PROTEOMIC ANALYSIS OF PROSTATE CANCER CELL LINE … · 2013-10-11 · cancer cell lines to identify secreted proteins that could serve as novel prostate ... project. As well as Drs

PROTEOMIC ANALYSIS OF PROSTATE CANCER CELL LINE CONDITIONED MEDIA FOR THE DISCOVERY OF CANDIDATE BIOMARKERS FOR

PROSTATE CANCER

by

Girish Sardana

A thesis submitted in conformity with the requirements for the Degree of Doctor of Philosophy

Graduate Department of Laboratory Medicine and Pathobiology University of Toronto

© Copyright by Girish Sardana 2008

PROTEOMIC ANALYSIS OF PROSTATE CANCER CELL LINE

CONDITIONED MEDIA FOR THE DISCOVERY OF CANDIDATE

BIOMARKERS FOR PROSTATE CANCER

Girish Sardana

Doctor of Philosophy 2008

Department of Laboratory Medicine and Pathobiology

University of Toronto

ABSTRACT

Early detection of prostate cancer is problematic due to the lack of a marker

that has high diagnostic sensitivity and specificity. The prostate specific antigen test,

in combination with digital rectal examination, is the gold standard for prostate

cancer diagnosis. However, this modality suffers from low specificity. Therefore,

specific markers for clinically relevant prostate cancer are needed. Our objective

was to proteomically characterize the conditioned media from human prostate

cancer cell lines to identify secreted proteins that could serve as novel prostate

cancer biomarkers.

An initial proof of principle study of the PC3 prostate cancer cell line was

conducted. From this study over 200 proteins were identified in the conditioned

media. Through gene ontology analysis and literature searches Mac-2 binding

protein was selected as a candidate biomarker for validation in the serum of prostate

cancer patients. A preliminarily validation showed that Mac-2 binding protein has

ii

discriminatory ability in prostate cancer diagnosis. However, an extended validation

did not confirm this.

Based on our proof of principle study we optimized our workflow and

extended our analysis by culturing three different prostate cell lines [PC3 (bone

metastasis), LNCaP (lymph node metastasis), and 22Rv1 (localized to prostate)].

We conducted a bottom-up analysis of each cell line by 2-dimensional liquid

chromatography and tandem mass spectrometry. Of the 2124 proteins identified,

12% (329) were classified as extracellular and 18% (504) as membrane-bound.

Among the identified proteins were known prostate cancer biomarkers such as PSA

and KLK2. To select the most promising candidates for further investigation, tissue

specificity, biological function, disease association based on literature searches, and

comparison of protein overlap with the proteome of seminal plasma and serum were

examined. Based on these results, several candidates were selected for validation in

serum of patients with and without prostate cancer. Of these four novel candidates:

follistatin, chemokine (C-X-C motif) ligand 16, pentraxin 3 and spondin 2 showed

discriminatory ability.

Of the four candidates, follistatin was further studied in an extended validation

in serum of patients with biopsy confirmed prostate cancer and tissues of prostate

cancer patients of low and high grade tumours by immunohistochemistry. In

addition, follistatin was also investigated in the tissue of colon and lung cancer

where intense staining was observed in one specimen of lung squamous carcinoma.

iii

ACKNOWLEDGEMENTS

First and foremost I would like to thank my supervisor Dr. Eleftherios P.

Diamandis for his guidance and vision for this project. It has been a maturing and

character building experience and I am truly grateful for the opportunity to have been

a part of the ACDC Lab.

I would like to also thank my advisory committee members Drs. Alex

Romaschin and Pui-Yuen Wong for their continued review of my work throughout my

project. As well as Drs. Robert Nam and Oliver John Semmes for their perspectives

and constructive criticism on my thesis. I would like to give special thanks to Drs.

John Marshall, Theodorus van der Kwast and Carsten Stephan for their

collaboration on this project. I am also grateful to the following for providing the

necessary resources to conduct my research: The Department of Laboratory

Medicine and Pathobiology, The University of Toronto, The Samuel Lunenfeld

Research Institute, Mt. Sinai Hospital and Proteomic Methods Inc.

There is a saying - “It’s the people that make the lab”. I would not have been

able to experience the true meaning of graduate studies without the friends and

colleagues that I met along the way. Special thanks to all the members of the ACDC

Lab past and present.

Finally, I would like to thank my family for their support during my graduate

studies.

iv

TABLE OF CONTENTS

ABSTRACT ii ACKNOWLEDGEMENTS iv TABLE of CONTENTS v LIST of TABLES viii LIST of FIGURES ix LIST of ABBREVIATIONS xi CHAPTER 1: INTRODUCTION 1.1 Prostate Cancer 2

1.1.1 Epidemiology and Statistics 2 1.1.2 Anatomy and Histology of the Prostate 5 1.1.3 Pathobiology of Prostate Cancer 7 1.1.4 Current Methods for Diagnosis 10 1.1.5 Current Treatments 13

1.2 Cancer Biomarkers 19

1.2.1 Introduction to Cancer Biomarkers 19 1.2.2 Methods for Identifying Novel Cancer Biomarkers 20

1.2.2.1 Genomic Approaches 21 1.2.2.2 Proteomic Approaches 24

1.2.3 Clinical and Analytical Properties of a Cancer Biomarker 27 1.2.4 Applications of Cancer Biomarkers 29 1.2.5 Current and Emerging Prostate Cancer Biomarkers 31

1.3 Prostate Cancer Model Systems 51 1.3.1 Mouse Models 51 1.3.2 Cell Culture Models 52

1.4 Continued Need for Novel Prostate Cancer Biomarkers 54 1.5 Rationale and Objectives of the Present Study 55

1.5.1 Rationale 55 1.5.2 Hypothesis 57 1.5.3 Objectives 57

v

CHAPTER 2: PROOF OF PRINCIPLE: PROTEOMIC ANALYSIS OF THE PC3 CELL LINE CONDITIONED MEDIA

2.1 Introduction 60 2.2 Materials and Methods 61

2.2.1 Roller Bottle Cell Culture 61 2.2.2 Dialysis 63 2.2.3 Fast Performance Liquid Chromatography of CM 63 2.2.4 Lyophilization and Digestion of Fractions 63 2.2.5 Liquid Chromatography–MS Analysis 64 2.2.6 Database, Genome Ontology, and Literature Search 65 2.2.7 ELISAs for Kallikrein 5, 6, and 11 66 2.2.8 Protein Recovery 66 2.2.9 Mac-2BP ELISA 66

2.3 Results 67

2.3.1 Total Protein, KLK5 and KLK6 Concentrations in Culture Over Time

67

2.3.2 Recovery of KLK5 and KLK6 During Sample Preparation 69 2.3.3 Proteins Identified by Mass Spectrometry 69 2.3.4 Overlap of Proteins Identified in Batches 1 and 2 73 2.3.5 Mac-2BP Concentrations in Patients with Prostate Cancer vs. Healthy Men and in the Conditioned Media of the PC3(AR)6 cell line

74

2.4 Discussion 79 CHAPTER 3: OPTIMIZATION OF CELL CULTURE AND PROTEOMIC WORKFLOW 3.1 Introduction 83 3.2 Materials and Methods 84

3.2.1 Cell Culture 84 3.2.2 Measurement of Total Protein, LDH and PSA, KLK5 and KLK6

84

3.2.3 Sample Preparation 85 3.3 Results 86

3.3.1 Cell Culture Optimization 86 3.3.2 Sample Preparation Optimization 90

3.4 Discussion 92

vi

CHAPTER 4: COMPARATIVE PROTEOMIC ANALYSIS OF THE CONDITIONED MEDIA OF THREE PROSTATE CANCER CELL LINES


4.2.1 Cell Culture 98 4.2.2 Measurement of Total Protein, Lactate Dehydrogenase and PSA, KLK5, KLK6

99

4.2.3 Conditioned Media Sample Preparation and Trypsin Digestion

99

4.2.4 Strong Cation Exchange High Performance Liquid Chromatography

101

4.2.5 Online Reversed Phase Liquid Chromatography – Tandem Mass Spectrometry

101

4.2.6 Database Searching and Bioinformatics 1024.2.7 Validation of Candidates 104

4.3 Results 105

4.3.1 Proteins Identified by Mass Spectrometry 1054.3.2 Identification of Internal Control Proteins 1064.3.3 Reproducibility between Replicates 1064.3.4 Differences in Proteins Identified Between Cell Lines 1094.3.5 Genome Ontology Distributions of Proteins 1094.3.6 Secreted and Membrane Proteins 1144.3.7 Overlap with Previous Data 1144.3.8 Overlap with Seminal Plasma Proteins 1144.3.9 Biological Network Analysis 1154.3.10 Overlap with Breast Cancer Secretome 1184.3.11 Validation of Follistatin, Chemokine (C-X-C motif) ligand 16, Pentraxin 3 and Spondin 2

118

4.4 Discussion 121 CHAPTER 5: VALIDATION OF CANDIDATE BIOMARKERS IN PROSTATE CANCER PATIENT SAMPLES


5.2.1 Conditioned Media, Serum Samples and Tissue Specimens

130

5.2.2 Quantification of candidates in biological fluids 1315.2.3 Immunohistochemistry of Follistatin 1345.2.4 Statistical Data Analysis 134

vii

5.3 Results 1365.3.1 Chemokine (C-X-C motif) ligand 5 1365.3.2 Chemokine (C-X-C motif) ligand 1 1385.3.3 Lipocalin-2 1405.3.4 Tumour necrosis factor receptor superfamily, member 6b, decoy

142

5.3.5 Extended Validation of Follistatin 144 5.4 Discussion 151 CHAPTER 6: SUMMARY AND FUTURE DIRECTIONS 6.1 Summary 157 6.2 Key Findings 157 6.3 Future Directions 159 REFERENCES 161 APPENDIX: Assay Reproducibility and Precision 201

viii

LIST OF TABLES

Table Title Page

1.1 List of candidate biomarkers for prostate cancer and their possible clinical utility.

50

2.1 Extracellular candidate tumor markers identified in culture medium of the PC3 (AR)6 roller bottle culture

71

4.1 Known prostate biomarkers identified in the conditioned media of PC3, LNCaP, and 22Rv1 cell lines

107

4.2 List of candidate biomarkers selected based on selection criteria. 127

ix

LIST OF FIGURES Figure Title Page

2.1 Schematic representation of the workflow for proteomic analysis of

roller bottle CM 62

2.2 Measurement of KLK5, KLK6 and total protein 68 2.3 Cellular location of proteins from PC3(AR)6 CM 70 2.4 Mac-2BP concentrations and the correlation between serum Mac-

2BP concentrations and PSA 76

2.5 Concentrations of KLK5, KLK6, and KLK11 in serum of 26 CaP patients and 17 healthy men

77

2.6 Serum concentrations of Mac-2BP in CaP, BPH and normal patients

78

3.1 Measurements of PSA and LDH in the CM of the 22Rv1 cell line 87 3.2 Measurements of PSA and LDH in the CM of the LNCaP cell line 88 3.3 Measurements of KLK5, KLK6 and LDH in the CM of the PC3 cell

line 89

3.4 Overview of the optimized sample preparation workflow 91 4.1 Workflow of proteomic method employed 100 4.2 Overlap of the 3 replicates from PC3, LNCaP and 22Rv1

conditioned media 108

4.3 Overlap of proteins identified between each cell line 110 4.4 Overlap of the extracellular and membrane proteins identified in

each cell line 111

4.5 Classification of proteins by cellular location 112 4.6 Classification of proteins by cellular location for each cell line 113 4.7 Molecular functions related to diseases associated with Follistatin 116 4.8 Molecular functions related to diseases associated with

Chemokine (C-X-C motif) ligand 16, Pentraxin 3 and Spondin 2 117

4.9 Concentrations of each candidate in the conditioned media of each cell line and the control flask

119

4.10 Validation of Follistatin, Chemokine (C-X-C motif) ligand 16, Pentraxin 3 and Spondin 2 in serum

120

5.1 Measurement of CXCL5 in CM of CaP cell lines and serum of CaP patients and normals

137

5.2 Measurement of CXCL1 in CM of CaP cell lines and serum of CaP patients and normals

139

5.3 Measurement of LCN2 in CM of CaP cell lines and serum of CaP patients and normals

141

5.4 Measurement of TNFRSF6B in CM of CaP cell lines and serum of CaP patients and normals

143

5.5 Follistatin serum levels and correlations with PSA, Gleason score and age

146

x

Figure Title Page

5.6 PSA, %fPSA, KLK2 and KLK11 serum concentrations in patient serum

147

5.7 Prostate tissue specimens stained for follistatin 148 5.8 Lung tissue specimens stained for follistatin 149 5.9 Colon tissue specimens stained for follistatin 150

xi

LIST OF ABBREVIATIONS

2-DE two-dimensional gel electrophoresis

A2M α-2-macroglobulin

ACN acetonitrile

ACT α-1-antichymotrypsin

AIPC androgen independent prostate cancer

AMACR α-methylacyl CoA racemase

ANN artificial neural networks

ANXA3 annexin A3

API α1-protease inhibitor

AR androgen receptor

AUC area under curve

BPH benign prostate hyperplasia

BPSA BPH associated PSA

CaP prostate cancer

CDCHO chemically-defined Chinese Hamster Ovary

CGH comparative genomic hybridization

cPSA complexed PSA

CM conditioned medium

CXCL16 chemokine (C-X-C motif) ligand 16

DiEtOH diethanolamine

DRE digital rectal exam

DTT dithiothreitol

xii

ELISA enzyme linked immunosorbent assay

EPCA early prostate cancer antigen

ESI electrospray ionization

EZH2 enhancer of zeste homolog gene 2

FCS fetal calf serum

fPSA free PSA

FISH fluorescence in situ hybridization

FPLC fast-performance liquid chromatography system

FPR false positive rate

FSH follicle-stimulating hormone

GO gene ontology

GSTP1 glutathione S-transferase pi

IGF insulin-like growth factors

IGFBP insulin-like growth factor binding proteins

IL-6 interleukin 6

iPSA intact PSA

KLK1 pancreatic/renal kallikrein 1

KLK2 human kallikrein-related peptidase 2

LCM laser-capture microdissection

LC-MS/MS liquid chromatography-tandem mass spectrometry

LDH lactate dehydrogenase

LH luteinizing hormone

LHRH LH releasing hormone

xiii

Mac-2BP mac 2 binding protein

MALDI matrix assisted laser desorption ionization

MS mass spectrometry

MS/MS tandem MS

NCBI national center of biotechnology information

NGEP new gene expressed in the prostate

NPV negative predictive value

%fPSA percent free PSA

PAP prostatic acid phosphatase

PBS phosphate-buffered saline

PCA3 prostate cancer antigen 3

PCPT the prostate cancer prevention trial

PIN prostatic intraepithelial neoplasia

PPV positive predictive value

PSA prostate specific antigen

PSAD PSA density

PSAD-TZ transitional zone PSA density

PSMA prostate specific membrane antigen

PSAV PSA velocity

proGRP progastrin-releasing peptide

PSCA prostate stem cell antigen

PSP94 prostate secretory protein 94

PSPBP PSP94-binding protein

xiv

xv

PTEN phosphatase and tensin homologue deleted from chromosome 10

PTX3 pentraxin 3

RB retinoblastoma

RT-PCR real-time polymerase chain reaction

ROC receiver operating characteristic curve

SAX strong anion exchange

SCX strong cation exchange

SELDI-TOF surface enhanced laser desorption ionization time of flight

SFM serum-free media

SILAC stable isotopic labelling with amino acids in cell culture

SPON2 spondin 2

TBST tris-buffered saline with tween 20

TGF-β1 transforming growth factor β1

TFA trifluoroacetic acid

tPSA total PSA

TMN tumour node metastasis

TRAMP transgenic adenocarcinoma of the mouse prostate

TRUS trans-rectal ultrasound

TZ transitional zone

uPA urokinase-plasminogen activation

Chapter 1: Introduction 1

CHAPTER 1:

INTRODUCTION


1.1 Prostate Cancer

The prostate is recognized for having a relatively high occurrence of tumours

in men, especially older men. As a result, prostate cancer (CaP) is the most

prevalent malignancy in men and is the second leading cause of cancer deaths in

North America with one in six men having a lifetime risk of being diagnosed and a

3.4% chance of death due to CaP(1). Most patients currently are being diagnosed

with early stages of the disease with no symptoms(2). As a result of this increase in

early stage tumour diagnoses or stage migration, the classical ways of prognosis,

such as the Partin Tables and Kattan Nomograms are no longer as effective. The

focus has now moved from early detection to the determination of the clinical

significance of these early stage tumours. One objective is to find ways of

distinguishing clinically relevant tumours that have the ability to metastasize.

Currently, 30% of tumours removed by radical prostatectomy are deemed clinically

insignificant and would not have required such invasive treatment. Most patients

diagnosed have a latent, non-aggressive form of CaP(3), thus it is important that

these patients are not over-treated. Currently, little is known about the molecular

pathogenesis of CaP(4). Control of CaP could be achieved through early detection

and appropriate selection of treatment; however we have yet to reach this level of

sophistication.

1.1.1 Epidemiology and Statistics

It is estimated that there will be over 500,000 new cases worldwide of CaP

making it the sixth cancer in the world in terms of incidence. It is the most common


cancer in North America, Europe and parts of Africa. Current statistics from the

Canadian Cancer Society show that 24,700 estimated new cases with 4,300 deaths

in 2008(5). Before the age of 40, CaP is rarely diagnosed however its clinical

incidence increases with age. The mean age of diagnosis of CaP is 72-74 with the

majority of patients diagnosed after the age of 65. Prostate cancer incidence rates

are expected to rise as life expectancies continue to increase. The rise in incidence

rates since the mid 1990’s has been attributed to the implementation of the prostate

specific antigen (PSA) test for population screening in Western countries. While

incidence rates have been rising, mortality rates in Canada have been shown to

decrease significantly by 2.9% from 1995 to 2004(5). While this has been presumed

to be a result of screening and improved treatment there are two prospective clinical

trials underway that are assessing this: the European Randomized study of

Screening for Prostate Cancer and Prostate, Lung, Colorectal and Ovarian Cancer

Screening Trial.

While CaP claims the highest incidence rates among all cancers in men, this

does not take into account the prevalence of small adenocarcinomas that are found

incidentally upon autopsy or during other surgical procedures. The prevalence of

these ‘latent’ tumours is thought to be over 50% in men over 50(6). This highlights

that men are more likely to die with CaP rather than from it. This attribute of CaP

makes it very difficult in its diagnosis where the high prevalence of clinically

insignificant tumours could result in an increase in unnecessary treatment.

Geographically the highest incidence in the world of CaP is centered mainly in

the United States and Scandinavian countries and the lowest in Mediterranean


countries and Japan(7). Within countries, CaP differs in ethnic groups. With respect

to the US, African-American males harbour an increased incidence versus

Caucasian, native Chinese and native Japanese(8). An increase in the incidence of

CaP in Japanese males who move to Western countries suggests that life-style

factors may play a role in the development of detectable CaP(9). Other factors that

may account for these differences include access to health care and how CaP is

reported.

Differences in genetic susceptibility are also attributed as being a

predisposing factor for CaP. The chance for developing CaP increases by two to

three times when a first degree relative is afflicted with the disease and increases

when there are more affected individuals in a family, between 10-15% of CaP

patients have a relative who is affected(10). Twin studies have also assessed the

link between genetics and environmental factors showing a possible autosomal

dominant inheritance of a high risk gene(11). Strong candidate genes for genetic

association of CaP are those involved with androgen regulation(12) and

metabolism(13).

Androgen hormones are required for the development of the normal prostate

and have been linked to CaP progression. As a result anti-androgen therapies have

proven to be of great use in treating CaP in 75-80% of patients with metastatic CaP.

While testosterone and dihydrotestosterone have shown to induce CaP tumours

there is very little epidemiological data to support androgen levels as a risk factor for

CaP(14).


In addition to genetics and hormonal influences, nutritional factors have been

shown to play a role in CaP incidence. Studies have shown that CaP is associated

with the Western lifestyle with respect to a diet that contains high fat, red meat and

dairy(15). Studies continue to show conflicting data but there are some consistently

reoccurring components such as α-linolenic acid(16) and calcium(17). Conversely,

protective effects of some diets such as those found in Asia have been attributed to

high intakes of phyto-oestrogens from soybeans which have a preventative effect on

CaP(18). Intake of lycopene from tomato based products has also been shown to

reduce risk of CaP by acting as an antioxidant(19). However, the most promising

data of nutritional prevention of CaP comes from selenium and vitamin E. Both are

considered antioxidants that also enhance the immune system and induce apoptosis

in cancerous cells. The value of these two nutrients in prevention of CaP is currently

being studied in the Selenium and Vitamin E Cancer Prevention Trial (SELECT) and

final results will be available by 2013(20).

1.1.2 Anatomy and Histology of the Prostate

The prostate is an accessory reproductive organ that consists of exocrine

glands that are organized in a fibromuscular architecture surrounding the proximal

urethra. It is located in the abdominal cavity inferior to the bladder where the urethra

from the bladder passes through the prostate before entering the penis. The portion

of the urethra in the prostate is termed the prostatic urethra. Connected to the

prostate are two seminal vesicles which connect via two deferential ducts that

originate from the epididymis. The seminal vesicles and deferential ducts connect to


the prostatic urethra via two ejaculatory ducts within the prostatic gland. Within the

pelvis the prostate is surrounded by striated musculature as well as fatty tissue and

neurovascular bundles.

The glandular tissue within the prostate can be divided into different regions

anatomically. These are the five lobes of the prostate: the anterior, middle, posterior

and two lateral lobes. The prostate can also be subdivided functionally into two

zones. These are the centrally located transition zone, the posteriorly and laterally

located peripheral zone. The transition zone is the main site for benign prostate

hyperplasia (BPH) development while the peripheral zone is where malignant

tumours are found(21).

The prostate produces substances that add to those produced by the seminal

vesicles and the spermatozoa. Approximately 60% of semen consists of

seminogelin – a protein produced by the seminal vesicles. Thirty percent is produced

by the prostate which adds proteins and enzymes that affect the liquidity of the

semen, and the remaining 10% are secretions from the bulbourethral glands that lie

‘downstream’ from the prostate. These secretions contribute to the protective liquid

environment of the semen and for nourishment of the sperm.

The histology of the prostate is similar in all regions consisting of ducts and

acini which are lined by a double epithelial layer. The luminal epithelial cells are

cuboidal or columnar in shape and the basal epithelial cells are flattened. Below

these two layers is the basal membrane. Throughout the gland are scattered

neuroendocrine cells that lie above the basal cell layer. The glands are supported

by a dense stroma of smooth muscle fibres and fibrous connective tissue which


exerts its main function during ejaculation when prostatic secretions are ejected

together with semen.

The luminal cells in the ducts and acini produce proteins that become

constituents of semen. These include enzymes such as prostatic acid phosphatase

(PAP) and PSA. Prostate specific antigen is responsible for liquefying the seminal

clot by digesting seminogelin and thus increasing the motility of sperm. Basal cell

function is still not yet known but have been hypothesized to act as either stem-like

cells to replenish the luminal epithelium or act as functional cell barrier between

luminal cells and the stroma and are involved in growth and differentiation of the

epithelium(22). Neuroendocrine cells in the prostate produce peptides such as

chromogranin A, serotonin and progastrin-releasing peptide. The exact function of

this is currently unknown, however rare tumours are derived from neuroendocrine

cells and these peptides have been shown to be useful as tumour markers in their

diagnosis(23,24).

1.1.3 Pathobiology of Prostate Cancer

Cancers including CaP have been shown to develop from accumulating

genetic alterations that translate into increased cell growth and propensity to invade

surrounding tissues. Morphologic alterations in prostate tissue histology are evident

that indicate it is a multistep process that involve genetic mutations including

deletions, amplifications, rearrangements and epigenetic modifications(25).

Phenotypic changes during progression of CaP have been characterized to follow

transition states from benign glands to the malignant phenotype. Initially, epithelial


cells are seen to progress to prostatic intraepithelial neoplasia (PIN) within benign

glands. Here we begin to see dysplastic epithelial cells that can be categorized as

high and low grade PIN(26). High grade PIN is considered a precursor to CaP

however not all PIN lesions detected progress to CaP(27). It is hypothesized that

tumour cells are themselves genetically instable and are thus prone to genetic

mutations(28). Transformed cells that develop a survival advantage then proliferate

over the surrounding cells and form a tumour.

Results from epidemiologic studies have shown a genetic susceptibility risk

for men with a family history of CaP as well as through twin studies(29,30). From

linkage analysis there have been several CaP susceptibility loci that have been

reported(31). Candidate genes have emerged that may play a role in CaP

progression. These include RNASEL which encodes a ribonuclease that can induce

apoptosis upon viral infection(32) and MSR1 which is a lipopolysaccharide receptor

found at sites of inflammation in the prostate(33). While the prevalence of familial

CaP is low, most tumours arise from somatic mutations that accumulate over several

decades. An alteration that occurs frequently in the CaP genome is a gain or loss of

a region of DNA. Common gains are at 7p, 7q, 8q, Xq and common losses are at 8p,

10q, 13q and 16q(34-36). The frequency of these alterations varies with reports;

however, by identifying functions of disrupted genes a molecular pathway may be

developed to explain the process of the growth, metastasis and progression to

androgen independent prostate cancer (AIPC). However, due the heterogeneity of

the tumours this process is complicated and requires careful dissection(37).


Deletion, inactivation by mutation or suppression of genes by promoter

methylation that function in regulating apoptosis and cell proliferation may act to

permit the progression of CaP. These tumour suppressor genes have been shown to

be important in several aspects of CaP cell growth. A gene lost early at the stage of

PIN due to promoter methylation is the glutathione S-transferase pi-1 (GSTP1) gene

which helps to neutralize oxidative stress in basal and epithelial cells(38). Additional

loss of function of the retinoblastoma (RB) and TP53 gene that regulate cell division

have been shown in CaP cell lines and metastatic disease(39). A prominent tumour

suppressor gene lost in many CaP tumours is the phosphatase and tensin

homologue deleted from chromosome 10 (PTEN). This gene encodes a

phosphatase that inactivates the PI3 kinase and Akt pathways that result in

regulating apoptotic activity(40).

In addition to inhibition of tumour suppressor genes, activation of proto-

oncogenes in CaP are observed. Gain of function mutations are evident in these

genes that result in their over-expression or constitutive activation resulting in either

increased proliferation or decreased cell death. Growth factors and their receptors

that have been shown to be overexpressed include epidermal growth factor,

transforming growth factor-α and β, platelet-derived growth factor and insulin-like

growth factor-1 and 2(41-44). The Her-2/neu gene encodes a transmembrane

tyrosine kinase where overexpression in breast cancer has shown prognostic and

therapeutic significance. Her-2/neu amplification is uncommon in CaP, however

overexpression has been observed in metastatic CaP and AIPC(45,46). Another

potent oncogene is c-myc amplification which is associated with AIPC(47). The


cyclin-dependent kinases, a family of protein kinases that regulate the cell cycle

have also been shown to have altered expression in AIPC lesions(48). Mutations in

genes involved in cell survival pathways such as telomerase, BCL-2 and the ras

family of oncogenes have also been implicated in CaP progression(47,49,50).

The androgen receptor (AR) plays a pivotal role in the development of the

prostate and the progression of CaP(51). Androgen resistant tumours are thought to

arise through selection pressure for cells containing mutations in the AR gene

allowing it to grow independently of testosterone. One mechanism is the increased

expression of the AR in AIPC that would permit the growth of the tumour in a low

testosterone environment(52). Mutations in the ligand binding domain of the AR

result in gain of function of the receptor by allowing it to be able to bind and be

activated by a variety of ligands(53).

1.1.4 Current Methods for Diagnosis and Prognosis

The diagnosis of CaP is usually based on symptoms, an abnormal digital

rectal exam (DRE) or PSA test. Confirmation of diagnosis is then performed by a

trans rectal ultrasound (TRUS)-guided biopsy, and histological examination of the

prostate tissue is performed to assess the presence of cancer. It is common

practice in North America that men over the age of 50 have a routine DRE and PSA

test performed. Men that are at high risk may have the exams performed earlier.

This form of screening for CaP has resulted in the stage migration to organ confined

tumours, however, the benefit in terms of increasing overall survival is hotly debated.


A CaP patient may present with symptoms stemming from urinary

obstruction. These can be hesitancy, nocturia, incomplete emptying, hematuria,

diminished urinary stream or impotence. These symptoms are also commonly

associated with BPH and warrant further examination by DRE and PSA test.

Currently, patients rarely present with symptoms of metastatic disease such as

anemia, bone pain or neurologic related symptoms associated with cord

compression or metastases to the brain.

A DRE is usually performed to assess the physical features of the prostate

and to determine areas of the prostate that have hardened and to determine if there

is extension to the area surrounding the prostate. This exam is subjective and the

results do not correlate well with tumour volume and extent of disease progression.

However, it is useful in determining if the tumour has spread to outside the capsule

of the prostate or whether it is confined.

Accompanying a DRE is usually the PSA test which is a serum blood test for

the presence of elevated levels of PSA which is a highly prostate specific protein

that is produced in large amounts by prostate epithelial cells. It has been shown that

normal and BPH tissue produce more PSA than CaP tissue with poorly differentiated

CaP producing less than well-differentiated CaP(54). Elevated serum levels of PSA

due to CaP growth result from disruption of the basement membrane surrounding

the epithelial cells and leakage of into capillaries and lymphatics. The normal range

of PSA is less than 4 ng/mL, however, it has been realised that PSA can not be used

as a dichotomous marker and levels are considered as a continuum. As a result,


there is considerable overlap between levels of PSA in benign conditions and CaP,

thus reducing the specificity of PSA.

Based on the results of the PSA and DRE a TRUS is performed of the

prostate to determine staging, prostate volume and to determine sites for needle

biopsy. Cancerous tissue is usually hypoechoic compared to the surrounding

prostate tissue and thus can be visualized with TRUS, however, this method can not

predict CaP with certainty(55). Based on the TRUS results, a needle biopsy is

performed transrectally using a TRUS-guided spring-loaded needle gun. Areas that

are felt or visualized to be abnormal are sampled with at least 10 to 12 cores. Each

core is then assessed by a pathologist to determine the spread of the tumour within

the prostate. Biopsies are usually taken from the peripheral or transition zone of the

prostate.

To assess the presence and differentiation of CaP the Gleason grading

system is used(56). Gleason grading is based on the pattern of the glands and

combines in the first and second most prominent patterns evident. Grading is on a

scale of 1 to 5 increasing from well differentiated tissue architecture to poorly

differentiated structures. The grading from each pattern is then summed to attain the

Gleason score from 2 to 10. Based on the Gleason score further imaging of the

surrounding tissues may be warranted to determine staging and spread of the

tumour outside the prostate. Imaging modalities that are currently used include:

ultrasound, computed tomography, magnetic resonance imaging in combination with

contrast agents are used to determine spread to the lymph nodes and bone. Based

on these results the tumour may be staged based on the Tumour Node Metastasis


(TNM) staging system. From this a prognosis may be determined and appropriate

treatment decisions can be made.

1.1.5 Current Treatments

Clinical treatment decisions for CaP, especially in CaP that is localized and

detected early, are problematic when determining whether to treat or not to treat the

patient. Several factors contribute to determining this decision namely the risk

benefit ratio of the adverse events of treatment versus the life expectancy of the

patient. Due to the slow growing nature of CaP it is important to determine the

prognosis of the patient. Recommendations for treatment based on risk stratification

has recently been suggested by the National Institute for Health and Clinical

Excellence(57). Ultimately it is up to the healthcare professional to ensure the

patient receives adequate information so as to make an informed decision.

Initially, a conservative management approach is taken that is referred to as

“watchful waiting”. Here patients enter a clinical follow-up program in which the

tumour size and progression is measured by serum PSA testing. If there is evidence

of CaP progression then curative or palliative treatments may be considered. For a

substantial number of CaP patients, especially older patients with localised and low

grade tumours, a watchful waiting approach might be the better option. Data from a

recent study from the Surveillance, Epidemiology, and End Results Program showed

that patients who undertook conservative management of clinically localized CaP

had a 10-year survival of 94% for Gleason 2 to 4 tumours and 45% for Gleason 8 to

10(58).


Radical Prostatectomy

Due to the use of PSA testing there has been an increase in the number of

organ confined tumours, the majority of which are deemed to be clinically

relevant(59). Thus, these tumours are at the stage where they may be removed to

cure a patient’s CaP. Advances in surgical techniques over the decades have

revolutionized the ability to cure locally confined CaP. Radical prostatectomy is a

surgical procedure where all or part of the prostate gland is removed either as a

result from a progressing tumour or if there is urinary obstruction due to enlargement

of the prostate from BPH or CaP that restricts the flow of urine in the urethra. The

ideal candidate for radical prostatectomy would have a life expectancy of 10 to 20

years post surgery, should be free of co-morbidities and have a clinically relevant

tumour that is confined to the prostate and thus curable by surgery. Current surgical

techniques allow for nerve sparring where erectile function is preserved and

incontinence can be prevented. However, these are still possible long term side

effects from the procedure. Radical prostatectomy is not recommended for patients

with metastatic disease.

Radiation Therapy

Similar to surgery, the objective of radiation therapy is to cure patients with

localized CaP by eliminating the tumour from the prostate. In this case, external-

beam radiation or implanted radioactive ‘seeds’ termed brachytherapy are delivered

to the prostate with the least possible harm to surrounding tissues. The ionizing

radiation given during radiation therapy damages the DNA in cells. Cancer cells


have a reduced ability to repair this damage and thus are unable to proliferate, while

normal cells that receive this radiation are able to repair any DNA damage.

Since the advent of external-beam radiation for treatment of localized CaP

there have been several advances that have allowed improvements in treatment

planning and delivery. As a result, the curative ability of external-beam radiation for

localized CaP has improved over the past decades(60). The type of radiation used is

usually X-rays or gamma rays and is given daily over the course of 6 to 10 weeks.

The implementation of three-dimensional conformal radiation therapy has been able

to incorporate more precise definition of tumour volume and thus calculation of dose

delivered. The ability to exclude normal tissues from receiving high-dose radiation

allows for higher tumour doses without increased surrounding tissue toxicity(61). To

obtain three dimensional images, computed tomography images are taken to

produce high-resolution three dimensional images of the prostate and surrounding

tissues. From these decisions on delivery and dose of radiation are determined(62).

A further advancement has been the implementation of intensity-modulated radiation

therapy which allows dose distributions that conform to the tumour area with

enhanced precision(63).

The use of brachytherapy relies on the principle that high does of radiation

can be given directly within tumours without the radiation leaking to surrounding

tissues. Currently, the mode of delivery is through TRUS-guided implantation of the

radioactive isotopes 125I or 103Pd with or without external-beam radiation(64). Use of

permanent metal seed implants in combination with external-beam radiation therapy

has shown improvements in survival(65). Some side effects of both external-beam


and brachytherapy include incontinence, impotence, infertility, bowel problems and

fatigue.

Androgen Ablation Therapy

Steroid hormones such as testosterone and other androgens have been

shown to play an essential role in the growth and development of the prostate.

Circulating testosterone is converted to its more potent form in the prostate by 5α-

reductase to dihydrotestosterone. This molecule then enters prostate epithelial cells

and binds the androgen-binding region of the AR which then binds to androgen

response elements on DNA promoting transcription of androgen-responsive genes

and prostate cell growth. The major source of testosterone production is from the

testes, with a small proportion coming from the adrenal glands. Testosterone

secretion is regulated by secretion of luteinizing hormone (LH) and follicle-

stimulating hormone (FSH) from the anterior pituitary which stimulates the Leydig

cells in the testes and causes them to produce testosterone. Secretion of LH is

regulated by LH releasing hormone (LHRH) also known as gonadotropin-releasing

hormone which is secreted from the hypothalamus. The secretion of LH and LHRH

is under the control of a negative feedback loop where the concentration of

testosterone in the circulation acts to inhibit the secretion of LH and LHRH.

The relationship with prostate and CaP growth and androgens was

discovered over sixty years ago(66) and was marked with the awarding of the 1966

Nobel Prize. Since then we know that CaP cells deprived of androgens slow in

proliferation and under apoptosis while normal prostate tissue undergoes atrophy

and a reduction in prostate gland size. Thus, methods to block the action of


androgens on the prostate were developed and have since been used in the

treatment of metastatic CaP.

The gold standard for androgen ablation therapy and treatment for metastatic

CaP has been bilateral orchiectomy (removal of the testes). By this method serum

levels of testosterone are reduced by 95% and due to the reduction in testosterone

there is permanent rise in LH and FSH. Administration of estrogens also reduces

testosterone levels by down-regulating the secretion of LH and FSH by the pituitary.

Diethylstilbestrol is used during this therapy with castrate levels of testosterone

being achieved between 3 to 9 weeks(67). Side effects from estrogen therapy

include nausea, vomiting, gynecomastia, fluid retention as well as cardiovascular

effects. Anti-androgen therapy has also been used to reduce the effects of

androgens to treat CaP. There are two classes of anti-androgens: steroidal and

non-steroidal. Steroidal anti-androgens include cyproterone acetate and megestrol

acetate which act by inhibiting C21-9 decarboxylase – an enzyme that synthesises

androgens, and by inhibiting LHRH release. Non-steroidal anti-androgens act by

competitively blocking the binding of dihydrotestosterone to the androgen receptor.

Three agents that are commonly used are flutamide, bicalutamide and nilutamide.

The agents will decrease the intracellular levels of testosterone and

dihydrotestosterone but in contrast to steroidal anti-androgens there will be

increased serum levels of testosterone due to increased LHRH. The chemical

androgen ablation therapy of choice has become LHRH agonists. Two agonists that

are used are leuprolide acetate and goserelin acetate which are taken

subcutaneously. They function by negatively inhibiting LHRH release in the pituitary


through a negative feedback loop and by downregulating the receptors for LH in the

pituitary. An initial rise in LH and FSH is observed initially and this is compensated

for by the administration of anti-androgens. A meta-analysis of monotherapies used

in androgen ablation showed that LHRH agonists were equivalent to patients treated

with orchiectomy(68).

Approximately 70% of CaP patients respond well initially to hormonal

treatment(69). However, most tumours eventually acquire androgen independence

and can grow without the presence of androgens. The prognosis of these patients is

poor with median survival time of less than 10 months(69). New chemotherapeutic

approaches have been developed that combat CaP at this stage.

Cytotoxic Chemotherapy for Androgen Independent CaP

Owing to the slow growing nature of early CaP, chemotherapeutic agents are

not effective in targeting these cells preferentially. As a result, uses of

chemotherapies are restricted to treatment of patients with AIPC. Current

chemotherapeutic agents used for treatment of AIPC have not been able to show a

survival benefit, however, they have been found useful for palliative care and to

improve overall quality of life. Current chemotherapies for AIPC include:

mitoxantrone and prednisone for the palliation of pain of bone metastases(70).

Mitoxantrone acts by inhibiting type II topoisomerase thus disrupting DNA synthesis

and repair. Doxorubicin and etoposide act in similar ways by intercalating with DNA

and disrupting topoisomerase activity. Estramustine, which is a derivative of

estrogen and acts as an alkylating agent that binds and disrupts DNA, has been

shown with etoposide to decrease PSA and show a non-statistical survival


benefit(71). Vinblastine is a microtubule inhibitor and functions by binding tubulin

and halting cell cycle progression at M phase. Paclitaxel and its semi-synthetic

derivative docetaxel are both from the drug class of taxanes and both bind tubulin in

microtubules with high affinity thus stabilizing the structure of the cytoskeleton and

preventing depolymerisation and as a result, freezing the cell from undergoing cell

division. Docetaxel has shown reductions in PSA and non-statistical increases in

progression-free survival(72).

1.2 Cancer Biomarkers

1.2.1 Introduction to Cancer Biomarkers

Cancer biomarkers, or tumour markers, can be defined as measurable

analytes in biological tissues or fluids that act as a surrogate indication for the

presence of a tumour. They may be produced by the tumour itself or by the host in

response to the presence of the tumour. The analyte may be measured qualitatively

or quantitatively by several methods such as proteomic, genomic or chemical to

determine its presence. The ideal tumour marker should be both specific and

sensitive for the detection of a certain type of tumour early in its progression.

Currently, most tumour markers are not sensitive or specific enough to be used in

routine population screening and are used in monitoring patients after therapy.

Tumour biomarkers have been shown to span many different forms. These

include small molecule, gene or protein based analytes. Small molecule metabolites

have been identified resulting from the metabolic pathways of cancer cells. In

addition, some cancers have shown higher concentrations of certain micronutrients.


Nucleic acid based analysis of cancers has resulted in a plethora of differing forms

of cancer biomarkers. These include DNA based mutations in cancer cells, single

nucleotide polymorphisms, altered expression of certain genes measurable by their

transcript levels and epigenetic modifications that alter gene expression. Changes in

peptide and protein levels have also been assayed as cancer biomarkers. These

include altered expression of peptides and proteins, post-translational modifications,

variants of protein isoforms, mutations in proteins, as well as autoantibodies. Both

genomics and proteomics have been utilized in multi-parametric approaches such as

gene and protein microarrays and mass spectrometric proteomic patterns. In

addition, computational approaches to the formation of multi-panel biomarkers

through the use of artificial neural networks have resulted in increased performance

of several biomarkers. With evolving technologies such as high throughput genomics

and proteomics, the search for novel biomarkers that are of high quality is ongoing.

1.2.2 Methods for Identifying Novel Cancer Biomarkers

The opportunity to identify candidate cancer biomarkers has materialized in

the past decade with the completion of the human genome sequence and the advent

of high throughput sequencing technologies, microarrays and mass spectrometric

approaches for the identification of proteins. As a result, there have been large

efforts in developing methods to identify novel tumour markers that show clinical

utility for diagnosis, prognosis or therapeutic monitoring. Here we discuss two main

approaches, genomics and proteomics that have been applied to the search for

cancer biomarkers.


1.2.2.1 Genomic Approaches

Of recent there has been an emergence of genetic tools that have made

available the comparison of genetic profiles of normal and cancerous tissue and cell

cultures. The gold standard to assess gene transcript levels has been the use of

real-time polymerase chain reaction (RT-PCR) which is able to quantitatively

determine levels of mRNA transcripts in several samples simultaneously. However,

the most common high throughput tool in use is the DNA microarray which can

assess either DNA gene copy number alterations or transcript levels of a gene’s

mRNA on a genome wide scale. Microarrays have thus provided several insights

into the molecular mechanisms of cancer biology, as well as led to the discovery of

new markers for disease and therapeutic response. The application of microarrays

to CaP was thought to help elucidate the genetic underpinnings of the differences

between indolent tumours and tumours with a clinically aggressive phenotype. Due

to the several challenges posed by the nature of CaP this has resulted in limited

success.

There are unfortunately limited models for CaP such as cell lines which are

mainly attained from metastatic sites and patients with AIPC(73). Several animal

models established have been studied but their relevance to human CaP

progression is controversial, albeit recent transgenic models have shown promise for

study(74). When using fresh frozen tissue samples for genomic analysis the tissue

samples are invariably contaminated with normal glandular architecture and stromal

cells, thus confounding results obtained due to high background from normal cells.

Issues also occur when selecting normal tissues for control since a field effect may


be present that affects normal epithelium or the presence of PIN lesions(75).

Therefore, tissue samples should be carefully examined histologically before being

used for RNA extraction in a genomic study. Improvements in tissue dissection

including laser-capture microdissection (LCM) have shown to improve the purity of

tissue samples obtained(76). However, this procedure has been subject to

contamination with artefacts.

In the quest for the identification of differentially expressed biomarkers,

several studies have compared gene expression of cancerous to normal tissues or

BPH tissue. A number of novel genes have been discovered that have shown

differential expression. One gene, AMACR, was found to be expressed in almost all

CaP and confirmed by immunohistochemistry and has since become a routine

diagnostic tool for pathologists(77). Due to the increased number of datasets

emerging from microarray studies of CaP, discrepancies between datasets were

seen. This may in part have been due to selection bias of samples used for each

dataset, or differences in sample preparation or the genes represented on

microarrays used. To overcome these differences a meta-analysis of independent

data sets was conducted to validate and establish a model for CaP gene

expression(78). From this 50 genes were identified to be overexpressed and 103

genes were underexpressed in CaP with 11 novel genes being found to score high

that were not noted in their original datasets. Value has also been gained in the

reanalysis of a public data set of Lapointe et al.(79). Tomlins et al.(80) used cancer

profile outlier analysis to identify translocations of gene fusions. This involved

searching multiple microarray data sets for gene expression differences that were


consistently different between normal and CaP tissues. Through this they discovered

the novel gene fusion of the TMPRSS2 serine protease gene to two ETS family

transcription factor genes which were found to be overexpressed in CaP; ERG and

ETV1, which is under current investigation as being a prognostic marker for CaP.

Correlating clinical outcome with transcript levels has also been an

approached used for stratification of tumour subtypes. An example is hepsin which is

seen to be elevated in CaP relative to normal, but tumours with relatively lower

levels of hepsin have a worse prognosis(81). A comparison of normal, localized and

metastatic tumours resulted in identification of EZH2 which was correlated by

immunohistochemistry with recurrent CaP after surgery(82). In contrast to identifying

single genes that discriminate CaP from normal, another approach has been to use

gene expression patterns to identify subtypes of tumours and predict prognosis. In

this instance unsupervised hierarchical clustering is used to separate tumours based

on clinical outcome(79). From the data used by Lapointe et al. they were able to

distinguish 4 genes positively and 19 genes negatively associated with recurrence of

CaP tumours. Of these MUC1 And zinc-α-2-glycoprotein were also confirmed by

immunostaining to have poor prognosis and tumour recurrence. The comparison of

gene expression data from cell lines and animal models with CaP tissue samples

has also proven useful. A tumour stem cell gene signature was identified by

Glinskey et al. that predicted failure of therapy(83).

In addition to gene expression changes, alterations in gene copy number due

to chromosomal aberrations are associated with malignant transformation(4).

Several platforms are available that are used to assess DNA copy number


alterations by comparative genomic hybridization (CGH) including BAC arrays, SNP

chips, and cDNA microarrays. Numerous studies have been performed to analyze

copy number alterations including those done using cell lines(84,85) and with tumour

samples(86,87). The significance of these changes will facilitate the identification of

genes correlating with clinical outcome and identify the critical regions of genetic

alteration in CaP progression.

1.2.2.2 Proteomic Approaches

High-throughput technologies that are used to proteomically profile biological

specimens have evolved tremendously over the last decade. A large focus has

been on the identification of novel biomarkers for disease. The rationale used for a

proteomic approach versus a genomic approach to biomarker discovery is that

proteins are the effector molecules of the cell, and differences caused during the

progression to a malignant phenotype will be evident in differences in the proteome.

These proteomic alterations are then candidates for identification as surrogate

biomarkers either in tissues or biological fluids.

Proteomic studies first involved the use of two dimensional gel

electrophoresis (2-DE), however, novel technologies that have emerged include

protein microarrays, ‘soft’ ionization techniques such as electrospray ionization (ESI)

and matrix assisted laser desorption ionization (MALDI), improvements in mass

spectrometer design such as the LTQ-Orbitrap (ThermoFisher Scientific), and

quantitative isotopic labelling approaches such as iTRAQ(88) and stable isotopic


labelling with amino acids in cell culture (SILAC)(89). All of these have combined to

contribute to the search for protein biomarkers.

Several approaches can be taken to biomarker discovery utilizing proteomics.

The most common, as in genomic studies, is to utilize cancer and normal tissue or

biological fluids from patients and perform a proteomic analysis on each to

determine differences in the abundance of proteins or protein modifications. The

most commonly used sources for searching for biomarkers for prostate cancer has

been serum, prostate tissue and seminal fluid. However, one of the main issues with

proteomic analysis of biological fluids is the large dynamic range of proteins present.

Usually, the proteins of interest are low abundance tumour proteins and the

presence of high abundance serum proteins is seen to suppress their ionization thus

precluding them from detection. In an effort to simplify this complex mixture

debulking or chromatographic methods are used to remove the interference of high

abundance proteins.

A common approach that was initially employed was the use of surface

enhanced laser desorption ionization time of flight (SELDI-TOF) mass spectrometry

(MS) analysis of serum. It was thought to have low sample preparation, was able to

simplify the sample mixture prior to MS analysis and have a high throughput. The

technology relies on the use of chips with affinity-based chromatographic surfaces

where serum samples would be deposited and then washed off. The bound proteins

and peptides to the chip surface would then be subjected to MALDI-TOF analysis.

This would generate a collection of unknown peptide mass spectra from each

sample of which discriminatory algorithms would be used to determine spectral


peaks that corresponded to normal and cancerous states. This method was used as

an approach in the discovery of CaP biomarkers(90-92), and has since been

debunked as a method for biomarker discovery due to its poor reproducibility and

artifactual results(93).

The proteomic analysis of tissues also presents with similar hurdles as with

serum but to a lesser extent. Current approaches have been to use LCM to capture

pure populations of cancerous and normal cells and perform a comparative

proteomic analysis(94). In addition, the use of archived formalin fixed prostate tissue

biopsies have also been used for proteomic analysis(95). A simpler approach, albeit

in vitro, has been to use CaP cell lines and study their response to stimuli(96) or

their secreted proteins in culture(97,98).

To identify protein biomarkers that may be secreted by the tumour and end up

in the circulation, it is relevant to study the proximal fluids that are produced by the

prostate gland. Since, proteins present in this fluid would contain secreted proteins

from tumour glands as well as normal secretions; it is highly likely that these

secreted proteins would be detectable in the circulation through tissue leakage. Two

recent studies highlight this approach, each using prostatic fluid(99) or urine(100)

containing prostatic fluid from patients with and without CaP. Bioinformatics analysis

of the proteins identified resulted in selection of novel candidates for further

validation.

It has also been speculated that antibodies are generated to tumour antigens

and are then amplified and present in the circulation. A novel approach to identify

autoantibodies to tumour antigens utilizes protein microarrays that are spotted with


antibodies from the serum of patients with CaP. This method measures the humoral

response to tumour proteins. This approach was used to detect a response against

AMACR a well known CaP biomarker(101). Wang et al. demonstrated that they were

able to generate phage-display library of proteins derived from CaP tissue. Using

this they identified a 22-phage peptide detector for autoantibodies in CaP

patients(102). Thus, the clinical significance of autoantibodies in CaP has been

shown to be prominent.

1.2.3 Clinical and Analytical Properties of a Cancer Biomarker Assay

Clinical efficacy and analytical robustness are two of the main criteria used in

selecting tumour makers for use. A tumour marker should be specific for its

application be it screening, prediction or therapeutic response monitoring. Overall it

should exhibit high sensitivity and specificity with a high area under curve (AUC) in a

receiver operating characteristic curve (ROC). It should also be well defined

biochemically with the availability of detection reagents such as antibodies. The

analytical method for measuring the marker should be accurate and precise with no

cross-reactivity and should be automated for efficient execution.

When developing a test for a tumour marker in a biological fluid for example it

is important that the reproducibility of the assay first be optimized. Next the dynamic

range of the assay should be established such that analytical sensitivity and limit of

detection be determined along with the maximal value measurable that is accurate.

After a working assay is developed the clinical validation of the marker may be

assessed. It is important to establish reference values of the tumour marker in


healthy individuals to determine upper limits of the marker in the healthy population.

This will in turn help in deciding cut-offs for the tumour marker in determining clinical

diagnosis.

The clinical sensitivity and specificity of a tumour marker assay can be

defined as the percentage of people tested with the disease that the test correctly

finds positive and percentage of people with the disease that the test correctly finds

negative, respectively. These values are in essence indications of the true positive

and false positive rates of the test. Calculations for each are shown below:

Sensitivity = True Positives / (True Positives + False Negatives)

Specificity = True Negatives / (False Positives + True Negatives)

Based on these values a ROC curve may be drawn depicting the range of sensitivity

and specificity across a range of cut-off values. The sensitivity values are plotted on

the y-axis with 1-specifity on the x-axis. This aids in determining the trade off

between sensitivity and specificity and helps in choosing a cut-off value. It is

important to note the consequences of a false positive or false negative when

choosing the appropriate cut-off. The added value of a ROC curve is that multiple

tests can be compared on the same graph to determine which has the highest AUC

which translates the ability of the test to discriminate those with and without the

disease. These values range from 0.5 (a test that has a 50% probability of being

correct) and 1 (a test that has a 100% of being correct).


A measure of depicting how clinically relevant the tumour marker assay is of

detecting disease in the patient is shown in its predictive value. The positive

predictive value (PPV) of a test is the percentage of people testing positive for the

test that indeed have the disease. The negative predictive value (NPV) of a test is

the percentage of people testing negative for the test who indeed don’t have the

disease. The calculations of which are shown below:

PPV = True Positives / (True Positives + False Positives)

NPV = True Negatives / (False Negatives + True Negatives)

Taken together in combination with the prevalence of the disease will determine if

the test is suitable for population screening or in just high risk populations.

1.2.4 Applications of Cancer Biomarkers

There are several uses for tumour markers in aiding clinical management of

patients with cancer. Based on the properties of the marker and the test used to

measure it dictates which application the marker is suitable for. These include

population screening, diagnosis, prognosis, prediction of therapeutic response and

monitoring effectiveness of therapy, determining tumour recurrence or remission, or

being used as an imaging marker in localizing the tumour. Below we discuss the

requirements of the marker and assay for each application.


Screening for cancer in the population requires that a marker is elevated

when the disease is in its early stages and curable. With the exception of PSA, most

markers are elevated in late stage disease and have a low diagnostic sensitivity. It is

also beneficial if the marker is specific to a particular tissue since elevations in other

tissues due to other diseases or inflammatory conditions will reduce its specificity

and increase its false positives. It is also required that the test for the marker be non-

invasive and low in cost to increase compliance, e.g. a blood test. Screening must

also demonstrate that it delivers a clear benefit to the patient in reducing morbidity

and mortality associated with the disease.

A tumour marker used in diagnosis of cancer involves the same

considerations as a marker for screening. Most markers used have either low

diagnostic sensitivity or specificity. However, for patients in high-risk groups where

prevalence is relatively high, using a diagnostic marker will aid in deciding whether

more invasive tests are required.

Determining the prognosis of a tumour is essential in deciding the course of

therapy to be taken. Most tumour markers are correlated with cancer prognosis and

other prognostic indications such as tumour grade and staging. However, the

clinical decision to determine a therapeutic course of action can not be taken based

on the levels of the tumour marker alone. This is evident with PSA and its

derivatives.

Tumour markers may also be used to predict and monitor therapeutic

response. With a marker such as PSA, its high tissue specificity allows it to be used

to monitor a patient after radical prostatectomy and other therapies. This determines


if the therapy was effective and also determines if there is tumour recurrence as the

PSA level will be seen to increase again. Tumour markers have also found use as

targets for therapy. The most prominent marker being Her-2/neu in breast cancer as

a target for the antibody based therapy Herceptin. Tumours overexpressing Her-

2/neu are responsive and tumours lacking will not respond, thus this tailors the type

of therapy to the patient.

1.2.5 Current and Emerging Prostate Cancer Biomarkers

The current biomarker used for CaP diagnosis is PSA. It is considered both

the best tumour marker available for any cancer and as a marker with many

shortcomings. Originally, it was used for monitoring patients with CaP and was

subsequently implemented for screening.

Prostate Specific Antigen-Derived Forms:

It has become clear that the operating characteristics of PSA need to be

improved. One approach has been to measure PSA derivates that include PSA rate

change over time (PSA velocity, PSAV), ratio of PSA to prostate volume (PSA

density, PSAD), and age specific PSA ranges. In addition, novel improvements in

measuring PSA and PSA related proteins have allowed the measuring of percent

free PSA (%fPSA) which is a ratio of free to total PSA (tPSA); complexed PSA which

measures how much PSA in serum is bound to either α-2-macroglobulin (A2M), α1-

protease inhibitor (API) or α-1-antichymotrypsin (ACT); as well as different cleavage

isoforms of PSA.


Age-Specific PSA:

Age-specific PSA ranges adjust PSA thresholds based on age with the

assumption that older men have higher normal PSA levels due to prostate

enlargement since PSA correlates linearly with prostate size(103). Thus, a lower cut-

off for younger men should be employed to improve sensitivity, and a higher cut-off

in older men would reduce false positives, increasing specificity. Recently, a

retrospective study showed that men <50 years with a PSA of 0.51-1ug/L compared

to <0.5ug/L had a 2.5 fold greater risk of CaP up to 25 years later. In addition, men

with PSA of 2-3ug/L had a 19 times greater chance of CaP(104). Thus, this test can

serve to stratify patients based on risk and to monitor and tailor therapy. This test

still remains controversial as studies have shown reduction in sensitivity and lower

diagnostic accuracy compared to the 4.0ug/L cut-off(105,106).

PSA Density:

PSA density takes into account the volume of the peripheral and transitional

zone (TZ) of the prostate gland and is calculated by dividing the serum PSA level by

prostate size determined by TRUS. A higher PSAD has been shown in men with

CaP versus BPH(107,108), while others have not been able to reproduce these

results(109,110). In BPH the TZ is mostly enlarged and not a frequent site of

carcinoma. Taking into account TZ size specifically (PSAD-TZ) has shown

improvements in specificity but has not been reproducible(111-114). A PSAD ratio of

more than 0.15 is used to indicate an increased risk of CaP(108,115,116). This test

would increase specificity and sensitivity in large prostates in men with small

prostates respectively. While the use of PSAD is not in clinical use for screening due


the cumbersome need for TRUS, utility may be derived in selecting men for repeat

biopsy.

PSA Velocity:

PSA velocity is calculated by measuring serial PSA measurements over time

and determining the rate increase. It has been seen that BPH patients show a linear

increase in PSA, while CaP patients display exponential increases in PSA(117).

However, variations in PSA levels due to biological or analytical variability can result

in imprecision of this test(118,119). In addition, infection and chronic inflammation

are confounding factors that can increase PSA levels. A consensus on the

appropriate parameters to use has still yet to be determined. Carter et al.(120,121)

showed that at least three measurements over a period of at least two years are

required and a PSAV >0.75ug/L per year was significantly associated with CaP with

men with 4-10ug/L showing specificities of >90% PSA. In addition, a 2.0ug/L yearly

PSAV has been associated with a shorter time to PSA relapse and mortality

following surgery for localized CaP(122-124). Thus, this could identify patients who

may not benefit from localized therapy. Initial results from the European Randomized

Study of Screening for Prostate Cancer confirmed PSAV values differ significantly

between men with and without CaP and was limited in value in predicting biopsy

outcome(125).

Free PSA:

Measuring the ratio of fPSA to PSA (%fPSA) has shown to be beneficial by

improving the operating characteristics of PSA in the ranges of 4-10ug/L and <4ug/L

in several retrospective and prospective studies(126-129). Higher %fPSA correlates


with decreased risk of CaP and increased risk of BPH and this ratio is now being

used routinely since the mid 1990’s and has been approved by the USA Food and

Drug Administration in 1998. Sensitivity for %fPSA testing has been shown at 90-

95% resulting in reduction of unnecessary biopsies by 15-20%(126,130). However,

confounding variables such as prostate volume(131), stage and grade(132), total

PSA(133), PIN lesions(134), race, age and drug treatment history should be

considered(135). A recent study utilizing 10-12 core biopsy versus the sextant

biopsy method depicted %fPSA as having reduced diagnostic efficacy(136).

Conflicting studies(137-139) have resulted in a debate as to what the optimal %fPSA

threshold should be.

Complexed PSA:

Prostate specific antigen circulating in serum has been shown to be mostly

bound to protease inhibitors, with the majority (~75%) consisting of the PSA-ACT

complex(140-142). Measuring these forms of complexed PSA (cPSA) have shown to

improve specificity marginally in the PSA range of 2-10ug/L in a prospective trial as

well as in meta-analyses of CaP patients(143,144). Currently there are two assays

available: one is for PSA bound to ACT (PSA-ACT) and one is for cPSA which

measures PSA-ACT and PSA bound to API (PSA-API) but not bound to

A2M(144,145). This assay is of limited use since PSA-ACT increases(146) where as

PSA-API decreases in CaP patients(147). Currently, it is very difficult to measure

PSA-A2M since the levels are low and A2M sterically hinders access to PSA

epitopes(148). Complexed PSA however, in combination with tPSA (cPSA/tPSA)

results in comparable sensitivity and specificity as %fPSA(149-152). In addition,


there have been conflicting multi-center studies depicting either benefits of cPSA

compared to tPSA(143,144) or showing no advantage when used

independently(147,153). Thus, the context of the usefulness of cPSA is still debated.

Measurements of PSA-A2M(154) and PSA-API(155) complexes in serum have been

reported(156) and have both shown elevation in BPH patients vs. CaP, while PSA-

API has shown to be increased and an improvement over tPSA but not fPSA.

Cleavage Isoforms of PSA:

Prostate specific antigen is secreted from prostate epithelial cells as a pro-

enzyme with a seven amino acid activation peptide that is cleaved after secretion.

However, there are several different isoforms and cleavage products of PSA that

result and that have been assayed(157,158). The non-clipped ‘intact’ PSA (iPSA)

has been assayed and shown to be informative in discriminating CaP from

BPH(159,160). This free form of PSA is enzymatically inactive and the ratio with

fPSA has been shown to be higher in CaP patients. proPSA contains 244 amino

acids (-7)(161) compared to fPSA (237 amino acids) and can be differentially

cleaved to produce different forms termed (-7), (-5), (-4), (-2), (-1)(162). proPSA in

combination with fPSA has been shown to improve detection of CaP in ranges less

than 4ug/L(163) and in the 4-10ug/L range and associated with aggressive CaP’s

more so than tPSA, fPSA and cPSA alone(157,164,165). (-2)proPSA has shown to

be increased in serum of CaP patients compared to BPH and can discriminate CaP

from BPH in patients with elevated fPSA(166). The efficacy of the (-5, -7) proPSA

forms has been evaluated and has shown inadequate ability compared with %fPSA

to enhance detection of CaP(167,168). However, in combination with fPSA it was


correlated with increased Gleason grade and metastatic disease and was able to

improve specificity in PSA levels of 2-4ug/L(169).

Benign or BPH associated PSA (BPSA) is a cleaved sub-form of fPSA that

has been correlated with presence of benign prostate tissue in the TZ of the

prostate(170). Benign PSA is unable to discriminate between BPH and CaP in

serum independently(171), however it could serve as a marker for BPH when used

in conjunction with %fPSA and therapeutic monitoring of BPH patients(172). The

proPSA/BPSA ratio was measured in CaP patients with fPSA less than 15% and

attained a sensitivity and specificity of 90% and 46% respectively(173). BPSA is

elevated in patients with symptomatic BPH versus asymptomatic patients and

correlates with prostate enlargement more closely than tPSA and %fPSA(174). In

healthy men BPSA is almost undetectable and is not affected by age.

Molecular derivatives of PSA still need to be validated in larger sample sets to

determine their clinical utility for diagnosing BPH or CaP. Currently, there is still

debate on which combination of derivatives produces the optimal diagnostic and

prognostic characteristics. Thus, there is a great need for novel identification of

biomarkers to aid in increasing the specificity of CaP detection and prognostication.

Human Kallikrein-Related Peptidase 2:

Human kallikrein-related peptidase 2 (KLK2) is a secreted serine protease

from the same gene family as PSA and shares an 80% sequence homology. Like

PSA it’s highest expression is in the prostate, albeit it is expressed in other

tissues,(175) it is also androgen regulated in the prostate however its concentrations

in the circulation are 100 times less than PSA, albeit its mRNA levels are half that of


PSA. Unlike PSA, KLK2 differs in enzymatic activity and is mostly found in the free

unbound form. Prostate cancer tissue data has shown that KLK2 is relatively

elevated during CaP progression and thus may be a useful CaP

biomarker(176,177). Serum studies have displayed improvements in diagnosis of

CaP(178) in combination with tPSA(179) and fPSA(180-185) specifically with

respect to extracapsular extension and tumour volume(186,187). KLK2 also showed

improved independent prognostic information versus PSA for risk of biochemical

recurrence in men with PSA <10ug/L(188). Like PSA, KLK2 is also found in a bound

and free form(189), where the bound form has been shown to diagnose CaP with

higher specificity in complex with protease inhibitor 6(190). Additional validation

studies are required to elucidate the full prognostic potential of KLK2.

Other Tissue Kallikreins:

Up until recently PSA, KLK2 and pancreatic/renal kallikrein 1 (KLK1) were the

only genes identified in the kallikrein locus on chromosome 19. Now we know that

the locus spans 300kb and consists of 15 genes that share significant homology and

sequence similarity at the DNA and amino acid sequence(175,191). In addition to

KLK2, other kallikreins have shown utility as biomarkers for CaP and other

diseases(192-196). There are 8 kallikreins that are expressed relatively highly in

prostate tissue – KLK2-4, 10-13, and 15(175). Of these KLK11 shows promise as a

serum biomarker for CaP(197). In combination with tPSA and %fPSA showed

improvement in prediction of CaP(198,199).


Prostate Specific Membrane Antigen:

Prostate specific membrane antigen (PSMA), formally known as folate

hydrolase 1, is a membrane glycoprotein highly expressed in epithelial cells of

normal and CaP. It was first discovered from a monoclonal antibody derived from

mice immunized with the membrane fraction of the CaP cell line LNCaP(200). It was

found that there was increased relative expression of PSMA in epithelial cells of CaP

tissue; as well PSMA was shown to have folate hydrolase and carboxypeptidase

activity. The function of PSMA in the prostate is currently unknown; however, in the

brain it has shown to release glutamate from N-acetylaspartylglutamate. A

commercial imaging test for PSMA was developed (Prostascint, Cytogen Corp.) by

using the 7E11 antibody to PSMA conjugated to indium-111 and used for

radioimmunoscintigraphy. This approach was validated against computed

tomography, magnetic resonance imaging, and biopsy results and was shown to be

more sensitive and specific in detecting metastatic spread to the lymph nodes(201).

Currently, there has been conflicting data showing reproducible use of PSMA as a

serum marker for CaP(202-205). Detection of PSMA versus PSA via RT-PCR for the

identification of circulating tumour cells did not show a strong correlation with

CaP(206). Finally, PSMA has been studied as a target for therapy by using

antibodies conjugated to radioisotopes or toxins, or the activation of dendritic cells

against PSMA(207-211). The use of PSMA has not been adopted into clinical

practice as yet and its role as a diagnostic and therapeutic tool is still evolving.


Prostate Cancer Antigen 3:

Also known as DD3, prostate cancer antigen 3 (PCA3) is a non-coding RNA

that is expressed almost exclusively in the prostate and shown to be highly

overexpressed in CaP tissue, including metastasis, versus BPH(212). Several

assays can measure PCA3 mRNA in urine sediment. The only commercially

available test is APTIMA® (Gen-Probe Inc., San Diego, CA) which utilizes

transcription-mediated amplification(213). A PCA3 score is derived from PCA3

mRNA levels normalized to PSA. A recent large multi-institutional study of patients

undergoing biopsy included PCA3 analysis in voided urine after prostatic massage;

an AUC of 0.66 versus 0.57 for PSA using a PCA3 score cut-off of 58 was

obtained(214). In a study of 233 patients undergoing repeat biopsy after a negative

biopsy, PCA3 had an AUC of 0.68 and a sensitivity of 58% and specificity of

72%(215). This test could potentially be useful in improving the specificity of PSA. A

combination of PCA3 and another three urinary biomarkers, GOLPH2, SPINK1 and

TMPRSS2:ERG fusion, improved the sensitivity and specificity of PCA3 alone.(216)

Early Prostate Cancer Antigens:

Changes in nuclear matrix proteins have been shown to be associated with

carcinogenesis. Early prostate cancer antigen (EPCA) is a nuclear matrix protein

that was initially detected by proteomic profiling of rat prostate tissue(217). It has

since then been shown promise as a diagnostic marker for CaP.

Immunohistochemical studies by auto-antibodies to EPCA in biopsies of CaP tissue

showed increased staining relative to non-cancerous specimens(218). A field effect

was also seen in non-cancerous adjacent areas to tumour tissue, in 86% of CaP


tissue and aids in identifying patients at risk who have a negative biopsy. A blood-

based assay for EPCA was developed showing a sensitivity of 92% with 94%

specificity in a small cohort of 12 CaP and 34 healthy patients(219). Another study

measuring the EPCA-2 protein in serum showed a 92% specificity and 94%

sensitivity for identifying CaP and was able to differentiate localized from metastatic

CaP with an AUC of 0.89(220). Larger and other independent studies are awaited to

confirm this promising data. However, methodological deficiencies were identified

with such markers, casting doubt on their actual validity(221).

Αlpha-Methylacyl-CoA Racemase:

Alpha-methylacyl CoA racemase (AMACR) is an enzyme involved in the

oxidative metabolism and synthesis of branched chain fatty acids found in dairy

products and red meat(222). As well as being strongly expressed in CaP tissue, it is

located in a gene region (5p13.3) that contains polymorphisms that are associated

with CaP(223). A meta-analysis of microarray data showed AMACR to be up-

regulated in CaP with high confidence(78). A multi-institutional study of

immunohistochemical staining of AMACR helped in distinguishing benign from

cancerous prostate tissue with a 97% sensitivity and 92% specificity(224). Recently,

decreased AMACR expression was also shown to have prognostic value in

predicting biochemical recurrence and death due to CaP(225). Circulating levels of

AMACR mRNA have been measured by RT-PCR in serum and urine(226). Protein

levels are low in serum, however, AMACR protein has been detected by Western

blot in urine(227). Increased levels of autoantibodies to AMACR were able to

discriminate CaP patients from healthy subjects, in the PSA range 4-10ug/L, and


showed a sensitivity of 62% and specificity of 72%(101). Additional studies are

underway to fully elucidate the potential of AMACR as a biomarker for CaP.

Urokinase Plasminogen Activator and Receptor:

The degradation of the extracellular matrix has been associated with cancer

progression and the urokinase-plasminogen activation (uPA) cascade has been

shown to participate in this. Plasminogen is converted to its active form plasmin

through the activation of the serine protease uPA, binding to its receptor uPAR. One

study demonstrated increased levels in BPH and CaP compared to normal, albeit

there was no statistically significant association with CaP(228). Detection of isoforms

of uPAR, in combination with PSA isoforms and KLK2 showed improved prediction

of biopsy outcome in patients with elevated PSA in both univariate and multivariate

models(229). Increased tissue levels in CaP were associated with osteoblastic

metastases(230), as well as advanced CaP progression(231). Recently a study has

shown increased serum levels of uPA and uPAR in patients with metastatic bone

CaP(232). These studies depicted preoperative plasma uPA as a predictor of

biochemical recurrence and metastatic disease indicating the presence of distant

disease at time of localized therapy. Large prospective studies are needed to

elucidate the full prognostic potential of uPA and uPAR for preoperative models of

disease progression and metastases.

Insulin-Like Growth Factors and Binding Proteins:

Serum levels of insulin-like growth factors (IGF) and their binding proteins

(IGFBP) have been found to be associated with CaP. The IGF family consists of two

ligands (IGF-1, IGF-2), two receptors (IGFR-1, IGFR-2) and six binding proteins


(IGFBP-1 to 6). Increased levels of IGF-1 and decreased levels of IGFBP-3 have

been correlated with increased risk of developing CaP(233). Another prospective

study found that IGF-1 levels were increased slightly with CaP risk but did not

outperform PSA(234). However, others have failed to reproduce these results and

found no association with CaP progression. In addition, the main IGFBP produced

by the prostate is IGFBP-2 and was shown to be increased in CaP. Albeit, levels in

localized tumours were correlated inversely to tumour size and indicators of CaP

progression(235). Serum levels of IGFBP-3 were shown to be inversely correlated

with presence of metastases to the bone, however did not show any difference to

localized CaP and normals(235).

TMPRSS2:ERG/ETV Gene Fusion:

Gene rearrangements have been implicated in cancers, in particular

haematological malignancies. One such rearrangement between the transcription

factor genes ERG (21q22.2) and ETV1 (7p21.1) with membrane-anchored serine

protease TMPRSS2 (21q22.3) was shown to occur in 80% of CaP by cancer outlier

profile analysis(80). The fusion product of these genes tested in urine has been

shown in 42% of CaP patients and in 20% of patients with PIN and rarely in

BPH(236). A prospective study which followed 252 men with stage T1a/b CaP for 9

years showed that the TMPRSS2:ERG fusion was associated more than the

TMPRSS2:ETV fusion in Gleason sums>7, metastatic disease and death due to

CaP(237). An isoform of the TMPRSS2:ERG fusion has been shown by

fluorescence in situ hybridization (FISH) analysis to be present in 80-95% of CaP

tissue, and could be a potential target for therapy.


Transforming Growth Factor Beta:

Transforming growth factor β1 (TGF-β1) is a widely acting growth factor

involved in a variety of molecular processes such as cellular differentiation, immune

response, angiogenesis and proliferation. Studies with model systems of CaP have

shown a role of TGF-β1 in CaP progression(238). Increased levels of TGF-β1 in CaP

tissue have been correlated with tumour grade and stage and lymph node

metastasis(239). An ELISA for TGF-β1 was used to measure pre-operative plasma

levels, which have shown to be increased in CaP patients(240) and correlate with

extracapsular extension, seminal vesicle invasion, metastasis(241) and biochemical

recurrence(242). Thus, TGF-β1 could prove useful as a prognostic marker for CaP.

Enhancer of Zeste Homolog Gene 2:

Enhancer of zeste homolog gene 2 (EZH2) is part of the polycomb family of

proteins involved with regulation of gene expression. It has been shown that EZH2 is

expressed more in metastatic CaP compared to localized CaP and BPH through

gene expression profiling of CaP tissue from autopsies of men who died from

metastatic CaP(82). In addition, it was found to outperform preoperative PSA and

Gleason score for determination of CaP progression. It was also shown in

combination with E-cadherin to predict CaP recurrence after localized therapy(243).

Development of a serum assay would aid in the validation of this candidate

biomarker to identify patients at risk of developing metastatic disease.

Glutathione S-Transferase Pi hypermethylation:

Hypermethylation of tumour suppressor genes at their promoter regions at

cytosine/guanine (CpG) nucleotide islands have been implicated in CaP. Glutathione


S-transferase pi (GSTP1) is an enzyme that protects DNA from free radical damage.

Reduced expression of the GSTP1 gene due to hypermethylation of the promoter

has been consistently shown in CaP and has been studied in urine sediment to

determine the need for biopsy(244). This assay has been improved through prostatic

massage before collecting urine(245). Panels of genes, including GSTP1, have been

studied in a similar manner(246).

Prostate Secretory Protein 94 and Binding Protein:

Prostate secretory protein 94 (PSP94), also known as h-microseminoprotein,

is a highly abundant protein in semen and plays a role in regulation of cell

proliferation and apoptosis. Bound forms of PSP 94 exist that are complexed with

PSP94-binding protein (PSPBP). Serum levels of PSP94 / free PSP94 and PSPBP

in CaP patients after local surgery were associated with Gleason sum, biochemical

recurrence and surgical margin status(247). Large prospective studies are still

required to validate this candidate biomarker.

Markers for Neuroendocrine Differentiation:

Chromogranin A is a peptide produced by the neuroendocrine cells in the

prostate and is currently used for diagnosis and prognosis of CaP tumours that show

neuroendocrine differentiation. Increased chromogranin A levels in serum have

been correlated with androgen independent CaP progression and poor

prognosis(23), and have been shown to precede PSA elevation(248), and improve

specificity in combination with fPSA(249).

Progastrin-releasing peptide (proGRP) is a growth factor found to be released

in the neuroendocrine type of CaP. Increased levels were detected in metastatic


CaP and was associated with its progression(24). It was also shown to be predictive

of the androgen-independent phenotype(250). Thus, both chromogranin A and

proGRP may be used to monitor patients with late stage disease for hormone-

refractory CaP in a subset of patients that display neuroendocrine differentiation.

E-cadherin:

Cell-cell adhesion plays an important role in normal tissue architecture and

carcinogenesis. E-cadherin is a cell adhesion molecule expressed in epithelial cells

and whose expression has been shown to predict CaP prognosis. An

immunohistochemical study showed reduced expression of E-cadherin in 50% of

CaP tumours, where normal prostate tissue showed uniform expression(251). This

was then further studied and correlated with grade, tumour stage, and survival. A

lower immunohistochemical expression of E-cadherin was associated with CaP

patients with a shorter survival(252).

Annexin A3:

Annexin A3 (ANXA3) is a calcium-binding protein that is part of the annexin

family of proteins. ANXA3 has been shown to be involved with the activation of the

immune response as well as membrane trafficking and lymphocyte

migration(253,254). Recently, ANXA3 has been studied as a promising prognostic

tissue marker for CaP and was shown to have lower expression in CaP versus BPH

and PIN and normal tissues by immunohistochemistry(255,256). ANXA3 was able to

stratify a large group of intermediate-risk patients in to high and low risk subgroups.


Prostate Stem Cell Antigen:

Prostate stem cell antigen (PSCA) is a membrane glycoprotein that shows

fairly specific expression to the prostate. PSCA was detected by

immunohistochemistry in CaP tissues and its RNA was found in blood samples(257).

Increased expression of PSCA was correlated with increased risk of CaP, Gleason

score, stage and presence of metastasis(258). PSCA was also investigated as a

target for therapy(259). However, larger validation studies are required to confirm its

clinical utility.

Hepsin:

Hepsin is a membrane serine protease that is highly expressed in prostate

tissue and was first identified in the human liver from cDNA libraries(260).

Overexpression of the hepsin gene has been shown in 90% of CaP tumours through

mRNA expression profiling studies(261). Immunohistochemical staining of hepsin in

one study showed it to be highly expressed in PIN lesions of the prostate, and

preferentially expressed in CaP versus BPH(262). Further studies in serum and

urine are required to fully elucidate its diagnostic potential.

Interleukin-6 Ligand and Receptor:

Interleukin 6 (IL-6) is a cytokine secreted by a variety of cell types and

involved in the immune and acute-phase responses. Elevated levels of IL-6 and its

receptor IL-6R have been shown in metastatic and AIPC(263), and has thus far have

been suggested as a candidate markers for CaP progression(264). Studies with IL-6

in combination with TGF-β1 have also shown promise in CaP diagnosis(265).


Circulating Tumour-associated DNA:

Dissemination of tumour cells is a prerequisite to metastasis; thus early

detection of these cells in the circulation can be useful in prognosis of CaP patients.

This has been shown by RT-PCR which has proven to be sensitive and used to

increase the accuracy of staging and prediction of disease recurrence by using

markers specific to the prostate(257,266).

Autoantibodies:

It is known that the immune system elicits an autoantibody response to some

antigens overexpressed by tumours. Humoral responses to huntingtin-interaction

protein 1, prostasomes and AMACR have been reported(101,267,268). By using

phage display and protein microarrays, Wang et al.(102) were able to identify

autoantibodies to peptides derived from CaP tissue in a new approach termed

“cancer immunomics”. They were able to generate a 22 phage peptide array that

was able to discriminate 68 CaP serum samples from 60 controls with 88.2%

specificity and 81.6% sensitivity with an AUC of 0.93, which was superior to PSA

(AUC 0.8). Studies are underway to further validate this detection tool in a larger

cohort. A recent study by the same group applied a similar approach, followed by

biological network analysis to determine deregulated pathways in CaP

progression(269). One concern is that the needle biopsies themselves may be

eliciting an autoimmune response.

Nomograms:

Nomograms are multi-parametric tools that combine clinical features such as

tumour grade/stage and biomarkers to provide physicians with standardized patient


care. They utilize evidence-based approaches for decisions regarding treatment at

each stage of disease management. The value of a nomogram is derived from its

performance characteristics and user-friendliness. There are numerous nomograms

that have been developed for CaP including a TGF-β1 and IL-6 standard nomogram

for biochemical recurrence(270), as well as nomograms predicting outcome of

biopsy(271). A review by Karakiewicz and Hutterer summarizes nomograms that

have been developed for CaP(272).

Multi-Parametric Tests / Artificial Neural Networks:

The use of combined biomarkers for improvement of disease prediction has

been used widely for many different disease states. There is heterogeneity among

individuals and as a result, disease states within these individuals differ in their

biology. Thus, the use of multi-parametric tests will most likely be more applicable

for population screening versus one marker. Recently, Parekh et al.(273) used a

biomarker panel consisting of 54 proteins that included adipokines,

metalloproteinases, adhesion molecules and growth factors. They used age-

matched controls and measured pre-diagnostic serum concentrations of patients

later diagnosed with CaP. Their results did not prove that the marker panel was able

to outperform the risk factors from the The Prostate Cancer Prevention Trial (PCPT)

calculator. Artificial neural networks (ANN) have been used to model complex

relationships between variables and find patterns in data. Stephan et al. have used

the ANN approach to assess various combinations of kallikrein biomarkers to

determine their clinical utility for CaP diagnosis(274).


Proteomic Patterns:

High throughput proteomic analysis of biological fluids, tissues and cell lines

has recently become a popular approach for the identification of novel biomarkers. In

particular, the application of SELDI-TOF has been used frequently to profile

biological samples. With respect to CaP, Adam et al.(91) used a decision tree

algorithm to ascertain a peak fingerprint that could discriminate CaP from normal

individuals with a sensitivity and specificity of 83% and 97%, respectively. Petricoin

et al.(275) used 266 serum samples from CaP patients and controls to achieve 95%

sensitivity and 78% specificity. Qu et al.(92) used a boosted decision tree algorithm

for analysis of their SELDI-TOF data and were able to achieve 97% sensitivity and

specificity. Other studies have also used proteomic profiling for CaP diagnosis

showing usefulness of the approach(276). However, the use of proteomic pattern

fingerprinting has come under scrutiny and as a result the NCI/EDRN has conducted

a multi-institutional study to objectively validate this approach and the results were

recently published(93). Even though stage 1 of the validation confirmed the

analytical reproducibility of the approach, stage 2 was unable to determine if the

approach could predict CaP in a case-control series across institutions. The cause of

this failure has been attributed to pre-analytical, analytical and bioinformatics biases,

as described in previous literature(277,278).


Table 1.1: List of candidate biomarkers for prostate cancer and their possible clinical utility.

Candidate CaP Biomarker Assessed Clinical Utility References

KLK2 Diagnostic and prognostic predictor of extracapsular extension, tumour volume and biochemical recurrence (186,188)

KLK11 Early predictor of CaP in serum (199) PSMA Imaging marker and target for therapy (201,209,210)

PSP94 Predictor of Gleason sum, surgical margin status and biochemical recurrence after local surgery (247)

PSCA Immunohistochemical marker associated with Gleason sum and stage. Target for therapy. (258,259)

PCA3 Urinary biomarker for detection of CaP (214,215)

EPCA/EPCA2 Immunohistochemical detection of CaP, serum marker to differentiate local from metastatic CaP (218-220)

Hepsin Immunohistochemical detection in PIN, CaP vs. BPH (262)

AMACR Increased detection of autoantibodies in CaP,

immunohistochemical expression as a prognostic factor for biochemical recurrence and death

(101,224,225)

EZH2 CaP tissue gene expression predicts progression (82)

uPA/uPAR Increased tissue and serum levels predicts biochemical recurrence and metastasis (230,232)

IGF/IGFBP IGF-1 slightly increased in CaP serum, IGFBP’s inversely correlated to CaP progression (234,235)

TGF-β1 Increased immunohistochemical and serum levels with

CaP progression and biochemical recurrence (239-241)

TMPRSS2:ERG/ETV Increased detection in urine of CaP, PIN patients vs. BPH, gene fusion present in CaP tissue by FISH (236,237)

GSTP1 Hypermethylation of promoter detected in urine to assess for biopsy (244)

IL-6 Elevated serum levels in late stage CaP (263,264)

Chromogranin A Monitoring of patients with androgen independent late stage CaP with neuroendocrine differentiation (23,248,249)

Annexin A3 Decreased expression in tissues of CaP by immunohistochemistry, prognostic risk marker (255,256)

Progastrin-releasing peptide

Monitoring of patients with metastatic CaP with neuroendocrine and androgen-independent phenotype (24,250)

E-cadherin Reduced immunohistochemical expression in CaP correlated with stage and reduced survival (251,252)


1.3 Prostate Cancer Model Systems

The utilization of mouse and cell culture based model systems for CaP has

allowed the portrayal of this disease phenotype and has enabled the understanding

of the cancer biology and aided in developing novel therapies. Here we discuss both

mouse model and cell culture based approaches to CaP model systems.

1.3.1 Mouse Models

Mouse models have been shown to be immensely valuable in comprehending

CaP cancer biology. Knockouts of genes in a prostate specific fashion have

elucidated the associations with CaP progression. One of the most commonly used

mouse models for CaP is the transgenic adenocarcinoma of the mouse prostate

(TRAMP) mouse in which the Simian virus SV40 large and small T antigens are

expressed under the regulation of a rat probasin promoter(279). In the TRAMP

mouse, CaP pathogenesis develops within 12 weeks of birth with initial PIN lesions

being seen and metastatic spread occurring at 30 months. Another model also

utilizing the SV40 large t-antigen is the LADY mouse model. This is an early model

of CaP progression with PIN lesions developing at 20 months and no metastatic

spread observed(280). Of use have also been models using c-myc(281) and a

heterozygous PTEN mouse model(282). The PTEN model shows that loss of PTEN

recapitulates features of human CaP progression and mice also lacking p27 present

with more aggressive tumours with earlier onset, albeit without metastasis. A rat

model of CaP, known as the Dunning rat model, was one of the first to be developed


and used widely(283). It has shown utility for the study of AIPC growth of CaP cells

and the molecular determinants of metastases.

Due to the in vivo nature of the mouse model the interplay between host

responses and the tissue microenvironment can result in a representative phenotype

that recapitulates the morphology in humans. However, due the complex nature of

these interactions, elucidating the mechanisms involved become more of a

challenge, added to this is the maintenance of a colony of mice. A canine model for

PIN development has also been used, albeit their long life span makes them

unsuitable for many studies(284). With respect to the identification of circulating

protein biomarkers for CaP in animal models, similar issues are present as with

analyzing human specimens. In this case a simplified cell culture based approach

may be warranted to study CaP cells in a homogenous environment.

1.3.2 Cell Culture Models

The use of cell lines derived from prostate tumours allows the analysis of

prostate tumour cells in a simplified and controlled environment. There have been

numerous cell lines derived from prostate tumours, albeit most have been from

metastatic sites(73). Prominent cell lines that are commonly used include the PC3,

LNCaP, DU-145, 22Rv1 and MDA PCa 2b. All are obtained from metastatic sites

except for 22Rv1 which was obtained from a localized tumour to the prostate.

Normal epithelial phenotype cells have been developed such as the RWPE-1 cell

line series, albeit their sustained growth in culture draws question to the extent of the

normal phenotype.


There are several advantages and disadvantages to using cancer cell lines

over animal models. These then dictate the nature of the experiment that can be

conducted. Firstly, the cost involved with maintaining them is significantly less than

maintaining mice. They are readily available and studies can be performed relatively

quickly. Large quantities and volumes of cells may be propagated to create high-

throughput studies. Cell lines are extremely versatile in the types of studies they may

be used in. Not only can they be grown in vitro but also can be injected into mice to

form xenograft models of CaP progression. They can be modified and studied over

time to determine sequential events that occur as a result of specific stimulus. As

well as the products produced from the cells such as their ‘secretome’ can be

analyzed readily. Disadvantages that are associated with cell lines are that they do

not represent the heterogeneity of the tumour microenvironment as well as the

necessarily heterogeneous nature of tumours with a patient and between patients.

As a result multiple cell lines may be required to address the full heterogeneity seen

in a tumour phenotype. Cell lines are also subject to genetic alterations in culture

that may alter their phenotype over the course of a long experiment. The path to the

progression of the tumour is lost and does not provide insight in the pathogenic

process necessarily.

Developing CaP cell lines has been regarded as being notoriously difficult.

As a result the CaP cell lines that are utilized today have been around for quite some

time and have been characterized very well. Each cell has shown to have its own

unique phenotype which has become useful in comparing the proteins secreted by

each in the search for CaP biomarkers(285).


1.4 Continued Need for Novel Prostate Cancer Biomarkers

The discovery of PSA and its introduction in the clinic in the early 1990’s has

had a profound impact on the early diagnosis of CaP and resulted in an increase in

CaP incidence(286). PSA is currently used as a dichotomous marker for diagnosis

but it is now being realized that its values represent a relative degree of risk for

CaP(287). The upper limit of normal set at 4ug/L fails to detect a significant number

of cancers and the PCPT determined that there is no level of PSA where cancer can

be ruled out(288). Measurement of total PSA has been shown to be useful as a

prognostic tool, with high preoperative values associated with advanced disease and

poor clinical outcome(104). The controversy surrounding the use of this marker is

being currently debated since it is not clear if PSA screening has led to a decline in

mortality due to CaP(289). In 2008 and 2009 two major randomized prospective

clinical trials will report if PSA screening reduces mortality: the European

Randomized study of Screening for Prostate Cancer and Prostate, Lung, Colorectal

and Ovarian Cancer Screening Trial. The relationship of PSA with tumour grade is

also not clear. It has been shown that tissue PSA decreases with increasing

Gleason sum(54), albeit serum levels increase due to disruption of the basement

membrane surrounding the prostate epithelial cells and the overall prostate tissue

architecture. PSA is not specific for prostate cancer and can serve as a marker for

benign prostate hyperplasia and prostate volume growth(290). Key statistics for the

test have been shown to be inadequate, with positive predictive values of 37% and

patients in the grey zone of 4-10ug/L having 25% chance of displaying CaP(291)

and 15% of men with levels of <4ug/L displaying CaP(287). Due to the inadequacies


of PSA there is a need for novel markers of CaP to prevent over treatment of

indolent tumours. In addition to diagnosis, prognostic, predictive and therapeutic

markers are needed to act as surrogate endpoints in forecasting disease severity,

choosing treatments and monitoring response to therapies, respectively.

1.5 Rationale and Objectives of the Present Study

1.5.1 Rationale

Advancements in sequencing technologies for genomics have led to the

completion of many genomes including the human(292). Coupled with advances in

mass spectrometry, particularly soft ionization techniques for MS such as ESI and

MALDI, has of recent made a large impact in the field of biomarker

discovery(293,294). These new advances have enabled the generation of large

databases housing theoretical protein sequences of all the genes in an organism.

With these advancements, the identification of complex mixtures of proteins by MS

has become routine. In our global analysis of proteins present in the CM of CaP cell

lines we will utilize this technology in our search for novel prostate tumour markers.

We have chosen a cell line based model for identifying secreted proteins for

our discovery approach. By collecting and concentrating CM produced from cell

lines, the proteins secreted from the cells will accumulate in the CM and be present

at higher levels than those found in tissues, thereby making their identification

through MS more facile. In addition, given that the cell lines to be used are specific

to epithelial metastatic CaP cells, the proteins present in the CM will originate from

the cancer cell specifically and not from the the stroma environment, thereby


avoiding unnecessary complications in our analysis. This will eliminate any

extraneous proteins identified that are not specific to the tumour. To select our

candidates, our focus will be on secreted or shed membrane proteins, since in the

past they have proven to be the most useful as biomarkers(295). In addition, using

prostate cell lines derived from clinically relevant tumours will aid in our discovery of

tumour markers that are specific to clinically relevant tumours.

Through qualitative analysis of the PC3, LNCaP and 22Rv1 malignant

prostate cell lines, proteins identified will be investigated for their potential as

candidate tumour markers. We assume that these molecules have not yet been

identified because their concentration in serum is too low and therefore cannot be

measured or purified, unless specific immunological reagents and highly sensitive

enzyme linked immunosorbent assay (ELISA) methods are available.

To facilitate our analysis of the CM, we will use protein and peptide free

chemically defined serum-free media (SFM) in our cell culture system. This will

provide us with a distinct advantage over fetal calf serum (FCS) containing media,

since all the proteins secreted into the media will be from the cell line and can thus

be easily identified without worry of contaminating proteins from the FCS.

Conditioned cell culture media is also significantly easier to characterize compared

to human serum or plasma which contains many substances that interfere with its

proteomic analysis such as lipids. In addition, there are many high abundance

proteins in human serum that will mask the detection by MS of lower abundant

proteins. There are also concerns regarding sample acquisition and handling that


makes blood samples an imperfect source when performing biomarker discovery

studies(296).

1.5.2 Hypothesis

Proteins produced by tumour cells are secreted or shed into the circulation

and may act as surrogate tumour markers. These tumour markers can be used to

aid in the diagnosis and prognosis of CaP patients. We hypothesize that candidate

protein tumour markers for the early detection of clinically significant CaP are

secreted in vitro by CaP cell lines into their tissue culture media. These proteins can

subsequently be identified through qualitative proteomic analysis of the CM by MS

based proteomics. Candidate tumour markers identified will then be selected to

validate their clinical utility in serum and tissues of patients with and without CaP.

1.5.3 Objectives

1) Perform a proof of principle study to demonstrate the proteomic analysis

of the secretome of the PC3(AR)6 cell line by mass spectrometry

a) Optimize the culture of PC3(AR)6 in large volumes of SFM for an

extended period of time by measuring control proteins KLK5, KLK6

and total protein

b) Collect and prepare the CM for fractionation by fast performance

liquid chromatography

c) Process the fractions for liquid chromatography tandem mass

spectrometry


d) Analyze the identified proteins and select candidates for further

validation through bioinformatics and literature searches

e) Validate candidates in serum of patients with and without CaP

2) Conduct and extended proteomic analysis of the CM of three CaP cell

lines and a control flask

a) Optimize culture of each cell line in SFM to optimize secreted

protein production and reduce release of intracellular proteins

b) Collect and prepare CM for peptide fractionation through high

performance liquid chromatography

c) Process samples for analysis by liquid chromatography tandem

mass spectrometry

d) Analyze the identified proteins and determine protein identification

probabilities and false positive rates

e) Subject the list of proteins to bioinformatics and literature searches

and select novel candidates for validation

3) Perform a pre-clinical validation on selected candidates in serum of

patients with and without CaP

4) Conduct and extended validation on the most promising candidate in the

serum and tissues of patients with and without CaP

Chapter 2: Proof of Principle: Proteomic Analysis of the PC3 Cell Line Conditioned Media

59

CHAPTER 2:

PROOF OF PRINCIPLE: PROTEOMIC ANALYSIS OF THE PC3 CELL LINE CONDITIONED MEDIA

The work presented in this chapter was published in Clinical Chemistry:

Sardana, G., Marshall J. and Diamandis, E.P. Discovery of Candidate Tumor Markers for Prostate Cancer via Proteomic Analysis of Cell

Culture–Conditioned Medium Clinical Chemistry, 2007 Mar;53(3):429-37

Copyright permission has been granted.


60

2.1 Introduction

Prostate cancer is the most common malignancy in men and the 2nd leading

cause of cancer-related deaths (297). Early diagnosis of cancer improves clinical

outcomes, but detection methods for clinically relevant preclinical CaP are limited

(298). Tumour markers can be used to detect cancer, determine prognosis, or

monitor treatment (299). The established CaP tumour marker PSA, has low

diagnostic specificity (300); increased concentrations are also seen in BPH and

prostatitis (301). The sensitivity and specificity of the PSA test for CaP have been

improved by modifications such as measuring PSAV and measuring the ratio of

fPSA to tPSA or KLK2 in addition to PSA (302-304). Other methods for detecting

CaP are not highly specific for CaP and are uncomfortable to patients (305). Serum

proteomic profiling has emerged as a new method for detecting CaP (306,307) but

has not been adequately validated (308,309).

Determining the clinical significance of a prostate tumour is a major concern

of CaP testing (310). An autopsy study of men who died of other causes revealed

CaP or precursor lesions in 29% of men 30 to 40 years old and 64% of those >60

years old (311). Because treatments for CaP (androgen ablation, radical

prostatectomy, radiation, and chemotherapy) have serious side effects, there is a

need to differentiate patients who require treatment from those who do not. Mass

spectrometry for biomarker discovery (312-314) has generated large databases of

protein sequences. We used a cell culture–based proteomic approach to search for

novel candidate prostate tumour markers in the proteins secreted into the CM of the

CaP cell line PC3(AR)6.


61

2.2 Materials and Methods

2.2.1 Roller Bottle Cell Culture

We grew the CaP epithelial cell line PCR(AR)6, kindly provided by Dr.

Theodore Brown (Toronto, Ontario, Canada), in a humidified incubator at 37°C and

5% CO2 in RPMI 1640 (Gibco) with 80 mL/L fetal calf serum (Hyclone) to confluence

(20 X 106 cells/flask) in two 175-cm2 tissue culture flasks (Nunc). The cells were

trypsinized and transferred to an 850-cm2 roller bottle flask with a vented cap

(Corning), placed on a roller culture apparatus (Wheaton Science Products), and

incubated for 2 days in 150 mL RPMI with 8% FCS to allow the cells to adhere

(Figure 2.1). Afterwards, the medium was discarded and the interior was rinsed

twice with 150 mL phosphate-buffered saline (PBS) (137 mmol/L NaCl, 10 mmol/L

phosphate, 2.7 mmol/L KCl, pH 7.4). Next, 400 mL of chemically defined Chinese

hamster ovary medium (CDCHO) (Gibco), supplemented with glutamine (8 mmol/L)

(Gibco), was added, and the roller bottle was incubated for 14 days. During the

culture period, we measured total protein by the Coomassie (Bradford) assay

(Pierce Biotechnology) and KLK5 and KLK6 by ELISA (192,315). The CM was

collected, spun down (3,000g) to remove cellular debris, and frozen at –20°C for

later use. We processed 2 replicates for MS analysis; these replicate cultures are

referred to as batch 1 and batch 2.


62

Figure 2.1

Figure 2.1: Schematic representation of the workflow for proteomic analysis of roller bottle CM. CM from roller bottles was collected and dialyzed overnight. The dialyzed medium was directly loaded onto a SAX column and eluted by FPLC. Ten fractions were collected, lyophilized, and trypsin-digested. The resulting peptides were ZipTip desalted and separated by reversed-phase C-18 chromatography coupled online to an ion-trap mass spectrometer. The acquired MS/MS data were searched by Mascot, and identified proteins were manually categorized by Genome Ontology and literature searches through NCBI.


63

2.2.2 Dialysis

The thawed CM was dialyzed in tubing with a molecular weight cut-off of 3.5

kDa (Spectra/Por) in 10 L of 20 mmol/L diethanolamine (DiEtOH) (Sigma-Aldrich),

pH 8.9, overnight at 4°C. A sample aliquot was taken after dialysis.

2.2.3 Fast Performance Liquid Chromatography of CM

The dialyzed CM was loaded onto an HR10/10 column (GE Amersham)

containing SOURCE15 strong anion exchange (SAX) beads (GE Amersham). An

AKTA fast-performance liquid chromatography system (FPLC) was used running

Unicorn v4.12 software equipped with a P-960 sample pump and Frac-900 fraction

collector (GE Amersham), at a flow rate of 1 mL/min followed by a 2-stage elution

gradient at a flow rate of 3 mL/min (0% to 60% elution buffer within 40 min followed

by a ramp from 60% to 100% within 10 min) using 20 mmol/L DiEtOH, pH 8.9

running buffer, and 1 mol/L NaCl elution buffer. Absorbance was monitored at 214

nm. The first 10 fractions of 10mL each were collected, taking sample aliquots from

each, and the flow-through. A SAX protein standard (Bio-Rad) was run each time to

evaluate the quality of the column before each sample loading.

2.2.4 Lyophilization and Digestion of Fractions

The collected fractions were lyophilized overnight to dryness, resuspended in

1 mL dH2O, and trypsin-digested using a 10X digest buffer [5% acetonitrile (ACN),

200 mmol/L urea, and 50 mmol/L tricine, pH 8.8] to digest ~100μg protein from each

lyophilized fraction; 1μg trypsin (Promega) was used per digest. The digests were


64

incubated overnight at 37°C and reduced the following day with dithiothreitol (DTT)

(1 mmol/L) for 1 h at 25°C. We added a final 1μg trypsin, and the samples were

incubated at 37°C for approximately 3 h.

2.2.5 Liquid Chromatography–MS Analysis

Digested samples were collected on a C-18 ZipTip (Millipore) to purify and

desalt the peptides. The ZipTip was primed with 50% ACN in 0.1% acetic acid and

washed with 0.1% acetic acid before the digested samples were passed through the

ZipTip. The peptides were eluted from the ZipTip with 2μL of 0.1% acetic acid in

65% ACN, and dH2O was added to give a final volume of 20μL. The desalted

peptides were injected at 2μL/min onto a C-18 reversed-phase (RP) chromatography

column (Vydak 300μm X 15 cm) via an Agilent 1100 series HPLC system coupled to

a Bruker HCT ion-trap ESI mass spectrometer (Bruker Daltronics) via a metal

electrospray needle. The sample was injected in 5% acetonitrile in 0.1% acetic acid

and, after loading for 5 min, a 1-min gradient to 12.5% acetonitrile was followed by a

90-min gradient to 65% acetonitrile in 0.1% acetic acid. We analyzed the eluted

peptides by tandem MS (MS/MS), and data were acquired and mass spectral peaks

were deconvoluted to consolidate different charge states observed for each peptide

with the software supplied by Bruker. The instrument was standardized with a tryptic

digest of alcohol dehydrogenase, cytochrome C, and glycogen phosphorylase to

assess mass accuracy and sensitivity of the instrument before and after each set of

runs.


65

2.2.6 Database, Genome Ontology, and Literature Search

The resulting MS/MS spectra was searched in MGF format from both batch 1

and 2 using the Mascot algorithm search engine (version 2.1) with default variables

and trypsin specified. Specifically, one missed cleavage was allowed, a variable

oxidation of methionine residues and a fixed modification of carbamidomethylation of

cysteines was set with a fragment tolerance of 0.4 Da and a parent tolerance of 3.0

Da. The database used was a custom-built non-redundant compilation of human,

mouse, and rat sequences from GenBank, Ensembl, and SwissProt, compiled

January 2005. A bioinformatics program was used through Protana Inc.

(Mississauga) to identify peptides from the MS/MS spectra present from each

fraction, giving each peptide a score. The identified peptides from all the fractions

within each respective batch were clustered with other peptides that were common

to a particular protein, and each group of peptides was then given a cluster score.

Any peptide with a Mascot score <20 and any protein with a score <40 was removed

from the data. The identified proteins were manually analyzed and classified by their

genome ontology cellular component classification and PubMed literature searches

were conducted on each protein.

The false-positive rate of protein identification was measured by searching a

random database, in which every sequence entry from the “normal” database was

randomly shuffled. The number of hits from each search was categorized based on

score, and for each scoring interval, the false-positive rate was calculated as number

of random hits/(number of random hits + number of normal hits).


66

2.2.7 ELISAs for Kallikrein 5, 6, and 11

Kallikreins 5, 6, and 11 were measured by sandwich-type ELISA, 96-well

plates, as described earlier (192,198,315) with use of biotinylated detection

antibodies, alkaline phosphatase-conjugated streptavidin and diflunisal phosphate

substrate. Plates were read by time-resolved fluorescence(316).

2.2.8 Protein Recovery

To assay for sample recovery of proteins during dialysis and fractionation, we

analyzed sample aliquots that had been taken during this procedure for KLK5 and

KLK6 by the aforementioned ELISA assays.

2.2.9 Mac-2BP ELISA

We obtained the s90K/Mac-2BP ELISA reagent set from Bender

Medsystems. Briefly, serum samples were diluted 1:500 and CM samples 1:10 in

sample diluent buffer (provided by the manufacturer) and loaded onto 96-well strips

pre-coated with anti–Mac-2BP antibody. Samples were incubated at 37°C for 45 min

with shaking at 100 rpm. The wells were washed 3 times with wash buffer (as

provided), a detection antibody was added, and the plate was incubated for 45 min.

The plate was washed and substrate solution (as provided) was added to each well,

and the plate was incubated with shaking at room temperature for 10 min, following

which a stop solution (as provided) was added to each well. Absorbance was

measured at 490 nm by a Wallac–Victor2 plate reader (Perkin-Elmer). An extended

validation of Mac-2BP was performed with 210 CaP, 53 BPH and 17 normal serum


67

samples donated by Carsten Stephan from the Department of Urology, Charité -

Universitätsmedizin Berlin, Germany.

2.3 Results

2.3.1 Total Protein, KLK5 and KLK6 Concentrations in Culture Over Time

To demonstrate the accumulation of secreted proteins over time in the roller

bottle culture, we measured 2 secreted proteins that are known to be produced by

the PC3(AR)6 cell line, namely KLK5 and KLK6, over the 14-day culture period

(Figure 2.2A, B). The concentrations of KLK5 and KLK6 increased with time in

culture and began to plateau after ~10 days. We measured the total protein in the

CM, which increased steadily throughout the culture period (Figure 2.2C).


68

Figure 2.2

Figure 2.2: Measurement of KLK5, KLK6 and total protein. KLK5 (A), KLK6 (B), and total protein (C) concentrations over time in CM of the PC3(AR)6 roller bottle culture. The concentrations of KLK5 and KLK6 were monitored by ELISA. Two replicates are shown.


69

2.3.2 Recovery of KLK5 and KLK6 During Sample Preparation The recoveries of KLK5 and KLK6 were 25% and 17%, respectively. The

major protein losses were seen after lyophilization. In addition, for KLK5, a

significant amount went in the flow-through after column loading.

2.3.3 Proteins Identified by Mass Spectrometry

After LC-MS/MS and searching by Mascot from both batches, we identified

262 proteins from the SAX FPLC fractions that had a Mascot score of at least 40.

Each protein identified was tabulated and cross-referenced with the genome

ontology database for cellular components (317) (Figure 2.3). A large percentage

(39%) of all proteins identified are classified as extracellular (23%) or membrane

(16%) proteins. Many identified proteins are classified as intracellular (50%),

whereas 11% were unclassified. From the list of the 262 proteins, we selected

candidate biomarkers (Table 2.1) based on the following criteria:

1. Proteins were searched manually against the Genome Ontology database

(317) for their cellular localization. Proteins that were classified as secreted

and membrane-bound were selected.

2. Literature searches through the National Center of Biotechnology Information

(NCBI) PubMed database were then performed to determine:

a. if these proteins are novel molecules that have yet to be explored as

potential biomarkers;

b. if these proteins are known to participate in critical pathways implicated

in cancer initiation and progression.


70

Figure 2.3

Figure 2.3: Cellular location of proteins from PC3(AR)6 CM. Classification of proteins by cellular location from PC3(AR)6 CM, batches 1 and 2. Each protein identified after Mascot searching was classified by its cellular location using Genome Ontology classifiers (www.geneontology.org).


71

Table 2.1: Extracellular candidate tumor markers identified in culture medium of the PC3 (AR)6 roller bottle culture.

Gene Protein Description Mascot Score

GDF15 Prostate differentiation factor; PTGF-β 40.81 COL2A1 Alpha 1 type II collagen, isoform 1, preproprotein; collagen II, 44.96 PAM* Peptidylglycine α-amidating monooxygenase, isoform c, preproprotein 45.64 IGFBP2 Insulin-like growth factor binding protein 2 precursor (IGFBP-2) (IBP-2) 45.7 S100A8* S100 calcium-binding protein A8; cystic fibrosis antigen; calgranulin A 47.9 TLR9* Toll-like receptor 9, isoform A precursor 49.44 TPT1* Tumor protein, translationally controlled 1 50.73 CTSL* Cathepsin L 52.09 PLG Plasminogen 54.92 LGALS3 Lectin, galactoside-binding, soluble, 3 (galectin 3); Lectin, galactose-binding 55.53 LTBP2* Latent transforming growth factor β binding protein 2 65.36 MUC5B* Mucin 5, subtype B, tracheobronchial 66.33 GALNT6* Polypeptide N-acetylgalactosaminyltransferase 6 66.35 IGFBP6 Insulin-like growth factor binding protein 6 72.26 INHBB Inhibin β B subunit precursor; Inhibin, β-2 74.86 TIMP2 Tissue inhibitor of metalloproteinase 2 precursor 75.21 LAMC2 Laminin, γ 2, isoform a precursor; nicein (100 kDa); 75.96 CXCL3* Macrophage inflammatory protein-2-β precursor (MIP2-β) 80.8 AZGP1 α-2-glycoprotein 1, zinc; α-2-glycoprotein 82.29 NPC2* Niemann-Pick disease, type C2; Niemann-Pick disease, type C2 gene 82.8 MDK* Midkine (neurite growth-promoting factor 2) 90.34 SERPINA1 Serine (or cysteine) proteinase inhibitor 93.92 KLK6* Kallikrein 6 (neurosin, zyme); protease M; protease, serine, 9 (neurosin) 95.71 SEMA3F* Semaphorin 3F precursor (Semaphorin IV) (Sema IV) (Sema III/F) 100.2 C19orf10* Chromosome 19 open reading frame 10; interleukin 25; interleukin 27 103.18 LRG1 Leucine-rich α-2-glycoprotein 109.66 COL6A2 α 2 type VI collagen, isoform 2C2 109.73 GALNT2 Polypeptide N-acetylgalactosaminyltransferase 2; UDP-GalNAc transferase 2 114.5 IGFBP5 Insulin-like growth factor binding protein 5 120.38 RNASET2* Ribonuclease 6 (EC 3.1.27.—) precursor 124.53 FBLN1 Fibulin 1 isoform C precursor 137.3 DAG1* Dystroglycan 1 precursor; α-dystroglycan; Dystrophin-associated glycoprotein-1 145.92 S100A9 S100 calcium-binding protein A9; calgranulin B 152.1 TGFB2 Transforming growth factor, β 2 152.59 SPINT1* Hepatocyte growth factor activator inhibitor precursor 154.24 B2M β-2-microglobulin 177.47 LCN2* Lipocalin 2 (oncogene 24p3) 181.58 KLK5* Kallikrein 5; stratum corneum tryptic enzyme 181.94 LCN7* P3ECSL; glucocorticoid-inducible protein; oxidized-LDL responsive gene 2 184.67 DF Complement factor D preproprotein; adipsin; properdin factor D 187.96 SFN* Stratifin 193.8 APP* Amyloid β (A4) precursor protein (protease nexin-II, Alzheimer disease) 215.47


72

TIMP1 Tissue inhibitor of metalloproteinase 1 precursor; Erythroid-potentiating activity 227.55 C1S Complement component 1, s subcomponent 254.89 GRN Granulin 258.53 PLTP* Phospholipid transfer protein 282.02 STC2* Stanniocalcin 2; stanniocalcin 2; stanniocalcin-related protein 303 PTX3* Pentaxin-related gene, rapidly induced by IL-1β; Pentraxin-3 308.44 CYR61* Cysteine-rich, angiogenic inducer, 61; cysteine-rich heparin-binding protein 61 327.5 IL6 Interleukin 6 (interferon, β 2) 336.16 PSAP* Saposin precursor 359.65 PLAU Plasminogen activator, urokinase 361.2 FSTL1* Follistatin-like 1 precursor; follistatin-related protein 367.11 P4HB* Procollagen-proline, 2-oxoglutarate 4-dioxygenase (proline 4-hydroxylase) 375.33 AGRN* Agrin 486.69 HSPG2* Basement membrane-specific heparan sulfate proteoglycan core protein precursor 506.82 INHBA* Inhibin β A subunit precursor; Inhibin, β-1; EDF 507.77 NUCB1* Nucleobindin 1 precursor (CALNUC) 631.69 LGALS3BP* Galectin 3 binding protein; L3 antigen; Mac-2-binding protein; serum protein 90K 632.55 COL6A1 Collagen, type VI, α 1 precursor; collagen VI, α-1 polypeptide 766.46 CTSD Cathepsin D (lysosomal aspartyl protease) 771.08 THBS1 Thrombospondin 1 precursor 865.57 BF Complement factor B preproprotein; B-factor, properdin; C3 proactivator 1071.18FN1 Fibronectin 1 isoform 2 preproprotein; cold-insoluble globulin 1978.92C3 Complement component 3 3339.99

**PPrrootteeiinnss lliinnkkeedd ttoo ccaanncceerr iinn pprreevviioouuss lliitteerraattuurree rreeppoorrttss..


73

We determined the overlap of the proteins identified between the 2 batches through

visual inspection of the Mascot data.

Proteins in this list (Table 2.1) that are marked with an asterisk have not been

previously evaluated as serum biomarkers for CaP as determined through PubMed

literature searches specific for the protein.

We calculated false-positive protein identification rates based on a random

database search. False-positive rates for specific scoring intervals were as follows:

40–50, 44%; 50–60, 35%; 60–70, 7%; 70–80, 8%; 90–150, 7%; and >150, 0%. The

presence of a higher false-positive rate in the 70–80 scoring interval is the result of a

statistical fluctuation attributable to additional protein identification in the random

database search.

2.3.4 Overlap of Proteins Identified in Batches 1 and 2

To determine the reproducibility of the method, the proteins identified from

batches 1 and 2 were manually compared for overlap. We identified 145 proteins in

both batches (55% overlap). Additionally, we identified 78 proteins only in batch 1

and 39 proteins only in batch 2. Combined, the total number of identified proteins

was 262. As expected, the more abundant proteins were preferentially identified in

both batches.


74

2.3.5 Mac-2BP Concentrations in Patients with Prostate Cancer vs. Healthy Men and in the Conditioned Media of the PC3(AR)6 Cell Line

From the proteins identified, Mac-2BP was chosen as one novel biomarker

candidate for further validation. The concentrations of Mac-2BP increased in the CM

over time (Figure 2.4A), as expected and in a similar fashion to KLK5 and KLK6

(Figure 2.2). We measured serum concentrations of Mac-2BP from 26 men with CaP

and 17 healthy men using a Mac-2BP ELISA. The median Mac-2BP concentrations

in CaP patients were almost twice as high as those in healthy men (Figure 2.4B),

with 50% of the CaP patients having increased Mac-2BP concentrations compared

with the healthy men (at the 100th percentile of healthy men as a cutoff). The

difference in medians of the 2 populations by Mann-Whitney test was highly

significant (P = 0.003). The negative correlation between Mac-2BP and PSA in these

26 patients was also significant, with a Spearman correlation coefficient (rs) of –0.63

(P < 0.001) (Figure 4C).

We further measured KLK5, KLK6, and KLK11 (a previously identified

prostate and ovarian cancer biomarker) (318) in the same set of patients and

controls, as above. The data (Figure 2.5) showed decreased concentrations of KLK5

in CaP (P < 0.0001 by Mann-Whitney test), decreased concentrations of KLK6 in

CaP (P = 0.03 by Mann-Whitney test), and increased concentrations of KLK11 in

CaP (P < 0.0001 by Mann-Whitney test).

Spearman correlations for all pairs of measured concentrations (Mac-2BP,

KLK5, KLK6, KLK11, and PSA) included only one statistically significant negative

correlation between Mac-2BP and PSA (Figure 2.4C).


75

An extended validation of Mac-2BP was performed using serum samples that

were confirmed cases of CaP, BPH and healthy males. As can be seen from Figure

2.6 the trend initially observed in the preliminary validation is reversed with the

median levels of CaP patients showing lower Mac-2BP serum concentrations than

normals. Levels are also lower in BPH patients. A significant difference is observed

by the Kruskal-Wallis non-parametric one-way ANOVA of p = 0.0028. Post-hoc

analysis by the Dunn’s multiple comparisons test showed there was a significant

difference between the CaP and normal group, p < 0.01.


76

Figure 2.4

Figure 2.4: Mac-2BP concentrations and the correlation between serum Mac-2BP concentrations and PSA. (A), Mac-2BP concentrations in CM of PC3(AR)6 cell line over time. (B), Mac-2BP concentrations in serum of 26 CaP patients and 17 healthy men. Horizontal lines indicate medians. P value was calculated with the Mann–Whitney test. (C), Correlation between serum Mac-2BP concentrations and PSA in the 26 cancer patients. rs = Spearman correlation coefficient.


77

Figure 2.5

Figure 2.5: Concentrations of KLK5, KLK6, and KLK11 in serum of 26 CaP patients and 17 healthy men. Horizontal lines indicate medians. P value was calculated with the Mann–Whitney test.


78

Figure 2.6

p = 0.0028

Figure 2.6: Serum concentrations of Mac-2BP in CaP, BPH and normal patients. Horizontal lines indicate median values. Significance was determined using the Kruskal-Wallis non-parametric one-way ANOVA test.


79

2.4 Discussion

We used a proteomic method for identification of secreted proteins from the

CM of the metastatic CaP cell line PC3(AR)6 as a model to discover novel markers

for CaP. Because large amounts of cells are needed for confident MS detection of

low-abundance cellular proteins, we determined if a cell line could be grown in a

large volume of SFM for an extended period. The use of protein- and peptide-free

chemically defined Chinese hamster ovary serum-free medium simplified analysis,

providing a distinct advantage over fetal calf serum–containing medium, which would

contaminate the CM.

The loss of KLK5 in the flow-through during SAX FPLC was attributable to

incomplete capture of KLK5 by the SAX column. The incomplete capture is

consistent with its relatively high pI of ~ 8. Appreciable losses of both KLK5 and

KLK6 also occurred after lyophilization of the FPLC fractions, possibly attributable to

incomplete solubilization of the freeze-dried protein (319).

Independent MS detection of KLK5 and KLK6, proteins known to be secreted

by PC3(AR)6, was confirmed for batch 1 and 2 in the expected fractions. The

complex mixture of proteins present at varying concentrations in CM necessitated

fractionation before MS to increase the depth of identification. However, not all

proteins in a mixture can be ionized and detected in 1 run, with the lower abundance

proteins not being identified in both batches(320).

Fifty percent of the proteins identified were intracellular. Their presence in the

CM is to be expected because of cell death and their high abundance within cells.

Because we were primarily interested in investigating proteins that are secreted or


80

shed from CaP cells in vivo, we selected the extracellular and membrane proteins

for further evaluation. Each protein was examined to establish if it had been

previously evaluated as a CaP biomarker or if it had any link to cancer. The selected

candidates are listed in Table 2.1.

We performed preliminary validation of Mac-2BP by ELISA. Mac-2BP has

been shown to be a serum prognostic marker in lymphoma (321), and lung (322),

breast (323), hepatocellular (324), ovarian (325), and colon (326,327) carcinoma.

Serum concentrations have not been evaluated in CaP, however, despite the

correlation of immunohistochemical staining for Mac-2BP with Gleason grade (328).

Serum Mac-2BP was increased in 50% of the CaP patients. (Figure 2.4). The

correlation of Mac-2BP concentrations and PSA concentrations in these patients

was weak and negative (Figure 2.4C). As can be seen in our extended validation of

Mac-2BP, the trend seen in the preliminary validation was reversed (Figure 2.6).

Even though Mac-2BP is seen to be elevated at the tissue level in CaP this was not

translated into an elevation seen in serum in our larger sample set. Upon further

study of Mac-2BP it was realised that it is ubiquitously expressed in almost every

tissue in the human body (Unigene database). As a result, the elevations due to

CaP most likely are not discernable among the background of endogenous Mac-2BP

in serum from other tissues. As a result we did concluded that Mac-2BP would not

be a useful serum tumour marker for CaP as it is not specific to the prostate and

eleveations due to CaP are not detectable in serum. Two kallikreins (KLK5 and

KLK6) were also present at lower concentrations in CaP, whereas KLK11 was

present at increased concentrations. These data support the theory that secreted


81

proteins are adjunct biomarkers for CaP, although with less diagnostic accuracy than

PSA. No correlation was seen between any pairs of these markers in serum.

The results of this and other similar studies (97,329) suggest that a wealth of

knowledge is obtainable by analyzing the CM of cell lines. We observed minimal

overlap of our data with those of Martin et al.(97) and Lin et al.(330), who studied the

proteins secreted and present in the LNCaP cell line, with 67 proteins from the

Martin et al. study and 27 proteins from the Lin et al. study overlapping. This

highlights the heterogeneity of cell lines and the data that can be derived from each.

Obviously, much work is needed to further evaluate the identified candidate

biomarkers, but this study will form the basis of future communications.

Chapter 3: Optimization of Cell Culture and Proteomic Workflow

82

CHAPTER 3:

OPTIMIZATION OF CELL CULTURE AND PROTEOMIC WORKFLOW

Sections of this chapter were published in:

Sardana, G., Jung, K., Stephan, C. and Diamandis, E.P. Proteomic Analysis of Conditioned Media from the PC3, LNCaP and 22Rv1 Prostate Cancer

Cell Lines: Discovery and Validation of Candidate Prostate Cancer Biomarkers Journal of Proteome Research, 2008 Aug 1;7(8):3329-3338


Chapter 3: Optimization of Cell Culture and Proteomic Workflow 83

3.1 Introduction

As with most proof of principle studies, based on the results obtained an

optimization is needed before further experiments are conducted. From the data

shown in the previous chapter there were some key issues that were identified that

required improvement.

First, proteins identified from one cell line are not indicative of the

heterogeneity of a disease phenotype. Second, the culture volume and amount of

total protein produced was not controlled for the capacity of the mass spectrometer

used. Third, the amount of cell death was not controlled for during culture with ~50%

of the identified proteins identified were from an intracellular source. Fourth, the

sample preparation workflow employed was not in agreement with the published

literature as an optimal method for sample preparation. Fifth, the amount of protein

loss during our sample preparation steps was not quantified. Sixth, the type of mass

spectrometer used was considered to have a relatively low sensitivity and accuracy

in identifying peptides. Seventh, our bioinformatics approach did not fully take

advantage of multiple search engines and assessing prediction probabilities as well

as determining false positive rates (FPR).

Taking all these concerns into account, a modified approach was devised

based on the best practices used for proteomic analysis from the literature. These

changes were then incorporated and applied to modify the proteomic workflow from

our proof of principle study.



3.2.1 Cell Culture

The PC3, LNCaP and 22Rv1 cell lines were purchased from the American

Type Culture Collection (Rockville, MD). All cell lines were grown in T-175 culture

flasks (Nunc) in RPMI 1640 culture medium (Gibco) supplemented with 8% fetal

bovine serum (FBS) (Hyclone). Cells were cultured in a humidified incubator at 37oC

and 5% CO2. Cells were seeded at varying densities in duplicate. PC3 and 22Rv1

cells were grown for 2 days in 30mL of RPMI + 8% FBS. Afterwards, the medium

was removed, and the flask was gently washed 3 times with 30mL of PBS (137 mM

NaCl, 10 mM phosphate, 2.7 mM KCl, pH 7.4). Thirty milliliters of CDCHO (Gibco)

medium supplemented with glutamine (8mmol/L) (Gibco) were added to the flasks

and incubated for 2 days. The LNCaP cell line was grown as above, except that the

cells were incubated for 3 days in RPMI + 8% FBS before the media were changed

to CDCHO. The cell lines were grown in duplicate.

After incubation in CDCHO, the CM was collected and spun down (3,000 X g)

to remove cellular debris. Measurement of total protein, lactate dehydrogenase

(LDH) and PSA, KLK5, KLK6 were taken from each culture.

3.2.2 Measurement of Total Protein, LDH and PSA, KLK5 and KLK6

The total protein of the CM was measured using the Coomassie (Bradford)

assay (Pierce Biotechnology) as recommended by the manufacturer.

Lactate dehydrogenase levels in the CM were measured via an enzymatic

assay based on conversion of lactate to pyruvate. NADH production from NAD+


during this reaction was monitored at 340mm with an automated method and

converted to Units per litre (U/L) (Roche Modular Systems).

Kallikrein 3 (PSA), KLK5 and KLK6 were measured with in-house ELISAs as

described earlier(192,315,331).

3.2.3 Sample Preparation

We devised a new approach to sample preparation where we monitored the

level of PSA by ELISA after each sample preparation step as a marker for overall

protein recovery. The sample preparation procedure was devised from the published

literature. Approximately 30mL of CM from the LNCaP cell line, which corresponded

to 1mg of total protein, were dialyzed overnight using a 3.5kDa cut-off dialysis tubing

(Spectra/Por) at 4oC in 5L of 1mM of ammonium bicarbonate solution with one buffer

exchange. The dialyzed CM was lyophilized overnight to dryness followed by

resolubilization with 322µL of 8M urea, 25µL of 200mM DTT, and 25µL of 1M

ammonium bicarbonate. The sample was vortexed thoroughly and incubated at 50oC

for 30min. One hundred and twenty five micro litres of 500mM iodoacetamide were

added and the sample was incubated in the dark at room temperature for 1h. The

sample was then desalted using a NAP-5 column (GE Healthcare) and lyophilized to

dryness.


3.3 Results

3.3.1 Cell Culture Optimization

Growth conditions of the three cell lines were optimized in order to reduce cell

death and maximize secreted protein levels. Cells were incubated in SFM for 2 days

at different seeding densities. Total protein, LDH and the concentration of KLK5 and

KLK6 in the CM of PC3, and PSA in the CM of LNCaP and 22Rv1 cells were

measured. The ratio of PSA, 5, and 6 concentrations with LDH levels for each

culture condition (Figures 3.1, 3.2, 3.3) were compared. The optimal seeding

concentrations were 7.5 X 106, 22 X 106, and 75 X 106 cells for PC3, LNCaP and

22Rv1 cell lines, respectively, as these gave the highest ratio of KLK production

(indicator of secreted proteins) to LDH (indicator of cell death). The total protein of

the CM for the optimized seeding densities was 33 µg/mL, 39 µg/mL and 39 µg/mL

for PC3, LNCaP and 22Rv1 cells, respectively. Thus, 30mL of media contained

approximately 1mg of total protein.


Figure 3.1

Figure 3.1: Measurements of PSA and LDH in the CM of the 22Rv1 cell line. Concentrations of PSA (A) and LDH (B) at different seeding densities. (C) Ratio of PSA to LDH at different seeding densities.


Figure 3.2

Figure 3.2: Measurements of PSA and LDH in the CM of the LNCaP cell line. Concentrations of PSA (A) and LDH (B) at different seeding densities. (C) Ratio of PSA to LDH at different seeding densities.


Figure 3.3

Figure 3.3: Measurements of KLK5, KLK6 and LDH in the CM of the PC3 cell line. Concentrations of KLK5 and KLK6 (A) and LDH (B) at different seeding densities. (C) Ratio of KLK5 and KLK6 to LDH at different seeding densities.


3.3.2 Sample Preparation Optimization

The consensus reached from the literature search to optimize our sample

preparation workflow is shown in Figure 3.4 with the total recovery of PSA after each

step also shown. The proteomic workflow of choice was found to be the MuDPiT

(multi-dimensional protein identification technology) approach first shown by

Washburn et al.(332) in 2001. This involved trypsin digestion of the sample followed

by orthogonal chromatography steps. Strong cation exchange (SCX) followed by

LC-MS/MS reversed-phase C18 capillary chromatography. To employ this approach

with our requirements for processing the CM, we utilized dialysis and lyophilization

as intermediate steps to concentrate the sample prior to denaturing with urea, DTT

and iodoacetamide. A desalting step was added to remove any salts in the sample,

followed by a lyophilization step prior to resuspension in Buffer A for SCX

chromatography.

In order to evaluate sample recovery PSA was measured by ELISA at each

step. The results are shown in Figure 3.4. As can be seen the overall recovery is

~20% with the majority of sample loss at the lyophilization steps. Based on the

results obtained here, we determined our starting total protein concentration should

be 1mg. Thus, ~200ug of total protein would be fractionated prior to LC-MS/MS, and

if ~10 fractions were collected, each would contain ~20ug of peptides. Considering

the capillary column used during electrospray ionization into the mass spectrometer

can only hold 1ug of peptides, the total concentration of peptides in each fraction

would be more than enough.


Figure 3.4

Figure 3.4: Overview of the optimized sample preparation workflow. Levels of PSA are measured by ELISA after each step of the sample preparation procedure to assess sample recovery.


3.4 Discussion

In order to improve our sample coverage and efficiency of identification of

proteins the methods employed in the proof of principle study needed to be

optimized. This required analyzing our approach at each step of our workflow and

determining the most robust procedure for sample preparation and analysis. We

addressed each of our concerns raised in the proof of principle study before

continuing on with our extended analysis of CaP cell lines.

Initially, our primary concern was the sample preparation procedure used to

process the CM prior to LC-MS/MS analysis was not the most commonly used for

this type of ‘shotgun’ proteomics application. Instead a multi-dimensional

chromatographic approach that fractionated peptides rather than proteins was

investigated(332). This MuDPiT approach reduced the sample handling of samples

by performing one initial trypsin digest prior to the first dimension of chromatography

and in addition employed a high resolution HPLC for the first chromatographic

dimension. It was agreed upon that this method provided a robust approach to

protein identification.

Our next concern was the culture volume used to grow the cells was too large

and we did not optimize the seeding density and duration of culture for the amount of

total protein produced to meet our requirements for LC-MS/MS. Based on the

amount of total protein required for LC-MS/MS analysis we worked backwards and

determined what would be the initial amount of total protein required before cell

culture and the sample processing steps. To aid in determining this we first

quantified our protein recovery for our new proteomic workflow. By measuring PSA


after each step up to trypsin digestion we were able to determine the overall

recovery to be ~20% (Figure 3.4). Using this information and based on the fact that

the reversed phase capillary used for LC-MS/MS can only bind 1ug of peptides, we

determined that having 1mg of total protein starting material would be sufficient for

our purposes.

To encompass more of the heterogeneity of the CaP phenotype we chose to

expand the number of cell lines studied. In this respect, we chose to use the PC3,

LNCaP and 22Rv1 CaP cell lines since each had been derived from a different

metastatic site or localized from the prostate and exhibited a unique morphology and

phenotype in culture. As a result each cell line had to be optimized to grow in culture

to produce the optimal amount of total protein while minimizing cell death and

maximizing secreted protein produced. We grew each cell line at different seeding

densities for a short period in SFM. The CM was then collected and the levels of

KLK’s for each cell line were measured as well as the level of LDH in the CM. The

amount of KLK and LDH in the CM acted as a surrogate markers for the amounts of

secreted and intracellular proteins produced by the cell respectively. To determine

the optimum seeding density for each cell line the ratio of KLK to LDH was

determined and the seeding density with the highest ratio was chosen. The results of

the optimization for each cell line are shown in Figures 3.1, 3.2, and 3.3. As can be

seen, each cell line had different growth rates and required differing seeding

densities. Based on these results the appropriate seeding density for each cell line

was chosen.


In addition to improvements in our sample preparation and workflow, we also

aimed to improve the mass spectrometric platform used. From the Bruker ion trap

used in our proof of principle study we transitioned to using the Thermo LTQ linear

ion trap which has enhanced sensitivity. This is due to its ability to trap more ions

and analyze them faster due to its shortened duty cycle. This translates into an

increase in the number of peptides identified as well as the number of spectra

identified per peptide.

To manage the large amounts of data that will be obtained from the analysis

of three cell lines and the replicates semi-automated database applications were

designed to assist in the annotation and comparison of the data. These included a

parsing program to assist with annotation by genome ontology of proteins identified

which was designed in collaboration with Adrian Pasulescu at the Samuel Lunenfeld

Research Institute; a database application to sort the genome ontology annotated

proteins into unique groups, by Peter Bowden at Ryerson University; a protein list

comparison database application to compare proteins identified between cell lines

an replicates, designed with David Ngyuen at Mt. Sinai Hospital. Together these

semi-automated applications will increase the efficiency at which data is analyzed.

Chapter 4: Comparative Proteomic Analysis of the Conditioned Media of Three Prostate Cancer Cell Lines

95

CHAPTER 4:

COMPARATIVE PROTEOMIC ANALYSIS OF THE CONDITIONED MEDIA OF THREE PROSTATE CANCER

CELL LINES

The work presented in this chapter was published in part in the Journal of Proteome Research:

Sardana, G., Jung, K., Stephan, C. and Diamandis, E.P. Proteomic Analysis of Conditioned Media from the PC3, LNCaP and 22Rv1 Prostate Cancer

Cell Lines: Discovery and Validation of Candidate Prostate Cancer Biomarkers Journal of Proteome Research, 2008 Aug 1;7(8):3329-3338



96

4.1 Introduction

Currently, serum PSA levels, combined with DRE, are the recommended

screening tests for early detection of CaP in asymptomatic men over the age of

50(333). However, there is considerable controversy surrounding the efficacy of the

PSA test in reducing the overall mortality of CaP(334,335). These concerns stem

from over-diagnosis and the lack of specificity of PSA in discriminating CaP from

BPH(335). While PSA has been shown to correlate very well with tumour

volume(336), it is unable to predict with certainty the biological aggressiveness of

the disease. Several refinements of the PSA test have been shown to increase its

sensitivity and specificity(337-341). However, there is still a need to develop non-

invasive tests to identify clinically relevant CaP(342).

Mass spectrometry-based proteomic technologies are currently in the

forefront of cancer biomarker discovery. Serum or plasma is usually the discovery

fluid of choice(343) however, several studies have employed MS for biomarker

discovery in various other biological fluids(344,345). In addition, a number of

alternative approaches have been used, such as MS spectra profiling(275,346),

which refers to the identification of discriminatory mass spectral features without

identification of the peptides being identified, peptide profiling(347,348), which refers

to the identification of differentially cleaved or modified peptide sequences, and

isotopic labelling(349). There are many inherent limitations to using MS for

biomarker discovery in complex biological mixtures such as serum(296). The main

concern is the suppression of ionization of low abundance proteins by high

abundance proteins such as albumin and immunoglobulins. Depletion strategies


97

have been used to remove high abundance proteins which have aided in improving

the detection limit of MS(350,351). One approach to overcome the limitations posed

by biological fluids is to study the secretome of cell lines grown in SFM. The proteins

identified from the CM are specific to the cell line being cultured as there are no

other contaminating proteins, therefore, greatly simplifying MS analysis. Studies

analyzing the CM from prostate(97), colon(352), endothelial(353), adipose(354),

nasopharyngeal(355) and retinal epithelial cells(356), have already been conducted,

thus demonstrating the versatility of this approach.

Previously, we have shown the PC3(AR)6 CaP cell line can be grown in a

serum-free environment and the secreted proteins present can be readily identified

by MS-based methods(98). In this study, we performed a detailed proteomic

analysis of the CM of three CaP cell lines; PC3, LNCaP and 22Rv1. From this

analysis we identified 2,124 proteins by using a bottom-up approach, consisting of

offline strong cation exchange (SCX) chromatography followed by capillary C-18

reversed-phase liquid chromatography-tandem mass spectrometry (LC-MS/MS).

Our extensive lists of proteins, and their cellular and biological classifications, may

form the basis for discovering novel CaP biomarkers.

The utility of this approach was examined by validating four candidates by

ELISA. Each was found elevated in serum of a subset of CaP patients; a positive

correlation was also observed with serum PSA levels. Furthermore, our approach

identified many known CaP biomarkers including PSA, KLK2, KLK11, prostatic acid

phosphatase and prostate specific membrane antigen, further supporting the view

that this unbiased approach may aid in new CaP biomarker discovery efforts.


98


4.2.1 Cell Culture

The PC3, LNCaP and 22Rv1 cell lines were purchased from the American

Type Culture Collection (Rockville, MD). All cell lines were grown in T-175 culture

flasks (Nunc) in RPMI 1640 culture medium (Gibco) supplemented with 8% fetal

bovine serum (FBS) (Hyclone). Cells were cultured in a humidified incubator at 37oC

and 5% CO2. Cells were seeded at 7.5 X 106, 22 X 106, and 75 X 106 cells for PC3,

LNCaP and 22Rv1 cell lines, respectively. PC3 and 22Rv1 cells were grown for 2

days in 30mL of RPMI + 8% FBS. Afterwards, the medium was removed, and the

flask was gently washed 3 times with 30mL of PBS (137 mM NaCl, 10 mM

phosphate, 2.7 mM KCl, pH 7.4). Thirty milliliters of CDCHO (Gibco) medium

supplemented with glutamine (8mmol/L) (Gibco) were added to the flasks and

incubated for 2 days. The LNCaP cell line was grown as above, except that the cells

were incubated for 3 days in RPMI + 8% FBS before the media were changed to

CDCHO. All cell lines were grown in triplicate and independently processed and

analyzed. A negative control was also prepared with the same procedures as above,

except no cells were seeded.

After incubation in CDCHO, the CM was collected and spun down (3,000 X g)

to remove cellular debris. Aliquots were taken for measurement of total protein,

lactate dehydrogenase (LDH) (a marker of cell death), and kallikreins 3, 5, 6 (internal

control proteins). The remainder was frozen at -80oC until further use.


99

4.2.2 Measurement of Total Protein, Lactate Dehydrogenase and PSA, KLK5, KLK6

The total protein of the CM was measured using the Coomassie (Bradford)

assay (Pierce Biotechnology) as recommended by the manufacturer.

Lactate dehydrogenase levels in the CM were measured via an enzymatic

assay based on conversion of lactate to pyruvate. NADH production from NAD+

during this reaction was monitored at 340mm with an automated method and

converted to Units per litre (U/L) (Roche Modular Systems).

Kallikrein 3 (PSA), KLK5 and KLK6 were measured with in-house ELISAs as

described earlier(192,315,331).

4.2.3 Conditioned Media Sample Preparation and Trypsin Digestion

Our general protocol is shown in Figure 4.1. Approximately 30mL of CM from

each cell line, which corresponded to 1mg of total protein, was dialyzed overnight

using a 3.5kDa cut-off dialysis tubing (Spectra/Por) at 4oC in 5L of 1mM of NH4HCO3

solution with one buffer exchange. The dialyzed CM was lyophilized overnight to

dryness followed by resolubilization with 322µL of 8M urea, 25µL of 200mM DTT,

and 25µL of 1M NH4HCO3. The sample was vortexed thoroughly and incubated at

50oC for 30min. 125uL of 500mM iodoacetamide was added, the sample was

incubated in the dark at room temperature for 1h. The sample was desalted with a

NAP-5 column (GE Healthcare), lyophilized and resuspended in 120µL of 50mM

NH4HCO3, 100µL methanol, 150µL H2O and 5µg of trypsin (Promega), vortexed,

and incubated for 12h at 37oC.


100

Figure 4.1

Figure 4.1: Workflow of proteomic method employed. For additional information, refer to Materials and Methods section.


101

4.2.4 Strong Cation Exchange High Performance Liquid Chromatography

The tryptic digests were lyophilized and resuspended in 120µL of 0.26M

formic acid in 10% ACN (mobile phase A). The sample was fractionated using an

Agilent 1100 HPLC system connected to a PolySULFOETHYL ATM column

containing a hydrophilic, anionic polymer (poly-2-sulfoethyl aspartamide) with a 200Å

pore size and a diameter of 5µm (The Nest Group Inc.). A one hour linear gradient

was used, with 1M ammonium formate and 0.26M formic acid in 10% acetonitrile

(mobile phase B) at a flow rate of 200µL/min. Fractions were collected via a fraction

collector every 5 min (12 fractions per run), and frozen at -80oC for further use. A

protein cation exchange standard, consisting of four peptides, was run at the

beginning of each day to assess column performance (American Peptide).

4.2.5 Online Reversed Phase Liquid Chromatography – Tandem Mass Spectrometry

Each 1mL fraction was C18 extracted using a Zip TiPC18 pipette tip (Millipore)

and eluted in 4µL of 90% ACN, 0.1% formic acid, 10% water, 0.02% trifluoroacetic

acid (TFA) (Buffer B). Upon which 80µL of 95% water, 0.1% formic acid, 5% ACN,

0.02% TFA (Buffer A) was added and half of this volume (40µL) was injected via an

auto-sampler on an Agilent 1100 HPLC. The peptides were first collected onto a 2

cm C18 trap column (inner diameter 200 µm), then eluted onto a resolving 5 cm

analytical C18 column (inner diameter 75 µm) with an 8 micron tip (New Objective).

The HPLC was coupled online to an LTQ 2-D Linear Ion Trap (Thermo Inc.). A 120

min gradient was used on the HPLC and peptides were ionized via nano-


102

electrospray ionization. The peptides were subjected to MS/MS and DTAs were

created using the Mascot Daemon v2.16 and extract_msn (Matrix Science).

Parameters for DTA creation were: min mass 300, max mass 4000, automatic

precursor charge selection, min peaks 10 per MS/MS scan for acquisition and min

scans per group of 1.

4.2.6 Database Searching and Bioinformatics

Mascot, v2.1.03 (Matrix Science)(357) and X!Tandem v2.0.0.4 (GPM, Beavis

Informatics Ltd.)(358) database search engines were used to search the spectra

from the LTQ runs. Each fraction was searched separately against both search

engines using the IPI Human database V3.16(359) with trypsin specified as the

digestion enzyme. One missed cleavage was allowed, a variable oxidation of

methionine residues and a fixed modification of carbamidomethylation of cysteines

was set with a fragment tolerance of 0.4 Da and a parent tolerance of 3.0 Da.

The resulting DAT files from Mascot and XML files from X!Tandem were

inputed into Scaffold v01_05_19 (Proteome Software)(360) and searched by

allocating all DAT files into one biological sample and all XML files into another

biological sample. This was repeated three times for each cell line. The cut-offs in

Scaffold were set for 95% peptide identification probability and 80% protein

identification probability. Identifications not meeting these criteria were not included

in the displayed results. The sample reports were exported to Excel, and an in-

house developed program was used to extract Gene Ontology (GO) terms for

cellular component for each protein and the proportion of each GO term in the


103

dataset. Proteins that were not able to be classified by GO terms were checked with

Swiss-Prot entries and against the Human Protein Reference Database(361) and

Bioinformatics Harvester (http://harvester.embl.de) to search for cellular component

annotations. Finally, the overlap between proteins identified from each cell line and

within each of the replicates per cell line was determined by in-house built software.

Each protein was also searched against the Plasma Proteome Database

(www.plasmaproteomedatabase.org). The list of proteins were also compared with

those found in seminal plasma by Pilch et al.(362) and in breast cancer cell line CM

by Kulasingam et al.(363) In addition, we used Ingenuity Pathway Analysis software

(Ingenuity Systems) to determine differences in biological networks in the

extracellular and membrane proteins of each cell line, as well as overlay molecular

functions with respect to disease conditions associated with each of the biomarker

candidates.

The same set of spectra produced by the LTQ was searched with the same

parameters as above, but against a reversed IPI Human database V3.16 which was

created using the Reverse.pl script from The Wild Cat Toolbox (Arizona Proteomics).

The DAT and XML files from this “reversed” search were input into Scaffold as

before and the identified peptides meeting the pre-set cut-offs were identified. The

false positive rate (FPR) was calculated as such: FPR = # False peptides / (# True

peptides + # False peptides).

http://harvester.embl.de/

http://www.plasmaproteomedatabase.org/


104

4.2.7 Validation of Candidates

A pre-clinical validation was conducted where the concentrations of four

candidate proteins (follistatin, chemokine (C-X-C motif) ligand 16, pentraxin 3 and

spondin 2) were measured by ELISA in the serum of patients with or without

prostate cancer and in the CM of each cell line. These candidates were selected for

validation using similar criteria as Kulasingam et al.(363) Follistatin (R&D Systems),

chemokine (C-X-C motif) ligand 16 (R&D Systems) and pentraxin 3 (Alexis

Biochemicals) were measured using their respective manufacturer’s protocol.

Spondin 2 was measured at diaDexus Corporation, San Francisco, CA using an in-

house developed assay. Serum samples from healthy males and CaP patients were

collected at the Toronto Medical Laboratories, Toronto, Canada. The patient groups

were classified by their PSA levels. Healthy males were patients with PSA levels

<1ug/L and CaP patients were those with PSA levels >10ug/L. The study was

carried out after Institutional Review Board approval, median age of patients were 75

and 67 for the cancer and non-cancer groups, respectively.


105

4.3 Results

4.3.1 Proteins Identified by Mass Spectrometry

A schematic of the sample preparation and bioinformatics utilized in this study

is shown in Figure 4.1. Each cell line was independently cultured in triplicate to

determine the reproducibility of our method in identifying proteins between the

replicates. After setting a cut-off of 95% peptide probability and 80% protein

probability in Scaffold, 2,124 proteins were identified that met the criteria from all

three cell lines combined. In total, from the three replicates per cell line, 1,157,

1,285, and 1,116 proteins were found in the PC3, LNCaP and 22Rv1 cell lines,

respectively.

A control flask was also prepared that did not contain any cells but were

treated with the same procedure. A total of 69 proteins were identified in the

negative control flask, which represented FCS-derived proteins from incomplete

washing of the tissue culture flasks. These proteins were removed from the list of

identified CM proteins of each cell line and were not considered further in the data

analysis.

Furthermore, to empirically determine false positive rate (FPR) of peptide

detection, the dataset was searched against a reversed IPI Human v3.16 database

using the same search parameters and Scaffold cut-offs. A total of 4, 4, and 6

proteins from the PC3, LNCaP and 22Rv1 cell lines, respectively were observed.

Each protein was identified by one peptide with Mascot and with no peptides

identified by X!Tandem (0% FPR). A FPR of <1% for all 3 cell lines was calculated

by Mascot.


106

4.3.2 Identification of Internal Control Proteins

As an internal control, we sought to identify by LC-MS/MS three proteins that

were known to be secreted by these cell lines and were monitored by ELISA during

our cell culture optimization. The approximate initial concentrations of these proteins

in CM are given in brackets, below. We confidently identified PSA in the CM of

LNCaP (~550 µg/L) and 22Rv1 (~3 µg/L) with several peptides (Table 4.1). PC3

cells do not secrete any detectable PSA by ELISA and, as expected, this protein

was not identified in its CM. One KLK6 peptide from the CM of PC3 (~1.5 µg/L) was

identified, but no peptides from KLK5 were identified by MS. Together, these data

suggest that the detection limit of our MS-based method for protein identification is in

the low µg/L range.

4.3.3 Reproducibility between Replicates

Next, we investigated the reproducibility of our method by culturing each cell

line in triplicate. The Mascot and X!Tandem results from the PC3, LNCaP and

22Rv1 cell lines are shown in Figure 4.2. In general, a 56% overlap of identified

proteins in all three replicates from both Mascot and X!Tandem for each of the cell

lines was observed. Approximately 20% of proteins were found in two replicates and

24% were exclusive to one replicate. This data highlights the ionization efficiency of

the mass spectrometer, and thus the need for replicate analysis of samples by MS to

obtain a more comprehensive list of the secretome of these cell lines.


107

Table 4.1: Known prostate biomarkers identified in the conditioned media of PC3, LNCaP, and 22Rv1 cell lines1

1 Protein name and number of peptides identified by Mascot and X!Tandem for each cell line are listed. KLK2 – Human kallikrein 2, PSA – Prostate specific antigen, KLK11 – Human kallikrein 11, ACPP – Prostatic acid phosphatase, Mac-2BP – Mac 2 binding protein, Zn-α2-GP – Zinc alpha 2 glycoprotein, PSMA – Prostate specific membrane antigen, NGEP – New gene expressed in prostate.


108

Figure 4.2

Figure 4.2: Overlap of the 3 replicates from PC3, LNCaP and 22Rv1 conditioned media: Number of proteins identified from each cell culture replicate that overlapped with other replicates for both Mascot and X!Tandem results are depicted. Each circle represents a replicate.


109

4.3.4 Differences in Proteins Identified Between Cell Lines Following this, the proteins identified from each of the three replicates of each

cell line were combined to form a non-redundant list per cell line. These three lists

were compared to determine their overlap. The results are shown in Figure 4.3.

About 54% of proteins were unique to one of the cell lines, 21% were common in all

3 cell lines, with 24% were identified in two of the cell lines. This data highlights the

heterogeneity of CaP cell lines and the need to investigate multiple cell lines, to

obtain a comprehensive picture of the CaP proteome. The overlap among the cell

lines for extracellular and membrane proteins, rather than total proteins yielded

similar heterogeneity (Figure 4.4).

4.3.5 Genome Ontology Distributions of Proteins

As shown in Figure 4.5, 12% of the proteins identified were classified as

extracellular, 18% as membrane and 12% as unclassified. The remainder of the

proteins identified in the CM were classified as intracellular, nucleus, golgi,

endoplasmic reticulum (ER), endosome or mitochondria (58%). Classification by

cellular localization is redundant since a protein can be classified in more than one

compartment. A similar distribution was found with the cellular component

distribution of each cell line (Figure 4.6).


110

Figure 4.3

Figure 4.3: Overlap of proteins identified between each cell line: Each circle represents a cell line.


111

Figure 4.4

Figure 4.4: Overlap of the extracellular and membrane proteins identified in each cell line: Extracellular and membrane proteins respectively were compared from each cell line. The numbers of proteins common and unique to each cell line are depicted.


112

Figure 4.5

Figure 4.5: Classification of proteins by cellular location: Each protein identified after MASCOT and X!Tandem searching was classified by its cellular location using Genome Ontology classifiers (www.geneontology.org).

http://www.geneontology.org/


113

Figure 4.6

Figure 4.6: Classification of proteins by cellular location for each cell line: Each protein identified after MASCOT and X!Tandem searching was classified by its cellular location using Genome Ontology classifiers (www.geneontology.org). Distribution of proteins identified in the conditioned media of (A) PC3, (B) LNCaP, and (C) 22Rv1 cell lines.

http://www.geneontology.org/


114

4.3.6 Secreted and Membrane Proteins After GO annotation, 329 proteins were classified as extracellular, 504 as

membranous and 339 as unclassified. From the list of extracellular and membrane

proteins several previously known CaP biomarkers were identified. These included

PSA(364), KLK2(337,338), KLK11(365), PAP(366), Mac-2BP(367), zinc-α-2-

glycoprotein(368), a new gene expressed in the prostate (NGEP)(369) and

PSMA(370) (Table 4.1). Comparing all extracellular, membrane and unclassified

proteins with the Plasma Proteome Database yielded 93 membrane proteins, 98

extracellular proteins, and 27 unclassified proteins as being found previously in

serum.

4.3.7 Overlap with Previous Data

Given that we had previously analyzed the CM of the PC3(AR)6 cell line(98),

the overlap between this set of data and our previous one was examined. In brief,

the current data yielded 5-fold more proteins owing in part to the improvements in

sample preparation, the type of mass spectrometer used (more sensitive) and the

powerful bioinformatics (higher confidence of protein probability).

4.3.8 Overlap with Seminal Plasma Proteins

Recently, Pilch and Mann(362) conducted an in-depth proteomic survey of

the seminal plasma proteome and identified 932 unique proteins. Comparing our list

of extracellular, membrane and unclassified proteins found in the CM of prostate

cancer cell lines with those reported in a biological fluid such as seminal


115

plasma(362) resulted in 108 extracellular, 120 membranous, and 40 unclassified

proteins as being common.

4.3.9 Biological Network Analysis

The extracellular and membrane proteins from each cell line were analyzed

using Ingenuity Pathway Analysis software v5.51 (Ingenuity Systems) and functional

networks were developed. Cancer, cellular movement and pathways involved in

tissue development were listed among the top networks for each of the cell lines.

However, in the case of 22Rv1 these were listed secondary to the “housekeeping”

functional networks. Next, the functions and disease association of each of the four

candidates selected were examined. Follistatin displayed the most direct

connections to cancer such as prostate(371,372), colon(373), and ovarian

cancer(374) (Figure 4.7). Chemokine (C-X-C motif) ligand 16 (CXCL16) displayed

connections to inflammation and chemotaxis(375) (Figure 4.8A). Pentraxin 3 (PTX3)

showed multiple connections to tissue and embryonic development(376) (Figure

4.8B). Few studies have been conducted on spondin 2 (SPON2); it was shown to

be involved in neuronal guidance and lung cancer(377) (Figure 4.8C).


116

Figure 4.7

Figure 4.7: Molecular functions related to diseases associated with Follistatin: Web diagram depicting the biological functions that follistatin is associated with, in the context of disease. Diagram generated through Ingenuity Pathway Analysis software (Ingenuity Systems).


117

Figure 4.8

Figure 4.8: Molecular functions related to diseases associated with Chemokine (C-X-C motif) ligand 16, Pentraxin 3 and Spondin 2: Web diagram depicting biological functions associated with in the context of disease for CXCL16 (A), PTX3 (B) and SPON2 (C). Diagram generated through Ingenuity Pathway Analysis software (Ingenuity Systems).


118

4.3.10 Overlap with Breast Cancer Secretome The proteins identified by Kulasingam et al.(363) in the CM of two breast

cancer cell lines were compared with those identified in our analysis of CaP CM.

There was an overlap of 256 proteins that were present in the BT474, MDA468,

PC3, LNCaP and 22Rv1 cell line CM. These proteins were then subjected to

biological network analysis in a similar manner as stated above. Pathways related to

cancer and cell growth were among the top identified.

4.3.11 Validation of Follistatin, Chemokine (C-X-C motif) ligand 16, Pentraxin 3 and Spondin 2

From the list of proteins identified, pre-clinical validation of four candidates

were performed in the CM of PC3, LNCaP and 22Rv1 cell lines (Figure 4.9) and in

42 serum samples from patients with or without CaP (Figure 4.10). The

concentration of each candidate in the CM correlates semi-quantitatively with the

number of unique spectra (shown in brackets) identified from each peptide of each

candidate after database searching by Mascot/X!Tandem: Follistatin (PC3, 100/128;

22Rv1, 0/1); CXCL16 (LNCaP, 3/4, 22Rv1, 2/2); PTX3 (PC3, 45/52); SPON2

(LNCaP, 152/149, 22Rv1, 2/2). A significant difference (Mann-Whitney Test) in sera

of patients with or without CaP in all four candidates (Figure 4.9A, C, E, G) by ELISA

was observed. In addition, the correlation between PSA levels and candidate levels

in serum of patients with CaP was significant and positive by Spearman analysis

(Fig 4.9B, D, F, H).


119

Figure 4.9

Figure 4.9: Concentrations of each candidate in the conditioned media of each cell line and the control flask. Levels of Follistatin (A), CXCL16 (B), Pentraxin 3 (C), Spondin 2 (D) in conditioned media is shown with corresponding number of spectra identified for each candidate by Mascot/X!Tandem.


120

Figure 4.10

Figure 4.10: Validation of Follistatin, Chemokine (C-X-C motif) ligand 16, Pentraxin 3 and Spondin 2 in serum: Levels of Follistatin, CXCL16, PTX3 and SPON2 measured in serum of patients with or without CaP (A, C, E, G respectively). Median values are shown by a horizontal line. p values were calculated using a Kruskal-Wallis test. Correlations of each candidate with PSA levels (B, D, F, H), p values calculated by the Spearman correlation. (r = Spearman correlation coefficient)


121

4.4 Discussion LC-MS/MS analysis allows for the elucidation of the identity of thousands of

proteins in complex mixtures, in a high-throughput fashion. This technology has

been applied to cancer biomarker discovery, where biological fluids, tissues or cell

cultures have been analyzed for differences in protein expression(97,352,378-380).

However, the dynamic range of current LC-MS/MS methods is not adequate to

identify all proteins in a complex mixture such as serum. Even with depletion of

highly abundant proteins and extensive fractionation, this is a major challenge still

faced today(296).

Prostate cancer, when diagnosed early, is associated with favourable clinical

outcomes(381). Our objective was to identify proteins secreted by three

tumourigenic CaP cell lines of differing origin and phenotype. The purpose of

examining the secretome of cell lines of differing origin was to obtain a more

complete picture of the proteome of CaP since CaP is a heterogeneous

disease(382,383) and it requires a diverse model system for biomarker discovery.

Thus, we chose to focus on shed and secreted proteins from three CaP cell lines,

since this approach is amenable to MS and these proteins will most likely be

produced by the tumour in measurable amounts to be detected via a blood test.

In our previous study of the CM of the PC3(AR)6 cell line(98) we found a large

number of intracellular proteins. In another study, similar data were seen with the

LNCaP cell line (97). We sought to reduce the amount of intracellular proteins by

optimizing the cell culture (Figures 3.1, 3.2, 3.3). Yet our current data (Figure 4.5)

revealed a similar distribution of proteins by cellular component. From this, we


122

deduce that cell death during culture is an unavoidable contaminant when analyzing

CM by proteomics. However, previous studies from our laboratory utilizing a similar

cell culture-based approach found that the proteins identified in cell lysates do not

contain as many extracellular proteins as the CM for that cell line(363). Furthermore,

the extracellular proteins identified in the cell lysate displayed minimal overlap to the

proteins identified in the CM illustrating that analyzing the CM, despite the amount of

cell death occurring in SFM, leads to a significant enrichment of secreted proteins

which may be novel serological markers. We also believe that our method may

include microsomal proteins that were not removed during centrifugation of the CM

after harvesting.

From our previous work we determined that replicate analysis expanded the

coverage and increased the number of identified proteins(98). As can be seen in

Figure 4.2, there are proteins that are uniquely identified by only one replicate. This

is most likely due to the incomplete ionization of certain peptides during an LC-

MS/MS run. Many studies have shown that cell lines do represent the tumour from

which they originated. Hence, the proteins that they secrete should reflect the

genetic alterations that they harbor. Given that the biological triplicates yielded a

more complete coverage of the secretome for a cell line, it is highly probable that the

differences in the proteins identified among the 3 cell lines (Figure 4.3) indeed

reflects the heterogeneity of that cell line.

We confirmed the presence of two internal control proteins (PSA and KLK6)

in the CM at µg/L concentrations as measured by ELISA and identified by MS

(Figures 3.1, 3.2, 3.3). We were not able to identify by MS KLK5 which was also


123

present in PC3 at about the same concentration as KLK6 (2 µg/L). It is possible that

the stringent peptide and protein probability cut-offs utilized in this study was too

strict to allow identification of KLK5 by MS. Alternatively, it is possible that a

concentration of 2 µg/L is close to the detection limit of our methodology and hence

it was not identified. We thus conclude that our method can identify proteins in CM of

approx. low µg/L or higher, a detection limit which nevertheless is 2-3 orders of

magnitude better than the ones achieved by using serum(384). Furthermore, based

on the currently used biomarkers in the clinic, this is the expected concentration

range that potential tumour markers should be observed in serum.

The use of multiple search engines has been shown to increase confidence,

as well as expand coverage(385). In this study, Mascot and X!Tandem were used

since these search engines use different algorithms to determine if a mass spectrum

matches an entry in the database. The use of both search engines served to provide

an independent confirmation of the results. The use of peptide(386) and protein

prophet algorithms(387) contained within Scaffold allowed for increased confidence

of the protein identification probabilities. To further increase confidence, we

performed a search against a reversed IPI human database(388) and obtained FPR

of <1% for the cell lines by Mascot. The low FPR highlights the fidelity of the search

approach used. In addition, to eliminate contaminants left over from the FCS we

processed a control flask that did not contain any cells and deleted from the list of

proteins identified from the CaP cell lines the FCS-derived proteins.

Moreover, from the extracellular, membrane and unclassified proteins

identified in this study, 98, 93, and 26 proteins in extracellular, membrane and


124

unclassified protein lists, respectively were found in the plasma proteome. It is

possible that the other proteins are of low abundance and have not yet been

identified in circulation. This list was also compared against the list of proteins found

in seminal plasma(362). We reasoned that if a protein identified from our CaP cell

lines is also detected in seminal plasma, it is likely to be secreted or shed at

relatively high concentration by prostate cells. We found 108 extracellular, 120

membranous and 40 unclassified proteins in our CM, as well as in the seminal

plasma proteome.

To discern functional differences between the “secretome” of the three cell

lines, we performed a biological analysis of the extracellular and membrane proteins

identified from each cell line’s CM. First, we compared the extracellular and

membrane proteins identified between each cell line (Figure 4.4). We found 18%

and 20% of proteins identified as extracellular and membrane, respectively, were

common to all three cell lines. This is a significantly lower overlap than the overall

number of proteins which were common between the cell lines, indicating that there

are significant differences specifically within the “secretome” from each cell line. We

further investigated functional differences between the cell lines using the Ingenuity

Pathway Analysis. The top ranked networks were those involved with cell

movement, signalling, cancer and the cell cycle. Furthermore, 22Rv1 was not

represented as having any extracellular proteins involved with cellular movement,

whereas PC3 and LNCaP did. This is interesting and relevant since 22Rv1 was

derived from an organ-confined prostate tumour while PC3 and LNCaP were derived

from metastatic tumours. In addition, each cell line differs with its sensitivity to


125

androgens, a major component to prostate development: PC3 does not express the

AR and is a model for androgen insensitive tumours, 22Rv1 expresses the AR but is

considered AR insensitive, and LNCaP, which expresses AR, is considered

androgen dependent. In addition, we also investigated the common proteins in

breast CM identified by Kulasingam et al.(363) with those identified in this study to

look for similarities between these two hormonally regulated cancers. Here we see

that cancer and cell proliferation networks are among the top pathways identified to

be common. This commonality between the cell lines highlights that these tumours

utilize similar processes during carcinogenesis and warrants further study to

delineate biomarkers that could be useful for diagnosis.

Using information from this study, we developed criteria to narrow down our

list of candidate biomarkers: (a) we considered proteins that showed relatively

prostate-specific mRNA expression by searching through the UniGene expressed

sequence tag online database; (b) we selected proteins that were classified as

extracellular or membrane and have been identified in seminal plasma(362); (c) we

performed literature searches to ensure that these proteins have not been validated

before as CaP biomarkers and highlighted proteins that have shown biological links

to CaP and other cancers; (d) we then selected candidates that had commercially

available immunoassays. The list of narrowed down candidates are listed in Table

4.2.

From these, follistatin, CXCL16, PTX3 and SPON2 were chosen for further

validation based on the above criteria. Each candidate showed several links to

CaP(371,389-391) or carcinogenesis(377,392-397). Each candidate showed a


126

significant difference between patients with or without CaP. In addition, a positive

correlation with increasing PSA was also observed (Figure 4.10 B, D, F, H). From

these results, we speculate that these candidates show an association with CaP

progression. Future studies will determine if in combination they can improve the

specificity of the PSA test.

To determine the biological association with CaP we profiled each of the

candidate’s links to functions and diseases (Figure 4.7, 4.8). With the exception of

SPON2, each of the candidates displayed several links to cancer development,

tissue development, inflammation or chemotaxis. All of these processes have shown

to play a role in the malignant development of tumours. However, the involvement of

each candidate with respect to CaP pathobiology will need to be further studied to

elucidate their role during progression.

In summary, we present a robust method of proteomic analysis of cell culture

CM and bioinformatics for new biomarker discovery. The four candidates validated

have not been previously shown to be serum markers for CaP but require further

study to fully elucidate their roles in CaP progression. Additional candidates from this

large database are worth validating in the future.


127

Table 4.2: List of candidate biomarkers selected based on selection criteria.

Protein name Gene 64 kDa protein C1orf116Abhydrolase domain containing 14A ABHD14AADAMTS-1 precursor ADAMTS1Cartilage intermediate layer protein 2 precursor CILP2CEGP1 protein SCUBE2c-Myc-responsive protein Rcl C6orf108Creatine kinase B-type CKBchemokine (C-X-C motif) ligand 16 CXCL16Epiplakin EPPK1Epoxide hydrolase 2 EPHX2GAGE-2 protein GAGE2Galanin precursor GALglucosaminyl (N-acetyl) transferase 2, I-branching enzyme isoform A GCNT2Glycoprotein endo-alpha-1,2-mannosidase (Fragment) MANEAGroup 3 secretory phospholipase A2 precursor PLA2G3Growth-regulated protein alpha precursor CXCL1Hypothetical protein RPL22L1Hypothetical protein C1orf80 C1orf80Secretogranin-1 precursor CHGBSecretogranin-3 precursor SCG3Semaphorin 3F variant (Fragment) SEMA3FSplice Isoform 1 of Follistatin precursor FSTSplice Isoform 1 of Heme-binding protein 2 HEBP2Splice Isoform PTPS of Receptor-type tyrosine-protein phosphatase S precursor PTPRSSpondin-2 precursor SPON2von Willebrand factor A domain-related protein isoform 1 VWA1glucosaminyl (N-acetyl) transferase 2, I-branching enzyme isoform A GCNT2Splice Isoform 1 of Zinc transporter SLC39A6 precursor SLC39A6Neurexin-3-beta precursor NRXN3Tumor-associated calcium signal transducer 2 precursor TACSTD2protein tyrosine phosphatase, receptor type, sigma isoform 3 precursor PTPRSlatent transforming growth factor beta binding protein 1 isoform LTBP-1L LTBP1Pentraxin-related protein PTX3 precursor PTX3Splice Isoform 1 of Kallikrein-15 precursor KLK15Interferon-induced 17 kDa protein precursor ISG15Lysyl oxidase homolog 1 precursor LOXL1Spondin-2 precursor SPON264 kDa protein C1orf116CEGP1 protein SCUBE2ATP-binding cassette, sub-family C (CFTR/MRP), member 10 ABCC10hypothetical protein LOC255101 isoform 1 CCDC108PREDICTED: ephrin receptor EphA6 EPHA6Similar to Hepatocyte nuclear factor 1a dimerization cofactor isoform PCBD2von Willebrand factor A domain-related protein isoform 1 VWA1Tetraspanin-1 TSPAN1NGEP long variant TMEM16GSorbitol dehydrogenase SORD

Chapter 5: Validation of Candidate Biomarkers in Prostate Cancer Patient Samples

128

CHAPTER 5:

VALIDATION OF CANDIDATE BIOMARKERS IN PROSTATE CANCER PATIENT SAMPLES


128

5.1 Introduction

In addition to the four candidates validated in the previous chapter there were

four additional candidates that were selected for further validation. Based on the

criteria outlined in the previous chapter for the selection of candidates identified in

the CM of the CaP cell lines, a preliminary validation was preformed on chemokine

(C-X-C motif) ligand 5 (CXCL5), lipocalin-2 (LCN2), chemokine (C-X-C motif) ligand

1 (CXCL1) and tumor necrosis factor receptor superfamily, member 6b, decoy

(TNFRSF6B). However, the preliminary results did not qualify these candidates for

further study. Their association to cancer and pathological conditions are

summarized below.

Chemokine (C-X-C motif) ligand 5 also known as epithelial neutrophil-

activating protein 78 is a member of the CXC chemokine family and has been shown

to act as an inflammatory chemoattractant and activator of neutrophil function. It has

been seen to be elevated in serum of patients with rheumatoid arthritis, inflammatory

bowl conditions such as Crohn’s disease and ulcerative colitis. High expression of

CXCL5 was also seen in tissues of patients with chronic pancreatitis(398). Unlike

other chemokines, expression of CXCL5 is not limited to immune cells and is seen in

epithelial, endothelial and fibroblastic cell types. Expression of CXCL5 has been

associated with both CaP and gastrointestinal malignancies and related

inflammatory conditions(399-402).

Lipocalin 2 also known as neutrophil gelatinase-associated lipocalin is a 25

kDa secreted glycoprotein that is part of the lipocalin family of proteins and has been

shown to function in cell proliferation and survival. An immunohistochemical study


129

has shown the expression of LCN2 in a variety of tissues including tumour tissues of

the prostate, lung, colon, pancreas and kidney(403). In addition, LCN2 has been

studied as a marker for kidney injury(404) and ovarian cancer(405) and has shown

to be a poor prognostic indicator in breast cancer(406). The role of LCN2 in

carcinoma progression has also been studied and has shown to be promote breast

cancer and esophageal squamous cell carcinoma invasion by stabilizing

extracellular matrix enzymes(407,408) and inhibit cell invasion in colon cancer

cells(409).

Chemokine (C-X-C motif) ligand 1 also known as growth-related oncogene-α

like CXCL5 is part of the CXC family of chemokines. As with CXCL5, CXCL1 is also

elevated in serum of patients with inflammatory bowl conditions(410). Studies in

several different cancers have revealed that CXCL1 is a mediator of angiogenic and

tumourigenic activity in CaP cells(400,411) as well as has shown to promote the

malignant phenotype in colon cancer(412).

Tumor necrosis factor receptor superfamily, member 6b, decoy, also known

as decoy receptor 3, is a soluble receptor that is part of the tumour necrosis factor

receptor family. This soluble receptor has been shown to be overexpressed in some

cancers and thought to compete with binding Fas-L death receptors, thus inhibiting

apoptosis(413). Amplification of the TNFRSF6B gene has been shown in lung and

colon cancers(414), lymphomas(415), gastric cancer(416), glioblastoma(417) and

hepatocellular carcinoma(418). Serum levels of the soluble decoy receptor have

been measured in several malignancies and been shown to have diagnostic and

prognostic ability(397,419,420).


130

An extended validation of follistatin was also performed in patient serum and

tissue samples. From all the candidates previously validated, follistatin was

determined to be the strongest. Here we measure its levels in serum of patients with

biopsy confirmed CaP, BPH, PIN, inflammatory conditions and disease free. In

addition, we show immunohistochemical staining of follistatin in prostate, colon and

lung tumours.


Four candidates were validated in the biological fluids of patients with and

without CaP as well as other disease conditions. In addition, an extended validation

of follistatin was conducted in a larger patient cohort, as well as its levels in CaP and

other cancer tissues were determined by immunohistochemistry.

5.2.1 Conditioned Media, Serum Samples and Tissue Specimens

Aliquots from the CM of the PC3, LNCaP, 22Rv1 and the control flask were

used for measuring of CXCL5, LCN2, CXCL1 and TNFRSF6B. The cell lines were

grown at their optimized seeding densities as stated in the previous chapter.

Serum samples used for measurement of CXCL5, LCN2, CXCL1 and

TNFRSF6B were collected from patients at the Toronto Medical Laboratories,

Toronto, Canada. The study was carried out after Institutional Review Board

approval. Samples were screened based on PSA levels and levels above 10ng/mL

were grouped in the CaP group, while PSA levels below 1ng/mL were considered

normal. Serum samples used for measurement of follistatin, KLK2, PSA, fPSA and


131

KLK11 were collected from patients admitted to The Prostate Biopsy Clinic, in the

Prostate Centre at Princess Margaret Hospital, Toronto, for prostate biopsy.

Informed consent was obtained from each patient for participation in this study and

blood samples were collected prior to biopsy in yellow top serum separator tubes.

Whole blood was allowed to clot for 30 minutes at room temperature and

subsequently aliquoted and stored at -80oC until needed. Patients were diagnosed

with either CaP, BPH, PIN, inflammation or free of disease (Benign) based on biopsy

results. Data regarding age, PSA level, Gleason score, number of biopsy cores

taken, reason for biopsy, tumour location and if the patient had initial or recurrent

CaP was recorded. The mean ages for each group were: Benign – 60, BPH – 63,

CaP – 64, PIN – 60, Inflammation – 60. The majority of the patients in each group

had PSA levels in the gray zone (4-10ug/L).

Paraffin embedded tissue samples of prostate, lung and colon carcinomas

were obtained from the Toronto General Hospital, Toronto, Canada pathology

laboratory after institutional review board approval.

5.2.2 Quantification of Candidates in Biological Fluids

Chemokine (C-X-C motif) ligand 5

The CXCL5 sandwich ELISA kit was purchased from R&D Systems (Cat#:

DX000). The assay was performed using the kit protocol. Briefly, 200ul of assay

diluent was added to each pre-coated well in a 96 well plate followed by 50ul of

standard or sample in duplicate. The plate was incubated at room temperature for 2

hours followed by washing 3 times using an automated plate waster. Two hundred


132

microlitres of the secondary detection antibody was added to each well and the plate

was incubated at room temperature for 2 hours. Two hundred microlitres of

substrate solution was added to each well and the plate was covered with foil and

left for 30 minutes at room temperature. Fifty microlitres of stop solution was then

added to each well and the absorbance was read by a plate reader at 450nm

wavelength with a correction of 570nm.

Lipocalin-2

The human lipocalin-2 sandwich ELISA kit was purchased from R&D Systems

(Cat#:DLCN20) and was performed according to the kit protocol. The methodology

is similar to CXCL5 with some minor differences.

Chemokine (C-X-C motif) ligand 1

The human CXCL1 ELISA kit was purchased from R&D systems

(Cat#:DGR00) and was performed according to the kit protocol. The methodology is

similar to that of CXCL5 and LCN2.

Follistatin

The human follistatin ELISA kit was purchased from R&D systems

(Cat#:DFN00) and was performed according to the kit protocol. The methodology is

similar to that of CXCL5, LCN2 and CXCL1.

Tumor necrosis factor receptor superfamily, member 6b, decoy

The human tumor necrosis factor receptor superfamily, member 6b, decoy

(TNFRSF6B) ELISA kit was purchased from R&D systems (Cat#:DY142) and was

performed according to the kit protocol with some minor modifications. Briefly,

200uL of the capture antibody was coated overnight at a concentration of 4ug/mL.


133

The antibody was aspirated off and 100uL of standard or sample was aliquoted in

duplicate and incubated shaking at room temperature for 2 hours. The plate was

then washed 6 times and the biotinylated detection antibody was added and

incubated for 2 hours shaking at room temperature. The plate was washed 6 times

and a streptavidin-alkaline phosphatase conjugate was added and incubated for 15

minutes. The plate was washed 6 times again and 100ul of a difusinal phosphate

solution was added for 10 mins followed by 100ul of developing solution. The plate

was read by time-resolved fluorescence.

Prostate specific antigen, Kallikrein-related peptidase 2, Kallikrein-related peptidase

11

The assays for human PSA, KLK2 and KLK11 were conducted using in-

house developed ELISAs and were conducted as published

previously(198,331,421).

Free PSA

Free PSA was measured using the Elecsys modular analytic system by

Roche at Mt. Sinai Hospital, Toronto. Briefly, 20uL of sample and two monoclonal

antibodies, one biotinylated and one labelled with a ruthenium complex react to form

a complex. Streptavidin-coated microparticles are added to the mixture and bind the

complex. The mixture is aspirated and dispensed into a measuring cell where the

microparticles are capture by a magnet onto an electrode. Unbound substances are

removed and a voltage is applied to the electrode inducing chemiluminescence that

is measured by a photomultiplier. The duration of the assay is 18 minutes.


134

5.2.3 Immunohistochemistry of Follistatin

Paraffin sections were dewaxed in 5 changes of xylene and brought down

with water through graded alcohols. Sections were then heat-retrieved in Tris-EDTA

buffer at pH 9.0 for 2 minutes at 120°C using a decloaking chamber (Biocare).

Sections were allowed to cool off at room temperature for 20 minutes before

washing well in running tap water. Endogenous peroxidase and biotin activities were

blocked respectively using 3% hydrogen peroxide and an avidin/biotin blocking kit

(Lab Vision, Cat# TA-015-BB). Sections were then treated for 10 mins with 10%

normal horse serum (Vector Labs, Cat# S2000) before incubated overnight with the

mouse monoclonal anti-Follistatin antibody (R&D, Clone 85918, Cat# MAB669) at

1/200 in a moist chamber. This was followed by 30 mins each with biotinylated horse

anti-mouse IgG (Vector Labs. Inc. Cat# BA-2001) and HRP-conjugated Ultra

Streptavidin (ID Labs. Inc. Cat# BP2378). Colour development was done with freshly

prepared NovaRed solution (Vector Labs. Inc: Cat# SK-4800) and counterstained

with Mayer’s hematoxylin. Finally, sections were dehydrated through graded

alcohols, cleared in xylene and mounted in Permount (Fisher: Cat# SP15-500).

5.2.4 Statistical Data Analysis

To determine if there were significant differences between the median serum

levels measured for normal and cancer patient groups tested for CXCL5, LCN2,

CXCL1 and TNFRSF6B the Mann-Whitney non-parametric test was employed using

95% confidence intervals and two-sided P values. To determine if there were

differences between the multiple groups measured by follistatin, KLK2, PSA, fPSA


135

and KLK11 the Kruskal-Wallis test was employed using 95% confidence intervals. A

Dunn’s multiple comparisons test was performed post-hoc to determine if there were

any differences between individual groups. A Spearman correlation was utilized with

two-tailed P values and 95% confidence intervals to determine if there was an

association between follistatin with age, PSA or Gleason score. All tests were

conducted using GraphPad Prism software (GraphPad Software Inc.).


136

5.3 Results

5.3.1 Chemokine (C-X-C motif) ligand 5

The CXCL5 protein was identified by MS in the CM of the PC3 cell line with

both Mascot and X!Tandem proteomic search engines. Two unique peptides and 8

unique spectra were identified by each search engine. To confirm our MS results we

measured the concentration of CXCL5 in the CM of each cell line and control flask

by ELISA (Figure 5.1A). As can bee seen our ELISA data corroborates with our MS

data as CXCL5 was only detected in the PC3 CM by both methods. An initial

validation of CXCL5 was also conducted in serum samples from patients considered

to not harbour CaP and have CaP (Figure 5.1B). From this preliminary validation we

observed no statistical difference in median values between the two groups

(p=0.7493). As a result, this candidate was not evaluated further.


137

Figure 5.1

A

B

Figure 5.1: Measurement of CXCL5 in CM of CaP cell lines and serum of CaP patients and normals. (A) CXCL5 is measured by ELISA in CM of each cell line and only identified in PC3 CM. Number of spectra identified by Mascot and X!Tandem are also shown. (B) Levels of CXCL5 in serum of normal and CaP patients. Horizontal bars represent median values. P value calculated by Mann-Whitney test.


138

5.3.2 Chemokine (C-X-C motif) ligand 1

Another candidate preliminarily validated was CXCL1. Similarly CXCL1 was

only identified in the CM of the PC3 cell line by MS. The MS results were validated

by measuring CXCL1 concentrations by ELISA (Figure 5.2A). The number of unique

spectra identified by Mascot and X!Tandem were 30 and 28 respectively, with 3 and

4 unique peptides identified by each search engine. The concentrations of CXCL1

were measured in the serum samples of men with and without CaP (Figure 5.2B),

where there was no significant difference determined by the Mann-Whitney test

(p=1.0000). Likewise this candidate was not studied further.


139

Figure 5.2

A

B

Figure 5.2: Measurement of CXCL1 in CM of CaP cell lines and serum of CaP patients and normals. (A) CXCL1 is measured by ELISA in CM of each cell line and only identified in PC3 CM. Number of spectra identified by Mascot and X!Tandem are also shown. (B) Levels of CXCL1 in serum of normal and CaP patients. Horizontal bars represent median values. P value calculated by Mann-Whitney test.


140

5.3.3 Lipocalin 2

Lipocalin 2 was identified in the CM of two of the CaP cell lines, PC3 and

LNCaP by MS. This was validated by measuring the concentration of LCN2 in the

CM of each cell line and the control flask. Once again the spectral counts derived

from Mascot and X!Tandem agreed with our ELISA data (Figure 5.3A). In addition,

LCN2 was also a protein identified by Pilch et al. in seminal plasma(362). We

examined the serum concentrations of LCN2 in the serum of patients with and

without CaP (Figure 5.3B). No significant difference was observed between groups

and the protein was not examined further.


141

Figure 5.3

A

B

Figure 5.3: Measurement of LCN2 in CM of CaP cell lines and serum of CaP patients and normals. (A) LCN2 is measured by ELISA in CM of each cell line and identified in PC3 and LNCaP CM. Number of spectra identified by Mascot and X!Tandem are also shown. (B) Levels of LCN2 in serum of normal and CaP patients. Horizontal bars represent median values. P value calculated by Mann-Whitney test.


142

5.3.4 Tumor necrosis factor receptor superfamily, member 6b, decoy

The final candidate investigated, TNFRSF6B was identified solely in the CM

of the PC3 CaP cell line by 1 peptide and 7 unique spectra each by Mascot and

X!Tandem. Results from the ELISA measurement of TNFRSF6B in the CM of each

cell line and control flask also agree with our MS data (Figure 5.4A). Serum

samples from patients with and without CaP were measured for TNFRSF6B by

ELISA and the results are shown in Figure 5.4B. As can be seen no significant

difference was observed between the median values of each group and this

candidate was not investigated further.


143

Figure 5.4

Figure 5.4: Measurement of TNFRSF6B in CM of CaP cell lines and serum of CaP patients and normals. (A) TNFRSF6B is measured by ELISA in CM of each cell line and identified in PC3 CM. Number of spectra identified by Mascot and X!Tandem are also shown. (B) Levels of TNFRSF6B in serum of normal and CaP patients. Horizontal bars represent median values. P value calculated by Mann-Whitney test.


144

5.3.5 Extended Validation of Follistatin

Measurement of follistatin in the serum of patients pre-biopsy was assessed

(Figure 5.5). Median levels of follistatin in each group are shown by horizontal lines

and were analyzed by the Kruskal-Wallis non-parametric one-way ANOVA test and

no significant difference was observed between the groups (p = 0.9521) (Figure

5.5A). To determine if there was a correlation of follistatin in patients diagnosed with

CaP with PSA, age or Gleason score a Spearman correlation was performed (Figure

5.5B, C, D respectively). No significant correlation was observed. In addition, post-

hoc analysis of the 95% confidence intervals showed overlapping intervals for each

median value, thus futher confirming the validity of our result.

The concentrations of PSA, %fPSA, KLK2 and KLK11 were also measured in

the same patient cohort (Figure 5.6). The Kruskal-Wallis non-parametric one-way

ANOVA test was performed. The only significant difference that was observed was

with KLK11 (Figure5.6D). When examined further by post-hoc pair-wise analysis of

the groups using the Mann-Whitney test it was determined that there were significant

differences between the benign and CaP group (p=0.0100) and the BPH and CaP

group (p=0.0085).

Follistatin expression was also preliminarily investigated in the tissues of

prostate, colon and lung cancers by immunohistochemical staining (Figure 5.7, 5.8.

5.9). In the prostate, staining was observed in the luminal epithelial cells and not in

the stroma of the benign, PIN, BPH and cancerous glands. All glands stained

positive however, staining of prostatic carcinoma (Figure 5.7A) in particular high

grade carcinoma (Figure 5.7E) was seen to be more intense relative to benign and


145

BPH glands (Figure 5.7A, B, D). Interestingly, certain basal cells of benign (Figure

5.7B) and BPH glands displayed a more intense staining (Figure 5.7D). In addition,

PIN lesions were seen to stain equally to low grade carcinoma (Figure 5.7C).

Follistatin staining was also evaluated in lung tissue and was localized to the

cytoplasm of normal bronchial epithelial cells (Figure 5.8A), pulmonary carcinoma

(Figure 5.8C) as well as pulmonary squamous carcinoma (Figure 5.8D, E).

Macrophages in the lung were also seen to stain positive (Figure 5.8B). In addition,

one out of four specimens from different patients of pulmonary squamous carcinoma

was seen to stain intensely (Figure 5.8E) with staining seen in the cytoplasm as well

as in the surrounding membrane of the epithelial cells. Tissue specimens from

normal and cancerous colon were also investigated for follistatin staining (Figure

5.9). Here we see a more intense stain in the luminal mucosal cells of the colon

carcinoma (Figure 5.9C, D) versus the normal mucosal epithelium (Figure 5.9A, B).


146

Figure 5.5

Figure 5.5: Follistatin serum levels and correlations with PSA, Gleason score and age. (A) Concentrations of follistatin in patients with benign, BPH, Inflammation, PIN and CaP. Horizontal bars represent median levels. P value calculated by Kruskal-Wallis one-way ANOVA test. (B, C, D) Correlation of follistatin with PSA Gleason score and age in CaP patients, respectively. r = Spearman coefficient, P values are calculated using the non-parametric Spearman correlation.


147

Figure 5.6

Figure 5.6: PSA, %fPSA, KLK2 and KLK11 serum concentrations in patient serum. (A, B, C, D) Serum concentrations of PSA, %fPSA, KLK2 and KLK11 respectively in serum of patients with benign, BPH, inflammation, PIN and CaP disease. Horizontal lines represent median values. P values are calculated using the Kruskal-Wallis non-parametric one-way ANOVA test.


148

Figure 5.7

Figure 5.7: Prostate tissue specimens stained for follistatin. (A) CaP specimen containing benign gland and cancerous gland with perineural invasion (CaP-PNI). (B) Benign and cancerous prostate glands, basal cell staining is evident. (C) CaP specimen displaying PIN lesions and cancerous glands. (D) Prostate specimen displaying a BPH gland with evident basal cell staining (E) Specimen of high grade carcinoma (HG-CaP).


149

Figure 5.8

Figure 5.8 Lung tissue specimens stained for follistatin. (A) Normal lung bronchial epithelium (BrEp). (B) Normal lung alveoli, macrophages (Mac) are seen to stain positive for follistatin. (C) Specimen of pulmonary carcinoma. (D, E) Specimens of pulmonary squamous carcinoma, Intense staining of follistatin is seen in the latter specimen, with concentrations of intense membranous staining.


150

Figure 5.9

Figure 5.9: Colon tissue specimens stained for follistatin. (A, B) Normal colon mucosa. (C, D) Specimens of colon cancer. Staining of follistatin is evident in the cytoplasm of the columnar mucosal epithelium.


151

5.4 Discussion

There have been significant efforts to stop the progression of CaP by

detecting it at an early stage, but these have been curtailed by lack of a sensitive

and specific marker for the disease when it is organ-confined. In addition, there is a

lack of an effective therapy that does not involve significant side effects and cause

morbidity. Improvements in the prognosis of patients with organ-confined disease

are needed to determine those patients that require active management. Therefore,

the importance in delineating the role of growth and regulatory mechanisms in CaP

translates into targeting the disease earlier through innovative therapies and

improved diagnostics.

While the role of androgens in stimulation of CaP is well studied, the roles of

associated growth factors are still emerging. Growth factors are seen as playing

almost independent roles to androgen stimulation and provide potential targets for

therapy for AIPC. One family of proteins, the transforming growth factor β (TGFβ)

family have been implicated in progression of many cancers and are key in acting as

tumour suppressors(422). Development of resistance to this pathway is an important

event in malignant transformation(423). While the involvement of TGFβ has been

well determined, the involvement of other members of the family such as follistatin is

less well known.

Follistatin is a glycosylated secreted protein that is part of a larger group of

proteins with a conserved “follistatin” domain that is rich in cysteines. Follistatin has

been shown to be alternatively spliced with two isoforms of proteins FS288 and

FS315. The biological function of follistatin was first shown to act to inhibit the


152

secretion of FSH by binding to proteins known as activins(424). The activins along

with inhibins regulate production of gonadotropins with activins stimulating

production and inhibins inhibiting production. While follistatin also binds to inhibin it

does so in a weaker fashion than to activin where the binding is irreversible(425).

The action of follistatin is not limited to the FSH secretion axis as it has been shown

to be expressed in a variety of tissues(426) as well as shown to bind other

molecules such as bone morphogenic proteins(427). With respect to CaP follistatin

has been shown to suppress the growth inhibiting effects of activins.

The localization of follistatin in prostate tissues has been shown in normal,

BPH and CaP. Using a polyclonal antibody Thomas et al.(390) were able to localize

follistatin in the stroma, basal and epithelial cells of high grade CaP. Expression in

BPH was also determined by the same group to be present in the stromal

tissue(428). While our study results do not show stromal staining, our results do

confirm basal cell staining in the prostate of some benign and BPH glands (Figure

5.7B, D) as well as staining of the epithelial cells in high grade CaP (Figure 5.7E).

While the immunohistochemical staining of follistatin in the prostate does not show

any obvious diagnostic significance, the interactions between its ligand activin has

been shown to play a role in CaP progression.

Co-localisation of activin and follistatin in the prostate support the theory that

follistatin reduces the growth inhibitory effects of activin(428). In addition, studies in

cell lines have demonstrated the direct effects of follistatin’s ability to neutralise the

action of activin(371,372). The PC3 cell line has been shown to secrete both activin

and follistatin(371), which was confirmed by our MS and ELISA data. In the PC3 cell


153

line the inhibitory effect of activin is thought to be neutralised by follistatin in an

autocrine manner that promotes cell growth. This was tested by treating PC3 cells

with a neutralising antibody to follistatin and adding activin. This resulted in a 60%

decrease in cell proliferation thus confirming the theory of follistatin’s neutralising

ability over activin(371).

Follistatin has also been shown to have a heparin-binding region and have

been shown to interact with the heparin-sulfate chains on proteglycans on the cell

surface(429). This was thought to increase its ability to bind activin and sequester it

and prevent it from binding its receptor(430). However, another mechanism showed

the activin-follistatin complex being internalised and degraded by lysosomes(431).

The localization of follistatin to the cell membrane was apparent in one specimen

from lung cancer (Figure 5.8E).

While follistatin has been shown to be expressed in several tissues(426),

immunohistochemical location in colon and lung tissues has not been shown.

Studies with respect to lung cancer have shown follistatin to play an inhibitory role in

the progression of metastasis of lung tumours(432). Similar to prostate, the

cancerous epithelial cells of both lung and colon stain more intensely than the

normal epithelium (Figure 5.8, 5.9). One specimen of lung cancer displayed an

intense staining in the epithelial tumour cells. The prognostic significance of this will

need to be evaluated by further study to elucidate the prognostic potential of

follistatin as a lung tissue biomarker.

While there have been studies on the expression of follistatin in CaP there

have not been any determining its potential as a prostate biomarker in serum given


154

that it has shown to play a role in promoting CaP progression. Our preliminary

validation of follistatin in the serum of patients with and without CaP (Figure 4.5A)

showed follistatin to have some discriminatory ability. However, this was not

confirmed in an extended validation conducted with biopsy confirmed patients

(Figure 5.5). The contradictory nature of the results is most likely due to the patients

used in the preliminary validation were selected from those who had PSA values

greater than 10ug/L. These patients in all probability harbour clinically advanced

CaP and those with highly elevated PSA greater that 50ug/L are likely to have

metastases. Measuring follistatin in this study set confirmed that follistatin is

elevated in a patient population with elevated PSA and that it also correlates

positively with increasing PSA (Figure 4.5B). The significance of these results were

not seen in serum of patients with biopsy confirmed CaP. In the extended patient

population the range of PSA values were mostly in the grey region of 4-10ug/L and

the tumours identified by biopsy had Gleason scores in the range of 5-9.

Correlations with PSA and Gleason score also did not show any significance with

follistatin measured in serum (Figure 5.7B, C). This could be attributed to the fact

that these are early stage tumours and levels of follistatin leaking into the circulation

are not enough to be measured by ELISA. To ensure there was no bias with age a

correlation of serum follistatin concentrations and age of CaP was shown to also be

non-significant.

The challenge of discriminating CaP from benign and BPH is seen even with

the use of PSA and %fPSA in this population (Figure 5.8A, B). There was no

significant difference seen between the groups using PSA or %fPSA. In addition,


155

two well studied biomarkers for CaP, KLK2 and KLK11 were also measured (Figure

5.8C, D). In this case KLK11 was the only one to show a significant difference

between the benign and CaP groups and BPH and CaP groups. However, the

clinical relevance of this distinction is minimal considering there is significant overlap

in the levels of KLK11 measured in each group.

In conclusion, we show that follistatin is unable to distinguish CaP patients

from benign and inflammatory conditions in a patient cohort with PSA levels in the

grey region. However, we show for the first time the expression of follistatin in

tissues of colon and lung cancer with intense staining in one specimen of lung

squamous carcinoma. The clinical significance of this finding will need to be

elucidated with further studies in lung tissue specimens and serum samples of

patients with and without lung cancer.

Chapter 6: Summary and Future Directions

156

CHAPTER 6:

SUMMARY AND FUTURE DIRECTIONS

Chapter 6: Summary and Future Directions 157

6.1 Summary

This thesis presents an optimized and validated approach to the discovery of

novel candidate CaP biomarkers through proteomic analysis of CaP cell line CM.

The use of CaP cell lines as model systems for the study of their secreted proteins

by MS approaches resulted in the identification of hundreds of proteins. These

proteins were assessed by literature and bioinformatics searches to elucidate their

potential as candidates for validation. Select candidates were validated in the serum

and tissues of patients with and without CaP.

6.2 Key Findings

1. Proof of Principle

a. Established a procedure for long term growth of the CaP cell line

PC3(AR)6 in large volume roller bottle culture in SFM

b. Processed CM in a suitable way for strong anion exchange fast

performance liquid chromatography

c. Prepared fractions for MS analysis by C18 reversed phase HPLC

coupled to a linear ion trap mass spectrometer.

d. Identified 262 proteins from combining proteins identified in two

replicates of culture by searching mass spectra with MASCOT

e. Measured the novel candidate Mac-2BP in the serum of CaP, BPH and

normal patients


2. Optimization of Culture Conditions

a. Seeding density for each of the three CaP cell lines used (PC3, LNCaP

and 22Rv1) was varied to optimize the amount of secreted proteins vs.

intracellular proteins. Marker proteins for secreted and intracellular

fractions were compared and total protein measured to determine

optimal seeding densities.

3. Comparative Proteomic Analysis of CM of Three CaP Cell Lines

a. Cultured and harvested the CM from each CaP cell line in triplicate

b. Concentrated and prepared the CM for each replicate via peptide

fractionation by HPLC SCX chromatography

c. Analyzed each fraction by MS/MS via C18 reversed phase capillary

electrophoresis coupled online to a linear ion trap mass spectrometer

d. Identified 2124 proteins by searching mass spectra with MASCOT and

X! Tandem proteomic search databases and combining results through

Scaffold

e. Determined candidate proteins for validation through a set of criteria

4. Validation of Candidate Biomarkers

a. Novel candidates were pre-clinically validated in serum of CaP and

healthy males.

b. Four novel candidates were selected for further pre-clinical validation

in serums of CaP and healthy males, each showed a positive


correlation with PSA serum levels and a statistically significant

increase in serum of CaP patients.

i. Chemokine (C-X-C) motif-16

ii. Spondin 2

iii. Pentraxin 3

iv. Follistatin

c. Follistatin was validated in an extended serum sample set along with

KLK2, KLK11 and %fPSA. No discriminatory ability was seen with any

marker.

d. Follistatin was evaluated as a tissue marker in the tissues of men with

high and low grade CaP and PIN and BPH and was not shown to offer

any discriminatory utility.

e. Evaluation of follistatin in the tissues of colon and lung cancers

revealed and intense positive staining in one lung cancer specimen

6.3 Future Directions

The approach of this study yielded several novel candidates for validation as

CaP biomarkers. The most promising of which was follistatin which showed slight

discriminatory staining between CaP and BPH glands. However, even though the

clinical utility of this as a tissue marker is not significant follistatin may play a role in

CaP progression and further elucidation of its role in CaP is required to determine

this. This would involve transcript levels and post translational modification analysis.

In addition, follistatin was shown to be expressed in other tissues and we show


staining in other cancers. Thus, more information on the role of follistatin in these

cancers is also required.

While we show that CXCL16, SPON2 and PTX3 show significant differences

between CaP and healthy males, their roles in CaP progression have not been fully

elucidated as well as their roles as tissue markers.

There is a need to examine the list of proteins produced in this study to a

greater extent to extract more novel candidates that would require validation. There

were several promising candidates that were not validated based on a lack of an

ELISA. Developments of antibodies and an ELISA to these candidates would be of

value. In addition, development of MS methods to quantify these candidates using

techniques such as multiple reaction monitoring would aid in their validation.

Further study of more CaP cell lines would add to the coverage of the

‘secretome’ of CaP and would aid in uncovering more candidates for validation. It is

likely that there will not be one biomarker for all tumours and thus a multi-parametric

panel is more than likely required to properly assess risk of CaP. Quantitative MS

proteomic comparisons of tumour tissue from matched tissue samples from tumour

and normal tissue by using isotopic tagging reagents such as iTRAQ would aid in

uncovering proteins that show quantitative differences between normal and CaP

specimens.

References

161

REFERENCES

References 162

Reference List

1. Jemal A, Siegel R, Ward E, Murray T, Xu J, Thun MJ. Cancer statistics, 2007. CA Cancer J Clin 2007;57:43-66.

2. Miller DC, Hafez KS, Stewart A, Montie JE, Wei JT. Prostate carcinoma presentation, diagnosis, and staging: an update form the National Cancer Data Base. Cancer 2003;98:1169-78.

3. Bill-Axelson A, Holmberg L, Ruutu M, Haggman M, Andersson SO, Bratell S et al. Radical prostatectomy versus watchful waiting in early prostate cancer. N Engl J Med 2005;352:1977-84.

4. Foley R, Hollywood D, Lawler M. Molecular pathology of prostate cancer: the key to identifying new biomarkers of disease. Endocr Relat Cancer 2004;11:477-88.

5. Canadian Cancer Society. Canadian Cancer Statistics. 1-4-2008. Ref Type: Report

6. Sakr WA, Grignon DJ, Crissman JD, Heilbrun LK, Cassin BJ, Pontes JJ, Haas GP. High grade prostatic intraepithelial neoplasia (HGPIN) and prostatic adenocarcinoma between the ages of 20-69: an autopsy study of 249 cases. In Vivo 1994;8:439-43.

7. Muir CS, Nectoux J, Staszewski J. The epidemiology of prostatic cancer. Geographical distribution and time-trends. Acta Oncol 1991;30:133-40.

8. Miller BA, Kolonel LN, and Berstein L. Racial/ethnic patterns of cancer in the United States 1988–1992. NIH Pub. No. 96-4101. 1996. National Cancer Institute.

Ref Type: Report

9. Shimizu H, Ross RK, Bernstein L, Yatani R, Henderson BE, Mack TM. Cancers of the prostate and breast among Japanese and white immigrants in Los Angeles County. Br J Cancer 1991;63:963-6.

10. Whittemore AS, Wu AH, Kolonel LN, John EM, Gallagher RP, Howe GR et al. Family history and prostate cancer risk in black, white, and Asian men in the United States and Canada. Am J Epidemiol 1995;141:732-40.

11. Cui J, Staples MP, Hopper JL, English DR, McCredie MR, Giles GG. Segregation analyses of 1,476 population-based Australian families affected by prostate cancer. Am J Hum Genet 2001;68:1207-18.

References 162

12. Chang BL, Zheng SL, Hawkins GA, Isaacs SD, Wiley KE, Turner A et al. Polymorphic GGC repeats in the androgen receptor gene are associated with hereditary and sporadic prostate cancer risk. Hum Genet 2002;110:122-9.

13. Nam RK, Toi A, Vesprini D, Ho M, Chu W, Harvie S et al. V89L polymorphism of type-2, 5-alpha reductase enzyme gene predicts prostate cancer presence and progression. Urology 2001;57:199-204.

14. Hsing AW, Reichardt JK, Stanczyk FZ. Hormones and prostate cancer: current perspectives and future directions. Prostate 2002;52:213-35.

15. Armstrong B, Doll R. Environmental factors and cancer incidence and mortality in different countries, with special reference to dietary practices. Int J Cancer 1975;15:617-31.

16. De Stefani E, Deneo-Pellegrini H, Boffetta P, Ronco A, Mendilaharsu M. Alpha-linolenic acid and risk of prostate cancer: a case-control study in Uruguay. Cancer Epidemiol Biomarkers Prev 2000;9:335-8.

17. Chan JM, Stampfer MJ, Ma J, Gann PH, Gaziano JM, Giovannucci EL. Dairy products, calcium, and prostate cancer risk in the Physicians' Health Study. Am J Clin Nutr 2001;74:549-54.

18. Strom SS, Yamamura Y, Duphorne CM, Spitz MR, Babaian RJ, Pillow PC, Hursting SD. Phytoestrogen intake and prostate cancer: a case-control study using a new database. Nutr Cancer 1999;33:20-5.

19. Giovannucci E, Rimm EB, Liu Y, Stampfer MJ, Willett WC. A prospective study of tomato products, lycopene, and prostate cancer risk. J Natl Cancer Inst 2002;94:391-8.

20. Lippman SM, Goodman PJ, Klein EA, Parnes HL, Thompson IM, Jr., Kristal AR et al. Designing the Selenium and Vitamin E Cancer Prevention Trial (SELECT). J Natl Cancer Inst 2005;97:94-102.

21. Blennerhassett JB, Vickery AL, Jr. Carcinoma of the prostate gland. An anatomical study of tumor location. Cancer 1966;19:980-4.

22. El Alfy M, Pelletier G, Hermo LS, Labrie F. Unique features of the basal cells of human prostate epithelium. Microsc Res Tech 2000;51:436-46.

23. Taplin ME, George DJ, Halabi S, Sanford B, Febbo PG, Hennessy KT et al. Prognostic significance of plasma chromogranin a levels in patients with hormone-refractory prostate cancer treated in Cancer and Leukemia Group B 9480 study. Urology 2005;66:386-91.

References 163

24. Yashi M, Nukui A, Kurokawa S, Ochi M, Ishikawa S, Goto K et al. Elevated serum progastrin-releasing peptide (31-98) level is a predictor of short response duration after hormonal therapy in metastatic prostate cancer. Prostate 2003;56:305-12.

25. Nelson WG, De Marzo AM, Isaacs WB. Prostate cancer. N Engl J Med 2003;349:366-81.

26. Bostwick DG. Prospective origins of prostate carcinoma. Prostatic intraepithelial neoplasia and atypical adenomatous hyperplasia. Cancer 1996;78:330-6.

27. Bostwick DG, Brawer MK. Prostatic intra-epithelial neoplasia and early invasion in prostate cancer. Cancer 1987;59:788-94.

28. Sarasin A. An overview of the mechanisms of mutagenesis and carcinogenesis. Mutat Res 2003;544:99-106.

29. Steinberg GD, Carter BS, Beaty TH, Childs B, Walsh PC. Family history and the risk of prostate cancer. Prostate 1990;17:337-47.

30. Lichtenstein P, Holm NV, Verkasalo PK, Iliadou A, Kaprio J, Koskenvuo M et al. Environmental and heritable factors in the causation of cancer--analyses of cohorts of twins from Sweden, Denmark, and Finland. N Engl J Med 2000;343:78-85.

31. Nupponen NN, Carpten JD. Prostate cancer susceptibility genes: many studies, many results, no answers. Cancer Metastasis Rev 2001;20:155-64.

32. Carpten J, Nupponen N, Isaacs S, Sood R, Robbins C, Xu J et al. Germline mutations in the ribonuclease L gene in families showing linkage with HPC1. Nat Genet 2002;30:181-4.

33. Xu J, Zheng SL, Komiya A, Mychaleckyj JC, Isaacs SD, Hu JJ et al. Germline mutations and sequence variants of the macrophage scavenger receptor 1 gene are associated with prostate cancer risk. Nat Genet 2002;32:321-5.

34. DeMarzo AM, Nelson WG, Isaacs WB, Epstein JI. Pathological and molecular aspects of prostate cancer. Lancet 2003;361:955-64.

35. Gonzalgo ML, Isaacs WB. Molecular pathways to prostate cancer. J Urol 2003;170:2444-52.

36. Visakorpi T. The molecular genetics of prostate cancer. Urology 2003;62:3-10.

References 164

37. De Marzo AM, DeWeese TL, Platz EA, Meeker AK, Nakayama M, Epstein JI et al. Pathological and molecular mechanisms of prostate carcinogenesis: implications for diagnosis, detection, prevention, and treatment. J Cell Biochem 2004;91:459-77.

38. Lee WH, Morton RA, Epstein JI, Brooks JD, Campbell PA, Bova GS et al. Cytidine methylation of regulatory sequences near the pi-class glutathione S-transferase gene accompanies human prostatic carcinogenesis. Proc Natl Acad Sci U S A 1994;91:11733-7.

39. Cordon-Cardo C. Mutations of cell cycle regulators. Biological and clinical implications for human neoplasia. Am J Pathol 1995;147:545-60.

40. Whang YE, Wu X, Suzuki H, Reiter RE, Tran C, Vessella RL et al. Inactivation of the tumor suppressor PTEN/MMAC1 in advanced human prostate cancer through loss of expression. Proc Natl Acad Sci U S A 1998;95:5246-50.

41. Scher HI, Sarkis A, Reuter V, Cohen D, Netto G, Petrylak D et al. Changing pattern of expression of the epidermal growth factor receptor and transforming growth factor alpha in the progression of prostatic neoplasms. Clin Cancer Res 1995;1:545-50.

42. Uehara H, Kim SJ, Karashima T, Shepherd DL, Fan D, Tsan R et al. Effects of blocking platelet-derived growth factor-receptor signaling in a mouse model of experimental prostate cancer bone metastases. J Natl Cancer Inst 2003;95:458-70.

43. Cardillo MR, Monti S, Di Silverio F, Gentile V, Sciarra F, Toscano V. Insulin-like growth factor (IGF)-I, IGF-II and IGF type I receptor (IGFR-I) expression in prostatic cancer. Anticancer Res 2003;23:3825-35.

44. Tu WH, Thomas TZ, Masumori N, Bhowmick NA, Gorska AE, Shyr Y et al. The loss of TGF-beta signaling promotes prostate cancer metastasis. Neoplasia 2003;5:267-77.

45. Osman I, Scher HI, Drobnjak M, Verbel D, Morris M, Agus D et al. HER-2/neu (p185neu) protein expression in the natural or treated history of prostate cancer. Clin Cancer Res 2001;7:2643-7.

46. Signoretti S, Montironi R, Manola J, Altimari A, Tam C, Bubley G et al. Her-2-neu expression and progression toward androgen independence in human prostate cancer. J Natl Cancer Inst 2000;92:1918-25.

References 165

47. Edwards J, Krishna NS, Witton CJ, Bartlett JM. Gene amplifications associated with the development of hormone-resistant prostate cancer. Clin Cancer Res 2003;9:5271-81.

48. Drobnjak M, Osman I, Scher HI, Fazzari M, Cordon-Cardo C. Overexpression of cyclin D1 is associated with metastatic prostate cancer to bone. Clin Cancer Res 2000;6:1891-5.

49. Paradis V, Dargere D, Laurendeau I, Benoit G, Vidaud M, Jardin A, Bedossa P. Expression of the RNA component of human telomerase (hTR) in prostate cancer, prostatic intraepithelial neoplasia, and normal prostate tissue. J Pathol 1999;189:213-8.

50. Kaur P, Kallakury BS, Sheehan CE, Fisher HA, Kaufman RP, Jr., Ross JS. Survivin and Bcl-2 expression in prostatic adenocarcinomas. Arch Pathol Lab Med 2004;128:39-43.

51. Buchanan G, Irvine RA, Coetzee GA, Tilley WD. Contribution of the androgen receptor to prostate cancer predisposition and progression. Cancer Metastasis Rev 2001;20:207-23.

52. Linja MJ, Savinainen KJ, Saramaki OR, Tammela TL, Vessella RL, Visakorpi T. Amplification and overexpression of androgen receptor gene in hormone-refractory prostate cancer. Cancer Res 2001;61:3550-5.

53. Buchanan G, Yang M, Harris JM, Nahm HS, Han G, Moore N et al. Mutations at the boundary of the hinge and ligand binding domain of the androgen receptor confer increased transactivation function. Mol Endocrinol 2001;15:46-56.

54. Aihara M, Lebovitz RM, Wheeler TM, Kinner BM, Ohori M, Scardino PT. Prostate specific antigen and gleason grade: an immunohistochemical study of prostate cancer. J Urol 1994;151:1558-64.

55. Lee F, Gray JM, McLeary RD, Meadows TR, Kumasaka GH, Borlaza GS et al. Transrectal ultrasound in the diagnosis of prostate cancer: location, echogenicity, histopathology, and staging. Prostate 1985;7:117-29.

56. Gleason DF, Mellinger GT. Prediction of prognosis for prostatic adenocarcinoma by combined histological grading and clinical staging. J Urol 1974;111:58-64.

57. Graham J, Baker M, Macbeth F, Titshall V. Diagnosis and treatment of prostate cancer: summary of NICE guidance. BMJ 2008;336:610-2.

58. Lu-Yao GL, Yao SL. Population-based study of long-term survival in patients with clinically localised prostate cancer. Lancet 1997;349:906-10.

References 166

59. Catalona WJ, Smith DS, Ratliff TL, Basler JW. Detection of organ-confined prostate cancer is increased through prostate-specific antigen-based screening. JAMA 1993;270:948-54.

60. Shipley WU, Thames HD, Sandler HM, Hanks GE, Zietman AL, Perez CA et al. Radiation therapy for clinically localized prostate cancer: a multi-institutional pooled analysis. JAMA 1999;281:1598-604.

61. Zelefsky MJ, Leibel SA, Gaudin PB, Kutcher GJ, Fleshner NE, Venkatramen ES et al. Dose escalation with three-dimensional conformal radiation therapy affects the outcome in prostate cancer. Int J Radiat Oncol Biol Phys 1998;41:491-500.

62. Fuks Z, Leibel SA, Kutcher GJ, Mohan R, Ling CC. Three-dimensional conformal treatment: a new frontier in radiation therapy. Important Adv Oncol 1991;151-72.

63. Cahlon O, Hunt M, Zelefsky MJ. Intensity-modulated radiation therapy: supportive data for prostate cancer. Semin Radiat Oncol 2008;18:48-57.

64. Blasko JC, Grimm PD, Ragde H. Brachytherapy and Organ Preservation in the Management of Carcinoma of the Prostate. Semin Radiat Oncol 1993;3:240-9.

65. Ragde H, Elgamal AA, Snow PB, Brandt J, Bartolucci AA, Nadir BS, Korb LJ. Ten-year disease free survival after transperineal sonography-guided iodine-125 brachytherapy with or without 45-gray external beam irradiation in the treatment of patients with clinically localized, low to high Gleason grade prostate carcinoma. Cancer 1998;83:989-1001.

66. Huggins C, Hodges CV. Studies on prostatic cancer. I. The effect of castration, of estrogen and androgen injection on serum phosphatases in metastatic carcinoma of the prostate. CA Cancer J Clin 1972;22:232-40.

67. Cox RL, Crawford ED. Estrogens in the treatment of prostate cancer. J Urol 1995;154:1991-8.

68. Seidenfeld J, Samson DJ, Hasselblad V, Aronson N, Albertsen PC, Bennett CL, Wilt TJ. Single-therapy androgen suppression in men with advanced prostate cancer: a systematic review and meta-analysis. Ann Intern Med 2000;132:566-77.

69. Stearns ME, McGarvey T. Prostate cancer: therapeutic, diagnostic, and basic studies. Lab Invest 1992;67:540-52.

70. Moore MJ, Osoba D, Murphy K, Tannock IF, Armitage A, Findlay B et al. Use of palliative end points to evaluate the effects of mitoxantrone and low-dose

References 167

prednisone in patients with hormonally resistant prostate cancer. J Clin Oncol 1994;12:689-94.

71. Pienta KJ, Redman BG, Bandekar R, Strawderman M, Cease K, Esper PS et al. A phase II trial of oral estramustine and oral etoposide in hormone refractory prostate cancer. Urology 1997;50:401-6.

72. Berry W, Dakhil S, Gregurich MA, Asmar L. Phase II trial of single-agent weekly docetaxel in hormone-refractory, symptomatic, metastatic carcinoma of the prostate. Semin Oncol 2001;28:8-15.

73. Peehl DM. Primary cell cultures as models of prostate cancer development. Endocr Relat Cancer 2005;12:19-47.

74. Calvo A, Gonzalez-Moreno O, Yoon CY, Huh JI, Desai K, Nguyen QT, Green JE. Prostate cancer and the genomic revolution: Advances using microarray analyses. Mutat Res 2005;576:66-79.

75. Bostwick DG. Prostatic intraepithelial neoplasia. Curr Urol Rep 2000;1:65-70.

76. Simone NL, Bonner RF, Gillespie JW, Emmert-Buck MR, Liotta LA. Laser-capture microdissection: opening the microscopic frontier to molecular analysis. Trends Genet 1998;14:272-6.

77. Luo J, Zha S, Gage WR, Dunn TA, Hicks JL, Bennett CJ et al. Alpha-methylacyl-CoA racemase: a new molecular marker for prostate cancer. Cancer Res 2002;62:2220-6.

78. Rhodes DR, Barrette TR, Rubin MA, Ghosh D, Chinnaiyan AM. Meta-analysis of microarrays: interstudy validation of gene expression profiles reveals pathway dysregulation in prostate cancer. Cancer Res 2002;62:4427-33.

79. Lapointe J, Li C, Higgins JP, van de RM, Bair E, Montgomery K et al. Gene expression profiling identifies clinically relevant subtypes of prostate cancer. Proc Natl Acad Sci U S A 2004;101:811-6.

80. Tomlins SA, Rhodes DR, Perner S, Dhanasekaran SM, Mehra R, Sun XW et al. Recurrent fusion of TMPRSS2 and ETS transcription factor genes in prostate cancer. Science 2005;310:644-8.

81. Magee JA, Araki T, Patil S, Ehrig T, True L, Humphrey PA et al. Expression profiling reveals hepsin overexpression in prostate cancer. Cancer Res 2001;61:5692-6.

82. Varambally S, Dhanasekaran SM, Zhou M, Barrette TR, Kumar-Sinha C, Sanda MG et al. The polycomb group protein EZH2 is involved in progression of prostate cancer. Nature 2002;419:624-9.

References 168

83. Glinsky GV, Glinskii AB, Stephenson AJ, Hoffman RM, Gerald WL. Gene expression profiling predicts clinical outcome of prostate cancer. J Clin Invest 2004;113:913-23.

84. Zhao H, Kim Y, Wang P, Lapointe J, Tibshirani R, Pollack JR, Brooks JD. Genome-wide characterization of gene expression variations and DNA copy number changes in prostate cancer cell lines. Prostate 2005;63:187-97.

85. Wolf M, Mousses S, Hautaniemi S, Karhu R, Huusko P, Allinen M et al. High-resolution analysis of gene copy number alterations in human prostate cancer using CGH on cDNA microarrays: impact of copy number on gene expression. Neoplasia 2004;6:240-7.

86. Henshall SM, Afar DE, Hiller J, Horvath LG, Quinn DI, Rasiah KK et al. Survival analysis of genome-wide gene expression profiles of prostate cancers identifies new prognostic targets of disease relapse. Cancer Res 2003;63:4196-203.

87. Lieberfarb ME, Lin M, Lechpammer M, Li C, Tanenbaum DM, Febbo PG et al. Genome-wide loss of heterozygosity analysis from laser capture microdissected prostate cancer using single nucleotide polymorphic allele (SNP) arrays and a novel bioinformatics platform dChipSNP. Cancer Res 2003;63:4781-5.

88. Bantscheff M, Boesche M, Eberhard D, Matthieson T, Sweetman G, Kuster B. Robust and sensitive iTRAQ quantification on an LTQ Orbitrap mass spectrometer. Mol Cell Proteomics 2008.

89. Hanke S, Besir H, Oesterhelt D, Mann M. Absolute SILAC for accurate quantitation of proteins in complex mixtures down to the attomole level. J Proteome Res 2008;7:1118-30.

90. Petricoin EF, III, Ornstein DK, Paweletz CP, Ardekani A, Hackett PS, Hitt BA et al. Serum proteomic patterns for detection of prostate cancer. J Natl Cancer Inst 2002;94:1576-8.

91. Adam BL, Qu Y, Davis JW, Ward MD, Clements MA, Cazares LH et al. Serum protein fingerprinting coupled with a pattern-matching algorithm distinguishes prostate cancer from benign prostate hyperplasia and healthy men. Cancer Res 2002;62:3609-14.

92. Qu Y, Adam BL, Yasui Y, Ward MD, Cazares LH, Schellhammer PF et al. Boosted decision tree analysis of surface-enhanced laser desorption/ionization mass spectral serum profiles discriminates prostate cancer from noncancer patients. Clin Chem 2002;48:1835-43.

References 169

93. McLerran D, Grizzle WE, Feng Z, Thompson IM, Bigbee WL, Cazares LH et al. SELDI-TOF MS whole serum proteomic profiling with IMAC surface does not reliably detect prostate cancer. Clin Chem 2008;54:53-60.

94. Diaz JI, Cazares LH, Corica A, John SO. Selective capture of prostatic basal cells and secretory epithelial cells for proteomic and genomic analysis. Urol Oncol 2004;22:329-36.

95. Hwang SI, Thumar J, Lundgren DH, Rezaul K, Mayya V, Wu L et al. Direct cancer tissue proteomics: a method to identify candidate cancer biomarkers from formalin-fixed paraffin-embedded archival tissues. Oncogene 2007;26:65-76.

96. Wright ME, Eng J, Sherman J, Hockenbery DM, Nelson PS, Galitski T, Aebersold R. Identification of androgen-coregulated protein networks from the microsomes of human prostate cancer cells. Genome Biol 2003;5:R4.

97. Martin DB, Gifford DR, Wright ME, Keller A, Yi E, Goodlett DR et al. Quantitative proteomic analysis of proteins released by neoplastic prostate epithelium. Cancer Res 2004;64:347-55.

98. Sardana G, Marshall J, Diamandis EP. Discovery of candidate tumor markers for prostate cancer via proteomic analysis of cell culture-conditioned medium. Clin Chem 2007;53:429-37.

99. Runsheng Li, Yan Guo, Bang Ming Han, Xiaowei Yan, Angelita G.Utleg, Wei Li et al. Proteomics cataloging analysis of human expressed prostatic secretions reveals rich source of biomarker candidates. Proteomics Clinical Applications 2008;2:543-55.

100. Dan Theodorescu, Eric Schiffer, Hartwig W.Bauer, Friedrich Douwes, Frank Eichhorn, Reinhard Polley et al. Discovery and validation of urinary biomarkers for prostate cancer. Proteomics Clinical Applications 2008;2:556-70.

101. Sreekumar A, Laxman B, Rhodes DR, Bhagavathula S, Harwood J, Giacherio D et al. Humoral immune response to alpha-methylacyl-CoA racemase and prostate cancer. J Natl Cancer Inst 2004;96:834-43.

102. Wang X, Yu J, Sreekumar A, Varambally S, Shen R, Giacherio D et al. Autoantibody signatures in prostate cancer. N Engl J Med 2005;353:1224-35.

103. Oesterling JE, Jacobsen SJ, Chute CG, Guess HA, Girman CJ, Panser LA, Lieber MM. Serum prostate-specific antigen in a community-based population of healthy men. Establishment of age-specific reference ranges. JAMA 1993;270:860-4.

References 170

104. Lilja H, Ulmert D, Bjork T, Becker C, Serio AM, Nilsson JA et al. Long-term prediction of prostate cancer up to 25 years before diagnosis of prostate cancer using prostate kallikreins measured at age 44 to 50 years. J Clin Oncol 2007;25:431-6.

105. Bassler TJ, Jr., Orozco R, Bassler IC, O'Dowd GJ, Stamey TA. Most prostate cancers missed by raising the upper limit of normal prostate-specific antigen for men in their sixties are clinically significant. Urology 1998;52:1064-9.

106. Etzioni R, Shen Y, Petteway JC, Brawer MK. Age-specific prostate-specific antigen: a reassessment. Prostate Suppl 1996;7:70-7.

107. Bangma CH, Kranse R, Blijenberg BG, Schroder FH. The value of screening tests in the detection of prostate cancer. Part I: Results of a retrospective evaluation of 1726 men. Urology 1995;46:773-8.

108. Benson MC, Whang IS, Olsson CA, McMahon DJ, Cooner WH. The use of prostate specific antigen density to enhance the predictive value of intermediate levels of serum prostate specific antigen. J Urol 1992;147:817-21.

109. Catalona WJ, Richie JP, deKernion JB, Ahmann FR, Ratliff TL, Dalkin BL et al. Comparison of prostate specific antigen concentration versus prostate specific antigen density in the early detection of prostate cancer: receiver operating characteristic curves. J Urol 1994;152:2031-6.

110. Mettlin C, Littrup PJ, Kane RA, Murphy GP, Lee F, Chesley A et al. Relative sensitivity and specificity of serum prostate specific antigen (PSA) level compared with age-referenced PSA, PSA density, and PSA change. Data from the American Cancer Society National Prostate Cancer Detection Project. Cancer 1994;74:1615-20.

111. Djavan B, Zlotta AR, Byttebier G, Shariat S, Omar M, Schulman CC, Marberger M. Prostate specific antigen density of the transition zone for early detection of prostate cancer. J Urol 1998;160:411-8.

112. Brawer MK, Aramburu EA, Chen GL, Preston SD, Ellis WJ. The inability of prostate specific antigen index to enhance the predictive the value of prostate specific antigen in the diagnosis of prostatic carcinoma. J Urol 1993;150:369-73.

113. Cookson MS, Floyd MK, Ball TP, Jr., Miller EK, Sarosdy MF. The lack of predictive value of prostate specific antigen density in the detection of prostate cancer in patients with normal rectal examinations and intermediate prostate specific antigen levels. J Urol 1995;154:1070-3.

References 171

114. Lin DW, Gold MH, Ransom S, Ellis WJ, Brawer MK. Transition zone prostate specific antigen density: lack of use in prediction of prostatic carcinoma. J Urol 1998;160:77-81.

115. Benson MC, Whang IS, Pantuck A, Ring K, Kaplan SA, Olsson CA, Cooner WH. Prostate specific antigen density: a means of distinguishing benign prostatic hypertrophy and prostate cancer. J Urol 1992;147:815-6.

116. Seaman E, Whang M, Olsson CA, Katz A, Cooner WH, Benson MC. PSA density (PSAD). Role in patient evaluation and management. Urol Clin North Am 1993;20:653-63.

117. Carter HB, Morrell CH, Pearson JD, Brant LJ, Plato CC, Metter EJ et al. Estimation of prostatic growth using serial prostate-specific antigen measurements in men with and without prostate disease. Cancer Res 1992;52:3323-8.

118. Roehrborn CG, Pickens GJ, Carmody T, III. Variability of repeated serum prostate-specific antigen (PSA) measurements within less than 90 days in a well-defined patient population. Urology 1996;47:59-66.

119. Link RE, Shariat SF, Nguyen CV, Farr A, Weinberg AD, Morton RA et al. Variation in prostate specific antigen results from 2 different assay platforms: clinical impact on 2304 patients undergoing prostate cancer screening. J Urol 2004;171:2234-8.

120. Carter HB, Pearson JD, Metter EJ, Brant LJ, Chan DW, Andres R et al. Longitudinal evaluation of prostate-specific antigen levels in men with and without prostate disease. JAMA 1992;267:2215-20.

121. Carter HB, Pearson JD, Waclawiw Z, Metter EJ, Chan DW, Guess HA, Walsh PC. Prostate-specific antigen variability in men without prostate cancer: effect of sampling interval on prostate-specific antigen velocity. Urology 1995;45:591-6.

122. D'Amico AV, Chen MH, Roehl KA, Catalona WJ. Preoperative PSA velocity and the risk of death from prostate cancer after radical prostatectomy. N Engl J Med 2004;351:125-35.

123. D'Amico AV, Renshaw AA, Sussman B, Chen MH. Pretreatment PSA velocity and risk of death from prostate cancer following external beam radiation therapy. JAMA 2005;294:440-7.

124. Patel DA, Presti JC, Jr., McNeal JE, Gill H, Brooks JD, King CR. Preoperative PSA velocity is an independent prognostic factor for relapse after radical prostatectomy. J Clin Oncol 2005;23:6157-62.

References 172

125. Raaijmakers R, Wildhagen MF, Ito K, Paez A, de Vries SH, Roobol MJ, Schroder FH. Prostate-specific antigen change in the European Randomized Study of Screening for Prostate Cancer, section Rotterdam. Urology 2004;63:316-20.

126. Catalona WJ, Partin AW, Slawin KM, Brawer MK, Flanigan RC, Patel A et al. Use of the percentage of free prostate-specific antigen to enhance differentiation of prostate cancer from benign prostatic disease: a prospective multicenter clinical trial. JAMA 1998;279:1542-7.

127. Catalona WJ, Smith DS, Wolfert RL, Wang TJ, Rittenhouse HG, Ratliff TL, Nadler RB. Evaluation of percentage of free serum prostate-specific antigen to improve specificity of prostate cancer screening. JAMA 1995;274:1214-20.

128. Catalona WJ, Smith DS, Ornstein DK. Prostate cancer detection in men with serum PSA concentrations of 2.6 to 4.0 ng/mL and benign prostate examination. Enhancement of specificity with free PSA measurements. JAMA 1997;277:1452-5.

129. Catalona WJ, Partin AW, Finlay JA, Chan DW, Rittenhouse HG, Wolfert RL, Woodrum DL. Use of percentage of free prostate-specific antigen to identify men at high risk of prostate cancer when PSA levels are 2.51 to 4 ng/mL and digital rectal examination is not suspicious for prostate cancer: an alternative model. Urology 1999;54:220-4.

130. Uemura H, Nakamura M, Hasumi H, Sugiura S, Fujinami K, Miyoshi Y et al. Effectiveness of percent free prostate specific antigen as a predictor of prostate cancer detection on repeat biopsy. Int J Urol 2004;11:494-500.

131. Stephan C, Lein M, Jung K, Schnorr D, Loening SA. The influence of prostate volume on the ratio of free to total prostate specific antigen in serum of patients with prostate carcinoma and benign prostate hyperplasia. Cancer 1997;79:104-9.

132. Southwick PC, Catalona WJ, Partin AW, Slawin KM, Brawer MK, Flanigan RC et al. Prediction of post-radical prostatectomy pathological outcome for stage T1c prostate cancer with percent free prostate specific antigen: a prospective multicenter clinical trial. J Urol 1999;162:1346-51.

133. Roehl KA, Antenor JA, Catalona WJ. Robustness of free prostate specific antigen measurements to reduce unnecessary biopsies in the 2.6 to 4.0 ng./ml. range. J Urol 2002;168:922-5.

134. Morote J, Raventos CX, Encabo G, Lopez M, de Torres IM. Effect of high-grade prostatic intraepithelial neoplasia on total and percent free serum prostatic-specific antigen. Eur Urol 2000;37:456-9.

References 173

135. Ornstein DK, Smith DS, Humphrey PA, Catalona WJ. The effect of prostate volume, age, total prostate specific antigen level and acute inflammation on the percentage of free serum prostate specific antigen levels in men without clinically detectable prostate cancer. J Urol 1998;159:1234-7.

136. Canto EI, Singh H, Shariat SF, Kadmon D, Miles BJ, Wheeler TM, Slawin KM. Effects of systematic 12-core biopsy on the performance of percent free prostate specific antigen for prostate cancer detection. J Urol 2004;172:900-4.

137. Catalona WJ, Southwick PC, Slawin KM, Partin AW, Brawer MK, Flanigan RC et al. Comparison of percent free PSA, PSA density, and age-specific PSA cutoffs for prostate cancer detection and staging. Urology 2000;56:255-60.

138. Shariat SF, Abdel-Aziz KF, Roehrborn CG, Lotan Y. Pre-operative percent free PSA predicts clinical outcomes in patients treated with radical prostatectomy with total PSA levels below 10 ng/ml. Eur Urol 2006;49:293-302.

139. Graefen M, Karakiewicz PI, Cagiannos I, Hammerer PG, Haese A, Palisaar J et al. Percent free prostate specific antigen is not an independent predictor of organ confinement or prostate specific antigen recurrence in unscreened patients with localized prostate cancer treated with radical prostatectomy. J Urol 2002;167:1306-9.

140. Stenman UH, Leinonen J, Alfthan H, Rannikko S, Tuhkanen K, Alfthan O. A complex between prostate-specific antigen and alpha 1-antichymotrypsin is the major form of prostate-specific antigen in serum of patients with prostatic cancer: assay of the complex improves clinical sensitivity for cancer. Cancer Res 1991;51:222-6.

141. Lilja H, Christensson A, Dahlen U, Matikainen MT, Nilsson O, Pettersson K, Lovgren T. Prostate-specific antigen in serum occurs predominantly in complex with alpha 1-antichymotrypsin. Clin Chem 1991;37:1618-25.

142. Christensson A, Bjork T, Nilsson O, Dahlen U, Matikainen MT, Cockett AT et al. Serum prostate specific antigen complexed to alpha 1-antichymotrypsin as an indicator of prostate cancer. J Urol 1993;150:100-5.

143. Partin AW, Brawer MK, Bartsch G, Horninger W, Taneja SS, Lepor H et al. Complexed prostate specific antigen improves specificity for prostate cancer detection: results of a prospective multicenter clinical trial. J Urol 2003;170:1787-91.

144. Brawer MK, Meyer GE, Letran JL, Bankson DD, Morris DL, Yeung KK, Allard WJ. Measurement of complexed PSA improves specificity for early detection of prostate cancer. Urology 1998;52:372-8.

References 174

145. Allard WJ, Zhou Z, Yeung KK. Novel immunoassay for the measurement of complexed prostate-specific antigen in serum. Clin Chem 1998;44:1216-23.

146. Kobayashi T, Kamoto T, Nishizawa K, Mitsumori K, Ogura K, Ide Y. Prostate-specific antigen (PSA) complexed to alpha1-antichymotrypsin improves prostate cancer detection using total PSA in Japanese patients with total PSA levels of 2.0-4.0 ng/mL. BJU Int 2005;95:761-5.

147. Lein M, Kwiatkowski M, Semjonow A, Luboldt HJ, Hammerer P, Stephan C et al. A multicenter clinical trial on the use of complexed prostate specific antigen in low prostate specific antigen concentrations. J Urol 2003;170:1175-9.

148. Zhang WM, Finne P, Leinonen J, Vesalainen S, Nordling S, Rannikko S, Stenman UH. Characterization and immunological determination of the complex between prostate-specific antigen and alpha2-macroglobulin. Clin Chem 1998;44:2471-9.

149. Roddam AW, Duffy MJ, Hamdy FC, Ward AM, Patnick J, Price CP et al. Use of prostate-specific antigen (PSA) isoforms for the detection of prostate cancer in men with a PSA level of 2-10 ng/ml: systematic review and meta-analysis. Eur Urol 2005;48:386-99.

150. Tanguay S, Begin LR, Elhilali MM, Behlouli H, Karakiewicz PI, Aprikian AG. Comparative evaluation of total PSA, free/total PSA, and complexed PSA in prostate cancer detection. Urology 2002;59:261-5.

151. Bjork T, Lilja H, Christensson A. The prognostic value of different forms of prostate specific antigen and their ratios in patients with prostate cancer. BJU Int 1999;84:1021-7.

152. Mitchell ID, Croal BL, Dickie A, Cohen NP, Ross I. A prospective study to evaluate the role of complexed prostate specific antigen and free/total prostate specific antigen ratio for the diagnosis of prostate cancer. J Urol 2001;165:1549-53.

153. Lein M, Jung K, Hammerer P, Graefen M, Semjonow A, Stieber P et al. A multicenter clinical trial on the use of alpha1-antichymotrypsin-prostate-specific antigen in prostate cancer diagnosis. Prostate 2001;47:77-84.

154. Zhang WM, Finne P, Leinonen J, Salo J, Stenman UH. Determination of prostate-specific antigen complexed to alpha(2)-macroglobulin in serum increases the specificity of free to total PSA for prostate cancer. Urology 2000;56:267-72.

References 175

155. Finne P, Zhang WM, Auvinen A, Leinonen J, Maattanen L, Rannikko S et al. Use of the complex between prostate specific antigen and alpha 1-protease inhibitor for screening prostate cancer. J Urol 2000;164:1956-60.

156. Zhang WM, Finne P, Leinonen J, Stenman UH. Characterization and determination of the complex between prostate-specific antigen and alpha 1-protease inhibitor in benign and malignant prostatic diseases. Scand J Clin Lab Invest Suppl 2000;233:51-8.

157. Mikolajczyk SD, Catalona WJ, Evans CL, Linton HJ, Millar LS, Marker KM et al. Proenzyme forms of prostate-specific antigen in serum improve the detection of prostate cancer. Clin Chem 2004;50:1017-25.

158. Mikolajczyk SD, Marker KM, Millar LS, Kumar A, Saedi MS, Payne JK et al. A truncated precursor form of prostate-specific antigen is a more specific serum marker of prostate cancer. Cancer Res 2001;61:6958-63.

159. Nurmikko P, Pettersson K, Piironen T, Hugosson J, Lilja H. Discrimination of prostate cancer from benign disease by plasma measurement of intact, free prostate-specific antigen lacking an internal cleavage site at Lys145-Lys146. Clin Chem 2001;47:1415-23.

160. Steuber T, Nurmikko P, Haese A, Pettersson K, Graefen M, Hammerer P et al. Discrimination of benign from malignant prostatic disease by selective measurements of single chain, intact free prostate specific antigen. J Urol 2002;168:1917-22.

161. Mikolajczyk SD, Grauer LS, Millar LS, Hill TM, Kumar A, Rittenhouse HG et al. A precursor form of PSA (pPSA) is a component of the free PSA in prostate cancer serum. Urology 1997;50:710-4.

162. Peter J, Unverzagt C, Krogh TN, Vorm O, Hoesel W. Identification of precursor forms of free prostate-specific antigen in serum of prostate cancer patients by immunosorption and mass spectrometry. Cancer Res 2001;61:957-62.

163. Catalona WJ, Bartsch G, Rittenhouse HG, Evans CL, Linton HJ, Horninger W et al. Serum pro-prostate specific antigen preferentially detects aggressive prostate cancers in men with 2 to 4 ng/ml prostate specific antigen. J Urol 2004;171:2239-44.

164. Catalona WJ, Bartsch G, Rittenhouse HG, Evans CL, Linton HJ, Amirkhan A et al. Serum pro prostate specific antigen improves cancer detection compared to free and complexed prostate specific antigen in men with prostate specific antigen 2 to 4 ng/ml. J Urol 2003;170:2181-5.

References 176

165. Khan MA, Partin AW, Rittenhouse HG, Mikolajczyk SD, Sokoll LJ, Chan DW, Veltri RW. Evaluation of proprostate specific antigen for early detection of prostate cancer in men with a total prostate specific antigen range of 4.0 to 10.0 ng/ml. J Urol 2003;170:723-6.

166. de Vries SH, Raaijmakers R, Blijenberg BG, Mikolajczyk SD, Rittenhouse HG, Schroder FH. Additional use of [-2] precursor prostate-specific antigen and "benign" PSA at diagnosis in screen-detected prostate cancer. Urology 2005;65:926-30.

167. Bangma CH, Wildhagen MF, Yurdakul G, Schroder FH, Blijenberg BG. The value of (-7, -5)pro-prostate-specific antigen and human kallikrein-2 as serum markers for grading prostate cancer. BJU Int 2004;93:720-4.

168. Lein M, Semjonow A, Graefen M, Kwiatkowski M, Abramjuk C, Stephan C et al. A multicenter clinical trial on the use of (-5, -7) pro prostate specific antigen. J Urol 2005;174:2150-3.

169. Stephan C, Meyer HA, Paul EM, Kristiansen G, Loening SA, Lein M, Jung K. Serum (-5, -7) proPSA for distinguishing stage and grade of prostate cancer. Anticancer Res 2007;27:1833-6.

170. Mikolajczyk SD, Millar LS, Wang TJ, Rittenhouse HG, Wolfert RL, Marks LS et al. "BPSA," a specific molecular form of free prostate-specific antigen, is found predominantly in the transition zone of patients with nodular benign prostatic hyperplasia. Urology 2000;55:41-5.

171. Linton HJ, Marks LS, Millar LS, Knott CL, Rittenhouse HG, Mikolajczyk SD. Benign prostate-specific antigen (BPSA) in serum is increased in benign prostate disease. Clin Chem 2003;49:253-9.

172. Mikolajczyk SD, Song Y, Wong JR, Matson RS, Rittenhouse HG. Are multiple markers the future of prostate cancer diagnostics? Clin Biochem 2004;37:519-28.

173. Khan MA, Sokoll LJ, Chan DW, Mangold LA, Mohr P, Mikolajczyk SD et al. Clinical utility of proPSA and "benign" PSA when percent free PSA is less than 15%. Urology 2004;64:1160-4.

174. Canto EI, Singh H, Shariat SF, Lamb DJ, Mikolajczyk SD, Linton HJ et al. Serum BPSA outperforms both total PSA and free PSA as a predictor of prostatic enlargement in men without prostate cancer. Urology 2004;63:905-10.

175. Yousef GM, Diamandis EP. The new human tissue kallikrein gene family: structure, function, and association to disease. Endocr Rev 2001;22:184-204.

References 177

176. Darson MF, Pacelli A, Roche P, Rittenhouse HG, Wolfert RL, Young CY et al. Human glandular kallikrein 2 (hK2) expression in prostatic intraepithelial neoplasia and adenocarcinoma: a novel prostate cancer marker. Urology 1997;49:857-62.

177. Lintula S, Stenman J, Bjartell A, Nordling S, Stenman UH. Relative concentrations of hK2/PSA mRNA in benign and malignant prostatic tissue. Prostate 2005;63:324-9.

178. Becker C, Piironen T, Pettersson K, Hugosson J, Lilja H. Clinical value of human glandular kallikrein 2 and free and total prostate-specific antigen in serum from a population of men with prostate-specific antigen levels 3.0 ng/mL or greater. Urology 2000;55:694-9.

179. Recker F, Kwiatkowski MK, Piironen T, Pettersson K, Huber A, Lummen G, Tscholl R. Human glandular kallikrein as a tool to improve discrimination of poorly differentiated and non-organ-confined prostate cancer compared with prostate-specific antigen. Urology 2000;55:481-5.

180. Nam RK, Diamandis EP, Toi A, Trachtenberg J, Magklara A, Scorilas A et al. Serum human glandular kallikrein-2 protease levels predict the presence of prostate cancer among men with elevated prostate-specific antigen. J Clin Oncol 2000;18:1036-42.

181. Kwiatkowski MK, Recker F, Piironen T, Pettersson K, Otto T, Wernli M, Tscholl R. In prostatism patients the ratio of human glandular kallikrein to free PSA improves the discrimination between prostate cancer and benign hyperplasia within the diagnostic "gray zone" of total PSA 4 to 10 ng/mL. Urology 1998;52:360-5.

182. Becker C, Piironen T, Pettersson K, Bjork T, Wojno KJ, Oesterling JE, Lilja H. Discrimination of men with prostate cancer from those with benign disease by measurements of human glandular kallikrein 2 (HK2) in serum. J Urol 2000;163:311-6.

183. Magklara A, Scorilas A, Catalona WJ, Diamandis EP. The combination of human glandular kallikrein and free prostate-specific antigen (PSA) enhances discrimination between prostate cancer and benign prostatic hyperplasia in patients with moderately increased total PSA. Clin Chem 1999;45:1960-6.

184. Partin AW, Catalona WJ, Finlay JA, Darte C, Tindall DJ, Young CY et al. Use of human glandular kallikrein 2 for the detection of prostate cancer: preliminary analysis. Urology 1999;54:839-45.

References 178

185. Stephan C, Jung K, Nakamura T, Yousef GM, Kristiansen G, Diamandis EP. Serum human glandular kallikrein 2 (hK2) for distinguishing stage and grade of prostate cancer. Int J Urol 2006;13:238-43.

186. Haese A, Graefen M, Steuber T, Becker C, Noldus J, Erbersdobler A et al. Total and Gleason grade 4/5 cancer volumes are major contributors of human kallikrein 2, whereas free prostate specific antigen is largely contributed by benign gland volume in serum from patients with prostate cancer or benign prostatic biopsies. J Urol 2003;170:2269-73.

187. Haese A, Graefen M, Steuber T, Becker C, Pettersson K, Piironen T et al. Human glandular kallikrein 2 levels in serum for discrimination of pathologically organ-confined from locally-advanced prostate cancer in total PSA-levels below 10 ng/ml. Prostate 2001;49:101-9.

188. Steuber T, Vickers AJ, Haese A, Becker C, Pettersson K, Chun FK et al. Risk assessment for biochemical recurrence prior to radical prostatectomy: significant enhancement contributed by human glandular kallikrein 2 (hK2) and free prostate specific antigen (PSA) in men with moderate PSA-elevation in serum. Int J Cancer 2006;118:1234-40.

189. Becker C, Piironen T, Kiviniemi J, Lilja H, Pettersson K. Sensitive and specific immunodetection of human glandular kallikrein 2 in serum. Clin Chem 2000;46:198-206.

190. Mikolajczyk SD, Millar LS, Marker KM, Rittenhouse HG, Wolfert RL, Marks LS et al. Identification of a novel complex between human kallikrein 2 and protease inhibitor-6 in prostate cancer tissue. Cancer Res 1999;59:3927-30.

191. Diamandis EP, Yousef GM, Clements J, Ashworth LK, Yoshida S, Egelrud T et al. New nomenclature for the human tissue kallikrein gene family. Clin Chem 2000;46:1855-8.

192. Diamandis EP, Scorilas A, Fracchioli S, Van Gramberen M, De Bruijn H, Henrik A et al. Human kallikrein 6 (hK6): a new potential serum biomarker for diagnosis and prognosis of ovarian carcinoma. J Clin Oncol 2003;21:1035-43.

193. Diamandis EP, Yousef GM, Petraki C, Soosaipillai AR. Human kallikrein 6 as a biomarker of alzheimer's disease. Clin Biochem 2000;33:663-7.

194. Luo LY, Grass L, Howarth DJ, Thibault P, Ong H, Diamandis EP. Immunofluorometric assay of human kallikrein 10 and its identification in biological fluids and tissues. Clin Chem 2001;47:237-46.

References 179

195. Diamandis EP, Okui A, Mitsui S, Luo LY, Soosaipillai A, Grass L et al. Human kallikrein 11: a new biomarker of prostate and ovarian carcinoma. Cancer Res 2002;62:295-300.

196. Luo LY, Bunting P, Scorilas A, Diamandis EP. Human kallikrein 10: a novel tumor marker for ovarian carcinoma? Clin Chim Acta 2001;306:111-8.


198. Nakamura T, Scorilas A, Stephan C, Jung K, Soosaipillai AR, Diamandis EP. The usefulness of serum human kallikrein 11 for discriminating between prostate cancer and benign prostatic hyperplasia. Cancer Res 2003;63:6543-6.

199. Stephan C, Meyer HA, Cammann H, Nakamura T, Diamandis EP, Jung K. Improved prostate cancer detection with a human kallikrein 11 and percentage free PSA-based artificial neural network. Biol Chem 2006;387:801-5.

200. Horoszewicz JS, Kawinski E, Murphy GP. Monoclonal antibodies to a new antigenic marker in epithelial prostatic cells and serum of prostatic cancer patients. Anticancer Res 1987;7:927-35.

201. Elgamal AA, Holmes EH, Su SL, Tino WT, Simmons SJ, Peterson M et al. Prostate-specific membrane antigen (PSMA): current benefits and future value. Semin Surg Oncol 2000;18:10-6.

202. Murphy GP, Barren RJ, Erickson SJ, Bowes VA, Wolfert RL, Bartsch G et al. Evaluation and comparison of two new prostate carcinoma markers. Free-prostate specific antigen and prostate specific membrane antigen. Cancer 1996;78:809-18.

203. Murphy GP, Elgamal AA, Su SL, Bostwick DG, Holmes EH. Current evaluation of the tissue localization and diagnostic utility of prostate specific membrane antigen. Cancer 1998;83:2259-69.

204. Beckett ML, Cazares LH, Vlahou A, Schellhammer PF, Wright GL, Jr. Prostate-specific membrane antigen levels in sera from healthy men and patients with benign prostate hyperplasia or prostate cancer. Clin Cancer Res 1999;5:4034-40.

205. Xiao Z, Adam BL, Cazares LH, Clements MA, Davis JW, Schellhammer PF et al. Quantitation of serum prostate-specific membrane antigen by a novel

References 180

protein biochip immunoassay discriminates benign from malignant prostate disease. Cancer Res 2001;61:6029-33.

206. Kurek R, Nunez G, Tselis N, Konrad L, Martin T, Roeddiger S et al. Prognostic value of combined "triple"-reverse transcription-PCR analysis for prostate-specific antigen, human kallikrein 2, and prostate-specific membrane antigen mRNA in peripheral blood and lymph nodes of prostate cancer patients. Clin Cancer Res 2004;10:5808-14.

207. Ghosh A, Heston WD. Tumor target prostate specific membrane antigen (PSMA) and its regulation in prostate cancer. J Cell Biochem 2004;91:528-39.

208. Gong MC, Chang SS, Sadelain M, Bander NH, Heston WD. Prostate-specific membrane antigen (PSMA)-specific monoclonal antibodies in the treatment of prostate and other cancers. Cancer Metastasis Rev 1999;18:483-90.

209. Mincheff M, Zoubak S, Makogonenko Y. Immune responses against PSMA after gene-based vaccination for immunotherapy-A: results from immunizations in animals. Cancer Gene Ther 2006;13:436-44.

210. Murphy GP, Tjoa BA, Simmons SJ, Jarisch J, Bowes VA, Ragde H et al. Infusion of dendritic cells pulsed with HLA-A2-specific prostate-specific membrane antigen peptides: a phase II prostate cancer vaccine trial involving patients with hormone-refractory metastatic disease. Prostate 1999;38:73-8.

211. Tjoa BA, Simmons SJ, Elgamal A, Rogers M, Ragde H, Kenny GM et al. Follow-up evaluation of a phase II prostate cancer vaccine trial. Prostate 1999;40:125-9.

212. Hessels D, Klein Gunnewiek JM, van O, I, Karthaus HF, van Leenders GJ, van Balken B et al. DD3(PCA3)-based molecular urine analysis for the diagnosis of prostate cancer. Eur Urol 2003;44:8-15.

213. Groskopf J, Aubin SM, Deras IL, Blase A, Bodrug S, Clark C et al. APTIMA PCA3 molecular urine test: development of a method to aid in the diagnosis of prostate cancer. Clin Chem 2006;52:1089-95.

214. van Gils MP, Hessels D, van Hooij O, Jannink SA, Peelen WP, Hanssen SL et al. The time-resolved fluorescence-based PCA3 test on urinary sediments after digital rectal examination; a Dutch multicenter validation of the diagnostic performance. Clin Cancer Res 2007;13:939-43.

215. Marks LS, Fradet Y, Deras IL, Blase A, Mathis J, Aubin SM et al. PCA3 molecular urine assay for prostate cancer in men undergoing repeat biopsy. Urology 2007;69:532-5.

References 181

216. Laxman B, Morris DS, Yu J, Siddiqui J, Cao J, Mehra R et al. A first-generation multiplex biomarker analysis of urine for the early detection of prostate cancer. Cancer Res 2008;68:645-9.

217. Getzenberg RH, Pienta KJ, Huang EY, Coffey DS. Identification of nuclear matrix proteins in the cancer and normal rat prostate. Cancer Res 1991;51:6514-20.

218. Uetsuki H, Tsunemori H, Taoka R, Haba R, Ishikawa M, Kakehi Y. Expression of a novel biomarker, EPCA, in adenocarcinomas and precancerous lesions in the prostate. J Urol 2005;174:514-8.

219. Paul B, Dhir R, Landsittel D, Hitchens MR, Getzenberg RH. Detection of prostate cancer with a blood-based assay for early prostate cancer antigen. Cancer Res 2005;65:4097-100.

220. Leman ES, Cannon GW, Trock BJ, Sokoll LJ, Chan DW, Mangold L et al. EPCA-2: a highly specific serum marker for prostate cancer. Urology 2007;69:714-20.

221. Diamandis EP. POINT: EPCA-2: a promising new serum biomarker for prostatic carcinoma? Clin Biochem 2007;40:1437-9.

222. Wanders RJ, Vreken P, Ferdinandusse S, Jansen GA, Waterham HR, van Roermund CW, Van Grunsven EG. Peroxisomal fatty acid alpha- and beta-oxidation in humans: enzymology, peroxisomal metabolite transporters and peroxisomal diseases. Biochem Soc Trans 2001;29:250-67.

223. Zheng SL, Chang BL, Faith DA, Johnson JR, Isaacs SD, Hawkins GA et al. Sequence variants of alpha-methylacyl-CoA racemase are associated with prostate cancer risk. Cancer Res 2002;62:6485-8.

224. Jiang Z, Wu CL, Woda BA, Iczkowski KA, Chu PG, Tretiakova MS et al. Alpha-methylacyl-CoA racemase: a multi-institutional study of a new prostate cancer marker. Histopathology 2004;45:218-25.

225. Rubin MA, Bismar TA, Andren O, Mucci L, Kim R, Shen R et al. Decreased alpha-methylacyl CoA racemase expression in localized prostate cancer is associated with an increased rate of biochemical recurrence and cancer-specific death. Cancer Epidemiol Biomarkers Prev 2005;14:1424-32.

226. Zehentner BK, Secrist H, Zhang X, Hayes DC, Ostenson R, Goodman G et al. Detection of alpha-methylacyl-coenzyme-A racemase transcripts in blood and urine samples of prostate cancer patients. Mol Diagn Ther 2006;10:397-403.

References 182

227. Rogers CG, Yan G, Zha S, Gonzalgo ML, Isaacs WB, Luo J et al. Prostate cancer detection on urinalysis for alpha methylacyl coenzyme a racemase protein. J Urol 2004;172:1501-3.

228. McCabe NP, Angwafo FF, III, Zaher A, Selman SH, Kouinche A, Jankun J. Expression of soluble urokinase plasminogen activator receptor may be related to outcome in prostate cancer patients. Oncol Rep 2000;7:879-82.

229. Steuber T, Vickers A, Haese A, Kattan MW, Eastham JA, Scardino PT et al. Free PSA isoforms and intact and cleaved forms of urokinase plasminogen activator receptor in serum improve selection of patients for prostate cancer biopsy. Int J Cancer 2007;120:1499-504.

230. Achbarou A, Kaiser S, Tremblay G, Ste-Marie LG, Brodt P, Goltzman D, Rabbani SA. Urokinase overproduction results in increased skeletal metastasis by prostate cancer cells in vivo. Cancer Res 1994;54:2372-7.

231. Miyake H, Hara I, Yamanaka K, Gohji K, Arakawa S, Kamidono S. Elevation of serum levels of urokinase-type plasminogen activator and its receptor is associated with disease progression and prognosis in patients with prostate cancer. Prostate 1999;39:123-9.

232. Shariat SF, Roehrborn CG, McConnell JD, Park S, Alam N, Wheeler TM, Slawin KM. Association of the circulating levels of the urokinase system of plasminogen activation with the presence of prostate cancer and invasion, progression, and metastasis. J Clin Oncol 2007;25:349-55.

233. Chan JM, Stampfer MJ, Giovannucci E, Gann PH, Ma J, Wilkinson P et al. Plasma insulin-like growth factor-I and prostate cancer risk: a prospective study. Science 1998;279:563-6.

234. Harman SM, Metter EJ, Blackman MR, Landis PK, Carter HB. Serum levels of insulin-like growth factor I (IGF-I), IGF-II, IGF-binding protein-3, and prostate-specific antigen as predictors of clinical prostate cancer. J Clin Endocrinol Metab 2000;85:4258-65.

235. Shariat SF, Lamb DJ, Kattan MW, Nguyen C, Kim J, Beck J et al. Association of preoperative plasma levels of insulin-like growth factor I and insulin-like growth factor binding proteins-2 and -3 with prostate cancer invasion, progression, and metastasis. J Clin Oncol 2002;20:833-41.

236. Laxman B, Tomlins SA, Mehra R, Morris DS, Wang L, Helgeson BE et al. Noninvasive detection of TMPRSS2:ERG fusion transcripts in the urine of men with prostate cancer. Neoplasia 2006;8:885-8.

References 183

237. Demichelis F, Fall K, Perner S, Andren O, Schmidt F, Setlur SR et al. TMPRSS2:ERG gene fusion associated with lethal prostate cancer in a watchful waiting cohort. Oncogene 2007;26:4596-9.

238. Morton DM, Barrack ER. Modulation of transforming growth factor beta 1 effects on prostate cancer cell proliferation by growth factors and extracellular matrix. Cancer Res 1995;55:2596-602.

239. Shariat SF, Menesses-Diaz A, Kim IY, Muramoto M, Wheeler TM, Slawin KM. Tissue expression of transforming growth factor-beta1 and its receptors: correlation with pathologic features and biochemical progression in patients undergoing radical prostatectomy. Urology 2004;63:1191-7.

240. Ivanovic V, Melman A, Davis-Joseph B, Valcic M, Geliebter J. Elevated plasma levels of TGF-beta 1 in patients with invasive prostate cancer. Nat Med 1995;1:282-4.

241. Sinnreich O, Kratzsch J, Reichenbach A, Glaser C, Huse K, Birkenmeier G. Plasma levels of transforming growth factor-1beta and alpha2-macroglobulin before and after radical prostatectomy: association to clinicopathological parameters. Prostate 2004;61:201-8.

242. Shariat SF, Walz J, Roehrborn CG, Montorsi F, Jeldres C, Saad F, Karakiewicz PI. Early postoperative plasma transforming growth factor-beta1 is a strong predictor of biochemical progression after radical prostatectomy. J Urol 2008;179:1593-7.

243. Rhodes DR, Sanda MG, Otte AP, Chinnaiyan AM, Rubin MA. Multiplex biomarker approach for determining risk of prostate-specific antigen-defined recurrence of prostate cancer. J Natl Cancer Inst 2003;95:661-8.

244. Gonzalgo ML, Nakayama M, Lee SM, De Marzo AM, Nelson WG. Detection of GSTP1 methylation in prostatic secretions using combinatorial MSP analysis. Urology 2004;63:414-8.

245. Crocitto LE, Korns D, Kretzner L, Shevchuk T, Blair SL, Wilson TG et al. Prostate cancer molecular markers GSTP1 and hTERT in expressed prostatic secretions as predictors of biopsy results. Urology 2004;64:821-5.

246. Roupret M, Hupertan V, Yates DR, Catto JW, Rehman I, Meuth M et al. Molecular detection of localized prostate cancer using quantitative methylation-specific PCR on urinary cells obtained following prostate massage. Clin Cancer Res 2007;13:1720-5.

247. Nam RK, Reeves JR, Toi A, Dulude H, Trachtenberg J, Emami M et al. A novel serum marker, total prostate secretory protein of 94 amino acids,

References 184

improves prostate cancer detection and helps identify high grade cancers at diagnosis. J Urol 2006;175:1291-7.

248. Sasaki T, Komiya A, Suzuki H, Shimbo M, Ueda T, Akakura K, Ichikawa T. Changes in chromogranin a serum levels during endocrine therapy in metastatic prostate cancer patients. Eur Urol 2005;48:224-9.

249. Fracalanza S, Prayer-Galetti T, Pinto F, Navaglia F, Sacco E, Ciaccia M et al. Plasma chromogranin A in patients with prostate cancer improves the diagnostic efficacy of free/total prostate-specific antigen determination. Urol Int 2005;75:57-61.

250. Yashi M, Muraishi O, Kobayashi Y, Tokue A, Nanjo H. Elevated serum progastrin-releasing peptide (31-98) in metastatic and androgen-independent prostate cancer patients. Prostate 2002;51:84-97.

251. Umbas R, Schalken JA, Aalders TW, Carter BS, Karthaus HF, Schaafsma HE et al. Expression of the cellular adhesion molecule E-cadherin is reduced or absent in high-grade prostate cancer. Cancer Res 1992;52:5104-9.

252. Umbas R, Isaacs WB, Bringuier PP, Schaafsma HE, Karthaus HF, Oosterhof GO et al. Decreased E-cadherin expression is associated with poor prognosis in patients with prostate cancer. Cancer Res 1994;54:3929-33.

253. Gerke V, Creutz CE, Moss SE. Annexins: linking Ca2+ signalling to membrane dynamics. Nat Rev Mol Cell Biol 2005;6:449-61.

254. Schartz NE, Chaput N, Andre F, Zitvogel L. From the antigen-presenting cell to the antigen-presenting vesicle: the exosomes. Curr Opin Mol Ther 2002;4:372-81.

255. Kollermann J, Schlomm T, Bang H, Schwall GP, Eichel-Streiber C, Simon R et al. Expression and Prognostic Relevance of Annexin A3 in Prostate Cancer. Eur Urol 2008.

256. Wozny W, Schroer K, Schwall GP, Poznanovic S, Stegmann W, Dietz K et al. Differential radioactive quantification of protein abundance ratios between benign and malignant prostate tissues: cancer association of annexin A3. Proteomics 2007;7:313-22.

257. Hara N, Kasahara T, Kawasaki T, Bilim V, Obara K, Takahashi K, Tomita Y. Reverse transcription-polymerase chain reaction detection of prostate-specific antigen, prostate-specific membrane antigen, and prostate stem cell antigen in one milliliter of peripheral blood: value for the staging of prostate cancer. Clin Cancer Res 2002;8:1794-9.

References 185

258. Gu Z, Thomas G, Yamashiro J, Shintaku IP, Dorey F, Raitano A et al. Prostate stem cell antigen (PSCA) expression increases with high gleason score, advanced stage and bone metastasis in prostate cancer. Oncogene 2000;19:1288-96.

259. Gu Z, Yamashiro J, Kono E, Reiter RE. Anti-prostate stem cell antigen monoclonal antibody 1G8 induces cell death in vitro and inhibits tumor growth in vivo via a Fc-independent mechanism. Cancer Res 2005;65:9495-500.

260. Leytus SP, Loeb KR, Hagen FS, Kurachi K, Davie EW. A novel trypsin-like serine protease (hepsin) with a putative transmembrane domain expressed by human liver and hepatoma cells. Biochemistry 1988;27:1067-74.

261. Stephan C, Yousef GM, Scorilas A, Jung K, Jung M, Kristiansen G et al. Hepsin is highly over expressed in and a new candidate for a prognostic indicator in prostate cancer. J Urol 2004;171:187-91.

262. Dhanasekaran SM, Barrette TR, Ghosh D, Shah R, Varambally S, Kurachi K et al. Delineation of prognostic biomarkers in prostate cancer. Nature 2001;412:822-6.

263. Shariat SF, Andrews B, Kattan MW, Kim J, Wheeler TM, Slawin KM. Plasma levels of interleukin-6 and its soluble receptor are associated with prostate cancer progression and metastasis. Urology 2001;58:1008-15.

264. Nakashima J, Tachibana M, Horiguchi Y, Oya M, Ohigashi T, Asakura H, Murai M. Serum interleukin 6 as a prognostic factor in patients with prostate cancer. Clin Cancer Res 2000;6:2702-6.

265. Shariat SF, Kattan MW, Traxel E, Andrews B, Zhu K, Wheeler TM, Slawin KM. Association of pre- and postoperative plasma levels of transforming growth factor beta(1) and interleukin 6 and its soluble receptor with prostate cancer progression. Clin Cancer Res 2004;10:1992-9.

266. Shariat SF, Kattan MW, Song W, Bernard D, Gottenger E, Wheeler TM, Slawin KM. Early postoperative peripheral blood reverse transcription PCR assay for prostate-specific antigen is associated with prostate cancer progression in patients undergoing radical prostatectomy. Cancer Res 2003;63:5874-8.

267. Bradley SV, Oravecz-Wilson KI, Bougeard G, Mizukami I, Li L, Munaco AJ et al. Serum antibodies to huntingtin interacting protein-1: a new blood test for prostate cancer. Cancer Res 2005;65:4126-33.

References 186

268. Minelli A, Ronquist G, Carlsson L, Mearini E, Nilsson O, Larsson A. Antiprostasome antibody titres in benign and malignant prostate disease. Anticancer Res 2005;25:4399-402.

269. Taylor BS, Pal M, Yu J, Laxman B, Kalyana-Sundaram S, Zhao R et al. Humoral response profiling reveals pathways to prostate cancer progression. Mol Cell Proteomics 2008;7:600-11.

270. Shariat SF, Walz J, Roehrborn CG, Zlotta AR, Perrotte P, Suardi N et al. External validation of a biomarker-based preoperative nomogram predicts biochemical recurrence after radical prostatectomy. J Clin Oncol 2008;26:1526-31.

271. Chun FK, Briganti A, Graefen M, Porter C, Montorsi F, Haese A et al. Development and external validation of an extended repeat biopsy nomogram. J Urol 2007;177:510-5.

272. Karakiewicz PI, Hutterer GC. Predictive models and prostate cancer. Nat Clin Pract Urol 2008;5:82-92.

273. Parekh DJ, Ankerst DP, Baillargeon J, Higgins B, Platz EA, Troyer D et al. Assessment of 54 biomarkers for biopsy-detectable prostate cancer. Cancer Epidemiol Biomarkers Prev 2007;16:1966-72.

274. Stephan C, Xu C, Finne P, Cammann H, Meyer HA, Lein M et al. Comparison of two different artificial neural networks for prostate biopsy indication in two different patient populations. Urology 2007;70:596-601.


276. Ornstein DK, Rayford W, Fusaro VA, Conrads TP, Ross SJ, Hitt BA et al. Serum proteomic profiling can discriminate prostate cancer from benign prostates in men with total prostate specific antigen levels between 2.5 and 15.0 ng/ml. J Urol 2004;172:1302-5.

277. Diamandis EP. Point: Proteomic patterns in biological fluids: do they represent the future of cancer diagnostics? Clin Chem 2003;49:1272-5.

278. Diamandis EP. Mass spectrometry as a diagnostic and a cancer biomarker discovery tool: opportunities and potential limitations. Mol Cell Proteomics 2004;3:367-78.

279. Foster BA, Gingrich JR, Kwon ED, Madias C, Greenberg NM. Characterization of prostatic epithelial cell lines derived from transgenic

References 187

adenocarcinoma of the mouse prostate (TRAMP) model. Cancer Res 1997;57:3325-30.

280. Kasper S, Sheppard PC, Yan Y, Pettigrew N, Borowsky AD, Prins GS et al. Development, progression, and androgen-dependence of prostate tumors in probasin-large T antigen transgenic mice: a model for prostate cancer. Lab Invest 1998;78:319-33.

281. Ellwood-Yen K, Graeber TG, Wongvipat J, Iruela-Arispe ML, Zhang J, Matusik R et al. Myc-driven murine prostate cancer shares molecular features with human prostate tumors. Cancer Cell 2003;4:223-38.

282. Di Cristofano A, De Acetis M, Koff A, Cordon-Cardo C, Pandolfi PP. Pten and p27KIP1 cooperate in prostate cancer tumor suppression in the mouse. Nat Genet 2001;27:222-4.

283. DUNNING WF. PROSTATE CANCER IN THE RAT. Natl Cancer Inst Monogr 1963;12:351-69.

284. Aquilina JW, McKinney L, Pacelli A, Richman LK, Waters DJ, Thompson I et al. High grade prostatic intraepithelial neoplasia in military working dogs with and without prostate cancer. Prostate 1998;36:189-93.

285. Hathout Y. Approaches to the study of the cell secretome. Expert Rev Proteomics 2007;4:239-48.

286. McDavid K, Lee J, Fulton JP, Tonita J, Thompson TD. Prostate cancer incidence and mortality rates and trends in the United States and Canada. Public Health Rep 2004;119:174-86.

287. Thompson IM, Pauler DK, Goodman PJ, Tangen CM, Lucia MS, Parnes HL et al. Prevalence of prostate cancer among men with a prostate-specific antigen level < or =4.0 ng per milliliter. N Engl J Med 2004;350:2239-46.

288. Thompson IM, Ankerst DP, Chi C, Lucia MS, Goodman PJ, Crowley JJ et al. Operating characteristics of prostate-specific antigen in men with an initial PSA level of 3.0 ng/ml or lower. JAMA 2005;294:66-70.

289. Ross KS, Carter HB, Pearson JD, Guess HA. Comparative efficiency of prostate-specific antigen screening strategies for prostate cancer detection. JAMA 2000;284:1399-405.

290. Stamey TA, Caldwell M, McNeal JE, Nolley R, Hemenez M, Downs J. The prostate specific antigen era in the United States is over for prostate cancer: what happened in the last 20 years? J Urol 2004;172:1297-301.

References 188

291. Bunting PS. Screening for prostate cancer with prostate-specific antigen: beware the biases. Clin Chim Acta 2002;315:71-97.

292. Venter JC, Adams MD, Myers EW, Li PW, Mural RJ, Sutton GG et al. The sequence of the human genome. Science 2001;291:1304-51.

293. Karas M, Hillenkamp F. Laser desorption ionization of proteins with molecular masses exceeding 10,000 daltons. Anal Chem 1988;60:2299-301.

294. Fenn JB, Mann M, Meng CK, Wong SF, Whitehouse CM. Electrospray ionization for mass spectrometry of large biomolecules. Science 1989;246:64-71.

295. Duffy MJ. Evidence for the clinical use of tumour markers. Ann Clin Biochem 2004;41:370-7.

296. Qian WJ, Jacobs JM, Liu T, Camp DG, Smith RD. Advances and challenges in liquid chromatography-mass spectrometry-based proteomics profiling for clinical applications. Mol Cell Proteomics 2006;5:1727-44.

297. Jemal A, Siegel R, Ward E, Murray T, Xu J, Smigal C, Thun MJ. Cancer statistics, 2006. CA Cancer J Clin 2006;56:106-30.

298. Etzioni R, Urban N, Ramsey S, McIntosh M, Schwartz S, Reid B et al. The case for early detection. Nat Rev Cancer 2003;3:243-52.

299. Duffy MJ. Evidence for the clinical use of tumour markers. Ann Clin Biochem 2004;41:370-7.

300. Ryan CJ, Small EJ. Prostate cancer update: 2005. Curr Opin Oncol 2006;18:284-8.

301. Schmid HP, Riesen W, Prikler L. Update on screening for prostate cancer with prostate-specific antigen. Crit Rev Oncol Hematol 2004;50:71-8.


303. Haese A, Vaisanen V, Lilja H, Kattan MW, Rittenhouse HG, Pettersson K et al. Comparison of predictive accuracy for pathologically organ confined clinical stage T1c prostate cancer using human glandular kallikrein 2 and prostate specific antigen combined with clinical stage and Gleason grade. J Urol 2005;173:752-6.

References 189

304. Stephan C, Jung K, Diamandis EP, Rittenhouse HG, Lein M, Loening SA. Prostate-specific antigen, its molecular forms, and other kallikrein markers for detection of prostate cancer. Urology 2002;59:2-8.

305. Ciatto S, Bonardi R, Mazzotta A, Lombardi C, Santoni R, Cardini S, Zappa M. Comparing two modalities of screening for prostate cancer: digital rectal examination + transrectal ultrasonography vs. prostate-specific antigen. Tumori 1995;81:225-9.

306. Semmes OJ, Malik G, Ward M. Application of mass spectrometry to the discovery of biomarkers for detection of prostate cancer. J Cell Biochem 2006;98:496-503.


308. Diamandis EP. Analysis of serum proteomic patterns for early cancer diagnosis: drawing attention to potential problems. J Natl Cancer Inst 2004;96:353-6.

309. Diamandis EP. Mass spectrometry as a diagnostic and a cancer biomarker discovery tool: opportunities and potential limitations. Mol Cell Proteomics 2004;3:367-78.

310. Stamey TA. The era of serum prostate specific antigen as a marker for biopsy of the prostate and detecting prostate cancer is now over in the USA. BJU Int 2004;94:963-4.

311. Sakr WA, Grignon DJ, Crissman JD, Heilbrun LK, Cassin BJ, Pontes JJ, Haas GP. High grade prostatic intraepithelial neoplasia (HGPIN) and prostatic adenocarcinoma between the ages of 20-69: an autopsy study of 249 cases. In Vivo 1994;8:439-43.

312. Scaros O, Fisler R. Biomarker technology roundup: from discovery to clinical applications, a broad set of tools is required to translate from the lab to the clinic. Biotechniques 2005;Suppl:30-2.

313. Jacobs JM, Adkins JN, Qian WJ, Liu T, Shen Y, Camp DG, Smith RD. Utilizing human blood plasma for proteomic biomarker discovery. J Proteome Res 2005;4:1073-85.

314. Anderson NL, Polanski M, Pieper R, Gatlin T, Tirumalai RS, Conrads TP et al. The human plasma proteome: a nonredundant list developed by combination of four separate sources. Mol Cell Proteomics 2004;3:311-26.

References 190

315. Yousef GM, Polymeris ME, Grass L, Soosaipillai A, Chan PC, Scorilas A et al. Human kallikrein 5: a potential novel serum biomarker for breast and ovarian cancer. Cancer Res 2003;63:3958-65.

316. Christopoulos TK, Diamandis EP. Enzymatically amplified time-resolved fluorescence immunoassay with terbium chelates. Anal Chem 1992;64:342-6.

317. Ashburner M, Ball CA, Blake JA, Botstein D, Butler H, Cherry JM et al. Gene ontology: tool for the unification of biology. The Gene Ontology Consortium. Nat Genet 2000;25:25-9.


319. Roy I, Gupta MN. Freeze-drying of proteins: some emerging concerns. Biotechnol Appl Biochem 2004;39:165-77.

320. Annesley TM. Ion suppression in mass spectrometry. Clin Chem 2003;49:1041-4.

321. Fornarini B, D'Ambrosio C, Natoli C, Tinari N, Silingardi V, Iacobelli S. Adhesion to 90K (Mac-2 BP) as a mechanism for lymphoma drug resistance in vivo. Blood 2000;96:3282-5.

322. Ozaki Y, Kontani K, Teramoto K, Fujita T, Tezuka N, Sawai S et al. Involvement of 90K/Mac-2 binding protein in cancer metastases by increased cellular adhesiveness in lung cancer. Oncol Rep 2004;12:1071-7.

323. Iacobelli S, Sismondi P, Giai M, D'Egidio M, Tinari N, Amatetti C et al. Prognostic value of a novel circulating serum 90K antigen in breast cancer. Br J Cancer 1994;69:172-6.

324. Iacovazzi PA, Guerra V, Elba S, Sportelli F, Manghisi OG, Correale M. Are 90K/MAC-2BP serum levels correlated with poor prognosis in HCC patients? Preliminary results. Int J Biol Markers 2003;18:222-6.

325. Zeimet AG, Natoli C, Herold M, Fuchs D, Windbichler G, Daxenbichler G et al. Circulating immunostimulatory protein 90K and soluble interleukin-2-receptor in human ovarian cancer. Int J Cancer 1996;68:34-8.

326. Greco C, Vona R, Cosimelli M, Matarrese P, Straface E, Scordati P et al. Cell surface overexpression of galectin-3 and the presence of its ligand 90k in the blood plasma as determinants in colon neoplastic lesions. Glycobiology 2004;14:783-92.

References 191

327. Grassadonia A, Tinari N, Iurisci I, Piccolo E, Cumashi A, Innominato P et al. 90K (Mac-2 BP) and galectins in tumor progression and metastasis. Glycoconj J 2004;19:551-6.

328. Bair EL, Nagle RB, Ulmer TA, Laferte S, Bowden GT. 90K/Mac-2 binding protein is expressed in prostate cancer and induces promatrilysin expression. Prostate 2006;66:283-93.

329. Volmer MW, Stuhler K, Zapatka M, Schoneck A, Klein-Scory S, Schmiegel W et al. Differential proteome analysis of conditioned media to detect Smad4 regulated secreted biomarkers in colon cancer. Proteomics 2005;5:2587-601.

330. Lin B, White JT, Lu W, Xie T, Utleg AG, Yan X et al. Evidence for the presence of disease-perturbed networks in prostate cancer cells by genomic and proteomic analyses: a systems approach to disease. Cancer Res 2005;65:3081-91.

331. Ferguson RA, Yu H, Kalyvas M, Zammit S, Diamandis EP. Ultrasensitive detection of prostate-specific antigen by a time-resolved immunofluorometric assay and the Immulite immunochemiluminescent third-generation assay: potential applications in prostate and breast cancers. Clin Chem 1996;42:675-84.

332. Washburn MP, Wolters D, Yates JR, III. Large-scale analysis of the yeast proteome by multidimensional protein identification technology. Nat Biotechnol 2001;19:242-7.

333. Barry MJ. Clinical practice. Prostate-specific-antigen testing for early diagnosis of prostate cancer. N Engl J Med 2001;344:1373-7.

334. Ryan CJ, Small EJ. Prostate cancer update: 2005. Curr Opin Oncol 2006;18:284-8.

335. Constantinou J, Feneley MR. PSA testing: an evolving relationship with prostate cancer screening. Prostate Cancer Prostatic Dis 2006;9:6-13.

336. Kabalin JN, McNeal JE, Johnstone IM, Stamey TA. Serum prostate-specific antigen and the biologic progression of prostate cancer. Urology 1995;46:65-70.

337. Martin BJ, Finlay JA, Sterling K, Ward M, Lifsey D, Mercante D et al. Early detection of prostate cancer in African-American men through use of multiple biomarkers: human kallikrein 2 (hK2), prostate-specific antigen (PSA), and free PSA (fPSA). Prostate Cancer Prostatic Dis 2004;7:132-7.

References 192

338. Stenman UH. New ultrasensitive assays facilitate studies on the role of human glandular kallikrein (hK2) as a marker for prostatic disease. Clin Chem 1999;45:753-4.


340. Haese A, Vaisanen V, Lilja H, Kattan MW, Rittenhouse HG, Pettersson K et al. Comparison of predictive accuracy for pathologically organ confined clinical stage T1c prostate cancer using human glandular kallikrein 2 and prostate specific antigen combined with clinical stage and Gleason grade. J Urol 2005;173:752-6.

341. Stephan C, Jung K, Diamandis EP, Rittenhouse HG, Lein M, Loening SA. Prostate-specific antigen, its molecular forms, and other kallikrein markers for detection of prostate cancer. Urology 2002;59:2-8.

342. Ciatto S, Bonardi R, Mazzotta A, Lombardi C, Santoni R, Cardini S, Zappa M. Comparing two modalities of screening for prostate cancer: digital rectal examination + transrectal ultrasonography vs. prostate-specific antigen. Tumori 1995;81:225-9.

343. Anderson NL, Anderson NG. The human plasma proteome: history, character, and diagnostic prospects. Mol Cell Proteomics 2002;1:845-67.

344. Alexander H, Stegner AL, Wagner-Mann C, Du Bois GC, Alexander S, Sauter ER. Proteomic analysis to identify breast cancer biomarkers in nipple aspirate fluid. Clin Cancer Res 2004;10:7500-10.

345. Rogers MA, Clarke P, Noble J, Munro NP, Paul A, Selby PJ, Banks RE. Proteomic profiling of urinary proteins in renal cancer by surface enhanced laser desorption ionization and neural-network analysis: identification of key issues affecting potential clinical utility. Cancer Res 2003;63:6971-83.

346. Petricoin EF, Ardekani AM, Hitt BA, Levine PJ, Fusaro VA, Steinberg SM et al. Use of proteomic patterns in serum to identify ovarian cancer. Lancet 2002;359:572-7.

347. Villanueva J, Philip J, Entenberg D, Chaparro CA, Tanwar MK, Holland EC, Tempst P. Serum peptide profiling by magnetic particle-assisted, automated sample processing and MALDI-TOF mass spectrometry. Anal Chem 2004;76:1560-70.

References 193

348. Villanueva J, Shaffer DR, Philip J, Chaparro CA, Erdjument-Bromage H, Olshen AB et al. Differential exoprotease activities confer tumor-specific serum peptidome patterns. J Clin Invest 2006;116:271-84.

349. Anderson NL, Anderson NG, Haines LR, Hardie DB, Olafson RW, Pearson TW. Mass spectrometric quantitation of peptides and proteins using Stable Isotope Standards and Capture by Anti-Peptide Antibodies (SISCAPA). J Proteome Res 2004;3:235-44.

350. Ramstrom M, Hagman C, Mitchell JK, Derrick PJ, Hakansson P, Bergquist J. Depletion of high-abundant proteins in body fluids prior to liquid chromatography fourier transform ion cyclotron resonance mass spectrometry. J Proteome Res 2005;4:410-6.

351. Fu Q, Bovenkamp DE, Van Eyk JE. A rapid, economical, and reproducible method for human serum delipidation and albumin and IgG removal for proteomic analysis. Methods Mol Biol 2007;357:365-71.

352. Volmer MW, Stuhler K, Zapatka M, Schoneck A, Klein-Scory S, Schmiegel W et al. Differential proteome analysis of conditioned media to detect Smad4 regulated secreted biomarkers in colon cancer. Proteomics 2005;5:2587-601.

353. Pellitteri-Hahn MC, Warren MC, Didier DN, Winkler EL, Mirza SP, Greene AS, Olivier M. Improved mass spectrometric proteomic profiling of the secretome of rat vascular endothelial cells. J Proteome Res 2006;5:2861-4.

354. Chen X, Cushman SW, Pannell LK, Hess S. Quantitative proteomic analysis of the secretory proteins from rat adipose cells using a 2D liquid chromatography-MS/MS approach. J Proteome Res 2005;4:570-7.

355. Wu CC, Chien KY, Tsang NM, Chang KP, Hao SP, Tsao CH et al. Cancer cell-secreted proteomes as a basis for searching potential tumor markers: nasopharyngeal carcinoma as a model. Proteomics 2005;5:3173-82.

356. An E, Lu X, Flippin J, Devaney JM, Halligan B, Hoffman E et al. Secreted proteome profiling in human RPE cell cultures derived from donors with age related macular degeneration and age matched healthy donors. J Proteome Res 2006;5:2599-610.

357. Perkins DN, Pappin DJ, Creasy DM, Cottrell JS. Probability-based protein identification by searching sequence databases using mass spectrometry data. Electrophoresis 1999;20:3551-67.

358. Craig R, Beavis RC. A method for reducing the time required to match protein sequences with tandem mass spectra. Rapid Commun Mass Spectrom 2003;17:2310-6.

References 194

359. Kersey PJ, Duarte J, Williams A, Karavidopoulou Y, Birney E, Apweiler R. The International Protein Index: an integrated database for proteomics experiments. Proteomics 2004;4:1985-8.

360. Searle, B. C., Brundege, J. M., and Turner, M. Improving Sensitivity by Combining Results from Multiple MS/MS Search Methodologies with the Scaffold Computer Algorithm. Human Proteome Organization (HUPO), 4th Annual World Congress, Munich, Germany . 2006.

Ref Type: Abstract

361. Peri S, Navarro JD, Amanchy R, Kristiansen TZ, Jonnalagadda CK, Surendranath V et al. Development of human protein reference database as an initial platform for approaching systems biology in humans. Genome Res 2003;13:2363-71.

362. Pilch B, Mann M. Large-scale and high-confidence proteomic analysis of human seminal plasma. Genome Biol 2006;7:R40.

363. Kulasingam V, Diamandis EP. Proteomics analysis of conditioned media from three breast cancer cell lines: a mine for biomarkers and therapeutic targets. Mol Cell Proteomics 2007;6:1997-2011.

364. Makarov DV, Carter HB. The discovery of prostate specific antigen as a biomarker for the early detection of adenocarcinoma of the prostate. J Urol 2006;176:2383-5.


366. Afzal S, Ahmad M, Mushtaq S, Mubarik A, Qureshi AH, Khan SA. Morphological features correlation with serum tumour markers in prostatic carcinoma. J Coll Physicians Surg Pak 2003;13:511-4.

367. Bair EL, Nagle RB, Ulmer TA, Laferte S, Bowden GT. 90K/Mac-2 binding protein is expressed in prostate cancer and induces promatrilysin expression. Prostate 2006;66:283-93.

368. Hale LP, Price DT, Sanchez LM, Demark-Wahnefried W, Madden JF. Zinc alpha-2-glycoprotein is expressed by malignant prostatic epithelium and may serve as a potential serum marker for prostate cancer. Clin Cancer Res 2001;7:846-53.

369. Das S, Hahn Y, Nagata S, Willingham MC, Bera TK, Lee B, Pastan I. NGEP, a prostate-specific plasma membrane protein that promotes the association of LNCaP cells. Cancer Res 2007;67:1594-601.

References 195

370. Elsasser-Beile U, Wolf P, Gierschner D, Buhler P, Schultze-Seemann W, Wetterauer U. A new generation of monoclonal and recombinant antibodies against cell-adherent prostate specific membrane antigen for diagnostic and therapeutic targeting of prostate cancer. Prostate 2006;66:1359-70.

371. McPherson SJ, Mellor SL, Wang H, Evans LW, Groome NP, Risbridger GP. Expression of activin A and follistatin core proteins by human prostate tumor cell lines. Endocrinology 1999;140:5303-9.

372. Wang QF, Tilly KI, Tilly JL, Preffer F, Schneyer AL, Crowley WF, Jr., Sluss PM. Activin inhibits basal and androgen-stimulated proliferation and induces apoptosis in the human prostatic cancer cell line, LNCaP. Endocrinology 1996;137:5476-83.

373. Nakagawa H, Liyanarachchi S, Davuluri RV, Auer H, Martin EW, Jr., de la CA, Frankel WL. Role of cancer-associated stromal fibroblasts in metastatic colon cancer to the liver and their expression profiles. Oncogene 2004;23:7366-77.

374. Di Simone N, Crowley WF, Jr., Wang QF, Sluss PM, Schneyer AL. Characterization of inhibin/activin subunit, follistatin, and activin type II receptors in human ovarian cancer cell lines: a potential role in autocrine growth regulation. Endocrinology 1996;137:486-94.

375. Hase K, Murakami T, Takatsu H, Shimaoka T, Iimura M, Hamura K et al. The membrane-bound chemokine CXCL16 expressed on follicle-associated epithelium and M cells mediates lympho-epithelial interaction in GALT. J Immunol 2006;176:43-51.

376. Camozzi M, Rusnati M, Bugatti A, Bottazzi B, Mantovani A, Bastone A et al. Identification of an antiangiogenic FGF2-binding site in the N terminus of the soluble pattern recognition receptor PTX3. J Biol Chem 2006;281:22605-13.

377. Manda R, Kohno T, Matsuno Y, Takenoshita S, Kuwano H, Yokota J. Identification of genes (SPON2 and C20orf2) differentially expressed between cancerous and noncancerous lung cells by mRNA differential display. Genomics 1999;61:5-14.

378. Jacobs JM, Adkins JN, Qian WJ, Liu T, Shen Y, Camp DG, Smith RD. Utilizing human blood plasma for proteomic biomarker discovery. J Proteome Res 2005;4:1073-85.

379. Leak LV, Liotta LA, Krutzsch H, Jones M, Fusaro VA, Ross SJ et al. Proteomic analysis of lymph. Proteomics 2004;4:753-65.

References 196

380. Drake RR, Cazare LH, Semmes OJ, Wadsworth JT. Serum, salivary and tissue proteomics for discovery of biomarkers for head and neck cancers. Expert Rev Mol Diagn 2005;5:93-100.

381. Etzioni R, Urban N, Ramsey S, McIntosh M, Schwartz S, Reid B et al. The case for early detection. Nat Rev Cancer 2003;3:243-52.

382. Hill R, Song Y, Cardiff RD, Van Dyke T. Heterogeneous tumor evolution initiated by loss of pRb function in a preclinical prostate cancer model. Cancer Res 2005;65:10243-54.

383. Shah RB, Mehra R, Chinnaiyan AM, Shen R, Ghosh D, Zhou M et al. Androgen-independent prostate cancer is a heterogeneous group of diseases: lessons from a rapid autopsy program. Cancer Res 2004;64:9209-16.

384. Barnidge DR, Goodmanson MK, Klee GG, Muddiman DC. Absolute quantification of the model biomarker prostate-specific antigen in serum by LC-Ms/MS using protein cleavage and isotope dilution mass spectrometry. J Proteome Res 2004;3:644-52.

385. Kapp EA, Schutz F, Connolly LM, Chakel JA, Meza JE, Miller CA et al. An evaluation, comparison, and accurate benchmarking of several publicly available MS/MS search algorithms: sensitivity and specificity analysis. Proteomics 2005;5:3475-90.

386. Keller A, Nesvizhskii AI, Kolker E, Aebersold R. Empirical statistical model to estimate the accuracy of peptide identifications made by MS/MS and database search. Anal Chem 2002;74:5383-92.

387. Nesvizhskii AI, Keller A, Kolker E, Aebersold R. A statistical model for identifying proteins by tandem mass spectrometry. Anal Chem 2003;75:4646-58.

388. Higdon R, Hogan JM, Van Belle G, Kolker E. Randomized sequence databases for tandem mass spectrometry peptide and protein identification. OMICS 2005;9:364-79.

389. Fujii Y, Kawakami S, Okada Y, Kageyama Y, Kihara K. Regulation of prostate-specific antigen by activin A in prostate cancer LNCaP cells. Am J Physiol Endocrinol Metab 2004;286:E927-E931.

390. Thomas TZ, Wang H, Niclasen P, O'Bryan MK, Evans LW, Groome NP et al. Expression and localization of activin subunits and follistatins in tissues from men with high grade prostate cancer. J Clin Endocrinol Metab 1997;82:3851-8.

References 197

391. Risbridger GP, Mellor SL, McPherson SJ, Schmitt JF. The contribution of inhibins and activins to malignant prostate disease. Mol Cell Endocrinol 2001;180:149-53.

392. Heydtmann M, Lalor PF, Eksteen JA, Hubscher SG, Briskin M, Adams DH. CXC chemokine ligand 16 promotes integrin-mediated adhesion of liver-infiltrating lymphocytes to cholangiocytes and hepatocytes within the inflamed human liver. J Immunol 2005;174:1055-62.

393. Hojo S, Koizumi K, Tsuneyama K, Arita Y, Cui Z, Shinohara K et al. High-level expression of chemokine CXCL16 by tumor cells correlates with a good prognosis and increased tumor-infiltrating lymphocytes in colorectal cancer. Cancer Res 2007;67:4725-31.

394. Wagsater D, Hugander A, Dimberg J. Expression of CXCL16 in human rectal cancer. Int J Mol Med 2004;14:65-9.

395. Willeke F, Assad A, Findeisen P, Schromm E, Grobholz R, von Gerstenbergk B et al. Overexpression of a member of the pentraxin family (PTX3) in human soft tissue liposarcoma. Eur J Cancer 2006;42:2639-46.

396. Klouche M, Brockmeyer N, Knabbe C, Rose-John S. Human herpesvirus 8-derived viral IL-6 induces PTX3 expression in Kaposi's sarcoma cells. AIDS 2002;16:F9-18.

397. Simon I, Liu Y, Krall KL, Urban N, Wolfert RL, Kim NW, McIntosh MW. Evaluation of the novel serum markers B7-H4, Spondin 2, and DcR3 for diagnosis and early detection of ovarian cancer. Gynecol Oncol 2007;106:112-8.

398. Walz A, Schmutz P, Mueller C, Schnyder-Candrian S. Regulation and function of the CXC chemokine ENA-78 in monocytes and its role in disease. J Leukoc Biol 1997;62:604-11.

399. Hochreiter WW, Nadler RB, Koch AE, Campbell PL, Ludwig M, Weidner W, Schaeffer AJ. Evaluation of the cytokines interleukin 8 and epithelial neutrophil activating peptide 78 as indicators of inflammation in prostatic secretions. Urology 2000;56:1025-9.

400. Engl T, Relja B, Blumenberg C, Muller I, Ringel EM, Beecken WD et al. Prostate tumor CXC-chemokine profile correlates with cell adhesion to endothelium and extracellular matrix. Life Sci 2006;78:1784-93.

401. Speetjens FM, Kuppen PJ, Sandel MH, Menon AG, Burg D, van de Velde CJ et al. Disrupted expression of CXCL5 in colorectal cancer is associated with

References 198

rapid tumor formation in rats and poor prognosis in patients. Clin Cancer Res 2008;14:2276-84.

402. Park JY, Park KH, Bang S, Kim MH, Lee JE, Gang J et al. CXCL5 overexpression is associated with late stage gastric cancer. J Cancer Res Clin Oncol 2007;133:835-40.

403. Friedl A, Stoesz SP, Buckley P, Gould MN. Neutrophil gelatinase-associated lipocalin in normal and neoplastic human tissues. Cell type-specific pattern of expression. Histochem J 1999;31:433-41.

404. Ding H, He Y, Li K, Yang J, Li X, Lu R, Gao W. Urinary neutrophil gelatinase-associated lipocalin (NGAL) is an early biomarker for renal tubulointerstitial injury in IgA nephropathy. Clin Immunol 2007;123:227-34.

405. Lim R, Ahmed N, Borregaard N, Riley C, Wafai R, Thompson EW et al. Neutrophil gelatinase-associated lipocalin (NGAL) an early-screening biomarker for ovarian cancer: NGAL is associated with epidermal growth factor-induced epithelio-mesenchymal transition. Int J Cancer 2007;120:2426-34.

406. Bauer M, Eickhoff JC, Gould MN, Mundhenke C, Maass N, Friedl A. Neutrophil gelatinase-associated lipocalin (NGAL) is a predictor of poor prognosis in human primary breast cancer. Breast Cancer Res Treat 2007.

407. Fernandez CA, Yan L, Louis G, Yang J, Kutok JL, Moses MA. The matrix metalloproteinase-9/neutrophil gelatinase-associated lipocalin complex plays a role in breast tumor growth and is present in the urine of breast cancer patients. Clin Cancer Res 2005;11:5390-5.

408. Zhang H, Xu L, Xiao D, Xie J, Zeng H, Wang Z et al. Upregulation of neutrophil gelatinase-associated lipocalin in oesophageal squamous cell carcinoma: significant correlation with cell differentiation and tumour invasion. J Clin Pathol 2007;60:555-61.

409. Lee HJ, Lee EK, Lee KJ, Hong SW, Yoon Y, Kim JS. Ectopic expression of neutrophil gelatinase-associated lipocalin suppresses the invasion and liver metastasis of colon cancer cells. Int J Cancer 2006;118:2490-7.

410. Mitsuyama K, Tsuruta O, Tomiyasu N, Takaki K, Suzuki A, Masuda J et al. Increased circulating concentrations of growth-related oncogene (GRO)-alpha in patients with inflammatory bowel disease. Dig Dis Sci 2006;51:173-7.

411. Moore BB, Arenberg DA, Stoy K, Morgan T, Addison CL, Morris SB et al. Distinct CXC chemokines mediate tumorigenicity of prostate cancer cells. Am J Pathol 1999;154:1503-12.

References 199

412. Wen Y, Giardina SF, Hamming D, Greenman J, Zachariah E, Bacolod MD et al. GROalpha is highly expressed in adenocarcinoma of the colon and down-regulates fibulin-1. Clin Cancer Res 2006;12:5951-9.

413. Ashkenazi A, Dixit VM. Apoptosis control by death and decoy receptors. Curr Opin Cell Biol 1999;11:255-60.

414. Pitti RM, Marsters SA, Lawrence DA, Roy M, Kischkel FC, Dowd P et al. Genomic amplification of a decoy receptor for Fas ligand in lung and colon cancer. Nature 1998;396:699-703.

415. Ohshima K, Haraoka S, Sugihara M, Suzumiya J, Kawasaki C, Kanda M, Kikuchi M. Amplification and expression of a decoy receptor for fas ligand (DcR3) in virus (EBV or HTLV-I) associated lymphomas. Cancer Lett 2000;160:89-97.

416. Takahama Y, Yamada Y, Emoto K, Fujimoto H, Takayama T, Ueno M et al. The prognostic significance of overexpression of the decoy receptor for Fas ligand (DcR3) in patients with gastric carcinomas. Gastric Cancer 2002;5:61-8.

417. Arakawa Y, Tachibana O, Hasegawa M, Miyamori T, Yamashita J, Hayashi Y. Frequent gene amplification and overexpression of decoy receptor 3 in glioblastoma. Acta Neuropathol 2005;109:294-8.

418. Shen HW, Gao SL, Wu YL, Peng SY. Overexpression of decoy receptor 3 in hepatocellular carcinoma and its association with resistance to Fas ligand-mediated apoptosis. World J Gastroenterol 2005;11:5926-30.

419. Wu Y, Han B, Sheng H, Lin M, Moore PA, Zhang J, Wu J. Clinical significance of detecting elevated serum DcR3/TR6/M68 in malignant tumor patients. Int J Cancer 2003;105:724-32.

420. Hwang SL, Lin CL, Cheng CY, Lin FA, Lieu AS, Howng SL, Lee KS. Serum concentration of soluble decoy receptor 3 in glioma patients before and after surgery. Kaohsiung J Med Sci 2004;20:124-7.

421. Black MH, Magklara A, Obiezu CV, Melegos DN, Diamandis EP. Development of an ultrasensitive immunoassay for human glandular kallikrein with no cross-reactivity from prostate-specific antigen. Clin Chem 1999;45:790-9.

422. Blobe GC, Schiemann WP, Lodish HF. Role of transforming growth factor beta in human disease. N Engl J Med 2000;342:1350-8.

References 200

423. Kim SJ, Im YH, Markowitz SD, Bang YJ. Molecular mechanisms of inactivation of TGF-beta receptors during carcinogenesis. Cytokine Growth Factor Rev 2000;11:159-68.

424. Robertson DM, Klein R, de Vos FL, McLachlan RI, Wettenhall RE, Hearn MT et al. The isolation of polypeptides with FSH suppressing activity from bovine follicular fluid which are structurally different to inhibin. Biochem Biophys Res Commun 1987;149:744-9.

425. Schneyer AL, Rzucidlo DA, Sluss PM, Crowley WF, Jr. Characterization of unique binding kinetics of follistatin and activin or inhibin in serum. Endocrinology 1994;135:667-74.

426. Wada M, Shintani Y, Kosaka M, Sano T, Hizawa K, Saito S. Immunohistochemical localization of activin A and follistatin in human tissues. Endocr J 1996;43:375-85.

427. Fainsod A, Deissler K, Yelin R, Marom K, Epstein M, Pillemer G et al. The dorsalizing and neural inducing gene follistatin is an antagonist of BMP-4. Mech Dev 1997;63:39-50.

428. Thomas TZ, Chapman SM, Hong W, Gurusingfhe C, Mellor SL, Fletcher R et al. Inhibins, activins, and follistatins: expression of mRNAs and cellular localization in tissues from men with benign prostatic hyperplasia. Prostate 1998;34:34-43.

429. Nakamura T, Sugino K, Titani K, Sugino H. Follistatin, an activin-binding protein, associates with heparan sulfate chains of proteoglycans on follicular granulosa cells. J Biol Chem 1991;266:19432-7.

430. de Winter JP, ten Dijke P, de Vries CJ, van Achterberg TA, Sugino H, de Waele P et al. Follistatins neutralize activin bioactivity by inhibition of activin binding to its type II receptors. Mol Cell Endocrinol 1996;116:105-14.

431. Hashimoto O, Nakamura T, Shoji H, Shimasaki S, Hayashi Y, Sugino H. A novel role of follistatin, an activin-binding protein, in the inhibition of activin action in rat pituitary cells. Endocytotic degradation of activin and its acceleration by follistatin associated with cell-surface heparan sulfate. J Biol Chem 1997;272:13835-42.

432. Ogino H, Yano S, Kakiuchi S, Muguruma H, Ikuta K, Hanibuchi M et al. Follistatin suppresses the production of experimental multiple-organ metastasis by small cell lung cancer cells in natural killer cell-depleted SCID mice. Clin Cancer Res 2008;14:660-7.

Appendix: Assay Reproducibility and Precision

201

APPENDIX

ASSAY REPRODUCIBILITY AND PRECISION

Appendix: Assay Reproducibility and Precision 202

Mac-2BP

0 25 50 75 100 125 150 175 2000

255075

100125150175200

n = 2

Theoretical Concentrationsng/mL

Expe

rim

enta

lC

once

ntra

tions

ng/m

L

KLK5

0.01 0.1 1 10 1000.01

0.1

1

10

100

n = 2

Theoretical Concentrationsug/L

Expe

rim

enta

lC

once

ntra

tions

ug/L


KLK6

0.1 1 10 1000.1

1

10

100

n = 2

Theoretical Concentrationsug/L

Expe

rim

enta

lC

once

ntra

tions

ug/L

KLK2

10 100 1000 1000010

100

1000

10000

n = 2

Theoretical Concentrationsng/L

Expe

rim

enta

lC

once

nrat

ions

ng/L


KLK11

100 1000 10000 100000100

1000

10000

100000

n = 2


Expe

rim

enta

lC

once

ntra

tions

ng/L

PSA

10 100 1000 1000010

100

1000

10000

n = 2

Theoretical Concentratinsng/L

Expe

rim

enta

lC

once

ntra

tions

ng/L


SPON 2

0 25 50 75 100 125 150 175 2000.00

0.25

0.50

0.75

1.00

1.25

1.50

n = 2

Theoretical Concentrationng/mL

Expe

rim

enta

l Cou

nts

CXCL16

0 100 200 300 400 500 600 700 800 90010000

250

500

750

1000

n = 2

Theoretical Controlsng/L

Expe

rim

enta

l Con

trol

sng

/L


PTX3

0 400 800 1200 1600 2000 24000

400

800

1200

1600

2000

2400

n = 2


Expe

cted

Con

cent

ratio

nsng

/L

Follistatin

0 4000 8000 12000 160000

4000

8000

12000

16000

n = 2


Expe

rim

enta

lC

once

ntra

tions

ng/L


CXCL1

0 100 200 300 400 500 600 700 800 90010000

100200300400500600700800900

1000

n = 2


Expe

rim

enta

lC

once

ntra

tions

ng/L

CXCL5

0 250 500 750 1000 1250 1500 1750 20000

250500750

10001250150017502000

n = 2


Expe

rim

enta

lC

once

ntra

tions

ng/L

Appendix: Assay Reproducibility and Precision

208

TNFRSF6B

0 2000 4000 6000 8000 10000 120000

2000

4000

6000

8000

10000

12000

n = 2


Expe

rim

enta

lC

once

ntra

tions

ng/L

Lipocalin 2

0 1 2 3 4 5 6 7 8 9 100123456789

10

n = 2Theoretical Concentration

ng/L

Expe

rim

enta

lC

once

ntra

tion

ng/L

PROTEOMIC ANALYSIS OF PROSTATE CANCER CELL LINE … · 2013-10-11 · cancer cell lines to identify secreted proteins that could serve as novel prostate ... project. As well as Drs

Documents