in abdomen), open radical perineal
prostatectomy (RPP) (entire prostate removed through a cut between scrotum and
anus), laparoscopic prostatectomy (entire prostate removed through small cuts made
in abdomen, using a laparoscope) and robotic laparoscopic surgery, remains the ‘gold
standard’ for treatment of localized prostate cancer [Catalona and Han, 2007]. For
patients with advanced cancer, transurethral resection of the prostate (TURP) can be
performed in which cancer tissues that block the flow of urine are removed.
Page | 10
Chapter 1 Introduction
Radiation therapy and/or hormone therapy can be performed for any stage of
prostate cancer. Radiation therapy consists of external beam radiation therapy and
internal beam radiation therapy (brachytherapy) [Catalona and Han, 2007]. Radiation
therapy makes use of high energy waves that can damage the DNA of cells and
focuses that energy onto targeted cancer cells. Briefly, external beam radiation
accelerates electrons to generate waves of high energy photon radiation while
brachytherapy involves planting radioactive substances in the prostate to eliminate the
cancer cells over a period of time.
Hormone therapy involves the use of drugs that will deprive the prostate cancer of androgens like testosterone, which the cancer cells are dependent on for
growth. Anti-androgen drugs include flutamide and bicalutamide. Gonadotrophin-
releasing hormone agonists such as leuprolide, goserelin, triptorelin can also inhibit
testosterone production in the testicles. Ketoconazole prevents testosterone production
in the adrenal gland. Sometimes, orchiectomy (surgery to remove testicles) is also
performed [Nelson JB, 2007]. These measures aim to limit testosterone supply to the
prostate so as to curb the growth of the prostate cancer cells. However, prostate cancer
cells could eventually progress to be androgen-independent and will continue to grow
and spread, resulting in hormone therapy failure. This is commonly seen during the advanced stages of prostate cancer and is known as hormone refractory prostate cancer.
Chemotherapy may be used for hormone refractory prostate cancer. Docetaxel, a drug that disrupts microtubules and induces apoptosis, is currently the standard of care for metastatic hormone-refractory prostate cancer [Hadaschik and Gleave, 2007].
Nevertheless, drug-resistant cancer cells eventually surface and results in patient mortality.
Page | 11
Chapter 1 Introduction
1.1.8 Current Challenges in Prostate Cancer
Carcinoma of the prostate is the most frequently diagnosed cancer among
American men, and is second only to lung cancer in malignancy-related deaths.
[Jemal et al, 2009] While comparatively less common in Asian populations [Hsing et
al, 2000] prostate cancer has seen a rise of 118% from 6.6 to 14.4 per 100000 person-
years among Chinese men in Singapore [Sim and Cheng, 2005], where it is now the
3rd commonest malignancy among all Singapore men [Seow et al, 2009]. PSA
detection, digital rectal examination and transrectal ultrasonography are current
procedures used in the initial diagnosis of prostate cancer, and they are not specific to
and cannot confirm prostate cancer diagnosis. Moreover, heterogeneity in disease
manifestations of prostate cancer, in which some prostate cancers remain latent while
others may progress to clinically significant metastatic forms poses a huge prognostic
and treatment challenge. It remains a clinical challenge to decide treatment plans for prostate cancer patients, so as to ensure that there is no over-treatment for those
whose cancer will not become life threatening, nor to under-treat aggressive cancers
with conservative treatment [Yao and Lu-Yao, 2002]. More biomarkers are in need to
warrant better prediction and treatment plans for each patient. In addition,
understanding the complexities of this disease at a molecular level may aid in the
identification and subsequent development of potential therapeutics for localized and
metastatic prostate cancer.
1.1.9 Genome-wide expression profiling of prostate cancer
Genome-wide expression profiling is a high throughput method that generates
comprehensive descriptions of the transcriptional state of cancer cells and allows the
identification of candidate molecules that are dysregulated in the malignant state. In
Page | 12
Chapter 1 Introduction addition, such distinct gene signatures may allow the classification of different tumours into different tumour subtypes, which could potentially have prognostic values. For example, the use of gene expression profiling has been used to demarcate various breast tumours into clinically relevant groups based on their molecular profile.
Perou and co-workers revealed that breast tumours could be clearly delineated into 4 distinct subgroups: luminal, normal breast-like, human epidermal growth factor receptor 2 (Her2) and basal-like [Perou et al, 2000]. Notably, this classification of breast cancer has key clinical implications, where it has been associated with significantly different clinical outcomes. Luminal tumours tend to have better prognosis than Her2/basal-like breast carcinomas. [Sørlie et al, 2001 and 2003;
Sotiriou et al, 2003] This has driven much research into the potential use of such classification in a clinical setting. For instance, immunohistochemical surrogates have been actively investigated to replace the high cost and expertise required for the microarray profiling.
In prostate cancer, many microarray studies have been previously performed in different populations, albeit using different platforms [Welsh et al, 2001;
Dhanasekaran et al, 2001; Luo et al, 2001; Lapointe et al, 2004; Chandran et al,
2005]. Several genes have been commonly observed to be differentially expressed across different microarray datasets, such as AMACR, HPN and FASN. The identification of some of these genes in microarray studies has led to their use as potential diagnostic markers. AMACR is an enzyme that plays important roles in the peroxisomal and mitochondrial β-oxidation pathways [Lloyd et al, 2008] and was upregulated in malignant prostate tissues. AMACR protein analysis, in combination with the basal cell marker p63, has become a useful immunohistochemical aid in the diagnosis of prostate cancer [Hameed et al, 2005].
Page | 13
Chapter 1 Introduction
In addition, Lapointe et al has identified three classes of prostate tumours based on
their distinct gene signature that was profiled through microarray studies. Subtype I
prostate tumours is the clinically least aggressive subclass while subtype II and III
represent the clinically aggressive tumour classes. Notably, subtype III tumours share
features of gene expression with lymph node metastases. MUC1, encoding mucin 1
transmembrane protein, is a gene highly expressed in subgroup II and III. MUC1 was
found to associate with higher recurrence risk. On the contrary, AZGP1 encoding
zinc-α-2-glycoprotein, characterized subtype I tumours and was associated with longer time to recurrence [Lapointe et al, 2004]. These data demonstrated the potential utility and benefits expression profiling can offer, where identification of distinct gene signatures and candidate genes may provide a basis for improved prognostication and treatment stratifications.
1.2 Glycosaminoglycans (GAG) and Proteoglycans (PG) – An introduction
1.2.1 Structure
Glycosaminoglycans are unbranched polysaccharides composed of repeating units of alternating amino sugars and uronic acid. Heparan sulphate, heparin, chondroitin sulphate, dermatan sulphate, keratan sulphate and hyaluronan are different types of glycosaminoglycans. Different glycosaminoglycans (except hyaluronan) are found attached to different core proteins, giving rise to structurally diverse molecular entities known as proteoglycans. Glycosaminoglycans are usually attached to a serine residue in a core protein through a tetrasaccharide linkage region made up of xylose-galactose-galactose-glucuronic acid [Yip et al, 2006] (Figure 1.1).
Chondroitin sulphate (CS) and dermatan sulphate (DS) are composed of repeating disaccharide units of N-acetyl-D-galactosamine (GalNAc) and uronic acid.
(Figure 1.2A,B) In chondroitin sulphate, all of the uronic acid moieties are β-D-
Page | 14
Chapter 1 Introduction
glucuronic acid (GlcA) while in dermatan sulphate, epimerization of the C5 position
of the uronic acid moiety results in a mixture of α-L-iduronic acid (IdoA) and GlcA
[Yip et al, 2006]. Sulphate modifications frequently occur at various positions of the
CS and DS sugar chains; O-sulphation can occur at the C4 and/or C6 position of
GalNAc and/or C2 position of GlcA and IduA [Kusche-Gullberg and Kjellén, 2003].
Heparan sulphate / heparin are composed of repeating disaccharide units of N-
acetyl-D-glucosamine (GlcNAc) and uronic acid (glucuronic acid or iduronic acid)
(Figure 1.2C). Heparan sulphate/heparin are modified at several positions; the first
being the N-deacetylation and N-sulphation of the GlcNAc, which is the prerequisite
for all other modifications such as C5-epimerization of glucuronic acid to iduronic
acid, sulphate modifications at C2 position of the uronic acids, 6-O-sulphation and 3-
O-sulphation of the C6 and C3 positions of the glucosamine residues respectively
[Carlsson et al, 2008].
The classification of proteoglycans can be on the basis of their location
(cellular surface, extracellular matrix or basement membrane), GAG-chain
composition and amino acid homology of their core protein. GAG representations on
protein cores are not necessarily homogenous; different GAG types can be found on one protein core and the number of GAGs attached to each protein core could differ.
For example, syndecan-1, a cell-surface proteoglycan, consists of heparan sulphate
and chondroitin sulphate chains. Table 1.4 shows a list of proteoglycans classified
accordingly to location, namely, cell surface proteoglycans and extracellular matrix
proteoglycans, as compiled in the Kyoto Encyclopedia of Genome and Genes
(KEGG) database [Hashimoto et al, 2006].
Page | 15
Chapter 1 Introduction
Figure 1.1 Structure of proteoglycans and glycosaminoglycans
Figure 1.2 Structure of glycosaminoglycans. [A] CS [B] DS [C] HS/heparin. The R1/2/3 denotes positions where sulphate modifications are possible.
Page | 16
Chapter 1 Introduction
Table 1.4 List of proteoglycans, classified according to location.
Gene Symbol Proteoglycan name Types of GAG on proteoglycan Cell surface proteoglycans Syndecan family (transmembrane HSPG) SDC1 syndecan 1 HS / CS SDC2 syndecan 2 (fibroglycan) HS SDC3 syndecan 3 (N-syndecan) HS SDC4 syndecan 4 (amphiglycan, ryudocan) HS / CS
Glypican family (GPI-anchored HSPG) GPC1 glypican 1 HS GPC2 glypican 2 (cerebroglycan) HS GPC3 glypican 3 (OCI-5) HS GPC4 glypican 4 (K-glypican) HS GPC5 glypican 5 HS / CS GPC6 glypican 6 HS
Others PTPRZ protein tyrosine phosphatase, receptor- CS / KS type, Z polypeptide TGFBR3 transforming growth factor, beta HS / CS receptor III CSPG4 chondroitin sulphate proteoglycan 4 CS CSPG5 chondroitin sulphate proteoglycan 5 CS (neuroglycan C) CD44 CD44 antigen CS / DS APP amyloid beta (A4) precursor protein CS (protease nexin-II, Alzheimer disease)
Extracellular matrix (ECM) proteoglycans Small leucine-rich proteoglycan (SLRP) family Class I BGN biglycan CS / DS DCN decorin CS / DS Class II FMOD fibromodulin KS LUM lumican KS KERA keratocan KS OMD osteomodulin KS Class III OGN osteoglycin (osteoinductive factor, KS mimecan) DSPG3 dermatan sulphate proteoglycan 3 CS / DS (epiphycan)
Aggrecan/Versican family AGC1, CSPG1 aggrecan 1 CS / KS VCAN chondroitin sulphate proteoglycan 2 CS / DS
Page | 17
Chapter 1 Introduction
(versican) NCAN chondroitin sulphate proteoglycan 3 CS (neurocan) BCAN brevican CS
Basement membrane proteoglycans HSPG2 heparan sulphate proteoglycan 2 HS / CS (perlecan) AGRN agrin HS COL18A collagen, type XVIII, alpha HS SMC3 (CSPG6) structural maintenance of chromosome CS 3 (chondroitin sulphate proteoglycan 6) LEPRE leucine proline-enriched proteoglycan CS (leprecan) COL15A collagen, type XV, alpha CS
Others SRGN serglycin Heparin / CS SPOCK1 sparc / osteonectin, cwcv and kazal- CS / HS like domeins SPOCK2 sparc / osteonectin, cwcv and kazal- CS / HS like domeins
Page | 18
Chapter 1 Introduction
1.2.2 Biosynthesis of chondroitin sulphate/dermatan sulphate and heparan sulphate With such a diverse structural repertoire of the glycosaminoglycans, many biosynthetic enzymes are co-ordinately involved in the generation of these chains. A summary of all genes involved in the synthesis of CS/DS and HS is shown in Table
1.5. Briefly, the common tetrasaccharide linkage is first synthesised by the addition of xylose to the serine residue via Xylotransferases 1 (Xyl1) and Xylotransferases 2
(Xyl2) [Götting et al, 2000]. Two subsequent galactoses are added sequentially to the xylose via the action of 2 different galactosyltransferases coded by Xylosylprotein beta 1,4-galactosyltransferase, polypeptide 7 (galactosyltransferase I) (B4GalT7)
[Almeida et al, 1999] and UDP-Gal:betaGal beta 1,3-galactosyltransferase polypeptide 6 (B3GalT6) [Bai et al, 2001] respectively. The glucuronic acid is added via glucuronyltransferase I coded by Beta-1,3-glucuronyltransferase 3 (B3GAT3)
[Kitagawa et al, 1998].
The addition of a galactosamine to the linkage region via Chondroitin sulphate
N-acetylgalactosaminyltransferase 1 (CSGalNAcT-1) will thus initiate the synthesis of the CS/DS instead of HS/heparin [Uyama et al, 2002; Gotoh et al, 2002; Sato et al,
2003]. The subsequent polymerisation of the CS/DS chain is aided by the galactosaminyltransferases and glucuronyltransferases, Chondroitin sulphate N- acetylgalactosaminyltransferase 2 (CSGalNAcT-2) [Sato et al, 2003; Uyama et al,
2003], Chondroitin synthase I (CHSY1) [Kitagawa et al, 2001], Chondroitin synthase
3 (CHSY3) [Yada et al, 2003], Chondroitin polymerising factor (CHPF) [Kitagawa et al, 2003] and Chondroitin sulphate glucuronyltransferase (CSGlcAT) [Gotoh et al,
2002; Izumikawa et al, 2008]. Epimerisation of glucuronic acid to iduronic acid is peformed by the Dermatan sulphate epimerase (DSE) [Maccarana et al, 2006].
Page | 19
Chapter 1 Introduction
Sulphate modifications to the CS/DS chain are added via the different
sulphotransferases, Carbohydrate (chondroitin 4) sulphotransferase 13 (CHST13
/C4ST1) [Yamauchi et al, 2000; Hiraoka et al, 2000], Carbohydrate (chondroitin 4)
sulphotransferase 11 (CHST11/C4ST2) [Hiraoka et al, 2000] , Carbohydrate
(chondroitin 4) sulphotransferase 12 (CHST12/C4ST3) [Kang et al, 2002],
Carbohydrate (N-acetylglucosamine 6-O) sulphotransferase 7 (CHST7/C6ST1)
[Fukuta et al, 1998], Carbohydrate (chondroitin 6) sulphotransferase 3
(CHST3/C6ST2) [Kitagawa et al, 2000], B cell RAG associated protein (GalNAcT4S-
6ST) [Ohtake et al, 2001], Carbohydrate (N-acetylgalactosamine 4-0)
sulphotransferase 14 (CHST14/D4ST1) [Evers et al, 2001] and Uronyl-2-
sulphotransferase (UST) [Kobayashi et al, 1999].
The transfer of GlcNAc by N-acetyl-glucosaminyltransferase I (GlcNAcT-I) to
the tetrasaccharide linkage region will initiate heparan sulphate/heparin formation.
This GlcNAcT-I reaction is catalysed by two EXT gene family members, EXTL2 and
EXTL3 [Fritz et al, 1994; Kitagawa et al, 1999; Kim et al, 2001]. Elongation of the
heparan sulphate/heparin chains is further performed by the N-acetyl-
glucosaminyltransferase II (GlcNAcT-II) coded by EXTL1 and EXTL3 and the
heparan sulphate polymerases (GlcA/GlcNAc transferases) coded by EXT1 and EXT2, which are bifunctional enzymes that have both GlcNAcT-II and glucuronyltransferase activities [Lind et al, 1998; McCormick et al, 2000; Senay et al, 2000]. Following
heparan sulphate/heparin chain elongation, HS/heparin GlcNAc N-deacetylase/N-
sulphotransferase (NDST1-4) catalyzes N-deacetylation and N-sulphation of GlcNAc
residue of HS and heparin [Aikawa et al, 2001; Kjellén L, 2003], which are
prerequisites for all of other modifications such as C5-epimerization of GlcA to
iduronic acid (IdoA) [Campbell et al, 1994; Li et al, 2001], 2-O-sulphation of the
Page | 20
Chapter 1 Introduction uronic acids, and 6-O- and 3-O-sulphation of glucosamine residues, which are catalyzed by uronyl C5-epimerase (GLCE), 2-O-sulphotransferase (HS2ST1), 6-O- sulphotransferase (HS6ST1-3), and 3-O-sulphotransferase (HS3ST1-5), respectively
[Kusche-Gullberg and Kjellén L, 2003].
Page | 21
Chapter 1 Introduction
Table 1.5 Genes coding for enzymes involved in the synthesis of CSPG/DSPG and HSPG.
Gene Gene name Function
Tetrasaccharide Linkage Region Synthesis XylTI xylosyltransferase I Xylosylation of core protein XylTII xylosyltransferase II Xylosylation of core protein B4GalT7 xylosylprotein beta 1,4-galactosyltransferase, polypeptide 7 Addition of first galactose in linkage region (galactosyltransferase I) B3GalT6 UDP-Gal:betaGal beta 1,3-galactosyltransferase polypeptide 6 Addition of second galactose in linkage region B3GAT3 beta-1,3-glucuronyltransferase 3 (glucuronosyltransferase I) Transfer GlcA to linkage region
Chondroitin/Dermatan Sulphate Synthesis CSGalNAcT-1 chondroitin sulphate N-acetylgalactosaminyltransferase 1 Initiates biosynthesis of CS/DS biosynthesis by adding the first GalNAc CSGalNAcT-2 chondroitin sulphate N-acetylgalactosaminyltransferase 2 Transfer GalNAc to growing oligosaccharide chain CSGLCA-T chondroitin sulphate glucuronyltransferase Transfer of GlcA to growing oligosaccharide chain CHPF chondroitin polymerising factor Chain polymerization: alternate addition of GlcA and GalNAc CHSY1 chondroitin synthase I Chain polymerization: alternate addition of GlcA and GalNAc CHSY3 chondroitin synthase 3 Chain polymerization: alternate addition of GlcA and GalNAc DSE, SART2 dermatan sulphate epimerase Epimerization of GlcA to IdoA CHST13 carbohydrate (chondroitin 4) sulphotransferase 13 Sulphates GalNAc in CS at 4-O-position CHST11 carbohydrate (chondroitin 4) sulphotransferase 11 Sulphates GalNAc in CS at 4-O-position CHST12 carbohydrate (chondroitin 4) sulphotransferase 12 Sulphates GalNAc in CS at 4-O-position
Page | 22
Chapter 1 Introduction
CHST7 carbohydrate (N-acetylglucosamine 6-O) sulphotransferase 7 Sulphates GalNAc in CS at 6-O-position CHST3 carbohydrate (chondroitin 6) sulphotransferase 3 Sulphates GalNAc in CS at 6-O-position GALNAC4S-6ST B cell RAG associated protein Sulphates GalNAc4S at 6-O-position CHST14, D4ST1 carbohydrate (N-acetylgalactosamine 4-0) sulphotransferase 14 Sulphates GalNAc in DS at 4-O-position UST uronyl-2-sulphotransferase Sulphates GlcA in GlcA-GalNAc6S or IdoA in IdoA- GalNAc or IdoA-GalNAc4S
Heparan Sulphate Synthesis EXTL2 exostoses (multiple)-like 2 Chain polymerization: alternate addition of GlcA and GlcNAc EXTL3 exostoses (multiple)-like 3 Chain polymerization: alternate addition of GlcA and GlcNAc EXTL1 exostoses (multiple)-like 1 Chain polymerization: alternate addition of GlcA and GlcNAc EXT1 exostoses (multiple) 1 Chain polymerization: alternate addition of GlcA and GlcNAc EXT2 exostoses (multiple) 2 Chain polymerization: alternate addition of GlcA and GlcNAc NDST1 N-deacetylase/N-sulphotransferase (heparan glucosaminyl) 1 N-deacetylation and N-sulphation of GlcNAc units NDST2 N-deacetylase/N-sulphotransferase (heparan glucosaminyl) 2 N-deacetylation and N-sulphation of GlcNAc units NDST3 N-deacetylase/N-sulphotransferase (heparan glucosaminyl) 3 N-deacetylation and N-sulphation of GlcNAc units NDST4 N-deacetylase/N-sulphotransferase (heparan glucosaminyl) 4 N-deacetylation and N-sulphation of GlcNAc units GLCE glucuronic acid epimerase Epimerisation of GlcA to IdoA HS2ST1 heparan sulphate 2-O-sulphotransferase 1 Sulphates GlcA in HS at 2-O-position HS6ST1 heparan sulphate 6-O-sulphotransferase 1 Sulphates GlcNAc in HS at 6-O-position HS6ST2 heparan sulphate 6-O-sulphotransferase 2 Sulphates GlcNAc in HS at 6-O-position HS6ST3 heparan sulphate 6-O-sulphotransferase 3 Sulphates GlcNAc in HS at 6-O-position
Page | 23
Chapter 1 Introduction
HS3ST1 heparan sulphate (glucosamine) 3-O-sulphotransferase 1 Sulphates GlcNAc in HS at 3-O-position HS3ST2 heparan sulphate (glucosamine) 3-O-sulphotransferase 2 Sulphates GlcNAc in HS at 3-O-position HS3ST3A1 heparan sulphate (glucosamine) 3-O-sulphotransferase 3A1 Sulphates GlcNAc in HS at 3-O-position HS3ST3B1 heparan sulphate (glucosamine) 3-O-sulphotransferase 3B1 Sulphates GlcNAc in HS at 3-O-position HS3ST5 heparan sulphate (glucosamine) 3-O-sulphotransferase 4 Sulphates GlcNAc in HS at 3-O-position HS3ST5 heparan sulphate (glucosamine) 3-O-sulphotransferase 5 Sulphates GlcNAc in HS at 3-O-position SULF1 sulphatase 1 Heparan sulphate 6-O-endosulphatases SULF2 sulphatase 2 Heparan sulphate 6-O-endosulphatases HPSE heparanase Cleaves heparan sulphate chains HPSE2 heparanase 2 Cleaves heparan sulphate chains
CS/HS/DS modifications PAPSS1 3'-phosphoadenosine 5'-phosphosulphate synthase 1 Sulphate donor cosubstrate for all sulphotransferase enzymes PAPSS2 3'-phosphoadenosine 5'-phosphosulphate synthase 2 Sulphate donor cosubstrate for all sulphotransferase enzymes
Page | 24
Chapter 1 Introduction
1.2.3 Enzymatic remodelling of chondroitin/dermatan sulphate and heparan sulphate chains
Physiologically, heparanase coded by the HPSE gene, is an endo-beta-D- glucuronidase that can cleave heparan sulphate side chains of heparan sulphate proteoglycans on cell surfaces and the extracellular matrix [Nasser NJ, 2008]. The degradation of heparan sulphate chains by heparanase at the cell surface and in the extracellular matrix results in the release of bound growth factors and induces tumour growth and angiogenesis. Expression of heparanase in tumour cells have been found to correlate with increased metastatic potential [Vlodavsky et al, 2002; Ilan et al,
2006].
Sulphatases are heparan-6-O-endosulphatases that are found localised to the cell surface or secreted into the extracellular matrix, and selectively remove the 6-O- sulphates from heparan sulphate. The heparan-6-O-endosulphatases are encoded by
HSulf1 and HSulf2. Similarly, HSulf1 and HSulf2 have been implicated in cancer growth [Dai et al, 2005].
Microbial-derived heparan sulphate and chondroitin/dermatan sulphate lyases include the heparinases and the chondroitinases. Heparinases have been classified as heparinase I acting primarily on heparin, heparinase II acting on both heparin and heparan sulphate, and heparinase III, which acts exclusively on heparan sulphate
[Ernst et al, 1995]. The chondroitinases include the chondroitinase AC (degrades chondroitin-4-sulphate and chondroitin-6-sulphate), chondroitinase B (degrades dermatan sulphate), chondroitinase C (degrades chondroitin-6-sulphate) and chondroitinase ABC (degrades chondroitin-4-sulphate, chondroitin-6-sulphate and dermatan sulphate). These enzymes are frequently used and commercially available, for exogenous applications to understand the endogenous roles of the CS/DS and HS.
Page | 25
Chapter 1 Introduction
1.2.4 Functions of chondroitin/dermatan sulphate and heparan sulphate and/or proteoglycans in cancer
The huge repertoire of structurally diverse chondroitin/dermatan sulphate
proteoglycans and heparan sulphate proteoglycans is generated by different protein
cores, post translational modifications of the protein backbone, the number and types
of glycosaminoglycan on the protein cores and post translational modifications of the
GAG chains. Important biological functions of the proteoglycans derive primarily from those of the attached glycosaminoglycan chains, protein component of the molecule and/or possibly function as a whole structural entity. Different sulphate moieties at different positions in the GAG/PG give rise to different structural characteristics of the GAG/PG, leading to specific functional properties of that
GAG/PG.
Being ubiquitously present on the cellular surface and in the ECM and
basement membrane, chondroitin/dermatan sulphates and heparan sulphate/heparin
are well positioned to elicit many cell-cell, cell-ECM and cell-BM signalling events, regulating homeostatic cellular behaviour. Numerous biologically active molecules
(e.g growth factors like FGF, HGF, PDGF, VEGF and chemokines) can bind to
HSPGs, which then function as co-receptors to elicit downstream signalling. HSPGs
can also bind to HS/heparin binding molecules, where they accumulate and
concentrate, and are released when appropriate signalling mechanisms are detected
[Sanderson RD, 2001]. For instance, the binding of FGF to HSPG and FGFR induces
receptor dimerization, activation and autophosphorylation of multiple tyrosine
residues in the cytoplasmic domain of the receptor. This leads to phosphorylation of
several molecules like Shc, phospholipase-Cgamma and STAT1 that elicit cell
Page | 26
Chapter 1 Introduction proliferation, cell differentiation, cell migration and cell survival [Eswarakumar et al,
2005].
Chondroitin/dermatan sulphates have also been reported to be able to interact with various growth factors such as heparin binding growth factors like FGF-2, pleiotrophin and midkine, and thus play important roles in cell signaling [Delehedde et al, 2001]. The CSPG, protein-tyrosine phosphatase zeta (PTPzeta), is a receptor for midkine and pleiotrophin, and upon binding, induce downstream signalling that includes ERK and PI3 kinase, thus promote the growth, survival, and migration of various cells [Qi et al, 2001]. The DSPG, decorin, modulates epidermal growth factor, insulin-like growth factor signalling and thus controls cell proliferation and angiogenesis.
Dysregulation of these molecules has thus important functional repercussions, as they could play many roles in diseases such as cancer. Indeed, many studies on the altered composition and structure of glycosaminoglycans and proteoglycans in numerous cancers such as gastric carcinoma, melanoma, rectum carcinoma, breast cancer, squamous cell laryngeal carcinoma and pancreatic cancer have been reported
[Pirinen et al, 2005; Pukkila et al, 2007; Pukkila et al, 2004; Suwiwat et al, 2004;
Voutilainen et al, 2003; Yang et al, 2004].
1.2.4.1 Heparan sulphate in cancer biology
The heparan sulphate proteoglycan syndecans consist of four family members, syndecan 1-4. Syndecan-1 has been more widely studied; elevated expression of syndecan-1 has been reported in breast cancer [Loussouarn et al, 2008], gall bladder cancer [Roh et al, 2008] and in the extracellular matrix of ovarian tumours [Davies et al, 2004], and is predictive of poor prognosis in all the cancers. Syndecan-1 may
Page | 27
Chapter 1 Introduction interact with heparin-binding growth factors, such as fibroblast growth factors, in the stromal milieu to stimulate carcinoma development and tumour angiogenesis
[Mundhenke et al, 2002].
Syndecan-1 expression in prostate adenocarcinoma cases was found to correlate with established features of biologically aggressive cancer such as higher
PSA levels, higher Gleason sum and occurrence of lymph nodes metastases [Shariat et al, 2008; Chen et al, 2004; Kiviniemi et al, 2004). In addition, syndecan-1 expression was also associated with significantly greater risk of PSA-progression after surgery [Shariat et al, 2008].
Glypicans, GPI-anchored cellular membrane proteoglycans, consist of 6 family members, glypican 1-6. Glypican-1 and glypican-3 are the more widely reported glypicans in cancer expression and biomarker studies; glypican-1 was reported to be highly expressed in breast cancer [Matsuda et al, 2001], gliomas [Su et al, 2006], pancreatic cancer [Kleef et al, 1998], and its expression is a poor prognostic factor for survival. The differential expression of glypican-3 could be tissue-specific, as glypican-3 was found to be overexpressed in hepatocellular carcinoma [Nakatsura et al, 2003] and malignant melanoma [Ikuta et al, 2005] while downregulated in breast cancer [Xiang et al, 2001], ovarian [Lin et al, 1999] and gastric cancer [Zhu et al, 2002]. As a member of the HSPG family, glypican -1 and -3 can modulate the mitogenic effects of the heparin-binding growth factors such as FGF2 and TGFβ – dependent signalling. Glypican-3 has also been shown to inhibit growth and invasion in breast cancer cells by activating the Wnt signalling pathway [Stigliano et al, 2009].
However, what remains to be clarified is whether there are tissue-specific functions of glypican-3, which may have led to differing expression changes in different cancers,
Page | 28
Chapter 1 Introduction
or perhaps there could be differences in importance of glypican-3 functions in
different cancer progression.
Increased heparanase has also been detected in many malignant tumours like
renal cell carcinoma [Mikami et al, 2008], endometrial cancer [Inamine et al, 2008]
and head and neck cancers [Doweck et al, 2006], and these have been significantly
correlated with metastatic potential. Heparanase is a heparan sulphate (HS) degrading
endoglycosidase that regulates extracellular matrix degradation and remodelling. The
secretion of heparanase by tumour cells will allow the malignant cells to degrade
heparan sulphate and penetrate the basement membrane and ECM. Heparanase can
also release bioactive cell-surface HS fragments that can stimulate cell migration and angiogenesis, thus contributing to aggressive cancer phenotype [Roy and Marchetti,
2009]. Interestingly, besides enzymatically cleaving heparan sulphate chains, heparanase was noted to exert other biological functions. Nonenzymatic functions of heparanase include enhanced cell adhesion [Goldshmidt et al, 2003] and induction of p38 and Akt phosphorylation [Gingis-Velitski et al, 2004; Ben-Zaken et al, 2007]. In addition, enzymatically active and inactive heparanase were noted to induce VEGF expression and enhance EGFR phosphorylation, resulting in angiogenesis and increased cell migration and cell proliferation [Cohen-Kaplan et al, 2008].
Upregulation of heparanase can accelerate the shedding of syndecan-1 from the cell surface thus influencing the expression and localization of syndecan-1. Recently,
Chen and Sanderson, 2009, discovered that heparanase also alters the level of nuclear syndecan-1 [Chen and Sanderson, 2009]. Although the function of nuclear heparan sulphate is not well understood, there is emerging evidence that it may act to repress transcriptional activity. The loss of nuclear syndecan-1 due to heparanase could lead
Page | 29
Chapter 1 Introduction
to transcriptional expression of MMP-9, VEGF and other effectors that condition the
tumour microenvironment to promote an aggressive cancer phenotype.
Human prostatic adenocarcinoma cells are reported to have increased
heparanase production due to promoter hypomethylation and upregulation of the
transcriptional factor early growth response-1 [Ogishima et al, 2005; de Mestre et al,
2005]. There was a strong correlation between tumour grade and heparanase
expression. Overexpression of heparanase in PC-3 cells results in dramatic five-fold
increase in the metastatic potential [Lerner et al, 2008]. The degradation of the
basement membrane by heparanase could be a potential factor in its increased
metastatic potential [Kosir et al, 1999].
The heparin-degrading endosulphatases, SULF1 and SULF2, have also been
reported to be differentially expressed in several cancers (ovarian, hepatocellular,
gastric, breast cancer) [Narita et al, 2007; Lai et al, 2008b; Chen et al, 2009]. SULF1 was found downregulated in many cancers, and was postulated to be a tumour suppressor gene; overexpression of SULF1 in cancer cells lead to reduced sulphation
of the HSPGS and subsequent diminished phosphorylation of receptor tyrosine kinases that require sulphated HSPGs as co-receptors for their cognate ligands [Lai et
al, 2008a]. Consequently, downstream signaling through the extracellular signal-
regulated kinase pathway is impeded and this results in reduced proliferation. In
addition, this decreased heparin-binding growth factor signalling promotes apoptosis and inhibits angiogenesis. On the other hand, HSULF2 was reported to be upregulated in breast and hepatocellular carcinoma (HCC) [Morimoto-Tomita et al, 2005; Lai et al, 2008b]. Expression of SULF2 activates MAPK and Akt pathways and promotes cell growth and angiogenesis [Lai et al, 2008c].
Page | 30
Chapter 1 Introduction
Perlecan, a heparan sulphate proteoglycan found in basement membranes, is highly expressed in prostate cancer cell lines, including androgen insensitive cell lines and cell lines selected for metastatic properties [Datta et al, 2006]. Perlecan expression in prostate cancer tissues also correlates with high tumour grade (high
Gleason score). Perlecan was reported to regulate prostate cell growth via the sonic hedgehog pathway [Datta et al, 2006], as well as activating heparin-binding growth factor receptors such as FGF2 and VEGF [Savorè et al, 2005].
1.2.4.2 Chondroitin/dermatan sulphate in cancer biology
Structural modifications to chondroitin sulphate chains could be a means by which cancer cells employ in their quest to invade and spread. Highly sulphated chondroitin sulphate, CS-E, was found in higher proportions in highly metastatic
Lewis lung carcinoma cells than in cells with low metastatic potential [Li et al, 2008].
Inhibition of the CSE epitopes strongly inhibited the invasiveness of the metastatic cell line. In a separate study, Sugahara et al, 2008, reported that CSE fragments enhance CD44 cleavage and promote CD44-mediated cell motility [Sugahara et al,
2008]. Fthenou et al, 2006, reported the role of chondroitin sulphate A in the regulation of fibrosarcoma cell adhesion and motility via the JNK and tyrosine kinase pathways and also enhances the PDGF-mediated signalling in fibrosarcoma cells
[Fthenou et al, 2006]. Melanoma chondroitin sulphate proteoglycan has also been found to enhance FAK and extracellular signal regulated signalling pathways [Yang et al, 2004].
Decorin, a dermatan sulphate proteoglycan found to be downregulated in many cancers (e.g colon adenocarcinoma [Adany and Iozzo, 1991]; breast cancer
[Leygue et al, 2000], ovarian cancer [Nash et al, 2002], suppresses tumour
Page | 31
Chapter 1 Introduction angiogenesis through down-regulation of vascular endothelial growth factor production by cancer cell transduction [Grant et al, 2002]. In addition, decorin caused cell cycle arrest through mitogen-activated protein (MAPK) kinase signal pathway, which led to endogenous p21 activation, a potent inhibitor of cyclin-dependent kinases [Moscatello et al, 1998; Ständer et al, 1999]. It has also been reported that decorin can result in EGFR degradation by endocytosis of the EGFR, thus resulting in attenuation of its signaling pathway [Seidler et al, 2006]. Downregulation of decorin in tumours promotes malignant characteristics in tumours.
High performance liquid chromatography analysis of glycosaminoglycans extracted from prostate cancer samples revealed that chondroitin-6-sulphate and dermatan sulphate were the predominantly upregulated components [Iida et al, 1997].
Moreover, elevated levels of sulphated chondroitin were reported to be associated with poorer disease outcome in early stage prostate cancer [Ricciardelli et al, 1997 and 1999]. In patients with metastatic disease, chondroitin sulphate is upregulated but the increased expression does not predict survival. Moreover, Sakko et al, 2008, have recently shown the importance of non-sulphated chondroitin as an independent predictor of PSA relapse [Sakko et al, 2008].
Previous studies have shown that expression of versican, a chondroitin sulphate proteoglycan, is increased in prostate cancer and inhibits cancer cell adhesion to fibronectin in the stromal matrix [Ricciardelli et al, 1998; Sakko et al, 2003].
Increased versican expression is also associated with increased risk of PSA relapse in early-stage prostate cancer [Ricciardelli et al, 1998]. TENB2, a chondroitin sulphate proteoglycan, has been found upregulated in prostate adenocarcinoma, with particular prominence in high grade tumours [Glynne-Jones et al, 2001]. The presence of an
Page | 32
Chapter 1 Introduction
EGF and two follistatin domains in TENB2 points to a role of this CSPG in growth factor signalling.
A heparan sulphate and chondroitin sulphate proteoglycan, TGFβ receptor type III (TGFβR3, betaglycan), was reported to have decreased expression in the majority of human prostate cancers as compared with benign prostate tissue at both the mRNA and protein level [Sharifi et al, 2007; Turley et al, 2007]. Loss of betaglycan expression correlates with advancing tumour stage and a higher probability of prostate-specific antigen (PSA) recurrence. There were decreased cell motility and cell invasion through Matrigel in vitro and prostate tumourigenicity in vivo when betaglycan was overexpressed in prostate cancer cells [Turley et al, 2007;
Bandyopadhyay et al, 2005].
It appears that proteoglycans play significant roles in prostate cancer cell behaviour and progression. However, the vast functional repertoire of different chondroitin sulphates, heparan sulphates and their proteoglycans in prostate cancer are not completely investigated and understood. Delineating specific chondroitin sulphate and heparan sulphate structures that are involved in prostate cancer progression will provide more avenues to identify potential biomarkers and therapeutic targets.
1.3 Objectives of project
Identifying biomarkers for improved detection of aggressive prostate cancer will improve the management of malignancy. In addition, identification of new molecular targets could facilitate innovation of novel therapies that could halt progression of localized prostate cancer towards metastasis or that could target metastatic prostate cancer. Hence, the objectives of this dissertation are to:
Page | 33
Chapter 1 Introduction
1. Establish the application of different chondroitin sulphate and heparan
sulphate subtypes as biomarkers of prostate cancer progression, by first
determining their expression patterns and levels in prostate cancer tissues via
immunohistochemistry, and then further determining whether they are
associated with current prognostic markers.
2. Identify genes coding for chondroitin sulphate and heparan sulphate
biosynthesis enzymes that are deregulated in prostate cancer by microarray
profiling and real-time polymerase chain reaction, and then determine their
functional significance in prostate cancer phenotype using silencing RNAs.
Page | 34
Chapter 2 Materials and Methods
Chapter 2
Materials and Methods
2.1 Genome-wide expression profiling of prostate adenocarcinoma tissues
2.1.1 Collection of prostate adenocarcinoma tissues
Prostate adenocarcinoma samples and control benign prostatic hyperplasia tissue specimens were obtained from 11 patients who underwent open radical retropubic prostatectomy (RRP) or transurethral resection of the prostate (TURP) from 2005 to 2006 at the Department of Urology, Tan Tock Seng Hospital. Ethics approval was obtained from the Institutional Review Board, Tan Tock Seng Hospital. The tissue samples were snap-frozen in liquid nitrogen and stored at -80oC.
2.1.2 Haematoxylin and eosin staining of prostate tissues
The frozen prostate tissue sections collected (in Section 2.1.1) were sectioned using a cryostat. The slide sections were then fixed in 4% paraformaldehyde overnight, washed thrice in 1X PBS and stained in Shandon’s haematoxylin for 5 minutes. After washing in deionized water, the sections were immersed in differentiating fluid (70% alcohol with
HCl), blued in tap water for 5 minutes, washed in deionized water and eosin-stained for 5 minutes. Dehydration of the sections was done in 3 changes of absolute ethanol followed by 3 changes of Histoclear.
2.1.3 RNA isolation, yield and quality from prostate tissues
Total RNA was isolated from each prostate adenocarcinoma and benign prostatic hyperplasia tissue using the RNeasy Fibrous Tissue Mini Kit (Qiagen, Hilden, Germany),
Page | 35
Chapter 2 Materials and Methods
as per the manufacturer’s protocol. RNA yield was quantified using Nanodrop while
RNA purity was determined using the absorbance ratio of A260/A280. The ratio of the absorbance values measured at 260nm (A260) to absorbance values measured at 280nm
(A280) provides an estimate of RNA purity, in consideration of contaminants that absorb in the UV spectrum. As pH changes can affect the absorbance, the guideline is that pure
RNA has an A260/A280 ratio of 1.9–2.1 in 10 mM Tris·Cl, pH 7.5 [Wilfinger et al, 1997].
RNA quality was checked using the Agilent RNA 6000 Nano Kit (Agilent Technologies
Inc, California, United States of America) on the Agilent 2100 Bioanalyzer System.
Quality of the RNA was assessed based on the electrophoretic trace of the RNA sample and the RNA integrity number (RIN). For the electrophoregram, 2 distinct peaks representing the 18S and 28S ribosomal subunits should be present in intact total RNA.
The RIN is assigned through the Agilent Bioanalyzer, where it determines the integrity of the RNA; RIN of ‘10’ stands for a totally undegraded perfect RNA sample while a RIN of ‘1’ represents a completely degraded sample [Schroeder et al, 2006].
2.1.4 Microarray Analysis
RNA labeling and hybridization on the Affymetrix Genechip U133 Plus 2 chips
(Affymetrix Inc, California, United States of America) were performed according to the manufacturer’s protocol. Briefly, the two-cycle target labeling protocol was used to synthesize biotin-labeled cRNA from 100ng of total RNA starting material. The biotinylated cRNA targets are then cleaned up, fragmented, and hybridized to Affymetrix
Genechip U133 Plus 2 expression arrays. Washing and staining of the chips was performed using the Fluidics Station 450 where it is operated using the GeneChip
Page | 36
Chapter 2 Materials and Methods operating software (GCOS). Scanning of the probe arrays was done using Affymetrix
GeneChip Scanner 3000.
2.1.5 Analysis of microarray data
Microarray data was analyzed using 4 different softwares: GeneChip Operating Software
(GCOS), dCHIP, Robust Multi-array Average (RMA) and Genespring. Raw data were normalized according to the algorithms in the software (Table 2.1). Fold changes were calculated based on mean values. Genes that were found to be common in all gene lists, as determined by the criteria used in individual software, were merged.
2.1.6 cDNA synthesis and quantitative real-time polymerase chain reaction
Total RNA was converted into cDNA using the SuperScript III First-strand cDNA synthesis kit (Invitrogen Corporation, California, United States of America) according to manufacturer’s protocol. The cDNA was then used for real-time PCR using the
QuantiTect SYBR Green PCR Kit (Qiagen) on the Roche LightCycler 2 machine (Roche
Diagnostics Corporation, Roche Applied Science, Indianapolis, United States of
America). The programme settings used for the real-time PCR (qPCR) run is shown in
Table 2.2. Primers used for real-time PCR (qPCR) are as shown in Table 2.3.
Page | 37
Chapter 2 Materials and Methods
Table 2.1 Methods of normalization and criteria for determining genes differentially expressed using different microarray analysis software, for microarray work involving human prostate tissues.
Software Normalization and Expression Criteria values GCOS 1. One array was chosen as 1. Change calls (i.e baseline array, to which normalized expression of other arrays are normalized genes that are against. Thus, for 6 control differentially expressed as BPH samples and 6 prostate compared to the baseline adenocarcinoma samples, array – increase or there were a total of 36 decrease) are the same in different comparisons. at least 50% of the 36 different comparisons. 2. Fold change ≥ 2 3. P-value < 0.05 dCHIP 1. Array with the median 1. Detection ‘P’ call present overall intensity was used as in at least 50% of either the baseline array against control BPH group or which other arrays are prostate adenocarcinoma normalized (Li and Wong, group. 2001). 2. Fold change ≥ 2 2. PM/MM difference model- 3. P-value < 0.05 based expression values were calculated. (Li and Wong, 2001). RMA 1. RMA does not use a base 1. Fold change ≥ 2 line array, but "quantile 2. P-value < 0.05 normalization", based on the complete data set (Wu and Irizarry, 2004) Genespring 1. Methods as in ‘RMA’ 1. Fold change ≥ 2 software 2. P-value < 0.05
Page | 38
Chapter 2 Materials and Methods
Table 2.2 Programme settings for qPCR
Program Temperature Duration Ramp rate Cycle(s) Initial Activation 95oC 15 min 20oC/sec 1 3-step cycling 45 Denaturation 94oC 15 sec 20oC/sec Annealing 60oC 25 sec 20oC/sec Extension 72oC 12 sec 20oC/sec Melting Curve 95 oC 0 sec 20oC/sec 1 65 oC 15 sec 20oC/sec 95 oC 0 sec 0.1 oC/sec Cooling 40 oC 30 sec 20oC/sec 1
Page | 39
Chapter 2 Materials and Methods
Table 2.3 Primer sequences used in real-time PCR (qPCR) for microarray data verification Gene Symbol Gene Name Gene Forward primer (5’-3’) Reverse primer (5’-3’) Amplicon ID Size ALDH1A2 aldehyde dehydrogenase 1 family, 8854 ccattggagtgtgtggacag gatgagggctcccatgtaga 152 member A2 AMACR alpha-methylacyl-coA racemase 23600 gcatgtgaggaaggggtaga caccttcctcttgcgattgt 145 B3GALT6 UDP-Gal:betaGal beta 1,3- 126792 gcctctactggggcttcttc cgcgtagggcaggtagtagt 101 galactosyltransferase polypeptide 6 CHST2 carbohydrate (N-acetylglucosamine-6- 9435 aagttcgagagcatggagga agttgggcgaagtactggtg 146 O) sulphotransferase 2 COL4A6 collagen, type IV, alpha 6 1288 cagacccagattccccagta gcagtgtgcatgaggaaaga 152 CSGalNAcT-1 chondroitin beta1,4 N- 55790 gagatgtgcattgagcagga gaagttggcagctttggaag 114 acetylgalactosaminyltransferase CXCL13 chemokine (C-X-C motif) ligand 13 (B- 10563 aatgagcctggactcagagc gacctccagaacaccttgga 126 cell chemoattractant) CYP3A5 cytochrome P450, family 3, subfamily 1577 taggcccagtgggatttatg caatgatggggaacatctcc 126 A, polypeptide 5 CYP4B1 cytochrome P450, family 4, subfamily 1580 cctaaggcccctgatgtgta cagtgaacacggccacatag 151 B, polypeptide 1 DCN decorin 1634 gcccagaagttcctgatgac gagttgtgtcagggggaaga 144 EFEMP2 EGF-containing fibulin-like 30008 gcccaaacctgtgtcaactt atgaaggctgctctcgacat 136 extracellular matrix protein 2 FASN fatty acid synthase 2194 tcatccccctgatgaagaag actccacaggtgggaacaag 120 FBP1 fructose-1,6-bisphosphatase 1 2203 tcatcgcactctggtctacg gccctctggtgaatgtctgt 186 FLRT3 fibronectin leucine rich transmembrane 23767 accctcaatcgagagcaaga tgagaagagcgatccattcc 147 protein 3
Page | 40
Chapter 2 Materials and Methods
GALNT7 UDP-N-acetyl-alpha-D- 51809 gcacgaagtattggtgctca ctataagcggcacagtgcaa 145 galactosamine:polypeptide N- acetylgalactosaminyltransferase 7 (GalNAc-T7) GATA3 GATA binding protein 3 2625 aaggcagggagtgtgtgaac cttcgcttgggcttaatgag 142 GJA1 gap junction protein, alpha 1, 43kDa 2697 gtgcctgaacttgccttttc ccctccagcagttgagtagg 166 (connexin 43) GLYATL1 glycine-N-acyltransferase-like 1 92292 ttctcacctggagccttgat ggtaaccccgtctttcccta 147 HOXC6 homeobox C6 3223 ccaggaccagaaagccagta ggtctggtaccgcgagtaga 121 HPN hepsin 3249 gctgtgtggcattgtgagtt tggcttcggagtgagtcttt 122 HS3ST3A1 heparan sulphate (glucosamine) 3-O- 9955 cagtgccctctccacctc gccaggcagtagaagacgtaa 108 sulphotransferase 3A1 HS6ST2 heparan sulphate 6-O-sulphotransferase 90161 ccctggtaggctgctacaac aaactcagtgaggccgaaga 116 2 KRT15 keratin 15 3866 taccagaagcagaccccaac gcattgtcgatctccaggat 131 LAMB3 laminin, beta 3 3914 caagcctgagacctactgcac ctgcagagagacagggttcac 172 LTF lactotransferrin /// similar to 4057 ccctacaaactgcgacctgt ctgtgtggcaggacttcaga 142 lactotransferrin MAL2 mal, T-cell differentiation protein 2 114569 gacatcctgcggacctactc gctgtcacggacacaaacat 137 MYC v-myc myelocytomatosis viral 4609 ttcgggtagtggaaaaccag cctcctcgtcgcagtagaaa 120 oncogene homolog (avian) MYO6 myosin VI 4646 gcagtcgctggaagaaagtt aatgcgaggtttgtgtctcc 143 PCA3 prostate cancer specific antigen 3 50652 caaaggcagggaacctcata ctaaagcgcactcaccatga 145 PDGFC platelet derived growth factor C 56034 gccaggttgtctcctggtta ctgacaccggtctttggtct 141 PDLIM5 PDZ and LIM domain 5 10611 aatccttgcccagatcactg cgtgctgaaacttcatggtg 144 PPAP2B phosphatidic acid phosphatase type 2B 8613 ctgtctcgcgtatcagacca cagggagagcgtcgtcttag 132 RARRES2 retinoic acid receptor responder 5919 ttaagctgcagcagacaagc ccgcagaacttgggtctcta 167
Page | 41
Chapter 2 Materials and Methods
(tazarotene induced) 2 SCUBE2 signal peptide, CUB domain, EGF-like 57758 ctggtcgaacttcctgcttc caggctggtatgttcccact 165 2 SDC1 syndecan 1 6382 gggactcagccttcagacag ctcgtcaatttccaggagga 128 SGEF Src homology 3 domain-containing 26084 gatgtctgtgaggcaagcaa ggtcaaatgtggatgctgtg 118 guanine nucleotide exchange factor SLC14A1 solute carrier family 14 (urea 6563 ccagtgggagttggtcagat aagactgagtcccgctgcta 144 transporter), member 1 (Kidd blood group) SLC2A5 solute carrier family 2 (facilitated 6518 ccaagaaagccctacagacg aacagcttcagcacggagat 118 glucose/fructose transporter), member 5 SPOCK3 sparc/osteonectin, cwcv and kazal-like 50859 cagggaagtggcaaacagat acatccagccaagtgagtcc 152 domains proteoglycan (testican) 3 SRD5A2 steroid-5-alpha-reductase, alpha 6716 tgaataccctgatgggtggt acaagccaccttgtggaatc 155 polypeptide 2 (3-oxo-5 alpha-steroid delta 4-dehydrogenase alpha 2) SRPRB signal recognition particle receptor, B 58477 cgttgctctttgtcaggttg gggaaggtcaatcaaggtca 125 subunit STEAP4 STEAP family member 4 79689 ttgccttccttcatgtcctc gctgaggaggtgctaaatgg 127 SULF2 sulphatase 2 55959 gagctgcaagggttacaagc cttcccacagttgtcccagt 158 TCEAL2 transcription elongation factor A (SII)- 140597 acagaaaacacgggcaagag tctttgatcctccctgcatc 130 like 2 TGFB3 transforming growth factor, beta 3 7043 ggatcaccacaaccctcatc gcagttctcctccaagttgc 133 TGFBR3 transforming growth factor, beta 7049 ccaagatgaatggcacacac ccatctggccaaccactact 151 receptor III (betaglycan, 300kDa) TP73L tumour protein p73-like 8626 gaggttgggctgttcatcat gaggagaattcgtggagctg 174 TPD52 tumour protein D52 7163 cacagagaccctctcggaag ccaagtttccgcttgatctc 135
Page | 42
Chapter 2 Materials and Methods
TRIM29 tripartite motif-containing 29 23650 acatcataccagccctcgtc agcctttcagggagaaggag 184 TRIM36 tripartite motif-containing 36 55521 gccagtttgccatctgtgta cacctggctttccttaccaa 136 UAP1 UDP-N-acteylglucosamine 6675 gactgtggagcaaaggtggt gtccgtctgagcttcgtttt 139 pyrophosphorylase 1 VCAN chondroitin sulphate proteoglycan 2 1462 ggtgcactttgtgagcaaga ttcgtgagacaggatgcttg 159 (versican) VSNL1 visinin-like 1 7447 tgagcatgaactcaagcagtg tgaactctcggaagtcaatgg 188 WIF1 WNT inhibitory factor 1 11197 tgctttaatggagggacctg ccctggtaacctttggaaca 158
Page | 43
Chapter 2 Materials and Methods
2.2 Expression analysis of chondroitin sulphate and heparan sulphate in prostate adenocarcinoma tissues using immunohistochemistry
2.2.1 Clinical samples and patient data collection
Archival prostate adenocarcinoma samples and paired benign prostatic tissue specimens from corresponding patients were obtained from the Department of Pathology, Singapore
General Hospital. The samples were collected from 163 patients who underwent radical prostatectomy from 2001 to 2004. Targeted areas of the specimens were identified and used for tissue microarray construction using the Beecher arrayer as previously described
[Tan et al, 2005], with a minimum of two cores of tissue taken from each formalin-fixed paraffin-embedded sample block. Clinicopathological data were collected for statistical analysis. Ethics approval was obtained from the Institutional Review Board, Singapore
General Hospital.
2.2.2 Immunohistochemical staining
Monoclonal antibodies targeting heparan sulphate and various chondroitin sulphate subtypes were used, as shown in Table 2.4. After deparaffinization and rehydration of the tissues, endogenous peroxidase activity was quenched with 3% aqueous hydrogen peroxide for 30 min and washed in Tris-buffered saline, pH7.6 thereafter. For the monoclonals 1B5 and 3B3, the sections were treated with 0.025 units/ml chondroitinase
ABC (Sigma-Aldrich) and incubated for 2 h at 37oC. Thereafter, the tissue sections were washed thrice with 1X PBS, and then blocked with serum (from the species in which the secondary antibody is made. For 1B5, 3B3, CS56, F58-10E4 staining, horse serum is used while for LY111 staining, goat serum is used) for 1 hr at room temperature. For the monoclonals LY111, CS56 and F58-10E4, the sections were blocked with serum for 1 h
Page | 44
Chapter 2 Materials and Methods at room temperature, without the need for chondroitinase ABC treatment. All sections were then incubated at 4°C overnight with the respective primary antibodies (at their optimal primary antibody dilution, Table 2.4) after blocking. After 3 washes, biotinylated secondary antibody was added and incubated for 1 h at room temperature, followed by visualization of the staining by the avidin-biotin-complex technique using diaminobenzidine as the substrate (ABC kit, Vector Laboratories). The sections were counterstained with haematoxylin. Control sections were processed using the same protocol, but the chondroitinase ABC incubation step was omitted (for 3B3 and 1B5) while the primary antibodies were not added (for LY111, CS56 and F58-10E4).
Table 2.4 Antibodies and optimal conditions used for immunohistochemistry
Antibody Company Antigen Primary Secondary DAB Antibody Antibody staining dilution dilution duration 1B5 Seikagaku Chondroitinase 1:100 1:600 10 min (monoclonal ABC digested mouse IgG) non-sulphated chondroitin LY111 Seikagaku Chondroitin-4- 1:100 or 1:400a 1:200 10 min (monoclonal sulphate mouse IgM) 3B3 Seikagaku Chondroitinase 1:100 1:600 10 min (monoclonal ABC digested mouse IgM) chondroitin-6- sulphate CS56 Sigma- Chondroitin- 1:1000 1:600 7 min (monoclonal Aldrich 2,6-sulphate mouse IgM) F58-10E4 Seikagaku Heparan 1:100 or 1:800 a 1:200 15 min (monoclonal sulphate mouse IgM) a Different batches of antibodies were used – different optimization conditions have to be used to detect similar immunohistochemical staining.
Page | 45
Chapter 2 Materials and Methods
2.2.3 Immunohistochemical evaluation
Sections were assessed for expression of chondroitin sulphate or heparan sulphate in secretory and basal cells of benign prostatic tissue, in cancerous cells of adenocarcinoma specimens, and in the stromal components of both groups. The intensity of staining was noted as absent (0), weak (1+), moderate (2+) or strong (3+). The percentage of cells stained was also recorded. For a tissue section stained heterogeneously with different intensities, a weighted average intensity was computed as follows: Σ(intensityx x percentage of cells stained with intensityx) ÷ total percentage of cells stained. An immunoreactivity score (IRS) was calculated for each section by multiplying the weighted average staining intensity by the percentage of cells stained. Sections from different cores of the same specimen were independently scored, and average values for staining intensity, percentage of cells stained and IRS calculated for each sample.
2.2.4 Statistical analysis
Statistical analysis was performed using SPSS for Windows Version 12.0. Paired T-test or one-way ANOVA was used to compare the means between variables. Associations between the immunohistochemical staining data and clinicopathological parameters were determined using Fisher’s exact test. A P-value of < 0.05 was considered to be statistically significant.
Page | 46
Chapter 2 Materials and Methods
2.3 In vitro cell culture
2.3.1 Cell lines
RWPE-1 cell line was purchased from ATCC and was propagated in Keratinocyte-serum free medium, supplemented with 0.05mg/ml bovine pituitary extract and 5ng/ml recombinant epidermal growth factor (Invitrogen Corporation). The LNCaP and PC-3 human prostate cancer cell lines were kind gifts from A/P Marie Veronique Clement
(Department of Biochemistry, National University of Singapore) Both LNCaP and PC-3 cells were cultured in RPMI-1640 (Invitrogen Corporation) supplemented with 10% Fetal
Bovine Serum (FBS) (Hyclone, Thermo Fisher Scientific Inc, Massachusetts, United
States of America), 2mM L-glutamine (Hyclone, Thermo Fisher Scientific Inc) and 1mM sodium pyruvate (Hyclone, Thermo Fisher Scientific Inc). All cell lines were incubated in
o a humidified incubator at 37 C in a 5% CO2, 95% air atmosphere.
2.3.2 Storage of cells
The cryopreservation medium used for LNCaP and PC-3 comprised of complete growth medium (see section 2.3.1) supplemented with an additional 20% FBS and 5% cryoprotective agent dimethylsulfoxide (DMSO) (Sigma-Aldrich, Missouri, United States of America). The cryopreservation medium used for RWPE-1 is complete growth medium (see section 2.3.1) supplemented with 15% FBS and 10% DMSO. The cell suspensions were then stored at -80 oC freezer overnight to ensure gradual freezing before storing in liquid nitrogen.
Page | 47
Chapter 2 Materials and Methods
2.3.3 Antibodies used
Antibodies against CSGalNacT-1 (ChGn MaxPab polyclonal antibody B01) and HIPK3
(H-80, rabbit polyclonal IgG) were purchased from Abnova Corporation, Taipei, Taiwan and Santa Cruz Biotechnologies, California, United States of America respectively.
Antibodies recognizing MAPRE-1 (1A11/4, mouse monoclonal IgG) and TPX2 (18D5-1, mouse monoclonal IgG) were obtained from Abcam Inc, Cambridge, Massachusetts,
United States of America. β-actin monoclonal antibody, horseradish peroxidise conjugated anti-mouse IgG secondary antibody, FITC-conjugated anti-mouse IgG secondary antibody were purchased from Sigma-Aldrich, Missouri, United States of
America. Alexa-Fluor 488 conjugated anti-rabbit immunoglobulins secondary antibody was purchased from Invitrogen Corporation, California, United States of America.
2.3.4 RNA Extraction, cDNA synthesis, qPCR of prostate tumour cell lines
Total RNA was isolated from the prostate cell lines, RWPE-1, LNCaP and PC-3 using the RNeasy Mini Kit or RNeasy Micro Kit (Qiagen, Hilden, Germany), as according to the manufacturer’s protocol. RNA yield was quantified using Nanodrop while RNA purity was determined using the absorbance ratio of A260/A280. Total RNA was converted into cDNA using the SuperScript III First-strand cDNA synthesis kit (Invitrogen
Corporation) according to manufacturer’s protocol. The cDNA was then used for real- time PCR using the QuantiTect SYBR Green PCR Kit (Qiagen) on the Roche
LightCycler 2 machine (Roche Diagnostics Corporation, Roche Applied Science).
Primers used for real-time PCR are as shown in Table 2.5. Programme settings used for the real-time PCR run are the same as Section 2.1.6, as shown in Table 2.2.
Page | 48
Chapter 2 Materials and Methods
Table 2.5 Primer sequences used in real-time PCR, for in vitro work involving prostate cancer cell lines.
Gene Symbol Gene Name Gene Forward primer (5’-3’) Reverse primer (5’-3’) Amplicon ID Size ACAN aggrecan 176 gaatcaactgctgcagacca atgctgctcaggtgtgactg 143 ACTB beta-actin 60 ggcatgggtcagaaggatt aggtgtggtgccagattttc 133 B3GalT6 UDP-Gal:betaGal beta 1,3- 126792 gcctctactggggcttcttc cgcgtagggcaggtagtagt 101 galactosyltransferase polypeptide 6 B3GAT3 beta-1,3-glucuronyltransferase 3 26229 ccacctggagagcagtcttc ctcctcctgcttcatcttgg 127 (glucuronosyltransferase I) B4GalT7 xylosylprotein beta 1,4- 11285 caggtggaccacttcaggtt cagcctcaggaaagccatag 145 galactosyltransferase, polypeptide 7 (galactosyltransferase I) CHPF chondroitin polymerising factor 79586 gaacgcacgtaccaggagat ggatggtgctggaataccc 114 CHST11 carbohydrate (chondroitin 4) 50515 atcaaccaccgcttgaaaag ttgatgatcttggtgccgta 140 sulphotransferase 11 CHST12 carbohydrate (chondroitin 4) 55501 gtgtccttcgccaacttcat ttccccacgaagtcgtagtc 131 sulphotransferase 12 CHST13 carbohydrate (chondroitin 4) 166012 atcgggtctcctctcctctc taaaacctgtgaccgcattg 131 sulphotransferase 13 CHST14, carbohydrate (N- 113189 attccccgagttcctgagat cagcctccagcctctcatag 139 D4ST1 acetylgalactosamine 4-0) sulphotransferase 14
Page | 49
Chapter 2 Materials and Methods
CHST3 carbohydrate (chondroitin 6) 9469 acgcccttttcttggttttt agagcttggggaatctgctt 106 sulphotransferase 3 CHST7 carbohydrate (N-acetylglucosamine 56548 ggggcaatctgtcacactct aggatgccttgttgaaatgg 236 6-O) sulphotransferase 7 CHSY1 chondroitin synthase 1 22856 ccctccttcatgaggtttca acaatgtcgtccaaggcttc 149
CHSY3 chondroitin synthase 3 337876 gaggacggctcattgacttc tgctgaagataggcatgacg 158 CPOX coproporphyrinogen oxidase 1371 ggcatcagctgtgtacttcaa ataacagagctcacgcccata 176 CSGalNAcT-1 chondroitin sulphate N- 55790 gagatgtgcattgagcagga gaagttggcagctttggaag 114 acetylgalactosaminyltransferase 1 CSGalNAcT-2 chondroitin sulphate N- 55454 tccccttggagagaaactga cggaagagggtcacatgtct 141 acetylgalactosaminyltransferase 2 CSGLCA-T chondroitin sulphate 54480 cccaacaagccctacaagaa agtgtagctcgggaggtcag 104 glucuronyltransferase CSPG4 chondroitin sulphate proteoglycan 4 1464 cagtatgggcatctcctggt cattgacacccctagccagt 148 CSPG5 chondroitin sulphate proteoglycan 5 10675 gctgcgtaggaccaacaaat cctctttcaggcaggacttg 150 (neuroglycan C) EXT1 exostoses (multiple) 1 2131 cacatcaaagtggacgaacg ttggtccaccatgttcttca 121 EXT2 exostoses (multiple) 2 2132 cgtgttagcccatttgaggt ccaccaagatgtttggcttt 131 EXTL1 exostoses (multiple)-like 1 2134 ggtcgtgtcgtgtccttttc gttctgtctctccgctcacc 136 EXTL2 exostoses (multiple)-like 2 2135 cgtctgtgtttcacccattg gggaaaggaggagaaagagaa 105 EXTL3 exostoses (multiple)-like 3 2137 ggacattgccatgaacttcc atcatgagacagggcctgag 113 GALNAC4S- B cell RAG associated protein 52363 tgagttcacgaccagacagc cgtaccaacaggggctctta 116 6ST GPC1 glypican 1 2817 gatggctgtctggatgacct agccgaggtcttctgtcctt 120 GPC2 glypican 2 221914 aggagggagtgtggtttcct cccacttccaacttccttca 135 GPC3 glypican 3 2719 aactccgaaggacaacgaga gcaccaggaagaagaagcac 113
Page | 50
Chapter 2 Materials and Methods
GPC4 glypican 4 2239 ctcccccaaaccatgttaaa aggcgagatcacagttggaa 149 GPC5 glypican 5 2262 tctcagaagccaagctgaaa ctgtgcccttcacatcttca 143 GPC6 glypican 6 10082 ggaccatgccacaaaaactt ggtccccacagatagctgaa 136 HIPK3 homeodomain interacting protein 10114 gctttcagcaccgtaaccat tggcttgagatcagcatgaa 185 kinase 3 HPSE heparanase 10855 gagaagtgattgattcagttac gcctggtgctctcaaccacctgga 144 HPSE2 heparanase 2 60495 gcagttacctggcaacattgc gtcccgtcgtccaccatcac 600+ HS2ST1 heparan sulphate 2-O- 9653 cgaagtccgagaaattgagc aatgaagtgcttgccgtttt 123 sulphotransferase 1 HS3ST1 heparan sulphate (glucosamine) 3-O- 9957 gaacgaggtccacttcttcg ttctccactgtgagctggtg 105 sulphotransferase 1 HS3ST2 heparan sulphate (glucosamine) 3-O- 9956 agtcggtccctgtcattgtc taaggtctgaggagggctga 137 sulphotransferase 2 HS3ST3A1 heparan sulphate (glucosamine) 3-O- 9955 cagtgccctctccacctc gccaggcagtagaagacgtaa 108 sulphotransferase 3A1 HS3ST3B1 heparan sulphate (glucosamine) 3-O- 9953 ctggcttcaggcaaggagat gacccactgaaaaagctgga 104 sulphotransferase 3B1 HS3ST4 heparan sulphate (glucosamine) 3-O- 9951 ggagcagtcagcattcttcc cagctttctcgccactacct 101 sulphotransferase 4 HS3ST5 heparan sulphate (glucosamine) 3-O- 222537 atcattgtcagggagccaac agggtctatggccagcttct 114 sulphotransferase 5 HS6ST1 heparan sulphate 6-O- 9394 ggcccttcatgcagtacaat tacagctgcatgtccaggtc 103 sulphotransferase 1 HS6ST2 heparan sulphate 6-O- 90161 ccctggtaggctgctacaac aaactcagtgaggccgaaga 116 sulphotransferase 2 HS6ST3 heparan sulphate 6-O- 266722 catctcccccttcacacagt ctcgtaaagctgcatgtcca 112 sulphotransferase 3
Page | 51
Chapter 2 Materials and Methods
HSPG2 heparan sulphate proteoglycan 2 / 429806 gggcaccttgtcttcaggta gacttggatggaacctctgc 129 perlecan LIFR leukaemia inhibitory factor receptor 3977 gcctgcatttacaaggtggt accaatgactgggctttcac 186 MAPRE-1 microtubule associated protein, 22919 cttacagggctgcttgaacc catctcagggctgaaacaca 166 RP/EB family, member 1 NCAN neurocan 1463 ggatcccaggtgtttgaaga gcttgaggagcccagtgtag 146 NDST1 N-deacetylase/N-sulphotransferase 3340 aaggattttggtgccaactg ttggagagctcgatgttgtg 142 (heparan glucosaminyl) 1 NDST2 N-deacetylase/N-sulphotransferase 8509 ctacacacggaccctcaggt ggaaaagacgggactcagtg 134 (heparan glucosaminyl) 2 NDST3 N-deacetylase/N-sulphotransferase 9348 aatgccttggaaagagcaaa tgtgcagcagctttgagagt 114 (heparan glucosaminyl) 3 NDST4 N-deacetylase/N-sulphotransferase 64579 agccaagatcatcaccatcc aagtcagatggagcccaatg 144 (heparan glucosaminyl) 4 PAPSS1 3'-phosphoadenosine 5'- 9061 tcctttgatgtggcgtatga gacctcagttggtccagcat 121 phosphosulphate synthase 1 PAPSS2 3'-phosphoadenosine 5'- 9060 cacccggtcctcctactaca gacggaaagatggcaacaat 125 phosphosulphate synthase 2 RCBTB1 regulator of chromosome 55213 tctgaacctggagcctgaat ggggctcctaggcttgtatc 177 condensation (RCC1) and BTB (POZ) domain containing protein 1 RPLPO ribosomal protein, large, P0 6175 ctgttgcatcagtaccccatt gccttgaccttttcagcaag 103 SDC1 syndecan 1 6382 gggactcagccttcagacag ctcgtcaatttccaggagga 128 SDC2 syndecan 2 6383 ctgctccaaaagtggaaacc tgggtccattttcctttctg 126 SDC3 syndecan 3 9672 accccaactccagagacctt cccacagctaccacctcatt 141 SDC4 syndecan 4 6385 cattagctcccgtcaccact agaggaggtgctcactccaa 104
Page | 52
Chapter 2 Materials and Methods
SMC3 structural maintenance of 9162 aagagtatggagcgctggaa tggggaagtgatccaagttc 158 chromosomes 3 SULF1 sulphatase 1 23213 aagtgacgggttcttggttg ggcaaggtcaaatgaggtgt 117 SULF2A sulphatase 2A 55959 tggccagaaatgaagagacc ggaaatcacccacatggttt 147 SULF2B sulphatase 2B 55959 gaggtggacctccaaaaaca ttgactgagagtgcgtgctt 130 TPX2 TPX, microtubule associated protein 22974 ttcccttgcctcattttgac gcttccaagtctgtgccttc 123 homologue (Xenopus laevis) UST uronyl-2-sulphotransferase 10090 gcaaagctgaacgtgaatga ttccaagcttcctgtgctct 148 VCAN versican 1462 ggtgcactttgtgagcaaga ttcgtgagacaggatgcttg 159 XylTI xylosyltransferase I 64131 ctctggatccacacccaagt ctttggggacagagttctcg 121 XylTII xylosyltransferase II 64132 ctacatgctggtggttcacg caggtagtcggaacgcttgt 116
Page | 53
Chapter 2 Materials and Methods
2.3.5 Silencing of prostate cell lines 3 x 105 PC-3 cells in 6 well plates were seeded and allowed to adhere overnight.
Following this, 40 nM of ON-TARGETplus SMARTpool CSGalNAcT-1 siRNA
(Dharmacon Inc, Colorado, United States of America) was added to the cells, together
with 3ul of DharmaFECT 2 siRNA transfection reagent (Dharmacon Inc), according to
protocol recommended by the manufacturer. ON-TARGETplus siCONTROL Non-
targeting SMARTpool siRNAs (Dharmacon Inc) were used as negative controls while
ON-TARGETplus siCONTROL GAPD SMARTpool siRNAs (Dharmacon, Inc) were used
as positive controls. Silencer Cy3 labeled Negative Control #1 siRNA (Ambion, Inc /
Applied Biosystems, Texas, United States of America) were used to monitor the
transfection efficiency. The sequences for the various siRNAs are listed in Table 2.6. The
siRNA-transfection reagent complexes were removed 24 hours later and replaced with
complete growth medium.
Table 2.6. ON-TARGETplus SMARTpool CSGalNAcT-1 siRNA sequences
Sequence Anti-sense Sense 1 5’-PUGACACGAUAUUUAUGGUUUU AACCAUAAAUAUCGUGUCAUU 2 5’-PUAUUGAUGAAGUCUGACCGUU CGGUCAGACUUCAUCAAUAUU 3 5’PAUACCUUCUUCCCUGGCUGUU CAGCCAGGGAAGAAGGUAUUU 4 5’PAUACUUGCGAUAAAGGUGCUU GCACCUUUAUCGCAAGUAUUU
5 x 104 RWPE-1 cells were seeded in 24 well plates and allowed to adhere overnight. 2ul
of Oligofectamine transfection reagent (Invitrogen Corporation) and a final concentration
of 100nM of HS3ST3A1 siRNA (Ambion, Inc / Applied Biosystems) or HS6ST2 siRNA
(Ambion, Inc / Applied Biosystems) were added to each well. Table 2.7 shows the
Page | 54
Chapter 2 Materials and Methods
sequences of the siRNA used. Ambion’s Silencer siRNA Negative Control #1 was used
as negative controls while Ambion’s Silencer GAPD siRNA was used as positive
controls. Silencer Cy3 labeled Negative Control #1 siRNA (Ambion, Inc / Applied
Biosystems) Complete growth medium with 2% FBS was added 4 hours after
transfection and transfection medium was replaced with complete growth medium 24
hours post-transfection.
Table 2.7 Ambion Silencer siRNA sequences siRNA Anti-sense (5’ -> 3’) Sense (5’ -> 3’) HS3ST3A1 UAGAAGACGUAAAGGGACGtg CGUCCCUUUACGUCUUCUAtt HS6ST2 CAGAUAUUGGGUCUUCCGCtg GCGGAAGACCCAAUAUCUGtt
2.3.6 Western blot Proteins were extracted using M-PER Mammalian Protein Extraction Reagent (Thermo
Fisher Scientific, Massachusetts, United States of America) and resolved by SDS-PAGE
and transferred onto polyvinylidene difluoride membrane (Immun-Blot PVDF
membrane, Bio-Rad Laboratories, California, United States of America) using the semi-
dry method. The membranes were blocked in Tris-buffered saline containing 0.2%
Tween 20 (Bio-Rad Laboratories) and 5% fat-free dry milk (Bio-Rad Laboratories). The
blot was then incubated with the primary antibody and then with horseradish peroxidise-
conjugated secondary antibody. Visualisation of the proteins was determined via
chemiluminescence detection using SuperSignal West Femto Maximun Sensitivity
Substrate (Thermo Fisher Scientific).
Page | 55
Chapter 2 Materials and Methods
2.3.7 GAG extraction and quantification
Glycoaminoglycans were purified from sonicated boiled cellular extracts using 3 volumes of 1.3% potassium acetate in 95% ethanol at -20oC overnight. The precipitates were collected by centrifugation at 14000g for 15min, treated with pronase at 55oC overnight and reboiled. GAGs were further extracted from the supernatant by using 3 volumes of
1.3% potassium acetate in 95% ethanol at -20oC overnight. The precipitates were collected by centrifugation at 14000g for 15min and resuspended in deionized water. The total GAGs were quantified using Blyscan Glycosaminoglycan Assay (Biocolor Ltd,
United Kingdom). Chondroitinase B (Sigma-Aldrich, Missouri, United States of
America) and chondroitinase AC (Sigma-Aldrich) was used to remove dermatan sulphate and chondroitin-4-sulphate and chondroitin-6-sulphate respectively and total GAG was requantified after enzymatic digestion – dermatan sulphate and chondroitin sulphate levels can be quantified by subtraction of total GAG after digestion from total GAG before digestion.
2.3.8 Proliferation assay
Proliferation assays were performed using the CellTiter 96 AQueous One Solution Cell
Proliferation Assay (Promega Corporation, Madison, Wisconsin, USA) according to
® manufacturer’s protocol. The CellTiter 96 AQueous One Solution Reagent contains a tetrazolium compound [3-(4,5-dimethylthiazol-2-yl)-5-(3-carboxymethoxyphenyl)-2-(4- sulfophenyl)-2H-tetrazolium, inner salt; MTS] which is bioreduced by NADPH or
NADH produced by metabolically active cells via dehydrogenase enzymes into a colored formazan product that is soluble in tissue culture medium. The quantity of formazan product is directly proportional to the number of viable living cells.
Page | 56
Chapter 2 Materials and Methods
2.3.9 Migration assay
Transwell Inserts (Corning Incorporated, New York, United States of America) of
6.5mm diameter, polycarbonate membrane with 8.0μm pore size were used for the assay.
The inserts were first hydrated by adding serum-free medium to the 24-well plate well and then to the Transwell inserts. The plate was then incubated for 2 hours at 37oC with
5% CO2. This initial equilibrium period was used to improve cell attachment.
0.6ml of 20% FBS complete growth medium was added to the lower chamber. 8 x
104 RWPE-1 cells resuspended in 0.2ml serum-free medium were added to the inside of
o the Transwell inserts. Migration of the cells was allowed to occur at 37 C in 5% CO2 conditions for 48 hours.
The medium was carefully removed from the upper and lower chambers of the Transwell
Inserts. The inserts were then washed with 1X PBS and fixed with 0.6ml of 4% paraformaldehyde for 10min. Subsequently, the inserts were rinsed with 1X PBS again and left to air dry. The inserts were then stained with 0.6ml of aqueous crystal violet
(0.5% w/v) for 30 minutes. The excess stain was then washed out by immersion of the inserts in a beaker of clean tap water. A cotton swab was used to clean the insides of the inserts so as to remove the unmigrated / uninvaded cells. The inserts were then air-dried and viewed under a microscope (Nikon SMZ 1500). Five different field views of each insert were taken and the cells were counted using an image analysis software,
AxioVision Release 4.6.
2.3.10 Invasion assay
BD BioCoat™ Matrigel™ Invasion Chambers (BD Biosciences, California, United
States of America) were used for the conduct of the invasion assay. The Falcon Cell
Page | 57
Chapter 2 Materials and Methods
Culture Inserts contain an 8.0μm pore size polyethylene terephthalate (PET) membrane with a thin layer of Matrigel Basement Membrane Matrix. The Matrigel Matrix serves as a reconstituted basement membrane in vitro. The layer occludes the pores of the membrane, blocking non-invasive cells from migrating through the membrane. In contrast, invasive cells are able to detach themselves from and invade through the
Matrigel Matrix and the 8.0μm membrane pores.
The Cell Culture inserts were first hydrated overnight at 4oC by adding serum-free medium to the 24-well plate well and then to the cell culture inserts. The plate was then
o incubated for 2 hours at 37 C with 5% CO2 prior to the start of the invasion proper. This initial equilibrium period was used to improve cell attachment.
A total of 0.5ml of 20% FBS complete growth medium was added to the lower chamber. 1 x 105 PC-3 cells resuspended in 0.5ml of serum-free medium were added to the inside of the Falcon cell culture inserts. Invasion of the cells was allowed to occur at
o 37 C in 5% CO2 conditions for 48 hours. Staining and counting of the invaded cells were performed as in Section 2.3.11, Migration Assay.
2.3.11 Adhesion assay
2X104 transfected PC-3 or RWPE-1 cells were seeded into pre-collagen type 1- coated ((BD Biosciences) 96 well plates. The cells were then incubated at humified 37oC,
5% CO2 level incubator for 30 mins. Unadhered cells were then removed and washed thrice with 1X PBS. Unwashed wells of cells were used as loading controls. Thereafter,
CellTiter 96 AQueous One Solution Cell Proliferation Assay (Promega Corporation) was added to quantify the number of cells that adhered to the collagen type I substrate.
Page | 58
Chapter 2 Materials and Methods
2.3.12 Cell cycle assay
Cell cycle analysis was performed on ethanol-fixed PC-3 cells and stained using
20ug/ml propidium iodide (PI) (Sigma-Aldrich) with 0.01% Triton X (Bio-Rad
Laboratories) and 0.2mg/ml DNase-free RNase A (Roche Diagnostics Corporation,
Roche Applied Science) for 30 mins. Thereafter, the PI-stained cells were put through the
Dako Cytomation Cyan LX (Dako, Denmark) and fluorescence data was collected and analysed using Dako Summit version 4.3 software (Dako).
2.3.13 Cell apoptosis assay
Apoptosis assay was conducted using the Guava Nexin Reagent (Millipore Corporation,
Massachusetts, United States of America) on the EasyCyte Mini Flow Cytometer (Guava
Technologies, Inc, California, United States of America), according to the manufacturer’s protocol. The Guava Nexin Assay contains 2 dyes, Annexin V-PE and 7-AAD. Annexin
V binds with high affinity to phosphatidylserine which externalizes to the outer cellular surface in the early stages of apoptosis [Vermes et al, 1995]. 7-AAD is a cell impermeant dye and is used as an indicator of cell membrane structural integrity, which in late stages of apoptosis, there is loss of membrane integrity [Schmidt et al, 1992].
2.3.14 Flow cytometry
Transfected PC-3 cells were fixed in cold methanol (for MAPRE-1 staining) or 95% ethanol with 5% acetic acid (for TPX-2 and HIPK3 staining) at -20oC for 10min. The cells were then washed in 1X PBS and permeabilized in 0.5% Tween-20 for 10min. Cell pellet was collected and the primary antibody, diluted in 3% BSA/PBS was added and incubated at room temperature for 1 hour. MAPRE-1 antibody was diluted 1:400, TPX2 antibody was diluted 1:200 and HIPK3 antibody was diluted 1:100. The cells were
Page | 59
Chapter 2 Materials and Methods washed thrice in 1X PBS and cell pellet was again collected by centrifugation at 1000rpm for 5min. Respective secondary antibodies were added and incubated at room temperature for 1 hour. The cells were washed thrice in 1X PBS and cell pellet was again collected by centrifugation at 1000rpm for 5min. The cells were then resuspended in 1X
PBS and 5000 stained cells were used for fluorescence analysis, performed on the
EasyCyte Mini Flow Cytometer (Guava Technologies, Inc) using the Guava Express Pro
Software (Guava Technologies, Inc).
2.3.15 Microarray analysis
RNA was isolated from the transfected PC-3 cells using the RNeasy Micro Kit (Qiagen) and the RNA quality was checked using the Agilent RNA 6000 Nano Kit (Agilent
Technologies Inc) on the Agilent 2100 Bioanalyzer System. RNA labeling and hybridization on the Affymetrix Genechip U133 Plus 2 chips (Affymetrix Inc, California,
United States of America) were performed, as in Section 2.1.4, except that the one-cycle target labeling protocol was used to synthesize biotin-labeled cRNA from 2.5ug of total
RNA starting material (from the non-targeting or CSGalNAcT-1 silenced PC-3 cells).
Microarray data was analyzed as shown in Table 2.8. Genes that were found to be common in all gene lists, as determined by the criteria used in individual software, were merged.
Page | 60
Chapter 2 Materials and Methods
Table 2.8 Methods of normalization and criteria for determining genes differentially expressed using different microarray analysis software, for microarray work involving CSGalNAcT-1 silenced PC-3 cells.
Software Normalization and Expression values Criteria GCOS 1. One array was chosen as baseline array, to which 1. Change calls (i.e normalized expression of genes that other arrays are normalized against. Thus, for 6 are differentially expressed as compared to the control BPH samples and 6 prostate baseline array – increase or decrease) are the same in adenocarcinoma samples, I have a total of 36 at least 50% of the 36 different comparisons. different comparisons. 2. Fold change ≥ 2 3. P-value < 0.05 dCHIP 1. Array with the median overall intensity was used as 1. Detection ‘P’ call present in at least 50% of either the baseline array against which other arrays are control BPH group or prostate adenocarcinoma normalized (Li and Wong 2001a). group. 2. PM/MM difference model-based expression values 2. Fold change ≥ 2 were calculated. (Li and Wong 2001a). 3. P-value < 0.05
Genespring 1. Five different algorithms were used in Genespring 1. Fold change ≥ 2 to generate 5 different lists – GC-RMA, RMA, 2. P-value < 0.05 dCHIP, GCOS and PLIER
Page | 61
Chapter 3 Expression profiling of prostate adenocarcinoma tissues
Chapter 3
Expression Profiling of Prostate Adenocarcinoma Tissues
The expression of most of the genes in the human genome can be analyzed
simultaneously through the application of expression microarray technology. This has
enabled the discovery of molecular biomarkers that can serve as diagnostic and prognostic markers of prostate cancer, and also permitted the identification of novel
therapeutic targets. In this study, genome wide expression profiling of prostate cancer
tissues and benign hyperplastic prostate tissues were performed so as to identify
potential biomarkers and therapeutic targets of prostate cancer.
3.1 Selection of tissues to be used for study
Six benign prostatic hyperplasia tissue sections and six prostate
adenocarcinoma tissue sections were histologically checked using haematoxylin-eosin
stain (Figure 3.1 and Figure 3.2) and subsequently used for the microarray study. The
tissue sections used for the study were all obtained from Chinese patients.
Clinicopathological data of the samples used are listed in Table 3.1. The mean age of
diagnosis for patients with benign prostatic hyperplasia is 68.3 while for patients with
prostate adenocarcinoma is 77.5 years. All the samples used belong to patients of
Chinese ethnicity. Three of the six cancer patients have bone metastasis.
3.2 RNA quality of tissue samples used for Affymetrix GeneChip microarrays
Total RNA was extracted from all the twelve prostate tissues to be used for the
microarray study. The mean A260/A280 and mean RIN numbers of BPH and
adenocarcinoma samples are shown in Table 3.2 and the electrophoregrams are shown
in Figure 3.3. As seen, all samples used for the microarray study were of good purity
Page | 62
Chapter 3 Expression profiling of prostate adenocarcinoma tissues
with mean A260/A280 of 2.04 (BPH samples) and of 2.01 (Cancer samples). The mean
RIN for BPH samples is 7.7 and is 7.5 for cancer samples.
Labelled cRNA was fragmented prior to hybridization to Affymetrix U133
plus2 Genechips. cRNA is usually fragmented to sizes ranging from 50 to 200 bases
to improve target specificity [Affymetrix, Inc, 2007]. The fragmented cRNA was checked using the Agilent Bioanalyzer and the electrophoregrams are shown in Figure
3.4. The electrophoregrams show that all the cRNA was successfully fragmented to sizes ranging from 50 to 200 bases.
Page | 63
Chapter 3 Expression profiling of prostate adenocarcinoma tissues
Figure 3.1 Haematoxylin and eosin staining of BPH prostate samples used for microarray study. Scale bar represents 100µm.
Page | 64
Chapter 3 Expression profiling of prostate adenocarcinoma tissues
Figure 3.2 Haematoxylin and eosin staining of prostate adenocarcinoma samples used for microarray study. Scale bar represents 100µm.
Page | 65
Chapter 3 Expression profiling of prostate adenocarcinoma tissues
Table 3.1 Clinical features of prostate adenocarcinoma cases used for expression profiling.
Sample Age Pre-op Metastasis Any pre- Pathological Clinical Operation Post-op PSA PSA Level op Tumour grade TMN PSA Relapse (ng/ml) therapy? staging (ng/ml) Benign Sample 1 80 34.3 N/A N/A Benign prostatic N/A TURP N/A N/A hyperplasia Benign Sample 2 / 61 16 No No Gleason 3+4 T2cN0M0 Open RRP <0.01 No Cancer Sample 2a Benign Sample 3 74 5.7 N/A N/A Benign prostatic N/A TURP N/A N/A hyperplasia Benign Sample 4 59 N/A N/A N/A Benign prostatic N/A TURP N/A N/A hyperplasia Benign Sample 5b 68 25.1 No No Gleason 3+4 T1cN0M0 Open RRP <0.1 No Benign Sample 6 68 N/A N/A N/A Benign prostatic N/A TURP N/A N/A hyperplasia Cancer Sample 1 69 235.8 Yes (bone) No Gleason 5+5 T1cNxM1 TURP N/A N/A Cancer Sample 3 80 587.8 Yes (bone) No Gleason 4+3 T1cNxM1 TURP 10.41 Yes Cancer Sample 4c 90 16.2 N/A N/A Benign prostatic N/A TURP N/A N/A hyperplasia Cancer Sample 5 83 22.8 No No Gleason 4+3 T1bN0M0 TURP <0.1 N/A Cancer Sample 6 82 127.6 Yes (bone) No Gleason 5+5 T1bNxM1 TURP 7.9 Yes a Benign sample 2 and Cancer sample 2 were 2 different prostate tissue types taken from the same patient, and the 2 different tissue types were used as separate samples. b Patient had prostate adenocarcinoma but no prostate adenocarcinoma tissue were seen in the prostatic chips collected from the patient. Only BPH sections were obtained. cCancer tissues were found in the prostatic chips that were collected, but were not sampled when the chips were collected for pathologic diagnosis. See Figure 3.2.
Page | 66
Chapter 3 Expression profiling of prostate adenocarcinoma tissues
Figure 3.3 Electrophoregrams of RNA samples extracted from six BPH sections and six cancer sections, as determined by Agilent Bioanalyzer. Good quality RNA samples are typified by intact 18S and 28S ribosomal subunit peaks, flat baseline throughout electrophoregrams and no well-defined peaks between 18S and 28S peaks. Y-axis represents ‘FU’, which stands for fluorescence units while x-axis represents time in seconds.
Page | 67
Chapter 3 Expression profiling of prostate adenocarcinoma tissues
Figure 3.4 Electrophoregrams of fragmented cRNA samples. A distinct peak should be observed from approximately 20 to 30 seconds on the x-axis. Y-axis represents ‘FU’, which stands for fluorescence units while x-axis represents time in seconds.
Page | 68
Chapter 3 Expression profiling of prostate adenocarcinoma tissues
Table 3.2 RNA purity and quality of samples used for microarray study
Samples A260/A280 RIN Mean (± SD) Mean (± SD) BPH (n=6) 2.04 ± 0.02 7.7 ± 0.3 Adenocarcinoma (n=6) 2.01 ± 0.07 7.5 ± 0.8
3.3 Quality control of microarray data
Staining, washing and scanning of the microarray chips were performed on
Fluidics Station 450 and the Affymetrix GeneChip Scanner 3000. The presence of housekeeping genes, spike control cRNAs, background scaling factor and noise (raw
Q) were checked to ensure the quality of the assays done.
On the Genechips, a number of housekeeping genes are controlled for by using a set of 3 probes directed against the 5’, middle, and 3’ region of the gene transcript. 3'/5' ratio, which is the signal intensity ratio of the 3' probe set over the 5' probe set, can be used to indicate the efficacy of the whole cRNA labelling process, from integrity of the starting RNA, efficiency of first strand cDNA synthesis to in vitro transcription of cRNA. In an ideal situation with total intact RNA, efficient oligo-dT based first strand cDNA synthesis and efficient hybridization, the 3'/5' ratio will be 1:1. However, it has been noted that a two-cycle cRNA synthesis and labelling process could be biased towards the 3’ region, resulting in higher 3'/5' ratio
[Affymetrix, Inc, 2007]. It may be more appropriate to detect any sample whose 3'/5' ratios deviates from the mean of the group within a particular study, which could highlight an unusual sample that deserves further investigation. The 3’/5’ ratios for
GAPDH gene lies from 2.59 to 3.92 for all samples.
Page | 69
Chapter 3 Expression profiling of prostate adenocarcinoma tissues
The scaling factor provides a measure of the ‘brightness’ of the array.
‘Brightness’, a measure of the image intensity, varies from array to array due to
experimental variation. Hence, it is necessary to normalize intensities of all arrays to
same level to prevent confounding by non-biological errors. The intensity values of
all probe sets on the array, except the top and bottom 2% of the probe set intensities,
were averaged in GCOS, giving rise an overall intensity. This overall intensity is then
multiplied by a scaling factor to bring it to an arbitrary target intensity value i.e 500.
In a particular set of experiments, the scaling factor value for all the arrays should be
very close to each other (within two-fold of each other) [Affymetrix, Inc, 2007]. The
scaling factors for all the samples are within two-fold of each other, the smallest
factor being 6.4 while the largest factor is 11.0, with a mean factor of 8.0.
Controls are needed to detect hybridization efficiency and Bio-B, C, D and
CRE serve as such controls and are spiked at the following concentrations: BioB: 1.5 pM, BioC: 5.0 pM, BioD: 25.0 pM, CRE: 100 pM. All controls are expected to be
called present at all times, with increasing signal values (BioC, BioD, and CRE,
respectively), with the exception of BioB, which at the lowest concentration, is
anticipated to be called present at least 70% of the time [Affymetrix, Inc, 2007]. All
the spiked in controls were detected as present in all arrays; in addition, there was
increasing average intensity values from BioB to CRE for all samples.
Small variations in the digitized signal can happen due to sampling of probe
array surface by the scanner, leading to noise (Q value). Examining pixel-to-pixel
variations in background intensities allows us to measure noise. The noise value for
all the arrays in one experiment should be very similar to each other [Affymetrix, Inc,
2007], and the Q values for all our samples are similar to one another, with a range of
1.1 to 1.7 only.
Page | 70
Chapter 3 Expression profiling of prostate adenocarcinoma tissues
3.4 Differential expression of genes in prostate adenocarcinoma
Several softwares such as GCOS, dCHIP, RMA and Genespring were used to
analyze the microarray data. The data was normalized and expression intensities of
different genes were derived using each of the above software. Genes that were commonly identified by the four softwares as differentially expressed in the prostate cancer sections using the criteria set in Table 2.1 were identified and merged. From this selection criterion, 192 genes were found to be upregulated and 673 genes to be downregulated.
Using Database for Annotation, Visualization, and Integrated Discovery
(DAVID) 2008, the 865 genes that were differentially expressed were functionally annotated and categorised, and the Gene Ontology (GO) biological process enriched terms associated with the genes are represented in Table 3.3. Gene enrichment
analysis, i.e the association of the input genes with GO terms, is statistically measured
by Fisher’s Exact Statistics in DAVID, as represented by an EASE score. Due to
repeated representation of similar annotations, DAVID has also developed a new tool
termed Functional Annotation Clustering, which groups similar annotations together.
This allows the reader to have a clearer idea of the functional categorization of the
gene list. In addition, the Group Enrichment Score, obtained after Functional
Annotation Clustering, is a measure of the particular annotation cluster’s biological
significance. Thus, the bigger the group enrichment score, the more statistically
significant is the genes in the group associated with GO terms. [Huang et al, 2009;
Dennis et al, 2003].
As seen in Table 3.3, the top-ranked annotation cluster contains genes relating
to development process and cell differentiation. A total of 243 genes are represented
in this cohort. Other functional annotation clusters include genes involved in cell
Page | 71
Chapter 3 Expression profiling of prostate adenocarcinoma tissues migration (41 genes), proliferation (69 genes), carbohydrate binding (26 genes), cellular response to stress (126 genes), homeostasis (59 genes), cell communication
(212 genes), metabolism (87 genes), of which 46 genes are involved in lipid metabolism, cell cycle (37 genes) and cell death (73 genes). Due to functional overlaps of some genes, one gene may be represented more than once in several annotation groups.
Table 3.4 shows a list of some genes that are aberrantly expressed in the prostate cancer tissues, grouped according to their functional annotations. This list is not exhaustive but serves to highlight certain genes that are frequently reported to be involved in tumourigenesis, and also genes that code enzymes involved in GAG synthesis or those that code for proteoglycans. For instance, it was noted that many genes involved in Wnt signalling, Hedgehog signalling, TGFβ signalling and homeobox genes are differently expressed in prostate cancer; and all of these signalling pathway molecules are known to be involved in development and differentiation. Genes that code for GAG synthesis enzymes such as CSGalNAcT-1 and CHST2, and genes coding for protein cores of proteoglycans such as DCN, VCAN and TGFBR3 are also differentially expressed and known to be involved in development and differentiation.
It has been reported that chondroitin sulphate and heparan sulphate can bind other proteins to elicit a concerted effect on cellular signalling [Deepa et al, 2004].
Interestingly, transcripts coding for proteins known to be able to bind GAGs have been identified to be aberrantly expressed in prostate cancer (Table 3.4). It can also be observed that many of the GAG synthetic and editing genes are dysregulated in prostate adenocarcinoma tissues, such as B3GalT6, CSGalNAcT-1, GALNAC4S-6ST,
Page | 72
Chapter 3 Expression profiling of prostate adenocarcinoma tissues
HS6ST2, HS3ST3A1, HS3ST3B1 and SULF2. In addition, CSPG, DSPG and HSPG such as syndecan-1 (SDC1), decorin (DCN), versican (VCAN) and transforming growth factor beta receptor 3 (TGFβR3) are found to be differentially expressed in prostate adenocarcinoma cases as compared to benign hyperplastic tissues.
There are also many genes that are drastically upregulated or downregulated in prostate adenocarcinoma tissues but are not present in the gene enrichment analysis table in Table 3.4. However, these genes may potentially be used as biomarkers or could be important functionally as well. For such reasons, I have included in the whole list of the 865 genes found to be more than 2 fold differentially expressed in prostate adenocarcinoma tissue, without any functional classification but in terms of its fold change value, in Appendix I.
Page | 73
Chapter 3 Expression profiling of prostate adenocarcinoma tissues
Table 3.3 Gene Ontology (GO) annotation terms and clusters that the 865 filtered genes are associated with.
Gene Ontology Annotations Number of genes Percentage of genes (n=865) EASE score (p-value) Annotation Cluster – Development and differentiation Total genes: 243 Group Enrichment Score: 15.4 anatomical structure development 184 21.27% 3.79E-24 developmental process 241 27.86% 1.22E-23 multicellular organismal development 192 22.20% 5.23E-23 system development 144 16.65% 1.97E-16 multicellular organismal process 243 28.09% 8.71E-16 organ development 106 12.25% 2.40E-12 cell differentiation 117 13.53% 7.73E-07 cellular developmental process 117 13.53% 7.73E-07
Annotation Cluster – Cell motility Total genes: 41 Group Enrichment Score: 5.03 localization of cell 40 4.62% 1.61E-06 cell motility 40 4.62% 1.61E-06 cell migration 25 2.89% 3.17E-04
Annotation Cluster – Cell proliferation Total genes: 69 Group Enrichment Score: 4.01 cell proliferation 57 6.59% 4.79E-05 negative regulation of cell proliferation 25 2.89% 5.57E-05 regulation of cell proliferation 38 4.39% 3.46E-04
Annotation Cluster - Carbohydrate binding Total genes: 26 Group Enrichment Score : 3.82 polysaccharide binding 16 1.85% 2.49E-05 glycosaminoglycan binding 15 1.73% 6.58E-05
Page | 74
Chapter 3 Expression profiling of prostate adenocarcinoma tissues pattern binding 16 1.85% 7.78E-05 heparin binding 12 1.39% 4.12E-04 carbohydrate binding 26 3.01% 0.001497
Annotation Cluster – Cellular response to stress Total genes: 126 Group Enrichment Score : 2.94 inflammatory response 35 4.05% 7.82E-08 response to wounding 40 4.62% 2.84E-06 response to external stimulus 49 5.66% 4.15E-05 defence response 43 4.97% 2.09E-04 response to stress 62 7.17% 0.006162 immune response 39 4.51% 0.544508 immune system process 47 5.43% 0.617306 response to stimulus 121 13.99% 0.71586
Annotation Cluster - Homeostasis Total genes: 59 Group Enrichment Score: 2.71 cellular cation homeostasis 20 2.31% 1.55E-04 cation homeostasis 20 2.31% 1.67E-04 cellular di-, tri-valent inorganic cation homeostasis 18 2.08% 2.94E-04 di-, tri-valent inorganic cation homeostasis 18 2.08% 3.16E-04 regulation of biological quality 58 6.71% 3.72E-04 cellular chemical homeostasis 20 2.31% 0.001254 cellular ion homeostasis 20 2.31% 0.001254 ion homeostasis 20 2.31% 0.003802 chemical homeostasis 22 2.54% 0.004482 cellular homeostasis 24 2.77% 0.00484 metal ion homeostasis 13 1.50% 0.00692 cellular metal ion homeostasis 13 1.50% 0.00692
Page | 75
Chapter 3 Expression profiling of prostate adenocarcinoma tissues calcium ion homeostasis 12 1.39% 0.010591 cellular calcium ion homeostasis 12 1.39% 0.010591 homeostatic process 26 3.01% 0.036701
Annotation Cluster – Cell communication Total genes: 212 Group Enrichment Score: 2.68 cell communication 212 24.51% 7.25E-05 signal transduction 185 21.39% 0.002471 cell surface receptor linked signal transduction 91 10.52% 0.051529
Annotation Cluster – Lipid metabolism Total genes: 46 Group Enrichment Score: 2.12 fatty acid metabolic process 17 1.97% 0.00219 lipid metabolic process 46 5.32% 0.008691 cellular lipid metabolic process 37 4.28% 0.023393
Annotation Cluster - Metabolism Total genes: 52 Group Enrichment Score: 1.83 monocarboxylic acid metabolic process 22 2.54% 0.001061 fatty acid metabolic process 17 1.97% 0.00219 amine metabolic process 31 3.58% 0.007742 carboxylic acid metabolic process 36 4.16% 0.011448 organic acid metabolic process 36 4.16% 0.012118 nitrogen compound metabolic process 32 3.70% 0.012476 amino acid and derivative metabolic process 20 2.31% 0.199456 amino acid metabolic process 15 1.73% 0.369168
Annotation Cluster – Amine metabolism Total genes: 6 Group Enrichment Score: 1.78 amino sugar metabolic process 6 0.69% 0.010263 N-acetylglucosamine metabolic process 5 0.58% 0.020122
Page | 76
Chapter 3 Expression profiling of prostate adenocarcinoma tissues glucosamine metabolic process 5 0.58% 0.022879
Annotation Cluster - Cell cycle Total genes: 37 Group Enrichment Score: 1.05 negative regulation of progression through cell cycle 18 2.08% 0.005422 regulation of progression through cell cycle 30 3.47% 0.063881 regulation of cell cycle 30 3.47% 0.06594 cell cycle process 33 3.82% 0.426408 cell cycle 37 4.28% 0.559598
Annotation Cluster – Cell death Total genes: 73 Group Enrichment Score: 0.88 cell development 73 8.44% 0.00157 regulation of apoptosis 30 3.47% 0.078811 regulation of programmed cell death 30 3.47% 0.086461 death 42 4.86% 0.123836 cell death 42 4.86% 0.123836 anti-apoptosis 11 1.27% 0.153858 programmed cell death 39 4.51% 0.169792 apoptosis 38 4.39% 0.200731 negative regulation of apoptosis 13 1.50% 0.234682 positive regulation of apoptosis 14 1.62% 0.24354 negative regulation of programmed cell death 13 1.50% 0.244507 positive regulation of programmed cell death 14 1.62% 0.248142 induction of apoptosis 11 1.27% 0.371252 induction of programmed cell death 11 1.27% 0.376754
Page | 77
Chapter 3 Expression profiling of prostate adenocarcinoma tissues
Table 3.4 A highlight of some genes that are differentially expressed in prostate adenocarcinoma tissues (n=6) as compared to BPH tissues (n=6), grouped by their functions. Gene Symbol Gene Name Gene ID Relative fold changea Developmental and Differentiation Wnt signalling pathway DIXDC1 Dix domain containing 1 85458 -2.1 DKK3 Dickkopf Homolog 3 (Xenopus Laevis) 27122 -2.3 FZD1 Frizzled Homolog 1 (Drosophila) 8321 -2.2 FZD10 Frizzled Homolog 10 (Drosophila) 11211 -9.5 FZD2 Frizzled Homolog 2 (Drosophila) 2535 -2.4 FZD7 Frizzled Homolog 7 (Drosophila) 8324 -2.4 PPAP2B Phosphatidic acid phosphatase Type 2B 8613 -2.9 ROR2 Receptor tyrosine kinase-like orphan receptor 2 4920 -5.0 RSPO3 R-spondin 3 homolog (Xenopus laevis) 84870 -3.9 SFRP1 Secreted Frizzled-Related Protein 1 6422 -3.8 TCF7L1 Transcription factor 7-Like 1 (T-cell specific, HMG-Box) 83439 -2.7 TGFB1/1 Transforming growth factor beta 1 induced transcript 1 7041 -3.3 TLE2 Transducin-like enhancer of split 2 (E(Sp1) Homolog, Drosophila) 7089 -2.4 WIF1 WNT inhibitory factor 1 11197 -36.7 WNT2 Wingless-type MMTV integration site family member 2 7472 -2.8 WNT4 Wingless-type MMTV integration site family member 4 54361 -3.1 WNT5B Wingless-type MMTV integration site family member 5B 81029 -2.0
Hedgehog signalling pathway BMP4 Bone morphogenetic protein 4 652 -2.9 BMP5 Bone morphogenetic protein 5 653 -2.7 GAS1 Growth arrest specific 1 2619 -2.5
Page | 78
Chapter 3 Expression profiling of prostate adenocarcinoma tissues
HHIP Hedgehog interacting protein 64399 -2.9 WNT2 Wingless-type MMTV integration site family member 2 7472 -2.8 WNT4 Wingless-type MMTV integration site family member 4 54361 -3.1 WNT5B Wingless-type MMTV integration site family member 5B 81029 -2.0 ZIC2 Zic family member 2 (Odd-paired Homolog, Drosophila) 7546 21.7
TGFβ signalling pathway BMP4 Bone morphogenetic protein 4 652 -2.9 BMP5 Bone morphogenetic protein 5 653 -2.7 COMP Cartilage oligomeric matrix protein 1311 4.4 DCN Decorin 1634 -2.0 FST Follistatin 10468 -3.9 TGFB3 Transforming growth factor, beta 3 7043 -5.1
Homeobox genes HOXC6 Homeobox C6 3223 12.3 HOXC4 Homeobox C4 3221 4.8 HOXB3 Homeobox B3 3213 -2.5 HOXB2 Homeobox B2 3212 -2.6 HOXD10 Homeobox D10 3236 -3.2 DLX1 Distal-less Homeobox 1 1745 7.8 NKX2-3 NK2 tanscription factor homolog C (Drosophila) 159296 6.6
GAG enzyme synthesis genes/proteoglycans CSGALNACT-1 Chondroitin beta 1,4 N-acetylgalactosaminyltransferase 55790 2.0 DCN Decorin 1634 -2.0 VCAN Chondroitin sulphate proteoglycan 2 (versican) 1462 -2.7
Page | 79
Chapter 3 Expression profiling of prostate adenocarcinoma tissues
TGFBR3 Transforming growth factor , beta receptor III (Betaglycan, 300kDa) 7049 -2.8 CHST2 Carbohydrate (N-Acetylglucosamine-6-O) Sulphotransferase 2 9435 -4.4
Cell proliferation MYC V-MYC Myelocytomatosis viral oncogene homolog (Avian) 4609 4.1 CSGALNACT-1 Chondroitin beta 1,4 N-acetylgalactosaminyltransferase 55790 2.0 FGFR1 Fibroblast growth factor receptor 1 2260 -2.0 PDGFC Spinal cord derived growth factor; Secretory growth factor-like protein 56034 -2.3 fallotein PTN Pleiotrophin 5764 -3.3 PGF Placental growth factor 5228 -3.6 FGFR2 Fibroblast growth factor receptor 2 2263 -4.7 TGFB3 Transforming growth factor, beta 3 7043 -5.1
Glycosaminoglycan binding proteins ABP1 Amiloride binding protein 1 (amine oxidase (Copper containing) 26 -27.8 ANG Angiogenin, ribonuclease, RNase A family, 5 283 -2.2 APOE Apolipoprotein E 348 2.1 BMP4 Bone morphogenetic protein 4 652 -2.9 COL13A1 Collagen, type XIII, alpha 1 1305 -4.6 CXCL6 Chemokine (C-X-C Motif) Ligand 6 (Granulocyte chemotactic protein 6372 -6.7 2) FGFR1 Fibroblast growth factor receptor 1 2260 -2.0 FGFR2 Fibroblast growth factor receptor 2 2263 -4.7 HMMR Hyaluronan-mediated motility receptor (RHAMM) 3161 2.4 PGF Placental growth factor 5228 -3.6 PTN Pleiotrophin 5764 -3.3
Page | 80
Chapter 3 Expression profiling of prostate adenocarcinoma tissues
RSPO3 R-spondin 3 homolog (Xenopus laevis) 84870 -3.9 SERPINE2 Serpin peptidase inhibitor, Clade E (Nexin, Plasminogen activator 5270 -2.1 inhibitor type I), member 2 TGFBR3 Transforming growth factor , beta receptor III (Betaglycan, 300kDa) 7049 -2.8 VCAN Chondroitin sulphate proteoglycan 2 (versican) 1462 -2.7
Metabolism APOE Apolipoprotein E 348 2.1 B3GALT6 UDP-Gal:BetaGal Beta 1,3-Galactosyltransferase Polypeptide 6 126792 2.2 CHST2 Carbohydrate (N-Acetylglucosamine-6-O) sulphotransferase 2 9435 -4.4 CHST9 Carbohydrate (N-Acetylgalactosamine-4-O) sulphotransferase 9 83539 -11.3 CSGALNACT-1 Chondroitin beta 1,4 N-acetylgalactosaminyltransferase 55790 2.0 FABP5 Fatty acid binding protein 5 (Psoriasis-associated) 2171 4.4 FASN Fatty acid synthase 2194 4.5 FBP1 Fructose-1,6-bisphosphatase 1 2203 4.1 GALNAC4S-6ST BRAG 51363 -4.0 HK2 Hexokinase 2 3099 3.8 HS3ST3B1 Heparan sulphate-3-O-sulphotransferase 3B1 5593 -2.2 SULF2 Sulphatase 2 55959 -2.7
GAG synthesis / Proteoglycan B3GALT6 UDP-Gal:BetaGal Beta 1,3-Galactosyltransferase Polypeptide 6 126792 2.2 CHST2 Carbohydrate (N-Acetylglucosamine-6-O) sulphotransferase 2 9435 -4.4 CHST9 Carbohydrate (N-Acetylgalactosamine-4-O) sulphotransferase 9 83539 -11.3 CSGALNACT-1 Chondroitin beta 1,4 N-acetylgalactosaminyltransferase 55790 2.0 DCN Decorin 1634 -2.0 GALNAC4S-6ST BRAG 51363 -4.0
Page | 81
Chapter 3 Expression profiling of prostate adenocarcinoma tissues
HS3ST3A1 Heparan sulphate-3-O-sulphotransferase 3A1 9955 -2.5 HS3ST3B1 Heparan sulphate-3-O-sulphotransferase 3B1 5593 -2.2 HS6ST2 Heparan sulphate 6-O-sulphotransferase 2 90161 -3.6 SDC1 Syndecan 1 6382 -2.8 SULF2 Sulphatase 2 55959 -2.7 TGFBR3 Transforming growth factor , beta receptor III (Betaglycan, 300kDa) 7049 -2.8 VCAN Chondroitin sulphate proteoglycan 2 (versican) 1462 -2.7 aRelative fold changes listed are those determined by dCHIP.
Page | 82
Chapter 3 Expression profiling of prostate adenocarcinoma tissues
3.5 Hierarchical clustering of microarray data
Unsupervised sample clustering using the 865 filtered genes was used to identify
novel sample clusters and their associated gene signature. This allows checking of the quality
of the microarray data to determine if similar samples cluster together and to identify
unexpected clustering (e.g of samples processed in the same batch cluster together.) A
dendrogram shown in Figure 3.5 illustrates the overall gene expression profile across the
samples (genes-horizontal, samples-vertical). This analysis revealed a major division between
the benign prostate sections and the cancerous tissue sections.
3.6 Principal component analysis (PCA) of microarray expression data
Principal component analysis (PCA) is a useful method to determine sample classes
based on the gene expression profiles. PCA maps samples in high N dimensions (N is the
number of genes) to two dimensions, identifying clusters of samples that are more similar to
one another in terms of gene expression. Using PCA, it can be clearly seen that the cancer
samples are separated distinctly from the BPH samples (Figure 3.6). The cancer samples are
observed to be further apart from one another, demonstrating a more heterogeneous genetic
makeup.
Page | 83
Chapter 3 Expression profiling of prostate adenocarcinoma tissues
Figure 3.5 Hierarchical clustering of differentially expressed genes (n=865) in prostate cancer tissues (n=6) compared to BPH (n=6) tissues. Rows: individual genes; columns: individual samples. Each cell in the matrix represents expression level of a single gene in a single sample. Red: transcripts above the median for that gene across all samples. Green: transcripts below the median for that gene across all samples. The dendrogram on the left represents genes grouped by similarity. Inset: BPH samples clustered separately from prostate cancer samples.
Page | 84
Chapter 3 Expression profiling of prostate adenocarcinoma tissues
Figure 3.6 Principal component analysis (PCA) of the six BPH samples and six prostate adenocarcinoma samples used in the study. PCA was generated in two- dimension and presented here, showing difference and similarities between samples. The BPH samples and cancer samples were distinctly separated from each other.
Page | 85
Chapter 3 Expression profiling of prostate adenocarcinoma tissues
3.7 Quantitative real-time polymerase chain reaction (qPCR) verification of microarray results
For the purpose of verification of the microarray data, quantitative real-time
polymerase chain reaction was performed on the human prostate tissue samples used
for the microarray study. Fifty-three random genes were quantified, in which all 53
genes (100%) were found to be differentially changed in the same direction as
determined from the microarray study (i.e both upregulated or downregulated through
microarray study and real-time PCR) (Table 3.5). From the microarray, 865 genes
from our gene list are genes determined to be differentially expressed by at least 2
fold change with a significant p-value <0.05. However, from real-time PCR, 5 of the
53 genes (about 9%) did not demonstrate a statistically significant p-value (p>0.05) albeit revealing a trend in the same differential direction as the microarray result.
Nevertheless, these data represent that approximately 91% of the genes from the microarray data are true positives.
Page | 86
Chapter 3 Expression profiling of prostate adenocarcinoma tissues
Table 3.5 Verification of several genes from microarray list, as determined by real-time PCR
Gene Gene name Gene Microarraya - Real-time PCR ID Relative Fold Changed Relative Fold p-value Changed ALDH1A2 aldehyde dehydrogenase 1 family, member A2 8854 -5.22 -4.65 0.0003 AMACR alpha-methylacyl-CoA racemase 23600 28.94 59.58 <0.0001 B3GALT6c UDP-Gal:betaGal beta 1,3-galactosyltransferase 126792 2.32 2.47 0.0158 polypeptide 6 CHST2 c carbohydrate (N-acetylglucosamine-6-O) 9435 -4.38 1.31 0.496 sulphotransferase 2 CSGalNAcT-1 c chondroitin beta1,4 N-acetylgalactosaminyltransferase 55790 2.75 4.27 0.0161 COL4A6 collagen, type IV, alpha 6 1288 -6.38 -4.51 0.0016 CXCL13 chemokine (C-X-C motif) ligand 13 (B-cell 10563 -27.09 -33.01 0.0002 chemoattractant) CYP3A5 cytochrome P450, family 3, subfamily A, polypeptide 5 1577 -29.34 N.Db <0.05 CYP4B1 cytochrome P450, family 4, subfamily B, polypeptide 1 1580 -24.84 -37.57 <0.0001 DCN c decorin 1634 -2.03 -4.55 0.07 EFEMP2 EGF-containing fibulin-like extracellular matrix protein 30008 -2.90 -1.96 0.0113 2 FASN fatty acid synthase 2194 4.30 7.87 0.0002 FBP1 fructose-1,6-bisphosphatase 1 2203 4.27 13.25 <0.0001 FLRT3 fibronectin leucine rich transmembrane protein 3 23767 -6.06 -4.85 0.0003 GALNT7 UDP-N-acetyl-alpha-D-galactosamine:polypeptide N- 51809 2.74 10.82 0.0006 acetylgalactosaminyltransferase 7 (GalNAc-T7) GATA3 GATA binding protein 3 2625 -9.67 -8.11 0.0003
Page | 87
Chapter 3 Expression profiling of prostate adenocarcinoma tissues
GJA1 gap junction protein, alpha 1, 43kDa (connexin 43) 2697 -3.23 -3.90 0.0007 GLYATL1 glycine-N-acyltransferase-like 1 92292 7.62 19.23 0.0002 HOXC6 homeobox C6 3223 12.08 30.03 0.0008 HPN hepsin (transmembrane protease, serine 1) 3249 6.19 6.39 0.22 HS3ST3A1c heparan sulphate (glucosamine) 3-O-sulphotransferase 9955 -2.52 -2.32 0.035 3A1 HS6ST2 c heparan sulphate 6-O-sulphotransferase 2 90161 -3.54 -3.59 0.0293 KRT15 keratin 15 3866 -43.86 0.01 0.0012 LAMB3 laminin, beta 3 3914 -11.79 -28.64 0.0022 LTF lactotransferrin 4057 -188.70 -70.36 <0.0001 MAL2 mal, T-cell differentiation protein 2 114569 4.66 10.33 0.0002 MYC v-myc myelocytomatosis viral oncogene homolog 4609 3.86 13.91 <0.0001 (avian) MYO6 myosin VI 4646 3.56 8.39 0.0021 PCA3 prostate cancer antigen 3 50652 19.36 140.07 0.047 PDGFC platelet derived growth factor C 56034 -1.97 -2.45 0.0093 PDLIM5 PDZ and LIM domain 5 10611 3.42 2.22 0.09 PPAP2B phosphatidic acid phosphatase type 2B 8613 -2.86 -1.75 0.0258 RARRES2 retinoic acid receptor responder (tazarotene induced) 2 5919 -4.23 -1.66 0.1991 SDC1 c syndecan 1 6382 -2.94 -2.92 0.063 SGEF Src homology 3 domain-containing guanine nucleotide 26084 4.89 2.71 0.0025 exchange factor SLC14A1 solute carrier family 14 (urea transporter), member 1 6563 -30.8 -41.79 0.0003 (Kidd blood group) SLC2A5 solute carrier family 2 (facilitated glucose/fructose 6518 -14.67 -60.13 0.0043 transporter), member 5 SPOCK3 sparc/osteonectin, cwcv and kazal-like domains 50859 -2.11 -3.69 0.0418
Page | 88
Chapter 3 Expression profiling of prostate adenocarcinoma tissues
proteoglycan (testican) 3 SRD5A2 steroid-5-alpha-reductase, alpha polypeptide 2 (3-oxo-5 6716 -8.63 -26.85 0.0297 alpha-steroid delta 4-dehydrogenase alpha 2) SRPRB signal recognition particle receptor, B subunit 58477 2.14 5.62 0.0004 STEAP4 STEAP family member 4 79689 3.01 11.50 0.0001 SULF2 c sulphatase 2 55959 -2.49 -2.05 0.0154 TCEAL2 transcription elongation factor A (SII)-like 2 140597 -3.83 -3.09 0.0304 TGFB3 transforming growth factor, beta 3 7043 -4.20 -2.87 0.0333 TGFBR3 c transforming growth factor, beta receptor III 7049 -2.55 -2.08 0.0362 (betaglycan, 300kDa) TP73L tumour protein p73-like 8626 -53.81 -753.95 0.0151 TPD52 tumour protein D52 7163 2.38 4.93 0.0011 TRIM29 tripartite motif-containing 29 23650 -5.54 -102.66 0.0182 TRIM36 tripartite motif-containing 36 55521 3.92 12.79 <0.0001 UAP1 UDP-N-acteylglucosamine pyrophosphorylase 1 6675 3.45 7.26 0.0013 VCAN c chondroitin sulphate proteoglycan 2, versican 1462 -2.59 -2.26 0.0442 VSNL1 visinin-like 1 7447 -4 -157.04 0.0021 WIF1 Wnt inhibitory factor 1 11197 -38.85 -45.10 0.0006 aFor relative fold changes of genes, as derived from microarray study, p-values are all <0.05. bNot detectable in 45 cycles of real-time PCR in prostate adenocarcinoma tissues, but detectable within 45 cycles for BPH tissues. cGenes highlighted in bold are genes coding for CS/DS/HS enzymes or genes coding for CSPG/DSPG/HSPG. dRelative fold changes are calculated using BPH tissues as control sections.
Page | 89
Chapter 3 Expression profiling of prostate adenocarcinoma tissues
3.8 Discussion
3.8.1 Aberrant expression of genes in prostate adenocarcinoma
Genome-wide expression profiling of human tissues using microarray platforms have been very beneficial in studies aimed at elucidating molecular pathways that have gone awry in diseases. Due to its high-throughput nature, thousands of genes can be screened in a short matter of time, giving rise to molecular pictures illustrating many genes that are differentially expressed in diseased conditions such as cancer, as compared to normal controls. This molecular picture obtained can then be applied in the field of cancer diagnosis and prognosis, such as the one developed in breast cancer, where the molecular patterns of thousands of gene expression have been used as prognostic markers of breast cancer outcome [Perou et al, 2000; Zajchowski et al,
2001].
The prerequisite of such a study for prognostic purposes would be to have a big sample population size and long-term follow up data. Prostate cancer incidence in
Singapore for the Chinese population occurs at a rate of 24.3 for every 100000 persons per year, from 2002-2006 [Seow et al, 2009]. From our experience with prostate tissue sample collection at Tan Tock Seng Hospital, we managed to collect prostate tissues from 62 different patients (of different races) with BPH or prostate cancer, over a span of approximately 2 years (2005 and 2006), out of which, after haematoxylin and eosin staining (to check for presence of cancer tissues) and RNA quality check using Agilent Bioanalyzer 2000 (only good quality RNA can be used for downstream microarray analysis), only 6 prostate adenocarcinoma tissue sections were found suitable for use in the microarray study. Long-term post-surgery follow up data for the prostate cancer patients are still in midst of acquisition. As such, the small sample sizes (n=6 for each of BPH control group and prostate cancer group) and lack
Page | 90
Chapter 3 Expression profiling of prostate adenocarcinoma tissues
of long term follow up data in this study currently do not allow delineation of
predictive models for prostate cancer progression.
Nevertheless, this study gives insight into some of the molecules that are dysregulated in prostate adenocarcinoma, which thereafter permits us to conduct further studies to determine which of these molecules are important players in the
prostate adenocarcinoma phenotype. In addition, molecules that are highly
differentially expressed can act as potential markers for prostate cancer diagnosis and/or prognosis. Our study has provided insights into prostate carcinogenesis in a
Singapore Chinese population using the Affymetrix GeneChip microarray platform.
Interestingly, some genes that have been reported to be differentially expressed in
previous studies [Welsh et al, 2001; Dhanasekaran et al, 2001; Luo et al, 2001;
Lapointe et al, 2004; Chandran et al, 2005], have also been found to be differentially expressed in our study. This suggests that these genes could be important factors in prostate cancer regardless of genetic makeup, geographic locality and technical operating issues. Due to difficulties in obtaining raw microarray data from most of those previous studies except those from Welsh et al, 2001, it was only possible to compare our filtered gene list with theirs. Welsh et al, 2001 compared the expression profiles of 25 prostate cancer tissues with 9 nonmalignant prostate tissues using the
Affymetrix GeneChip microarray platform. It was revealed that 247 genes (28.6% of my gene list) were found common to be differently expressed by 2 fold change in both their study and my study. Considering genes that were highlighted to be differentially expressed in prostate cancer from the other studies where raw data was not easily available, alpha-methylacyl-CoA-racemase (AMACR), hepsin (HPN) and fatty acid synthase (FASN) were found to be consistently upregulated in prostate cancer across all studies including those of mine.
Page | 91
Chapter 3 Expression profiling of prostate adenocarcinoma tissues
Using DAVID and based on the Gene Ontology Biological Function categories, the list of differentially expressed genes could be divided into different functional groups. As seen from the results above, the top-ranked annotation cluster in our gene list are the genes relating to development process and cell differentiation. A total of
243 genes are represented in this cohort. This is followed by, in decreasing group enrichment scores, genes involved in cell migration (41 genes), proliferation (69 genes), carbohydrate binding (26 genes), cellular response to stress (126 genes), homeostasis (59 genes), cell communication (212 genes), metabolism (87 genes), of which 46 genes are involved in lipid metabolism, cell cycle (37 genes) and cell death
(73 genes). Due to functional overlaps of some genes, one gene may be represented more than once in several annotation groups.
3.8.2 Prostate cancer as a result of development gone wrong?
It has long been hypothesized that the biology of cancer cell is a recapitulation of the biology of normal development and tissue renewal. Observational studies by
Pierce and Speers, 1988, described the similarities seen in both developing tissues and tumours, that both tissue types had numerous undifferentiated cells with high proliferative capacity and the presence of post-mitotic differentiated cells. This led to the proposal that both tissues types are sustained by a population of self-renewing stem cells that first give rise to undifferentiated, highly proliferative daughter cells and in the case of the developing and renewing tissue, the daughter cells will subsequently differentiate and remain growth quiescent. What is different in the tumour tissue is the inability of the progeny cells to undergo proper differentiation, resulting in the accumulation of a heterogeneous mass of undifferentiated cancer stem cells with great proliferative ability, as well as more differentiated cancer cells with limited proliferative capacity [Pierce and Speers, 1988]. The suggestion that the
Page | 92
Chapter 3 Expression profiling of prostate adenocarcinoma tissues
mechanisms regulating cell proliferation and differentiation in tumours is
fundamentally similar to the mechanisms regulating those of the developing and
renewing tissues is thus put forth; dysregulation of the normal stem cell self-renewal
pathway could result in unregulated control of cell proliferation and leads to
neoplastic transformation [Marker PC, 2008]. It has also been proposed that mutations
could cause the de-differentiation of mature differentiated cells and subsequent
acquisition of cancer stem cell-like properties.
Evidence for the existence of normal stem cells in prostate gland have been
identified: by using cell surface markers integrin α2β1 and CD133 [Collins et al,
2005; Gu et al, 2007; Patrawala et al, 2007], stem cells are found to be randomly
distributed throughout the acini and ductal regions, usually at the base of branching
points [Collins et al, 2001; Hudson et al, 2001; Richardson et al, 2004]. The adult prostate epithelial stem cells have been shown to sit on and be anchored to the basement membrane by integrin α2β1 [Collins et al, 2001].
Cancer stem cells do not necessarily originate from normal stem cells; the origins of the cancer stem cell are still not well-established but many postulations are made.
These include the ideas that cancer stem cells can arise from aberrant mutated stem cells, or due to accumulations of mutations in progeny daughter cells and perhaps even as a result of acquisition of stem-cell like properties of well-differentiated progeny cells that are undergoing de-differentiation [Moltzahn et al, 2008]. In the case of prostate cancer, prostate cancer stem cells are postulated to arise from the normal prostate stem cells; Collins et al, 2005, isolated cell populations from primary prostate cancers and found that only cells expressing α2β1/ CD133+/ CD44+ , which were
identical to normal prostate stem cells, could self-renew in vitro [Lang et al, 2009;
Collins et al, 2005]. Transcript levels of α2β1, CD133 and CD44 were not differently
Page | 93
Chapter 3 Expression profiling of prostate adenocarcinoma tissues
expressed in the prostate adenocarcinoma tissues as compared to BPH sections, which
could possibly indicate that there are no higher numbers of prostate cancer stem cells
in prostate tumours, but rather support the view that prostate cancer stem cells arise
from the normal prostate stem cells and hence, possibly hint at the dysregulation of
the normal stem cell self-renewal mechanisms, leading to prostate tumourigenesis and
progression.
Several pathways have been shown to regulate a balance between normal stem
cell self-renewal and apoptosis. These include the WNT signalling pathway, TGFβ
pathway, Sonic hedgehog and Notch signalling pathway [Lang et al, 2009; Moltzahn
et al, 2008]. Interestingly, these same signal transduction pathways have been
acknowledged to be able to trigger epithelial-mesenchymal transition (EMT)
programmes. EMT refers to “a complex molecular programme in which epithelial
cells sheds their well-differentiated characteristics, including cell-cell adhesion, planar
apical-basal polarity, lack of motility, and acquire mesenchymal features such as
motility, invasion and apoptosis resistance” [Polyak and Weinberg, 2009]. EMTs are
crucial in embryonic development and increasingly acknowledged to be important in
carcinogenesis as well, where they contribute to invasive properties, and generates
stem cell-like cells [Hugo et al, 2007; Yang & Weinberg, 2008; Mani et al, 2008].
Therefore, it appears that dysregulation of these pathways have significant
repercussions, from disrupting homeostatic balance of normal stem cell self renewal,
differentiation and apoptosis to triggering EMTs, resulting in neoplastic transformation and subsequent metastasis. In short, carcinogenesis appears to be development of the organ that has gone awry.
Indeed, results from my study appear to agree with the idea that cancer could be a dysregulated recapitulation of development and differentiation and due to activation
Page | 94
Chapter 3 Expression profiling of prostate adenocarcinoma tissues
of EMT programmes. Two hundred and forty-three genes in our differentially
expressed gene list are genes involved in development and differentiation. These
perhaps suggest that the genes regulating development and differentiation are
dysregulated resulting in tumourigenesis or from another view, these genes may be
selected for use during prostate cancer progression. Many of the genes in our list
belong to the WNT signalling pathway, hedgehog signalling pathway and TGFβ
pathway, and in addition to that, many of these genes are homeobox genes, all of
which are established to be implicated in organogenesis, tumourigenesis and EMTs.
3.8.2.1 Dysregulation of Wnt signalling pathway
The WNT signalling pathway is a highly conserved pathway, crucial in
embryonic development in many animal species and continues to be important in the
regeneration of many adult tissues [Klaus and Birchmeier, 2008]; dysregulation of
WNT pathway has been implicated in many cancers to date [Akiri et al, 2009; Bisson
and Prowse, 2009; Bandapalli et al, 2009] and in addition, specification and
maintenance of precursor cell and stem cell lineages have been shown to involve the
Wnt signalling pathway too [Bisson and Prowse, 2009; Evans T, 2009; Goessling et
al, 2009]. From our gene list, 17 genes, known to be involved in the WNT signalling
pathway, have been found to be differentially expressed in prostate adenocarcinoma as compared to benign prostate hyperplastic tissues. These 17 genes include WNT inhibitory factor 1 (WIF1), frizzled homolog 10 (Drosophila) (FZD10), frizzled
homolog 2 (Drosophila) (FZD2), frizzled homolog 7 (Drosophila) (FZD7), frizzled
homolog 1 (Drosophila) (FZD1), wingless-type MMTV integration site family, member 4 (WNT4), wingless-type MMTV integration site family member 2 (WNT2),
wingless-type MMTV integration site family, member 5B (WNT5B), secreted
frizzled-related protein 1 (SFRP1), dickkopf homolog 3 (Xenopus laevis) (DKK3), R-
Page | 95
Chapter 3 Expression profiling of prostate adenocarcinoma tissues spondin 3 homolog (Xenopus laevis) (RSPO3), DIX domain containing 1 (DIXDC1), receptor tyrosine kinase-like orphan receptor 2 (ROR2), transforming growth factor beta 1 induced transcript 1 (TGFB1/1), phosphatidic acid phosphatase type 2B
(PPAP2B), transcription factor 7-like 1 (T-cell specific, HMG-box) (TCF7L1), transducin-like enhancer of split 2 (E(sp1) homolog, Drosophila) (TLE2).
It is interesting to note that all 17 genes are downregulated in prostate adenocarcinoma tissues, the most prominent gene being WIF1, which is downregulated by approximately 37 fold. WIF1 encodes a secreted protein that binds and inhibits WNT proteins and have been reported to be a tumour suppressor gene that is frequently silenced in tumours [Chim et al, 2008; Kansara et al, 2009; Licchesi et al, 2008; Elston et al, 2008]. RNA and protein levels of WIF1 were also found to be reduced in prostate cancers, in a separate study conducted by Wissman et al.
[Wissmann et al, 2003]. In addition, SFRP1 and DKK, soluble WNT antagonists, are oftenly found downregulated in several tumours and suggested to be potential tumour suppressor genes [Gauger et al, 2009; Chung et al, 2009; Maehata et al, 2008; Suzuki et al, 2008]. Loss of DKK3 has been reported to result in loss of prostate glandular architecture and overproliferation of prostate epithelial cells, a hallmark of prostate cancer progression [Kawano et al, 2006].
Frizzled (FZD) are cell surface receptors of WNT ligands, and both receptors and ligands encompass a huge family of variants; there are 10 known frizzled (FZD) and
12 known WNT ligands in the pathway, but exact roles and interplay between each
FZD and WNT ligands in the activation or repression of the WNT transduction signal remains to be elucidated. MYC, a well-known proto-oncogene, is a target of the WNT signalling pathway [He et al, 1998], and this gene has been shown to be upregulated by 4 folds in the cohort of prostate adenocarcinomas studied here. All in, these results
Page | 96
Chapter 3 Expression profiling of prostate adenocarcinoma tissues
have clearly demonstrated that many components of the WNT signalling pathway is
dysregulated in prostate adenocarcinoma, and this dysregulation could potentially
disturb the balance between stem cell renewal, differentiation and apoptosis
ultimately resulting in neoplastic transformation of the prostate gland.
3.8.2.2 Dysregulation of homeobox genes
Homeobox genes encode nuclear proteins that act as transcription factors, which
are important in regulating morphogenesis and cell differentiation during embryonic
development [Nunes et al, 2003]. Homeobox genes are categorized into clustered (or
HOX genes) and non-clustered genes. HOX genes are organized in 4 genomic clusters, HOX-A, HOX-B, HOX-C and HOX-D, located at different genomic loci while non-clustered homeobox genes are dispersed everywhere at different positions on different chromosomes [Cillo et al, 2001]. From our study, several homeobox genes that are differentially expressed in prostate adenocarcinomas, that includes homeobox C6 (HOXC6), distal-less homeobox 1 (DLX1), homeobox C4 (HOXC4),
homeobox B2 (HOXB2), homeobox B3 (HOXB3), homeobox D10 (HOXD10) and
NK2 transcription factor related, locus 3 (Drosophila) (NKX2-3) have been identified.
HOXC6 was upregulated by nearly 12 fold in prostate adenocarcinoma tissues used in
our study, and other studies have also reported the overexpression of HOXC6 in
prostate cancers that correlated with cancer progression [McCabe et al, 2008]. Loss of
HOXC6 was found to induce apoptosis in prostate cancer cells, which infers that a
high expression of HOXC6 could confer apoptosis resistance in prostate
adenocarcinoma [Ramachandran et al, 2005]. In addition, HOXC6 was found to
indirectly regulate Notch and WNT signalling pathways in prostate cancer in vivo
[McCabe et al, 2008]. HOXC4 was also shown to be upregulated in prostate cancers
in our study, consistent with another study performed by Miller GJ et al. [Miller et al,
Page | 97
Chapter 3 Expression profiling of prostate adenocarcinoma tissues
2003]. However, the other HOX genes found in a different cluster from HOX-C,
namely HOXB2, HOXB3, HOXD10 were downregulated in our prostate
adenocarcinoma tissues. The expression and functions of these HOX genes in prostate
cancer have rarely been reported to date. Again, differential expression of these
homeobox genes in prostate adenocarcinoma could result in aberrations in cell
differentiation and apoptosis, piloting the way for carcinogenesis.
3.8.3 Prostate cancer – sustaining growth and motility while evading death signals
The hallmarks of cancer cells are to sustain cell growth and proliferation, evade
apoptosis, detach from cell-cell adhesions, invade into the surrounding stroma and migrate to metastatic sites [Hanahan and Weinberg, 2000]. Malignant transformation
is viewed as a multistep process comprising of the above processes, and these steps
reflect genetic alterations that drive the cells to malignancy. As anticipated, many
genes known to be involved in cell cycle, proliferation, migration and apoptosis, were
observed to be differentially expressed in the prostate adenocarcinoma tissues used in
our study. Delineating the functional significance of the differential expression of these genes will be important though, so as to demarcate genes that could play central roles, not necessarily causative factors, in prostate tumourigenesis, from genes that are differentially expressed as a consequence of the transformation process. In doing so, these could provide us potential novel therapeutic targets that could impede the progression of prostate cancer.
3.8.4 Prostate cancer as a result of aberration of metabolic processes?
Cancer has hardly been recognized as a disease due to aberration in metabolic pathways but until recently, tumour bioenergetics has been thrown into the spotlight.
Certainly, the Warburg effect, exemplified by the dependence of cancer cells on
Page | 98
Chapter 3 Expression profiling of prostate adenocarcinoma tissues
glycolysis (anaerobic breakdown of glucose into ATP) even in the presence of
oxygen, has long been recognized since nearly 80 years ago. Otto Warburg thought
that this phenomenon was a main cause of cancer, which has subsequently been
discredited over the years due to identification of cancers that do not rely heavily on
aerobic glycolysis for energy [Zhivotovsky and Orrenius, 2009]. However, in recent years, reports were made linking metabolic changes to malignant transformation, but
this comes with controversy and debate [Gatenby and Gillies, 2004; Bernards and
Weinberg, 2002; Garber K, 2006]. Some refute that metabolic changes are
consequences of cancer rather than as contributors to it. No doubt such metabolic
changes could provide cancer cells with growth and survival advantages in the tumour
microenvironment, as glycolysis gives an added boost of energy to cancer cells and
provides sustenance and autonomy. Hence, it will be valuable to target components of
the metabolic pathway for anti-tumour therapeutics.
From our gene list, 104 genes associated with various metabolic pathways have
been shown to be differentially expressed in prostate adenocarcinoma. The
upregulation of fructose-1,6-bisphosphatase (FBP1) and hexokinase 2 (HEX2) seems
to highlight the prominence of high glucose utilization in the prostate tumour cells;
fructose-1,6-bisphosphatase is an enzyme that converts fructose-1,6-bisphosphate to fructose during gluconeogenesis [Marcus et al, 1987], and increased expression of this
gene implies increased glucose synthesis. Moreover, hexokinases phosphorylate
glucose to yield glucose-6-phosphate (Glc-6-P), which is the first essential step of
glucose metabolism. This allows glucose entry into the cells through maintaining a
concentration gradient, and is fundamental to all major pathways of glucose
utilization, including glycolysis, the pentose phosphate pathway (PPP) and
Page | 99
Chapter 3 Expression profiling of prostate adenocarcinoma tissues glycogenesis [Mathupala et al, 2009]. Again, this emphasizes the high glucose requirement of the tumour cells.
Of significance too, is the huge representation of lipid metabolism genes (46 genes) in our differentially expressed gene list, with the prominent upregulation of fatty acid synthase (FASN) in the prostate adenocarcinoma tissues. FASN is a lipogenic enzyme catalyzing the terminal steps in the de novo synthesis of fatty acids; the endogenous fatty acid production could be an important source of energy supply for tumour growth, or a crucial source of structural constituents of the membrane for actively proliferating cells, and/or to synthesize lipid-based factors that could act as autocrine/paracrine factors, all of which could confer growth advantage and survival for the tumour cells [Menendez and Lupu, 2006]. Many tumours were observed to undergo exacerbated fatty acid synthesis with a concomitant FASN overexpression
[Innocenzi et al, 2003; Takahiro et al, 2003; Wang et al, 2004; Visca et al, 2004]. In prostate cancer, this FASN increase was associated with a four-fold increase risk of disease recurrence and with higher Gleason sum [van de Sande et al, 2005]. Inhibition of FASN leading to reduced FA synthesis has resulted in cell cycle arrest, reduced growth and induced apoptosis [Migita et al, 2009], suggesting that FASN may play a role that integrates metabolism and cell survival signalling pathways. This positions
FASN as a potential drug target for cancer therapy.
3.8.5 Aberrant expression of GAGs and PGs genes in prostate cancer
Essentially, we have also identified many genes coding for glycosaminoglycan biosynthetic and editing enzymes and proteoglycans that are differentially expressed in prostate adenocarcinomas. The genes coding for GAG synthesis and modifying enzymes that are significantly upregulated (p<0.05) by more than 2 folds include
UDP-Gal:betaGal beta 1,3-galactosyltransferase polypeptide 6 (B3GalT6),
Page | 100
Chapter 3 Expression profiling of prostate adenocarcinoma tissues
chondroitin sulphate N-acetylgalactosaminyltransferase 1 (CSGalNAcT1) and sulphatase 2 (SULF2) while those downregulated by more than 2 fold are B cell RAG associated protein (GalNAcT4S-6ST), heparan sulphate 6-O-sulphotransferase 2
(HS6ST2), heparan sulphate (glucosamine) 3-O-sulphotransferase 3A1 (HS3ST3A1) and heparan sulphate (glucosamine) 3-O-sulphotransferase 3B1 (HS3ST3B1).
Proteoglycans like syndecan-1 (SDC1), decorin (DCN), versican (VCAN) and
transforming growth factor beta receptor type III (TGFBR3) are downregulated by at
least 2 fold from the microarray result.
B3GalT6 codes galactosyltransferase II, which is responsible for the addition of
the second galactose in the tetrasaccharide linkage region of the GAG chain to the
core protein. The increased expression of B3GalT6 could represent the increased
synthesis and attachment of various glycosaminoglycans to the protein core.
CSGalNAcT-1, on the other hand, is required for the initiation of chondroitin sulphate
biosynthesis after the formation of the linkage region, which is common in HSPG,
CSPG and DSPG. Again, the upregulation of CSGalNAcT-1 could represent increased
biosynthesis of chondroitin sulphate chains on the proteoglycan. Expression changes
and their roles in cancer biology have rarely been reported to date.
Specific sulphation motifs have been known to function as molecular
recognition elements for various growth factors and this functional information
encoded by glycosaminoglycans is analogous to the information provided by DNA,
RNA and proteins [Gama et al, 2006]. For instance, chondroitin-4,6-sulphate has been
shown to be highly expressed by highly metastatic lung carcinoma cells and involved
in cell proliferation, migration and invasion while chondroitin-4-sulphate did not
appear to be crucial [Li et al, 2008]. Different sulphation patterns of heparan sulphate
had different binding properties with different members of the fibroblast growth
Page | 101
Chapter 3 Expression profiling of prostate adenocarcinoma tissues
family too [Ashikari-Hada S et al, 2004 and 2009; Sugaya et al, 2008]. In our study,
only one chondroitin sulphotransferase, the GalNAcT4S-6ST, was found to be
downregulated by at least 2 fold. In fact, GalNAcT4S-6ST was downregulated by 4
fold in the prostate adenocarcinoma tissue as compared to benign prostate tissues.
Interestingly, the expression of GalNAcT4S-6ST was found overexpressed in other
tumour types: highly metastatic lung carcinoma cells compared to low metastatic
cells, colorectal cancer [Ito et al, 2007], pancreatic ductal adenocarcinoma [Sugahara
et al, 2008] and ovarian adenocarcinoma [ten Dam et al, 2007]. It remains to be
elucidated the exact tissue-specific function of GalNAcT4S-6ST and chondroitin-4,6-
sulphate moieties in different tumours.
The heparan sulphotransferases differentially expressed in our prostate
adenocarcinoma tissues belong mainly to 2 families: the heparan-3-O-
sulphotransferases (HS3ST3A1 and HS3ST3B1) and the heparan-6-O-
sulphotransferases (HS6ST2). Notably, sulphatase 2 (SULF2), an enzyme that
selectively removes the 6-O-sulphate groups from heparan sulphate, was upregulated
in prostate adenocarcinoma. This, combined with a downregulation of the HS6ST2
gene, could indicate that the expression of heparan-6-O-sulphate moiety is reduced in
prostate adenocarcinoma and perhaps play an important role in carcinogenesis, thus
warrants further investigation. HS3ST3A1 and HS3ST3B1 are both downregulated in
prostate adenocarcinoma tissues and these merit further investigation into their roles and their corresponding heparan-3-O-sulphate motifs in prostate tumourigenesis.
Syndecan-1 (SDC1) is a proteoglycan containing both heparan sulphate and
chondroitin sulphate chains. From our microarray study, SDC-1 demonstrated nearly
3 fold downregulation in the prostate adenocarcinoma tissues with p<0.05; however,
real-time PCR verification showed also a similar 3 fold downregulation albeit this did
Page | 102
Chapter 3 Expression profiling of prostate adenocarcinoma tissues
not reach statistical significance (P=0.063). This could result from a small sample size
(n=6 per group) and inherent variation of SDC-1 expression in different prostate
tumours, resulting in sampling variation. This is actually consistent with previous
studies, in which they found that syndecan-1 is altered, but not uniformly absent in
prostate adenocarcinomas. Shariat et al, 2008, reported that although SDC-1 is found
expressed only in a small percentage of prostate adenocarcinoma tissues (37.1%), it
has prognostic value. Patients with SDC-1 expression were at significantly greater risk
of aggressive progression after surgery [Chen et al, 2004; Kiviniemi et al, 2004]. Most
recently, Shimada et al, 2009, have found that SDC-1 was found at higher levels in
androgen-independent cell lines, DU145 and PC-3, as compared to androgen-
dependent LNCaP, and yet the expression of SDC-1 was increased in LNCaP after
long-term androgen deprivation. Using siRNA studies, SDC-1 repression resulted in
increased apoptosis and reduced angiogenesis, indicating that SDC-1 might participate
in the progression from androgen-dependent to androgen-independent status. Since
SDC1 expression appears to favour aggressive progression of cancer cells, it would perhaps be interesting to investigate further why only a small percentage of prostate cancers (37.1%) express SDC1. SDC1 expression could possibly be temporal and switched on only after a certain tumour progression phase, which allows the progression of androgen-dependent cancer to an androgen-independent one. If so, these results suggest a role for SDC-1 in prostate cancer progression to hormone- resistant cancer, and could prove to be a valuable therapeutic target, especially in late stage prostate cancer.
Decorin (DCN) and versican (VCAN), both proteoglycans containing chondroitin sulphate and dermatan sulphate chains, were found to be downregulated in prostate adenocarincoma tissues in our study. Likewise, Banerjee et al, 2003,
Page | 103
Chapter 3 Expression profiling of prostate adenocarcinoma tissues reported that stromal areas of BPH stained intensively for DCN and decreased with advancing stages of prostate cancer [Banerjee et al, 2003]. They also reported on versican expression to be moderate in all BPH tissues, but weak in most of the moderately differentiated prostate cancer, yet moderately focal in poorly- differentiated prostate cancer. In contrary to the above study, Ricciardelli et al, 1998, reported the increased expression of versican and decorin in the peritumoural fibromuscular stroma in the prostatic tissue of men with early-stage prostate cancer compared with tissue from men without cancer. The increase in versican concentration but not that of decorin was associated with increased risk of prostate- specific antigen (PSA) progression [Ricciardelli et al, 1998]. Variation of versican and decorin expression across different studies could be attributed to the different patient cohort used and the inherent highly variable nature of VCAN and DCN expression in prostate tumours. Decorin has been known to modulate signalling mechanisms in the tumour microenvironment [Wegrowski and Maquart, 2004;
Goldoni and Iozzo, 2008]. Decorin binds to and modulates the signalling of the epidermal growth factor receptor and other members of the ErbB family of receptor tyrosine kinases [Santra et al, 2000]. p21, an inhibitor of cyclin-dependent kinases, is also upregulated as a result of decorin-induced signalling, leading to growth suppression [Ständer et al, 1999; de Luca et al, 1996]. Consequently, decorin suppresses angiogenesis and metastatic spreading, and can enhance apoptosis through caspase-3 activation as well. The role of versican in cancer biology appears more complex. It has been reported to cause both apoptotic resistance and sensitivity
[LaPierre et al, 2007], and is generally found to be overexpressed in many cancers such as laryngeal cancer [Vynios et al, 2008], cervical cancer [Kodama et al, 2007] and endometrial cancer [Kodama et al, 2007]. Versican upregulation has often been
Page | 104
Chapter 3 Expression profiling of prostate adenocarcinoma tissues
associated with worse prognosis in such studies too. However, post translational
modifications in versican might add another dimension of complexity in tumour
progression as well. Post-translational modifications, such as the type, length and the
sulphation pattern of the GAG chains on versican, were found to be different in
human colon adenocarcinoma versus human normal colon and pancreatic cancer
versus normal pancreas [Theocharis AD, 2002]. These specific structural alterations
of versican may influence the biology of cancer cells too, in addition to transcript
changes and more study is necessitated to understand the complex biology of versican
in cancer.
The transforming growth factor-beta (TGFβ) pathway has been implicated in the progression of prostate adenocarcinoma [Jones et al, 2009; Bello-DeOcampo and
Tindall, 2003]. This pathway has been paralleled to a double edge sword, where it inhibits tumour growth in early stages of prostate tumourigenesis but subsequently promotes tumour capabilities in metastatic prostate cancer [Bello-DeOcampo and
Tindall, 2003]. This pathway is known to inhibit prostate cell proliferation, differentiation and induce apoptosis but when growth inhibitory signals are lost, it promotes invasion and metastasis. From our study, we observed a loss of TGFβRIII
(betaglycan) and TGFβIII in prostate adenocarcinomas as compared to BPH samples, while there were no other observable changes in the expression of other members of the TGFβ family such as TGFβRI and TGFβRII. This result can be supported with findings from other studies. Kim et al, 1996 and 1998, and Guo et al, 1997, reported loss of TGFβRI and TGFβRII expression in only 30% of prostate tumours [Kim et al,
1996 & 1998; Guo et al, 1997] but the TGFβRIII is the most commonly perturbed member of the TGFβ pathway in prostate cancer [Sharifi et al, 2007; Turley et al,
2007]. Loss of TGFβRIII correlates with increased tumour grade and PSA relapse.
Page | 105
Chapter 3 Expression profiling of prostate adenocarcinoma tissues
Functional studies revealed that TGFβRIII downregulation impedes tumour cell
motility and invasiveness [Mythreye and Blobe, 2009]. Future work could be done to
investigate the mechanisms by which TGFβRIII regulate prostate cancer cell
behaviour.
Heparanase transcript levels (both HPSE and HPSE2) were not observed to be
differentially expressed in the prostate adenocarcinomas studied here. This is in
contrast to a study by Lerner et al, 2008, in which they report heparanase
overexpression in prostate cancers using immunohistochemistry, and the
overexpression of heparanase was associated with high tumour grade [Lerner et al,
2008]. Functional studies using prostate cancer cell lines showed that heparanase
regulates tumour growth and faciliatates angiogenesis [Zhou et al, 2008; Kosir et al,
1999]. The discrepancy of our findings with those in literature is not surprising, as
heparanase is not unanimously overexpressed in all prostate cancers. 14% of the
prostate tumours studied by Lerner et al, 2008, had no heparanase staining as well.
The small sample size in our study could attribute to the incongruity as well, as our
samples could represent the small percentage of tumours that did not express
heparanase.
3.8.6 Conclusions
This study has allowed the identification of 865 candidate genes that are altered in
malignant prostate tissues obtained from a local Singapore Chinese population. This
will set the stage for future work on identifying good prognostic markers, especially
pertaining to a local context. In addition, the delineation of prostate tumours into different subclasses can potentially be achieved in the future with the collection of more prostate tissues for a bigger sample size.
Page | 106
Chapter 4 Chondroitin sulphate in prostate tissue
Chapter 4
Expression and Functional Analysis of Chondroitin Sulphate in Prostate Cancer
4.1 Expression analysis of chondroitin sulphate in prostate cancer
4.1.1 Clinical and demographic data
A total of one hundred and sixty-three prostate tissue samples (consisting of both benign hyperplastic prostatic tissue and prostate adenocarcinoma tissue specimens) were used for immunohistochemical evaluation. The age of the patients ranged from 45 to 90 years, with a mean of 65.5 years and a median of 64 years. The study cohort predominantly consisted of Chinese patients (127 cases), while the remaining cases included Malays (20 cases), Indians (4 cases) and other minor ethnic groups (7 cases).
Clinicopathological features of the cancer samples used are shown in Table 4.1. Due to loss of tissue sections during immunohistochemical processing, the number of cases used to determine the expression of the various chondroitin sulphate subtypes in prostate adenocarcinoma could vary and is detailed in the corresponding sections.
Page | 107
Chapter 4 Chondroitin sulphate in prostate tissue
Table 4.1 Clinical features of prostate adenocarcinoma cases used for immunohistochemistry
Clinicopathological parameter Frequency distribution Clinicopathological parameter Frequency distribution
Age (years) Median 64 Maximum tumour dimension Minimum 45 <2.5cm 46 Maximum 90 ≥2.5cm 40 Not available 6 Pathological T staging Core biopsies 71 T2a 5 T2b 5 Tumour volume T2c 39 <4.0 cm3 37 T3a 29 ≥4.0 cm3 32 T3b 4 Not available 23 T3c 5 Core biopsies 71 At least T2ca 4 Not available 1 Gleason sum Core biopsies 71 Gleason sum 3 6 Gleason sum 4 19 Extraprostatic extension Gleason sum 5 5 Yes 38 Gleason sum 6 36 No 48 Gleason sum 7 44 Cannot be assessedb 4 Gleason sum 8 2 Not available 2 Gleason sum 9 44 Core biopsies 71 Gleason sum 10 7
Page | 108
Chapter 4 Chondroitin sulphate in prostate tissue
Clinicopathological parameter Frequency distribution Clinicopathological parameter Frequency distribution Lymphovascular invasion Apex (distal) margin involvement Present 18 Involved 41 Absent 138 Not involved 49 Not available 7 Not available 2 Core Biopsies 71 Obturator lymph node metastasis Yes 0 Bladder base (proximal) margin involvement No 49 Involved 10 Not sampled 43 Not involved 81 Core biopsies 71 Not available 1 Core Biopsies 71 Seminal vesicle involvement Yes 8 Clinical staging No 82 T1b 1 Not available 2 T1c 44 Core Biopsies 71 T2a 4 T2b 4 Perineural invasion T2c 1 Seen 93 Not available 38 Not seen 68 Core Biopsies 71 Not available 2 Lobe occurence Preoperative PSA Both lobes 77 ≤ 8 ng/ml 27 Single lobe 12 > 8 ng/ml 25 Not available 3 Not available 111 Core Biopsies 71
Page | 109
Chapter 4 Chondroitin sulphate in prostate tissue
Clinicopathological parameter Frequency distribution Resection margin involvement (other than at apex and bladder base/proximal margin) Involved 38 Not involved 53 Not available 1 Core Biopsies 71
aPathological T staging is classified as ‘at least T2c’ due to inability to ascertain the capsular status of the tumour. bCapsular integrity cannot be ascertained as the site where the positive surgical margin is does not disclose the capsule of the gland.
Page | 110
Chapter 4 Chondroitin sulphate in prostate tissue
4.1.2 Expression analysis of non-sulphated chondroitin in prostate adenocarcinoma
4.1.2.1 Localization of non-sulphated chondroitin in prostatic tissues
Expression analysis of the non-sulphated chondroitin was examined in a total
of 116 cases of benign prostatic hyperplastic tissues and 142 cases of prostate
adenocarcinoma cases, in which 95 of these cases are matched cases (i.e both BPH
samples and prostate adenocarcinoma tissues were taken from the same patient).
In benign prostatic tissue, non-sulphated chondroitin was expressed at
moderate to high intensity in basal cells and in the periglandular stroma (Figure
4.1A). Weaker staining intensity was detected in the cytoplasm of secretory cells and in the stromal connective tissue away from the immediate periglandular zones. In
prostate cancer, immunohistochemical staining for non-sulphated chondroitin was
seen in the cytoplasm of malignant cells and in the surrounding stroma (Figure 4.1B
to 4.1D). The expression data of non-sulphated chondroitin in the 116 cases of benign prostatic tissues and 142 cases of prostate adenocarcinoma samples are shown in
Table 4.2.
4.1.2.2 Comparison of non-sulphated chondroitin expression in benign prostatic and adenocarcinoma tissues As shown in Figure 4.2 and Table 4.3, when the staining differences in the
means of 95 cases of matched benign and prostate adenocarcinoma tissues were
compared, it was found that cancer cells were more intensely stained compared
against benign secretory cells (P=0.0008). Furthermore, an average of 40.11% of
cancer cells were positively stained for non-sulphated chondroitin, in contrast to an
average of 24.90% of secretory cells in the benign prostatic group (P=0.0016). The
immunoreactivity score (IRS) of the prostate adenocarcinoma cells were also
Page | 111
Chapter 4 Chondroitin sulphate in prostate tissue
significantly higher (mean of 71.77) than the benign hyperplastic prostates (mean of
36.18). However, no significant changes in staining intensity or percentage of cells stained in the peritumoural stroma, as compared to the periglandular stroma were observed.
Page | 112
Chapter 4 Chondroitin sulphate in prostate tissue
Figure 4.1 Non-sulphated chondroitin expression in prostate tissues. [A] BPH [B] Gleason sum 4 Section [C] Gleason sum 7 section [D] Gleason sum 10 section [E] Negative Control ‘BL’ represents basal cell layer, ‘Epi’ represents benign epithelial cell, ‘CaP’ represents cancer cells and ‘SM’ represents stromal cells.
Page | 113
Chapter 4 Chondroitin sulphate in prostate tissue
Table 4.2 Expression analysis of non-sulphated chondroitin in benign hyperplastic prostate tissues and prostate adenocarcinoma tissues.
Number of cases stained Number of Stained Tissue Total Weighted Average Intensity Percentage of cells cases Immunoreactivity Score (IRS) Component number (WAI) stained (%) unstained ≤1 1 Adenocarcinoma Cancerous cells 142 52 26 22 42 50 40 54 28 8 36.62% 18.31% 15.49% 29.58% 35.21% 28.17% 38.03% 19.72% 5.63% Peritumoural 142 40 41 20 41 53 49 59 23 20 Stromal cells 28.17% 28.87% 14.08% 28.87% 37.32% 34.51% 41.55% 16.20% 14.08% Page | 114 Chapter 4 Chondroitin sulphate in prostate tissue Figure 4.2 Comparisons of non-sulphated chondroitin expression in prostate adenocarcinoma samples (n=95) against matched BPH samples (n=95). [A] Weighted Average Intensity [B] Percentage of cells positively stained [C] Immunoreactivity Score ‘**’ represents p<0.01; ‘***’ represents p<0.001. P-values were determined by Paired T-test. Page | 115 Chapter 4 Chondroitin sulphate in prostate tissue Table 4.3 Mean values of non-sulphated chondroitin staining in prostate adenocarcinoma samples (n=95) versus paired BPH samples (n=95). Weighted Average Intensity Percentage of cells stained (%) IRS Mean SEM P-value Mean SEM p-value Mean SEM P-value Epithelial / Cancer cell region Benign secretory epithelial 0.83 0.09 0.0008 24.90 3.18 0.0016 36.18 5.07 0.0001 cells Adenocarcinoma cells 1.28 0.10 40.11 3.82 71.77 7.74 Stromal region Periglandular stromal cells 0.95 0.10 0.7377 26.84 3.60 0.212 46.48 7.24 0.7323 Peritumoural stromal cells 0.91 0.09 32.30 3.81 49.59 6.99 Page | 116 Chapter 4 Chondroitin sulphate in prostate tissue 4.1.2.3 Comparison of non-sulphated chondroitin expression in different grades of prostate adenocarcinoma tissues To take a step further, the expression levels of non-sulphated chondroitin in both the cytoplasm of the prostate adenocarcinoma cells and the peritumoural stroma were associated with different tumour grades of prostate adenocarcinoma. The tumour grades were classified into 3 groups: Gleason sum ≤5, Gleason sum 6 and 7 and Gleason sum ≥8 and the means of chondroitin sulphate staining of each group were compared. As seen from Figure 4.3, a decreasing expression of non-sulphated chondroitin staining (in terms of staining intensity, percentage of cells positively stained and immunoreactivity score) in the prostate adenocarcinoma cells were observed with increasing tumour grade while there was an increasing non-sulphated chondroitin expression (increased staining intensity, percentage of cells positively stained and immunoreactivity score) in the peritumoural stroma with increasing tumour grade. The mean values of non-sulphated chondroitin staining in the various groups of tumour grade are shown in Table 4.4. By using One-way ANOVA, it was demonstrated that the means are significantly different. These results suggest a potential role of non-sulphated chondroitin involvement in prostate cancer progression and warrants further investigation. Page | 117 Chapter 4 Chondroitin sulphate in prostate tissue Figure 4.3 Comparisons of non-sulphated chondroitin expression in different grades of prostate adenocarcinoma tissues: Gleason sum 3,4 & 5 (n=22) vs Gleason sum 6 & 7 (n=68) vs Gleason sum 8, 9 & 10 (n=52), in both [A,B,C] cytoplasm of cancer cell and [D,E,F] peritumoural stroma. [A,D] Weighted Average Intensity [B,E] Percentage of cells positively stained [C,F] Immunoreactivity score. ‘*’ represents p< 0.05; ‘**’ represents p<0.01; ‘***’ represents p<0.001. P-values were determined by One-way ANOVA with Bonferroni’s Multiple Comparison Post test. Page | 118 Chapter 4 Chondroitin sulphate in prostate tissue Table 4.4 Mean values of non-sulphated chondroitin staining in different grades of prostate adenocarcinoma tissues: Gleason sum 3,4 & 5 (n=22) vs Gleason sum 6 & 7 (n=68) vs Gleason sum 8, 9 & 10 (n=52). P-values were determined by One-way ANOVA with Bonferroni’s Mulitple Comparison Post test. Gleason Sum Weighted Average Intensity Percentage of cells positively stained Immunoreactivity Score Mean SEM P-value Mean SEM P-value Mean SEM P-value Adenocarcinoma 3,4 & 5 1.26 0.16 0.0092 54.09 8.82 <0.0001 82.02 16.12 0.0063 6 & 7 1.30 0.11 39.01 4.17 70.26 8.83 8, 9 & 10 0.77 0.15 17.47 4.18 34.96 8.45 Peritumoural Stroma 3,4 & 5 0.47 0.15 <0.0001 24.35 8.02 0.0001 33/04 14.13 <0.0001 6 & 7 1.08 0.10 33.86 4.40 56.07 8.99 8, 9 & 10 1.70 0.14 59.90 5.65 124.4 13.75 Page | 119 Chapter 4 Chondroitin sulphate in prostate tissue 4.1.3 Expression analysis of chondroitin-4-sulphate in prostate adenocarcinoma. 4.1.3.1 Localization of chondroitin-4-sulphate in prostate tissues Expression analysis of the chondroitin-4-sulphate was examined in a total of 117 cases of benign prostatic hyperplastic tissues and 146 cases of prostate adenocarcinoma cases, in which 100 of these cases are matched cases (i.e both BPH samples and prostate adenocarcinoma tissues were taken from the same patient). Chondroitin-4-sulphate was not detected in the benign secretory epithelial cells in about 71.79% of benign prostate hyperplastic cases; while those that expressed cytoplasmic staining in the benign epithelial cells, intensity of staining varied from weak to strong (Figure 4.4A). Strong (WAI ≥ 2) periglandular stroma staining was observed in about 45.30% of the benign prostate tissues. 56.41% of the BPH cases had more than 50% of the periglandular stroma tissue stained. In prostate cancer, immunohistochemical staining for chondroitin-4-sulphate was observed more prominently in the peritumoural stroma while there was some staining in the cytoplasm of the malignant cells for about 20.55% of the adenocarcinoma tissues, with a majority of adenocarcinoma cells (63.01%) not expressing any chondroitin-4- sulphate in the cytoplasm of the cancer cells (Figure 4.4B to 4.4D). The expression data of chondroitin-4-sulphate in the 117 cases of benign prostatic tissues and 146 cases of prostate adenocarcinoma samples are shown in Table 4.5. Page | 120 Chapter 4 Chondroitin sulphate in prostate tissue Figure 4.4 Chondroitin-4-sulphate expression in BPH tissue and prostate adenocarcinoma sections. [A] BPH [B] Gleason sum 4 section [C] Gleason sum 7 section [D] Gleason sum 9 section [E] Negative control ‘BL’ represents basal cell layer, ‘Epi’ represents benign epithelial cell, ‘CaP’ represents cancer cells and ‘SM’ represents stromal cells. Page | 121 Chapter 4 Chondroitin sulphate in prostate tissue Table 4.5 Expression analysis of chondroitin-4-sulphate in benign hyperplastic prostate tissues and prostate adenocarcinoma tissues Number of cases stained Number Weighted Average Intensity Percentage of cells Stained Tissue Total Immunoreactivity Score (IRS) of cases (WAI) stained (%) Component number unstained ≤1 1 BPH Secretory epithelial 117 84 12 5 16 23 10 23 6 4 cells 71.79% 10.26% 4.27% 13.68% 19.66% 8.55% 19.66% 5.13% 3.42% Periglandular 117 11 34 19 53 40 66 60 21 25 Stromal cells 9.40% 29.06% 16.24% 45.30% 34.19% 56.41% 51.28% 17.95% 21.37% Adenocarcinoma Cancerous cells 146 92 30 9 15 31 23 41 9 4 63.01% 20.55% 6.16% 10.27% 21.23% 15.75% 28.08% 6.16% 2.74% Peritumoural 146 19 59 23 45 58 69 90 16 21 Stromal cells 13.01% 40.41% 15.75% 30.82% 39.73% 47.26% 61.64% 10.96% 14.38% Page | 122 Chapter 4 Chondroitin sulphate in prostate tissue 4.1.3.2 Comparison of chondroitin-4-sulphate expression in benign prostatic and adenocarcinoma tissues Although the majority of the benign epithelial cells and adenocarcinoma cells demonstrated negative chondroitin-4-sulphate staining, comparing the means of 100 cases of matched benign and prostate adenocarcinoma tissues revealed that cancer cells were more intensely stained compared against benign secretory cells (P=0.003). Furthermore, an average of 23.5% of cancer cells were positively stained for chondroitin-4-sulphate, in contrast to an average of 10.5% of secretory cells in the benign prostatic group (P=0.0002). The immunoreactivity score (IRS) of the prostate adenocarcinoma cells were also significantly higher (mean of 35.9) than the benign epithelial cells (mean of 19.8). (Figure 4.5 and Table 4.6) In addition, chondroitin-4-sulphate was found to be downregulated in the peritumoural stroma tissue compared to the periglandular stroma region. The mean staining intensity of the periglandular stroma was 1.61 while the mean staining intensity of the peritumoural stroma was 1.27 (P=0.002). There was no significant difference in the percentage of periglandular cells and the peritumoural cells positively stained. Page | 123 Chapter 4 Chondroitin sulphate in prostate tissue Figure 4.5 Comparisons of chondroitin-4-sulphate expression in prostate adenocarcinoma samples (n=100) against matched BPH samples (n=100). [A] Weighted Average Intensity [B] Percentage of cells positively stained [C] Immunoreactivity Score ‘*’ represents p<0.05; ‘**’ represents p<0.01; ‘***’ represents p<0.001. P-values were determined by Paired T-test. Page | 124 Chapter 4 Chondroitin sulphate in prostate tissue Table 4.6 Mean values of chondroitin-4-sulphate staining in prostate adenocarcinoma samples (n=100) versus paired BPH samples (n=100). P- values were determined by Paired T-test. Weighted Average Intensity Percentage of cells stained (%) IRS Mean SEM P-value Mean SEM p-value Mean SEM P-value Epithelial / Cancer cell region Benign secretory epithelial 0.40 0.08 0.003 10.50 2.50 0.0002 19.78 5.38 0.018 cells Adenocarcinoma cells 0.68 0.08 23.53 3.47 35.87 6.00 Stromal region Periglandular stromal cells 1.61 0.08 0.002 62.20 3.83 0.229 104.30 7.79 0.049 Peritumoural stromal cells 1.27 0.08 57.38 4.07 87.33 8.04 Page | 125 Chapter 4 Chondroitin sulphate in prostate tissue 4.1.3.3 Comparison of chondroitin-4-sulphate expression in different grades of prostate adenocarcinoma tissues Following this, the expression levels of chondroitin-4-sulphate in both the cytoplasm of the prostate adenocarcinoma cells and the peritumoural stroma were associated with different tumour grades of prostate adenocarcinoma. Due to technical difficulties encountered as a result of different batches of antibodies employed for use for the prostate TMAs (Gleason sum 6 and 7) and the core biopsies (Gleason sum 3 to 5 and 8 to 10), only the 2 extreme grades of prostate adenocarcinoma cases were compared (i.e Gleason sum 3 to 5 versus Gleason sum 8 to 10) since immunohistochemical staining was performed at the same time and using the same batch of antibodies for this set of tissues. As seen from Figure 4.6, there was very weak mean staining in the cytoplasm of the adenocarcinoma tissues of Gleason sum 3 to 5 and 8 to 10, and there was no significant difference between the two extreme grades. There was also a very small percentage of adenocarcinoma cells positively stained for chondroitin-4-sulphate in both Gleason sum 3 to 5 and Gleason sum 8 to 10 groups, and there was no significant difference between the two extreme grades either. Conversely, there was stronger staining in the peritumoural stroma than the cytoplasmic adenocarcinoma cells; in the Gleason sum 3 to 5 group, the mean staining in the peritumoural stroma was 0.95 and this mean was increased to 1.65 in the Gleason sum 8 to 10 group (P=0.003). No significant difference was observed in the percentage of peritumoural stroma cells positively stained between the Gleason sum 3 to 5 group and Gleason sum 8 to 10 group (Table 4.7). Page | 126 Chapter 4 Chondroitin sulphate in prostate tissue Figure 4.6 Comparisons of chondroitin-4-sulphate expression in different grades of prostate adenocarcinoma tissues: Gleason sum 3, 4 & 5 (n=26) vs Gleason sum 8,9 & 10 (n=51), in both [A,B,C] cytoplasm of cancer cell and [D,E,F] peritumoural stroma. [A,D] Weighted Average Intensity [B,E] Percentage of cells positively stained [C,F] Immunoreactivity score. ‘**’ represents p<0.01; P-values were determined by One- way ANOVA with Bonferroni’s Multiple Comparison Post test. Page | 127 Chapter 4 Chondroitin sulphate in prostate tissue Table 4.7 Mean values of chondroitin-4-sulphate staining in different grades of prostate adenocarcinoma tissues: Gleason sum 3, 4 & 5 (n=26) vs Gleason sum 8,9 & 10 (n=51). P-values were determined by One-way ANOVA with Bonferroni’s Mulitple Comparison Post Test. Weighted Average Intensity Percentage of cells positively stained Immunoreactivity Score Gleason Sum Mean SEM P-value Mean SEM P-value Mean SEM P-value Adenocarcinoma 3,4 & 5 0.12 0.06 0.829 5.00 3.89 0.925 5.00 3.89 0.461 8, 9 & 10 0.14 0.08 5.49 3.12 13..33 7.75 Peritumoural Stroma 3,4 & 5 0.95 0.19 0.003 30.77 7.36 0.469 51.92 14.14 0.388 8, 9 & 10 1.65 0.13 37.00 4.83 66.35 9.44 Page | 128 Chapter 4 Chondroitin sulphate in prostate tissue 4.1.4 Expression analysis of chondroitin-6-sulphate in prostate adenocarcinoma. 4.1.4.1 Localization of chondroitin-6-sulphate in prostate tissues Expression analysis of chondroitin-6-sulphate was examined in a total of 120 cases of benign prostatic hyperplastic tissues and 128 cases of prostate adenocarcinoma cases, in which 89 of these cases are matched cases (i.e both BPH samples and prostate adenocarcinoma tissues were taken from the same patient). In benign prostatic tissue, weak staining intensity was detected in the cytoplasm of secretory cells and in the periglandular stroma region. In prostate cancer, weak immunohistochemical staining for chondroitin-6-sulphate was seen in the cytoplasm of malignant cells while strong expression of chondroitin-6-sulphate was observed in the surrounding peritumoural stroma (Figure 4.7). The expression data of chondroitin-6-sulphate in the 120 cases of benign prostatic tissues and 128 cases of prostate adenocarcinoma samples are shown in Table 4.8. 4.1.4.2 Comparison of chondroitin-6-sulphate expression in benign prostatic and adenocarcinoma tissues Staining differences in the means of 89 cases of matched benign and prostate adenocarcinoma tissues were compared. It was found that a higher percentage of peritumoural stroma cells (approximate mean of 60.47%) were stained for chondroitin-6-sulphate compared against periglandular stroma cells (approximate mean of 46.85%) (P=0.002). (Figure 4.8) The immunoreactivity score (IRS) of the peritumoural stroma cells were also significantly higher (mean of 89.89) than the benign hyperplastic prostates (mean of 57.99). However, there were no significant changes in staining intensity or percentage of cells stained in the adenocarcinoma cells, as compared to the benign epithelial cells (Figure 4.8 and Table 4.9). Page | 129 Chapter 4 Chondroitin sulphate in prostate tissue Figure 4.7 Chondroitin-6-sulphate expression in BPH tissue and prostate adenocarcinoma sections. [A] BPH [B] Gleason sum 4 section [C] Gleason sum 7 section [D] Gleason sum 9 section [E] Negative control ‘BL’ represents basal cell layer, ‘Epi’ represents benign epithelial cell, ‘CaP’ represents cancer cells and ‘SM’ represents stromal cells. Page | 130 Chapter 4 Chondroitin sulphate in prostate tissue Table 4.8 Expression analysis of chondroitin-6-sulphate in benign hyperplastic prostate tissues and prostate adenocarcinoma tissues Number of cases stained Number Weighted Average Intensity Percentage of Stained Tissue Total Immunoreactivity Score (IRS) of cases (WAI) cells stained (%) Component number unstained ≤1 1 BPH Secretory 120 13 72 21 14 37 70 86 15 6 epithelial cells 10.83% 60.00% 17.50% 11.67% 30.83% 58.33% 71.67% 12.50% 5.00% Periglandular 120 22 64 15 19 61 37 83 12 3 stromal cells 18.33% 53.33% 12.50% 15.83% 50.83% 30.83% 69.17% 10.00% 2.50% Adenocarcinoma Cancerous cells 128 46 44 23 15 27 55 55 22 5 35.94% 34.38% 17.97% 11.72% 21.09% 42.97% 42.97% 17.19% 3.91% Peritumoural 128 10 49 21 47 43 89 63 26 28 stromal cells 7.81% 38.28% 16.41% 36.72% 33.59% 69.53% 49.22% 20.31% 21.88% Page | 131 Chapter 4 Chondroitin sulphate in prostate tissue Figure 4.8 Comparisons of chondroitin-6-sulphate expression in prostate adenocarcinoma samples (n=89) against BPH samples (n=89). [A] Weighted Average Intensity [B] Percentage of cells positively stained [C] Immunoreactivity Score ‘**’ represents p<0.01; ‘***’ represents p<0.001. P-values were determined by Paired T-test. Page | 132 Chapter 4 Chondroitin sulphate in prostate tissue Table 4.9 Mean values of chondroitin-6-sulphate staining in prostate adenocarcinoma samples (n=89) versus paired BPH samples (n=89). P-values were determined by Paired T-test. Weighted Average Intensity Percentage of cells stained (%) IRS Mean SEM P-value Mean SEM p-value Mean SEM P-value Epithelial / Cancer cell region Benign secretory epithelial 1.07 0.06 0.585 67.36 4.27 0.087 80.18 5.69 0.937 cells Adenocarcinoma cells 1.12 0.08 59.16 4.43 79.57 6.95 Stromal region Periglandular stromal cells 1.06 0.07 0.177 46.85 3.98 0.002 57.99 5.91 <0.0001 Peritumoural stromal cells 2.00 0.68 60.47 4.00 89.89 7.98 Page | 133 Chapter 4 Chondroitin sulphate in prostate tissue 4.1.4.3 Comparison of chondroitin-6-sulphate expression in different grades of prostate adenocarcinoma tissues The expression levels of chondroitin-6-sulphate in both the cytoplasm of the prostate adenocarcinoma cells and the peritumoural stroma are also associated with different tumour grades of prostate adenocarcinoma. The tumour grades were classified into 3 groups: Gleason sum ≤5, Gleason sum 6 and 7 and Gleason sum ≥8 and the means of chondroitin sulphate staining of each group were compared. With increasing tumour grade, there was increasing chondroitin-6-sulphate expression in the peritumoural stroma (P<0.0001) but decreasing chondroitin-6- sulphate expression in the cytoplasm of the adenocarcinoma cells (P<0.0001). (Figure 4.9) The mean values of non-sulphated chondroitin staining in the various groups of tumour grade are shown in Table 4.10. One-way ANOVA with Bonferroni’s post-test was performed to determine statistical significance. These results suggest a potential role of chondroitin-6-sulphate involvement in prostate cancer progression and warrants further investigation. Page | 134 Chapter 4 Chondroitin sulphate in prostate tissue Figure 4.9 Comparisons of chondroitin-6-sulphate expression in different grades of prostate adenocarcinoma tissues: Gleason sum 3,4 & 5 (n=19) vs Gleason sum 6 & 7 (n=65) vs Gleason sum 8,9 & 10 (n=44), in both [A,B,C] cytoplasm of cancer cell and [D,E,F] peritumoural stroma. [A,D] Weighted Average Intensity [B,E] Percentage of cells positively stained [C,F] Immunoreactivity score. ‘**’ represents p<0.01; ‘***’ represents p<0.001. P-values were determined by One-way ANOVA with Bonferroni’s Multiple Comparison Post test. Page | 135 Chapter 4 Chondroitin sulphate in prostate tissue Table 4.10 Mean values of chondroitin-6-sulphate staining in different grades of prostate adenocarcinoma tissues: Gleason sum 3, 4 & 5 (n=19) vs Gleason sum 6 & 7 (n= 65) vs Gleason sum 8, 9 & 10 (n=44). P-values were determined by One-way ANOVA with Bonferroni’s Mulitple Comparison Post Test. Weighted Average Intensity Percentage of cells positively stained Immunoreactivity Score Gleason Sum Mean SEM P-value Mean SEM P-value Mean SEM P-value Adenocarcinoma 3,4 & 5 1.48 0.28 <0.0001 52.11 9.66 <0.0001 99.47 21.78 <0.0001 6 & 7 1.11 0.05 69.38 4.67 83.05 6.41 8, 9 & 10 0.32 0.12 9.66 3.96 16.44 6.21 Peritumoural Stroma 3,4 & 5 1.16 0.24 <0.0001 32.11 7.59 <0.0001 47.37 10.78 <0.0001 6 & 7 1.18 0.07 65.41 4.55 85.56 8.30 8, 9 & 10 2.35 0.10 81.16 4.49 194.90 13.28 Page | 136 Chapter 4 Chondroitin sulphate in prostate tissue 4.1.5 Expression analysis of chondroitin-2,6-sulphate in prostate adenocarcinoma 4.1.5.1 Localization of chondroitin-2,6-sulphate in prostate tissues Expression analysis of the chondroitin-2,6-sulphate was examined in a total of 115 cases of benign prostatic hyperplastic tissues and 142 cases of prostate adenocarcinoma cases, in which 95 of these cases are matched cases (i.e both BPH samples and prostate adenocarcinoma tissues were taken from the same patient). Approximately 50.43% of the benign cases did not show chondroitin-2,6- sulphate staining in the benign epithelial cells; while the other 49.57% of the cases demonstrated chondroitin-2,6-sulphate in the cytoplasm of the benign epithelial cells, the intensity of staining varied from weak to strong staining. Chondroitin-2,-6- sulphate was observed to be expressed in the periglandular stroma in 95.65% of the cases, with 77.39% of the cases demonstrating moderate to strong staining. 74.78% of the cases had positive chondroitin-2,6-sulphate staining in more than 50% of the periglandular stroma cells as well. In prostate adenocarcinoma tissues, we observed a similar staining pattern for chondroitin-2,6-sulphate. Approximately 58.45% of the cases had no staining in the cytoplasm of the prostate adenocarcinoma cells, while those that showed cytoplasmic staining was in the range of weak to moderate. About 61.97% of the cases revealed strong chondroitin-2,6-sulphate staining in the peritumoural stroma, with 86.62% of the cases showing positive staining in more than 50% of the peritumoural stroma cells. (Figure 4.10 and Table 4.11) Page | 137 Chapter 4 Chondroitin sulphate in prostate tissue Figure 4.10 Chondroitin-2,6-sulphate expression in BPH and prostate adenocarcinoma tissues.[A] BPH [B] Gleason sum 4 section [C] Gleason sum 8 section [D] Gleason sum 9 section [E] Negative control ‘BL’ represents basal cell layer, ‘Epi’ represents benign epithelial cell, ‘CaP’ represents cancer cells and ‘SM’ represents stromal cells. Page | 138 Chapter 4 Chondroitin sulphate in prostate tissue Table 4.11 Expression Analysis of chondroitin-2,6-sulphate in benign hyperplastic prostate tissues and prostate adenocarcinoma tissues Number Number of cases stained Stained Tissue Total of cases Weighted Average Intensity Percentage of Component number Immunoreactivity Score (IRS) unstained (WAI) cells stained (%) ≤1 1 Adenocarcinoma Cancerous cells 142 83 26 4 29 34 25 39 15 5 58.45% 18.31% 2.82% 20.42% 23.94% 17.61% 27.46% 10.56% 3.52% Peritumoural 142 2 17 35 88 17 123 31 29 80 Stromal cells 1.41% 11.97% 24.65% 61.97% 11.97% 86.62% 21.83% 20.42% 56.34% Page | 139 Chapter 4 Chondroitin sulphate in prostate tissue 4.1.5.2 Comparison of chondroitin-2,6-sulphate expression in benign prostatic and adenocarcinoma tissues The expression levels of chondroitin-2,6-sulphate in 95 cases of matched benign prostate hyperplastic tissues and prostate adenocarcinoma tissues were compared. There were no significant difference in terms of staining intensity, percentage of cells positively stained and immunoreactivity score in the adenocarcinoma cells and in the peritumoural stroma (Figure 4.11). However, if the expression levels of chondroitin-2,6-sulphate in matched benign and adenocarcinoma were compared based on the Gleason sum of the adenocarcinoma tissue, we observed that adenocarcinoma tumours of Gleason sum 6 and 7 had increased expression of chondroitin-2,6-sulphate in the cytoplasmic adenocarcinoma cells as compared to their matched benign epithelial cells (n=60). However, decreased expression of chondroitin-2,6-sulphate in the cytoplasmic adenocarcinoma cells as compared to the matched benign epithelial cells was observed in Gleason sum 3 to 5 tumours (n=16) and Gleason sum 8 to 10 tumours (n=18), albeit without statistical significance. None of the prostate adenocarcinoma cells with Gleason sum 8, 9 and 10 were stained positively for chondroitin-2,6-sulphate. Therefore, to analyze the difference in chondroitin-2,6-sulphate expression in all the adenocarcinoma tissues collectively as a whole, regardless of Gleason sum as compared to their matched benign tissues, results in insignificant difference in terms of staining intensity, percentage of cells positively stained and immunoreactivity score. A trend of increased chondroitin-2,6-sulphate expression in the peritumoural stroma as compared to matched periglandular stroma was revealed, although this did not reach statistical significance of p<0.05; the mean IRS of periglandular stroma Page | 140 Chapter 4 Chondroitin sulphate in prostate tissue staining is 159.80 while the mean IRS of peritumoural stroma is 180.20 (P=0.07) (Table 4.12). Page | 141 Chapter 4 Chondroitin sulphate in prostate tissue Figure 4.11 Comparisons of chondroitin-2,6-sulphate expression in prostate adenocarcinoma samples (n=95) against paired BPH samples (n=95). [A,D] Weighted Average Intensity [B,C] Percentage of cells positively stained [C,F] Immunoreactivity Score [D,E,F] Tumours in different Gleason sum group were compared against their matched BPH samples, giving rise to various differential expression patterns of chondroitin-2,6-sulphate. No expression of chondroitin-2,6-sulphate were detected in all prostate adenocarcinoma cells with Gleason sum >8. ‘*’ represents p<0.05; P-value was determined using Paired T-test. Page | 142 Chapter 4 Chondroitin sulphate in prostate tissue Table 4.12 Mean values of chondroitin-2,6-sulphate staining in prostate adenocarcinoma samples (n=95) versus paired BPH samples (n=95). P-values were determined by Paired T-test. Weighted Average Intensity Percentage of cells stained IRS Mean SEM P-value Mean SEM p-value Mean SEM P-value Epithelial / Cancer cell region Benign secretory epithelial cells 0.78 0.10 0.34 22.87 3.27 0.58 39.99 6.41 0.41 Adenocarcinoma cells 0.90 0.11 25.32 3.55 47.03 7.51 Stromal region Periglandular stromal cells 1.83 0.08 0.11 81.73 2.99 0.07 159.80 9.23 0.07 Peritumoural stromal cells 1.99 0.07 88.26 2.62 180.20 8.95 Page | 143 Chapter 4 Chondroitin sulphate in prostate tissue 4.1.5.3 Comparison of chondroitin-2,6-sulphate expression in different grades of prostate adenocarcinoma tissues Subsequently, the expression levels of chondroitin-2,6-sulphate in both the cytoplasm of the prostate adenocarcinoma cells and the peritumoural stroma were associated with different tumour grades of prostate adenocarcinoma. We have classified the tumour grades into 3 groups: Gleason sum ≤5, Gleason sum 6 and 7 and Gleason sum ≥8 and compared the means of chondroitin-2,6-sulphate staining of each group (Figure 4.12). Interestingly, we observed that the expression of chondroitin-2,6-sulphate was highest in adenocarcinoma cells of Gleason sum 6 and 7 (mean IRS = 62.58) while tumours of Gleason sum 8 to 10 had the lowest expression of chondroitin-2,6-sulphate in the cytoplasm of the adenocarcinoma cells (mean IRS = 11.18) (P<0.0001). In the peritumoural stroma region, it was shown that chondroitin-2,6-sulphate staining was increased with increasing tumour grades, with the highest expression in Gleason sum 8 to 10 tumours (Table 4.13). Page | 144 Chapter 4 Chondroitin sulphate in prostate tissue Figure 4.12 Comparisons of chondroitin-2,6-sulphate expression in different grades of prostate adenocarcinoma tissues: Gleason sum 3,4 & 5 (n=22) vs Gleason sum 6 & 7 (n=69) vs Gleason sum 8,9 & 10 (n=51), in both [A,B,C] cytoplasm of cancer cell and [D,E,F] peritumoural stroma. [A,D] Weighted Average Intensity [B,E] Percentage of cells positively stained [C,F] Immunoreactivity score. ‘*’ represents p<0.05; ‘**’ represents p<0.01; ‘***’ represents p<0.001. P-values were determined by One-way ANOVA with Bonferroni’s Multiple Comparison Post test Page | 145 Chapter 4 Chondroitin sulphate in prostate tissue Table 4.13 Mean values of chondroitin-2,6-sulphate staining in different grades of prostate adenocarcinoma tissues: Gleason sum 3, 4 & 5 (n=22) vs Gleason sum 6 & 7 (n= 69) vs Gleason sum 8, 9 & 10 (n=51). P-values were determined by One-way ANOVA with Bonferroni’s Multiple Comparison Post Test. Gleason Sum Weighted Average Intensity Percentage of cells positively stained Immunoreactivity Score Mean SEM P-value Mean SEM P-value Mean SEM P-value Adenocarcinoma 3,4 & 5 0.55 0.17 <0.0001 16.36 6.36 0.0028 23.18 10.64 <0.0001 6 & 7 1.23 0.13 33.70 4.23 62.58 8.93 8, 9 & 10 0.12 0.07 5.39 3.06 11.18 7.03 Peritumoural Stroma 3,4 & 5 1.70 0.18 0.001 71.59 8.56 0.001 141.8 22.05 0.0003 6 & 7 1.91 0.09 90.07 2.88 174.6 9.93 8, 9 & 10 2.31 0.09 94.90 2.28 223.1 10.56 Page | 146 Chapter 4 Chondroitin sulphate in prostate tissue 4.1.6 Comparisons of chondroitin sulphate staining amongst each subtype Correlations in the immunohistochemical staining amongst the different chondroitin sulphate subtypes in the adenocarcinoma tissues were determined. Association studies were performed using the Fisher’s Exact Test and it was revealed that there were significant associations between different chondroitin sulphate subtype staining; both in the cytoplasm of adenocarcinoma cells and in the peritumoural stroma. From Table 4.14, it can be seen that higher non-sulphated chondroitin staining in the cytoplasmic adenocarcinoma cells is correlated with higher chondroitin-4-sulphate expression, chondroitin-6-sulphate expression and chondroitin-2,6-sulphate expression in the cytoplasm of the adenocarcinoma cells. Similarly, higher chondroitin-2,6-sulphate expression is associated with higher chondroitin-4-sulphate and chondroitin-6-sulphate expression in the adenocarcinoma cells; the same association is observed for chondroitin-4-sulphate and chondroitin-6- sulphate expression too. Likewise, different chondroitin sulphate subtypes staining in the peritumoural stroma were associated with one another (Table 4.15). With a higher expression of non-sulphated chondroitin in the peritumoural stroma, chondroitin-4-sulphate, chondroitin-6-sulphate and chondroitin-2,6-sulphate were also highly expressed. Chondroitin-2,6-sulphate expression was also associated with chondroitin-4-sulphate and chondroitin-6-sulphate expression in the peritumoural stroma. Although there was no statistical significance, it can be seen that high levels of chondroitin-6-sulphate in the peritumoural stroma is correlated with high chondroitin-4-sulphate expression. Page | 147 Chapter 4 Chondroitin sulphate in prostate tissue Table 4.14 Comparisons of chondroitin sulphate staining amongst each subtype in the adenocarcinoma cells Adenocarcinoma Cell Chondroitin-4-sulphate Chondroitin-6-sulphate Chondroitin-2,6-sulphate IRS ≤ IRS > p- IRS IRS p- IRS IRS p- Total Total Total 100 100 value ≤ 100 > 100 value ≤ 100 > 100 value Non-sulphated chondroitin IRS ≤ 100 104 100 4 0.00 92 84 8 0.00 101 94 7 0.00 100.0% 96.2% 3.8% 100.0% 91.3% 8.7% 100.0% 93.1% 6.9% IRS > 100 33 24 9 33 14 19 33 21 12 100.0% 72.7% 27.3% 100.0% 42.4% 57.6% 100.0% 63.6% 36.4% Chondroitin-2,6-sulphate IRS ≤ 100 117 114 3 0.00 105 88 17 0.001 - - - - 100.0% 97.4% 2.6% 100.0% 83.8% 16.2% - - - IRS > 100 19 9 10 18 8 10 - - - 100.0% 47.4% 52.6% 100.0% 44.4% 55.6% - - - Chondroitin-6-sulphate IRS ≤ 100 99 93 6 0.006 ------100.0% 93.94% 6.06% ------IRS > 100 26 19 7 ------100.0% 73.08% 26.92% ------ Page | 148 Chapter 4 Chondroitin sulphate in prostate tissue Table 4.15 Comparisons of chondroitin sulphate staining amongst each subtype in the peritumoural stroma Peritumoural Stroma Chondroitin-4-sulphate Chondroitin-6-sulphate Chondroitin-2,6-sulphate IRS IRS p- IRS IRS p- IRS IRS p- Total Total Total ≤ 100 > 100 value ≤ 100 > 100 value ≤ 100 > 100 value Non-sulphated chondroitin IRS ≤ 100 95 78 17 0.005 87 65 22 0.00 92 28 64 0.003 100.0% 82.1% 17.9% 100.0% 74.7% 25.3% 100.0% 30.4% 69.6% IRS > 100 41 24 17 38 7 31 41 3 38 100.0% 58.5% 41.5% 100.0% 18.4% 81.6% 100.0% 7.3% 92.7% Chondroitin-2,6-sulphate IRS ≤ 100 31 30 1 0.00 29 25 4 0.00 - - - - 100.0% 96.8% 3.2% 100.0% 86.2% 13.8% - - - IRS > 100 105 70 35 93 45 48 - - - 100.0% 66.7% 33.3% 100.0% 48.4% 51.6% - - - Chondroitin-6-sulphate IRS ≤ 100 71 56 15 0.214 ------100.0% 78.9% 21.1% ------IRS > 100 53 36 17 ------100.0% 67.9% 32.1% ------ Page | 149 Chapter 4 Chondroitin sulphate in prostate tissue 4.1.7 Association of chondroitin sulphate staining with clinicopathological parameters Most pathological parameters such as pathological T stage, extraprostatic extension, involvement of seminal vesicles, tumour size and volume, resection margin including distal and proximal margin involvement, obturator lymph node metastasis, presence of tumour in single/both prostate lobes are only available when patients undergo prostatectomies. Gleason grade and sum, lymphovascular invasion, perineural invasion, pre-operative PSA and clinical staging are other parameters that will be available for all core biopsies and prostatectomies cases. In our cohort of study, we have clinicopathological data for most prostatectomy cases (corresponding to Gleason sum 6 to 8 tumours) but only Gleason sum, presence of lymphovascular invasion and perineural invasion data for the biopsy cases (corresponding to Gleason sum 3 to 5 and 8 to 10). (refer to Table 4.1) Also, due to loss of some tissues during immunohistochemical processing and the unavailability of some clinicopathological data for some cases, the actual number of cases studied could vary and are shown in the respective tables. Analysis of absolute staining intensities for non-sulphated chondroitin, chondroitin-4-sulphate, chondroitin-6-sulphate and chondroitin-2,6-sulphate in both prostate adenocarcinoma cells and peritumoural stroma showed significant associations with several clinicopathological parameters and known prognostic factors. Table 4.16 shows that pathological T stage, extraprostatic extension, involvement of seminal vesicles, and pre-operative PSA levels were all correlated with strong expression of non-sulphated chondroitin in the peritumoural stroma. A Page | 150 Chapter 4 Chondroitin sulphate in prostate tissue strong expression of non-sulphated chondroitin in adenocarcinoma cells was also associated with involvement of both prostatic lobes. Chondroitin-4-sulphate staining in the cytoplasm of the adenocarcinoma cells was also associated with pathological T stage, extraprostatic extension and clinical staging. Similarly, a strong expression of chondroitin-2,6-sulphate in the cytoplasm of the adenocarcinoma cell and in the peritumoural stroma correlated with perineural invasion and seminal vesicle involvement respectively. Moreover, examination of the percentage of tumour cells stained positively for non-sulphated chondroitin revealed correlations with pathological T stage and extraprostatic extension. When a higher percentage of peritumoural stroma cells are stained positively for chondroitin-6-sulphate, this was associated with pathological T stage and extraprostatic extension too. In a similar manner, high percentage of adenocarcinoma cells and peritumoural stroma cells stained positively for chondroitin-2,6-sulphate was correlated with clinical staging and tumour volume respectively (Table 4.17). No significant correlations were found when the expression patterns of non- sulphated chondroitin, chondroitin-4-sulphate, chondroitin-6-sulphate or chondroitin- 2,6-sulphate were analyzed with the other clinicopathological parameters listed in Table 4.1 (P >0.05), namely resection margin including distal and proximal margin involvement and lymphovascular invasion. Results of all association studies performed to determine correlation between chondroitin sulphate staining and all clinicopathological parameters (regardless of statistical significance) are shown in Appendix II. No correlation with obturator lymph node involvement was performed as none of the patients demonstrated metastases to nodes. Page | 151 Chapter 4 Chondroitin sulphate in prostate tissue Table 4.16 Association studies of chondroitin sulphate staining intensity with clinicopathological parameters. Epitope Stained tissue component Clinicopathological Total Weighted average Weighted average P-value parameter Intensity<2 Intensity ≥2 Non-sulphated chondroitin Adenocarcinoma cells Lobe occurrence Single lobe 11 11 (100.0%) 0 (0.0%) 0.013 Both lobes 64 38 (59.4%) 26 (40.6%) Total 75 Peritumoural stroma Pathological T staging T2a to T2c 39 34 (87.2%) 5 (12.8%) 0.049 T3a to T3b 33 22 (66.7%) 11(33.3%) Total 72 Extraprostatic extension Yes 33 22 (66.7%) 11 (33.3%) 0.049 No 39 34 (87.2%) 5 (12.8%) Total 72 Seminal vesicle involvement Involved 7 3 (42.9%) 4 (57.1%) 0.034 Not involved 68 56 (82.4%) 12 (17.6%) Total 75 Page | 152 Chapter 4 Chondroitin sulphate in prostate tissue Non-sulphated chondroitin Peritumoural stroma Pre-op PSA levels ≤ 8ng/ml 21 21 (100.0%) 0 (0.0%) 0.021 > 8ng/ml 20 15 (75.0%) 5 (25.0%) Total 41 Chondroitin-4-sulphate Adenocarcinoma cells Extraprostatic extension Yes 35 24 (68.6%) 11 (31.4%) 0.041 No 38 34 (89.5%) 4 (10.5%) Total 73 Pathological T staging T2a to T2c 38 34 (89.5%) 4 (10.5%) 0.041 T3a to T3b 35 24 (68.6%) 11 (31.4%) Total 73 Clinical staging T1a to T1c 36 32 (88.9%) 4 (11.1%) 0.015 T2a to T4 7 3 (42.9%) 4 (57.1%) Total 43 Chondroitin-2,6-sulphate Adenocarcinoma cells Perineural invasion Seen 89 65 (73.0%) 24 (27.0%) 0.017 Not Seen 52 47 (90.4%) 5 (9.6%) Total 141 Page | 153 Chapter 4 Chondroitin sulphate in prostate tissue Peritumoural stroma Seminal vesicle involvement Involved 7 0 (0%) 7 (100.0%) 0.018 Not involved 72 34 (47.2%) 38 (52.8%) Total 79 Page | 154 Chapter 4 Chondroitin sulphate in prostate tissue Table 4.17 Association studies of percentage of cells positively stained for different chondroitin sulphate with clinicopathological parameters. Antigen Stained tissue component Clinicopathological Total % of positively % of positively P-value parameter stained cells ≤50% stained cells >50% Non-sulphated chondroitin Adenocarcinoma cells Pathological T stage T2a to T2c 40 30 (75.0%) 10 (25.0%) 0.05 T3a to T3b 33 17 (51.5%) 16 (48.5%) Total 73 Extraprostatic extension Yes 33 17 (51.5%) 16 (48.5%) 0.05 No 40 30 (75.0%) 10 (25.0%) Total 73 Chondroitin-6-sulphate Peritumoural stroma Extraprostatic extension Yes 34 9 (26.5%) 25 (73.5%) 0.004 No 35 22 (62.9%) 13 (37.1%) Total 69 Pathological T staging T2a to T2c 35 22 (62.9%) 13 (37.1%) 0.004 T3a to T3b 34 9 (26.5%) 25 (73.5%) Total 69 Page | 155 Chapter 4 Chondroitin sulphate in prostate tissue Chondroitin-2,6-sulphate Adenocarcinoma cells Clinical staging T1a to T1c 37 25 (67.6%) 12 (32.4%) 0.045 T2a to T4 8 2 (25.0%) 6 (75.0%) Total 45 Peritumoural stroma Tumour volume <4.00 cm3 28 7 (25.0%) 21 (75.0%) 0.025 ≥4.00 cm3 29 1 (3.4%) 28 (96.6%) Total 57 Page | 156 Chapter 4 Chondroitin sulphate in prostate tissue 4.2 Gene transcript analysis of chondroitin-sulphate biosynthesis and editing genes in prostate cancer cell lines Upregulation in the CS/DS biosynthetic and editing enzymes and the CSPG/DSPG was observed from the genome wide expression analysis in Chapter 3. This was further supported by immunohistochemical staining of the prostate tissue sections in which an increased expression of various chondroitin sulphate subtypes in prostate adenocarcinoma tissues was seen. Next, to employ a good in vitro model for use to investigate some of the functions of these differentially expressed genes, the expression of the CS/DS biosynthetic genes and some CSPGs in 3 prostate cell lines, RWPE-1, a normal epithelial prostate cell line and LNCaP and PC-3, both of which are prostate adenocarcinoma cell lines, were quantified. Both cancer cell lines were derived from metastatic sites: LNCaP is from the supraclavicular lymph node, while PC-3 is from bone metastasis. Table 4.18 shows the expression fold changes of CS/DS and CSPG/DSPG genes in LNCaP and PC-3 as compared to RWPE-1, determined from real-time PCR and from microarray study. It can be observed that many of these genes are either upregulated or downregulated in the same manner in both cancer cell lines; however, some genes are not found to be differentially expressed in the same manner in both cancer cell lines. There is about 10 fold upregulation of CSGalNAcT-1 in LNCaP while in PC-3, CSGalNAcT-1 is upregulated by about 20 fold. Microarray results revealed an upregulation of CSGalNacT-1 gene by 2 fold in the prostate adenocarcinoma tissues. This was further verified to be upregulated by 4.3 fold through real-time RT-PCR. CSGalNAcT-1 was reported to be responsible for the initiation of CS/DS biosynthesis after the formation of the tetrasaccharide linkage region xylose-galactose-galactose- Page | 157 Chapter 4 Chondroitin sulphate in prostate tissue glucuronic acid. CSGalNAcT-2, which shares similar homology to CSGalNAcT-1 and is reported to be involved in the elongation of the CS/DS chain, was not differently expressed in prostate adenocarcinoma tissues compared to BPH tissues. We hypothesized that the upregulation of CSGalNAcT-1 could result in increased expression of the chondroitin sulphate chains in prostate adenocarcinoma and could potentially contribute to cancer cell behaviour. Similarly, we do not observe a significant change in CSGalNAcT-2 expression in PC-3 but there was a downregulation of about 5 fold of CSGalNAcT-2 in LNCaP. PC-3 was thus used to further investigate the functions of CSGalNAcT-1 as its expression of CSGalNAcT-1 and CSGalNAcT-2 was more similar to the data obtained from the microarray study. Page | 158 Chapter 4 Chondroitin sulphate in prostate tissue Table 4.18 Expression fold changes of CS/DS biosynthesis and editing enzymes and CSPG/DSPG genes in 1) LNCaP (n=3) and PC-3 (n=3) as compared to RWPE-1 (n=3), determined by qPCR and 2) in prostate adenocarcinoma tissues (n=6) compared to BPH tissues (n=6), obtained from microarray study. P-values of qPCR were obtained using unpaired T-test. Relative fold changes and p-value shown here for microarray study were determined by dCHIP. Gene Gene name Prostate Cancer Cell Lines Prostate Adenocarcinoma tissues LNCaP PC-3 Microarray Study Relative Fold p-value Relative Fold p-value Relative Fold p-value Change Change Change XylTI xylosyltransferase I N.D <0.05 N.D <0.05 -1.5 >0.05 XylTII xylosyltransferase II 1.40 0.39 1.65 0.26 -1.3 >0.05 B4GalT7 xylosylprotein beta 1,4- 2.24 0.01 2.65 0.00 1.0 >0.05 galactosyltransferase, polypeptide 7 (galactosyltransferase I) B3GalT6 UDP-Gal:betaGal beta 1,3- 2.94 0.06 3.36 0.08 2.3 <0.05 galactosyltransferase polypeptide 6 B3GAT3 beta-1,3-glucuronyltransferase 3 5.10 0.00 2.61 0.03 1.0 >0.05 (glucuronosyltransferase I) CSGalNAcT-1 chondroitin sulphate N- 10.36 0.0009 20.49 0.0002 2.0 <0.05 acetylgalactosaminyltransferase 1 CSGalNAcT-2 chondroitin sulphate N- -5.23 0.01 1.11 0.88 1.2 >0.05 acetylgalactosaminyltransferase 2 CSGLCA-T chondroitin sulphate 3.17 0.00 2.45 0.04 1.3 >0.05 glucuronyltransferase CHPF chondroitin polymerising factor 2.18 0.21 1.15 0.83 1.3 >0.05 Page | 159 Chapter 4 Chondroitin sulphate in prostate tissue CHSY1 chondroitin synthase I -1.75 0.26 -3.90 0.041 1.0 >0.05 CHSY3 chondroitin synthase 3 N.D <0.05 3.30 0.00 -1.3 >0.05 SART2 (DSE) dermatan sulphate epimerase -1.6 >0.05 CHST13 carbohydrate (chondroitin 4) 263.8 0.0009 8.13 0.002 -1.0 >0.05 sulphotransferase 13 CHST11 carbohydrate (chondroitin 4) -4.17 0.331 -1.05 0.64 -1.3 >0.05 sulphotransferase 11 CHST12 carbohydrate (chondroitin 4) 4.99 0.0013 1.12 0.70 -1.5 >0.05 sulphotransferase 12 CHST7 carbohydrate (N-acetylglucosamine 6- -1.53 0.25 -2.99 0.06 -1.4 >0.05 O) sulphotransferase 7 CHST3 carbohydrate (chondroitin 6) -4.88 0.02 -6.72 0.001 1.1 >0.05 sulphotransferase 3 GALNAC4S-6ST B cell RAG associated protein 0.78 0.24 0.69 0.27 -4.0 <0.05 CHST14 (D4ST1) carbohydrate (N-acetylgalactosamine 1.31 0.36 1.95 0.18 -1.3 >0.05 4-0) sulphotransferase 14 UST uronyl-2-sulphotransferase 2.13 0.03 0.93 0.78 -1.7 >0.05 ACAN aggrecan 13.83 0.006 N.D <0.05 -1.13 >0.05 VCAN versican -296.1 0.0007 -4.72 0.009 -2.73 <0.05 NCAN neurocan 9.15 0.00 -1.31 0.50 -1.04 >0.05 CSPG4 chondroitin sulphate proteoglycan 4 -6.98 0.0002 -2.02 0.20 -3.07 >0.05 CSPG5 chondroitin sulphate proteoglycan 5 141.7 0.0006 5.09 0.010 1.64 >0.05 (neuroglycan C) SMC3 structural maintenance of -22.82 0.0004 -168.1 <0.000 -1.04 >0.05 chromosomes 3 1 N.D refers to not detectable in 45 cycles of real-time PCR Page | 160 Chapter 4 Chondroitin sulphate in prostate tissue 4.3 Functional Analysis of CSGalNAcT-1 in prostate cancer cell, PC-3 4.3.1 CSGalNAcT-1 is effectively silenced in PC-3 – gene transcript and protein levels To investigate the functional significance of CSGalNAcT-1 upregulation in prostate cancer, the gene was silenced in PC-3 prostate cancer cells using Dharmacon’s ON-TARGET PLUS SMARTpool double-stranded siRNA oligonucleotides. CSGalNAcT-1 was 90% silenced (Figure 4.13A) and protein levels of CSGalNAcT-1 were also reduced by 2.5 fold. (Figure 4.13C, 4.13D) CSGalNAcT- 2, with similar homology to CSGalNAcT-1, was not silenced, indicating that there was no unspecific off-target silencing. Using Cy3-conjugated non-targeting siRNA, transfection efficiency was observed to be 100% (Figure 4.13B). 4.3.2 Reduction of CS/DS chains after CSGalNAcT-1 siRNA The levels of CS/DS expression in the CSGalNAcT-1 silenced cells (cell lysate) were quantified by first measuring the total glycosaminoglycans (GAG) content using the Blyscan Assay and then by measuring the total GAG content again after chondroitinase AC and chondroitinase B digestion. Chondroitin sulphate was significantly decreased by about 37% after CSGalNAcT-1 siRNA treatment. Although dermatan sulphate differential expression did not reach statistical significance (p=0.05), it was observed that there was a 37% decrease in dermatan sulphate expression in the CSGalNAcT-1 silenced cells too (Figure 4.13E). These data are in line with the report that CSGalNAcT-1 is important in the CS/DS biosynthetic process [Sato et al, 2003]. Page | 161 Chapter 4 Chondroitin sulphate in prostate tissue Figure 4.13 CSGalNAcT-1 is effectively silenced and CS/DS expression is reduced. [A] CSGalNAcT-1 was 90% silenced in PC-3 prostate cancer cells using double stranded siRNA oligonucleotides. CSGalNAcT-2, with similar homology to CSGalNAcT-1, was not silenced, indicating that there was no unspecific off-target silencing. [B] Transfection efficiency was 100%, using Cy3-tagged non-targeting siRNA. [C, D] Protein levels of CSGalNAcT-1 silenced PC-3 were also reduced by 2.5 fold. [E] Using Blyscan Assay, the total GAG content (of the cell lysate) before and after each of chondroitinase AC and chondroitinase ABC treatment were qunatified. Chondroitin sulphate was significantly decreased by about 37% after CSGalNAcT-1 siRNA treatment. Dermatan sulphate was also decreased by about 37% in the CSGalNAcT-1 silenced cells, although this did not reach statistical significance (p=0.05). Non-targeting siRNA (n=3); CSGalNAcT-1 siRNA (n=3); ‘*’ represents p<0.05; ‘**’ represents p<0.01. Statistical test was done using unpaired T-test. Experiment was repeated to confirm results. Page | 162 Chapter 4 Chondroitin sulphate in prostate tissue 4.3.3 CSGalNAcT-1 silencing inhibited PC-3 proliferation CSGalNAcT-1 silencing in PC-3 revealed that the proliferation of PC-3 was inhibited by approximately 10% after 72 hours and by 25% after 96 hours of silencing (Figure 4.14A). I questioned whether the decrease in proliferative ability was a result of cell cycle arrest and/or due to increased apoptosis. 4.3.4 Apoptotic analysis of CSGalNAcT-1 silenced PC-3 cells Apoptosis assay was also performed at 72 hours by using the Guava Nexin Assay which contains 2 dyes, Annexin V-PE and 7-AAD. Annexin V binds with high affinity to phosphatidylserine which externalizes to the outer cellular surface in the early stages of apoptosis [Vermes et al, 1995]; 7-AAD is a cell impermeant dye and is used as an indicator of cell membrane structural integrity, which in late stages of apoptosis, there is loss of membrane integrity [Schmidt et al, 1992]. It was found that there was about 5% increase in apoptotic cells after CSGalNAcT-1 silencing, of which 3% increase were in early apoptosis and 2% increase were in late apoptosis (Figure 4.14B). 4.3.5 Cell cycle analysis of CSGalNAcT-1 silenced PC-3 cells Cell cycle analysis was performed at 72 hours via propidium iodide staining and it was observed that there was an arrest in the G1 phase. There were about 3% more cells in the G1 phase in the CSGalNAcT-1 silenced group as compared to the non- targeting siRNA group. Also, there were 3% lesser cells in the S phase of the CSGalNAcT-1 silenced group as compared to the non-targeting siRNA group. There was no significant difference in the number of cells in the G2/M phase in both silenced groups (Figure 4.14C). Page | 163 Chapter 4 Chondroitin sulphate in prostate tissue 4.3.6 CSGalNAcT-1 silencing impeded PC-3 invasion. The invasive ability of the CSGalNAcT-1 silenced cells were also investigated and we report a significant 35% decrease in the number of CSGalNAcT-1 silenced cells that invaded across the Matrigel™, as compared to the number of invaded non- targeting silenced cells at 72 hours post-transfection (Figure 4.14D). 4.3.7 CSGalNAcT-1 silencing has no effect on PC-3 adhesion to collagen type I. We also determined the adhesive capability of the CSGalNAcT-1 silenced cells and report that there was no significant change in the adhesion of CSGalNAcT-1 silenced cells to collagen type I as compared to the non-targeting silenced cells (Figure 4.14E). Page | 164 Chapter 4 Chondroitin sulphate in prostate tissue Figure 4.14 Effects of CSGAlNAcT-1 silencing on PC-3 cell behaviour. [A] CSGalNAcT-1 silencing inhibited proliferation of PC-3 by approximately 10% after 72 hours post-transfection and by 25% after 96 hours of silencing. [B] CSGalNAcT-1 silencing results in 5% increase in apoptotic cells at 72 hours post-transfection, of which 3% increase were in early apoptosis and 2% increase were in late apoptosis. [C] CSGalNAcT-1 silenced PC-3 cells were arrested at the G1 phase. [D] Cell invasion was inhibited after silencing CSGalNAcT-1; there were 35% less invaded CSGalNAcT-1 silenced cells than non-targeting silenced cells at 72 hours post- transfection. [E] No significant change was observed in the ability of CSGalNAcT-1 silenced cells to adhere to collagen type I, as compared to non-targeting silenced cells. Non-targeting siRNA (n=3); CSGalNAcT-1 siRNA (n=3); ‘*’ represents p<0.05; ‘**’ represents p<0.01; ‘***’ represents p<0.001. Statistical test was done using unpaired T-test. Page | 165 Chapter 4 Chondroitin sulphate in prostate tissue 4.3.8 Microarray analysis of CSGalNACT-1 silenced PC-3 cells To gain insight into some of the potential mechanisms CSGalNAcT-1 can mediate cellular behaviour, a genome-wide expression analysis using Affymetrix Human U133 Plus 2 GeneChips was performed. Total RNA was extracted from the non- targeting siRNA and CSGalNAcT-1 siRNA transfected cells to be used for the microarray study. The purity of the RNA was assessed based on A260/A280 while the quality of the isolated RNA was checked using the Agilent Bioanalyzer. Quality of the RNA was assessed based on the electrophoretic trace of the RNA sample (similar to Figure 3.3) and the RNA integrity number (RIN) (Table 4.19). Labelled cRNA was synthesized using the Affymetrix one-cycle target labeling kit and the labelled cRNA was fragmented prior to hybridization to Affymetrix U133 plus2 Genechips. The fragmented cRNA was checked using the Agilent Bioanalyzer and the electrophoregrams analyzed, as performed in Section 3.2, Figure 3.4. Table 4.19 RNA purity and quality of samples used for microarray study Samples A260/A280 RIN Mean SD Mean SD Non-targeting siRNA (n=3) 2.07 0.02 9.53 0.12 CSGalNAcT-1 siRNA (n=3) 2.07 0.04 9.30 0.44 Page | 166 Chapter 4 Chondroitin sulphate in prostate tissue 4.3.8.1 Quality control of microarray data Staining, washing and scanning of the microarray chips were performed on Fluidics Station 450 and the Affymetrix GeneChip Scanner 3000. The presence of housekeeping genes, spike control cRNAs, background scaling factor and noise (raw Q) were checked to ensure the quality of the assays done. The 3’/5’ ratios for β-actin gene lies from 1.1 to 1.51 for all samples. The scaling factors for all the samples are within two-fold of each other, the smallest factor being 3.80 while the largest factor is 5.68, with a mean factor of 5.02. All the spiked in controls were detected as present in all arrays; in addition, there was increasing average intensity values from BioB to CRE for all samples. Q values for all our samples are similar to one another, with a range of 1.0 to 1.2 only. 4.3.8.2 Differential expression of genes in CSGalNAcT-1 silenced PC-3 Several softwares such as GCOS, dCHIP and Genespring were used to analyze the microarray data. The data was normalized and expression intensities of different genes were derived using each of the above software. Genes that were commonly identified by the three softwares as differentially expressed in the prostate cancer sections by more than or equal to 2 fold change with p<0.05 and, in which more than 50% of the tumours showed the same change were obtained. From this selection criterion across the 3 softwares, 10 genes were found downregulated after CSGalNAcT-1 silencing in PC-3 (Table 4.20). Real-time polymerase chain reaction (qPCR) was performed on 6 genes and it was observed that all 6 of these genes were downregulated in CSGalNAcT-1 silenced PC-3 cells, showing 100% correlation with the microarray results. Page | 167 Chapter 4 Chondroitin sulphate in prostate tissue Table 4.20 List of genes differentially expressed after CSGalNAcT-1 silencing and their fold changes as determined from microarray study and real-time PCR. Non-targeting siRNA (n=3); CSGalNAcT-1 siRNA (n=3). Relative fold change and p-values for microarray study are all p<0.05, and obtained from dCHIP. P-values for real-time PCR are performed using unpaired T-test. Gene Gene name Gene Functions Microarray Real-time PCR ID Result (Relative Relative p-value Fold Change) Fold Change MAPRE-1 microtubule-associated protein, RP/EB family, 22919 Cell cycle -3.48 -3.80 0.001 member 1 RCBTB1 regulator of chromosome condensation (RCC1) 55213 Putative adaptor for cullin 3 -2.64 -3.39 0.002 and BTB (POZ) domain containing protein 1 E3 ligase LIFR Leukemia inhibitory factor receptor 3977 Cell proliferation and -2.35 -1.86 0.008 migration HIPK3 Homeodomain interacting protein kinase 3 10114 Apoptosis -3.56 -2.09 0.005 CPOX coproporphyrinogen oxidase 1371 Heme biosynthetic pathway -2.36 -2.52 0.0003 ATL3 atlastin GTPase 3 25923 ER and golgi -2.30 - - morphogenesis TPX2 TPX2, microtubule-associated protein homolog 22974 Cell cycle -2.46 -2.37 0.0001 (Xenopus laevis) FAM96A family with sequence similarity 96, member A 84191 unknown -2.55 - - TPPC2 trafficking protein particle complex 2 6399 ER-Golgi transport -2.22 - - - CDNA FLJ37302 fis, clone BRAMY2016009 - unknown -2.38 - - Page | 168 Chapter 4 Chondroitin sulphate in prostate tissue 4.3.9 Cell cycle and apoptotic proteins were downregulated in CSGalNAcT-1 silenced PC-3 cells It was found that the downregulation of CSGalNAcT-1 resulted in the suppression of several genes important in microtubule and cell cycle dynamics, such as the MAPRE-1 gene which encodes the EB1 proteins and the TPX2 gene. The HIPK3 gene, important in mediating Fas-associated apoptosis resistance in PC-3 cells, was also found to be downregulated through CSGalNAcT-1 silencing. These genes were also verified by real-time PCR to be downregulated in CSGalNAcT-1 silenced cells. The protein levels of these genes in CSGalNAcT-1 silenced PC-3 cells were next investigated. MAPRE-1 was found to be downregulated by 3.5 fold from microarray results and downregulated by 3.8 fold from real-time PCR. Using flow cytometry to determine the protein expression of the EB1 protein, the median log fluorescence intensity for the scrambled siRNA cells is 10.15 and for the CSGalNACT-1 silenced cells, is 9.03 at 48 hours. At 72 hours, the median log fluorescence intensity for the scrambled siRNA cells is 9.16 and for CSGalNACT-1 silenced cells, is 8.79. This correspond to a decrease in the median expression of the protein in CSGalNAcT-1 silenced cells by about 11% at 48 hours and a difference of 4% at 72 hours. (Figure 4.15 and 4.16) TPX2 was also found to be downregulated by 2.5 fold from microarray data and downregulated by 2.4 fold, as verified by real-time PCR. Similarly, using flow cytometry, it was revealed that there was about 30% less CSGalNAcT-1 silenced cells that were TPX2 positively stained at 48 hours than the non-targeting control group. At 48 hours, the median percentage of positively stained scrambled siRNA cells is 94.0% while it is 65.5% for TPX2 positively stained-CSGalNAcT-1 silenced cells. The Page | 169 Chapter 4 Chondroitin sulphate in prostate tissue expression of TPX2 in both non-targeting and CSGalNAcT-1 silenced cells were generally lower at 72 hours than at 48 hours, where the median percentage of positively stained scrambled siRNA cells is 38.2% while it is 25.7% for TPX2 positively stained-CSGalNAcT-1 silenced cells. Nevertheless, it was detected that there were 33% less TPX2 positively stained CSGalNAcT-1 silenced cells than control cells at 72 hours. (Figure 4.17 and 4.18) HIPK3 were also downregulated by 2.5 fold (microarray) and 2.1 fold (qPCR). Flow cytometry results showed that the median log expression of HIPK3 expression, at 48 hours, for scrambled siRNA cells was 9.27 and for CSGalNAcT-1 silenced cells, was 9.03. This translated to a decrease of 3% in CSGalNAcT-1 silenced PC-3 cells. This difference was increased to 10% at 72 hours, where the median log expression of HIPK3 expression for scrambled siRNA cells was 7.24 and for CSGalNAcT-1 silenced cells, was 6.46. (Figure 4.19 and 4.20) Page | 170 Chapter 4 Chondroitin sulphate in prostate tissue Figure 4.15 CSGalNAcT-1 silencing downregulated MAPRE-1 protein. Non-targeting cells and CSGalNAcT-1 PC-3 cells were fixed and stained with MAPRE-1 antibody and an FITC-labeled anti-mouse IgG secondary antibody, and then analyzed via a flow cytometer, at 48 hours and 72 hours post transfection. From the overlay diagram at 48 hours, the median of the graph of CSGalNAcT-1 silenced cells (red) was shifted to the left of the graph of the non-targeting silenced cells (black line), representing lower expression of the MAPRE-1 protein in the CSGalNAcT-1 silenced cells. ‘FSC’ is forward scatter, which is a measure of cellular size. Page | 171 Chapter 4 Chondroitin sulphate in prostate tissue MAPRE-1 Analysis 15 Non-targeting siRNA * * CSGalNAcT-1 siRNA 10 5 Log Fluorescence Intensity 0 48 hours 72 hours Figure 4.16 CSGalNAcT-1 silencing downregulated MAPRE-1 protein by about 11% at 48 hours and 4% at 72 hours. Non-targeting silenced PC-3 (n=3); CSGalNaCT-1 silenced PC-3 (n=3). Statistical test was performed using unpaired T-test. ‘*’ represents p<0.05. Page | 172 Chapter 4 Chondroitin sulphate in prostate tissue Figure 4.17 CSGalNAcT-1 silencing downregulated TPX2 protein. Non-targeting and CSGalNAcT-1 silenced PC-3 cells were fixed and stained with TPX2 antibody and an FITC-labeled anti-mouse IgG secondary antibody, and then analysed via a flow cytometer. The TPX2- positively stained cells were gated from the negative control cells. From the overlay diagrams, there were a smaller number of TPX-2 positively stained CSGalNAcT-1 silenced PC-3 cells (red) compared to the non-targeting silenced cells (black line), representing lower expression of the TPX2 protein in the CSGalNAcT-1 silenced cells at 48 hours and 72 hours. ‘FSC’ is forward scatter, which is a measure of cellular size. Page | 173 Chapter 4 Chondroitin sulphate in prostate tissue TPX2 Analysis 100 *** Non-targeting siRNA 80 CSGalNAcT-1 siRNA 60 *** 40 (% of total cells) total of (% 20 Positively stained cells stained Positively 0 48 hours 72 hours Figure 4.18 CSGalNAcT-1 silencing downregulated TPX-2 protein. There were 30% less TPX2-positive CSGalNAcT-1 silenced cells than the non-targeting control group at 48 hours post-transfection; at 72 hours post-transfection, there were 33% less TPX2 positively stained CSGalNAcT-1 silenced cells than control cells. Non-targeting silenced PC-3 (n=3); CSGalNaCT-1 silenced PC-3 (n=3). Statistical test was performed using unpaired T-test. ‘***’ represents p<0.001. Page | 174 Chapter 4 Chondroitin sulphate in prostate tissue Figure 4.19 CSGalNAcT-1 silencing downregulated HIPK3 protein. Non-targeting and CSGalNAcT-1 PC-3 cells were fixed and stained with HIPK3 antibody and an Alexa-Fluor 488 conjugated anti-rabbit IgG secondary antibody, and then analysed via a flow cytometer, 48 hours and 72 hours post-transfection. The HIPK3 positively stained cells were gated from the negative control cells. From the overlay diagram at 72 hours, the log median fluorescence intensity of the graph of CSGalNAcT-1 silenced cells (red) was shifted to the left of the graph of the non-targeting silenced cells (black line), representing lower expression of the HIPK3 protein in the CSGalNAcT-1 silenced cells. Page | 175 Chapter 4 Chondroitin sulphate in prostate tissue HIPK3 Analysis 10 ** Non-targeting siRNA ** 8 CSGalNAcT-1 siRNA 6 4 2 Log Fluorescence Intensity 0 48 hours 72 hours Figure 4.20 CSGalNAcT-1 silencing downregulated HIPK3 protein. The median log fluorescence intensity of HIPK3 stained positive cells was 3% lower in CSGalNAcT-1 silenced PC-3 cells at 48 hours and this difference was increased to 10% at 72 hours Non-targeting silenced PC- 3 (n=3); CSGalNaCT-1 silenced PC-3 (n=3). Statistical test was performed using unpaired T-test. ‘**’ represents p<0.01. Page | 176 Chapter 4 Chondroitin sulphate in prostate tissue 4.4 Discussion 4.4.1 Increased chondroitin sulphate expression and their potential use as a biomarker of prostate cancer progression In this study, the expression patterns of different chondroitin sulphate subtypes such as non-sulphated chondroitin, chondroitin-4-sulphate, chondroitin-6-sulphate and chondroitin-2,6-sulphate in benign prostatic tissue and paired prostate cancer samples from corresponding patients were examined. On the whole, chondroitin sulphate appears to be more highly expressed in malignant tissues as compared to non- malignant tissues. However, chondroitin sulphate expression and localization appears to alter in different Gleason grades of prostate cancer tissue, with increasing expression of chondroitin sulphate in the peritumoural stroma and a concomitant decreasing expression of chondroitin sulphate in the cytoplasm of the malignant cells, as tumour grade increases. The growth promoting effects of chondroitin sulphate has long been reported. Back in 1965, Takeuchi J demonstrated that chondroitin sulphate treatment promoted the growth of solid Ehrlich ascites tumour in a dose-dependent manner. [Takeuchi J, 1965] In yet another approach, Denholm et al, 2001, reported inhibition of melanoma cells’ proliferation through removing chondroitin sulphate and dermatan sulphate chains using chondroitinase AC and chondroitinase B respectively [Denholm et al, 2001]. Chondroitinase AC, but not chondroitinase B, treatment also enhanced apoptosis of melanoma cells. These studies demonstrate the growth promoting and apoptosis resistance effects of chondroitin sulphate in tumour biology. Chondroitin sulphate may regulate prostate cancer cell behavior and influence patient prognosis through several mechanisms. Platelet derived growth factors (PDGFs) are involved in regulating proliferation of cancer cells and stimulating Page | 177 Chapter 4 Chondroitin sulphate in prostate tissue tumour angiogenesis, and chondroitin-4-sulphate has been shown to enhance PDGF action by enhancing signaling of the tyrosine kinase receptors [Fthenou et al, 2006]. Indeed, PDGF and its receptor have been reported to be upregulated in prostate adenocarcinoma, and expression of PDGF receptor has been shown to mark prostate cancer cells with the greatest disposition for bony metastasis [Dolloff et al, 2005; Fudge et al, 1994]. Stromal cell-derived factor-1 (SDF-1) is a chemoattractant for prostate carcinoma cells [Cooper et al, 2003]. Chondroitin sulphate has been shown to be required for cellular binding of SDF-1 [Netelenbos et al, 2003]. In prostate carcinoma, treatment of cancer cells with SDF-1 upregulated expression of αvβ3 integrin receptors, resulting in increased cell adhesion and invasion [Sun et al, 2007]. Furthermore, SDF-1 has been shown to be critical for spread of prostate cancer, and a positive correlation has been reported between SDF-1 signaling and bony metastasis [Sun et al, 2005; Taichman et al, 2002]. Several reports have highlighted the importance chondroitin sulphate plays in development as well. They have demonstrated that chondroitin sulphate is essential for cell proliferation, cytokinesis and migration through inhibiting biosynthesis of chondroitin by RNA interference, and this may underlie the mechanistic processes leading to the association of chondroitin sulphate with adverse pathologic parameters. [Mizuguchi et al, 2003; Izumikawa et al, 2004] The differing expression of chondroitin sulphate in different tumour grades of prostate cancer reflects yet again, that tumours are of different heterogeneous genetic makeup. It was interesting to note that in the prostate adenocarcinoma samples, there was a decreasing expression of chondroitin-6-sulphate and non-sulphated chondroitin Page | 178 Chapter 4 Chondroitin sulphate in prostate tissue in the adenocarcinoma cells with increasing Gleason grades. Chondroitin-2,6-sulphate expression was the highest in the Gleason 6 and 7 grade cancers and the lowest in Gleason 8 to 10 grade cancers. In contrast, all chondroitin sulphate subtypes expression in the peritumoural stroma was observed to increase with increasing Gleason grades, the highest being in the Gleason sum 8 to 10 cancers. Proteoglycans and glycosaminoglycans in the tumour stroma have been well-known to play intricate and crucial roles in the carcinogenic process, such as growth factor binding, disturbances in the collagen fibrillar network that permits neoplastic formation and invasion [Iozzo RV, 1999] and matrix degradation via enhancement or inhibition of proteolytic enzymes [Hornebeck et al, 2002]. The significance of the spatial and temporal expression of the CS/CSPG in prostate adenocarcinoma tissues remains to be explored. 4.4.2 Unraveling the functional significance of CSGalNAcT-1 upregulation in prostate cancer Next, I was interested to understand the functional relevance of the upregulation of chondroitin sulphate in prostate adenocarcinoma. Using the prostate cancer cell lines, LNCaP and PC-3 cells as our model of study, the expression of genes coding for chondroitin sulphate biosynthetic and modifying enzymes were screened by using real-time PCR. This was performed to detect the differential expression of genes that could bring about an increased expression of chondroitin sulphate in prostate cancer. Many CS/CSPGs synthetic genes were revealed to be differentially expressed in the prostate cancer cell lines as compared to the normal prostate epithelial cell, RWPE-1. Of interest to this study, I would like to highlight the CSGalNAcT-1 gene, which was found to be overexpressed in prostate Page | 179 Chapter 4 Chondroitin sulphate in prostate tissue adenocarcinoma tissues from our microarray data (about 2 fold) and in prostate cancer cell lines, LNCaP (about 10 fold) and PC-3 (about 20 fold). Since CSGalNAcT-1 is necessary for the initiation of CS/DS biosynthesis [Sato et al, 2003], the upregulation of the CS chains observed through immunohistochemistry could be a ttributed to the increased expression of this gene. Silencing CSGalNAcT-1 in PC-3 resulted in decreased expression of the CS/DS chains and also resulted in decreased proliferation, increased apoptosis and reduced invasion. Interestingly, by using microarray screen again, we have also detected previously unknown downstream targets of CSGalNAcT-1 which include the microtubule-associated proteins, EB1 encoded by MAPRE-1 and TPX2, as well as HIPK3 that is involved in FAS-mediated apoptosis resistance. MAPRE-1 codes for the EB1 protein which is a family member of the microtubule tip-tracking proteins [Li and Yi, 2001]. EB1 has also been found to interact with the cytoplasmic microtubules along their full length with accentuations at their tips in interphase and was localised to the centrosomes and the mitotic spindle during mitosis [Morrison et al, 1998; Berrueta et al, 1998]. It has been implicated in stimulation of growth at the microtubule plus ends which is important for search- capture of ER-Golgi transport vesicle, in the microtubule capture at the cell cortex during cytoskeletal reorientation, at the kinetochores of prometaphase and at the cell cortex where astral microtubules can attach [Tirnauer et al, 2002; Rogers et al, 2002; Vaughan KT, 2005]. It is an important regulator of microtubule dynamics including mitotic spindle formation and proper chromosome segregation [Tirnauer and Bierer, 2000]. In addition, EB1 has been found to interact with the adenomatous polyposis coli (APC) protein, a well-known tumour suppressor gene [Su et al, 1995]. APC and EB1 have been reported to function together in mitosis to regulate spindle dynamics Page | 180 Chapter 4 Chondroitin sulphate in prostate tissue and chromosome alignment [Green et al, 2005], however, there was a separate study reporting that the EB1-APC interaction is not crucial to promoting microtubule net growth at the microtubule plus ends [Kita et al, 2006]. In addition, EB1 has been reported to promote Aurora B kinase activity through blocking its inactivation by protein phosphatase 2A [Sun et al, 2008]. Aurora B kinase, belonging to the Aurora kinase family well-known to be implicated in tumourigenesis and are important regulators of the cell cycle [Vader and Lens, 2008], is known to be an important regulator of kinetochore-microtubule attachments; it is crucial for correction of syntelic attachments, and ensuring that the chromosomes are properly bi-orientated and will segregate equally in anaphase [Andrews PD, 2005]. Downregulation of the EB1 protein could thus result in reduced activity of Aurora B kinase and lead to cell cycle arrest. Recently, Liu et al have reported that EB1 acts as an oncogene that promotes cellular growth and inhibits apoptosis via activation of the beta-catenin/TCF pathway [Liu et al, 2009]. In addition, another study conducted by Wang et al also reported the promotion of cellular growth in oesophageal squamous cell carcinoma by overexpressing EB1, through the beta-catenin/TCF pathway [Wang et al, 2005]. They revealed that EB1 disrupted the interaction between APC and beta-catenin. Beta- catenin/TCF are effector molecules of the Wnt signaling pathway. Beta-catenin can be found as a protein complex, being bound with APC, glycogen synthase kinase 3β (GSK3β), axin and protein phosphatase 2A (PP2A), a multi-subunit complex. [Hart et al, 1998; Ratcliffe et al, 2000] In the presence of Wnt signalling, beta-catenin is protected from degradation via the ubiquitin pathway [Aberle et al, 1997] and this stabilized beta-catenin enters the nucleus, complexes with the transcription factors of the Lef1/TCF family, promotes transcription of downstream targets such as c-myc, c- Page | 181 Chapter 4 Chondroitin sulphate in prostate tissue jun and cyclin D1 [Shtutman et al, 1999; He et al, 1998]. Our study has further confirmed the growth promoting and apoptosis-resistance effects of EB1, as downregulation of EB1 also resulted in growth inhibition and increased apoptosis in the prostate cancer cells. Even more so, we have demonstrated that CSGalNAcT-1 could regulate EBl proteins as well; although beta-catenin/TCF, APC, GSK3B, axin, c-myc, c-jun, cyclin D1 transcript levels were not differentially changed in our microarray data (see Table 4.22), these could represent changes at a post-translational level. Alternatively, perhaps CSGalNAcT-1 mediated-EB1 downregulation inhibits cell growth and enhance apoptosis via other pathways. Future work is needed to confirm how CSGalNAcT-1 mediates EB1 downregulation and also, how EB1 plays a role in CSGalNAcT-1 mediated growth inhibition and enhance apoptosis. Possibly, it would also be interesting to investigate any possible interactions between chondroitin sulphate/dermatan sulphate and or CSGalNAcT-1 interactions with the Wnt signalling pathway in prostate tumourigenesis; Wnt-1-induced secreted protein 1 (WISP-1) has been reported to bind to decorin and biglycan (dermatan sulphate proteoglycans) [Desnoyers et al, 2001] thus regulating their function; Nadanaka et al has also recently reported the importance of chondroitin sulphate E structures synthesized by C4ST1in Wnt3a signaling [Nadanaka et al, 2008]. TPX2 is a microtubule-associated protein important in promoting microtubule assembly around chromosomes that is crucial for bipolar spindle formation [Gruss et al, 2002]. In Xenopus laevis egg extracts, TPX2 is bound to importins α and β, which when RanGTP binds, TPX2 is released and promotes microtubule assembly [Gruss et al, 2001; Gruss et al, 2004]. TPX2 also activates Aurora A [Kufer et al, 2002; Eyers et al, 2003], a serine/threonine kinase that is important in centrosome maturation, centrosome separation and bipolar spindle assembly. Proper localization and function Page | 182 Chapter 4 Chondroitin sulphate in prostate tissue of centrosomal components, accurate migration of the centrosomes to define the 2 poles of the bipolar spindles and the astral microtubules connections with the centrosome to the cell cortex to maintain proper bipolarity requires Aurora A kinase [Vader and Lens, 2008]. Moreover, Aurora A kinase has also been reported to phosphorylate CENP-A which causes the recruitment of Aurora B kinase to the inner centromere at prometaphase [Kunitoku et al, 2003]. The endpoint of mitosis is to ensure diploid chromosome number in the daughter cell and as such, cell cycle progression is a tightly controlled process that involves many checkpoints to ensure proper cell division occurring at high fidelity. The microtubule network thus plays an imperative role in the proper segregation of cellular contents; we know that EB1 and TPX2 are microtubule-associated proteins, which together with their binding partners and downstream players such as the Aurora kinases, regulate microtubule dynamics and spindle assembly. Downregulation of EB1 and TPX2 could thus result in disruption in cell cycle progression. In our study, we have ascertained CSGalNAcT-1 silencing resulted in downregulation of EB1 and TPX2 which could perhaps affect the activation of the Aurora kinases and the beta- catenin/TCF pathway, resulting in decreased proliferative ability yet increased apoptotic capacity of the PC-3 cells. HIPK3, homeodomain interacting protein kinase 3, is a serine/threonine kinase that phosphorylates FADD [Rochat-Steiner et al, 2000] and was found to be elevated in Fas-resistant DU145 and PC-3 cells in comparison with more sensitive PPC-1 and ALVA31 prostate carcinoma cells [Curtin and Cotter, 2004]. Curtin and Cotter also reported that inhibition of HIPK3 significantly reduced FADD phosphorylation and increased the sensitivity of DU145 cells to Fas-mediated apoptosis. They also further demonstrated that HIPK3 expression is regulated by JNK through the transcription Page | 183 Chapter 4 Chondroitin sulphate in prostate tissue factor c-Jun. In our study, we did not detect any differential change in JNK or c-Jun expression (see Table 4.14) in the CSGalNAcT-1 silenced PC-3 cells; this could possibly indicate 1) a JNK-independent pathway of HIPK3 activation that is mediated by the CSGalNAcT-1 and/or CS/DS chains or 2) post-translational modifications of JNK. Nevertheless, our findings have shown that CSGalNAcT-1 silencing could downregulate HIPK3, rendering the cells more sensitive to apoptosis, consistent with our data that there was increased apoptosis after CSGalNAcT- 1 removal. CSGalNAcT-1 suppression also reduced PC-3 cells invasiveness through the Matrigel in vitro. Leukaemia inhibitory factor receptor (LIFR) has been reported to be implicated in carcinogenesis [Grant et al, 2001; Tomida et al, 2000] and in our study, have been found to be downregulated after CSGalNAcT-1 silencing. The leukemia inhibitory factor receptor (LIF-R) subunit is a component of cell-surface receptor complexes for the multifunctional cytokines like leukaemia inhibitory factor (LIF), cardiotrophin-1, ciliary neurotrophic factor, and human oncostatin M [Plun-Favreau et al, 2003]. LIFR activation by the cytokine LIF has been shown to promote cellular proliferation via protein tyrosine kinases, protein kinase C and STAT3 phosphorylation in pancreatic cancer cells [Kamohara et al, 2007]. In prostate cancer, oncostatin M stimulated the growth of 22Rv1 prostate cancer cells through the p38 MAPK signaling pathway and phosphatidylinositol-3-kinase pathway [Godoy Tundidor et al, 2005]. Targets of the p38 MAPK signaling pathway and PI3K signaling includes the matrix metalloproteinase family members, which play important roles in tumour invasion by degrading extracellular matrix and regulating cytokine activities [Shukla et al, 2007; Hsieh et al, 2007]. In our study, CSGalNAcT-1 silencing suppresses LIFR mRNA and also downregulation of matrix metalloproteinase 13 by about 1.6 fold, which could potentially affect the Page | 184 Chapter 4 Chondroitin sulphate in prostate tissue invasiveness of the PC-3 cells. In addition, other proteolytic enzymes are found to be downregulated after CSGalNAcT-1 silencing, such as carboxypeptidase M (1.6 fold, p<0.001) and Sumo1/sentrin specific peptidase 7 (1.6 fold, p<0.01) which although relatively not much is known about their effects on invasion currently, might serve to effect the aggression of the cancer cells into the surrounding tissue. Our present study provides mechanistic insights into how CSGalNAcT-1 could attenuate prostate tumour activity by inhibiting proliferation, activating apoptosis and inhibiting cell invasion (Figure 4.21). A longer period of study will be useful in determining the long term effects of CSGalNAcT-1 silencing in the prostate adenocarcinoma cells as from our study, we noticed only small percentages of changes in proliferation, apoptosis and EB1, TPX2 and HIPK3 protein levels at 72 hours. This could possibly be due to the complex synthesis of the CS/DS chains from gene transcript level to proteoglycans end-product; moreover, this could be hampered by the transient silencing of the CSGalNAcT-1 gene using our methods which will not allow us to conduct long-term studies. In addition, future work will also be needed to determine exactly the mechanisms by which CSGalNAcT-1 affects cell cycle, elicits apoptosis and reduces tumour cell invasion. Since CSGalNAcT-1 is involved in chondroitin sulphate synthesis, the question remains if CSGalNAcT-1 elicits signaling via chondroitin sulphate chains or by novel non-enzymatic mechanisms that bypasses the chondroitin sulphate chains. Nevertheless, this study reveals important tumour- attenuating properties of the CSGalNaCT-1 and/or the CS/DS chains and could prove to be an interesting novel therapeutic target. Page | 185 Chapter 4 Chondroitin sulphate in prostate tissue Table 4.21 Transcript data of genes possibly involved in CSGalNAcT-1-mediated signaling. Relative fold change values are those from dCHIP. P-value of genes listed are p >0.05 Non-targeting CSGalNAcT-1 siRNA PC-3 siRNA PC-3 Relative cells (n=3) cells (n=3) fold Gene Gene Name Mean Mean change – Signal Signal Microarray SEM SEM Intensity Intensity Result (Log2) (Log2) glycogen synthase GSK3β 8.91 0.04 8.94 0.006 1.02 kinase 3 beta CTNNB1 beta-catenin 5.39 0.11 5.25 0.05 0.90 adenomatous APC 2.59 0.20 2.11 0.15 0.72 polyposis coli AXIN1 axin 1 5.43 0.04 5.64 0.11 1.16 AXIN2 axin 2 5.78 0.11 5.96 0.05 1.13 protein phosphatase 2 (formerly 2A), PPP2CA 3.95 0.26 3.94 0.24 0.99 catalytic subunit, alpha isoform protein phosphatase 2 (formerly 2A), PPP2CB 10.32 0.02 10.35 0.04 1.02 catalytic subunit, beta isoform protein phosphatase 2 (formerly 2A), PPP2R1A 10.14 0.03 10.16 0.05 1.01 regulatory subunit A, alpha isoform protein phosphatase 2 (formerly 2A), PPP2R1B regulatory subunit A, 8.46 0.02 8.5 0.01 1.03 beta isoform protein phosphatase PPP2R5A 2, regulatory subunit 8.25 0.05 8.28 0.009 1.02 B', alpha isoform protein phosphatase PPP2R5B 2, regulatory subunit 8.16 0.06 8.16 0.05 1 B', beta isoform protein phosphatase PPP2R5C 2, regulatory subunit 3.7 0.21 3.35 0.26 0.78 B', gamma isoform protein phosphatase PPP2R5D 2, regulatory subunit 7.38 0.04 7.3 0.06 0.95 B', delta isoform protein phosphatase PPP2R5E 2, regulatory subunit 8.59 0.02 8.61 0.06 1.01 B', epsilon isoform AURKB Aurora kinase B 9.4 0.02 9.5 0.06 1.10 Page | 186 Chapter 4 Chondroitin sulphate in prostate tissue lymphoid enhancer- LEF1 4.04 0.28 4.50 0.14 1.38 binding factor 1 transcription factor 7 TCF7 (T-cell specific, 3.71 0.10 3.89 0.06 1.13 HMG-box) transcription factor TCF7L1 7-like 1 (T-cell 6.78 0.06 6.97 0.04 1.14 specific, HMG-box) transcription factor TCF7L2 7-like 2 (T-cell 8.50 0.09 8.42 0.06 0.94 specific, HMG-box) v-myc MYC/c- myelocytomatosis 10.38 0.02 10.34 0.04 0.97 myc viral oncogene homolog (avian) v-jun sarcoma virus c-Jun 17 oncogene 6.84 0.02 6.88 0.04 1.00 homolog (avian) CCND1 cyclin D1 10.24 0.03 10.43 0.02 1.14 serine/threonine AURKA kinase 6 (aurora 10.57 0.01 10.54 0.03 0.98 kinase A) centromere protein CENP-A 10.38 0.01 10.4 0.01 1.00 A, 17kDa Fas (TNFRSF6)- FADD associated via death 10.19 0.03 10.07 0.13 0.92 domain MAPK8 / Mitogen-activated 6.52 0.03 6.49 0.13 0.98 JNK protein kinase 8 CASP8 caspase 8 6.39 0.003 6.12 0.55 0.83 CASP3 caspase 3 8.62 0.02 8.61 0.04 0.99 Page | 187 Chapter 4 Chondroitin sulphate in prostate cancer Figure 4.21 Hypothetical diagram on the mechanisms of CSGalNAcT-1 role in PC-3 cell behaviour. Proposed hypothetical diagram of possible pathways CSGalNAcT-1 could be involved in. In this study, CSGalNAcT-1 has been shown to regulate the MAPRE-1 gene encoding EB1 protein. EB1 has been reported to promote cell growth and inhibit apoptosis via the beta-catenin/TCF pathway. Beta-catenin, glycogen synthase kinase 3 beta (GSK3B), adenomatous polyposis coli (APC), axin and protein phosphatase 2A (PP2A) can be found as a protein complex; in the presence of Wnt signalling, dishevelled (dvl) proteins transduce the signal to inhibit GSK3B, leading to beta-catenin accumulation in the cytosol. Beta-catenin enters the nucleus and cooperate with transcription factors of the LEF/TCF family to promote transcription of downstream targets such as c-myc, c-jun and cyclin D1. These targets are important factors for cell growth and cell cycle. EB1 can also protect Aurora kinase B (AURKB) from phosphorylation by PP2A; Aurora kinase B, as well as Aurora kinase A (AURKA), is an important mediator of cell cycle. TPX2 was also found to be regulated by CSGalNAcT-1. Crucial for bipolar spindle assembly, TPX2 also activates AURKA. AURKA could phosphorylate CENP-A which could promote AURKB to the inner centromere. CSGalNAcT-1 also regulates the HIPK3 gene, which phosphorylates Fas-associated via death domain (FADD) and results in Fas-mediated apoptosis resistance. This has been reported to be mediated by JNK through c-jun. As such, CSGalNAcT-1 silencing downregulates MAPRE-1/EB1, TPX2 and HIPK3 protein, which could impede cell cycle progression, reduce cell growth and promote apoptosis. Page | 188 Chapter 5 Heparan sulphate in prostate cancer Chapter 5 Expression and Functional Analysis of Heparan Sulphate in Prostate Cancer 5.1 Expression analysis of heparan sulphate in prostate cancer 5.1.1 Localization of heparan sulphate in BPH and adenocarcinoma tissues Expression analysis of the heparan sulphate was examined in a total of 106 cases of benign prostatic hyperplastic tissues and 132 cases of prostate adenocarcinoma cases, in which 85 of these cases are matched cases (i.e both BPH samples and prostate adenocarcinoma tissues were taken from the same patient). Using monoclonal antibody 10E4, heparan sulphate was not expressed in the benign secretory epithelial cells in about 90.57% of the benign cases. Heparan sulphate staining was positive in the periglandular stroma for 93.4% of the benign cases, with a range of weak to strong staining. 86.36% of the prostate adenocarcinoma tissues also do not express heparan sulphate in the cytoplasm of the cancer cells. Peritumoural stroma cells were positively stained for heparan sulphate in about 86.3% of the adenocarcinoma tissues, demonstrating weak to strong staining. The expression data of heparan sulphate in 106 cases of benign prostatic tissues and 132 cases of prostate adenocarcinoma samples are shown in Table 5.1 and Figure 5.1. 5.1.2 Comparison of heparan sulphate expression in benign prostatic and adenocarcinoma tissues When staining differences in the means of 85 cases of matched benign and prostate adenocarcinoma tissues were compared, no significant changes in staining intensity or percentage of cells stained positively for heparan sulphate in the adenocarcinoma cells and the peritumoural stroma, as compared to the benign Page | 189 Chapter 5 Heparan sulphate in prostate cancer epithelial cells and the periglandular stroma, respectively, were observed. (Figure 5.2 and Table 5.2) 5.1.3 Comparison of heparan sulphate expression in different grades of prostate adenocarcinoma tissues Following this, we wanted to determine whether the expression levels of heparan sulphate in both the cytoplasm of the prostate adenocarcinoma cells and the peritumoural stroma are associated with different tumour grades of prostate adenocarcinoma. Due to technical difficulties encountered as a result of different batches of antibodies employed for use for the prostate TMAs (Gleason sum 6 and 7) and the core biopsies (Gleason sum 3 to 5 and 8 to 10), we compared the 2 extreme grades of prostate adenocarcinoma cases only (i.e Gleason sum 3 to 5 versus Gleason sum 8 to 10) since immunohistochemical staining was performed at the same time and using the same batch of antibodies for this set of tissues. There was very weak mean staining in the cytoplasm of the adenocarcinoma tissues of Gleason sum 3 to 5 and 8 to 10, and there was no significant difference between the two extreme grades. There was also a very small percentage of adenocarcinoma cells positively stained for heparan sulphate in both Gleason sum 3 to 5 and Gleason sum 8 to 10 groups, and there was no significant difference between the two extreme grades either. In addition, no significant difference was also observed in the staining intensity, percentage of peritumoural stroma cells positively stained and immunoreactivity score, between the Gleason sum 3 to 5 group and Gleason sum 8 to 10 group (Figure 5.3 and Table 5.3). Page | 190 Chapter 5 Heparan sulphate in prostate cancer Table 5.1 Expression profiles of 10E4 specific-heparan sulphate epitope in prostate tissues. Number stained Stained tissue Number Weighted Average Intensity Percentage of cells Total number Immunoreactivity Score (IRS) component unstained (WAI) (%) ≤1 1 Benign Secretory 96 4 1 5 8 2 9 0 1 106 epithelial cells 90.57% 3.77% 0.94% 4.72% 7.55% 1.89% 8.49% 0.00% 0.94% Periglandular 7 37 14 48 32 67 48 23 28 106 Stromal cells 6.60% 34.91% 13.21% 45.28% 30.19% 63.21%45.28% 21.70% 26.42% Adenocarcinoma Cancerous 114 7 2 9 16 2 17 1 0 132 cells 86.36% 5.30% 1.52% 6.82% 12.12% 1.52% 12.88% 0.76% 0.00% Peritumoural 18 33 22 58 36 77 51 24 38 131 Stromal cellsa 13.74% 25.19% 16.79% 44.27% 27.48% 58.78%38.93% 18.32% 29.01% aOne cancer section had too little stroma cells for analysis. Page | 191 Chapter 5 Heparan sulphate in prostate cancer Figure 5.1 10E4-specific heparan sulphate epitope staining in prostate tissues. [A] BPH [B] Gleason sum 4 Section [C] Gleason sum 7 section [D] Gleason sum 9 section [E] Negative Control ‘Epi’ represents benign epithelial cell, ‘CaP’ represents cancer cells and ‘SM’ represents stromal cells. Page | 192 Chapter 5 Heparan sulphate in prostate cancer Figure 5.2 Comparisons of 10E4-specific heparan sulphate epitope expression in prostate adenocarcinoma samples (n=85) against matched BPH samples (n=85). [A] Weighted Average Intensity [B] Percentage of cells positively stained [C] Immunoreactivity Score. Statistical tests were performed using paired T-test. Page | 193 Chapter 5 Heparan sulphate in prostate cancer Table 5.2 Mean values of 10E4-specific heparan sulphate epitope staining in prostate adenocarcinoma samples (n=85) versus paired BPH samples (n=85). Statistical tests were performed using paired T-test. Weighted Average Intensity Percentage of cells stained IRS Mean SEM P-value Mean SEM p-value Mean SEM P-value Epithelial / Cancer cell region Benign secretory epithelial 0.18 0.07 0.092 3.82 1.76 0.199 7.41 3.91 0.609 cells Adenocarcinoma cells 0.32 0.08 6.15 1.95 9.53 2.88 Stromal region Periglandular stromal cells 1.59 0.09 0.907 66.29 3.88 0.368 119.70 9.61 0.446 Peritumoural stromal cells 1.60 0.09 69.79 4.17 128.00 9.84 Page | 194 Chapter 5 Heparan sulphate in prostate cancer Figure 5.3 Comparisons of 10E4-specific heparan sulphate epitope expression in different Gleason grades of prostate adenocarcinoma tissues, in both peritumoural stroma and cytoplasm of cancer cell. Gleason sum 3, 4 & 5 (n=21); Gleason sum 8, 9 & 10 (n=46) [A,D] Weighted Average Intensity [B,E] Percentage of cells positively stained [C,F] Immunoreactivity score. Statistical test was performed using unpaired T-test. Page | 195 Chapter 5 Heparan sulphate in prostate cancer Table 5.3 Mean values of 10E4-specific heparan sulphate staining in different grades of prostate adenocarcinoma tissues: Gleason sum 3,4 & 5 (n=21) vs Gleason sum 8,9 & 10 (n=46). P-values were determined by unpaired T-test. Gleason Sum Weighted Average Intensity Percentage of cells positively stained Immunoreactivity Score Mean SEM P-value Mean SEM P-value Mean SEM P-value Adenocarcinoma 3,4 & 5 0.48 0.33 0.808 0.19 0.15 0.402 0.95 0.74 0.467 8, 9 & 10 1.52 0.82 0.15 0.08 2.17 1.07 Peritumoural Stroma 3,4 & 5 1.64 0.29 0.696 45.24 9.75 0.953 108.10 25.82 0.600 8, 9 & 10 1.52 0.16 45.87 5.64 94.02 13.70 Page | 196 Chapter 5 Heparan sulphate in prostate cancer 5.1.4 Association of heparan sulphate staining with clinicopathological parameters Heparan sulphate staining in the peritumoural stroma was found associated with clinical staging (P=0.046). (Table 5.4 and 5.5) No significant correlations were found when the expression patterns of heparan sulphate were analyzed with the other clinicopathological parameters listed in Table 2 (P >0.05), namely resection margin including distal and proximal margin involvement, pathological T staging, extraprostatic extension, seminal vesicle involvement, perineural invasion, lymphovascular invasion, tumour volume and pre-operative PSA levels. We were unable to correlate with obturator lymph node involvement as none of our patients demonstrated metastases to nodes. Page | 197 Chapter 5 Heparan sulphate in prostate cancer Table 5.4 Association studies of 10E4-epitope specific heparan sulphate staining intensity with clinicopathological parameters. Stained tissue Clinicopathological Total Weighted Weighted P- component parameter Average Average value Intensity Intensity <2 ≥2 Adenocarcinoma Pathological T staging T2a to T2c 35 31 4 1.000 100.0% 88.6% 11.4% T3a to T3c 34 31 3 100.0% 91.2% 8.8% Total 69 Extraprostatic extension Yes 34 31 3 1.000 100.0% 91.2% 8.8% No 35 31 4 100.0% 88.6% 11.4% Total 69 Seminal vesicle involvement Involved 7 6 1 0.578 100.0% 85.7% 14.3% Uninvolved 65 58 7 100.0% 89.2% 10.8% Total 72 Pre-operative PSA levels ≤8ng/ml 20 17 3 0.695 100.0% 85.0% 15.0% >8ng/ml 20 15 5 100.0% 75.0% 25.0% Total 40 Lymphovascular invasion Present 9 9 0 0.584 100.0% 100.0% .0% Absent 60 52 8 100.0% 86.7% 13.3% Total 69 Perineural invasion Seen 60 54 6 0.626 100.0% 90.0% 10.0% Not seen 13 11 2 100.0% 84.6% 15.4% Total 73 Page | 198 Chapter 5 Heparan sulphate in prostate cancer Circumferential resection margin Involved 33 29 4 1.000 100.0% 87.9% 12.1% Uninvolved 39 35 4 100.0% 89.7% 10.3% Total 72 Apex margin Involved 34 32 2 0.267 100.0% 94.1% 5.9% Uninvolved 38 32 6 100.0% 84.2% 15.8% Total 72 Proximal margin Involved 8 6 2 0.210 100.0% 75.0% 25.0% Uninvolved 65 59 6 100.0% 90.8% 9.2% Total 73 Clinical stage T1a to T1c 33 28 5 0.172 100.0% 84.8% 15.2% T3a to T3b 8 5 3 100.0% 62.5% 37.5% Total 41 Lobe occurence Both Lobes 62 55 7 0.584 100.0% 88.7% 11.3% Singles Lobe 9 9 0 100.0% 100.0% .0% Total 71 Tumour Volume <4cm3 28 27 1 0.614 100.0% 96.4% 3.6% ≥4 cm3 31 28 3 100.0% 90.3% 9.7% Total 59 Maximum tumour dimension <2.5cm 35 32 3 0.711 100.0% 91.4% 8.6% ≥2.5cm 37 32 5 100.0% 86.5% 13.5% Total 72 Page | 199 Chapter 5 Heparan sulphate in prostate cancer Peritumoural Pathological T staging Stroma T2a to T2c 34 22 12 0.144 100.0% 64.7% 35.3% T3a to T3c 34 15 19 100.0% 44.1% 55.9% Total 68 Extraprostatic extension Yes 34 15 19 0.144 100.0% 44.1% 55.9% No 34 22 12 100.0% 64.7% 35.3% Total 68 Seminal vesicle involvement Involved 7 4 3 1.000 100.0% 57.1% 42.9% Uninvolved 64 35 29 100.0% 54.7% 45.3% Total 71 Pre-operative PSA levels ≤8ng/ml 19 7 12 0.748 100.0% 36.8% 63.2% >8ng/ml 20 9 11 100.0% 45.0% 55.0% Total 39 Lymphovascular invasion Present 9 4 5 0.722 100.0% 44.4% 55.6% Absent 59 33 26 100.0% 55.9% 44.1% Total 68 Perineural invasion Seen 60 34 26 0.756 100.0% 56.7% 43.3% Not seen 12 6 6 100.0% 50.0% 50.0% Total 72 Circumferential resection margin Involved 33 17 16 0.638 100.0% 51.5% 48.5% Uninvolved 38 22 16 100.0% 57.9% 42.1% Total 71 Page | 200 Chapter 5 Heparan sulphate in prostate cancer Apex margin Involved 34 18 16 0.637 100.0% 52.9% 47.1% Uninvolved 37 22 15 100.0% 59.5% 40.5% Total 71 Proximal margin Involved 8 5 3 0.725 100.0% 62.5% 37.5% Uninvolved 64 35 29 100.0% 54.7% 45.3% Total 72 Clinical stage T1a to T1c 32 12 20 0.250 100.0% 37.5% 62.5% T3a to T3b 8 5 3 100.0% 62.5% 37.5% Total 40 Lobe occurence Both Lobes 61 33 28 0.721 100.0% 54.1% 45.9% Singles Lobe 9 6 3 100.0% 66.7% 33.3% Total 70 Tumour Volume <4cm3 28 18 10 0.597 100.0% 64.3% 35.7% ≥4 cm3 31 17 14 100.0% 54.8% 45.2% Total 59 Maximum tumour dimension <2.5cm 34 21 13 0.341 100.0% 61.8% 38.2% ≥2.5cm 37 18 19 100.0% 48.6% 51.4% Total 71 Page | 201 Chapter 5 Heparan sulphate in prostate cancer Table 5.5 Association of percentage of positively 10E4-stained cells with clinicopathological data. Stained tissue Clinicopathological Total Percentage Percentage P- component parameter of cells of cells value positively positively stained stained ≤50% >50% Adenocarcinoma Pathological T staging T2a to T2c 35 35 0 0.239 100.0% 100.0% .0% T3a to T3c 34 32 2 100.0% 94.1% 5.9% Total 69 Extraprostatic extension Yes 34 32 2 0.239 100.0% 94.1% 5.9% No 35 35 0 100.0% 100.0% .0% Total 69 Seminal vesicle involvement Involved 7 6 1 0.186 100.0% 85.7% 14.3% Uninvolved 65 64 1 100.0% 98.5% 1.5% Total 72 Pre-operative PSA levels ≤8ng/ml 20 20 0 -- 100.0% 100.0% 0.0% >8ng/ml 20 20 0 100.0% 100.0% 0.0% Total 40 Lymphovascular invasion Present 9 8 1 0.246 100.0% 88.9% 11.1% Absent 60 59 1 100.0% 98.3% 1.7% Total 69 Perineural invasion Seen 60 58 2 1.000 100.0% 96.7% 3.3% Not seen 13 13 0 100.0% 100.0% 0.0% Total 73 Page | 202 Chapter 5 Heparan sulphate in prostate cancer Circumferential resection margin Involved 33 32 1 1.000 100.0% 97.0% 3.0% Uninvolved 39 38 1 100.0% 97.4% 2.6% Total 72 Apex margin Involved 34 34 0 0.495 100.0% 100.0% .0% Uninvolved 38 36 2 100.0% 94.7% 5.3% Total 72 Proximal margin Involved 8 8 0 1.000 100.0% 100.0% .0% Uninvolved 65 63 2 100.0% 96.9% 3.1% Total 73 Clinical stage T1a to T1c 33 33 0 0.195 100.0% 100.0% .0% T3a to T3b 8 7 1 100.0% 87.5% 12.5% Total 41 Lobe occurence Both Lobes 62 60 2 1.000 100.0% 96.8% 3.2% Singles Lobe 9 9 0 100.0% 100.0% .0% Total 71 Maximum tumour dimension <2.5cm 35 34 1 1.000 100.0% 97.1% 2.9% ≥2.5cm 37 36 1 100.0% 97.3% 2.7% Total 72 Tumour Volume <4cm3 28 27 1 1.000 100.0% 96.4% 3.6% ≥4 cm3 31 30 1 100.0% 96.8% 3.2% Total 59 Page | 203 Chapter 5 Heparan sulphate in prostate cancer Peritumoural Pathological T staging Stroma T2a to T2c 34 7 27 0.752 100.0% 20.6% 79.4% T3a to T3c 34 5 29 100.0% 14.7% 85.3% Total 68 Extraprostatic extension Yes 34 5 29 0.752 100.0% 14.7% 85.3% No 34 7 27 100.0% 20.6% 79.4% Total 68 Seminal vesicle involvement Involved 7 1 6 1.000 100.0% 14.3% 85.7% Uninvolved 64 11 53 100.0% 17.2% 82.8% Total 71 Pre-operative PSA levels ≤8ng/ml 19 3 16 0.661 100.0% 15.8% 84.2% >8ng/ml 20 2 18 100.0% 10.0% 90.0% Total 39 Lymphovascular invasion Present 9 2 7 0.654 100.0% 22.2% 77.8% Absent 59 10 49 100.0% 16.9% 83.1% Total 68 Perineural invasion Seen 60 13 47 0.107 100.0% 21.7% 78.3% Not seen 12 0 12 100.0% .0% 100.0% Total 72 Circumferential resection margin Involved 33 6 27 1.000 100.0% 18.2% 81.8% Uninvolved 38 7 31 100.0% 18.4% 81.6% Total 71 Page | 204 Chapter 5 Heparan sulphate in prostate cancer Apex margin Involved 34 7 27 0.762 100.0% 20.6% 79.4% Uninvolved 37 6 31 100.0% 16.2% 83.8% Total 71 Proximal margin Involved 8 2 6 0.629 100.0% 25.0% 75.0% Uninvolved 64 11 53 100.0% 17.2% 82.8% Total 72 Clinical stage T1a to T1c 32 2 30 0.046 100.0% 6.3% 93.8% T3a to T3b 8 3 5 100.0% 37.5% 62.5% Total 40 Lobe occurence Both Lobes 61 11 50 0.670 100.0% 18.0% 82.0% Singles Lobe 9 2 7 100.0% 22.2% 77.8% Total 70 Maximum tumour dimension <2.5cm 34 7 27 0.762 100.0% 20.6% 79.4% ≥2.5cm 37 6 31 100.0% 16.2% 83.8% Total 71 Tumour Volume <4cm3 28 8 20 0.197 100.0% 28.6% 71.4% ≥4 cm3 31 4 27 100.0% 12.9% 87.1% Total 59 Page | 205 Chapter 5 Heparan sulphate in prostate cancer 5.2 Gene transcript analysis of heparan sulphate biosynthesis genes in prostate cancer cell lines To understand the expression profiles of the enzymes responsible for heparan sulphate biosynthesis in the prostate cell lines, RWPE-1, LNCaP and PC-3, quantitative real-time polymerase chain reaction (qPCR) was performed using a series of primers targeting members of the heparan sulphate biosynthetic pathway. Relative fold changes of the heparan sulphate biosynthetic genes differentially expressed in the prostate cancer cell lines, LNCaP and PC-3, as compared to the normal prostate epithelial cell lines, RWPE-1, are shown in Table 5.6. It was observed that the heparan sulphate editing enzymes such as the N-deacetylase/N-sulphotransferases, heparan sulphotransferases, sulphatases and heparanases were differentially expressed in the prostate adenocarcinoma cell lines. Heparan sulphate proteoglycans such as glypican 4 is upregulated in both prostate cancer cell lines, while the other glypican family members and perlecan (HSPG2) are not differentially changed in the same manner in both cancer cell lines. Amongst the heparan sulphotransferase differentially expressed in the prostate cancer cell lines, the HS3ST3A1 and HS6ST2 genes were found to be downregulated in both LNCaP and PC-3. From the microarray study performed using human prostate adenocarcinoma tissues in Chapter 3, HS3ST3A1 and HS6ST2 were also found to be downregulated in prostate adenocarcinoma. As such, these evidences point to a potential role of HS3ST3A1 and HS6ST2 in prostate tumour cell behaviour. Silencing studies in the normal prostate epithelial cell line, RWPE-1, was performed using double-stranded RNA oligonucleotides, to determine whether HS3ST3A1 and HS6ST2 are potential tumour suppressor genes. Page | 206 Chapter 5 Heparan sulphate in prostate cancer Table 5.6 Relative fold changes of genes coding for heparan sulphate biosynthesis enzymes differentially expressed [1] in prostate cancer cell lines, LNCaP (n=3) and PC-3 (n=3) as compared to normal prostate epithelial cell line, RWPE-1 (n=3). (using β-actin as normalizer); P- values obtained by performing unpaired T-test and [2] in prostate adenocarcinoma tissues (n=6) compared to BPH tissues (n=6); relative fold change and p-values listed is from dCHIP. Gene Gene Name Prostate cancer cell lines Prostate Symbol adenocarcinoma tissues LNCaP PC-3 Microarray Result Relative fold p-value Relative fold p-value Relative p-value Change Change fold Change EXT1 exostoses (multiple) 1 -1.30 0.347 -1.15 0.761 -1.58 >0.05 EXT2 exostoses (multiple) 2 -4.17 0.403 -6.25 0.185 -1.22 >0.05 EXTL1 exostoses (multiple)-like 1 -1.75 0.05 -1.67 0.054 1.06 >0.05 EXTL2 exostoses (multiple)-like 2 6.77 0.004 -1.27 0.753 1.03 >0.05 EXTL3 exostoses (multiple)-like 3 4.8 0.001 1.30 0.232 -1.03 >0.05 NDST1 N-deacetylase/N-sulphotransferase (heparan -2.27 0.02 -2.17 0.001 -1.18 >0.05 glucosaminyl) 1 NDST2 N-deacetylase/N-sulphotransferase (heparan -1.19 0.618 3.40 0.027 1.02 >0.05 glucosaminyl) 2 NDST3 N-deacetylase/N-sulphotransferase (heparan upregulated b <0.05 upregulated b <0.05 -1.03 >0.05 glucosaminyl) 3 NDST4 N-deacetylase/N-sulphotransferase (heparan -1.52 0.515 -2.13 0.203 -1.20 >0.05 glucosaminyl) 4 PAPSS1 3'-phosphoadenosine 5'-phosphosulphate 1.13 0.883 -1.85 0.395 1.17 >0.05 synthase 1 Page | 207 Chapter 5 Heparan sulphate in prostate cancer PAPSS2 3'-phosphoadenosine 5'-phosphosulphate 1.42 0.195 1.765 0.063 1.57 >0.05 synthase 2 HS2ST1 heparan sulphate 2-O-sulphotransferase 1 4.10 0.005 -1.41 0.221 1.27 >0.05 HS3ST1 heparan sulphate (glucosamine) 3-O- not detectedc <0.05 -3.57 0.001 1.18 >0.05 sulphotransferase 1 HS3ST2 heparan sulphate (glucosamine) 3-O- not detectedc <0.05 -11.1 0.20 2.30 >0.05 sulphotransferase 2 HS3ST3A heparan sulphate (glucosamine) 3-O- -50.00 <0.001 -5.00 0.002 -2.23 <0.05 1 sulphotransferase 3A1 HS3ST3B heparan sulphate (glucosamine) 3-O- -33.33 <0.001 -100.0 <0.001 -2.46 <0.05 1 sulphotransferase 3B1 HS3ST4 heparan sulphate (glucosamine) 3-O- N.D a 1.06 >0.05 sulphotransferase 4 HS3ST5 heparan sulphate (glucosamine) 3-O- -2.78 0.880 352.1 0.01 1.17 >0.05 sulphotransferase 5 HS6ST1 heparan sulphate 6-O-sulphotransferase 1 1.78 0.069 -1.61 0.003 -1.18 >0.05 HS6ST2 heparan sulphate 6-O-sulphotransferase 2 -11.11 0.003 -2.33 0.007 -3.63 <0.05 HS6ST3 heparan sulphate 6-O-sulphotransferase 3 upregulation b <0.05 upregulationb <0.05 1.30 >0.05 SULF1 sulphatase 1 not detectedc <0.05 -45.45 <0.0001 1.08 >0.05 SULF2A sulphatase 2 isoform variant 1 -111.11 0.001 -1.14 0.706 2.79 <0.05 SULF2B sulphatase 2 isoform variant 2 -28.57 0.0003 -1.10 0.76 HPSE heparanase -33.33 0.004 -25.00 <0.0001 1.41 >0.05 HPSE2 heparanase 2 N.D a -1.39 >0.05 HSPG2 heparan sulphate proteoglycan 2 -8.33 <0.0001 1.61 0.086 -1.65 >0.05 GPC1 glypican 1 1.21 0.503 -4.76 0.005 -2.41 >0.05 GPC2 glypican 2 -1.09 0.669 -3.57 0.0003 -1.06 >0.05 Page | 208 Chapter 5 Heparan sulphate in prostate cancer GPC3 glypican 3 20.0 0.003 -10.87 0.003 -1.49 >0.05 GPC4 glypican 4 967 0.001 165 0.004 -1.90 >0.05 GPC5 glypican 5 N.D a -1.35 >0.05 GPC6 glypican 6 -16.67 0.005 3.72 0.01 -1.21 >0.05 aExpression not detected in all 3 prostate cell lines within 45 cycles of qPCR bExpression not detected in RWPE-1 but detected in LNCaP and PC-3, within 45 cycles of qPCR. cExpression not detected within 45 cycles of qPCR. Page | 209 Chapter 5 Heparan sulphate in prostate cancer 5.3 Functional Analysis of HS3ST3A1 in prostate cancer 5.3.1 HS3ST3A1 is effectively silenced in RWPE-1 To investigate the functional significance of HS3ST3A1 downregulation in prostate cancer, HS3ST3A1 was silenced in normal prostate epithelial cell, RWPE-1, using Silencer double-stranded siRNA oligonucleotides from Ambion Technologies, accompanied by Oligofectamine as the delivery reagent. HS3ST3A1 was 87% silenced as compared to the non-targeting siRNA control. (Figure 5.4A) Transfection efficiency was determined to be 100% by using Cy3-conjugated negative control siRNA. (Figure 5.4B) Unspecific silencing of HS3ST3B1, a close homolog of HS3ST3A1, was checked and found HS3ST3B1 to be unsilenced in HS3ST3A1- silenced RWPE-1 cells (p=0.2). 5.3.2 HS3ST3A1 silencing inhibited RWPE-1 proliferation Cell proliferation is performed by using the CellTiter 96 AQueous One Solution Cell Proliferation Assay (Promega Corporation, Madison, Wisconsin, USA) that contains a tetrazolium compound [3-(4,5-dimethylthiazol-2-yl)-5-(3- carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H-tetrazolium, inner salt; MTS] which is bioreduced by NADPH or NADH produced by metabolically active cells via dehydrogenase enzymes into a colored formazan product. [Barltrop et al, 1991] Measurement of absorbance of the formazan product at 490nm [Cory et al, 1991] revealed inhibition of cell proliferation by about 37% after silencing HS3ST3A1 in RWPE-1 cells, 72 hours post-transfection as compared to non-targeting siRNA controls. (Figure 5.4C) Page | 210 Chapter 5 Heparan sulphate in prostate cancer 5.3.3 HS3ST3A1 silencing inhibited RWPE-1 migration Acquisition of migration and invasion ability into the basement membrane and extracellular matrix are phenotypes of malignant cells. To investigate whether the downregulation of HS3ST3A1 in a normal prostate epithelial cell line would result in progression to malignancy, the migratory ability of the RWPE-1 cell line was tested after silencing HS3ST3A1, using the transwell migration chamber. It was found that migration of RWPE-1 was inhibited by 26% after HS3ST3A1 silencing, at 72 hours post transfection, as compared to non-targeting siRNA control. (Figure 5.4D) 5.3.4 HS3ST3A1 silencing did not affect RWPE-1 adhesion to collagen type I Proteoglycans are known to mediate cell-cell adhesion, cell adhesion to the basement membrane and extracellular matrix. Such complex adhesion capacities are important components of how a malignant cell migrates and invades into the surrounding tissue as well. The effects of HS3ST3A1 suppression on the adhesion ability of RWPE-1 cells was investigated, and revealed that HS3ST3A1 silencing did not affect RWPE-1 adhesion to collagen type I. (Figure 5.4E) Page | 211 Chapter 5 Heparan sulphate in prostate cancer Figure 5.4 Effects of HS3ST3A1 silencing on RWPE-1 cell behaviour. [A] Silencing efficiency of HS3ST3A1 siRNA is 87%, with no unspecific silencing of HS3ST3B1. [B] Transfection efficiency is 100%, using Cy3-tagged negative control siRNA. [C] HS3ST3A1 inhibited RWPE-1 proliferation. [D] HS3ST3A1 inhibited RWPE-1 motility. [E] HS3ST3A1 has no effect on RWPE-1 adhesion to collagen type I. Non-targeting silenced RWPE-1 (n=3); HS3ST3A1 silenced RWPE-1 (n=3). ‘***’ represents p <0.001; Statistical test was performed using unpaired T-test. Page | 212 Chapter 5 Heparan sulphate in prostate cancer 5.4 Functional Analysis of HS6ST2 in prostate cancer 5.4.1 HS6ST2 is effectively silenced in RWPE-1 To investigate the functional significance of HS6ST2 downregulation in prostate cancer, HS6ST2 was silenced in normal prostate epithelial cell, RWPE-1, using Silencer double-stranded siRNA oligonucleotides from Ambion Technologies and Oligofectamine as the delivery reagent. HS6ST2 was 87% silenced as compared to the non-targeting siRNA control. (Figure 5.5A) Transfection efficiency is the same as Section 5.3.1 as transfection conditions are exactly the same. Unspecific silencing of HS6ST1, an isoform of HS6ST2, was checked and found to be unsilenced (p=0.2). Another isoform member, HS6ST3 was found to be inherently expressed at a low level/not expressed in normal RWPE-1 cells (not detectable within 45 cycles of qPCR) and hence, unspecific silencing of HS6ST3 in HS6ST2-silenced RWPE-1 cells were not determined. 5.4.2 HS6ST2 silencing prevented the proliferation of RWPE-1 cells Proteoglycans are known to mediate cell proliferation through binding with several growth factors and acting as co-receptors. The downregulation of HS6ST2 in prostate cancer cells may allow normal epithelial prostate cells to acquire cancerous phenotypes, such as dysregulation in the control of cell proliferation. As such, the effects of HS6ST2 silencing in normal prostate epithelial cells, RWPE-1, was determined and found that contrary to our hypothesis, HS6ST2 silencing inhibited proliferation of the RWPE-1 cells by about 19%, at 72 hours post transfection (Figure 5.5B). Page | 213 Chapter 5 Heparan sulphate in prostate cancer 5.4.2 HS6ST2 silencing reduced the migratory ability of RWPE-1 cells Next, the migratory ability of the RWPE-1 cell line was tested after silencing HS6ST2, using the transwell migration chamber. It was found that migration of RWPE-1 was inhibited by 56% after HS6ST2 silencing, at 72 hours post transfection, as compared to non-targeting siRNA control (Figure 5.5C). 5.4.3 HS6ST2 silencing did not affect RWPE-1 adhesion to collagen type I Similarly, I wanted to understand whether HS6ST2 is important in the adhesion of RWPE-1 cells to the ECM/BM, since adhesion capacities are important components in the migration and invasion of malignant cells during metastasis as well. It was revealed that HS6ST2 downregulation did not affect the adhesion ability of RWPE-1 cells to collagen type I (Figure 5.5D). Page | 214 Chapter 5 Heparan sulphate in prostate cancer Figure 5.5 Effects of HS6ST2 silencing on RWPE-1 cell behaviour. [A] Silencing efficiency of HS6ST2 is 87%, with no unspecific silencing of HS6ST1. HS6ST3 was inherently found to be not expressed/expressed at low levels in RWPE-1 (not detectable within 45 cycles of qPCR), hence checking of unspecific silencing of HS6ST3 was not performed. [B] HS6ST2 silencing inhibited RWPE-1 cell proliferation. [C] HS6ST2 silencing impeded RWPE-1 motility. [D] HS6ST2 silencing did not appear to have any effect on RWPE-1 adhesion to collagen type I. Non- targeting silenced RWPE-1 (n=3); HS6ST2 silenced RWPE-1 (n=3). ‘***’ represents p <0.001; Statistical test was performed using unpaired T-test. Page | 215 Chapter 5 Heparan sulphate in prostate cancer 5.5 Discussion 5.5.1 Utility of 10E4-specific heparan sulphate as a prostate cancer biomarker Heparan sulphates are a group of structurally diverse molecules that can participate in many important cellular processes. In this study, monoclonal 10E4, which recognizes mixed HS domains containing both N-acetylated and N-sulphated disaccharide units were used [van den Born J et al, 2005]. Using this monoclonal 10E4, it was demonstrated that specific heparan sulphate epitopes recognized by this antibody, was not differentially expressed in the prostate adenocarcinoma tissues as compared to the benign hyperplastic prostate tissues, both in the cytoplasm of the adenocarcinoma cells and the peritumoural stroma. In addition, there were no significant differences in 10E4 staining in the low grade cancers (Gleason sum<6) as compared to high grade cancers (Gleason sum >8). Also, it was demonstrated that strong 10E4 staining is associated with higher clinical staging; however, 10E4 is not associated with other clinicopathological parameters such as resection margin including distal and proximal margin involvement, pathological T staging, extraprostatic extension, seminal vesicle involvement, perineural invasion, lymphovascular invasion, tumor volume and pre- operative PSA levels. Comparatively, detection of heparan sulphate staining using 10E4 in prostate cancer did not seem to correlate well with clinicopathological parameters as compared to chondroitin sulphate staining in prostate cancer. Chondroitin sulphate staining, as discussed in Chapter 4, correlated with many parameters of poor outcome, and is a potential biomarker of prostate cancer prognosis, and from our study, appears to be a better biomarker than the heparan sulphates. Indeed, reports of heparan sulphate expression in prostate adenocarcinoma and their potential use as a biomarker of Page | 216 Chapter 5 Heparan sulphate in prostate cancer prostate cancer prognosis is scarce, as compared to reports on chondroitin sulphate in prostate cancer. 5.5.2 HS3ST3A1 and HS6ST2 are not tumour suppressors in prostate cancer. Heparan sulphate can be sulphated at various positions on the N- acetylglucosamine (C3 position and C6 position) and uronic acid (C2 position). Sulphation is performed by different sulphotransferases from mainly 3 different families that include heparan sulphate-2-O-sulphotransferases, heparan sulphate-3-O- sulphotransferases and heparan sulphate-6-O-sulphotransferases. The sulphation pattern of the glycosaminoglycans and proteoglycans are important, as they encode molecular diversity and elicit different biologies [Gama et al, 2006]. There are six family members belonging to the heparan sulphate 3-O- sulphotransferases. They include HS3ST1, HS3ST2, HS3ST3A1, HS3ST3B1, HS3ST4 and HS3ST5 [Shworak et al, 1999; Mochizuki et al, 2003]. These proteins are comprised of a divergent amino-terminal region and a very homologous carboxyl- terminal sulphotransferase domain of approximately 260 residues. Southern analyses suggest the HS3ST1, HS3ST2, and HS3ST4 genes are single copy while HS3ST3A1 and HS3ST3B1 genes are each duplicated in humans. Each isoform has different, unique substrate specificities from one another [Liu et al, 1999; Xia et al, 2002; Chen et al, 2003], thus generating defined saccharide sequences in structurally complex HS polysaccharide. In my study, I have observed differential expression of the HS-3OST isoforms in prostate adenocarcinoma tissues as compared to BPH tissues, as well as in prostate cancer cell lines as compared to normal prostate epithelial cell lines. However, the same isoform might not necessarily be differentially expressed in the same manner in both prostate tissues and prostate cancer cell lines: HS3ST1 is downregulated in Page | 217 Chapter 5 Heparan sulphate in prostate cancer LNCaP and PC3 but there was no significant change in prostate adenocarcinoma tissues, HS3ST2 is suppressed in both LNCaP and PC3 but upregulated in prostate adenocarcinoma tissues (albeit no statistical significance) and HS3ST5 is downregulated in LNCaP, upregulated in PC3 but not differentially expressed in prostate cancer tissues compared to BPH tissues. These could be due to different tissue/cell types, as the prostate adenocarcinoma tissues contain a heterogeneous mixture of cell populations while the prostate cancer cell lines represent a homogenous cell population originating from the epithelial cells of the secretory glands. Alternatively, these could also suggest the inherent heterogeneous nature of different prostate tumours, with different genetic makeup and instability. HS3ST4 is not detectable in prostate cancer cell lines and not differentially expressed in prostate cancer tissues compared to BPH tissues. On the other hand, both HS3ST3A1 and HS3ST3B1 were consistently found downregulated in prostate adenocarcinoma tissues as well as in prostate cancer cell lines. HS3ST3A1 was downregulated by 2.2 fold in prostate cancer tissue, 50 fold in LNCaP and 5 fold in PC3 while HS3ST3B1 was downregulated by 2.5 fold in cancer tissue, 33 fold in LNCaP and 100 fold in PC3. The HS3ST3A1 and HS3ST3B1 are integral membrane proteins, consisting of conserved carboxyl-terminal sulphotransferase domain and the divergent amino- terminal regions. Intriguingly, the entire sulphotransferase domain sequence of the HS3ST3B1 cDNA (774 base pairs) was 99.2% identical to the same sulphotransferase region of HS3ST3A1 [Liu et al, 1999]. Also, HS3ST3A1 and HS3ST3B1 were reported to sulphate and generate identical 3-O-sulphated disaccharides; both enzymes transfer sulphate to the 3-OH position of the glucosamine within IdoA2S-GlcNS or IdoA2S- GlcNH2 [Liu et al, 1999]. Thus, sulphation specificity corresponds to the nearly identical sulphotransferase domains and is not perturbed by the unique amino terminal Page | 218 Chapter 5 Heparan sulphate in prostate cancer regions of each sulphotransferase. As such, the unique amino-terminal regions of HS3ST3A1 and HS3ST3B1 may contribute distinctive biologic roles to the nearly identical sulphotransferase domains. In fact, the cytoplasmic tail at the amino terminal of HS3ST3B1contains a polyproline tract, which are known to form a rigid left- handed-helix where such motifs are critical elements bound by protein interaction modules such as SH3 and WW domains. Protein-protein interactions within the amino-terminal regions may regulate the spatial organization or functional interaction of the enzymes, thus contributing to the different biologic activities of the different enzymes [Feng et al, 1994; Chen et al, 1997; Rau et al, 1995]. Actual delineation of the functional significance between HS3ST3A1 and HS3ST3B1 requires future investigation. Nevertheless, downregulation of both HS3ST3A1 and HS3ST3B1 in the prostate adenocarcinoma tissues and in prostate cancer cell lines indicates that heparan sulphate sequences containing specific 3-O-sulphated disaccharides coded by HS3ST3A1 and HS3ST3B1 are suppressed in prostate cancer, since HS3ST3A1 and HS3ST3B1 generate identical 3-O-sulphated disaccharides. This specific disaccharide sequence in heparan sulphate or heparan sulphate proteoglycans could be functionally important in prostate tumourigenesis, and thus warrants further investigation. In addition, it was found that the heparan sulphate 6-O-sulphotransferase (HS6ST) members are also dysregulated in prostate adenocarcinoma. HS6ST catalyze the transfer of sulphate from 3'-phosphoadenosine 5'-phosphosulphate (PAPS) to position 6 of the N-sulphoglucosamine residue of heparan sulphate. The HS6ST family comprised of three isoforms and one alternatively spliced form [Habuchi et al, 2003]. Individual isoforms showed characteristic preference for the uronic acid residue neighboring the N-sulphoglucosamine. [Habuchi et al, 2000] HS6ST1 Page | 219 Chapter 5 Heparan sulphate in prostate cancer predominantly sulphated the 6-OH position of GlcNSO3 in HexA-GlcNSO3 unit. On the other hand, HS6ST2 preferably transferred sulphate to the 6-OH position of GlcNSO3 in IdoA(2SO4)-GlcNSO3 unit or highly sulphated regions. A splice variant of HS6ST2, which is 40 amino acids shorter, termed HS6ST2-S has substrate specificity somewhat intermediate between HS6ST1 and HS6ST2 [Habuchi et al, 2003]. In this study, only HS6ST2 was found downregulated by 3.5 fold in prostate adenocarcinoma tissue and suppressed by 11 fold in LNCaP and by 2.3 fold in PC3, as compared to BPH tissues and normal prostate epithelial cell, RWPE-1 respectively. HS6ST1 and HS6ST3 were both not found differentially expressed in prostate adenocarcinoma tissues as compared to BPH tissues, but were found differentially expressed in prostate cancer cell lines; HS6ST1 was downregulated by 1.6 fold in PC3 only and HS6ST3 were found expressed in both prostate cancer cells, LNCaP and PC3, while it was not detectable within 45 cycles of amplification through real-time PCR in RWPE-1. Again, this could be due to differences in the heterogeneous nature of tissue sections and homogeneous cell populations using prostate cancer cell lines, as well as possibly due to tumour heterogeneity. Since HS6ST2 expression is suppressed in both prostate adenocarcinoma tissues and prostate cancer cell lines, HS6ST2 could be a tumour suppressor that allows prostate cancer acquisition after its suppression. Surprisingly, I discovered that silencing HS3ST3A1 or HS6ST2 in normal prostate epithelial cells, RWPE-1, did not appear to acquire cancerous phenotypes; in contrary, both HS3ST3A1-silenced RWPE-1 cells and HS6ST2-silenced RWPE-1 cells showed decreased proliferative capacity and decreased cell motility. These findings demonstrated that HS3ST3A1 and HS6ST2 are not tumour suppressors, but are important in normal cellular functions. Page | 220 Chapter 5 Heparan sulphate in prostate cancer Indeed, several studies have reported on the importance of heparan sulphate in mediating cellular processes such as growth and apoptosis. Recently, Kobayashi et al, 2010, have shown that HS6ST2 inhibition in chick limb region resulted in limb bud truncation [Kobayashi et al, 2010]. Todorovicc et al, 2005, have showed that the heparan sulphate proteoglycan, syndecan-4, together with matrix proteins and integrin receptors, triggers apoptosis in fibroblasts via p53 activation of Bax [Todorovicc et al, 2005]. These findings explain the importance of heparan sulphate and their biosynthetic enzymes in normal cellular function. Many regulatory mechanisms appear to control the expression of heparan sulphate too. Most recently, Chau et al, 2009, reported that perturbation of p53 directly influences SULF2, an extracellular heparan sulphate 6-O-endosulphatase that regulates growth factor binding to their cognate receptors [Chau et al, 2009]. p53 has been reported to directly regulate heparanase expression [Baraz et al, 2006]. These studies have revealed that timely expression of heparan sulphate are critical for normal functioning of cellular processes, and regulatory mechanisms are in place to tightly regulate the expression of these glycosaminoglycans. Although it has not been reported on the direct regulation of p53 on HS3ST3A1 and/or HS6ST2, given the observation that p53 directs the regulation of many genes including CS/HS modifying genes and CSPG/HSPG, it suggests then that perhaps, the biological functions of HS/CS and HSPG/CSPG could also be governed by p53. Undeniably, the structural diversity of heparan sulphates and their proteoglycans encompass a huge array of possible dysregulations in diseases, resulting in specific functional repercussions. In prostate cancer, much work can be further performed to determine specific HS structures that are differentially expressed by utilising different antibodies that recognize different heparan sulphate structures. Page | 221 Chapter 5 Heparan sulphate in prostate cancer In addition, functional studies investigating the downregulation of HS6ST2 and HS3ST3A1 in prostate cancer, as compared to normal prostate cells, can be performed by overexpressing these genes in cancer cell lines to determine if they are important in regulating tumour cell phenotypes. Page | 222 Chapter 6 Conclusions and future work Chapter 6 Conclusions and Future Work 6.1 Identification of potential biomarkers in prostate cancer Heterogeneity in disease manifestations of prostate cancer, in which some prostate cancers never become life threatening while some progress to metastatic forms, poses a huge prognostic and treatment challenge. Identifying biomarkers of clinical outcome have the potential to greatly improve the management of prostate cancer. In this study, genome-wide expression profiles of prostate adenocarcinoma tissues were compared against BPH tissues, and it was revealed that 865 genes involved in development and differentiation, cell motility, cell proliferation, cellular response to stress, homeostasis, cell communication, cell cycle, apoptosis and metabolism were aberrantly expressed by at least 2 fold in the cancerous tissues. This represents a disarray of genetic networks that could contribute to or arise from the cancer cell phenotype. Nonetheless, this gene list provides an excellent avenue to investigate the potential application of some of these molecules as biomarkers of prostate cancer prognosis, and also to identify potential novel therapeutic targets. Chondroitin sulphate and heparan sulphate are ubiquitously present on the cellular surface and in the ECM, and are hence well positioned to elicit many cell-cell, cell-ECM signalling mechanisms, regulating important cellular behaviour. Of interest to this study is the application of chondroitin sulphate and heparan sulphate as possible biomarkers of prostate cancer. Since chondroitin sulphate and heparan sulphate are structurally diverse molecules with different sulphate modifications, four different chondroitin sulphate subtypes – non-sulphated chondroitin, chondroitin-4- sulphate, chondroitin-6-sulphate and chondroitin-2,6-sulphate and one heparan Page | 223 Chapter 6 Conclusions and future work sulphate epitope that consists of mixed N-sulphated and N-acetylated regions, were explored in prostate adenocarcinoma tissues. Non-sulphated chondroitin and chondroitin-4-sulphate were upregulated in prostate adenocarcinoma cells as compared to paired benign epithelial secretory cells of the benign prostate hyperplastic tissues; in addition, expression of chondroitin-4- sulphate was downregulated but chondroitin-6-sulphate expression was upregulated in peritumoural stroma as compared to periglandular stroma. In addition, we found that with increasing tumour grades, there was increasing expression of all chondroitin sulphate subtypes in the peritumoural stroma, with the highest expression in high grade cancers (Gleason sum >8). These results represent a potential role of chondroitin sulphate in prostate tumourigenesis and possibly, spatiotemporal roles during prostate cancer progression to malignancy. Heparan sulphate expression was determined in prostate adenocarcinoma tissues using the monoclonal 10E4 that recognizes ‘mixed’ N-acetylated and N- sulphated domains of HS. By using this monoclonal antibody, no significant changes in heparan sulphate expression were detected in prostate adenocarcinoma tissues as compared to paired benign prostate hyperplastic tissues. Nevertheless, given the huge repertoire of heparan sulphate structures, other heparan sulphate epitopes can be investigated to determine whether they are differentially expressed in prostate cancer. Association studies between expression of chondroitin sulphate or heparan sulphate with clinicopathological parameters was also performed. It was found that high expression of non-sulphated chondroitin, chondroitin-4-sulphate, chondroitin-6- sulphate and chondroitin-2,6-sulphate in the cancer cells and in cancer stroma correlated with pathological T stage, extraprostatic extension, perineural invasion, Page | 224 Chapter 6 Conclusions and future work seminal vesicle involvement, tumour volume, clinical staging, pre-operative PSA levels and occurrence of cancer in both prostate lobes. All these pathologic parameters augur a worse prognosis for patients, and highlight the association of chondroitin sulphate expression with conventional adverse predictors. On the other hand, heparan sulphate staining as detected by monoclonal 10E4 did correlate with only clinical staging, but not with the other parameters. Comparatively, this shows that chondroitin sulphate could be a better predictive biomarker than heparan sulphate in prostate cancer outcome; however, we could first determine other heparan sulphate epitopes expression in prostate cancer and evaluate their use as potential biomarkers before reaching a definitive conclusion. To determine if chondroitin sulphate will be a good indicator of prostate cancer progression, long term follow up data such as PSA relapse and survival will be important. Future work will include data acquisition of such parameters and subsequent analysis with the current chondroitin sulphate expression patterns. 6.2 Establishing the functional significance of CS/HS and their biosynthesis enzymes in prostate cancer. Identifying new molecular targets could facilitate innovation of novel therapies that could halt progression of localized prostate cancer towards metastasis or that could target metastatic prostate cancer. From the genome wide expression screen, CSGalNAcT-1 was seen to be upregulated in prostate adenocarcinoma tissues. Since CSGalNAcT-1 is responsible for the initiation of CS/DS synthesis and a concomitant upregulation of chondroitin sulphate expression in prostate cancer tissues was observed through immunohistochemistry, I hypothesized that the upregulation of CSGalNAcT-1 and its corresponding increase in chondroitin sulphate expression could be functionally important in prostate cancer. Prostate cancer cell lines, LNCaP Page | 225 Chapter 6 Conclusions and future work and PC3, also demonstrated upregulation of CSGalNAcT-1 as compared to normal prostate epithelial cells, RWPE-1. CSGalNAcT-1 silenced PC3 cells showed inhibition in proliferation and invasion, increase in apoptosis and cell cycle arrest via downregulation of MAPRE-1, TPX2 and HIPK3. These results highlighted CSGalNAcT-1 as a pro-tumour protein and revealed CSGalNAcT-1 as a potential therapeutic target. Future work however, could be done to investigate whether CSGalNAcT-1 elicits cell signalling via direct effects on the downstream genes or it was performed via indirect effects, through the CS/DS chains that were reduced in expression as a result of CSGalNAcT-1 silencing. Heparan sulphate is a multi-faceted molecule contributed by its diverse structures, which is facilitated by different sulphate modifications. Heparan sulphate sulphotransferase enzymes, HS3ST3A1 and HS6ST2, were shown to be downregulated in prostate adenocarcinoma tissues compared to BPH tissues, and also revealed to be downregulated in prostate cancer cell lines, LNCaP and PC3, compared to normal prostate epithelial cell lines, RWPE-1. The hypothesis that these enzymes could be potential tumour suppressors in prostate cancer is made. HS3ST3A1 and HS6ST2 were silenced in normal prostate epithelial cells, RWPE-1. HS3ST3A1-silenced RWPE-1 and HS6ST2-silenced RWPE-1 both showed inhibition in proliferation and motility. These results showed that HS3ST3A1 and HS6ST2 are important enzymes in normal prostate physiology and also represent that HS3ST3A1 and HS6ST2 are not key tumour suppressor genes. However, we cannot disregard the importance of HS3ST3A1 and HS6ST2 suppression in prostate cancer as they could be functionally important in prostate tumour progression only after other anti-cancer signals are lost. Future work could include overexpression of HS3ST3A1 and HS6ST2 in prostate cancer cells so as to understand the roles of HS6ST2 and HS3ST3A1 in a tumour environment. Page | 226 Chapter 6 Conclusions and future work All in, the importance of chondroitin sulphate and heparan sulphate in normal prostate physiology and prostate cancer tumourigenesis has been revealed. Essentially, the potential role of chondroitin sulphate as a predictive biomarker and CSGalNAcT-1 as a novel therapeutic target has been highlighted. Due to the large structural and functional repertoire of chondroitin sulphate and heparan sulphate, this necessitates the need to elucidate the exact structural entity that is important in prostate cancer progression. Certainly, this remains challenging yet promising in the field of prostate cancer research. Page | 227 Chapter 7 References Chapter 7 References Aberle H, Bauer A, Stappert J, Kispert A, Kemler R. (1997) b-catenin is a target for the ubiquitin–proteasome pathway. EMBO J 16:3797–3804 Adany R, Iozzo RV. (1991) Hypomethylation of the decorin proteoglycan gene in human colon cancer. Biochem J 276:301-306 Affymetrix, Inc. “GeneChip® Expression Analysis Data Analysis Fundamentals Manual” [Internet] Available from: http://www.affymetrix.com/support/downloads/manuals/data_analysis_fundamentals_ manual.pdf. [cited August 2007] Aikawa J, Grobe K, Tsujimoto M, Esko JD. (2001) Multiple isozymes of heparan sulfate/heparin GlcNAc N-deacetylase/GlcN N-sulfotransferase. Structure and activity of the fourth member, NDST4. J Biol Chem 276:5876-5882 Akiri G, Cherian MM, Vijayakumar S, Liu G, Bafico A, Aaronson SA. (2009) Wnt pathway aberrations including autocrine Wnt activation occur at high frequency in human non-small-cell lung carcinoma. Oncogene 28:2163-2172 Almeida R, Levery SB, Mandel U, Kresse H, Schwientek T, Bennett EP, Clausen H. (1999) Cloning and expression of a proteoglycan UDP-galactose:beta-xylose beta1,4- galactosyltransferase I. A seventh member of the human beta4-galactosyltransferase gene family. J Biol Chem 274:26165-26171 Andrews PD. (2005) Aurora kinases: shining lights on the therapeutic horizon? Oncogene 24: 5005–5015 Ashikari-Hada S, Habuchi H, Sugaya N, Kobayashi T, Kimata K. (2009) Specific inhibition of FGF-2 signaling with 2-O-sulfated octasaccharides of heparan sulfate. Glycobiology 19:644-654 Page | 228 Chapter 7 References Ashikari-Hada S, Habuchi H, Kariya Y, Itoh N, Reddi AH, Kimata K. (2004) Characterization of growth factor-binding structures in heparin/heparan sulfate using an octasaccharide library. J Biol Chem 279:12346-12354 Ayala AG, Ro JY, Babaian R, Troncoso P, Grignon DJ. (1989) The prostatic capsule: does it exist? Its importance in the staging and treatment of prostatic carcinoma. Am J Surg Pathol 13:21-27 Ayala G, Wang D, Wulf G, Frolov A, Li R, Sowadski J, Wheeler TM, Lu KP, Bao L. (2003) The prolyl isomerase Pin1 is a novel prognostic marker in human prostate cancer. Cancer Res 63:6244-6251 Bai X, Zhou D, Brown JR, Crawford BE, Hennet T, Esko JD. (2001) Biosynthesis of the linkage region of glycosaminoglycans: cloning and activity of galactosyltransferase II, the sixth member of the beta 1,3-galactosyltransferase family (beta 3GalT6). J Biol Chem 276:48189-48195 Bandyopadhyay A, Wang L, López-Casillas F, Mendoza V, Yeh IT, Sun L. (2005) Systemic administration of a soluble betaglycan suppresses tumor growth, angiogenesis, and matrix metalloproteinase-9 expression in a human xenograft model of prostate cancer. Prostate 63:81-90 Banerjee AG, Liu J, Yuan Y, Gopalakrishnan VK, Johansson SL, Dinda AK, Gupta NP, Trevino L, Vishwanatha JK. (2003) Expression of biomarkers modulating prostate cancer angiogenesis: differential expression of annexin II in prostate carcinomas from India and USA. Mol Cancer 2:34-38 Baraz L, Haupt Y, Elkin M, Peretz T, Vlodavsky I. (2006) Tumor suppressor p53 regulates heparanase gene expression. Oncogene. 25:3939-3947 Barltrop JA, Owen TC, Cory AH, Cory JG. (1991) 5-(3-carboxymethoxyphenyl)-2- (4,5-dimethylthiazolyl)-3-(4-sulfophenyl)tetrazolium, inner salt (MTS) and related analogs of 3-(4,5-dimethylthiazolyl)-2,5-diphenyltetrazolium bromide (MTT) reducing to purple water-soluble formazans as cell-viability indicators. Bioorganic & Medicinal Chemistry Letters 1:611-614 Page | 229 Chapter 7 References Bello-DeOcampo D, Tindall DJ. (2003) TGF-betal/Smad signaling in prostate cancer. Curr Drug Targets 4:197-207 Ben-Zaken O, Gingis-Velitski S, Vlodavsky I, Ilan N. (2007) Heparanase induces Akt phosphorylation via a lipid raft receptor. Biochem Biophys Res Commun 361:829– 834 Bernards R, Weinberg RA. (2002) A progression puzzle. Nature 418:823 Berrueta L, Kraeft SK, Tirnauer JS, Schuyler SC, Chen LB, Hill DE, Pellman D, Bierer BE. (1998) The adenomatous polyposis coli-binding protein EB1 is associated with cytoplasmic and spindle microtubules Proc. Natl. Acad. Sci. USA 95: 10596– 10601 Bisson I, Prowse DM. (2009) WNT signaling regulates self-renewal and differentiation of prostate cancer cells with stem cell characteristics. Cell Res 19:683- 697 Bostwick DG (1994) Gleason grading of prostatic needle biopsies. Correlation with grade in 316 matched prostatectomies. Am J Surg Pathol 18:796-803 Brooks JD. Anatomy of the lower urinary tract and male genitalia. In: Wein AJ, Kavoussi LR, Novick AC, Partiri AW, Peters CA, editors. Campbell-Walsh Urology. Volume 1. 9th Edition. Philadelphia: Saunders Elsevier; 2007. P38-78. Campbell P, Hannesson HH, Sandback D, Roden L, Lindahl U, Li JP. (1994) Biosynthesis of heparin/heparan sulfate: purification of the D-glucuronyl C-5 epimerase from bovine liver. J Biol Chem 269:26953–26958 Carlsson P, Presto J, Spillmann D, Lindahl U, Kjellén L. (2008) Heparin/heparan sulfate biosynthesis: processive formation of N-sulfated domains. J Biol Chem 283:20008-20014 Catalona WJ, Han M. Definitive therapy for localised prostate cancer – an overview. In: Wein AJ, Kavoussi LR, Novick AC, Partiri AW, Peters CA, editors. Campbell- Walsh Urology. Volume 3. 9th Edition. Philadelphia: Saunders Elsevier; 2007. P2932- 2946 Page | 230 Chapter 7 References Chandran UR, Dhir R, Ma C, Michalopoulos G, Becich M, Gilbertson J. (2005) Differences in gene expression in prostate cancer, normal appearing prostate tissue adjacent to cancer and prostate tissue from cancer free organ donors. BMC Cancer 5:45-56 Chau BN, Diaz RL, Saunders MA, Cheng C, Chang AN, Warrener P, Bradshaw J, Linsley PS, Cleary MA. (2009) Identification of SULF2 as a novel transcriptional target of p53 by use of integrated genomic analyses. Cancer Res. 69:1368-1374 Chen D, Adenekan B, Chen L, Vaughan ED, Gerald W, Feng Z, Knudsen BS. (2004) Syndecan-1 expression in locally invasive and metastatic prostate cancer. Urology 63:402-407 Chen HI, Einbond A, Kwak SJ, Linn H, Koepf E, Peterson S, Kelly JW, Sudol M. (1997) Characterization of the WW domain of human yes-associated protein and its polyproline-containing ligands.J. Biol. Chem. 272, 17070–17077 Chen J, Duncan MB, Carrick K, Pope RM, Liu J. (2003) Biosynthesis of 3-O-sulfated heparan sulfate: unique substrate specificity of heparan sulfate 3-O-sulfotransferase isoform 5. Glycobiology 13:785-794 Chen L, Sanderson RD. (2009) Heparanase regulates levels of syndecan-1 in the nucleus. PLoS ONE 4:e4947 Chen Z, Fan JQ, Li J, Li QS, Yan Z, Jia XK, Liu WD, Wei LJ, Zhang FZ, Gao H, Xu JP, Dong XM, Dai J, Zhou HM. (2009) Promoter hypermethylation correlates with the Hsulf-1 silencing in human breast and gastric cancer. Int J Cancer 124:739-744 Chim CS, Pang R, Liang R. (2008) Epigenetic dysregulation of the Wnt signalling pathway in chronic lymphocytic leukaemia. J Clin Pathol 61:1214-1219 Chung MT, Lai HC, Sytwu HK, Yan MD, Shih YL, Chang CC, Yu MH, Liu HS, Chu DW, Lin YW. (2009) SFRP1 and SFRP2 suppress the transformation and invasion abilities of cervical cancer cells through Wnt signal pathway. Gynecol Oncol 112:646-653 Page | 231 Chapter 7 References Cillo C, Can tile M, Faiella A, Boncinelli E. (2001) Homeobox genes in normal and malignant cells. J Cell Physiol 188:161-169 Cohen-Kaplan V, Doweck I, Naroditsky I, Vlodavsky I, Ilan N. (2008) Heparanase augments epidermal growth factor receptor phosphorylation: correlation with head and neck tumor progression. Cancer Res 68:10077-85 Collins AT, Berry PA, Hyde C, Stower MJ, Maitland NJ. (2005) Prospective identification of tumorigenic prostate cancer stem cells. Cancer Res 65:10946–10951 Collins AT, Habib FK, Maitland NJ, Neal DE. (2001) Identification and isolation of human prostate epithelial stem cells based on α2β1-integrin expression. J Cell Sci 114:3865–3872 Cooper CR, Chay CH, Gendernalik JD, Lee HL, Bhatia J, Taichman RS, McCauley LK, Keller ET, Pienta KJ. (2003) Stromal factors involved in prostate carcinoma metastasis to bone. Cancer 97:739-747 Cory AH, Owen TC, Barltrop JA, Cory JG. (1991) Use of an aqueous soluble tetrazolium/formazan assay for cell growth assays in culture. Cancer Commun. 3:207- 212 Curtin JF and Cotter TG. (2004) JNK regulates HIPK3 expression and promotes resistance to Fas-mediated apoptosis in DU 145 Prostate carcinoma cells. J. Biol Chem 279:17090-17100 Dai Y, Yang Y, MacLeod V, Yue X, Rapraeger AC, Shriver Z, Venkataraman G, Sasisekharan R, Sanderson RD. (2005) HSulf-1 and HSulf-2 are potent inhibitors of myeloma tumor growth in vivo. J Biol Chem 280:40066-40073 Davies EJ, Blackhall FH, Shanks JH, David G, McGown AT, Swindell R, Slade RJ, Martin-Hirsch P, Gallagher JT, Jayson GC. (2004) Distribution and clinical significance of heparan sulfate proteoglycans in ovarian cancer. Clin Cancer Res 10:5178-5186 Page | 232 Chapter 7 References de Luca A, Santra M, Baldi A, Giordano A, Iozzo RV. (1996) Decorin-induced growth suppression is associated with up-regulation of p21, an inhibitor of cyclin- dependent kinases. J Biol Chem 271:18961-18965 de Mestre AM, Rao S, Hornby JR, Soe-Htwe T, Khachigian LM, Hulett MD. (2005) Early growth response gene 1 (EGR1) regulates heparanase gene transcription in tumor cells. J Biol Chem 280:35136-35147 Deepa SS, Yamada S, Zako M, Goldberger O, Sugahara K. (2004) Chondroitin sulfate chains on syndecan-1 and syndecan-4 from normal murine mammary gland epithelial cells are structurally and functionally distinct and cooperate with heparan sulfate chains to bind growth factors. A novel function to control binding of midkine, pleiotrophin, and basic fibroblast growth factor. J Biol Chem 279:37368-37376 Delehedde M, Lyon M, Sergeant N, Rahmoune H, Fernig DG. (2001) Proteoglycans: pericellular and cell surface multireceptors that integrate external stimuli in the mammary gland. J Mammary Gland Biol Neoplasia 6:253-273 Denholm EM., Lin YQ, Silver PJ. (2001) Anti-tumour activities of chondroitinase AC and chondroitinase B: inhibition of angiogenesis, proliferation and invasion. Eur J Pharmacol 416:213–221 Dennis G Jr, Sherman BT, Hosack DA, Yang J, Gao W, Lane HC, Lempicki RA. (2003) DAVID: Database for Annotation, Visualization, and Integrated Discovery. Genome Biol. 4:P3 Desnoyers L, Arnott D, Pennica D (2001) WISP-1 binds to decorin and biglycan. J Biol Chem 276:47599-47607 Dhanasekaran SM, Barrette TR, Ghosh D, Shah R, Varambally S, Kurachi K, Pienta KJ, Rubin MA, Chinnaiyan AM. (2001) Delineation of prognostic biomarkers in prostate cancer. Nature 412:822-826 Diamandis EP. (2000) Prostate-specific antigen: a cancer fighter and a valuable messenger? Clin Chem 46:896–900 Page | 233 Chapter 7 References Dolloff NG, Shulby SS, Nelson AV, Stearns ME, Johannes GJ, Thomas JD, Meucci O, Fatatis A. (2005) Bone-metastatic potential of human prostate cancer cells correlates with Akt/PKB activation by alpha platelet-derived growth factor receptor. Oncogene 24:6848-6854 Doweck I, Kaplan-Cohen V, Naroditsky I, Sabo E, Ilan N, Vlodavsky I. (2006) Heparanase localization and expression by head and neck cancer: correlation with tumor progression and patient survival. Neoplasia 8:1055-1061 Drobnjak M, Osman I, Scher HI, Fazzari M, Cordon-Cardo C. (2000) Overexpression of cyclin D1 is associated with metastatic prostate cancer to bone. Clin Cancer Res 6:1891-1895 Elston MS, Gill AJ, Conaglen JV, Clarkson A, Shaw JM, Law AJ, Cook RJ, Little NS, Clifton-Bligh RJ, Robinson BG, McDonald KL. (2008) Wnt pathway inhibitors are strongly down-regulated in pituitary tumors. Endocrinology 149:1235-1242 Epstein JI, Allsbrook WC Jr, Amin MB, Egevad LL; ISUP Grading Committee. (2005) The 2005 International Society of Urological Pathology (ISUP) Consensus Conference on Gleason Grading of Prostatic Carcinoma. Am J Surg Pathol 29:1228- 1242 Ernst S, Langer R, Cooney CL, Sasisekharan R. (1995) Enzymatic degradation of glycosaminoglycans. Crit Rev Biochem Mol Biol 30:387-444 Ernst T, Hergenhahn M, Kenzelmann M, Cohen CD, Bonrouhi M, Weninger A, Klären R, Gröne EF, Wiesel M, Güdemann C, Küster J, Schott W, Staehler G, Kretzler M, Hollstein M, Gröne HJ. (2002) Decrease and gain of gene expression are equally discriminatory markers for prostate carcinoma: a gene expression analysis on total and microdissected prostate tissue. Am J Pathol 160:2169-2180 Eswarakumar VP, Lax I, Schlessinger J. (2005) Cellular signaling by fibroblast growth factor receptors. Cytokine Growth Factor Rev 16:139-149 Evans T. (2009) Fishing for a WNT-PGE2 link: beta-catenin is caught in the stem cell net-work. Cell Stem Cell 4:280-282 Page | 234 Chapter 7 References Evers MR, Xia G, Kang HG, Schachner M, Baenziger JU. (2001) Molecular cloning and characterisation of a dermatan-specific N-acetylgalactosamine 4-O- sulfotransferase. J Biol Chem 276:36344-36353 Eyers PA, Erikson E, Chen LG, Maller JL. (2003) A novel mechanism for activation of the protein kinase Aurora A. Curr Biol 13:691-697. Feng S, Chen JK, Yu H, Simon JA, Schreiber SL. (1994) Two binding orientations for peptides to the Src SH3 domain: development of a general model for SH3-ligand interactions. Science 266:1241–1247 Fisher ER, Jeffrey W. (1965) Ultrastructure of human normal and neoplastic prostate. Am J Clin Pathol 44:119-134 Fritz TA, Gabb MM, Wei G, Esko JD. (1994) Two N-acetylglucosaminyltransferases catalyze the biosynthesis of heparan sulfate. J. Biol. Chem. 269:28809–28814. Fthenou E, Zafiropoulos A, Tsatsakis A, Stathopoulos A, Karamanos NK, Tzanakakis GN. (2006) Chondroitin sulfate A chains enhance platelet derived growth factor- mediated signalling in fibrosarcoma cells. Int J Biochem Cell Biol 38:2141-2150 Fudge K, Wang CY, Stearns ME. (1994) Immunohistochemistry analysis of platelet- derived growth factor A and B chains and platelet-derived growth factor alpha and beta receptor expression in benign prostatic hyperplasias and Gleason-graded human prostate adenocarcinomas. Mod Pathol 7:549-554 Fukuta M, Kobayashi Y, Uchimura K, Kimata K, Habuchi O. (1998) Molecular cloning and expression of human chondroitin-6-sulphotransferase. Biochim Biophy Acta 1399:57-61 Gama CI, Tully SE, Sotogaku N, Clark PM, Rawat M, Vaidehi N, Goddard WA 3rd, Nishi A, Hsieh-Wilson LC. (2006) Sulphation patterns of glycosaminoglycans encode molecular recognition and activity. Nat Chem Biol 2:467-473 Garber K. (2006) Energy deregulation: licensing tumors to grow. Science 312:1158- 1159 Page | 235 Chapter 7 References Gatenby RA, Gillies RJ. (2004) Why do cancers have high aerobic glycolysis? Nat Rev Cancer 4:891-899 Gauger KJ, Hugh JM, Troester MA, Schneider SS. (2009) Down-regulation of SFRP1 in a mammary epithelial cell line promotes the development of a CD44high/CD24low population which is invasive and resistant to anoikis. Cancer Cell Int 9:11 Gingis-Velitski S, Zetser A, Flugelman MY, Vlodavsky I, Ilan N. (2004) Heparanase induces endothelial cell migration via protein kinase B/Akt activation. J Biol Chem 279:23536–23541 Gleason DF. (1966) Classification of prostatic carcinomas. Cancer Chemother Rep 50:125-128 Glynne-Jones E, Harper ME, Seery LT, James R, Anglin I, Morgan HE, Taylor KM, Gee JM, Nicholson RI. (2001) TENB2, a proteoglycan identified in prostate cancer that is associated with disease progression and androgen independence. Int J Cancer 94:178-184 Godoy-Tundidor S, Cavarretta IT, Fuchs D, Fiechtl M, Steiner H, Friedbichler K, Bartsch G, Hobisch A, Culig Z. (2005) Interleukin-6 and oncostatin M stimulation of proliferation of prostate cancer 22Rv1 cells through the signaling pathways of p38 mitogen-activated protein kinase and phosphatidylinositol 3-kinase. Prostate 64:209- 216 Goessling W, North TE, Loewer S, Lord AM, Lee S, Stoick-Cooper CL, Weidinger G, Puder M, Daley GQ, Moon RT, Zon LI. (2009) Genetic interaction of PGE2 and Wnt signaling regulates developmental specification of stem cells and regeneration. Cell 136:1136-1147. Goldoni S, Iozzo RV. (2008) Tumor microenvironment: Modulation by decorin and related molecules harboring leucine-rich tandem motifs. Int J Cancer 123:2473-2479 Goldshmidt O, Zcharia E, Cohen M, Aingorn H, Cohen I, Nadav L, Katz BZ, Geiger B, Vlodavsky I. (2003) Heparanase mediates cell adhesion independent of its enzymatic activity. FASEB J 17:1015–1025 Page | 236 Chapter 7 References Gotoh M, Sato T, Akashima T, Iwasaki H, Kameyama A, Mochizuki H, Yada T, Inaba N, Zhang Y, Kikuchi N, Kwon YD, Togayachi A, Kudo T, Nishihara S, Watanabe H, Kimata K, Narimatsu H. (2002) Enzymatic synthesis of chondroitin with a novel chondroitin sulfate N-acetylgalactosaminyltransferase that transfers N- acetylgalactosamine to glucuronic acid in initiation and elongation of chondroitin sulfate synthesis. J Biol Chem. 277:38189-38196 Gotoh M, Yada T, Sato T, Akashima T, Iwasaki H, Mochizuki H, Inaba N, Togayachi A, Kudo T, Watanabe H, Kimata K, Narimatsu H. (2002) Molecular cloning and characterization of a novel chondroitin sulfate glucuronyltransferase that transfers glucuronic acid to N-acetylgalactosamine J Biol Chem. 277:38179-38188 Götting C, Kuhn J, Zahn R, Brinkmann T, Kleesiek K. (2000) Molecular cloning and expression of human UDP-d-Xylose:proteoglycan core protein beta-d- xylosyltransferase and its first isoform XT-II. J Mol Biol 304:517-528 Grant DS, Yenisey C, Rose RW, Tootell M, Santra M, Iozzo RV. (2002) Decorin suppresses tumor cell-mediated angiogenesis. Oncogene 21:4765-777 Grant SL, Douglas AM, Goss GA, Begley CG. (2001) Oncostatin M and leukemia inhibitory factor regulate the growth of normal human breast epithelial cells. Growth Factors 19:153-162 Green RA, Wollman R, Kaplan KB. (2005) APC and EB1 Function Together in Mitosis to Regulate Spindle Dynamics and Chromosome Alignment. Molec Biol Cell 16:4609–4622 Gruss OJ and Vernos I. (2004) The mechanism of spindle assembly: functions of Ran and its target TPX2. J Cell Biol 166:949-955 Gruss OJ, Wittmann M, Yokoyama H, Pepperkok R, Kufer T, Silljé H, Karsenti E, Mattaj IW, Vernos I. (2002) Chromosome-induced microtubule assembly mediated by TPX2 is required for spindle formation in HeLa cells. Nat Cell Biol 4:871-879 Gruss OJ, Carazo-Salas RE, Schatz CA, Guarguaglini G, Kast J, Wilm M, Le Bot N, Vernos I, Karsenti E, Mattaj IW. (2001) Ran induces spindle assembly by reversing the inhibitory effect of importin alpha on TPX2 activity. Cell 104:83-93 Page | 237 Chapter 7 References Gu G, Yuan J, Wills M, Kasper S. (2007) Prostate cancer cells with stem cell characteristics reconstitute the original human tumor in vivo. Cancer Res 67:4807– 4815 Guo Y, Jacobs SC, Kyprianou N. (1997) Down-regulation of protein and mRNA expression for transforming growth factor-beta (TGF-beta1) type I and type II receptors in human prostate cancer. Int J Cancer 71:573-579 Habuchi H, Habuchi O, Kimata K. (2004) Sulfation pattern in glycosaminoglycan: does it have a code? Glycoconj J. 21:47-52 Habuchi H, Miyake G, Nogami K, Kuroiwa A, Matsuda Y, Kusche-Gullberg M, Habuchi O, Tanaka M, Kimata K. (2003) Biosynthesis of heparan sulphate with diverse structures and functions: two alternatively spliced forms of human heparan sulphate 6-O-sulphotransferase-2 having different expression patterns and properties. Biochem J. 371:131-42 Habuchi H, Tanaka M, Habuchi O, Yoshida K, Suzuki H, Ban K, Kimata K. (2000) The occurrence of three isoforms of heparan sulfate 6-O-sulfotransferase having different specificities for hexuronic acid adjacent to the targeted N-sulfoglucosamine. J Biol Chem. 275:2859-2868 Habuchi H, Kobayashi M, Kimata K. (1998) Molecular characterization and expression of heparan-sulfate 6-sulfotransferase. Complete cDNA cloning in human and partial cloning in Chinese hamster ovary cells. J Biol Chem. 273:9208-9213 Hadaschik BA, Gleave ME. (2007) Therapeutic options for hormone-refractory prostate cancer in 2007. Urol Oncol 25:413-419 Hameed O, Sublett J and Humphrey PA. (2005) Immunohistochemical stains for p63 and alpha-methylacyl-CoA racemase, versus a cocktail comprising both, in the diagnosis of prostate carcinoma: a comparison of the immunohistochemical staining of 430 foci radical prostatectomy and needle biopsy tissues. Am J Surg Pathol 29:579- 587 Page | 238 Chapter 7 References Hammerich KH, Ayala GE and Wheeler TM. Anatomy of the prostate gland and surgical pathology of prostate cancer. In Hricak H, Scardino PT, editors. Prostate Cancer. New York: Cambridge University Press; 2009. P1-14 Hanahan D, Weinberg RA. (2000) The hallmarks of cancer. Cell 100:57-70 Hart MJ, de los Santos R, Albert IN, Rubinfeld B, Polakis P. (1998) Downregulation of b-catenin by human Axin and its association with the APC tumor suppressor, b- catenin and GSK3b. Curr Biol 8:573–581 Hashimoto K, Goto S, Kawano S, Aoki-Kinoshita KF, Ueda N, Hamajima M, Kawasaki T, Kanehisa M. (2006) KEGG as a glycome informatics resource. Glycobiology 16:63R-70R He TC, Sparks AB, Rago C, Hermeking H, Zawel L, da Costa LT, Morin PJ, Vogelstein B, Kinzler KW. (1998) Identification of c-myc as a target of the APC pathway. Science 281:1509–1512 Hiraoka N, Nakagawa H, Ong E, Akama TO, Fukuda MN, Fukuda M. (2000) Molecular cloning and expression of two distinct human chondroitin-4-O- sulfotransferases that belong to the HNK-1 sulfotransferase gene family. J.Biol Chem 275:20188-20196 Hornebeck W, Emonard H, Monboisse JC, Bellon G. (2002) Matrix directed regulation of pericellular proteolysis and tumor progression. Semin Cancer Biol 12:231–234 Hsieh YS, Yang SF, Lue KH, Chu SC, Li TJ, Lu KH. (2007) Upregulation of urokinase-type plasminogen activator and inhibitor and gelatinase expression via 3 mitogen-activated protein kinases and PI3K pathways during the early development of osteoarthritis. J Rheumatol. 4:785-793 Hsing AW, Tsao L, Devesa SS. (2000) International trends and patterns of prostate cancer incidence and mortality. Int J Cancer 85:60-67 Page | 239 Chapter 7 References Hsu W, Zeng L, Costantini F. (1999) Identification of a Domain of Axin That Binds to the Serine/Threonine Protein Phosphatase 2A and a Self-binding Domain. J Biol Chem 274:3439–3445 Huang DW, Sherman BT, Lempicki RA. (2009) Systematic and integrative analysis of large gene lists using DAVID Bioinformatics Resources. Nature Protoc 4:44-57 Hudson DL, O’Hare M, Watt FM, Masters JR. (2001) Proliferative heterogeneity in the human prostate: evidence for epithelial stem cells. Lab Invest 80:1243–1250 Hugo H, Ackland ML, Blick T, Lawrence MG, Clements JA, Williams ED, Thompson EW. (2007) Epithelial–mesenchymal and mesenchymal–epithelial transitions in carcinoma progression. J. Cell Physiol 213:374–383 Iida S, Suzuki K, Matsuoka K, Takazono I, Shimada A, Inoue M, Yahara J, Noda S. (1997) Analysis of glycosaminoglycans in human prostate by high-performance liquid chromatography. Br J Urol 79:763-769 Ikuta Y, Nakatsura T, Kageshita T, Fukushima S, Ito S, Wakamatsu K, Baba H, Nishimura Y. (2005) Highly sensitive detection of melanoma at an early stage based on the increased serum secreted protein acidic and rich in cysteine and glypican-3 levels. Clin Cancer Res. 11:8079-8088 Ilan N, Elkin M, Vlodavsky I. (2006) Regulation, function and clinical significance of heparanase in cancer metastasis and angiogenesis. Int J Biochem Cell Biol 38:2018- 2039 Inamine M, Nagai Y, Hirakawa M, Mekaru K, Yagi C, Masamoto H, Aoki Y. (2008) Heparanase expression in endometrial cancer: analysis of immunohistochemistry. J Obstet Gynaecol 28:634-637 Innocenzi D, Alò PL, Balzani A, Sebastiani V, Silipo V, La Torre G, Ricciardi G, Bosman C, Calvieri S. (2003) Fatty acid synthase expression in melanoma. J Cutan Pathol 30:23-28 Iozzo RV. (1999) The biology of the small leucine-rich proteoglycans. Functional network of interactive proteins. J Biol Chem 274:18843–18846 Page | 240 Chapter 7 References Ito Y, Watanabe M, Nishizawa T, Omachi T, Kobayashi T, Kasama S, Habuchi O, Nakayama J. (2007) The utility of formalin-fixed and paraffin-embedded tissue blocks for quantitative analysis of N-acetylgalactosamine 4-sulfate 6-O- sulfotransferase mRNA expressed by colorectal cancer cells. Acta Histochem Cytochem 40:53-59 Izumikawa T, Koike T, Shiozawa S, Sugahara K, Tamura J, Kitagawa H (2008) Identification of chondroitin sulfate glucuronyltransferase as chondroitin synthase-3 involved in chondroitin polymerization: chondroitin polymerization is achieved by multiple enzyme complexes consisting of chondroitin synthase family members. J Biol Chem. 283:11396-11406 Izumikawa T, Kitagawa H, Mizuguchi S, Nomura KH, Nomura K, Tamura J, Gengyo-Ando K, Mitani S, Sugahara K. (2004) Nematode chondroitin polymerizing factor showing cell-/organ-specific expression is indispensable for chondroitin synthesis and embryonic cell division. J Biol Chem 279:53755-53761 Jemal A, Siegel R, Ward E, Hao Y, Xu J. (2009) Thun MJ. Cancer Statistics CA Cancer J Clin. [Epub ahead of print] Jones E, Pu H, Kyprianou N. (2009) Targeting TGF-beta in prostate cancer: therapeutic possibilities during tumor progression. Expert Opin Ther Targets 13:227- 234 Kamohara H, Ogawa M, Ishiko T, Sakamoto K, Baba H. (2007) Leukemia inhibitory factor functions as a growth factor in pancreas carcinoma cells: Involvement of regulation of LIF and its receptor expression. Int J Oncol 30:977-983 Kang HG, Evers MR, Xia G, Baenziger JU, Schachner M. (2002) Molecular cloning and characterisation of chondroitin-4-O-sulfotransferase-3. A novel membrance of the HNK-1 family of sulfotransferases. J Biol Chem 277:34766-34772 Kansara M, Tsang M, Kodjabachian L, Sims NA, Trivett MK, Ehrich M, Dobrovic A, Slavin J, Choong PF, Simmons PJ, Dawid IB, Thomas DM. (2009) Wnt inhibitory factor 1 is epigenetically silenced in human osteosarcoma, and targeted disruption accelerates osteosarcomagenesis in mice. J Clin Invest 19:837-851 Page | 241 Chapter 7 References Kawano Y, Kitaoka M, Hamada Y, Walker MM, Waxman J, Kypta RM. (2006) Regulation of prostate cell growth and morphogenesis by Dickkopf-3. Oncogene 25:6528-6537 Kim BT, Kitagawa H, Tamura J, Saito T, Kusche-Gullberg M, Lindahl U, Sugahara K. (2001) Human tumor suppressor EXT gene family members EXTL1 and EXTL3 encode 1,4-Nacetylglucosaminyltransferases that likely are involved in heparan sulfate/heparin biosynthesis. Proc Natl Acad Sci U.S.A. 98:7176–7181 Kim IY, Ahn HJ, Lang S, Oefelein MG, Oyasu R, Kozlowski JM, Lee C. (1998) Loss of expression of transforming growth factor-beta receptors is associated with poor prognosis in prostate cancer patients. Clin Cancer Res 4:1625-1630 Kim IY, Ahn HJ, Zelner DJ, Shaw JW, Lang S, Kato M, Oefelein MG, Miyazono K, Nemeth JA, Kozlowski JM, Lee C. (1996) Loss of expression of transforming growth factor beta type I and type II receptors correlates with tumor grade in human prostate cancer tissues. Clin Cancer Res 2:1255–1261 Kita K, Wittmann T, Nathke IS, Waterman-Storer CM. (2006) Adenomatous Polyposis Coli on Microtubule Plus Ends in Cell Extensions Can Promote Microtubule Net Growth with or without EB1. Molec Biol Cell 17:2331–2345 Kitagawa H, Izumikawa T, Uyama T, Sugahara K. (2003) Molecular cloning of a chondroitin polymerizing factor that cooperates with chondroitin synthase for chondroitin polymerization. J Biol Chem. 278:23666-23671 Kitagawa H, Uyama T, Sugahara K. (2001) Molecular Cloning and Expression of a Human Chondroitin Synthase. J. Biol Chem 276:38721-38726 Kitagawa H, Fujita M, Ito N, Sugahara K. (2000) Molecular cloning and expression of a novel chondroitin-6-O-sulfotransferase. J. Biol Chem 275:21075-21080 Kitagawa H, Shimakawa H, Sugahara K. (1999) The tumor suppressor EXT-like gene EXTL2 encodes an 1,4-N-acetylhexosaminyltransferase that transfers N- acetylgalactosamine and N-acetylglucosamine to the common glycosaminoglycan- protein linkage region. The key enzyme for the chain initiation of heparan sulfate. J Biol Chem 274:13933–13937 Page | 242 Chapter 7 References Kitagawa H, Tone Y, Tamura J, Neumann KW, Ogawa T, Oka S, Kawasaki T, Sugahara K. (1998) Molecular Cloning and Expression of Glucuronyltransferase I Involved in the Biosynthesis of the Glycosaminoglycan-Protein Linkage Region of Proteoglycans J. Biol. Chem. 273:6615 - 6618 Kiviniemi J, Kallajoki M, Kujala I, Matikainen MT, Alanen K, Jalkanen M, Salmivirta M (2004) Altered expression of syndecan-1 in prostate cancer. APMIS 112:89-97 Kjellén L. (2003) Glucosaminyl N-deacetylase/N-sulphotransferases in heparan sulphate biosynthesis and biology. Biochem Soc Trans 31:340-342 Klaus A, Birchmeier W. (2008) Wnt signalling and its impact on development and cancer. Nat Rev Cancer. 8:387-98 Kleeff J, Ishiwata T, Kumbasar A, Friess H, Büchler MW, Lander AD, Korc M. (1998) The cell-surface heparan sulfate proteoglycan glypican-1 regulates growth factor action in pancreatic carcinoma cells and is overexpressed in human pancreatic cancer. J Clin Invest 102:1662-1673 Kobayashi M, Sugumaran G, Liu J, Shworak NW, Silbert JE, Rosenberg RD. (1999) Molecular cloning and characterisation of a human uronyl 2-sulfotransferase that sulfates iduronyl and glucuronyl residues in dermatan/chondroitin sulfate. J Biol Chem 274: 10474-10480 Kobayashi T, Habuchi H, Nogami K, Ashikari-Hada S, Tamura K, Ide H, Kimata K. (2010) Functional analysis of chick heparan sulphate 6-O-sulphotransferase in limb bud development. Dev Growth Diff 52:146-156 Kodama J, Hasengaowa, Kusumoto T, Seki N, Matsuo T, Nakamura K, Hongo A, Hiramatsu Y. (2007) Versican expression in human cervical cancer. Eur J Cancer 43:1460-1466 Kodama J, Hasengaowa, Kusumoto T, Seki N, Matsuo T, Ojima Y, Nakamura K, Hongo A, Hiramatsu Y. (2007) Prognostic significance of stromal versican expression in human endometrial cancer. Ann Oncol 18:269-274 Page | 243 Chapter 7 References Kosir MA, Wang W, Zukowski KL, Tromp G, Barber J. (1999) Degradation of basement membrane by prostate tumor heparanase. J Surg Res 81:42-47 Kufer TA, Silljé HH, Körner R, Gruss OJ, Meraldi P, Nigg EA. (2002) Human TPX2 is required for targeting Aurora-A kinase to the spindle Cell Biol 158:617-623 Kunitoku N, Sasayama T, Marumoto T, Zhang D, Honda S, Kobayashi O, Hatakeyama K, Ushio Y, Saya H and Hirota T. (2003). CENP-A phosphorylation by Aurora-A in prophase is required for enrichment of Aurora-B at inner centromeres and for kinetochore function. Dev. Cell. 5:853–864 Kusche-Gullberg M, Kjellén L. (2003) Sulfotransferases in glycosaminoglycan biosynthesis Curr Opin Struct Biol 13:605–611 Lai JP, Sandhu DS, Shire AM, Roberts LR. (2008a) The tumor suppressor function of human sulfatase 1 (SULF1) in carcinogenesis. J Gastrointest Cancer 39:149-58 Lai JP, Sandhu DS, Yu C, Han T, Moser CD, Jackson KK, Guerrero RB, Aderca I, Isomoto H, Garrity-Park MM, Zou H, Shire AM, Nagorney DM, Sanderson SO, Adjei AA, Lee JS, Thorgeirsson SS, Roberts LR. (2008b) Sulfatase 2 up-regulates glypican 3, promotes fibroblast growth factor signaling, and decreases survival in hepatocellular carcinoma. Hepatology 47:1211-22 Lai JP, Thompson JR, Sandhu DS, Roberts LR. (2008c) Heparin-degrading sulfatases in hepatocellular carcinoma: roles in pathogenesis and therapy targets. Future Oncol 4:803-14 Lang SH, Frame FM, Collins AT. (2009) Prostate cancer stem cells. J Pathol 217:299- 306 LaPierre DP, Lee DY, Li SZ, Xie YZ, Zhong L, Sheng W, Deng Z, Yang BB. (2007) The ability of versican to simultaneously cause apoptotic resistance and sensitivity. Cancer Res 67:4742-4750 Lapointe J, Li C, Higgins JP, van de Rijn M, Bair E, Montgomery K, Ferrari M, Egevad L, Rayford W, Bergerheim U, Ekman P, DeMarzo AM, Tibshirani R, Botstein D, Brown PO, Brooks JD, Pollack JR. (2004) Gene expression profiling Page | 244 Chapter 7 References identifies clinically relevant subtypes of prostate cancer. Proc Natl Acad Sci U S A 101:811-816 Le Page C, Koumakpayi IH, Alam-Fahmy M, Mes-Masson AM, Saad F. (2006) Expression and localisation of Akt-1, Akt-2 and Akt-3 correlate with clinical outcome of prostate cancer patients. Br J Cancer 94:1906-1912 Lerner I, Baraz L, Pikarsky E, Meirovitz A, Edovitsky E, Peretz T, Vlodavsky I, Elkin M. (2008) Function of heparanase in prostate tumorigenesis: potential for therapy. Clin Cancer Res 14:668-676 Leygue E, Snell L, Dotzlaw H, Troup S, Hiller-Hitchcock T, Murphy LC, Roughley PJ, Watson PH. (2000) Lumican and decorin are differentially expressed in human breast carcinoma. J Pathol 192:313-320 Li C, Wong WH (2001) Model-based analysis of oligonucleotide arrays: Expression index computation and outlier detection. Proc. Natl. Acad. Sci. 98:31-36 Li F, Ten Dam GB, Murugan S, Yamada S, Hashiguchi T, Mizumoto S, Oguri K, Okayama M, van Kuppevelt TH, Sugahara K. (2008) Involvement of highly sulfated chondroitin sulfate in the metastasis of the Lewis lung carcinoma cells. J Biol Chem 283:34294-34304 Li J, Kleeff J, Kayed H, Felix K, Penzel R, Büchler MW, Korc M, Friess H. (2004) Glypican-1 antisense transfection modulates TGF-beta-dependent signaling in Colo- 357 pancreatic cancer cells. Biochem Biophys Res Commun 320:1148-1155 Li JP, Gong F, Darwish KE, Jalkanen M, Lindahl U. (2001) Characterization of the D-glucuronyl C-5 epimerase involved in the biosynthesis of heparin and heparan sulfate. J Biol Chem 276:20069–20077 Li KS, Yi Q. (2001) Characterization of Human MAPRE Genes and Their Proteins. Genomics 71:142–149 Licchesi JD, Westra WH, Hooker CM, Machida EO, Baylin SB, Herman JG. (2008) Epigenetic alteration of Wnt pathway antagonists in progressive glandular neoplasia of the lung. Carcinogenesis 29:895-904 Page | 245 Chapter 7 References Lin H, Huber R, Schlessinger D, Morin PJ. (1999) Frequent silencing of the GPC3 gene in ovarian cancer cell lines. Cancer Res 59:807-810 Lind T, Tufaro F, McCormick C, Lindahl U, Lidholt K. (1998) The putative tumor suppressors EXT1 and EXT2 are glycosyltransferases required for the biosynthesis of heparan sulfate. J Biol Chem 273: 26265–26268 Liu J, Shriver Z, Blaiklock P, Yoshida K, Sasisekharan R, Rosenberg RD. (1999) Heparan sulfate D-glucosaminyl 3-O-sulfotransferase-3A sulfates N-unsubstituted glucosamine residues. J Biol Chem. 274:38155-38162 Liu J, Shworak NW, Sinaÿ P, Schwartz JJ, Zhang L, Fritze LM, Rosenberg RD. (1999) Expression of heparan sulfate D-glucosaminyl 3-O-sulfotransferase isoforms reveals novel substrate specificities. J Biol Chem. 274:5185-5192 Liu J, Shworak NW, Fritze LM, Edelberg JM, Rosenberg RD. (1996) Purification of heparan sulfate D-glucosaminyl 3-O-sulfotransferase. J Biol Chem. 271:27072-27082 Liu M, Yang S, Wang Y, Zhu H, Yan S, Zhang W, Quan L, Bai J, Xu N. (2009) EB1 acts as an oncogene via activating beta-catenin/TCF pathway to promote cellulat growth and inhibit apoptosis. Mol Carcinog 48:212-219 Lloyd MD, Darley DJ, Wierzbicki AS, Threadgill MD. (2008) a-Methylacyl-CoA racemase – an ‘obscure’ metabolic enzyme takes centre stage. FEBS Journal 275:1089–1102 Loussouarn D, Campion L, Sagan C, Frenel JS, Dravet F, Classe JM, Pioud-Martigny R, Berton-Rigaud D, Bourbouloux E, Mosnier JF, Bataille FR, Campone M. (2008) Prognostic impact of syndecan-1 expression in invasive ductal breast carcinomas. Br J Cancer 98:1993-1998 Lowsley O (1912) The development of the human prostate gland with reference to the development of other structures at the neck of the urinary bladder. Am J Anat. 13:299-350 Page | 246 Chapter 7 References Luo J, Duggan DJ, Chen Y, Sauvageot J, Ewing CM, Bittner ML, Trent JM, Isaacs WB. (2001) Human prostate cancer and benign prostatic hyperplasia: molecular dissection by gene expression profiling. Cancer Res 61:4683-4688 Luo JH, Yu YP, Cieply K, Lin F, Deflavia P, Dhir R, Finkelstein S, Michalopoulos G, Becich M. (2002) Gene expression analysis of prostate cancers. Mol Carcinog 33:25- 35 Maccarana M, Olander B, Malmström J, Tiedemann K, Aebersold R, Lindahl U, Li JP, Malmström A. (2006) Biosynthesis of dermatan sulfate: chondroitin-glucuronate C5-epimerase is identical to SART2. J Biol Chem. 281:11560-11568 Maehata T, Taniguchi H, Yamamoto H, Nosho K, Adachi Y, Miyamoto N, Miyamoto C, Akutsu N, Yamaoka S, Itoh F. (2008) Transcriptional silencing of Dickkopf gene family by CpG island hypermethylation in human gastrointestinal cancer. World J Gastroenterol 14:2702-2714 Mani SA, Guo W, Liao MJ, Eaton EN, Ayyanan A, Zhou AY, Brooks M, Reinhard F, Zhang CC, Shipitsin M, Campbell LL, Polyak K, Brisken C, Yang J, Weinberg RA. (2008) The epithelial–mesenchymal transition generates cells with properties of stem cells. Cell 133:704–715 Mao P, Nakao K, Bora R, Geller J. (1965) Human benign prostatic hyperplasia. Arch Pathol 79:270-283 Marcus F, Rittenhouse J, Gontero B, Harrsch PB. (1987) Function, structure and evolution of fructose-1,6-bisphosphatase. Arch Biol Med Exp (Santiago) 20:371-378 Marianne J. Ratcliffe, Keiji Itoh and Sergei Y. Sokol. (2000) A Positive Role for the PP2A Catalytic Subunit in Wnt Signal Transduction. J Biol Chem. 275:35680–35683 Marker PC. (2008) Does prostate cancer co-opt the developmental program? Differentiation 76:736–744 Mathupala SP, Ko YH, Pedersen PL. (2009) Hexokinase-2 bound to mitochondria: cancer's stygian link to the "Warburg Effect" and a pivotal target for effective therapy. Semin Cancer Biol. 19:17-24 Page | 247 Chapter 7 References Matsuda K, Maruyama H, Guo F, Kleeff J, Itakura J, Matsumoto Y, Lander AD, Korc M. (2001) Glypican-1 is overexpressed in human breast cancer and modulates the mitogenic effects of multiple heparin-binding growth factors in breast cancer cells. Cancer Res 61:5562-5569 McCabe CD, Spyropoulos DD, Martin D, Moreno CS. (2008) Genome-wide analysis of the homeobox C6 transcriptional network in prostate cancer. Cancer Res 68:1988- 1996 McCormick C, Duncan G, Goutsos KT, Tufaro F. (2000) The putative tumor suppressors EXT1 and EXT2 form a stable complex that accumulates in the Golgi apparatus and catalyzes the synthesis of heparan sulfate. Proc. Natl. Acad. Sci. U.S.A. 97:668–673 McNeal JE. (1988) Normal histology of the prostate. Am J Surg Pathol 12:619-633 McNeal JE. (1980) Anatomy of the prostate: an historical survey of divergent views. The Prostate 1:3-13 McNeal JE. (1972) The prostate and prostatic urethra: a morphologic synthesis. J Urol 107:1008-1016 McNeal JE, Villers A, Redwine E, Freiha FS, Stamey TA. (1990) Capsular penetration in prostate cancer: significance for natural history and treatment. Am J Surg Pathol 14:240-247 Menendez JA, Lupu R. (2006) Oncogenic properties of the endogenous fatty acid metabolism: molecular pathology of fatty acid synthase in cancer cells. Curr Opin Clin Nutr Metab Care 9:346-357 Mian BM, Troncoso P, Okihara K, Bhadkamkar V, Johnston D, Reyes AO, Babaian RJ. (2002) Outcome of patients with Gleason score 8 or higher prostate cancer following radical prostatectomy alone. J Urol 167:1675-1680 Migita T, Ruiz S, Fornari A, Fiorentino M, Priolo C, Zadra G, Inazuka F, Grisanzio C, Palescandolo E, Shin E, Fiore C, Xie W, Kung AL, Febbo PG, Subramanian A, Mucci L, Ma J, Signoretti S, Stampfer M, Hahn WC, Finn S, Loda M. (2009) Fatty Page | 248 Chapter 7 References acid synthase: a metabolic enzyme and candidate oncogene in prostate cancer. J Natl Cancer Inst 101:519-532 Mikami S, Oya M, Shimoda M, Mizuno R, Ishida M, Kosaka T, Mukai M, Nakajima M, Okada Y. (2008) Expression of heparanase in renal cell carcinomas: implications for tumor invasion and prognosis. Clin Cancer Res 14:6055-6061 Miller GJ, Miller HL, van Bokhoven A, Lambert JR, Werahera PN, Schirripa O, Lucia MS, Nordeen SK. (2003) Aberrant HOXC expression accompanies the malignant phenotype in human prostate. Cancer Res 63:5879-5888 Mizuguchi S, Uyama T, Kitagawa H, Nomura KH, Dejima K, Gengyo-Ando K, Mitani S, Sugahara K, Nomura K. (2003) Chondroitin proteoglycans are involved in cell division of Caenorhabditis elegans. Nature 423:443-448 Mochizuki H, Yoshida K, Gotoh M, Sugioka S, Kikuchi N, Kwon YD, Tawada A, Maeyama K, Inaba N, Hiruma T, Kimata K, Narimatsu H. (2003) Characterization of a heparan sulfate 3-O-sulfotransferase-5, an enzyme synthesizing a tetrasulfated disaccharide. J Biol Chem. 278:26780-26787 Moltzahn FR, Volkmer JP, Rottke D, Ackermann R. (2008) "Cancer stem cells"- lessons from Hercules to fight the Hydra. Urol Oncol. 26:581-589 Montironi R, Mazzucchelli R, Lopez-Beltran A, Cheng L. (2007) Prostate cancer origins, diagnosis, and prognosis in clinical practice in Clinical Pathology of Urologic Tumors, eds Mikuz G. Informa Healthcare, London, United Kingdom. Monzavi-Karbassi B, Stanley JS, Hennings L, Jousheghany F, Artaud C, Shaaf S, Kieber-Emmons T. (2007) Chondroitin sulfate glycosaminoglycans as major P- selectin ligands on metastatic breast cancer cell lines. Int J Cancer 120:1179-1191 Moore RA, HAnzel RF. (1936) Chemical composition of prostatic corpora amylacea and calculi. Arch Pathol 22:41-45 Morimoto-Tomita M, Uchimura K, Bistrup A, Lum DH, Egeblad M, Boudreau N, Werb Z, Rosen SD. (2005) Sulf-2, a proangiogenic heparan sulfate endosulfatase, is upregulated in breast cancer. Neoplasia 7:1001-1010 Page | 249 Chapter 7 References Morrison EE, Wardleworth BN, Askham JM, Markham AF, DM Meredith. (1998) EB1, a protein which interacts with the APC tumour suppressor, is associated with the microtubule cytoskeleton throughout the cell cycle. Oncogene 17:3471- 3477 Moscatello DK, Santra M, Mann DM, McQuillan DJ, Wong AJ, Iozzo RV. (1998) Decorin suppresses tumor cell growth by activating the epidermal growth factor receptor. J Clin Invest 101: 406-412 Mundhenke C, Meyer K, Drew S, Friedl A. (2002) Heparan sulfate proteoglycans as regulators of fibroblast growth factor-2 receptor binding in breast carcinomas. Am J Pathol 160:185-194 Mythreye K, Blobe GC. (2009) The type III TGF-beta receptor regulates epithelial and cancer cell migration through beta-arrestin2-mediated activation of Cdc42. Proc Natl Acad Sci U S A 106:8221-8226 Nadanaka S, Ishida M, Ikegami M, & Kitagawa H (2008) Chondroitin 4-O- sulfotransferase-1 modulates Wnt-3a signaling through control of E disaccharide expression of chondroitin sulfate. J Biol Chem 283(40):27333-27343 Nakatsura T, Yoshitake Y, Senju S, Monji M, Komori H, Motomura Y, Hosaka S, Beppu T, Ishiko T, Kamohara H, Ashihara H, Katagiri T, Furukawa Y, Fujiyama S, Ogawa M, Nakamura Y, Nishimura Y. (2003) Glypican-3, overexpressed specifically in human hepatocellular carcinoma, is a novel tumor marker. Biochem Biophys Res Commun 306:16-25 Narita K, Chien J, Mullany SA, Staub J, Qian X, Lingle WL, Shridhar V. (2007) Loss of HSulf-1 expression enhances autocrine signaling mediated by amphiregulin in breast cancer. J Biol Chem 282:14413-20 Nash MA, Deavers MT, Freedman RS. (2002) The expression of decorin in human ovarian tumors. Clin Cancer Res 8:1754-1760 Nasser NJ. (2008) Heparanase involvement in physiology and disease. Cell Mol Life Sci 65:1706-15 Page | 250 Chapter 7 References Nelson JB. Hormone therapy for prostate cancer. In: Wein AJ, Kavoussi LR, Novick AC, Partiri AW, Peters CA, editors. Campbell-Walsh Urology. Volume 3. 9th Edition. Philadelphia: Saunders Elsevier; 2007. P3082-3100. Netelenbos T, van den Born J, Kessler FL, Zweegman S, Merle PA, van Oostveen JW, Zwaginga JJ, Huijgens PC, Dräger AM. (2003) Proteoglycans on bone marrow endothelial cells bind and present SDF-1 towards hematopoietic progenitor cells. Leukemia 17:175-184. Nunes FD, de Almeida FC, Tucci R, de Sousa SC. (2003) Homeobox genes: a molecular link between development and cancer. Pesqui Odontol Bras 17:94-98 Ogishima T, Shiina H, Breault JE, Tabatabai L, Bassett WW, Enokida H, Li LC, Kawakami T, Urakami S, Ribeiro-Filho LA, Terashima M, Fujime M, Igawa M, Dahiya R. (2005) Increased heparanase expression is caused by promoter hypomethylation and up-regulation of transcriptional factor early growth response-1 in human prostate cancer. Clin Cancer Res 11:1028-1036 Ohtake S, Ito Y, Fukuta M, Habuchi O. (2001) Human N-acetylgalactosamine 4- sulfate 6-O-sulfotransferase cDNA is related to human B cell recombination activating gene-associated gene. J. Biol Chem 276:43894-43900 Patrawala L, Calhoun-Davis T, Schneider-Broussard R, Tang DG. (2007) Hierarchical organization of prostate cancer cells in xenograft tumors: The CD44+α2β1+ cell population is enriched in tumor-initiating cells. Cancer Res 67:6796–6805 Pavão MS, Vilela-Silva AC, Mourão PA. (2006) Biosynthesis of chondroitin sulfate: from the early, precursor discoveries to nowadays, genetics approaches. Adv Pharmacol 53:117-40 Perou CM, Sørlie T, Eisen MB, van de Rijn M, Jeffrey SS, Rees CA, Pollack JR, Ross DT, Johnsen H, Akslen LA, Fluge O, Perhamenschikov A, Williams C, Zhu SX, Lonning PE, Borresen-Dale AL, Brown PO, Botstein D. (2000) Molecular portraits of human breast tumours. Nature 406: 747-752 Page | 251 Chapter 7 References Petersen RO, Sesterhenn IA, Davis CJ Jr. Urologic Pathology 3rd Edition. Baltimore: Lippincott Williams and Wilkins; 2009. 636p. Pierce GB, Speers WC. (1988) Tumors as caricatures of the process of tissue renewal: prospects for therapy by directing differentiation. Cancer Res 48:1996–2004 Pirinen R, Leinonen T, Böhm J, Johansson R, Ropponen K, Kumpulainen E, Kosma VM. (2005) Versican in nonsmall cell lung cancer: relation to hyaluronan, clinicopathologic factors, and prognosis. Hum Pathol 36:44-50 Plun-Favreau H, Perret D, Diveu C, Froger J, Chevalier S, Lelièvre E, Gascan H, Chabbert M. (2003) Leukemia inhibitory factor (LIF), cardiotrophin-1, and oncostatin M share structural binding determinants in the immunoglobulin-like domain of LIF receptor. J Biol Chem. 78:27169-27179 Polyak K, Weinberg RA. (2009) Transitions between epithelial and mesenchymal states: acquisition of malignant and stem cell traits. Nat Rev Cancer 9:265-273 Prabhakar V, Sasisekharan R. (2006) The biosynthesis and catabolism of galactosaminoglycans. Adv Pharmacol 53:69-115 Pukkila M, Kosunen A, Ropponen K, Virtaniemi J, Kellokoski J, Kumpulainen E, Pirinen R, Nuutinen J, Johansson R, Kosma VM. (2007) High stromal versican expression predicts unfavourable outcome in oral squamous cell carcinoma. J Clin Pathol 60:267-272 Pukkila MJ, Kosunen AS, Virtaniemi JA, Kumpulainen EJ, Johansson RT, Kellokoski JK, Nuutinen J, Kosma VM. (2004) Versican expression in pharyngeal squamous cell carcinoma: an immunohistochemical study. J Clin Pathol 57:735-739 Qi M, Ikematsu S, Maeda N, Ichihara-Tanaka K, Sakuma S, Noda M, Muramatsu T, Kadomatsu K. (2001) Haptotactic migration induced by midkine. Involvement of protein-tyrosine phosphatase zeta. Mitogen-activated protein kinase, and phosphatidylinositol 3-kinase. J Biol Chem 276:15868-15875 Quinn DI, Henshall SM, Brenner PC, Kooner R, Golovsky D, O'Neill GF, Turner JJ, Delprado W, Grygiel JJ, Sutherland RL, Stricker PD. (2003) Prognostic significance Page | 252 Chapter 7 References of preoperative factors in localized prostate carcinoma treated with radical prostatectomy: importance of percentage of biopsies that contain tumor and the presence of biopsy perineural invasion. Cancer 97:1884-1893 Ramachandran S, Liu P, Young AN, Yin-Goen Q, Lim SD, Laycock N, Amin MB, Carney JK, Marshall FF, Petros JA, Moreno CS. (2005) Loss of HOXC6 expression induces apoptosis in prostate cancer cells. Oncogene 24:188-198 Ratcliffe MJ, Itoh K, Sokol SY. (2000) A positive role for the PP2A catalytic subunit in Wnt signal transduction. J Biol Chem 275:35680-35683 Rau C, Zheng N, Hazelwood C, Rau G. (1995) Scanning tunneling microscopy study of poly-L-proline. Physiol Chem Phys Med NMR 27:55–61 Ricciardelli C, Russell DL, Ween MP, Mayne K, Suwiwat S, Byers S, Marshall VR, Tilley WD, Horsfall DJ. (2007) Formation of hyaluronan- and versican-rich pericellular matrix by prostate cancer cells promotes cell motility. Biol Chem. 282:10814-10825 Ricciardelli C, Brooks JH, Suwiwat S, Sakko AJ, Mayne K, Raymond WA, Seshadri R, LeBaron RG, Horsfall DJ. (2002) Regulation of stromal versican expression by breast cancer cells and importance to relapse-free survival in patients with node- negative primary breast cancer. Clin Cancer Res 8:1054-1060 Ricciardelli C, Quinn DI, Raymond WA, McCaul K, Sutherland PD, Stricker PD, Grygiel JJ, Sutherland RL, Marshall VR, Tilley WD, Horsfall DJ. (1999) Elevated levels of peritumoral chondroitin sulfate are predictive of poor prognosis in patients treated by radical prostatectomy for early-stage prostate cancer. Cancer Res. 59:2324- 2328 Ricciardelli C, Mayne K, Sykes PJ, Raymond WA, McCaul K, Marshall VR, Horsfall DJ. (1998) Elevated levels of versican but not decorin predict disease progression in early-stage prostate cancer Clin Cancer Res. 4:963-971 Ricciardelli C, Mayne K, Sykes PJ, Raymond WA, McCaul K, Marshall VR, Tilley WD, Skinner JM, Horsfall DJ. (1997) Elevated stromal chondroitin sulfate Page | 253 Chapter 7 References glycosaminoglycan predicts progression in early-stage prostate cancer. Clin Cancer Res 3:983-992 Richardson GD, Robson CN, Lang SH, Neal DE, Maitland NJ, Collins AT. (2004) CD133, a novel marker for human prostatic epithelial stem cells. J Cell Sci 117:3539– 3545 Rochat-Steiner V, Becker K, Micheau O, Schneider P, Burns K, Tschopp J. (2000) FIST/HIPK3: a Fas/FADD-interacting serine/threonine kinase that induces FADD phosphorylation and inhibits Fas-mediated Jun NH2-terminal kinase activation. J Exp Med 192:1165-1174 Rogers SL, Rogers GC, Sharp DJ, Vale RD. (2002) Drosophila EB1 is important for proper assembly, dynamics, and positioning of the mitotic spindle. J. Cell Biol.158:873–884. Roh YH, Kim YH, Choi HJ, Lee KE, Roh MS. (2008) Syndecan-1 expression in gallbladder cancer and its prognostic significance. Eur Surg Res 41:245-250 Roy M, Marchetti D. (2009) Cell surface heparan sulfate released by heparanase promotes melanoma cell migration and angiogenesis. J Cell Biochem 106: 200-209 Sakko AJ, Butler MS, Byers S, Reinboth BJ, Stahl J, Kench JG, Horvath LG, Sutherland RL, Stricker PD, Henshall SM, Marshall VR, Tilley WD, Horsfall DJ, Ricciardelli C. (2008) Immunohistochemical level of unsulfated chondroitin disaccharides in the cancer stroma is an independent predictor of prostate cancer relapse. Cancer Epidemiol Biomarkers Prev. 17:2488-2497 Sakko AJ, Ricciardelli C, Mayne K, Suwiwat S, LeBaron RG, Marshall VR, Tilley WD, Horsfall DJ. (2003) Modulation of prostate cancer cell attachment to matrix by versican. Cancer Res 63:4786-4791 Sakr WA, Wheeler TM, Blute M, Bodo M, Calle-Rodrigue R, Henson DE, Mostofi FK, Seiffert J, Wojno K, Zincke H. (1996) Staging and reporting of prostate cancer-- sampling of the radical prostatectomy specimen. Cancer 78:366-368 Page | 254 Chapter 7 References Sanderson RD. (2001) Heparan sulfate proteoglycans in invasion and metastasis. Semin Cell Dev Biol 12:89-98 Santra M, Eichstetter I, Iozzo RV. (2000) An anti-oncogenic role for decorin. Down- regulation of ErbB2 leads to growth suppression and cytodifferentiation of mammary carcinoma cells. J Biol Chem 275:35153-35161 Sato T, Gotoh M, Kiyohara K, Akashima T, Iwasaki H, Kameyama A, Mochizuki H, Yada T, Inaba N, Togayachi A, Kudo T, Asada M, Watanabe H, Imamura T, Kimata K, Narimatsu H. (2003) Differential roles of two N-acetylgalactosaminyltransferases, CSGalNAcT-1, and a novel enzyme, CSGalNAcT-2. Initiation and elongation in synthesis of chondroitin sulfate. J Biol Chem. 278:3063-3071 Savorè C, Zhang C, Muir C, Liu R, Wyrwa J, Shu J, Zhau HE, Chung LW, Carson DD, Farach-Carson MC. (2005) Perlecan knockdown in metastatic prostate cancer cells reduces heparin-binding growth factor responses in vitro and tumor growth in vivo. Clin Exp Metastasis 22:377-390 Schmidt I, Krall WJ, Uittenbogaart CH, Braun J, Giorgi JV. (1992) Dead cell discrimination with 7-amino-actinomycin D in combination with dual color immunofluorescence in single laser flow cytometry. Cytometry. 13:204-208 Schroeder A, Mueller O, Stocker S, Salowsky R, Leiber M, Gassmann M, Lightfoot S, Menzel W, Granzow M, Ragg T. (2006) The RIN: an RNA integrity number for assigning integrity values to RNA measurements. BMC Molec Biol 7:3 Seidler DG, Goldoni S, Agnew C, Cardi C, Thakur ML, Owens RT, McQuillan DJ, Iozzo RV. (2006) Decorin protein core inhibits in vivo cancer growth and metabolism by hindering epidermal growth factor receptor function and triggering apoptosis via caspase-3 activation. J Biol Chem 281: 26408-26418 Senay C, Lind T, Muguruma K, Tone Y, Kitagawa H, Sugahara K, Lidholt K, Lindahl U, Kusche-Gullberg M. (2000) The EXT1/EXT2 tumor suppressors: catalytic activities and role in heparan sulfate biosynthesis. EMBO Rep 1:282–286 Page | 255 Chapter 7 References Seow A, Koh WP, Chia KS, Ling A, Chow KY, Jin A, Lee HP. (2009) Trends in cancer incidence in Singapore 2002 - 2006. Singapore Cancer Registry, Interim Report, Singapore Shariat SF, Khoddami SM, Saboorian H, Koeneman KS, Sagalowsky AI, Cadeddu JA, McConnell JD, Holmes MN, Roehrborn CG. (2004) Lymphovascular invasion is a pathological feature of biologically aggressive disease in patients treated with radical prostatectomy. J Urol 171:1122-1127 Shariat SF, Svatek RS, Kabbani W, Walz J, Lotan Y, Karakiewicz PI, Roehrborn CG. (2008) Prognostic value of syndecan-1 expression in patients treated with radical prostatectomy. BJU Int 101:232-237 Sharifi N, Hurt EM, Kawasaki BT, Farrar WL. (2007) TGFBR3 loss and consequences in prostate cancer. Prostate 67:301-311 Shimada K, Nakamura M, De Velasco MA, Tanaka M, Ouji Y, Konishi N. (2009) Syndecan-1, a new target molecule involved in progression of androgen-independent prostate cancer. Cancer Sci 100:1248-1254 Shtutman M, Zhurinsky J, Simcha I, Albanese C, D'Amico M, Pestell R, Ben-Ze'ev A. (1999). The cyclin D1 gene is a target of the beta-catenin LEF1 pathway. Proc. Natl. Acad. Sci.U.S.A.96:5522–5527 Shukla S, Maclennan GT, Hartman DJ, Fu P, Resnick MI, Gupta S. (2007) Activation of PI3K-Akt signaling pathway promotes prostate cancer cell invasion. Int J Cancer. 121:1424-1432 Shworak NW, Liu J, Fritze LM, Schwartz JJ, Zhang L, Logeart D, Rosenberg RD. (1997) Molecular cloning and expression of mouse and human cDNAs encoding heparan sulfate D-glucosaminyl 3-O-sulfotransferase. J Biol Chem. 272:28008-28019 Shworak NW, Liu J, Petros LM, Zhang L, Kobayashi M, Copeland NG, Jenkins NA, Rosenberg RD. (1999) Multiple isoforms of heparan sulfate D-glucosaminyl 3-O- sulfotransferase. Isolation, characterization, and expression of human cdnas and identification of distinct genomic loci. J Biol Chem. 274:5170-5184 Page | 256 Chapter 7 References Silbert JE, Sugumaran G. (2002) Biosynthesis of chondroitin/dermatan sulfate. IUBMB Life 54:177-186 Sim HG, Cheng CW. (2005) Changing demography of prostate cancer in Asia. Eur J Cancer 41:834-845 Snell RS. (2008) Clinical anatomy by regions Eighth Edition. Baltimore, USA: Lippincott Williams & Wilkins. 926p. Sørlie T, Perou CM, Tibshirani R, Aas T, Geisler S, Johnsen H, Hastie T, Eisen MB, van de Rijn M, Jeffrey SS, Thorsen T, Quist H, Matese JC, Brown PO, Botstein D, Eystein Lønning P, Børresen-Dale AL. (2001) Gene expression patterns of breast carcinomas distinguish tumor subclasses with clinical implications. Proc Natl Acad Sci U S A. 98: 10869-10874 Sørlie T, Tibshirani R, Parker J, Hastie T, Marron JS, Nobel A, Deng S, Johnsen H, Pesich R, Geisler S, Demeter J, Perou CM, Lønning PE, Brown PO, Børresen-Dale AL, Botstein D. (2003) Repeated observation of breast tumor subtypes in independent gene expression data sets. Proc Natl Acad Sci U S A. 100:8418-8423 Sotiriou C, Neo SY, McShane LM, Korn EL, Long PM, Jazaeri A, Martiat P, Fox SB, Harris AL, Liu ET. (2003) Breast cancer classification and prognosis based on gene expression profiles from a population-based study. Proc Natl Acad Sci U S A. 100: 10393-10398 Ständer M, Naumann U, Wick W, Weller M. (1999) Transforming growth factor-beta and p-21: multiple molecular targets of decorin-mediated suppression of neoplastic growth. Cell Tissue Res 296:221-227 Stigliano I, Puricelli L, Filmus J, Sogayar MC, Bal de Kier Joffé E, Peters MG. (2009) Glypican-3 regulates migration, adhesion and actin cytoskeleton organization in mammary tumor cells through Wnt signaling modulation. Breast Cancer Res Treat 114:251-262 Su G, Meyer K, Nandini CD, Qiao D, Salamat S, Friedl A. (2006) Glypican-1 is frequently overexpressed in human gliomas and enhances FGF-2 signaling in glioma cells. Am J Pathol 168:2014-2026 Page | 257 Chapter 7 References Su LK, Burrell M, Hill DE, Gyuris J, Brent R, Wiltshire R, Trent J, Vogelstein B, Kinzler KW. (1995) APC Binds to the Novel Protein EB1. Cancer Res 55:2972-2977 Sugahara K, Mikami T, Uyama T, Mizuguchi S, Nomura K, Kitagawa H. (2003) Recent advances in the structural biology of chondroitin sulfate and dermatan sulfate. Curr Opin Struct Biol 13:612-620 Sugahara KN, Hirata T, Tanaka T, Ogino S, Takeda M, Terasawa H, Shimada I, Tamura J, ten Dam GB, van Kuppevelt TH, Miyasaka M. (2008) Chondroitin sulfate E fragments enhance CD44 cleavage and CD44-dependent motility in tumor cells. Cancer Res 68:7191-7199 Sugaya N, Habuchi H, Nagai N, Ashikari-Hada S, Kimata K. (2008) 6-O-sulfation of heparan sulfate differentially regulates various fibroblast growth factor-dependent signalings in culture.J Biol Chem 283:10366-10376 Sun L, Gao J, Dong X, Liu M, Li D, Shi X, Dong JT, Lu X, Liu C, Zhou J. (2008) EB1 promotes Aurora B kinase activity through blocking its inactivation by protein phosphatise 2A. Proc Nat Acad Science 105:7153-7158 Sun YX, Fang M, Wang J, Cooper CR, Pienta KJ, Taichman RS. (2007) Expression and activation of alpha(v)beta(3) integrins by SDF-1/CXC12 increases the aggressiveness of prostate cancer cells. Prostate 67:61-73 Sun YX, Schneider A, Jung Y, Wang J, Dai J, Wang J, Cook K, Osman NI, Koh- Paige AJ, Shim H, Pienta KJ, Keller ET, McCauley LK, Taichman RS. (2005) Skeletal localization and neutralization of the SDF-1(CXCL12)/CXCR4 axis blocks prostate cancer metastasis and growth in osseous sites in vivo. J Bone Miner Res 20:318-329 Suwiwat S, Ricciardelli C, Tammi R, Tammi M, Auvinen P, Kosma VM, LeBaron RG, Raymond WA, Tilley WD, Horsfall DJ. (2004) Expression of extracellular matrix components versican, chondroitin sulfate, tenascin, and hyaluronan, and their association with disease outcome in node-negative breast cancer. Clin Cancer Res 10:2491-2498 Page | 258 Chapter 7 References Suzuki H, Toyota M, Carraway H, Gabrielson E, Ohmura T, Fujikane T, Nishikawa N, Sogabe Y, Nojima M, Sonoda T, Mori M, Hirata K, Imai K, Shinomura Y, Baylin SB, Tokino T. (2008) Frequent epigenetic inactivation of Wnt antagonist genes in breast cancer. Br J Cancer 98:1147-1156 Taichman RS, Cooper C, Keller ET, Pienta KJ, Taichman NS, McCauley LK. (2002) Use of the stromal cell-derived factor-1/CXCR4 pathway in prostate cancer metastasis to bone. Cancer Res 62:1832-1837 Takahiro T, Shinichi K, Toshimitsu S. (2003) Expression of fatty acid synthase as a prognostic indicator in soft tissue sarcomas. Clin Cancer Res 9:2204–2212 Takeuchi, J. (1965) Growth-promoting Effect of Chondroitin Sulphate on Solid Ehrlich Ascites Tumour. Nature. 7:537-538 Tan PH, Jayabaskar T, Yip G, Tan Y, Hilmy M, Selvarajan S, Bay BH. (2005) p53 and c-kit (CD117) protein expression as prognostic indicators in breast phyllodes tumors: a tissue microarray study. Mod Pathol 18:1527-1534 ten Dam GB, van de Westerlo EM, Purushothaman A, Stan RV, Bulten J, Sweep FC, Massuger LF, Sugahara K, van Kuppevelt TH. (2007) Antibody GD3G7 selected against embryonic glycosaminoglycans defines chondroitin sulfate-E domains highly up-regulated in ovarian cancer and involved in vascular endothelial growth factor binding. Am J Pathol 171:1324-1333 Theocharis AD. (2002) Human colon adenocarcinoma is associated with specific post-translational modifications of versican and decorin. Biochim Biophys Acta 1588:165-172 Tirnauer JS and Bierer BE. (2000) EB1 Proteins Regulate Microtubule Dynamics, Cell Polarity, and Chromosome Stability. J Cell Biol. 149:761–766 Tirnauer JS, Grego S, Salmon ED, Mitchison TJ. (2002) EB1-microtubule interactions in Xenopus egg extracts: role of EB1 in microtubule stabilization and mechanisms of targeting to microtubules. Mol. Biol. Cell. 13:3614–3626 Page | 259 Chapter 7 References Todovoricc V, Chen CC, Hay N, Lau LF. (2005) The matrix protein CCN1 (CYR61) induces apoptosis in fibroblasts. J Cell Biol 17:559-568 Tomida M. (2000) Structural and functional studies on the leukemia inhibitory factor receptor (LIF-R): gene and soluble form of LIF-R, and cytoplasmic domain of LIF-R required for differentiation and growth arrest of myeloid leukemic cells. Leuk Lymphoma. 37:517-525 Turley RS, Finger EC, Hempel N, How T, Fields TA, Blobe GC. (2007) The type III transforming growth factor-beta receptor as a novel tumor suppressor gene in prostate cancer. Cancer Res 67:1090-1098 Uyama T, Kitagawa H, Tanaka J, Tamura J, Ogawa T, Sugahara K. (2003) Molecular cloning and expression of a second chondroitin N-acetylgalactosaminyltransferase involved in the initiation and elongation of chondroitin/dermatan sulfate. J Biol Chem. 278:3072-3078 Uyama T, Kitagawa H, Tamura J, Sugahara K. (2002) Molecular cloning and expression of human chondroitin N-acetylgalactosaminyltransferase: the key enzyme for chain initiation and elongation of chondroitin/dermatan sulfate on the protein linkage region tetrasaccharide shared by heparin/heparan sulfate. J Biol Chem. 277:8841-8846 Vader G, Lens SMA. (2008) The Aurora kinase family in cell division and cancer Biochim Biophys Acta 1786:60–72 van de Sande T, Roskams T, Lerut E, Joniau S, Van Poppel H, Verhoeven G, Swinnen JV. (2005) High-level expression of fatty acid synthase in human prostate cancer tissues is linked to activation and nuclear localization of Akt/PKB. J Pathol 206:214–219 van den Born J, Salmivirta K, Henttinen T, Ostman N, Ishimaru T, Miyaura S, Yoshida K, Salmivirta M. (2005) Novel heparan sulfate structures revealed by monoclonal antibodies. J Biol Chem. 280:20516-20523 Page | 260 Chapter 7 References Vargas SO, Jiroutek M, Welch WR, Nucci MR, D'Amico AV, Renshaw AA. (1999) Perineural invasion in prostate needle biopsy specimens. Correlation with extraprostatic extension at resection. Am J Clin Pathol 111:223-8 Vaughan KT. (2005) TIP maker and TIP marker; EB1 as a master controller of microtubule plus ends. J Cell Biol. 171:197–200 Veltri R, Rodriguez R. Molecular biology, endocrinology, and physiology of the prostate and seminal vesicles. In: Wein AJ, Kavoussi LR, Novick AC, Partiri AW, Peters CA, editors. Campbell-Walsh Urology. Volume 3. 9th Edition. Philadelphia: Saunders Elsevier; 2007. p2677-2726 Vermes I, Haanen C, Steffens-Nakken H, Reutelingsperger C. (1995) A novel assay for apoptosis. Flow cytometric detection of phosphatidylserine expression on early apoptotic cells using fluorescein labelled Annexin V. J Immunol Meth. 184:39-51 Visca P, Sebastiani V, Botti C, Diodoro MG, Lasagni RP, Romagnoli F, Brenna A, De Joannon BC, Donnorso RP, Lombardi G, Alo PL. (2004) Fatty acid synthase (FAS) is a marker of increased risk of recurrence in lung carcinoma. Anticancer Res 24:4169–4173 Vlodavsky I, Goldshmidt O, Zcharia E, Atzmon R, Rangini-Guatta Z, Elkin M, Peretz T, Friedmann Y. (2002) Mammalian heparanase: involvement in cancer metastasis, angiogenesis and normal development. Semin Cancer Biol 12:121-129 Voutilainen K, Anttila M, Sillanpää S, Tammi R, Tammi M, Saarikoski S, Kosma VM. (2003) Versican in epithelial ovarian cancer: relation to hyaluronan, clinicopathologic factors and prognosis. Int J Cancer 107:359-364 Vynios DH, Theocharis DA, Papageorgakopoulou N, Papadas TA, Mastronikolis NS, Goumas PD, Stylianou M, Skandalis SS. (2008) Biochemical changes of extracellular proteoglycans in squamous cell laryngeal carcinoma. Connect Tissue Res 49:239-243 Wang Y, Zhou X, Zhu H, Liu S, Zhou C, Zhang G, Xue L, Lu N, Quan L, Bai J, Zhan Q, Xu N. (2005) Overexpression of EB1 in human esophageal squamous cell carcinoma (ESCC) may promote cellular growth by activating beta-catenin/TCF pathway Oncogene. 24:6637-45 Page | 261 Chapter 7 References Wang YY, Kuhajda FP, Li J, Finch TT, Cheng P, Koh C, Li T, Sokoll LJ, Chan DW. (2004) Fatty acid synthase as a tumor marker: its extracellular expression in human breast cancer. J Exp Ther Oncol 4:101–110 Wegrowski Y, Maquart FX. (2004) Involvement of stromal proteoglycans in tumour progression. Crit Rev Oncol Hematol 49:259–268 Welsh JB, Sapinoso LM, Su AI, Kern SG, Wang-Rodriguez J, Moskaluk CA, Frierson HF Jr, Hampton GM. (2001) Analysis of gene expression identifies candidate markers and pharmacological targets in prostate cancer. Cancer Res 61:5974-5978 Wilfinger WW, Mackey M, Chomczynski P. (1997) Effect of pH and ionic strength on the spectrophotometric assessment of nucleic acid purity. BioTechniques 22:474- 476 Wissmann C, Wild PJ, Kaiser S, Roepcke S, Stoehr R, Woenckhaus M, Kristiansen G, Hsieh JC, Hofstaedter F, Hartmann A, Knuechel R, Rosenthal A, Pilarsky C. (2003) WIF1, a component of the Wnt pathway, is down-regulated in prostate, breast, lung, and bladder cancer. J Pathol 201:204-212 Wu Z, Irizarry RA. (2004) Preprocessing of oligonucleotide array data. Nature Biotech. 22:656-658 Xia G, Chen J, Tiwari V, Ju W, Li JP, Malmstrom A, Shukla D, Liu J. (2002) Heparan sulfate 3-O-sulfotransferase isoform 5 generates both an antithrombin- binding site and an entry receptor for herpes simplex virus, type 1. J Biol Chem. 277:37912-37919 Xiang YY, Ladeda V, Filmus J. (2001) Glypican-3 expression is silenced in human breast cancer. Oncogene 20:7408-7412 Yada T, Sato T, Kaseyama H, Gotoh M, Iwasaki H, Kikuchi N, Kwon YD, Togayachi A, Kudo T, Watanabe H, Narimatsu H, Kimata K. (2003) Chondroitin sulfate synthase-3. Molecular cloning and characterization J Biol Chem. 278:39711-39725 Page | 262 Chapter 7 References Yamauchi S, Mita S, Matsubara T, Fukuta M, Habuchi H, Kimata K, Habuchi O. (2000) Molecular cloning and expression of chondroitin-4-sulfotransferase. J Biol Chem 275:8975-8981 Yang J, Price MA, Neudauer CL, Wilson C, Ferrone S, Xia H, Iida J, Simpson MA, McCarthy JB. (2004) Melanoma chondroitin sulfate proteoglycan enhances FAK and ERK activation by distinct mechanisms. J Cell Biol 165:881-891 Yang J, Weinberg RA. (2008) Epithelial–mesenchymal transition: at the crossroads of development and tumor metastasis. Dev Cell 14:818–829 Yao SL, Lu-Yao G. (2002) Understanding and appreciating overdiagnosis in the PSA era. J Natl Cancer Inst 94:958–960 Yip GW, Smollich M, Gotte M. (2006) Therapeutic value of glycosaminoglycans in cancer. Mol Cancer Ther 5:2139-2148 Zajchowski DA, Bartholdi MF, Gong Y, Webster L, Liu HL, Munishkin A, Beauheim C, Harvey S, Ethier SP, Johnson PH. (2001) Identification of gene expression profiles that predict the aggressive behavior of breast cancer cells. Cancer Res. 61:5168–78 Zhivotovsky B, Orrenius S. (2009) The Warburg Effect returns to the cancer stage. Semin Cancer Biol 19:1-3 Zhou Y, Song B, Qin WJ, Zhang G, Zhang R, Luan Q, Pan TJ, Yang AG, Wang H. (2008) Heparanase promotes bone destruction and invasiveness in prostate cancer. Cancer Lett 268:252-259 Zhu Z, Friess H, Kleeff J, Wang L, Wirtz M, Zimmermann A, Korc M, Büchler MW. (2002) Glypican-3 expression is markedly decreased in human gastric cancer but not in esophageal cancer. Am J Surg 184:78-83 Page | 263 Appendix I APPENDIX I Gene Gene name Gene ID Fold Symbol change AMACR alpha-methylacyl-CoA racemase 23600 23.10 ZIC2 Zic family member 2 (odd-paired homolog, 7546 21.71 Drosophila) ELF5 E74-like factor 5 (ets domain transcription 2001 14.62 factor) PCA3 prostate cancer antigen 3 50652 13.74 HOXC6 homeobox C6 3223 12.38 GLYATL1 glycine-N-acyltransferase-like 1 92292 10.13 IGSF4D immunoglobulin superfamily, member 4D 253559 9.45 TMEM45B transmembrane protein 45B 120224 8.46 HIST3H2A histone 3, H2a 92815 8.28 DLX1 distal-less homeobox 1 1745 7.84 ALOX15B arachidonate 15-lipoxygenase, type B 247 7.41 CLDN7 claudin 7 1366 7.21 AGR2 anterior gradient 2 homolog (Xenopus 10551 6.77 laevis) NKX2-3 NK2 transcription factor related, locus 3 159296 6.68 (Drosophila) SPP1 secreted phosphoprotein 1 (osteopontin, 6696 6.23 bone sialoprotein I, early T-lymphocyte activation 1) SLC45A2 solute carrier family 45, member 2 51151 6.02 PGC progastricsin (pepsinogen C) 5225 5.78 CLDN3 claudin 3 1365 5.74 C15orf48 chromosome 15 open reading frame 48 84419 5.62 ASPN asporin (LRR class 1) 54829 5.58 HPN hepsin (transmembrane protease, serine 1) 3249 5.46 SGEF Src homology 3 domain-containing 26084 5.24 guanine nucleotide exchange factor CGREF1 cell growth regulator with EF-hand domain 10669 5.17 1 TRIM36 tripartite motif-containing 36 55521 5.10 TRPM4 transient receptor potential cation channel, 54795 5.03 subfamily M, member 4 MAL2 mal, T-cell differentiation protein 2 114569 4.89 HOXC4 homeobox C4 3221 4.82 SLC4A4 solute carrier family 4, sodium bicarbonate 8671 4.76 cotransporter, member 4 BAMBI BMP and activin membrane-bound 25805 4.72 inhibitor homolog (Xenopus laevis) CENPN centromere protein N 55839 4.53 FASN fatty acid synthase 2194 4.50 FABP5 fatty acid binding protein 5 (psoriasis- 2171 4.47 associated) /// similar to Fatty acid-binding protein, epidermal (E-FABP) (Psoriasis- Page | 264 Appendix I associated fatty acid-binding protein homolog) (PA-FABP) GCNT1 glucosaminyl (N-acetyl) transferase 1, core 2650 4.44 2 (beta-1,6-N- acetylglucosaminyltransferase) COMP cartilage oligomeric matrix protein 1311 4.44 GDF15 growth differentiation factor 15 9518 4.44 AK5 adenylate kinase 5 26289 4.41 LOC285463 hypothetical protein LOC285463 285463 4.35 RAB17 RAB17, member RAS oncogene family 64284 4.26 TRGC2 T cell receptor gamma constant 2 /// T cell 6967 4.20 receptor gamma variable 2 /// T cell receptor gamma variable 9 /// TCR gamma alternate reading frame protein /// hypothetical protein LOC642083 ITGBL1 integrin, beta-like 1 (with EGF-like repeat 9358 4.14 domains) FBP1 fructose-1,6-bisphosphatase 1 2203 4.14 MYC v-myc myelocytomatosis viral oncogene 4609 4.11 homolog (avian) FMO5 flavin containing monooxygenase 5 2330 4.11 B3GAT1 beta-1,3-glucuronyltransferase 1 27087 4.03 (glucuronosyltransferase P) SLC43A1 solute carrier family 43, member 1 8501 4.00 KIAA0114 KIAA0114 57291 3.97 HK2 hexokinase 2 3099 3.89 MYO6 myosin VI 4646 3.86 PPP1R9A protein phosphatase 1, regulatory 55607 3.86 (inhibitor) subunit 9A GJB1 gap junction protein, beta 1, 32kDa 2705 3.78 (connexin 32, Charcot-Marie-Tooth neuropathy, X-linked) tcag7.216 similar to matrilin 2 precursor 285929 3.73 HIST1H2A histone 1, H2am 8336 3.68 M CA13 carbonic anhydrase XIII 377677 3.68 STX19 syntaxin 19 415117 3.68 NUDT16P nudix (nucleoside diphosphate linked 152195 3.63 moiety X)-type motif 16 pseudogene KIAA0984 KIAA0984 protein 23329 3.63 REPS2 RALBP1 associated Eps domain containing 9185 3.51 2 RRM2 ribonucleotide reductase M2 polypeptide 6241 3.48 PYCR1 pyrroline-5-carboxylate reductase 1 5831 3.41 CCDC4 coiled-coil domain containing 4 389206 3.36 MYRIP myosin VIIA and Rab interacting protein 25924 3.32 CTNND2 catenin (cadherin-associated protein), delta 1501 3.32 2 (neural plakophilin-related arm-repeat protein) GARNL4 similar to GTPase activating Rap/RanGAP 23108 3.14 Page | 265 Appendix I domain-like 4 DNAH5 dynein, axonemal, heavy polypeptide 5 1767 3.14 SLC7A11 solute carrier family 7, (cationic amino acid 23657 3.10 transporter, y+ system) member 11 PLA2G7 phospholipase A2, group VII (platelet- 7941 3.10 activating factor acetylhydrolase, plasma) /// phospholipase A2, group VII (platelet- activating factor acetylhydrolase, plasma) CYP2J2 cytochrome P450, family 2, subfamily J, 1573 3.07 polypeptide 2 FHIT fragile histidine triad gene 2272 3.05 RAP1GAP RAP1 GTPase activating protein 5909 3.05 GPR160 G protein-coupled receptor 160 26996 3.05 STEAP4 STEAP family member 4 79689 3.03 MPP6 membrane protein, palmitoylated 6 51678 3.03 (MAGUK p55 subfamily member 6) CHMP4C chromatin modifying protein 4C 92421 3.03 ZDHHC11 zinc finger, DHHC-type containing 11 79844 3.01 NEDD4L neural precursor cell expressed, 23327 2.99 developmentally down-regulated 4-like PDLIM5 PDZ and LIM domain 5 10611 2.99 TARP T cell receptor gamma constant 2 /// T cell 445347 2.99 receptor gamma variable 2 /// T cell receptor gamma variable 9 /// TCR gamma alternate reading frame protein /// hypothetical protein LOC642083 ELL3 elongation factor RNA polymerase II-like 3 80237 2.99 UAP1 UDP-N-acteylglucosamine 6675 2.99 pyrophosphorylase 1 MBOAT2 membrane bound O-acyltransferase domain 129642 2.93 containing 2 GALNT7 UDP-N-acetyl-alpha-D- 51809 2.87 galactosamine:polypeptide N- acetylgalactosaminyltransferase 7 (GalNAc-T7) SIM2 single-minded homolog 2 (Drosophila) 6493 2.81 ABCC4 ATP-binding cassette, sub-family C 10257 2.77 (CFTR/MRP), member 4 FLJ20021 hypothetical LOC90024 90024 2.75 MCCC2 methylcrotonoyl-Coenzyme A carboxylase 64087 2.71 2 (beta) ZNF96 zinc finger protein 96 9753 2.68 LOC440731 hypothetical LOC440731 440731 2.68 TLCD1 TLC domain containing 1 116238 2.60 RIMS1 regulating synaptic membrane exocytosis 1 22999 2.58 PCSK6 proprotein convertase subtilisin/kexin type 5046 2.57 6 PCDHGA4 protocadherin gamma subfamily A, 4 56111 2.57 C9orf58 chromosome 9 open reading frame 58 83543 2.57 SMS spermine synthase 6611 2.55 Page | 266 Appendix I AADACL1 arylacetamide deacetylase-like 1 57552 2.55 MAOA monoamine oxidase A 4128 2.55 ECT2 epithelial cell transforming sequence 2 1894 2.53 oncogene SEC24A SEC24 related gene family, member A (S. 10802 2.53 cerevisiae) /// CXYorf1-related protein ACAD8 acyl-Coenzyme A dehydrogenase family, 27034 2.50 member 8 ELOVL7 ELOVL family member 7, elongation of 79993 2.50 long chain fatty acids (yeast) RHPN2 rhophilin, Rho GTPase binding protein 2 85415 2.46 MGC72075 hypothetical protein MGC72075 340277 2.46 GPR98 G protein-coupled receptor 98 84059 2.46 HMMR hyaluronan-mediated motility receptor 3161 2.46 (RHAMM) IL20RA interleukin 20 receptor, alpha 53832 2.43 SLC27A2 solute carrier family 27 (fatty acid 11001 2.43 transporter), member 2 GOLPH2 golgi phosphoprotein 2 51280 2.43 ZCCHC6 zinc finger, CCHC domain containing 6 79670 2.43 RBM35B RNA binding motif protein 35B 80004 2.41 LOC286189 hypothetical protein LOC286189 286189 2.39 PCCB propionyl Coenzyme A carboxylase, beta 5096 2.38 polypeptide TSPAN13 Tetraspanin 13 27075 2.38 SLC19A1 solute carrier family 19 (folate transporter), 6573 2.38 member 1 SLC19A2 solute carrier family 19 (thiamine 10560 2.38 transporter), member 2 PPAT phosphoribosyl pyrophosphate 5471 2.38 amidotransferase C9orf65 chromosome 9 open reading frame 65 158471 2.36 MTAC2D1 membrane targeting (tandem) C2 domain 123036 2.35 containing 1 C2orf37 chromosome 2 open reading frame 37 80067 2.33 SRPRB signal recognition particle receptor, B 58477 2.33 subunit LOC389834 hypothetical gene supported by AK123403 389834 2.31 /// hypothetical protein LOC642398 PDIA5 protein disulfide isomerase family A, 10954 2.31 member 5 SLC26A2 solute carrier family 26 (sulfate 1836 2.30 transporter), member 2 TUFT1 tuftelin 1 7286 2.30 B3GALT6 UDP-Gal:betaGal beta 1,3- 126792 2.28 galactosyltransferase polypeptide 6 KM-HN-1 KM-HN-1 protein 256309 2.28 PSAT1 phosphoserine aminotransferase 1 29968 2.28 RPL22L1 ribosomal protein L22-like 1 /// similar to 200916 2.27 ribosomal protein L22 like 1 Page | 267 Appendix I TMEM144 transmembrane protein 144 55314 2.25 IGFBP2 insulin-like growth factor binding protein 3485 2.25 2, 36kDa RAB11FIP1 RAB11 family interacting protein 1 (class 80223 2.23 I) C1orf97 chromosome 1 open reading frame 97 /// 84791 2.22 chromosome 1 open reading frame 97 ANKRD22 ankyrin repeat domain 22 118932 2.22 PIGW phosphatidylinositol glycan anchor 284098 2.20 biosynthesis, class W PVRL3 poliovirus receptor-related 3 25945 2.20 ZNF30 zinc finger protein 30 90075 2.20 KIAA1244 KIAA1244 57221 2.20 SLC12A8 solute carrier family 12 (potassium/chloride 84561 2.20 transporters), member 8 PAOX polyamine oxidase (exo-N4-amino) 196743 2.20 LARP2 La ribonucleoprotein domain family, 55132 2.19 member 2 GALNT3 UDP-N-acetyl-alpha-D- 2591 2.19 galactosamine:polypeptide N- acetylgalactosaminyltransferase 3 (GalNAc-T3) BMSC- PNC1 protein 84275 2.19 MCP C4orf13 Chromosome 4 open reading frame 13 84068 2.19 LOC401097 Similar to LOC166075 401097 2.17 ASNS asparagine synthetase 440 2.17 TSPAN1 tetraspanin 1 10103 2.17 SLC7A1 solute carrier family 7 (cationic amino acid 6541 2.16 transporter, y+ system), member 1 APOE apolipoprotein E 348 2.16 PAICS phosphoribosylaminoimidazole 10606 2.16 carboxylase, phosphoribosylaminoimidazole succinocarboxamide synthetase FAM83H family with sequence similarity 83, 286077 2.16 member H MTHFD2 methylenetetrahydrofolate dehydrogenase 10797 2.14 (NADP+ dependent) 2, methenyltetrahydrofolate cyclohydrolase RIPK2 receptor-interacting serine-threonine kinase 8767 2.14 2 ILDR1 immunoglobulin-like domain containing 286676 2.14 receptor 1 HSD17B4 hydroxysteroid (17-beta) dehydrogenase 4 3295 2.14 LYPLAL1 lysophospholipase-like 1 127018 2.13 TRIB3 tribbles homolog 3 (Drosophila) 57761 2.13 WIPI1 WD repeat domain, phosphoinositide 55062 2.13 interacting 1 ZBTB10 zinc finger and BTB domain containing 10 65986 2.13 Page | 268 Appendix I DHCR24 24-dehydrocholesterol reductase 1718 2.10 C7orf24 chromosome 7 open reading frame 24 79017 2.08 LYPLA1 lysophospholipase I 10434 2.07 RBM35A RNA binding motif protein 35A 54845 2.07 RFC5 CS box-containing WD protein 5985 2.06 JOSD3 Josephin domain containing 3 79101 2.06 LRRC8E leucine rich repeat containing 8 family, 80131 2.06 member E LRBA LPS-responsive vesicle trafficking, beach 987 2.06 and anchor containing KIAA0746 KIAA0746 protein 23231 2.06 SHRM Transcribed locus 57619 2.04 PUS7 pseudouridylate synthase 7 homolog (S. 54517 2.04 cerevisiae) TMTC4 transmembrane and tetratricopeptide repeat 84899 2.04 containing 4 NLN neurolysin (metallopeptidase M3 family) 57486 2.04 STYK1 serine/threonine/tyrosine kinase 1 55359 2.04 THSD1 thrombospondin, type I, domain containing 55901 2.03 1 /// thrombospondin, type I, domain containing 1 pseudogene ChGn chondroitin beta1,4 N- 55790 2.03 acetylgalactosaminyltransferase ZDHHC23 zinc finger, DHHC-type containing 23 254887 2.01 FLJ10357 hypothetical protein FLJ10357 55701 -2.00 KLHL5 kelch-like 5 (Drosophila) 51088 -2.01 NLGN4Y neuroligin 4, Y-linked 22829 -2.01 ZNF532 zinc finger protein 532 55205 -2.01 VAV3 vav 3 oncogene 10451 -2.03 TCEA2 transcription elongation factor A (SII), 2 6919 -2.03 IFI16 interferon, gamma-inducible protein 16 3428 -2.04 C6orf60 chromosome 6 open reading frame 60 79632 -2.04 NUDT10 nudix (nucleoside diphosphate linked 170685 -2.04 moiety X)-type motif 10 TCF8 transcription factor 8 (represses interleukin 6935 -2.04 2 expression) ATP2A2 ATPase, Ca++ transporting, cardiac 488 -2.04 muscle, slow twitch 2 FXYD6 FXYD domain containing ion transport 53826 -2.06 regulator 6 CRISPLD2 cysteine-rich secretory protein LCCL 83716 -2.06 domain containing 2 SERPINB1 serpin peptidase inhibitor, clade B 1992 -2.06 (ovalbumin), member 1 SLC27A3 solute carrier family 27 (fatty acid 11000 -2.06 transporter), member 3 LOC153682 Hypothetical protein LOC644009 153682 -2.06 MAML2 mastermind-like 2 (Drosophila) 84441 -2.06 PRG1 proteoglycan 1, secretory granule 5552 -2.06 SNX7 sorting nexin 7 51375 -2.07 Page | 269 Appendix I TPBG trophoblast glycoprotein 7162 -2.07 PLEKHH2 pleckstrin homology domain containing, 130271 -2.07 family H (with MyTH4 domain) member 2 SLC8A1 solute carrier family 8 (sodium/calcium 6546 -2.07 exchanger), member 1 ZNF512 zinc finger protein 512 84450 -2.07 FGFR1 fibroblast growth factor receptor 1 (fms- 2260 -2.07 related tyrosine kinase 2, Pfeiffer syndrome) FLJ13197 hypothetical protein FLJ13197 79667 -2.07 STARD4 START domain containing 4, sterol 134429 -2.07 regulated RGMB RGM domain family, member B 285704 -2.08 PPP1R3C protein phosphatase 1, regulatory 5507 -2.10 (inhibitor) subunit 3C LOC283481 hypothetical protein LOC283481 283481 -2.10 PELI2 pellino homolog 2 (Drosophila) 57161 -2.10 PGAP1 GPI deacylase 80055 -2.10 WNT5B wingless-type MMTV integration site 81029 -2.10 family, member 5B /// wingless-type MMTV integration site family, member 5B EID3 E1A-like inhibitor of differentiation 3 493861 -2.10 SGCB sarcoglycan, beta (43kDa dystrophin- 6443 -2.10 associated glycoprotein) FAT4 FAT tumor suppressor homolog 4 79633 -2.10 (Drosophila) CG018 hypothetical gene CG018 90634 -2.10 DCN decorin 1634 -2.10 TMEM16A transmembrane protein 16A 55107 -2.11 PYGM phosphorylase, glycogen; muscle (McArdle 5837 -2.11 syndrome, glycogen storage disease type V) C20orf194 chromosome 20 open reading frame 194 25943 -2.11 NES nestin 10763 -2.11 LOC285989 zinc finger protein 789 285989 -2.11 FGD5 FYVE, RhoGEF and PH domain 152273 -2.11 containing 5 NALP1 NACHT, leucine rich repeat and PYD 22861 -2.11 (pyrin domain) containing 1 GPRASP1 G protein-coupled receptor associated 9737 -2.11 sorting protein 1 GNB5 guanine nucleotide binding protein (G 10681 -2.13 protein), beta 5 SYTL4 synaptotagmin-like 4 (granuphilin-a) 94121 -2.13 TFPI tissue factor pathway inhibitor (lipoprotein- 7035 -2.13 associated coagulation inhibitor) PGRMC1 progesterone receptor membrane 10857 -2.13 component 1 TUBB6 tubulin, beta 6 84617 -2.13 TRPC4 transient receptor potential cation channel, 7223 -2.13 Page | 270 Appendix I subfamily C, member 4 CUGBP2 CUG triplet repeat, RNA binding protein 2 10659 -2.13 LAMA4 laminin, alpha 4 3910 -2.13 ALDH1L2 aldehyde dehydrogenase 1 family, member 160428 -2.13 L2 RASSF5 Ras association (RalGDS/AF-6) domain 83593 -2.14 family 5 ZDHHC15 zinc finger, DHHC-type containing 15 158866 -2.16 TSC22D3 TSC22 domain family, member 3 1831 -2.16 SLC24A3 solute carrier family 24 57419 -2.16 (sodium/potassium/calcium exchanger), member 3 PRKAR2B protein kinase, cAMP-dependent, 5577 -2.16 regulatory, type II, beta C4orf31 chromosome 4 open reading frame 31 79625 -2.16 C1R complement component 1, r subcomponent 715 -2.16 /// similar to Complement C1r subcomponent precursor (Complement component 1, r subcomponent) SERPINE2 serpin peptidase inhibitor, clade E (nexin, 5270 -2.17 plasminogen activator inhibitor type 1), member 2 DIXDC1 DIX domain containing 1 85458 -2.17 CASP1 caspase 1, apoptosis-related cysteine 834 -2.17 peptidase (interleukin 1, beta, convertase) ARHGEF4 Rho guanine nucleotide exchange factor 50649 -2.17 (GEF) 4 MEIS3P1 Meis1 homolog 3 (mouse) pseudogene 1 4213 -2.17 FBXL7 F-box and leucine-rich repeat protein 7 23194 -2.17 C1S complement component 1, s subcomponent 716 -2.19 BTG2 BTG family, member 2 7832 -2.19 EVI1 ecotropic viral integration site 1 2122 -2.19 ITGB1BP2 integrin beta 1 binding protein (melusin) 2 26548 -2.19 MBNL2 muscleblind-like 2 (Drosophila) 10150 -2.19 MSN moesin 4478 -2.19 C7 complement component 7 730 -2.19 FAM46A family with sequence similarity 46, 55603 -2.20 member A OR2A20P Olfactory receptor, family 2, subfamily A, 401428 -2.20 member 7 PSCDBP pleckstrin homology, Sec7 and coiled-coil 9595 -2.20 domains, binding protein /// pleckstrin homology, Sec7 and coiled-coil domains, binding protein SERPING1 serpin peptidase inhibitor, clade G (C1 710 -2.20 inhibitor), member 1, (angioedema, hereditary) GPM6B glycoprotein M6B 2824 -2.20 RSNL2 restin-like 2 79745 -2.20 FZD1 frizzled homolog 1 (Drosophila) 8321 -2.22 Page | 271 Appendix I CD200 CD200 molecule 4345 -2.22 HEPH hephaestin 9843 -2.22 PPP1R14A protein phosphatase 1, regulatory 94274 -2.22 (inhibitor) subunit 14A CPVL carboxypeptidase, vitellogenic-like /// 54504 -2.22 carboxypeptidase, vitellogenic-like TSPAN2 tetraspanin 2 10100 -2.22 RASSF6 Ras association (RalGDS/AF-6) domain 166824 -2.22 family 6 DLL1 delta-like 1 (Drosophila) 28514 -2.23 HS3ST3B1 heparan sulfate (glucosamine) 3-O- 9953 -2.23 sulfotransferase 3B1 MAGI2 membrane associated guanylate kinase, 9863 -2.23 WW and PDZ domain containing 2 MYADM myeloid-associated differentiation marker 91663 -2.23 ITIH5 inter-alpha (globulin) inhibitor H5 80760 -2.23 BDH2 3-hydroxybutyrate dehydrogenase, type 2 56898 -2.25 NRP1 neuropilin 1 8829 -2.25 RAI2 retinoic acid induced 2 10742 -2.25 PLEKHA6 pleckstrin homology domain containing, 22874 -2.25 family A member 6 HTRA1 HtrA serine peptidase 1 5654 -2.25 DSC3 desmocollin 3 1825 -2.25 CCL19 chemokine (C-C motif) ligand 19 6363 -2.25 BOC Boc homolog (mouse) 91653 -2.27 LAMB2 laminin, beta 2 (laminin S) 3913 -2.27 MATN2 matrilin 2 4147 -2.27 TRPC1 transient receptor potential cation channel, 7220 -2.27 subfamily C, member 1 PCOLCE procollagen C-endopeptidase enhancer 5118 -2.27 ALDH1A1 aldehyde dehydrogenase 1 family, member 216 -2.27 A1 TRIM6 tripartite motif-containing 6 117854 -2.27 LDHB lactate dehydrogenase B 3945 -2.27 ZNF423 zinc finger protein 423 23090 -2.27 SELENBP1 selenium binding protein 1 /// selenium 8991 -2.27 binding protein 1 SELM selenoprotein M 140606 -2.27 SRPX sushi-repeat-containing protein, X-linked 8406 -2.28 LOC619208 hypothetical protein LOC619208 619208 -2.28 TIMP1 TIMP metallopeptidase inhibitor 1 7076 -2.28 ANG angiogenin, ribonuclease, RNase A family, 283 -2.28 5 /// ribonuclease, RNase A family, 4 HMGCLL1 3-hydroxymethyl-3-methylglutaryl- 54511 -2.28 Coenzyme A lyase-like 1 FOXF1 forkhead box F1 2294 -2.28 GPR124 G protein-coupled receptor 124 25960 -2.28 GALNTL1 UDP-N-acetyl-alpha-D- 57452 -2.28 galactosamine:polypeptide N- acetylgalactosaminyltransferase-like 1 /// Page | 272 Appendix I similar to UDP-N-acetyl-alpha-D- galactosamine:polypeptide N- acetylgalactosaminyltransferase-like 1 SLMAP sarcolemma associated protein 7871 -2.28 MXRA7 matrix-remodelling associated 7 439921 -2.28 EPHA4 EPH receptor A4 2043 -2.30 IL8 interleukin 8 3576 -2.30 PAK1 p21/Cdc42/Rac1-activated kinase 1 (STE20 5058 -2.30 homolog, yeast) DKK3 dickkopf homolog 3 (Xenopus laevis) 27122 -2.31 LOC493869 similar to RIKEN cDNA 2310016C16 493869 -2.31 GRK5 G protein-coupled receptor kinase 5 2869 -2.31 EHBP1L1 hypothetical protein MGC15523 /// EH 254102 -2.31 domain binding protein 1-like 1 TTMB TTMB protein 399474 -2.33 PARD6G par-6 partitioning defective 6 homolog 84552 -2.33 gamma (C. elegans) TWIST2 twist homolog 2 (Drosophila) 117581 -2.33 ACTN1 actinin, alpha 1 87 -2.33 EFHA2 EF-hand domain family, member A2 286097 -2.33 ADCY5 adenylate cyclase 5 111 -2.33 APCDD1 adenomatosis polyposis coli down- 147495 -2.33 regulated 1 ANK2 ankyrin 2, neuronal 287 -2.33 IFITM3 interferon induced transmembrane protein 10410 -2.35 3 (1-8U) MSRB3 methionine sulfoxide reductase B3 253827 -2.35 LTBP3 latent transforming growth factor beta 4054 -2.35 binding protein 3 DKFZP761 hypothetical protein DKFZP761M1511 54492 -2.35 M1511 ARMCX2 armadillo repeat containing, X-linked 2 9823 -2.35 ACAA2 acetyl-Coenzyme A acyltransferase 2 10449 -2.35 (mitochondrial 3-oxoacyl-Coenzyme A thiolase) FLJ21963 FLJ21963 protein 79611 -2.35 IFITM2 interferon induced transmembrane protein 10581 -2.35 2 (1-8D) BTG3 BTG family, member 3 10950 -2.36 HDAC9 histone deacetylase 9 9734 -2.36 BHLHB2 basic helix-loop-helix domain containing, 8553 -2.36 class B, 2 EFNB3 ephrin-B3 1949 -2.36 BACH2 BTB and CNC homology 1, basic leucine 60468 -2.36 zipper transcription factor 2 /// BTB and CNC homology 1, basic leucine zipper transcription factor 2 NDN necdin homolog (mouse) 4692 -2.36 KNTC2 kinetochore associated 2 10403 -2.36 C1orf54 chromosome 1 open reading frame 54 79630 -2.36 Page | 273 Appendix I ICAM2 intercellular adhesion molecule 2 3384 -2.36 C14orf78 chromosome 14 open reading frame 78 113146 -2.36 PDGFC platelet derived growth factor C 56034 -2.38 KCNE3 potassium voltage-gated channel, Isk- 10008 -2.38 related family, member 3 GNAZ guanine nucleotide binding protein (G 2781 -2.38 protein), alpha z polypeptide KIAA1713 KIAA1713 80816 -2.38 GBP3 guanylate binding protein 3 2635 -2.38 MID1 midline 1 (Opitz/BBB syndrome) 4281 -2.38 BIN1 bridging integrator 1 274 -2.39 CAPG capping protein (actin filament), gelsolin- 822 -2.39 like TRAF3IP2 TRAF3 interacting protein 2 10758 -2.39 LOC399959 hypothetical gene supported by BX647608 399959 -2.39 RP1- hypothetical protein LOC441168 441168 -2.39 93H18.5 TP53INP2 tumor protein p53 inducible nuclear protein 58476 -2.39 2 S100A6 S100 calcium binding protein A6 6277 -2.39 (calcyclin) PLEKHO1 pleckstrin homology domain containing, 51177 -2.39 family O member 1 ISL1 ISL1 transcription factor, 3670 -2.39 LIM/homeodomain, (islet-1) RGS13 regulator of G-protein signalling 13 6003 -2.39 CFL2 cofilin 2 (muscle) /// cofilin 2 (muscle) 1073 -2.41 SLC40A1 solute carrier family 40 (iron-regulated 30061 -2.41 transporter), member 1 LRFN5 leucine rich repeat and fibronectin type III 145581 -2.41 domain containing 5 PLEKHC1 pleckstrin homology domain containing, 10979 -2.41 family C (with FERM domain) member 1 tcag7.981 juxtaposed with another zinc finger gene 1 221895 -2.41 FZD7 frizzled homolog 7 (Drosophila) 8324 -2.41 FBXO32 F-box protein 32 114907 -2.43 PTGIS prostaglandin I2 (prostacyclin) synthase /// 5740 -2.43 prostaglandin I2 (prostacyclin) synthase JAM3 junctional adhesion molecule 3 83700 -2.43 EVA1 epithelial V-like antigen 1 10205 -2.43 C9orf26 chromosome 9 open reading frame 26 (NF- 90865 -2.43 HEV) FLJ39155 EGF-like, fibronectin type III and laminin 133584 -2.43 G domains DMD dystrophin (muscular dystrophy, Duchenne 1756 -2.43 and Becker types) CAV1 caveolin 1, caveolae protein, 22kDa 857 -2.43 ADD3 adducin 3 (gamma) 120 -2.45 FAM101B family with sequence similarity 101, 359845 -2.45 member B Page | 274 Appendix I MXRA6 Tensin 1 /// Tensin 1 54589 -2.45 ITGB8 integrin, beta 8 3696 -2.45 LOC203274 hypothetical protein LOC203274 203274 -2.45 IRAK1BP1 interleukin-1 receptor-associated kinase 1 134728 -2.45 binding protein 1 CKMT2 creatine kinase, mitochondrial 2 1160 -2.45 (sarcomeric) MCAM melanoma cell adhesion molecule 4162 -2.46 C9orf3 chromosome 9 open reading frame 3 84909 -2.46 DSCR1L2 ESTs 11123 -2.46 HLF hepatic leukemia factor 3131 -2.46 MEIS1 Meis1, myeloid ecotropic viral integration 4211 -2.46 site 1 homolog (mouse) TLE2 transducin-like enhancer of split 2 (E(sp1) 7089 -2.46 homolog, Drosophila) GBP2 guanylate binding protein 2, interferon- 2634 -2.46 inducible /// guanylate binding protein 2, interferon-inducible HS3ST3A1 heparan sulfate (glucosamine) 3-O- 9955 -2.46 sulfotransferase 3A1 VIM vimentin 7431 -2.46 IER3 immediate early response 3 8870 -2.48 SPON1 spondin 1, extracellular matrix protein 10418 -2.48 TMEM16E ESTs 203859 -2.48 MEST mesoderm specific transcript homolog 4232 -2.48 (mouse) PROCR protein C receptor, endothelial (EPCR) 10544 -2.48 EMP3 epithelial membrane protein 3 2014 -2.48 CCR2 chemokine (C-C motif) receptor 2 /// 1231 -2.48 chemokine (C-C motif) receptor 2 NT5DC2 5'-nucleotidase domain containing 2 64943 -2.48 LOC283070 hypothetical protein LOC283070 283070 -2.48 SPOCK3 sparc/osteonectin, cwcv and kazal-like 50859 -2.48 domains proteoglycan (testican) 3 FZD2 frizzled homolog 2 (Drosophila) 2535 -2.48 BNC2 basonuclin 2 54796 -2.48 RBPMS RNA binding protein with multiple splicing 11030 -2.50 OLFML1 olfactomedin-like 1 283298 -2.50 PPP1R12B protein phosphatase 1, regulatory 4660 -2.50 (inhibitor) subunit 12B BVES blood vessel epicardial substance 11149 -2.50 COL18A1 collagen, type XVIII, alpha 1 80781 -2.50 DUSP6 dual specificity phosphatase 6 1848 -2.50 GAS1 growth arrest-specific 1 2619 -2.51 LGALS3BP lectin, galactoside-binding, soluble, 3 3959 -2.51 binding protein C16orf45 chromosome 16 open reading frame 45 89927 -2.53 CDC42EP3 CDC42 effector protein (Rho GTPase 10602 -2.53 binding) 3 HOXB3 homeobox B3 3213 -2.53 Page | 275 Appendix I CAP2 CAP, adenylate cyclase-associated protein, 10486 -2.53 2 (yeast) EPB41L3 erythrocyte membrane protein band 4.1- 23136 -2.53 like 3 HEY1 hairy/enhancer-of-split related with YRPW 23462 -2.55 motif 1 ASB2 ankyrin repeat and SOCS box-containing 2 51676 -2.55 RRAS related RAS viral (r-ras) oncogene 6237 -2.55 homolog INMT indolethylamine N-methyltransferase 11185 -2.57 PCDH18 protocadherin 18 54510 -2.57 RND3 Rho family GTPase 3 390 -2.57 LAMP3 lysosomal-associated membrane protein 3 27074 -2.57 SERPINA7 Nik related kinase 6906 -2.57 CARD15 Transcribed locus, weakly similar to 64127 -2.57 XP_520634.1 PREDICTED: similar to naked cuticle homolog 1; naked cuticle-1; Dvl-binding protein [Pan troglodytes] MRGPRF MAS-related GPR, member F 219928 -2.57 FLJ10847 hypothetical protein FLJ10847 55244 -2.58 ISYNA1 myo-inositol 1-phosphate synthase A1 51477 -2.58 MT1X metallothionein 1X 4501 -2.58 TLR3 toll-like receptor 3 7098 -2.58 SVIL supervillin 6840 -2.58 EDNRA endothelin receptor type A 1909 -2.58 SCRN1 secernin 1 9805 -2.58 LOC402560 hypothetical LOC402560 402560 -2.60 DNALI1 dynein, axonemal, light intermediate 7802 -2.60 polypeptide 1 CHRM1 cholinergic receptor, muscarinic 1 1128 -2.60 PRKCB1 protein kinase C, beta 1 5579 -2.60 FEZ1 fasciculation and elongation protein zeta 1 9638 -2.60 (zygin I) NAV2 neuron navigator 2 89797 -2.60 C21orf63 chromosome 21 open reading frame 63 59271 -2.60 EYA4 ESTs 2070 -2.60 NT5E 5'-nucleotidase, ecto (CD73) 4907 -2.60 GNG2 guanine nucleotide binding protein (G 54331 -2.60 protein), gamma 2 LARGE like-glycosyltransferase 9215 -2.62 PDE8B phosphodiesterase 8B 8622 -2.62 RBP7 retinol binding protein 7, cellular 116362 -2.62 FOXC1 forkhead box C1 2296 -2.62 NID1 nidogen 1 4811 -2.62 PDZRN4 PDZ domain containing RING finger 4 29951 -2.62 SYNGR1 synaptogyrin 1 9145 -2.62 AKAP7 A kinase (PRKA) anchor protein 7 9465 -2.62 AMT aminomethyltransferase 275 -2.64 INPP1 inositol polyphosphate-1-phosphatase 3628 -2.64 RASL12 RAS-like, family 12 51285 -2.64 Page | 276 Appendix I PLCL2 phospholipase C-like 2 23228 -2.64 SORBS1 sorbin and SH3 domain containing 1 10580 -2.64 STOM stomatin 2040 -2.64 MMP2 matrix metallopeptidase 2 (gelatinase A, 4313 -2.66 72kDa gelatinase, 72kDa type IV collagenase) SERPINB9 serpin peptidase inhibitor, clade B 5272 -2.66 (ovalbumin), member 9 PALMD palmdelphin 54873 -2.66 TMEM16D transmembrane protein 16D 121601 -2.66 DACH1 dachshund homolog 1 (Drosophila) 1602 -2.68 Gcom1 GRINL1A combined protein 145781 -2.68 SPATA18 spermatogenesis associated 18 homolog 132671 -2.68 (rat) HOXB2 homeobox B2 3212 -2.68 PTPLA protein tyrosine phosphatase-like (proline 9200 -2.68 instead of catalytic arginine), member A DDIT4L DNA-damage-inducible transcript 4-like 115265 -2.68 CPA3 carboxypeptidase A3 (mast cell) 1359 -2.68 GMNN geminin, DNA replication inhibitor 51053 -2.68 CX3CL1 chemokine (C-X3-C motif) ligand 1 6376 -2.69 MGC39900 hypothetical protein MGC39900 286527 -2.69 AYTL1 acyltransferase like 1 54947 -2.69 MET met proto-oncogene (hepatocyte growth 4233 -2.69 factor receptor) NHS Nance-Horan syndrome (congenital 4810 -2.71 cataracts and dental anomalies) S100A8 S100 calcium binding protein A8 6279 -2.71 (calgranulin A) SEMA6D sema domain, transmembrane domain 80031 -2.71 (TM), and cytoplasmic domain, (semaphorin) 6D NEU1 FYN oncogene related to SRC, FGR, YES 4758 -2.71 /// sialidase 1 (lysosomal sialidase) EFEMP2 EGF-containing fibulin-like extracellular 30008 -2.71 matrix protein 2 ST5 suppression of tumorigenicity 5 6764 -2.73 METTL7A methyltransferase like 7A 25840 -2.73 FAM92A1 family with sequence similarity 92, 137392 -2.73 member A1 TUBA1 tubulin, alpha 1 7277 -2.73 AKR1B1 aldo-keto reductase family 1, member B1 231 -2.73 (aldose reductase) CSPG2 chondroitin sulfate proteoglycan 2 1462 -2.73 (versican) GSTP1 glutathione S-transferase pi 2950 -2.73 OGN osteoglycin (osteoinductive factor, 4969 -2.73 mimecan) LRRC34 leucine rich repeat containing 34 151827 -2.75 GBP1 guanylate binding protein 1, interferon- 2633 -2.75 Page | 277 Appendix I inducible, 67kDa GSTM3 glutathione S-transferase M3 (brain) 2947 -2.75 ROBO1 roundabout, axon guidance receptor, 6091 -2.75 homolog 1 (Drosophila) FCGRT Fc fragment of IgG, receptor, transporter, 2217 -2.77 alpha SULF2 sulfatase 2 55959 -2.77 TCF7L1 transcription factor 7-like 1 (T-cell specific, 83439 -2.77 HMG-box) /// transcription factor 7-like 1 (T-cell specific, HMG-box) PTGS2 prostaglandin-endoperoxide synthase 2 5743 -2.77 (prostaglandin G/H synthase and cyclooxygenase) PRICKLE2 prickle-like 2 (Drosophila) 166336 -2.77 C1orf165 chromosome 1 open reading frame 165 79656 -2.79 BMP5 bone morphogenetic protein 5 653 -2.79 LMO4 LIM domain only 4 8543 -2.79 IFITM1 interferon induced transmembrane protein 8519 -2.81 1 (9-27) NID2 nidogen 2 (osteonidogen) 22795 -2.81 PDE4A phosphodiesterase 4A, cAMP-specific 5141 -2.81 (phosphodiesterase E2 dunce homolog, Drosophila) RGC32 response gene to complement 32 28984 -2.81 RBM24 RNA binding motif protein 24 221662 -2.81 FLNA filamin A, alpha (actin binding protein 280) 2316 -2.81 TMEM158 transmembrane protein 158 25907 -2.81 GPX7 glutathione peroxidase 7 2882 -2.83 RNASE4 ribonuclease, RNase A family, 4 6038 -2.83 COL4A3 collagen, type IV, alpha 3 (Goodpasture 1285 -2.83 antigen) PRSS16 protease, serine, 16 (thymus) 10279 -2.83 PDE5A phosphodiesterase 5A, cGMP-specific 8654 -2.83 FLRT2 fibronectin leucine rich transmembrane 23768 -2.85 protein 2 EDG2 endothelial differentiation, 1902 -2.85 lysophosphatidic acid G-protein-coupled receptor, 2 RBMS3 RNA binding motif, single stranded 27303 -2.85 interacting protein SDC1 syndecan 1 6382 -2.85 CTTNBP2 cortactin binding protein 2 83992 -2.85 JPH4 junctophilin 4 84502 -2.85 PRTFDC1 phosphoribosyl transferase domain 56952 -2.85 containing 1 C6orf32 chromosome 6 open reading frame 32 9750 -2.87 OLFML3 olfactomedin-like 3 56944 -2.87 TBC1D1 TBC1 (tre-2/USP6, BUB2, cdc16) domain 23216 -2.87 family, member 1 WNT2 wingless-type MMTV integration site 7472 -2.87 Page | 278 Appendix I family member 2 SLCO3A1 solute carrier organic anion transporter 28232 -2.87 family, member 3A1 TGFBR3 transforming growth factor, beta receptor 7049 -2.87 III (betaglycan, 300kDa) CXorf6 chromosome X open reading frame 6 10046 -2.87 FER1L3 fer-1-like 3, myoferlin (C. elegans) 26509 -2.89 R7BP R7 binding protein 401190 -2.89 COL6A2 collagen, type VI, alpha 2 1292 -2.89 TRHDE thyrotropin-releasing hormone degrading 29953 -2.89 enzyme EFHC2 EF-hand domain (C-terminal) containing 2 80258 -2.89 FLJ32363 FLJ32363 protein 375444 -2.89 TRIM22 tripartite motif-containing 22 10346 -2.89 NPAL3 NIPA-like domain containing 3 57185 -2.89 UBD ubiquitin D 10537 -2.91 APOBEC3 apolipoprotein B mRNA editing enzyme, 60489 -2.91 G catalytic polypeptide-like 3G RAB9B RAB9B, member RAS oncogene family 51209 -2.91 CDCA7L cell division cycle associated 7-like 55536 -2.93 ENPP2 ectonucleotide 5168 -2.93 pyrophosphatase/phosphodiesterase 2 (autotaxin) WFDC1 WAP four-disulfide core domain 1 58189 -2.93 APOD apolipoprotein D 347 -2.95 HHIP hedgehog interacting protein 64399 -2.95 TRIP6 thyroid hormone receptor interactor 6 7205 -2.95 CAV2 caveolin 2 858 -2.95 DES desmin /// family with sequence similarity 1674 -2.95 48, member A AOC3 amine oxidase, copper containing 3 8639 -2.95 (vascular adhesion protein 1) AOX1 aldehyde oxidase 1 316 -2.95 BST2 bone marrow stromal cell antigen 2 684 -2.97 PLK2 polo-like kinase 2 (Drosophila) 10769 -2.97 RBP1 retinol binding protein 1, cellular 5947 -2.97 TMLHE trimethyllysine hydroxylase, epsilon 55217 -2.97 ADAMTS9 ADAM metallopeptidase with 56999 -2.97 thrombospondin type 1 motif, 9 PPAP2B phosphatidic acid phosphatase type 2B 8613 -2.97 CRYAB crystallin, alpha B 1410 -2.99 BMP4 bone morphogenetic protein 4 652 -2.99 TSHZ3 teashirt family zinc finger 3 57616 -2.99 C20orf100 chromosome 20 open reading frame 100 84969 -2.99 ECRG4 chromosome 2 open reading frame 40 84417 -2.99 ZNF711 zinc finger protein 711 7552 -3.01 S100A4 S100 calcium binding protein A4 (calcium 6275 -3.01 protein, calvasculin, metastasin, murine placental homolog) LGALS1 lectin, galactoside-binding, soluble, 1 3956 -3.01 Page | 279 Appendix I (galectin 1) ARL4C ADP-ribosylation factor-like 4C 10123 -3.01 HSD11B1 hydroxysteroid (11-beta) dehydrogenase 1 3290 -3.01 THSD4 thrombospondin, type I, domain containing 79875 -3.03 4 DMN desmuslin 23336 -3.03 EMILIN1 elastin microfibril interfacer 1 11117 -3.03 LPHN2 latrophilin 2 23266 -3.03 SYNPO2 synaptopodin 2 171024 -3.05 ST6GAL2 ESTs 84620 -3.05 C3 complement component 3 /// similar to 718 -3.05 Complement C3 precursor SERTAD4 SERTA domain containing 4 56256 -3.05 CYBA cytochrome b-245, alpha polypeptide 1535 -3.05 CPM carboxypeptidase M 1368 -3.07 GPR126 G protein-coupled receptor 126 57211 -3.07 JUB jub, ajuba homolog (Xenopus laevis) 84962 -3.10 EBI2 Epstein-Barr virus induced gene 2 1880 -3.10 (lymphocyte-specific G protein-coupled receptor) C20orf42 chromosome 20 open reading frame 42 55612 -3.10 KIAA1462 KIAA1462 57608 -3.10 SMTN smoothelin 6525 -3.10 QPRT quinolinate phosphoribosyltransferase 23475 -3.10 (nicotinate-nucleotide pyrophosphorylase (carboxylating)) PTGDS prostaglandin D2 synthase 21kDa (brain) 5730 -3.12 ARHGAP20 Rho GTPase activating protein 20 57569 -3.12 C13orf21 chromosome 13 open reading frame 21 387923 -3.14 CSRP1 cysteine and glycine-rich protein 1 1465 -3.14 TF transferrin 7018 -3.14 COX7A1 cytochrome c oxidase subunit VIIa 1346 -3.16 polypeptide 1 (muscle) TBX5 T-box 5 6910 -3.16 LOC130576 hypothetical protein LOC130576 130576 -3.16 G0S2 G0/G1switch 2 50486 -3.16 LTBP4 latent transforming growth factor beta 8425 -3.16 binding protein 4 DBNDD2 dysbindin (dystrobrevin binding protein 1) 55861 -3.16 domain containing 2 GMPR guanosine monophosphate reductase /// 2766 -3.18 guanosine monophosphate reductase CCDC8 coiled-coil domain containing 8 83987 -3.18 CXCL12 chemokine (C-X-C motif) ligand 12 6387 -3.18 (stromal cell-derived factor 1) KLHL14 kelch-like 14 (Drosophila) 57565 -3.18 WNT4 wingless-type MMTV integration site 54361 -3.18 family, member 4 /// wingless-type MMTV integration site family, member 4 CYTL1 cytokine-like 1 54360 -3.20 Page | 280 Appendix I PLP2 proteolipid protein 2 (colonic epithelium- 5355 -3.20 enriched) CDO1 cysteine dioxygenase, type I 1036 -3.20 HOXD10 homeobox D10 3236 -3.23 ANXA6 annexin A6 309 -3.25 NAV3 neuron navigator 3 89795 -3.25 FOXQ1 forkhead box Q1 94234 -3.25 C1orf106 chromosome 1 open reading frame 106 55765 -3.25 ZNF179 zinc finger protein 179 7732 -3.25 CCL2 chemokine (C-C motif) ligand 2 6347 -3.25 ACTG2 actin, gamma 2, smooth muscle, enteric 72 -3.27 PNMA1 paraneoplastic antigen MA1 9240 -3.29 SOX7 SRY (sex determining region Y)-box 7 83595 -3.29 F10 coagulation factor X 2159 -3.29 TGFB1I1 transforming growth factor beta 1 induced 7041 -3.29 transcript 1 STOX2 storkhead box 2 56977 -3.32 RNF175 ring finger protein 175 285533 -3.32 BIC BIC transcript 114614 -3.34 LDB3 LIM domain binding 3 11155 -3.34 PYCARD PYD and CARD domain containing 29108 -3.34 NPTX2 neuronal pentraxin II 4885 -3.34 PTN pleiotrophin (heparin binding growth factor 5764 -3.36 8, neurite growth-promoting factor 1) /// pleiotrophin (heparin binding growth factor 8, neurite growth-promoting factor 1) PIP5K1B phosphatidylinositol-4-phosphate 5-kinase, 8395 -3.36 type I, beta SCN4B sodium channel, voltage-gated, type IV, 6330 -3.39 beta ABCA8 ATP-binding cassette, sub-family A 10351 -3.39 (ABC1), member 8 FOXF2 forkhead box F2 2295 -3.39 CNN1 calponin 1, basic, smooth muscle 1264 -3.41 HSPB8 heat shock 22kDa protein 8 26353 -3.41 ANKRD35 ankyrin repeat domain 35 148741 -3.41 IGFBP6 insulin-like growth factor binding protein 6 3489 -3.41 IGJ Immunoglobulin J polypeptide, linker 3512 -3.41 protein for immunoglobulin alpha and mu polypeptides TSLP thymic stromal lymphopoietin 85480 -3.41 PRKCA protein kinase C, alpha 5578 -3.43 ODZ2 odz, odd Oz/ten-m homolog 2 (Drosophila) 57451 -3.43 CHN1 chimerin (chimaerin) 1 1123 -3.43 FLJ10159 hypothetical protein FLJ10159 55084 -3.43 PCDH9 protocadherin 9 5101 -3.43 STARD8 START domain containing 8 9754 -3.46 ITM2C integral membrane protein 2C /// integral 81618 -3.46 membrane protein 2C PRIM2A primase, polypeptide 2A, 58kDa 5558 -3.46 Page | 281 Appendix I RP4- KIAA0980 protein 22981 -3.46 691N24.1 CYP4X1 cytochrome P450, family 4, subfamily X, 260293 -3.46 polypeptide 1 ANXA1 annexin A1 301 -3.46 LOC285382 hypothetical gene supported by AK091454 285382 -3.48 C11orf70 chromosome 11 open reading frame 70 /// 85016 -3.48 chromosome 11 open reading frame 70 LOC284244 hypothetical protein LOC284244 284244 -3.48 CLIC6 chloride intracellular channel 6 54102 -3.48 ZNF185 zinc finger protein 185 (LIM domain) 7739 -3.48 LOC387882 hypothetical protein 387882 -3.48 RAB34 RAB34, member RAS oncogene family 83871 -3.51 ALS2CR4 amyotrophic lateral sclerosis 2 (juvenile) 65062 -3.51 chromosome region, candidate 4 LMOD1 leiomodin 1 (smooth muscle) 25802 -3.51 FRMD6 FERM domain containing 6 122786 -3.51 FAM107A family with sequence similarity 107, 11170 -3.53 member A ATRNL1 attractin-like 1 26033 -3.53 ACTC actin, alpha, cardiac muscle 1 70 -3.56 CHRDL1 chordin-like 1 91851 -3.56 MAB21L1 mab-21-like 1 (C. elegans) 4081 -3.56 LOC143381 hypothetical protein LOC143381 143381 -3.58 CDCA7 cell division cycle associated 7 /// cell 83879 -3.58 division cycle associated 7 C14orf159 chromosome 14 open reading frame 159 80017 -3.58 LOH11CR2 loss of heterozygosity, 11, chromosomal 4013 -3.58 A region 2, gene A PGF placental growth factor, vascular 5228 -3.61 endothelial growth factor-related protein GPRC5B G protein-coupled receptor, family C, 51704 -3.61 group 5, member B KIAA1644 KIAA1644 protein 85352 -3.61 CX3CR1 chemokine (C-X3-C motif) receptor 1 1524 -3.63 HS6ST2 heparan sulfate 6-O-sulfotransferase 2 90161 -3.63 SPRY2 sprouty homolog 2 (Drosophila) 10253 -3.63 CKM creatine kinase, muscle 1158 -3.63 VAMP5 vesicle-associated membrane protein 5 10791 -3.66 (myobrevin) TNFRSF19 tumor necrosis factor receptor superfamily, 55504 -3.66 member 19 ANGPT1 angiopoietin 1 284 -3.68 PLAC9 placenta-specific 9 219348 -3.68 PAK3 p21 (CDKN1A)-activated kinase 3 5063 -3.68 TMEM61 transmembrane protein 61 199964 -3.68 TCEAL2 transcription elongation factor A (SII)-like 140597 -3.68 2 MYOCD myocardin 93649 -3.71 FREM2 FRAS1 related extracellular matrix protein 341640 -3.71 Page | 282 Appendix I 2 POU2AF1 POU domain, class 2, associating factor 1 5450 -3.73 ARMCX1 armadillo repeat containing, X-linked 1 51309 -3.73 SLC22A17 solute carrier family 22 (organic cation 51310 -3.76 transporter), member 17 SNCAIP synuclein, alpha interacting protein 9627 -3.76 (synphilin) GPR161 G protein-coupled receptor 161 23432 -3.78 EYA1 eyes absent homolog 1 (Drosophila) 2138 -3.81 COCH coagulation factor C homolog, cochlin 1690 -3.81 (Limulus polyphemus) PGM5 phosphoglucomutase 5 5239 -3.84 B3GALT2 UDP-Gal:betaGlcNAc beta 1,3- 8707 -3.84 galactosyltransferase, polypeptide 2 TCF21 transcription factor 21 6943 -3.84 P2RY5 purinergic receptor P2Y, G-protein 10161 -3.86 coupled, 5 EDG3 endothelial differentiation, sphingolipid G- 1903 -3.86 protein-coupled receptor, 3 CPLX3 lectin, mannose-binding, 1 like /// 594855 -3.86 complexin 3 SFRP1 secreted frizzled-related protein 1 6422 -3.89 MCC mutated in colorectal cancers 4163 -3.89 RSPO3 R-spondin 3 homolog (Xenopus laevis) 84870 -3.89 TRO trophinin /// trophinin 7216 -3.89 FLJ25076 similar to CG4502-PA 134111 -3.89 VIT vitrin 5212 -3.92 PHLDA1 pleckstrin homology-like domain, family 22822 -3.92 A, member 1 FHL2 four and a half LIM domains 2 2274 -3.92 ATP1B1 ATPase, Na+/K+ transporting, beta 1 481 -3.94 polypeptide FST follistatin 10468 -3.94 MT1M metallothionein 1M 4499 -3.94 PTGER3 prostaglandin E receptor 3 (subtype EP3) 5733 -3.94 PLAC8 placenta-specific 8 51316 -3.94 FHOD3 formin homology 2 domain containing 3 80206 -3.94 RARRES1 retinoic acid receptor responder (tazarotene 5918 -3.94 induced) 1 STAC SH3 and cysteine rich domain 6769 -3.97 ZNF655 zinc finger protein 655 79027 -4.00 GALNAC4S B cell RAG associated protein 51363 -4.03 -6ST TMEM35 transmembrane protein 35 59353 -4.03 EPHB1 EPH receptor B1 2047 -4.03 FLJ20920 hypothetical protein FLJ20920 80221 -4.06 CFB complement factor B 629 -4.06 CCDC85A coiled-coil domain containing 85A 114800 -4.06 VSNL1 visinin-like 1 7447 -4.06 ACOX2 acyl-Coenzyme A oxidase 2, branched 8309 -4.08 Page | 283 Appendix I chain ETV5 ets variant gene 5 (ets-related molecule) 2119 -4.08 MCF2 MCF.2 cell line derived transforming 4168 -4.08 sequence NDP Norrie disease (pseudoglioma) 4693 -4.14 TPM2 tropomyosin 2 (beta) 7169 -4.17 KCNAB1 potassium voltage-gated channel, shaker- 7881 -4.17 related subfamily, beta member 1 SNAI2 snail homolog 2 (Drosophila) 6591 -4.23 RARRES2 retinoic acid receptor responder (tazarotene 5919 -4.23 induced) 2 TAGLN transgelin 6876 -4.29 MEIS2 Meis1, myeloid ecotropic viral integration 4212 -4.29 site 1 homolog 2 (mouse) GPX3 glutathione peroxidase 3 (plasma) 2878 -4.32 S100A16 S100 calcium binding protein A16 140576 -4.35 PENK proenkephalin 5179 -4.35 SLITRK6 SLIT and NTRK-like family, member 6 84189 -4.38 COL27A1 collagen, type XXVII, alpha 1 85301 -4.38 GJA1 gap junction protein, alpha 1, 43kDa 2697 -4.41 (connexin 43) CHST2 carbohydrate (N-acetylglucosamine-6-O) 9435 -4.41 sulfotransferase 2 LOC390940 similar to R28379_1 390940 -4.44 LOC255480 hypothetical protein LOC255480 255480 -4.44 LRP4 low density lipoprotein receptor-related 4038 -4.44 protein 4 SERPINF1 serpin peptidase inhibitor, clade F (alpha-2 5176 -4.44 antiplasmin, pigment epithelium derived factor), member 1 IHPK3 inositol hexaphosphate kinase 3 117283 -4.44 CSRP2 cysteine and glycine-rich protein 2 1466 -4.50 EMILIN3 elastin microfibril interfacer 3 90187 -4.53 ZNF662 zinc finger protein 662 389114 -4.53 LRRN3 leucine rich repeat neuronal 3 54674 -4.56 DEPDC7 DEP domain containing 7 91614 -4.56 ASPA aspartoacylase (Canavan disease) 443 -4.59 RP11- hypothetical protein FLJ14503 256714 -4.59 393H10.2 FAIM2 Fas apoptotic inhibitory molecule 2 23017 -4.63 SLC16A5 solute carrier family 16, member 5 9121 -4.63 (monocarboxylic acid transporter 6) ALDH1A2 aldehyde dehydrogenase 1 family, member 8854 -4.63 A2 LCN2 lipocalin 2 (oncogene 24p3) 3934 -4.63 JPH2 junctophilin 2 57158 -4.66 CLU clusterin 1191 -4.66 ZNF204 zinc finger protein 204 7754 -4.66 COL13A1 collagen, type XIII, alpha 1 1305 -4.66 KCNK3 potassium channel, subfamily K, member 3 3777 -4.69 Page | 284 Appendix I FLJ10781 hypothetical protein FLJ10781 55228 -4.69 PLCL1 phospholipase C-like 1 5334 -4.76 SCRG1 scrapie responsive protein 1 11341 -4.76 FGFR2 fibroblast growth factor receptor 2 2263 -4.79 (bacteria-expressed kinase, keratinocyte growth factor receptor, craniofacial dysostosis 1, Crouzon syndrome, Pfeiffer syndrome, Jackson-Weiss syndrome) KCNJ8 potassium inwardly-rectifying channel, 3764 -4.79 subfamily J, member 8 ANGPTL1 angiopoietin-like 1 9068 -4.79 ATP1A2 ATPase, Na+/K+ transporting, alpha 2 (+) 477 -4.82 polypeptide LEPREL1 leprecan-like 1 55214 -4.89 DEFB1 defensin, beta 1 1672 -4.89 ST6GALNA ST6 (alpha-N-acetyl-neuraminyl-2,3-beta- 10610 -4.89 C2 galactosyl-1,3)-N-acetylgalactosaminide alpha-2,6-sialyltransferase 2 GABRP gamma-aminobutyric acid (GABA) A 2568 -4.89 receptor, pi KCTD14 potassium channel tetramerisation domain 65987 -4.92 containing 14 MYL9 myosin, light polypeptide 9, regulatory 10398 -4.92 EFS embryonal Fyn-associated substrate 10278 -4.96 EDG8 endothelial differentiation, sphingolipid G- 53637 -4.96 protein-coupled receptor, 8 D4S234E DNA segment on chromosome 4 (unique) 27065 -4.96 234 expressed sequence CXCL1 chemokine (C-X-C motif) ligand 1 2919 -4.96 (melanoma growth stimulating activity, alpha) FAM19A1 family with sequence similarity 19 407738 -4.96 (chemokine (C-C motif)-like), member A1 KRT17 keratin 17 3872 -4.99 CXorf57 chromosome X open reading frame 57 55086 -4.99 PIGR polymeric immunoglobulin receptor 5284 -5.03 MAGED4 melanoma antigen family D, 4 /// 81557 -5.06 melanoma antigen family D, 4 /// similar to melanoma antigen family D, 4 isoform 1 /// similar to melanoma antigen family D, 4 isoform 1 TNC tenascin C (hexabrachion) 3371 -5.10 IGSF1 immunoglobulin superfamily, member 1 3547 -5.10 ROR2 receptor tyrosine kinase-like orphan 4920 -5.10 receptor 2 CP ceruloplasmin (ferroxidase) 1356 -5.10 TGFB3 transforming growth factor, beta 3 7043 -5.17 POPDC2 popeye domain containing 2 64091 -5.21 PLN phospholamban 5350 -5.21 RPE65 retinal pigment epithelium-specific protein 6121 -5.28 Page | 285 Appendix I 65kDa ELMOD1 ELMO/CED-12 domain containing 1 55531 -5.31 ATP8B4 ATPase, Class I, type 8B, member 4 79895 -5.35 COL4A6 collagen, type IV, alpha 6 1288 -5.35 PAGE4 P antigen family, member 4 (prostate 9506 -5.35 associated) NTN4 netrin 4 59277 -5.39 C9orf125 chromosome 9 open reading frame 125 /// 84302 -5.50 chromosome 9 open reading frame 125 CITED1 Cbp/p300-interacting transactivator, with 4435 -5.58 Glu/Asp-rich carboxy-terminal domain, 1 KCNMB1 potassium large conductance calcium- 3779 -5.58 activated channel, subfamily M, beta member 1 LOC284837 hypothetical protein LOC284837 284837 -5.66 CHI3L2 chitinase 3-like 2 /// chitinase 3-like 2 1117 -5.70 PTCHD1 patched domain containing 1 139411 -5.70 FLRT3 fibronectin leucine rich transmembrane 23767 -5.74 protein 3 TRIM29 tripartite motif-containing 29 23650 -5.78 PPARGC1A peroxisome proliferative activated receptor, 10891 -5.86 gamma, coactivator 1, alpha IQCA IQ motif containing with AAA domain 79781 -5.86 DPYS dihydropyrimidinase 1807 -5.94 SMOC1 SPARC related modular calcium binding 1 64093 -5.98 MGC13057 hypothetical protein MGC13057 84281 -6.11 ANKRD38 ankyrin repeat domain 38 163782 -6.15 NELL2 NEL-like 2 (chicken) /// NEL-like 2 4753 -6.32 (chicken) ABCC3 ATP-binding cassette, sub-family C 8714 -6.36 (CFTR/MRP), member 3 SMARCD3 SWI/SNF related, matrix associated, actin 6604 -6.45 dependent regulator of chromatin, subfamily d, member 3 CXCL6 chemokine (C-X-C motif) ligand 6 6372 -6.50 (granulocyte chemotactic protein 2) SERPINB5 serpin peptidase inhibitor, clade B 5268 -6.59 (ovalbumin), member 5 CPXM carboxypeptidase X (M14 family), member 56265 -6.59 1 SP8 Sp8 transcription factor 221833 -6.59 CFC1 cripto, FRL-1, cryptic family 1 /// similar to 55997 -6.68 cryptic WNK4 WNK lysine deficient protein kinase 4 65266 -6.92 DMKN dermokine 93099 -7.06 FBLN1 fibulin 1 2192 -7.16 ORM2 orosomucoid 1 /// orosomucoid 2 5005 -7.21 LSAMP limbic system-associated membrane protein 4045 -7.26 CPA6 carboxypeptidase A6 57094 -7.46 CD177 CD177 molecule 57126 -7.73 Page | 286 Appendix I COL9A1 collagen, type IX, alpha 1 1297 -7.78 SV2B synaptic vesicle glycoprotein 2B 9899 -7.84 SRD5A2 steroid-5-alpha-reductase, alpha 6716 -7.84 polypeptide 2 (3-oxo-5 alpha-steroid delta 4-dehydrogenase alpha 2) SCGB1A1 secretoglobin, family 1A, member 1 7356 -7.89 (uteroglobin) ITGB6 clone LNG06541, highly similar to IR200 3694 -8.11 RNF165 ring finger protein 165 494470 -8.22 GATA3 GATA binding protein 3 2625 -8.28 GPR87 G protein-coupled receptor 87 53836 -8.34 KRT222P keratin 222 pseudogene 125113 -8.82 CDH3 cadherin 3, type 1, P-cadherin (placental) 1001 -9.06 SLITRK5 SLIT and NTRK-like family, member 5 26050 -9.25 BCL11A B-cell CLL/lymphoma 11A (zinc finger 53335 -9.45 protein) FZD10 frizzled homolog 10 (Drosophila) 11211 -9.58 WFDC2 WAP four-disulfide core domain 2 10406 -9.78 LIX1 Lix1 homolog (mouse) 167410 -9.85 VILL villin-like 50853 -9.92 LAMB3 laminin, beta 3 3914 -9.99 KCNJ15 potassium inwardly-rectifying channel, 3772 -10.13 subfamily J, member 15 CCK cholecystokinin 885 -10.78 CAPN6 calpain 6 827 -10.93 H19 H19, imprinted maternally expressed 283120 -11.24 untranslated mRNA CHST9 carbohydrate (N-acetylgalactosamine 4-0) 83539 -11.39 sulfotransferase 9 NEFH neurofilament, heavy polypeptide 200kDa 4744 -11.88 BEX1 brain expressed, X-linked 1 55859 -12.04 LOC572558 hypothetical locus LOC572558 572558 -12.13 MMP7 matrix metallopeptidase 7 (matrilysin, 4316 -12.13 uterine) KRT23 keratin 23 (histone deacetylase inducible) 25984 -12.38 DUOX1 dual oxidase 1 53905 -13.18 RNF128 ring finger protein 128 79589 -13.45 KRT5 keratin 5 (epidermolysis bullosa simplex, 3852 -14.72 Dowling-Meara/Kobner/Weber-Cockayne types) SLC2A5 solute carrier family 2 (facilitated 6518 -15.24 glucose/fructose transporter), member 5 RP13- KIAA1210 protein 57481 -17.75 347D8.3 LOC92196 similar to death-associated protein 92196 -19.16 NEFL neurofilament, light polypeptide 68kDa 4747 -19.16 SOX2 SRY (sex determining region Y)-box 2 6657 -19.56 SCUBE2 signal peptide, CUB domain, EGF-like 2 57758 -20.25 KRT14 keratin 14 (epidermolysis bullosa simplex, 3861 -21.26 Dowling-Meara, Koebner) Page | 287 Appendix I CYP4B1 cytochrome P450, family 4, subfamily B, 1580 -22.47 polypeptide 1 CXCL13 chemokine (C-X-C motif) ligand 13 (B-cell 10563 -25.11 chemoattractant) EDN3 endothelin 3 1908 -25.63 CYP3A5 cytochrome P450, family 3, subfamily A, 1577 -25.81 polypeptide 5 ABP1 amiloride binding protein 1 (amine oxidase 26 -27.86 (copper-containing)) KRT15 keratin 15 3866 -32.90 TP73L tumor protein p73-like 8626 -33.36 SLC14A1 solute carrier family 14 (urea transporter), 6563 -34.30 member 1 (Kidd blood group) WIF1 WNT inhibitory factor 1 11197 -36.76 LTF lactotransferrin /// similar to 4057 -135.30 lactotransferrin Page | 288 Appendix II APPENDIX II Section 1A Association of non-sulphated chondroitin staining intensity with all clinicopathological parameters. Stained tissue Clinicopathological Total Weighted Weighted P- component parameter Average Average value Intensity <2 Intensity ≥2 Adenocarcinoma Pathological T staging T2a to T2c 40 28 12 0.225 100.0 70.0% 30.0% T3a to T3c 33 18 15 100.0 54.5% 45.5% Total 73 Extraprostatic extension Yes 33 18 15 0.225 100.0 54.5% 45.5% No 40 28 12 100.0 70.0% 30.0% Total 73 Seminal vesicle involvement Involved 7 2 5 0.090 100.0 28.6% 71.4% Uninvolved 69 47 22 100.0 68.1% 31.9% Total 76 Pre-operative PSA levels ≤8ng/ml 22 15 7 1.000 100.0 68.2% 31.8% >8ng/ml 20 14 6 100.0 70.0% 30.0% Total 42 Lymphovascular invasion Present 16 10 6 0.372 100.0 62.5% 37.5% Absent 120 89 31 100.0 74.2% 25.8% Total 136 Page | 289 Appendix II Perineural invasion Seen 88 58 30 0.125 100.0 65.9% 34.1% Not seen 53 42 11 100.0 79.2% 20.8% Total 141 Circumferential resection margin Involved 35 23 12 1.000 100.0 65.7% 34.3% Uninvolved 41 27 14 100.0 65.9% 34.1% Total 76 Apex margin Involved 34 20 14 0.470 100.0 58.8% 41.2% Uninvolved 42 29 13 100.0 69.0% 31.0% Total 76 Proximal margin Involved 9 3 6 0.059 100.0 33.3% 66.7% Uninvolved 68 47 21 100.0 69.1% 30.9% Total 77 Clinical stage T1a to T1c 36 26 10 0.676 100.0 72.2% 27.8% T3a to T3b 8 5 3 100.0 62.5% 37.5% Total 44 Lobe occurence Both Lobes 64 38 26 0.013 100.0 59.4% 40.6% Singles Lobe 11 11 0 100.0 100.0% .0% Total 75 Maximum tumour dimension <2.5cm 39 29 10 0.093 100.0 74.4% 25.6% ≥2.5cm 37 20 17 100.0 54.1% 45.9% Total 76 Page | 290 Appendix II Tumour Volume <4cm3 31 23 8 1.000 100.0 74.2% 25.8% ≥4 cm3 29 22 7 100.0 75.9% 24.1% Total 60 Peritumoural Pathological T staging Stroma T2a to T2c 39 34 5 0.049 100.0 87.2% 12.8% T3a to T3c 33 22 11 100.0 66.7% 33.3% Total 72 Extraprostatic extension Yes 33 22 11 0.049 100.0 66.7% 33.3% No 39 34 5 100.0 87.2% 12.8% Total 72 Seminal vesicle involvement Involved 7 3 4 0.034 100.0 42.9% 57.1% Uninvolved 68 56 12 100.0 82.4% 17.6% Total 75 Pre-operative PSA levels ≤8ng/ml 21 21 0 0.021 100.0 100.0% .0% >8ng/ml 20 15 5 100.0 75.0% 25.0% Total 41 Lymphovascular invasion Present 16 12 4 0.776 100.0 75.0% 25.0% Absent 120 83 37 100.0 69.2% 30.8% Total 136 Page | 291 Appendix II Perineural invasion Seen 88 58 30 0.125 100.0 65.9% 34.1% Not seen 53 42 11 100.0 79.2% 20.8% Total 133 Circumferential resection margin Involved 35 25 10 0.171 100.0 71.4% 28.6% Uninvolved 40 34 6 100.0 85.0% 15.0% Total 75 Apex margin Involved 34 24 10 0.160 100.0 70.6% 29.4% Uninvolved 41 35 6 100.0 85.4% 14.6% Total 75 Proximal margin Involved 9 7 2 1.000 100.0 77.8% 22.2% Uninvolved 67 53 14 100.0 79.1% 20.9% Total 76 Clinical stage T1a to T1c 35 30 5 0.565 100.0 85.7% 14.3% T3a to T3b 8 8 0 100.0 100.0% .0% Total 43 Lobe occurence Both Lobes 63 49 14 1.000 100.0 77.8% 22.2% Singles Lobe 11 9 2 100.0 81.8% 18.2% Total 74 Maximum tumour dimension <2.5cm 38 32 6 0.271 100.0 84.2% 15.8% ≥2.5cm 37 27 10 100.0 73.0% 27.0% Total 75 Page | 292 Appendix II Tumour Volume <4cm3 31 27 4 0.204 100.0 87.1% 12.9% ≥4 cm3 29 21 8 100.0 72.4% 27.6% Total 60 Page | 293 Appendix II Section 1B Association of percentage of cells positively stained for non- sulphated chondroitin staining with all clinicopathological parameters. Stained tissue Clinicopathological Total Percentage Percentage P- component parameter of cells of cells value positively positively stained stained ≤50% >50% Adenocarcinoma Pathological T staging T2a to T2c 40 30 10 0.050 100.0% 75.0% 25.0% T3a to T3c 33 17 16 100.0% 51.5% 48.5% Total 73 Extraprostatic extension Yes 33 17 16 0.050 100.0% 51.5% 48.5% No 40 30 10 100.0% 75.0% 25.0% Total 73 Seminal vesicle involvement Involved 7 3 4 0.238 100.0% 42.9% 57.1% Uninvolved 69 46 23 100.0% 66.7% 33.3% Total 139 Pre-operative PSA levels ≤8ng/ml 22 10 12 0.758 100.0% 45.5% 54.5% >8ng/ml 20 11 9 100.0% 55.0% 45.0% Total 42 Lymphovascular invasion Present 16 10 6 0.365 100.0% 62.5% 37.5% Absent 120 90 30 100.0% 75.0% 25.0% Total 136 Page | 294 Appendix II Perineural invasion Seen 88 64 24 1.000 100.0% 72.7% 27.3% Not seen 53 38 15 100.0% 71.7% 28.3% Total 133 Circumferential resection margin Involved 35 23 12 1.000 100.0% 65.7% 34.3% Uninvolved 41 27 14 100.0% 65.9% 34.1% Total 76 Apex margin Involved 34 22 12 1.000 100.0% 64.7% 35.3% Uninvolved 42 27 15 100.0% 64.3% 35.7% Total 76 Proximal margin Involved 9 4 5 0.264 100.0% 44.4% 55.6% Uninvolved 68 46 22 100.0% 67.6% 32.4% Total 77 Clinical stage T1a to T1c 36 18 18 0.701 100.0% 50.0% 50.0% T3a to T3b 8 5 3 100.0% 62.5% 37.5% Total 44 Lobe occurence Both Lobes 64 41 23 0738 100.0% 64.1% 35.9% Singles Lobe 11 8 3 100.0% 72.7% 27.3% Total 75 Maximum tumour dimension <2.5cm 39 27 12 0.473 100.0% 69.2% 30.8% ≥2.5cm 37 22 15 100.0% 59.5% 40.5% Total 76 Page | 295 Appendix II Tumour Volume <4cm3 31 23 8 0.408 100.0% 74.2% 25.8% ≥4 cm3 29 18 11 100.0% 62.1% 37.9% Total 60 Peritumoural Pathological T staging Stroma T2a to T2c 39 31 8 0.188 100.0% 79.5% 20.5% T3a to T3c 33 21 12 100.0% 63.6% 36.4% Total 72 Extraprostatic extension Yes 33 21 12 0.188 100.0% 63.6% 36.4% No 39 31 8 100.0% 79.5% 20.5% Total 72 Seminal vesicle involvement Involved 7 3 4 0.077 100.0% 42.9% 57.1% Uninvolved 68 52 16 100.0% 76.5% 23.5% Total 75 Pre-operative PSA levels ≤8ng/ml 21 20 1 0.184 100.0% 95.2% 4.8% >8ng/ml 20 16 4 100.0% 80.0% 20.0% Total 41 Lymphovascular invasion Present 16 9 7 0.582 100.0% 56.3% 43.8% Absent 120 78 42 100.0% 65.0% 35.0% Total 136 Perineural invasion Seen 88 57 31 1.000 100.0% 64.8% 35.2% Not seen 53 35 18 100.0% 66.0% 34.0% Total 141 Page | 296 Appendix II Circumferential Involved 35 26 9 1.000 100.0% 74.3% 25.7% Uninvolved 40 30 10 100.0% 75.0% 25.0% Total 75 Apex margin Involved 34 26 8 0.611 100.0% 76.5% 23.5% Uninvolved 41 29 12 100.0% 70.7% 29.3% Total 75 Proximal margin Involved 9 7 2 1.000 100.0% 77.8% 22.2% Uninvolved 67 49 18 100.0% 73.1% 26.9% Total 76 Clinical stage T1a to T1c 35 31 4 0.308 100.0% 88.6% 11.4% T3a to T3b 8 6 2 100.0% 75.0% 25.0% Total 43 Lobe occurence Both Lobes 63 46 17 1.000 100.0% 73.0% 27.0% Singles Lobe 11 8 3 100.0% 72.7% 27.3% Total 74 Maximum tumour dimension <2.5cm 38 27 11 0.795 100.0% 71.1% 28.9% ≥2.5cm 37 28 9 100.0% 75.7% 24.3% Total 75 Tumour Volume <4cm3 31 24 7 0.769 100.0% 77.4% 22.6% ≥4 cm3 29 21 8 100.0% 72.4% 27.6% Total 60 Page | 297 Appendix II Section 2A Association of chondroitin-4-sulphate staining intensity with all clinicopathological parameters. Stained tissue Clinicopathological Total Weighted Weighted P- component parameter Average Average value Intensity Intensity <2 ≥2 Adenocarcinoma Pathological T staging T2a to T2c 38 34 4 0.041 100.0% 89.5% 10.5% T3a to T3c 35 24 11 100.0% 68.6% 31.4% Total 73 Extraprostatic extension Yes 35 24 11 0.041 100.0% (68.6%) (31.4%) No 38 34 4 100.0% (89.5%) (10.5%) Total 73 Seminal vesicle involvement Involved 8 5 3 0.188 100.0% 62.5% 37.5% Uninvolved 68 56 12 100.0% 82.4% 17.6% Total 76 Pre-operative PSA levels ≤8ng/ml 21 16 5 0.697 100.0% 76.2% 23.8% >8ng/ml 21 18 3 100.0% 85.7% 14.3% Total 42 Lymphovascular invasion Present 17 15 2 0.681 100.0% 88.2% 11.8% Absent 122 110 12 100.0% 90.2% 9.8% Total 139 Page | 298 Appendix II Perineural invasion Seen 90 75 15 0.001 100.0% 83.3% 16.7% Not seen 54 54 0 100.0% 100.0% .0% Total 144 Circumferential resection margin Involved 35 27 8 0.574 100.0% 77.1% 22.9% Uninvolved 41 34 7 100.0% 82.9% 17.1% Total 76 Apex margin Involved 35 29 6 1.000 100.0% 82.9% 17.1% Uninvolved 41 33 8 100.0% 80.5% 19.5% Total 76 Proximal margin Involved 9 7 2 1.000 100.0% 77.8% 22.2% Uninvolved 68 55 13 100.0% 80.9% 19.1% Total 77 Clinical stage T1a to T1c 36 32 4 0.015 100.0% 88.9% 11.1% T3a to T3b 7 3 4 100.0% 42.9% 57.1% Total 43 Lobe occurence Both Lobes 65 52 13 0.677 100.0% 80.0% 20.0% Singles Lobe 10 9 1 100.0% 90.0% 10.0% Total 75 Maximum tumour dimension <2.5cm 38 32 6 0.565 100.0% 84.2% 15.8% ≥2.5cm 38 29 9 100.0% 76.3% 23.7% Total 76 Page | 299 Appendix II Tumour Volume <4cm3 29 23 6 0.748 100.0% 79.3% 20.7% ≥4 cm3 30 25 5 100.0% 83.3% 16.7% Total 59 Peritumoural Pathological T staging Stroma T2a to T2c 38 31 7 0.574 100.0% 81.6% 18.4% T3a to T3c 35 26 9 100.0% 74.3% 25.7% Total 73 Extraprostatic extension Yes 35 26 9 0.574 100.0% 74.3% 25.7% No 38 31 7 100.0% 81.6% 18.4% Total 73 Seminal vesicle involvement Involved 8 5 3 0.354 100.0% 62.5% 37.5% Uninvolved 68 55 13 100.0% 80.9% 19.1% Total 76 Pre-operative PSA levels ≤8ng/ml 21 18 3 1.000 100.0% 85.7% 14.3% >8ng/ml 21 18 3 100.0% 85.7% 14.3% Total 42 Lymphovascular invasion Present 17 10 7 0.402 100.0% 58.8% 41.2% Absent 122 86 36 100.0% 70.5% 29.5% Total 139 Perineural invasion Seen 90 62 28 1.000 100.0% 68.9% 31.1% Not seen 54 38 16 100.0% 70.4% 29.6% Total 144 Page | 300 Appendix II Circumferential resection margin Involved 35 26 9 0.407 100.0% 74.3% 25.7% Uninvolved 41 34 7 100.0% 82.9% 17.1% Total 76 Apex margin Involved 35 25 10 0.166 100.0% 71.4% 28.6% Uninvolved 41 35 6 100.0% 85.4% 14.6% Total 76 Proximal margin Involved 9 8 1 0.675 100.0% 88.9% 11.1% Uninvolved 68 53 15 100.0% 77.9% 22.1% Total 77 Clinical stage T1a to T1c 36 30 6 0.567 100.0% 83.3% 16.7% T3a to T3b 7 7 0 100.0% 100.0% .0% Total 43 Lobe occurence Both Lobes 65 50 15 0.679 100.0% 76.9% 23.1% Singles Lobe 10 9 1 100.0% 90.0% 10.0% Total 75 Maximum tumour dimension <2.5cm 38 33 5 0.158 100.0% 86.8% 13.2% ≥2.5cm 38 27 11 100.0% 71.1% 28.9% Total 76 Tumour Volume <4cm3 29 25 4 0.506 100.0% 86.2% 13.8% ≥4 cm3 30 23 7 100.0% 76.7% 23.3% Total 59 Page | 301 Appendix II Section 2B Association of percentage of cells positively stained for chondroitin-4-sulphate with all clinicopathological parameters. Stained tissue Clinicopathological Total Percentage Percentage P- component parameter of cells of cells value positively positively stained≤50 stained>50 % % Adenocarcinoma Pathological T staging T2a to T2c 38 30 8 0.076 100.0% 78.9% 21.1% T3a to T3c 35 20 15 100.0% 57.1% 42.9% Total 73 Extraprostatic extension Yes 35 20 15 0.076 100.0% 57.1% 42.9% No 38 30 8 100.0% 78.9% 21.1% Total 73 Seminal vesicle involvement Involved 8 3 5 0.050 100.0% 37.5% 62.5% Uninvolved 68 50 18 100.0% 73.5% 26.5% Total 76 Pre-operative PSA levels ≤8ng/ml 21 13 8 0.505 100.0% 61.9% 38.1% >8ng/ml 21 16 5 100.0% 76.2% 23.8% Total 42 Lymphovascular invasion Present 17 11 6 0.024 100.0% 64.7% 35.3% Absent 122 107 15 100.0% 87.7% 12.3% Total 139 Page | 302 Appendix II Perineural invasion Seen 90 71 19 0.035 100.0% 78.9% 21.1% Not seen 54 50 4 100.0% 92.6% 7.4% Total 144 Circumferential resection margin Involved 35 24 11 0.800 100.0% 68.6% 31.4% Uninvolved 41 30 11 100.0% 73.2% 26.8% Total 76 Apex margin Involved 35 25 10 0.807 100.0% 71.4% 28.6% Uninvolved 41 28 13 100.0% 68.3% 31.7% Total 76 Proximal margin Involved 9 6 3 1.000 100.0% 66.7% 33.3% Uninvolved 68 48 20 100.0% 70.6% 29.4% Total 77 Clinical stage T1a to T1c 36 27 9 0.172 100.0% 75.0% 25.0% T3a to T3b 7 3 4 100.0% 42.9% 57.1% Total 43 Lobe occurence Both Lobes 65 47 18 0.467 100.0% 72.3% 27.7% Singles Lobe 10 6 4 100.0% 60.0% 40.0% Total 75 Maximum tumour dimension <2.5cm 38 26 12 1.000 100.0% 68.4% 31.6% ≥2.5cm 38 27 11 100.0% 71.1% 28.9% Total 76 Page | 303 Appendix II Tumour Volume <4cm3 29 20 9 0.567 100.0% 69.0% 31.0% ≥4 cm3 30 23 7 100.0% 76.7% 23.3% Total 59 Peritumoural Pathological T staging Stroma T2a to T2c 38 12 26 0.804 100.0% 31.6% 68.4% T3a to T3c 35 10 25 100.0% 28.6% 71.4% Total 73 Extraprostatic extension Yes 35 10 25 0.804 100.0% 28.6% 71.4% No 38 12 26 100.0% 31.6% 68.4% Total 73 Seminal vesicle involvement Involved 8 1 7 0.423 100.0% 12.5% 87.5% Uninvolved 68 23 45 100.0% 33.8% 66.2% Total 76 Pre-operative PSA levels ≤8ng/ml 21 9 12 1.000 100.0% 42.9% 57.1% >8ng/ml 21 8 13 100.0% 38.1% 61.9% Total 42 Lymphovascular invasion Present 17 6 11 0.128 100.0% 35.3% 64.7% Absent 122 68 54 100.0% 55.7% 44.3% Total 139 Perineural invasion Seen 90 40 50 0.025 100.0% 44.4% 55.6% Not seen 54 35 19 100.0% 64.8% 35.2% Total 144 Page | 304 Appendix II Circumferential resection margin Involved 35 9 26 0.234 100.0% 25.7% 74.3% Uninvolved 41 16 25 100.0% 39.0% 61.0% Total 76 Apex margin Involved 35 12 23 1.000 100.0% 34.3% 65.7% Uninvolved 41 13 28 100.0% 31.7% 68.3% Total 76 Proximal margin Involved 9 2 7 0.710 100.0% 22.2% 77.8% Uninvolved 68 23 45 100.0% 33.8% 66.2% Total 77 Clinical stage T1a to T1c 36 14 22 1.000 100.0% 38.9% 61.1% T3a to T3b 7 3 4 100.0% 42.9% 57.1% Total 43 Lobe occurence Both Lobes 65 19 46 0.484 100.0% 29.2% 70.8% Singles Lobe 10 4 6 100.0% 40.0% 60.0% Total 75 Maximum tumour dimension <2.5cm 38 16 22 0.083 100.0% 42.1% 57.9% ≥2.5cm 38 8 30 100.0% 21.1% 78.9% Total 76 Tumour Volume <4cm3 29 13 16 0.288 100.0% 44.8% 55.2% ≥4 cm3 30 9 21 100.0% 30.0% 70.0% Total 59 Page | 305 Appendix II Section 3A Association of chondroitin-6-sulphate staining intensity with all clinicopathological parameters Stained tissue Clinicopathological Total Weighted Weighted P- component parameter Average Average value Intensity Intensity <2 ≥2 Adenocarcinoma Pathological T staging T2a to T2c 36 34 2 1.000 100.0% 94.4% 5.6% T3a to T3c 34 33 1 100.0% 97.1% 2.9% Total 70 Extraprostatic extension Yes 34 33 1 1.000 100.0% 97.1% 2.9% No 36 34 2 100.0% 94.4% 5.6% Total 70 Seminal vesicle involvement Involved 7 7 0 1.000 100.0% 100.0% .0% Uninvolved 66 63 3 100.0% 95.5% 4.5% Total 73 Pre-operative PSA levels ≤8ng/ml 21 20 1 1.000 100.0% 95.2% 4.8% >8ng/ml 20 19 1 100.0% 95.0% 5.0% Total 41 Lymphovascular invasion Present 16 16 0 0.211 100.0% 100.0% .0% Absent 106 92 14 100.0% 86.8% 13.2% Total 122 Page | 306 Appendix II Perineural invasion Seen 82 77 5 0.034 100.0% 93.9% 6.1% Not seen 45 36 9 100.0% 80.0% 20.0% Total 127 Circumferential resection margin Involved 32 30 2 0.581 100.0% 93.8% 6.3% Uninvolved 40 39 1 100.0% 97.5% 2.5% Total 72 Apex margin Involved 34 31 3 0.100 100.0% 91.2% 8.8% Uninvolved 38 38 0 100.0% 100.0% .0% Total 72 Proximal margin Involved 9 8 1 0.330 100.0% 88.9% 11.1% Uninvolved 64 62 2 100.0% 96.9% 3.1% Total 73 Clinical stage T1a to T1c 34 32 2 1.000 100.0% 94.1% 5.9% T3a to T3b 8 8 0 100.0% 100.0% .0% Total 42 Lobe occurence Both Lobes 61 59 2 0.370 100.0% 96.7% 3.3% Singles Lobe 10 9 1 100.0% 90.0% 10.0% Total 71 Maximum tumour dimension <2.5cm 34 32 2 0.599 100.0% 94.1% 5.9% ≥2.5cm 38 37 1 100.0% 97.4% 2.6% Total 72 Page | 307 Appendix II Tumour Volume <4cm3 27 26 1 0.482 100.0% 96.3% 3.7% ≥4 cm3 29 29 0 100.0% 100.0% .0% Total 56 Peritumoural Pathological T staging Stroma T2a to T2c 35 31 4 1.000 100.0% 88.6% 11.4% T3a to T3c 34 30 4 100.0% 88.2% 11.8% Total 69 Extraprostatic extension Yes 34 30 4 1.000 100.0% 88.2% 11.8% No 35 31 4 100.0% 88.6% 11.4% Total 69 Seminal vesicle involvement Involved 7 7 0 1.000 100.0% 100.0% .0% Uninvolved 65 57 8 100.0% 87.7% 12.3% Total 135 Pre-operative PSA levels ≤8ng/ml 20 19 1 0.605 100.0% 95.0% 5.0% >8ng/ml 20 17 3 100.0% 85.0% 15.0% Total 40 Lymphovascular invasion Present 16 11 5 0.590 100.0% 68.8% 31.3% Absent 105 63 42 100.0% 60.0% 40.0% Total 121 Perineural invasion Seen 82 58 24 0.007 100.0% 70.7% 29.3% Not seen 44 20 24 100.0% 45.5% 54.5% Total 126 Page | 308 Appendix II Circumferential resection margin Involved 32 30 2 0.281 100.0% 93.8% 6.3% Uninvolved 39 33 6 100.0% 84.6% 15.4% Total 71 Apex margin Involved 34 30 4 1.000 100.0% 88.2% 11.8% Uninvolved 37 33 4 100.0% 89.2% 10.8% Total 71 Proximal margin Involved 9 9 0 0.584 100.0% 100.0% .0% Uninvolved 63 55 8 100.0% 87.3% 12.7% Total 72 Clinical stage T1a to T1c 33 30 3 1.000 100.0% 90.9% 9.1% T3a to T3b 8 7 1 100.0% 87.5% 12.5% Total 41 Lobe occurence Both Lobes 60 53 7 1.000 100.0% 88.3% 11.7% Singles Lobe 10 9 1 100.0% 90.0% 10.0% Total 70 Maximum tumour dimension <2.5cm 33 28 5 0.459 100.0% 84.8% 15.2% ≥2.5cm 38 35 3 100.0% 92.1% 7.9% Total 72 Tumour Volume <4cm3 27 25 2 0.671 100.0% 92.6% 7.4% ≥4 cm3 29 25 4 100.0% 86.2% 13.8% Total 56 Page | 309 Appendix II Section 3B Association of percentage of cells with positive chondroitin-6- sulphate staining with all clinicopathological parameters. Stained tissue Clinicopathological Total Percentage Percentage P- component parameter of cells of cells value positively positively stained≤50 stained>50 % % Adenocarcinoma Pathological T staging T2a to T2c 36 11 25 0.794 100.0% 30.6% 69.4% T3a to T3c 34 9 25 100.0% 26.5% 73.5% Total 70 Extraprostatic extension Yes 34 9 25 0.794 100.0% 26.5% 73.5% No 36 11 25 100.0% 30.6% 69.4% Total 70 Seminal vesicle involvement Involved 7 3 4 0.671 100.0% 42.9% 57.1% Uninvolved 66 20 46 100.0% 30.3% 69.7% Total 73 Pre-operative PSA levels ≤8ng/ml 21 7 14 1.000 100.0% 33.3% 66.7% >8ng/ml 20 7 13 100.0% 35.0% 65.0% Total 41 Lymphovascular invasion Present 16 9 7 1.000 100.0% 56.3% 43.8% Absent 106 62 44 100.0% 58.5% 41.5% Total 116 Page | 310 Appendix II Perineural invasion Seen 82 42 40 0.134 100.0% 51.2% 48.8% Not seen 45 30 15 100.0% 66.7% 33.3% Total 127 Circumferential resection margin Involved 32 12 20 0.448 100.0% 37.5% 62.5% Uninvolved 40 11 29 100.0% 27.5% 72.5% Total 72 Apex margin Involved 34 7 27 0.124 100.0% 20.6% 79.4% Uninvolved 38 15 23 100.0% 39.5% 60.5% Total 72 Proximal margin Involved 9 3 6 1.000 100.0% 33.3% 66.7% Uninvolved 64 20 44 100.0% 31.3% 68.8% Total 73 Clinical stage T1a to T1c 34 12 22 0.697 100.0% 35.3% 64.7% T3a to T3b 8 2 6 100.0% 25.0% 75.0% Total 42 Lobe occurence Both Lobes 61 19 42 1.000 100.0% 31.1% 68.9% Singles Lobe 10 3 7 100.0% 30.0% 70.0% Total 71 Maximum tumour dimension <2.5cm 34 9 25 0.610 100.0% 26.5% 73.5% ≥2.5cm 38 13 25 100.0% 34.2% 65.8% Total 72 Page | 311 Appendix II Tumour Volume <4cm3 27 8 19 0.580 100.0% 29.6% 70.4% ≥4 cm3 29 11 18 100.0% 37.9% 62.1% Total 56 Peritumoural Pathological T staging Stroma T2a to T2c 35 22 13 0.004 100.0% 62.9% 37.1% T3a to T3c 34 9 25 100.0% 26.5% 73.5% Total 69 Extraprostatic extension Yes 34 9 25 0.004 100.0% 26.5% 73.5% No 35 22 13 100.0% 62.9% 37.1% Total 69 Seminal vesicle involvement Involved 7 2 5 0.442 100.0% 28.6% 71.4% Uninvolved 65 31 34 100.0% 47.7% 52.3% Total 72 Pre-operative PSA levels ≤8ng/ml 20 10 10 0.751 100.0% 50.0% 50.0% >8ng/ml 20 8 12 100.0% 40.0% 60.0% Total 40 Lymphovascular invasion Present 16 6 10 1.000 100.0% 37.5% 62.5% Absent 105 41 64 100.0% 39.0% 61.0% Total 121 Perineural invasion Seen 82 30 52 0.184 100.0% 36.6% 63.4% Not seen 44 22 22 100.0% 50.0% 50.0% Total 126 Page | 312 Appendix II Circumferential resection margin Involved 32 11 21 0.094 100.0% 34.4% 65.6% Uninvolved 39 22 17 100.0% 56.4% 43.6% Total 71 Apex margin Involved 34 14 20 0.635 100.0% 41.2% 58.8% Uninvolved 37 18 19 100.0% 48.6% 51.4% Total 71 Proximal margin Involved 9 2 7 0.166 100.0% 22.2% 77.8% Uninvolved 63 31 32 100.0% 49.2% 50.8% Total 72 Clinical stage T1a to T1c 33 17 16 0.059 100.0% 51.5% 48.5% T3a to T3b 8 1 7 100.0% 12.5% 87.5% Total 41 Lobe occurence Both Lobes 60 27 33 1.000 100.0% 45.0% 55.0% Singles Lobe 10 6 4 100.0% 60.0% 40.0% Total 70 Maximum tumour dimension <2.5cm 33 17 16 0.480 100.0% 51.5% 48.5% ≥2.5cm 38 16 22 100.0% 42.1% 57.9% Total 71 Tumour volume <4cm3 27 17 10 0.119 100.0% 63.0% 37.0% ≥4 cm3 29 12 17 100.0% 41.4% 58.6% Total 56 Page | 313 Appendix II Section 4A Association of chondroitin-2,6-sulphate staining intensity with all clinicopathological parameters. Stained tissue Clinicopathological Total Weighted Weighted P- component parameter Average Average value Intensity Intensity <2 ≥2 Adenocarcinoma Pathological T staging T2a to T2c 38 26 12 0.228 100.0% 68.4% 31.6% T3a to T3c 34 18 16 100.0% 52.9% 47.1% Total 72 Extraprostatic extension Yes 34 18 16 0.228 100.0% 52.9% 47.1% No 38 26 12 100.0% 68.4% 31.6% Total 72 Seminal vesicle involvement Involved 7 3 4 0.413 100.0% 42.9% 57.1% Uninvolved 68 44 24 100.0% 64.7% 35.3% Total 75 Pre-operative PSA levels ≤8ng/ml 22 14 8 0.760 100.0% 63.6% 36.4% >8ng/ml 21 12 9 100.0% 57.1% 42.9% Total 43 Lymphovascular invasion Present 18 14 4 0.757 100.0% 77.8% 22.2% Absent 118 95 23 100.0% 80.5% 19.5% Total Page | 314 Appendix II Perineural invasion Seen 89 65 24 0.017 100.0% 73.0% 27.0% Not seen 52 47 5 100.0% 90.4% 9.6% Total 141 Circumferential resection margin Involved 35 22 13 0.817 100.0% 62.9% 37.1% Uninvolved 40 24 16 100.0% 60.0% 40.0% Total 75 Apex margin Involved 35 20 15 0.635 100.0% 57.1% 42.9% Uninvolved 40 26 14 100.0% 65.0% 35.0% Total 75 Proximal margin Involved 9 5 4 0.725 100.0% 55.6% 44.4% Uninvolved 67 42 25 100.0% 62.7% 37.3% Total 76 Clinical stage T1a to T1c 37 25 12 0.226 100.0% 67.6% 32.4% T3a to T3b 8 3 5 100.0% 37.5% 62.5% Total 45 Lobe occurence Both Lobes 64 37 27 0.079 100.0% 57.8% 42.2% Singles Lobe 10 9 1 100.0% 90.0% 10.0% Total 74 Maximum tumour dimension <2.5cm 37 25 12 0.345 100.0% 67.6% 32.4% ≥2.5cm 38 21 17 100.0% 55.3% 44.7% Total 75 Page | 315 Appendix II Tumour Volume <4cm3 28 20 8 0.576 100.0% 71.4% 28.6% ≥4 cm3 29 18 11 100.0% 62.1% 37.9% Total 57 Peritumoural Pathological T staging Stroma T2a to T2c 40 16 24 1.000 100.0% 40.0% 60.0% T3a to T3c 34 14 20 100.0% 41.2% 58.8% Total 74 Extraprostatic extension Yes 36 16 20 0.817 100.0% 44.4% 55.6% No 40 16 24 100.0% 40.0% 60.0% Total 76 Seminal vesicle involvement Involved 7 0 7 0.018 100.0% 0% 100.0% Uninvolved 72 34 38 100.0% 47.2% 52.8% Total 79 Pre-operative PSA levels ≤8ng/ml 22 11 11 0.763 100.0% 50.0% 50.0% >8ng/ml 22 9 13 100.0% 40.9% 59.1% Total 44 Lymphovascular invasion Present 18 9 9 0.298 100.0% 50.0% 50.0% Absent 118 42 76 100.0% 35.6% 64.4% Total 136 Perineural invasion Seen 89 35 54 0.858 100.0% 39.3% 60.7% Not seen 52 19 33 100.0% 36.5% 63.5% Total 141 Page | 316 Appendix II Circumferential resection margin Involved 36 17 19 0.499 100.0% 47.2% 52.8% Uninvolved 44 17 27 100.0% 38.6% 61.4% Total 80 Apex margin Involved 37 15 22 0.822 100.0% 40.5% 59.5% Uninvolved 43 19 24 100.0% 44.2% 55.8% Total 80 Proximal margin Involved 10 4 6 1.000 100.0% 40.0% 60.0% Uninvolved 71 31 40 100.0% 43.7% 56.3% Total 81 Clinical stage T1a to T1c 38 18 20 0.710 100.0% 47.4% 52.6% T3a to T3b 8 3 5 100.0% 37.5% 62.5% Total 46 Lobe occurence Both Lobes 67 26 41 0.344 100.0% 38.8% 61.2% Singles Lobe 11 6 5 100.0% 54.5% 45.5% Total 78 Maximum tumour dimension <2.5cm 39 18 21 0.490 100.0% 46.2% 53.8% ≥2.5cm 38 14 24 100.0% 36.8% 63.2% Total 77 Tumour Volume <4cm3 30 16 14 0.299 100.0% 53.3% 46.7% ≥4 cm3 29 11 18 100.0% 37.9% 62.1% Total 59 Page | 317 Appendix II Section 4B Association of percentage of cells positively stained for chondroitin-2,6-sulphate with all clinicopathological parameters. Stained tissue Clinicopathological Total Percentage Percentage P- component parameter of cells of cells value positively positively stained stained ≤50% >50% Adenocarcinoma Pathological T staging T2a to T2c 38 28 10 0.087 100.0% 73.7% 26.3% T3a to T3c 34 18 16 100.0% 52.9% 47.1% Total 72 Extraprostatic extension Yes 34 18 16 0.087 100.0% 52.9% 47.1% No 38 28 10 100.0% 73.7% 26.3% Total 72 Seminal vesicle involvement Involved 7 3 4 0.227 100.0% 42.9% 57.1% Uninvolved 68 46 22 100.0% 67.6% 32.4% Total 75 Pre-operative PSA levels ≤8ng/ml 22 13 9 1.000 100.0% 59.1% 40.9% >8ng/ml 21 12 9 100.0% 57.1% 42.9% Total 43 Lymphovascular invasion Present 18 12 6 0.091 100.0% 66.7% 33.3% Absent 118 100 18 100.0% 84.7% 15.3% Total 136 Page | 318 Appendix II Perineural invasion Seen 89 71 18 0.367 100.0% 79.8% 20.2% Not seen 52 45 7 100.0% 86.5% 13.5% Total 141 Circumferential resection margin Involved 35 23 12 1.000 100.0% 65.7% 34.3% Uninvolved 40 26 14 100.0% 65.0% 35.0% Total 75 Apex margin Involved 35 24 11 0.633 100.0% 68.6% 31.4% Uninvolved 40 25 15 100.0% 62.5% 37.5% Total 75 Proximal margin Involved 9 7 2 0.710 100.0% 77.8% 22.2% Uninvolved 67 43 24 100.0% 64.2% 35.8% Total 76 Clinical stage T1a to T1c 37 25 12 0.045 100.0% 67.6% 32.4% T3a to T3b 8 2 6 100.0% 25.0% 75.0% Total 45 Lobe occurence Both Lobes 64 43 21 0.725 100.0% 67.2% 32.8% Singles Lobe 10 6 4 100.0% 60.0% 40.0% Total 74 Maximum tumour dimension <2.5cm 37 24 13 1.000 100.0% 64.9% 35.1% ≥2.5cm 38 25 13 100.0% 65.8% 34.2% Total 75 Page | 319 Appendix II Tumour Volume <4cm3 28 20 8 1.000 100.0% 71.4% 28.6% ≥4 cm3 29 20 9 100.0% 69.0% 31.0% Total 57 Peritumoural Pathological T staging Stroma T2a to T2c 38 8 30 0.090 100.0% 21.1% 78.9% T3a to T3c 34 2 32 100.0% 5.9% 94.1% Total 72 Extraprostatic extension Yes 34 2 32 0.090 100.0% 5.9% 94.1% No 38 8 30 100.0% 21.1% 78.9% Total 72 Seminal vesicle involvement Involved 7 0 7 0.584 100.0% .0% 100.0% Uninvolved 68 10 58 100.0% 14.7% 85.3% Total 75 Pre-operative PSA levels ≤8ng/ml 22 4 18 0.345 100.0% 18.2% 81.8% >8ng/ml 21 1 20 100.0% 4.8% 95.2% Total 43 Lymphovascular invasion Present 18 1 17 0.466 100.0% 5.6% 94.4% Absent 118 17 101 100.0% 14.4% 85.6% Total 136 Perineural invasion Seen 89 10 79 0.318 100.0% 11.2% 88.8% Not seen 52 9 43 100.0% 17.3% 82.7% Total 141 Page | 320 Appendix II Circumferential resection margin Involved 35 3 32 0.203 100.0% 8.6% 91.4% Uninvolved 40 8 32 100.0% 20.0% 80.0% Total 75 Apex margin Involved 35 6 29 0.746 100.0% 17.1% 82.9% Uninvolved 40 5 35 100.0% 12.5% 87.5% Total 75 Proximal margin Involved 9 1 8 1.000 100.0% 11.1% 88.9% Uninvolved 67 10 57 100.0% 14.9% 85.1% Total 76 Clinical stage T1a to T1c 37 4 33 1.000 100.0% 10.8% 89.2% T3a to T3b 8 1 7 100.0% 12.5% 87.5% Total 45 Lobe occurence Both Lobes 64 9 55 1.000 100.0% 14.1% 85.9% Singles Lobe 10 1 9 100.0% 10.0% 90.0% Total 74 Maximum tumour dimension <2.5cm 37 8 29 0.113 100.0% 21.6% 78.4% ≥2.5cm 38 3 35 100.0% 7.9% 92.1% Total 75 Tumour Volume <4cm3 28 7 21 0.025 100.0% 25.0% 75.0% ≥4 cm3 29 1 28 100.0% 3.4% 96.6% Total 57 Page | 321