INVESTIGATION OF METABOLIC REWIRING IN PROSTATE … · little is known about therapy induced metabolic alterations that help to facilitate cancer cell survival and drive disease progression

INVESTIGATION OF METABOLIC

REWIRING IN PROSTATE CANCER CELLS

DURING THE ADAPTIVE RESPONSE TO

ANDROGEN-TARGETED THERAPIES

Kaylyn Davis Tousignant

B.S. Biological Science and Allied Health

School of Biomedical Sciences

Faculty of Health

Queensland University of Technology

Submitted in fulfilment of the requirement for the degree of

Doctor of Philosophy

2020

Investigation of metabolic rewiring in prostate cancer cells during the adaptive response to androgen-

targeted therapies i

Keywords

Prostate, prostate cancer, lipid metabolism, lipogenesis, lipid uptake, lipid

transporters, phospholipids, free fatty acids, lipid remodelling, phospholipases, AR-

targeted therapies, drug resistance, lipidome, lipidomics, transcriptomics, tumour

progression model.


targeted therapies ii

Abstract

It is well established that androgen signalling is fundamental to prostate

cancer (PCa) growth, and that suppressing the androgen axis results in tumour

regression. Consequently, AR-targeted therapies (ATT) remain the mainstay treatment

for patients with advanced PCa. Unfortunately, as in many cancer types, acquired

treatment resistance by cancer cells ultimately results in relapse and disease

progression. While metabolic reprogramming is a well-described hallmark of cancer,

little is known about therapy induced metabolic alterations that help to facilitate cancer

cell survival and drive disease progression. The present study investigated ATT-

induced metabolic rewiring in PCa. Here, in vitro models of long-term ATTs were

characterised by quantitative fluorescent microscopy, lipid and protein mass

spectrometry, 13C metabolomics, transcriptomics, and real-time cell confluence

imaging for adaptive changes in lipid metabolism, the proteome and transcriptome,

mitochondrial activity and proliferation over the 21-day treatment course. ATTs drove

cells into growth arrest, lowered ATP levels and only modestly increased cell death.

Surprisingly, cell quiescence was associated with increased lipid content, and

enhanced uptake of cholesterol, low-density lipoprotein, and lysophospholipids.

Lipidomics analysis revealed extensive lipid remodelling, including a decrease in lipid

storage (triacylglycerols and cholesterol esters) and increases in essential fatty acids,

phospholipids and sphingomyelin as well as in the elongation and desaturation of fatty

acids. Lipid uptake and remodelling via PLA2G2A mediated activity was identified as

a novel adaptive response pathway associated with PCa cell survival. PLA2G2A was

further investigated for its therapeutic potential as a co-treatment with current ATTs.

The findings described in this study suggest that enhanced lipid uptake and

remodelling may serve as novel therapeutic targets to complement current ATTs in

order to prevent therapy resistance and progression to castrate-resistant prostate

cancer.


targeted therapies iii

Table of Contents

Abstract .................................................................................................................................... ii

Table of Contents .................................................................................................................... iii

List of Figures ...........................................................................................................................v

List of Tables ......................................................................................................................... vii

Acknowledgements ............................................................................................................... viii

List of Abbreviations ................................................................................................................x

Statement of Original Authorship .......................................................................................... xii

Awards and publications ....................................................................................................... xiii

Introduction ......................................................................................................... 15

1.1 Introductory statement ..................................................................................................15

1.2 Prostate cancer ..............................................................................................................16

1.3 Lipid metabolism ..........................................................................................................23

1.4 Lipid metabolism in prostate cancer .............................................................................41

1.5 Thesis outline ................................................................................................................46

Materials and Methods ....................................................................................... 48

2.1 Cell culture ...................................................................................................................48

2.2 RNA extraction and quantitative real-time polymerase chain reaction (PCR) .............48

2.3 Detection of lipid content using quantitative fluorescent microscopy (qFM) ..............51

2.4 Measurement of lipid uptake using quantitative fluorescent microscopy (qFM) .........51

2.5 Measurment of glucose uptake .....................................................................................52

2.6 Cell viability, live/dead staining and live-cell imaging assays .....................................52

2.7 Protein extraction and Western blot analysis ................................................................53

2.8 Immunofluorescence staining .......................................................................................54

2.9 Membrane fraction protein mass spectrometry ............................................................55

2.10 Isobaric mass tagging protein mass spectrometry ........................................................55

2.11 Cistrome analysis of AR ChIPseq peaks ......................................................................56

2.12 RNA sequencing analysis .............................................................................................56

2.13 Microarray gene expression profiling using the 180k VPC custom arrays ..................57

2.14 Microarray data analysis ...............................................................................................58

2.15 Lipid extraction .............................................................................................................58

2.16 Lipidomics analysis ......................................................................................................59

2.17 Metabolomics ...............................................................................................................60

2.18 Phospholipase A2 Activity ...........................................................................................61

2.19 Enzyme-linked immunosorbent assay (ELISA) ...........................................................61


targeted therapies iv

2.20 Statistical analysis ........................................................................................................ 61

Androgen regulation of lipid uptake .................................................................. 63

3.1 Introduction .................................................................................................................. 63

3.2 Results .......................................................................................................................... 65

3.3 Discussion .................................................................................................................... 84

Enhanced lipid uptake fuels the extensive remodelling of the PCa lipidome in

response to androgen-targeted therapies ............................................................... 91

4.1 Introduction .................................................................................................................. 91

4.2 Results .......................................................................................................................... 92

4.3 Discussion .................................................................................................................. 133

PLA2G2A is a novel target to fight the development of therapy resistance in

PCa cells .................................................................................................................. 140

5.1 Introduction ................................................................................................................ 140

5.2 Results ........................................................................................................................ 142

5.3 Discussion .................................................................................................................. 163

Overall discussion and future directions ......................................................... 168

6.1 Delineation of the lipid transporter landscape and androgen regulation of lipid uptake

in PCa ................................................................................................................................... 168

6.2 Lipid remodelling is a novel adaptive phenotype in response to ATTs ..................... 170

6.3 Lipid uptake is the major contributor to the increased lipid accumulation induced by

ATT-treatment ...................................................................................................................... 173

6.4 Mechanistic insights into the role of secreted phospholipase PLA2G2A in prostate

cancer ................................................................................................................................... 174

6.5 PLA2G2A represents a novel therapeutic target to combat ATT-induced lipid

remodelling and delay progression to CRPC ....................................................................... 176

6.6 Summary .................................................................................................................... 177

Appendices .............................................................................................................. 179

Appendix A : Supplementary Figures .................................................................................. 179

Appendix B : Resources and funding ................................................................................... 186

Appendix C : Coursework .................................................................................................... 186

Appendix D : Collaborative Arrangements .......................................................................... 186

Appendix E : Intellectual Property ....................................................................................... 186

Bibliography ........................................................................................................... 187


targeted therapies v

List of Figures

Figure 1.1 PCa incidence .......................................................................................... 16

Figure 1.2 Gleason grading system ........................................................................... 18

Figure 1.3 Adaptive changes to lipid metabolism are associated with

progression to CRPC.................................................................................... 20

Figure 1.4 Lipid structure representation of major lipid species ............................... 24

Figure 1.5 Example of fatty acid nomenclature counting from carboxyl end ........... 25

Figure 1.6 Structural representation of cholesterol and cholesteryl esters ................ 27

Figure 1.7 Example of membrane lipid bilayer composition .................................... 28

Figure 1.8 Mechanisms of cellular lipid acquisition ................................................. 32

Figure 1.9 DGAT activity and lipid droplet biogenesis ............................................ 35

Figure 1.10 Androgen regulation of lipid metabolism .............................................. 42

Figure 1.11 LNCaP longitudinal xenograft study shows increased FA content

throughout progression to CRPC ................................................................. 44

Figure 3.1 Androgens increase lipid content of AR-positive PCa cell lines ............. 66

Figure 3.2 Androgens strongly increase lipid uptake ................................................ 69

Figure 3.3 Androgen-enhanced lipid uptake is independent of cell cycle

progression and proliferation ....................................................................... 72

Figure 3.4 Delineation of the lipid transporter landscape in PCa ............................. 76

Figure 3.5 AR binding sites and the androgen regulated expression of lipid

... transporters………………………………………………………………81

Figure 3.6 Androgens regulated the transcript and protein expression of lipid

transporters in vitro and in vivo ................................................................... 83

Figure 3.7 Androgen receptor regulates lipid uptake and lipogenesis ...................... 88

Figure 4.1 Long-term in vitro model to study the adaptive response of LNCaP

cells to treatment with ATTs ...................................................................................... 94

Figure 4.2 Transcriptomic profiling of LNCaP cells undergoing ATT .................... 97

Figure 4.3 ATTs induce vast lipid remodelling in PCa cells .................................... 99

Figure 4.4 Integrated analysis of LNCaP transcriptome and lipidome ................... 101

Figure 4.5 Proteomic analysis of Enz treated LNCaP cells .................................... 103

Figure 4.6 Increased lipid content is an adaptive response to ATT ........................ 106

Figure 4.7 Lipidomics analysis ............................................................................... 108

Figure 4.8 Analysis of lipid composition in cell culture media .............................. 109

Figure 4.9 Enhanced lipid uptake in response to ATT ............................................ 112


targeted therapies vi

Figure 4.10 Enz-induced upregulation of transcripts encoding lipid

transporters ................................................................................................. 115

Figure 4.11 De novo lipogenesis decreases in the early adaptive response to

ATT ............................................................................................................ 116

Figure 4.12 Fatty acid remodelling contributes to the adaptive response of PCa

cells to ATTs .............................................................................................. 118

Figure 4.13 RSL3 sensitivity ................................................................................... 122

Figure 4.14 Enz induces expression of PLA2G2A in PCa cells ............................. 125

Figure 4.15 Schematic of fluorescent PLA2G2A substrates ................................... 126

Figure 4.16 Exogenous PLA2G2A promotes lysolipid uptake in Enz treated

PCa cells ..................................................................................................... 128

Figure 4.17 Lipid uptake in conditioned media ...................................................... 130

Figure 4.18 Summary of lipid metabolic pathways altered by Enzalutamide ......... 132

Figure 4.19 ATTs induce rewiring of metabolic networks in PCa cells to fuel

survival ....................................................................................................... 133

Figure 5.1 Phospholipid metabolism via Lands Cycle and Kennedy Pathway ....... 140

Figure 5.2 The multifunctional role of sPLA2 in cancer ........................................ 141

Figure 5.3 PLA2G2A is upregulated in PCa and is associated with higher

Gleason score ............................................................................................. 144

Figure 5.4 PLA2G2A is androgen repressed in PCa cells ...................................... 146

Figure 5.5 Transcript levels of PLA2 family members in PCa cells ....................... 148

Figure 5.6 Targeting PLA2G2A in PCa cells by gene silencing ............................. 152

Figure 5.7 Characterisation of small molecule inhibitors of PLA2G2A in vitro .... 155

Figure 5.8 Small molecule inhibition of PLA2G2A activity .................................. 156

Figure 5.9 KH064 and arachidonic acid combination treatment ............................. 158

Figure 5.10 Co-targeting AR and PLA2G2A in PCa cells ...................................... 160

Figure 5.11 Targeting PLA2G2A in an LNCaP tumour xenograft model of

CRPC progression ...................................................................................... 161

Figure 5.12 Co-targeting PLA2G2A and AR in vivo .............................................. 162

Figure 6.1 The role of lipid rafts in cell signalling .................................................. 171

Figure A1 Characterisation of the effects of chronic Enz treatment on LNCaP

cells ............................................................................................................ 179

Figure A2 Androgen regulation of lipid transporter transcript levels in DuCaP

and VCaP cells ........................................................................................... 181

Figure A3 LDLR and SCARB1 in PCa cells .......................................................... 183

Figure A4 Androgen regulation of LDLR and SCARB1 in LNCaP cells .............. 183

Figure A5 Top 100 deregulated protein IDs measured by mass spectrometry ....... 184

Figure A6 PLA2G2A in LNCaP cells following up to 21 days Enz treatment ....... 185


targeted therapies vii

List of Tables

Table 2.1 Table of forward and reverse primer sequences used for qRT-PCR ......... 50

Table 2.2 Internal standards for lipid quantitation .................................................... 59

Table 4.2 Fatty acids detected by GCMS FAME in LNCaP cells following

Enzalutamide treatment ............................................................................. 100

https://connectqutedu-my.sharepoint.com/personal/n9774416_qut_edu_au/Documents/Desktop/Examiner%20response/Final%20Thesis_Tousignant_revisions_121219.docx#_Toc27138282

https://connectqutedu-my.sharepoint.com/personal/n9774416_qut_edu_au/Documents/Desktop/Examiner%20response/Final%20Thesis_Tousignant_revisions_121219.docx#_Toc27138282


targeted therapies viii

Acknowledgements

First, I would like to express my deepest gratitude to my principal supervisor,

Prof. Colleen Nelson, for welcoming me in to the APCRC-Q and for your continuous

guidance throughout my PhD journey. In addition, thank you for providing me with a

Supervisor’s Scholarship to financially support my time here. When I first came to

Australia, I could have never imagined that this is where I’d be four years later. Thank

you for being a role model to me throughout this adventure.

To Dr. Martin Sadowski, there is nothing I can say to thank you enough for

everything you’ve taught me in the past few years. Your knowledge and skill set were

critical in the development of this project, but more importantly, your passion,

enthusiasm and encouragement are what made my experience here so valuable.

Whenever I had doubts about data or aspects of the project itself, you showed me how

to turn those thoughts around and create something positive out of them. I truly

wouldn’t have been able to finish this PhD while also maintaining my love for science

(and sanity) without your help throughout the way.

To Dr. Jenni Gunter and Dr. Lisa Philp, a huge thank you to you both for what

you’ve taught me in lab and for your continuous feedback on lab presentations, thesis

chapters and manuscripts. Most importantly, thank you for your help and support with

the in vivo work, I literally could not have done that without you. On that note, I owe

Mr. Mahmudul Haque a thank you for doing daily IPs for me when I physically could

not. I still owe you one!

To all the APCRC-Q members, I am grateful to have had the chance to do my

PhD amongst such a welcoming, knowledgeable and diverse group of people. The

valuable feedback from lab presentations and team meetings has positively contributed

to my thesis so much, and I wish all the best for each and every one of you! Thank you

for the making this time such a memorable experience.

To the team at CARF- Dr. Steven Blanksby, Dr. Berwyck Poad, Dr. Rajesh

Gupta and Mr. Reuben Young, a huge thank you to you all for your guidance and

expertise with the lipidomics analysis, which became a critical part of this thesis. I’d

also like to thank Dr. Ali Talebi for his contribution in our metabolomics analysis.

Finally, I would like to sincerely thank the Movember Foundation and the Prostate


targeted therapies ix

Cancer Foundation of Australia for providing the funding to support this project

through a Movember Revolutionary Team Award.


targeted therapies x

List of Abbreviations

AA Arachidonic acid

ACACA Acetyl-CoA carboxylase- α

ACAT Acetyl-CoA acetyltransferase

ACLY ATP citrate lyase

ACSL Acyl-CoA synthetase

ACBP Acyl-CoA binding proteins

ACSS1/2 Acyl-CoA synthetase short-chain family member 1/2

ADT Androgen deprivation therapy

AKT Serine/Threonine Kinase

AR Androgen receptor

ATGL Adipose triglyceride lipase

ATP Adenosine triphosphate

ATT Androgen targeted therapies

BSA Bovine serum albumin

cDNA Complementary deoxyribonucleic acid

CE Cholesteryl ester

CO2 Carbon dioxide

CRPC Castrate-resistant prostate cancer

CSS Charcoal stripped serum

DAPI 6-Diamidino-2-phenylindole

DGAT1/2 Diacylglycerol O-acyltransferase 1/2

DHT 5α-Dihydrotestosterone

DMEM Dulbecco’s Modified Eagle’s Medium

DMSO Dimethylsulfoxide

DNA Deoxyribonucleic acid

DNL de novo lipogenesis

Enz Enzalutamide

EMT Epithelial to mesenchymal transition

EtOH Ethanol

ELOVL1-7 Fatty acid elongase 1-7

FA Fatty acids

FAME Fatty acyl methyl ester

FASN Fatty acid synthase

FATP Fatty acid transport proteins

FBS Fetal bovine serum

GCMS Gas chromatography-mass spectrometry

GOT2 Glutamic-oxaloacetic transaminase 2

GSEA Gene set enrichment analysis

GSVA Gene set variation analysis

HMGCR HMG-CoA Reductase


targeted therapies xi

HMGCS HMG-CoA Synthase

LA/ALA Linoleic acid/alpha linolenic acid

LCMS Liquid chromatography-mass spectrometry

LDLR Low density lipoprotein receptor

LD Lipid droplet

LPL Lipoprotein lipase

LRP8 LDL receptor related protein 8

MGL Monoacylglycerol lipase

mTOR Mechanistic target of rapamycin

NBD (22-(N-(7-Nitrobenz-2-Oxa-1,3-Diazol-4-yl) Amino-23,24-Bisnor-5-Cholen-

3β-Ol)

P13K Phosphoinositide 3-kinase

PA Phosphatidic acid

PC Phosphatidylcholine

PCa Prostate cancer

PC Phosphatidylcholine

PE Phosphatidylethanolamine

PFA paraformaldehyde

PG Phosphatidylglycerol

PI Phosphatidylinositol

PLA2G2A Phospholipase 2 group 2 member A

PS Phosphatidylserine

PSA Prostate specific antigen

PTEN Phosphatase and tensin homolog

PUFA Polyunsaturated fatty acid

qFM Quantitative fluorescent microscopy

qRT-PCR Quantitative real-time polymerase chain reaction

RIPA Radioimmunoprecipitation assay buffer

RNA Ribonucleic acid

ROS Reactive oxygen species

RPMI Roswell Park Memorial Institute

SCAP SREBP cleavage activating protein

SCARB1 Scavenger receptor class B member 1

SCD1 Stearoyl-CoA desaturase 1

SDS-PAGE Sodium dodecyl sulfate-polyacrylamide gel electrophoresis

SLC27A1-6 Solute carrier family 27 member 1-6

SM Sphingomyelin

sPLS-DA Sparse partial least squares discriminant analysis

SREBP1/2 Sterol response element binding protein 1/2

SQS Squalene Sythase

TAG Triacylglycerol

TBS Tris-buffered saline

TNT Tunnelling nanotubes


targeted therapies xii

Statement of Original Authorship

The work contained in this thesis has not been previously submitted to meet

requirements for an award at this or any other higher education institution. To the best

of my knowledge and belief, the thesis contains no material previously published or

written by another person except where due reference is made.

Signature:

Date:

QUT Verified Signature


targeted therapies xiii

Awards and publications

Tousignant KD, Talebi A, Rockstroh A, Fard A, Poad B, Gupta R, Gunter

J, Swinnen J, Blanksby S, Nelson C, Sadowski M. Enhanced lipid uptake and lipid

remodeling are adaptive responses to androgen-targeted therapies in prostate cancer.

Awarded first prize for poster presentation delivered at Brisbane Cell and

Developmental Biology Meeting, Brisbane QLD. November 2017.

Egbewande FA, Sadowski MC, Levrier C, Tousignant KD, White JM, Coster MJ,

Nelson CC, Davis RA. Identification of Gibberellic Acid Derivatives That Deregulate

Cholesterol Metabolism in Prostate Cancer Cells. Journal of Natural Products, 2018.

81(4): p.838-845.

Tousignant KD, Rockstroh A, Fard AT, Lehman ML, Wang C, McPherson SJ, Philp

LK, Bartonicek N, Dinger ME, Nelson CC, Sadowski MC. Lipid uptake is an

androgen-enhanced lipid supply pathway associated with prostate cancer disease

progression and bone metastasis. Molecular Cancer Research, Feb 2019.

DOI: 10.1158/1541-7786. MCR-18-1147. Awarded HDR publication award by QUT

Faculty of Health.

Tousignant KD, Rockstroh A, Poad B, Talebi A, Fard AT, Gupta R, Zang T, Lehman

M, Swinnen J, Blanksby, Nelson C and Sadowski CM. Lipid uptake fuels therapy-

induced lipid remodeling in prostate cancer. Awarded first prize for poster

presentation delivered at Australian Cancer Metabolism Meeting, Sydney NSW. May

2019.

Introduction 15

Introduction

1.1 INTRODUCTORY STATEMENT

Current treatments for advanced prostate cancer (PCa) target the androgen

receptor (AR), a transcription factor that controls the expression of a large subset of

genes associated with growth and proliferation of PCa cells. Despite initial disease

regression with androgen/AR-targeted therapies (ATT), almost all patients with

advanced PCa develop recurrent disease and progress to castrate-resistant PCa

(CRPC). This is due to adaptive changes within the tumour cells, which involve AR

re-activation driving disease progression. Enhanced de novo lipogenesis (DNL), the

de novo synthesis of fatty acids and cholesterol, is a hallmark of PCa cells that is

regulated by the AR and is critical for cancer cell survival. However, therapeutic

targeting of DNL has had only limited success in pre-clinical studies due to the

abundance of exogenous lipids in the circulation. Re-activation of DNL is a key

adaptive metabolic response to ATT during the progression to CRPC. However, lipid

uptake is poorly characterised in many cancer types, including PCa. The aims of this

project were to examine ATT-induced adaptive changes that supply cancer cells with

lipids through enhanced uptake from extracellular sources and through direct

intercellular exchange, to identity lipid transporters contributing to ATT adaptation,

to examine how these lipid supply routes interact with DNL, and to explore their

therapeutic potential in preventing ATT resistance and CRPC. This project provides

insights for the development of better co-treatment strategies targeting AR and lipid

metabolism in advanced PCa patients in the hope of delaying progression to CRPC.

Introduction 16

1.2 PROSTATE CANCER

1.2.1 Prostate cancer worldwide incidence and mortality rates

Prostate cancer (PCa) is the second most commonly diagnosed cancer (Bray et

al., 2018) and the third leading cause of cancer mortality in men worldwide (Litwin &

Tan, 2017). Bray et. al (2018) report that in 2018 PCa accounted for nearly 1.3 million

(7.1% of total) new cancer cases and around 360,000 (3.8% of total) cancer-related

deaths. A number of risk factors for PCa have been established including age, race,

genetic polymorphisms and family history. Recent meta-analysis studies have

identified strong global epidemiological trends in which there is higher PCa incidence

in developed countries, such as the United States and Western European countries

(Wong et al., 2016). This could be due to a combination of more advanced diagnostic

methods and lifestyle factors that contribute to a longer life expectancy. Overall, PCa

incidence has increased in the past ten years, while PCa-associated mortality during

this time has decreased in most countries (Wong et al., 2016).

Figure 1.1 PCa incidence

Worldwide comparison of the incidence and mortality rates of PCa as of 2018. Data are

expressed as age-standardised rate (ASR) per 100,000 persons (Bray et al., 2018).

Introduction 17

1.2.2 Diagnosis and grading

The prostate is a small gland that sits just below the bladder and surrounds the

urethra. This androgen-dependent gland plays an important role in the male

reproductive system by producing the majority of fluid that makes up the semen. This

seminal fluid provides nutrients and a protective environment that facilitates the

survival and transport of sperm. PCa occurs when abnormal cells begin to reproduce

uncontrollably, resulting in a malignant tumour (PCF, 2018a).

Although PCa incidence is high, the indolent nature of many tumours translates

to high treatment success rates. Only 1 in 350 men under age 50 will be diagnosed,

however the rate increases dramatically with age, reaching a 1 in 11 risk of diagnosis

in men above the age of 70 (PCF, 2018b). In Australia, there is a 99% 5-year survival

rate for men diagnosed with localised PCa, largely due to advancements in early

diagnostic measures over the past decade. The widely adopted screening marker for

PCa is Prostate Specific Antigen, or PSA, which is an androgen-regulated protein

uniquely produced by both normal and malignant human prostatic epithelial cells

(Elzanaty, Rezanezhad, & Dohle, 2017). In healthy men, serum levels of PSA are

undetectable or found at very low concentrations (PCF, 2018b). Generally, in the

clinic, a serum PSA level of >4.0 ng/mL warrants further evaluation, however men

with a PSA level of less than 10 ng/mL are still considered to be at low risk of PCa

(Litwin & Tan, 2017; Pezaro, Woo, & Davis, 2014a; Wolf et al., 2010). PSA can be

elevated due to non-cancer reasons, so patients with elevated PSA levels will

subsequently undergo a needle core biopsy in order to confirm the presence and

aggressiveness of PCa.

Alongside PSA testing, a patient will usually undergo a digital rectal

examination (DRE) as a method of screening and early detection (Wolf et al., 2010).

Given that PCa is a slow growing and often asymptomatic tumour occurring in older

men, often those with non-aggressive low-risk disease will be subject to “watchful

waiting”. This involves monitoring the tumour status via regular PSA measurements

as opposed to surgery or treatment interventions which can be associated with serious

adverse side effects; often these low-risk men may not experience serious health

effects if their disease is left untreated. Although useful for tumour monitoring post-

diagnosis, there has been controversy over the use of PSA as a diagnostic marker due

to a spike in invasive and unnecessary treatments in men with low-risk disease (Pezaro,

Introduction 18

Woo, & Davis, 2014b). This has driven the investigation of more sensitive and specific

biomarkers to discriminate indolent vs aggressive disease (Pezaro et al., 2014a; Wolet

al., 2010).

Prostate cancer biopsies are graded according to a system known as the Gleason

grading system, which was originally developed in 1966 and remains the standard

method used by clinicians for the diagnosis and management of PCa (Gleason, 1966;

Shah & Zhou, 2016). Prostate tissue biopsies are examined microscopically by a

pathologist and assigned a Gleason grade between 1 and 5 based on the appearance of

the prostate cells and architecture of the prostate glandular structures (Harnden,

Shelley, Coles, Staffurth, & Mason, 2007; Litwin & Tan, 2017). A grade of 1 describes

well differentiated, closely packed cells with uniform shaped glands. Higher grades

represent more abnormal cells, characterised by poor differentiation, less defined

boundaries, and variation in size, shape and separation of the glands. A grade of 5

describes undifferentiated cells with a complete absence of gland formation and

clusters of cells (Harnden et al., 2007) (Fig 1.2). Based on their incidence within the

Figure 1.2 Gleason grading system

A score of 1 describes well differentiated PCa cells with small, uniform glands. Moving

towards 5, PCa cells become poorly differentiated with a complete lack of glands.

Image from Harnden et al. (2007).

Introduction 19

biopsy, the pathologist assigns two grades which represent the cellular features that

make up the two largest areas within the tumour; the sum of these two grades is the

total Gleason score (Litwin & Tan, 2017). A PCa with a Gleason score of less than 6

is considered relatively low risk and slow growing, 7 represents an intermediate risk

PCa, and a score of 8-10 indicates a high risk, fast-growing and aggressive tumour

(Litwin & Tan, 2017; Shah & Zhou, 2016).

1.2.3 Current treatment of localised prostate cancer

Once diagnosed, treatment options are generally guided by PSA levels, which

gives a relative idea of growth rate and aggressiveness of disease. Active surveillance,

also known as “watchful-waiting”, is increasingly being recommended for men

diagnosed with low-risk PCa (Heidenreich et al., 2011), primarily to reduce to risk of

overtreatment and intervention-related adverse side effects. These patients are

followed with recurring PSA tests and are treated if progression is initiated. Localised

tumours that are considered intermediate or high risk are treated with radical

prostatectomy and radiation therapy (Heidenreich et al., 2008; Heidenreich et al.,

2011; Litwin & Tan, 2017), both of which are initially effective therapies in most

patients. Recent technological advances are allowing for the exploration of new

therapies for localised disease with reduced adverse side effects, such as cryotherapy,

high-intensity ultrasound, and laser ablation (Litwin & Tan, 2017). However, there are

limited clinical data to draw conclusions on their efficacy thus far (Litwin & Tan,

2017).

1.2.4 Progression to advanced PCa and CRPC

Despite initial tumour regression and repressed PSA levels following surgery

and/or radiation, an additional 25-40% of PCa patients progress to advanced PCa

within 5 years of initial diagnosis (Kirby, Hirst, & Crawford, 2011) (Fig 1.3). It is well

established that both normal and malignant development, and the function and

maintenance of the prostate gland are highly dependent on androgens, especially 5α-

dihydrotestosterone (DHT), which serves as the ligand for the androgen receptor (AR)

(Heinlein & Chang, 2004; Lonergan & Tindall, 2011). Once activated, the AR

Introduction 20

mediates the transcription process to activate and repress a large set of target genes

that control growth, proliferation, differentiation, and cell survival (Dutt & Gao, 2009;

Lonergan & Tindall, 2011; Soekmadji, Russell, & Nelson, 2013). Thus, the mainstay

treatment for advanced PCa is Androgen Deprivation Therapy (ADT), which blocks

the production of testicular testosterone and starves the tumour of androgens in order

to inhibit the activation of AR. Hormone therapy can occur in the form of surgical

castration (orchiectomy) or chemical castration using gonadotropin-releasing hormone

(GnRH) agonists (Cook & Sheridan, 2000). Hormone therapy, when used in

combination therapy with prostatectomy or radiation therapy, was also found to

increase patient survival and decrease disease recurrence in a meta-analysis of seven

randomized trials (Bria et al., 2009; Heidenreich et al., 2011).

Figure 1.3 Adaptive changes to lipid metabolism are associated with progression

to CRPC

Following surgery or radiation to treat primary PCa, 25-40% of men will progress to

advanced PCa and will undergo androgen-deprivation therapy. Despite an initial drop

in PSA and tumour volume following ADT, these tumours eventually progress to

lethal castrate-resistant prostate cancer. Figure adapted from Professor Colleen

Nelson.

Introduction 21

Unfortunately in most patients with advanced PCa, treatment eventually fails

after 18-24 months on ADT and their disease progresses to the more aggressive

castrate-resistant PCa (CRPC) (Heinlein & Chang, 2004; Kirby et al., 2011). CRPC is

characterised by increased tumour size and rising PSA following chemical or surgical

castration. Metastases are present in over 84% of CRPC patients (mCRPC), many of

which metastasise to the bone, and patient survival is typically around 14 months from

mCRPC diagnosis (Kirby et al., 2011).

It was originally thought that the progression to CRPC was androgen

independent, however, evidence from recent years has demonstrated that the tumour

remains largely androgen sensitive and that tumour cells adopt a number of

mechanisms to survive within an androgen depleted environment (Dutt & Gao, 2009;

Levina et al., 2015; Lonergan & Tindall, 2011). These mechanisms include AR

amplification, relaxed ligand specificity, constitutively active AR splice variants, or

overexpression of AR co-activators and increased synthesis of adrenal and

intratumoural steroids (Dutt & Gao, 2009; Lonergan & Tindall, 2011). This has led to

the development of 2nd and 3rd generation anti-androgen therapies such as

Abiraterone and Enzalutamide. Abiraterone (Zytiga) inhibits CYP17, which is an

enzyme with a critical role in the synthesis of steroid hormones (Ingrosso et al., 2018).

Enzalutamide (Xtandi) is an AR-antagonist used to treat mCRPC and serves as a potent

competitive inhibitor of the AR (Saad, 2013; Tran et al., 2009). Enzalutamide also

prevents AR translocation to the nucleus and binds chromosomal DNA, preventing

transcription of AR regulated genes. While initially effective, these new therapies still

confer only modest survival advantages before tumours once again become treatment

resistant (Lonergan & Tindall, 2011). Consequently, CRPC is currently considered

incurable, making it crucial to identify and target adaptive response pathways activated

by ATTs in order to prevent treatment resistance and progression to CRPC.

1.2.5 Adaptive response pathways driving CRPC development

The mechanisms resulting in castrate resistant PCa are still not well understood.

One assumption is that ADT provides a selective advantage to androgen-independent

cells which continue to proliferate and repopulate the tumour (Heinlein & Chang,

2004). Alternatively, PCa cells activate adaptive response pathways to help to evade

androgen depletion. AR amplification is found in 20-30% of CPRC patients (Heinlein

Introduction 22

& Chang, 2004; Risbridger, Davis, Birrell, & Tilley, 2010), leading to increased AR

transcriptional activity that maintains AR signalling pathways. The CRPC phenotype

is also associated with changes in the tumour environment, endocrine signalling,

cellular plasticity and cellular metabolism, all of which provide alternative cell

survival and growth mechanisms. ATT induces changes in the tumour

microenvironment including increased bone remodelling and the activation of

androgen-repressed genes (Sieh et al., 2012). It is estimated that up to 90% of CRPC

patients develop bone metastasis (Coleman, 2001; Gartrell et al., 2015), making PCa

the most prevalent malignancy to metastasize to bone in men. This may be attributed

in part to the highly lipid and nutrient-rich environment found within the bone

microenvironment (Diedrich, Herroon, Rajagurubandara, & Podgorski, 2018).

Furthermore, ATTs have been shown to induce endocrine alterations in patients

including increased levels of insulin and leptin (Gunter, Sarkar, Lubik, & Nelson,

2013), which are associated with more rapid disease progression, and upregulation of

ghrelin by insulin (Seim et al., 2013), which promotes growth and cellular metabolism.

ATT also alters the plasticity of tumour cells in part by potentiating an epithelial to

mesenchymal transition (EMT) (Nouri et al., 2014; Sun et al., 2012), a process

involved in metastatic spread, thus enhancing tumour progression. Adaptations are

also seen in metabolic pathways, of which de novo lipogenesis is the most well

described (Brusselmans & Swinnen, 2009; Currie, Schulze, Zechner, Walther, &

Farese, 2013; Ettinger et al., 2004; Suburu & Chen, 2012). Collectively, these

physiological adaptations allow PCa cells to adapt to ATT and facilitate cancer cell

survival, metastasis and treatment resistance.

The ultimate objective of our research team is to identify critical adaptive

responses and characterise their potential as therapeutic targets in order to delay the

progression to CRPC. Our group has accumulated strong evidence that adaptive

changes in metabolic pathways, including lipid metabolism, are induced by ATT.

Introduction 23

1.3 LIPID METABOLISM

1.3.1 Lipid classification and function

In its simplest definition, a “lipid” can be defined as any member of a group of organic

molecules that are insoluble in water but soluble in organic solvents (Fahy, Cotter,

Sud, & Subramaniam, 2011), i.e. they share the common property of hydrophobicity.

The study of lipids and their dynamic roles in human physiology has become of

increasing importance in recent decades. Beyond providing essential fatty acids, lipids

serve a critical role in energy generation and storage, as well as intracellular signalling,

protein modification, eicosanoid production (Calder, 2017) and steroid hormone

synthesis (Currie et al., 2013; Swinnen, Brusselmans, & Verhoeven, 2006).

Additionally, fatty acids serve as the main building blocks for cellular membranes,

thus compartmentalising and helping to regulate many functions of the

cell including signalling, nutrient transport, cell division, respiration, and cell

death mechanisms (Butler, Centenera, & Swinnen, 2016; Swinnen et al., 2006). Their

role in maintaining cell membrane fluidity and structure can also impact cell function

(van Meer, Voelker, & Feigenson, 2008).

The emerging field of lipidomics has allowed for a comprehensive analysis and

classification system of lipid molecules, often separated into “simple” and “complex”

groups as determined by the number of distinct entities generated upon hydrolysis.

Simple lipids yield only two products while complex lipids yield three or more

products upon hydrolysis. Lipids have been further divided into eight categories: fatty

acyls, glycerolipids, glycerophospholipids, sphingolipids, sterol lipids, prenol lipids,

saccharolipids and polyketides (Fahy et al., 2011; Fahy et al., 2005) (Fig. 1.4). They

range in structure from simple, short hydrocarbon chains, i.e. fatty acids, to more

complex molecules including triacylglycerides, phospholipids, and sterol molecules.

The complexity within each lipid class can be further increased by the addition of

headgroups, elongation of the acyl chain or the addition of double bonds (Burdge &

Calder, 2015).

Introduction 24

Given their overwhelming structural diversity, a comprehensive Lipid

Classification System has been developed by the LIPID MAPS consortium (Fahy et

al., 2005) to classify lipids into eight lipid categories, each with its own sub class

hierarchy. Briefly, fatty acids are composed of a hydrocarbon chain with a carboxyl

group at one end and a methyl group at the other (Gurr, Frayn, & Harwood, 2002). The

hydrocarbon chain can vary in length ranging from 2-30 carbons, and in the number

of double bonds (unsaturation), both of which are used to identify distinct fatty acids.

Two numbering systems exist: if counting from the carboxyl end (COOH-), then the

C-1, C-2, C-3,… notation is used (Fig 1.5), whereas if counting from the methyl (-

CH3) end, then the methyl carbon serves as ω-1, and counting continues as ω-2, ω-3,

etc. The position of the double bond is then assigned using one of the two notations.

Figure 1.3 Lipid structure representation of major lipid species

Image from (Fahy et al., 2005).

Introduction 25

In complex lipids such as glycerolipids and glycerophospholipids, the

stereospecific numbering (sn) method is used to describe the acylated glycerol group,

typically sn-1 or sn-2 (Burdge & Calder, 2015). While systemic names describe the

structure of distinct lipid species, corresponding common or trivial names and

abbreviations have been assigned to provide a more convenient way to define lipids.

De novo fatty acid synthesis, or the biosynthesis of fatty acids within the cell

from acetyl Co-A and NADPH, produces palmitic acid (C16:0), which can then be

elongated within the cell to generate longer-chain fatty acids by a series of reactions

catalysed by elongases (Guillou, Zadravec, Martin, & Jacobsson, 2010), to be

discussed in further detail. Fatty acid desaturation, which occurs primarily within the

endoplasmic reticulum, inserts one or more double bonds to produce monounsaturated

(one double bond) or polyunsaturated (more than one double bond) fatty acids. A

number of mammalian desaturase enzymes exist, each with distinct specificities.

Of all the fatty acid species present in humans, only two cannot be synthesised

de novo. These fatty acids, namely Linoleic acid (18:2 ω -6) (LA) and Linolenic acid

(18:3 ω -3) (ALA), are now known as essential fatty acids and must be acquired from

Figure 1.4 Example of fatty acid nomenclature counting from carboxyl end

Examples of the nomenclature whereby C-1 is the carbon at the carboxyl end

(COOH-) and double bonds are assigned to the first carbon of the double bond.

Figure from Burdge & Calder, 2015.

Introduction 26

dietary sources. This is because mammals lack the delta12- and delta15- desaturase

enzymes, which insert double bonds at carbon atoms beyond the ninth carbon in the

fatty acid chain (Burdge & Calder, 2015). These essential fatty acids and their

metabolites serve critical roles in mammalian cells. Once taken up from dietary

sources, LA and ALA are converted to longer chain metabolites such as arachidonic

acid (20:4n-6) and eicosapentaenoic acid (20:5n-3), both of which play a major role in

pro- and anti-inflammatory pathways, respectively (Schmitz & Ecker, 2008).

Increasing evidence has revealed that the ratio of dietary ω -6 and ω -3 fatty acids is

critical in regulating the dynamic relationship between pro- and anti-inflammatory

pathways and has huge implications for human health (Chilton et al., 2017; James,

Gibson, & Cleland, 2000; Schmitz & Ecker, 2008).

Fatty acids serve as the building blocks for more complex lipid species. In

mammalian tissues, these are predominantly triacylglycerols (TAGs), phospholipids

(PLs) and cholesterol (Burdge & Calder, 2015). TAGS and PLs are molecules

consisting of a glycerol backbone to which fatty acids are bound via an ester bond.

In PLs, the sn-3 carbon is linked via a phosphoester bond to a phosphate, where a polar

headgroup is attached, giving rise to several PL subclasses. Sphingolipids are similar

in structure, but are characterised by their sphingoid base backbone rather than

glycerol, and are linked to a head group such as ethanolamine, serine or choline, and a

fatty acid linked via an amide bond (Merrill, Sullards, Allegood, Kelly, & Wang,

2005). Within each sphingolipid subclass, the varying combinations of sphingoid base

type, fatty acid sidechain, and headgroup result in a highly diverse number of lipid

species with distinct functions. Generally, sphingolipids are considered to be highly

bioactive compounds.

Cholesterol, a member of the sterol family, is another complex lipid species that

plays a critical role in cell membrane structure and function, as well as sex hormone

synthesis. Cholesterol is characterised by a planar structure consisting of four fused

hydrocarbon rings, with a hydrocarbon tail linked to one end and a hydroxyl group

linked to the other (Gurr et al., 2002). Free cholesterol (Fig 1.6a) is primarily found in

cell membranes or can be linked to a fatty acid to form cholesteryl esters (Fig 1.6b)

and stored in lipid droplets.

Introduction 27

1.3.2 Lipids in membrane structure and function

The cellular membrane is critical not only for its structural function and barrier

formation but also in signal transduction, cell adhesion, nutrient transport, fusion-

fission, endocytosis and protein sorting (Armstrong et al., 2013; Epand, 2015; van

Meer et al., 2008) and the arrangement of membrane lipids has major functional

implications. Membrane lipids are arranged in a bilayer composed primarily of

glycerophospholipids including phosphatidylcholine (PC), phosphatidylethanolamine

(PE), phosphatidylserine (PS), phosphatidylinositol (PI), and phosphatidic acid (PA)

(Gurr et al., 2002; van Meer et al., 2008), with PCs being the most predominant

structural membrane lipid (>50% of total phospholipids). Additional structural lipids

include sphingolipids and sterol lipids which contribute largely to the formation and

signalling function of lipid rafts and membrane fluidity (Armstrong et al., 2013;

Epand, 2015; van Meer et al., 2008; Zhuang, Kim, Adam, Solomon, & Freeman,

2005). The planar bilayer is arranged with hydrophobic fatty acid tails facing each

other and the hydrophilic headgroups on the outside (Fig 1.7), and asymmetry across

the bilayer is maintained by ATP-dependent flippases (Andersen et al., 2016; van Meer

et al., 2008). The variation in headgroups, fatty acyl chain length and degree of

desaturation and location within a membrane allow for vast functional diversity. For

example, PS and PI are found predominantly on the cytoplasmic face of the plasma

Figure 1.5 Structural representation of (a) cholesterol and (b) cholesteryl

esters

Figure from Fahy et al. (2005).

Introduction 28

membrane bilayer, likely due to their ability to function as secondary messengers

(Epand, 2015). Sphingomyelin is a major sphingolipid found in mammalian cell

membranes and, together with its phosphorylated form sphingosine phosphate, serves

as a signalling lipid but is more often described for its role in the formation of

membrane lipid aggregates known as lipid rafts (Epand, 2015; Kinoshita, Suzuki,

Murata, & Matsumori, 2018). Cholesterol is most abundant in the cell membrane and,

in addition to its role in lipid raft formation (Armstrong et al., 2013; Grouleff,

Irudayam, Skeby, & Schiøtt, 2015; van Meer et al., 2008; Zhuang et al., 2005), is

widely acknowledged for its effect on membrane packing and fluidity (Zalba & Ten

Hagen, 2017). Enrichment of cholesterol in the plasma membrane encourages

membrane packing and a more rigid, less permeable membrane (Quinn & Wolf, 2009;

Rubenstein, Smith, & McConnell, 1979; Zalba & Ten Hagen, 2017). Enrichment of

saturated FA has a similar effect, while the distorted hydrophobic chain caused by the

double bonds found in polyunsaturated fatty acids (PUFAs) prevents tight packing and

Figure 1.6 Example of membrane lipid bilayer composition

The liquid-ordered (Lo) phase usually consists of saturated lipids and cholesterol and

is therefore tightly packed and relatively rigid. The liquid-disordered (Ld) phase is

more fluid and loosely packed and occurs at temperatures above the transition

temperature, which is determined by lipid configuration of membrane. Figure from

Zalba & Hagen (2017).

Introduction 29

increases membrane permeability (Subczynski & Wisniewska, 2000; Zalba & Ten

Hagen, 2017). Composition of lipid rafts largely affects signalling pathway activation;

sphingomyelin and cholesterol rich domains promote cell proliferation, whereas

ceramide enrichment promotes apoptosis (Tekpli, Holme, Sergent, & Lagadic-

Gossmann, 2013; Zalba & Ten Hagen, 2017). The effects of membrane lipid

composition and the associated physiological implications have drawn major attention

to membrane lipids and their roles in many human diseases.

1.3.3 Lipid transport

While de novo fatty acid synthesis is predominant during development and in

some specialised processes later in life, most cells obtain enough FA to meet their

energy demand from circulating FA derived from dietary sources (Menendez & Lupu,

2007). In adults, FA synthesis occurs in the lungs for surfactant production, in the

lactating breast to produce FAs for milk lipids, and in steroidogenic tissues including

the prostate (Menendez & Lupu, 2007; Brusselmans & Swinnen, 2009). Additionally,

liver and adipose tissues convert excess carbohydrates to FA, which are then stored as

triglycerides in adipocytes (Brusselmans & Swinnen, 2009). Apart from these tissues,

the expression of lipogenic enzymes remains low in most cells after development.

Fatty acid uptake can occur through three possible routes, with protein-

mediated uptake being the most prevalent (Doege & Stahl, 2006). Circulating fatty

acids from the diet are transported in water-soluble lipoprotein complexes with diverse

compositions (high-density, low-density and very low-density lipoproteins).

Lipoprotein lipase (LPL) and other serum lipases facilitate the hydrolysis

of lipoproteins to release free fatty acids, which are bound to albumin in plasma

(Doege & Stahl, 2006). Fatty acids are then bound to plasma membrane proteins and

uptake is facilitated via fatty acid transporters, including fatty acid transport protein

family (FATP/SLC27), fatty acid translocase (FAT/CD36), and glutamic-oxaloacetic

transaminase 2 (GOT2) (Doege & Stahl, 2006; Sahoo, Aurich, Jonsson, & Thiele,

2014). Fatty acid transporters differ in their tissue expression patterns and substrate

specificity (Doege & Stahl, 2006). For example, FABPpm/GOT2 and SLC27A1-6 are

lipid transporters involved in the plasma membrane-localised transport of medium

chain fatty acids and free long chain fatty acids, respectively (Anderson & Stahl, 2013;

Doege & Stahl, 2006; Go & Mani, 2012; Pinthus et al., 2007; Sahoo et al., 2014).

While protein-mediated transport of fatty acids is widely accepted as the

Introduction 30

primary pathway of cellular free FA acquisition, FA uptake can also occur through

passive diffusion due to their lipophilic nature, however this occurs very minimally

(Doege & Stahl, 2006). In fact, at physiological concentrations, unbound free fatty

acids are found at relatively low levels (7.5 nM) (Richieri & Kleinfeld, 1995).

Once inside the cell, long-chain acyl-CoA synthetase (ACSL) converts fatty

acids into acyl-CoA esters. Acyl-CoA binding proteins (ACBP) then bind to acyl-CoA

esters, unloading the transporters (Doege & Stahl, 2006; Gossett et al., 1996). These

fatty acids are ultimately used for the synthesis of phospholipids to become the

building blocks of membranes, stored in lipid droplets, used for energy production via

β-oxidation, activated as lipid signalling molecules or used for protein modification

(Balaban, Lee, Schreuder, & Hoy, 2015; Brusselmans & Swinnen, 2009; Currie et al.,

2013).

Protein-mediated lipid uptake by receptor-mediated endocytosis of lipid

transporters and their cognate lipoprotein cargo provide cells with various lipid

components including phospholipids, cholesterol esters, triacylglycerol and free fatty

acids (Doege & Stahl, 2006; Sahoo et al., 2014). Lipoproteins are internalised via

lipoprotein receptors such as low- or very-low density lipoprotein receptors (LDLR

and VLDLR) and Scavenger receptor Class B Member 1/2 (SCARB1 and SCARB2)

(Go & Mani, 2012; Schneider, 2016). Various scavenger receptors have also been

shown to be associated with the uptake of modified (acetylated or oxidised) LDL

particles including SCARF1, SCARF2 and CXCL16 (Miller, Choi, Fang, & Tsimikas,

2010; Y. Tamura et al., 2004). Once internalised, the membrane-enclosed organelle is

delivered to lysosomes, where it is disassembled to release its lipid constituents

(Schneider, 2016). Lipoprotein receptors can then be recycled back to the cell surface

to bind and internalise new ligands. LDLR is especially critical in maintaining cellular

cholesterol homeostasis, made evident by the coordinated regulation of LDLR and

cholesterol synthesis enzymes (Schneider, 2016). Sterol-level sensing mechanisms by

sterol-response element-binding proteins (SREBPs) will increase the production of de

novo cholesterol synthesis enzymes when extracellular sources are unavailable or will

suppress production of LDLR or de novo synthesis enzymes when there is an excess

of cholesterol in order to prevent toxic cholesterol overloading (Schneider, 2016;

Lagor & Millar, 2009). This elegant mechanism of nutrient sensing highlights the

importance of maintaining cellular lipid homeostasis.

Introduction 31

More recently, alternative mechanisms of phospholipid and lysophospholipid

transport have been described (Andersen et al., 2016; Lopez-Marques, Theorin,

Palmgren, & Pomorski, 2014). In these models, phospholipids are translocated from

the external to the cytosolic leaflet of cell membranes via P4-ATPases using energy

provided by ATP hydrolysis. This is primarily thought to maintain membrane

asymmetry which is involved in membrane protein sorting, membrane curvature and

fluidity, and cell signalling (Andersen et al., 2016; Lopez-Marques et al., 2014).

Furthermore, lysolipid uptake has been proved to play a critical role in providing

nutrients and activating signalling pathways in cells undergoing nutrient stress or

oncogenic transformation (Kamphorst et al., 2013; Rolin & Maghazachi, 2011). The

abundance and specificity of lipid transporters together with strict regulatory

mechanisms observed in controlling lipid homeostasis in mammalian cells illustrate

the diverse and pivotal role that lipids play in many physiological and biochemical

processes.

Introduction 32

Figure 1.7 Mechanisms of cellular lipid acquisition

Circulating lipoproteins are hydrolised by lipoprotein lipases (LPL) to release free

fatty acids (FA), which then enter the cell via fatty acid transport proteins (FATP)/fatty

acid translocase (FAT) and are delivered to fatty acid binding proteins (FABP). Fatty

acids can also be synthesised endogenously by converting glucose to citrate via the

Citric Acid Cycle, which is then converted to acetyl-CoA by ATP Citrate Lyase.

Following conversion to Malonyl-CoA by Acetyl-CoA Carboxylase-α (ACACA),

Fatty Acid Synthase (FASN) forms fatty acids from acetyl-CoA and malonyl-CoA.

Alternatively, HMG-CoA synthase (HMGCS) and HMG-CoA reductase (HMGCR)

use acetyl-CoA for the synthesis of cholesterol.

Introduction 33

1.3.4 De novo lipogenesis

Lipid uptake is the primary mode of lipid acquisition in most healthy adult

cells, as described above. However, in many disease states including cancer, de

novo lipogenesis (DNL) is found to be highly upregulated (Swinnen et al., 2002a;

Swinnen et al., 2000; Swinnen et al., 2006). De novo lipogenesis uses a group of

lipogenic enzymes to synthesise fatty acids from circulating glucose or other carbon

sources (Hosios et al., 2016). Glucose is first converted to pyruvate via the Glycolytic

pathway. Pyruvate then enters the mitochondria and is metabolized to citrate via the

Kreb’s Cycle, which is then transported to the cytoplasm and converted to acetyl-CoA

by the enzyme ATP Citrate Lyase (ACLY). Acetyl-CoA serves as a precursor for both

fatty acids and cholesterol and is converted to malonyl-CoA by Acetyl-CoA

Carboxylase-α (ACACA). Finally, Fatty Acid Synthase (FASN) forms saturated long-

chain fatty acids from acetyl-CoA and malonyl-CoA via a series of successive

condensation reactions. These fatty acids have similar fates as those acquired

exogenously (Balaban et al., 2015; Brusselmans & Swinnen, 2009; Currie et al., 2013).

Alternatively, cholesterol can also be synthesised by the cell via conversion

of acetyl-CoA to mevalonate by HMG-CoA synthase (HMGCS) and HMG-CoA

reductase (HMGCR) (Brusselmans & Swinnen, 2009). Mevalonate is used to form

farnesyl diphosphate, which is then modified by Squalene Sythase (SQS) to form

cholesterol (Brusselmans & Swinnen, 2009). In addition to its role in membrane

composition, cholesterol also serves as a precursor of intratumoural steroidogenesis.

Both the exogenous uptake of lipids and de novo synthesis of FA and cholesterol are

critical in maintaining metabolic homeostasis.

1.3.5 Lipid droplet function and biogenesis

Recent advances in lipid droplet (LD) biology have uncovered their versatile and

pivotal functions in maintaining cellular homeostasis. The adaptability of LDs under

stressful environmental conditions allows them to play a major role in nutrient

homeostasis, lipotoxicity and oxidative stress. This is exemplified by the accumulation

of lipid droplets in cancer cells exposed to hypoxia or nutrient depletion (Cabodevilla

et al., 2013; Koizume & Miyagi, 2016; Petan, Jarc, & Jusović, 2018). These cytosolic

organelles are composed of a neutral lipid core of primarily triacylglycerides and sterol

esters (Petan et al., 2018; Wilfling, Haas, Walther, & Jr, 2014), and more recently

discovered acylceramides (Senkal et al., 2017), surrounded by a phospholipid

Introduction 34

monolayer including integral membrane proteins (Wilfling et al., 2014). The de novo

synthesis of LDs occurs in the ER and is initiated by the generation of neutral lipids

via esterification of fatty acyl substrates by Diacylglycerol O-Acyltransferases

(DGAT1 and DGAT2), two structurally unrelated proteins with distinct substrate

specificities and localisation (Yen, Stone, Koliwad, Harris, & Farese, 2008). While the

predominant function of DGAT1, which is found exclusively in the ER, is thought to

be TAG synthesis, it has been shown to have acyltransferase activities for a variety of

substrates including diacylglycerols, wax esters and retinyl esters (Yen, Monetti,

Burri, & Farese, 2005). Conversely, DGAT2, localised in the ER and LDs, was

previously shown to be unable to perform the additional acyltransferase activities

described above and has been shown to be the primary enzyme involved in the bulk of

TAG synthesis (Cases et al., 2001; Stone et al., 2004; Yen et al., 2008). Only recently

have additional acyltransferase activities by DGAT2 been described, i.e. acylceramide

synthesis (Senkal et al., 2017). Triglycerides and other neutral lipids not only act as a

major cellular energy reservoir for lipid storage, but have more recently been

discovered to serve a protective role in preventing lipotoxicity and subsequent

activation of ER stress pathways (Chitraju et al., 2017; Listenberger et al., 2003).

Following neutral lipid generation and accumulation within the ER bilayer, a lens of

neutral lipids is formed, followed by budding of the LD into the cytosol (Wilfling et

al., 2014).

Lipid droplets serve several stress-induced functions including protection

against lipotoxicity (described above), energy homeostasis, lipid mediator production,

regulation of autophagy, ER & membrane homeostasis, and serving as a source of fatty

acids for ß-oxidation (Petan et al., 2018). By sequestering toxic lipids including fatty

acids, cholesterol and ceramides, LDs help to prevent lipotoxic cell damage and their

complex relationship with autophagy and lipolysis is critical in maintaining cellular

energy levels. Evidence accumulating over the past decade has drawn a clear link

between LDs and many prevalent human diseases, including cancer (Petan et al.,

2018), however their functional significance in different cancer types requires further

investigation.

Introduction 35

1.3.6 Altered lipid metabolism is a hallmark of cancer

It has long been established that cancer cells convert most glucose to lactate

regardless of the oxygen supply available, a phenomenon known as “the Warburg

effect” (Warburg, Wind, & Negelein, 1927), which is now widely accepted as a

hallmark of cancer. This metabolic preference is also seen in normal tissues

undergoing proliferation, suggesting that cancer cells adopt metabolic pathways

conducive to proliferation rather than quiescence or differentiation. Oncogenic

mutations allow the shift towards scavenging of nutrients such as lipids, amino acids,

and nucleotides to create biomass and promote proliferation, rather than to sustain

efficient energy production (Finicle, Jayashankar, & Edinger, 2018; Vander Heiden,

Cantley, & Thompson, 2009). Multicellular organisms have evolved metabolic control

systems that utilize different cellular metabolism pathways in proliferating vs

nonproliferating cells. This is meant to prevent uncontrolled proliferation and to

increase energy production. Mammalian cells normally require growth factor signals

Figure 1.8 DGAT activity and lipid droplet biogenesis

Diacylglycerol O-Acyltransferase 1 and 2 (DGAT1/2) catalyse the conversion of

diacylglycerol and fatty acyl CoA to triacylglycerol (TAG), which is then stored in

lipid droplets. Figure from Yen et al. (2008).

Introduction 36

to stimulate the uptake of nutrients from their environment. Many cancer cells have

obtained oncogenic mutations to alter these receptor-mediated signalling pathways and

to avoid controlled proliferation systems. By modifying growth factors at the plasma

membrane, lipids help to regulate the generation of bioactive molecules and

extracellular vesicles, which in turn increases the intercellular communication between

cancer and healthy cells (Menendez & Lupu, 2007). This oncogenic adaptation

supports the increased requirement of malignant cells for nucleotides, amino acids, and

lipids that serve as building blocks for new cells.

While the Warburg effect is characteristic of many cancer types, advances in

metabolic research has allowed for a much more comprehensive analysis of altered

metabolic pathways in cancer. Interestingly, a study conducted using over 9,000

primary and metastatic tumour samples found that “Warburg effect genes” had a

similar mutation frequency in both primary and metastatic tumours, whiles genes

involved in FA oxidation, lipogenesis and cellular FA uptake genes were found to have

a higher mutation frequency in metastatic tumours (Aritro Nath & Chan, 2016a). This

accumulation of lipid metabolic genes in metastatic tumours suggests that, in addition

to increased proliferation as described by the Warburg effect, enhanced FA uptake

may also play a role in epithelial-mesenchymal transition (EMT) and tumour

metastasis. This presents a novel therapeutic approach for targeting cancer progression

and metastasis.

Fatty Acid Synthase (FASN) is a major lipogenic enzyme involved in the

production of long-chain FA and is found to be overexpressed in many

cancers (Beloribi-Djefaflia, Vasseur, & Guillaumond, 2016; Currie et al., 2013;

Kuhajda, 2000; Menendez & Lupu, 2007), including PCa (Brusselmans & Swinnen,

2009; Flavin, Zadra, & Loda, 2011; Fritz et al., 2010; Swinnen et al., 2006). FASN

expression within cancer cells facilitates the synthesis of phospholipids which

contribute to lipid-raft formation, thus influencing signal transduction, growth factor

signalling, intercellular trafficking, cell polarisation, and cell migration (Swinnen et

al., 2003). Additionally, the lipogenic enzyme ACACA is upregulated in PCa

(Swinnen et al., 2000). Targeting FASN and the fatty acid synthesis pathway is

considered a promising cancer treatment as cancer cells more heavily rely on FASN-

mediated de novo synthesis, while the uptake of circulating exogenous lipids is

sufficient for the requirements of most normal cells (Currie et al., 2013; Daniëls et al.,

2014).

Introduction 37

When cultured in lipid-reduced growth conditions, which results in attenuated

proliferation rates, cancer cells activate de novo lipid synthesis pathways. This results

in increased expression of lipogenic enzymes such as ACLY, Acyl-CoA Synthetase

Short-Chain Family Member 2 (ACSS2), FASN, and HMGCR (Daniëls et al., 2014).

The observation that cancer cells differentially activate and thrive on de novo lipid

synthesis pathways in a low-lipid environment suggests that there is functional cross

talk between these two pathways in order to meet the high lipid demand of proliferating

cancer cells (Daniëls et al., 2014).

Lipogenesis vs exogenous uptake: functional crosstalk between pathways

While the increase in de novo lipid synthesis in cancer cells is well established,

the relationship between de novo and exogenous FA uptake remains unclear. Because

increased FASN gene copy number, transcriptional activation or protein expression are

common characteristics observed in PCa (Swinnen et al., 2000), fatty acid and

cholesterol synthesis have been considered an attractive therapeutic target. However

the antineoplastic effects observed by inhibiting lipogenesis can be rescued by the

uptake of exogenous lipids (Griffiths et al., 2013; Kuemmerle et al., 2011),

highlighting that lipid uptake is a mechanism of clinical resistance to lipogenesis

inhibitors and that cellular capacity for lipid uptake is sufficient to substitute for the

blockade of lipogenesis. Indeed, it was recently reported that lung cancer cells

expressing a strong lipogenic phenotype generated up to 70% of their cellular lipid

carbon biomass from exogenous fatty acids and only 30% from de novo synthesis

supplied by glucose and glutamine as carbon sources (Hosios et al., 2016). The notion

that extracellular fatty acids are the predominant carbon source for lipid synthesis,

rather than glutamine or glucose, has since been validated in breast cancer cell lines

(Balaban et al., 2017).

While altered cellular lipid metabolism is a hallmark of the malignant

phenotype, PCa is in fact unique in that it is characterised by a relatively low glucose

uptake and glycolytic rate, compared to many solid tumours conforming to the

“Warburg effect” phenotype (Effert et al., 1996; Zadra, Photopoulos, & Loda, 2013).

Concordantly, PCa cells showed a dominant uptake of fatty acids over glucose, with

the uptake of palmitic acid measured to be ~20 times higher than that of glucose in

both malignant and benign PCa cells (Liu, Zuckier, & Ghesani, 2010b). Together,

Introduction 38

these data suggest that the increase in de novo lipid synthesis characteristic of PCa is

not solely responsible for changes in cellular fatty acid content within PCa cells.

Another recent study showed a marked difference in the reliance on DNL

contribution of palmitate between different cell lines of the same cancer type, where it

was found that the percentage of intracellular palmitate coming from exogenous

sources varied from 34% to as high as 78% across four colon cancer cell lines (Foletta

et al., 2016). This study demonstrates considerable heterogeneity in the contribution

of DNL and lipid uptake in different cancer types, yet exogenous uptake is a significant

and previously underappreciated supply route in cancer cells exhibiting a lipogenic

phenotype. Given that lipid homeostasis is critical for cell survival, there are many

finely-tuned sensing and regulatory mechanisms to maintain cellular lipid homeostasis

(Agmon & Stockwell, 2017), however there is currently little to no understanding of

the sensing mechanisms that regulate the contribution of DNL versus exogenous lipid

uptake in healthy or disease states. Further exploration into the ratio of DNL to

exogenous lipid uptake and associated regulatory mechanisms is required in order to

understand the relationships between these pathways and how to best manipulate these

pathways for successful therapeutic intervention.

1.3.7 Altered lipid metabolism helps to drive resistance to anti-cancer

therapies

A major challenge in cancer therapeutics stems not from the lack of initial

treatment options, but in the acquired resistance by cancer cells that ultimately results

in relapse and disease progression. However, the underlying molecular mechanisms of

drug resistance are still not well understood. While metabolic reprogramming is a well-

described hallmark of cancer, little is known about therapy induced metabolic

alterations that help to facilitate cancer cell survival and drive disease progression

(Corsetto et al., 2017).

Recent studies have shown that anti-cancer treatments activate metabolic

networks that contribute to drug resistance in renal cell carcinoma (Lue et al., 2017)

and breast cancer (Hangauer et al., 2017; Vijayaraghavalu, Peetla, Lu, & Labhasetwar,

2012) models. In PCa, by removing growth and proliferative signals via AR-

antagonists, cells enter a quiescent state of negligible growth and have increased

expression of several dedifferentiation markers (Hangauer et al., 2017). This could

contribute to evasion of selective drug pressure targeting fast-growing cell

Introduction 39

populations. Despite reduced proliferation, pathways such as phospholipid

metabolism, LD formation and mitochondrial respiration (Lue et al., 2017;

Vijayaraghavalu et al., 2012) have been shown to be increased in cancer cells as an

adaptive response to therapy. Increased LD formation could help to prevent

lipotoxicity, ROS damage and subsequent ER stress (discussed in section 1.3.5) as

well as serve as an energy reserve for cells undergoing nutrient stress. Altered

membrane lipid composition could also have protective benefits, for example, by

increasing saturated FA content in order to decrease membrane permeability and drug

uptake (discussed in section 1.3.2). There is very limited knowledge surrounding

therapy-induced metabolic reprogramming in cancer, however identifying and

targeting these metabolic vulnerabilities could serve a pivotal role in overcoming

therapeutic resistance.

1.3.8 Tunnelling Nanotubes as a mechanism of lipid acquisition

While protein-mediated uptake of FA and cholesterol is well described as the

primary mechanism of exogenous uptake, less is known about direct cell-cell

exchange of lipids as a rescue mechanism. A relatively unexplored area of cancer

biology is the role of tunnelling nanotubes (TNT) in cancer cell communication. These

thin, actin-based extensions allow for the exchange of cellular cytoplasmic material

including proteins, mitochondria, LDs, viral particles and miRNA (Gerdes,

Bukoreshtliev, & Barroso, 2007; Rustom, Saffrich, Markovic, Walther, & Gerdes,

2004). Furthermore, preliminary data from our laboratory shows the transfer of lipid

droplets via TNTs in PCa cells, especially those exposed to the stress of DNL

inhibition (personal communication, Dr. Sadowski). While there are few quantitative

data relating to TNTs, they are characterised as being 50-200 nm in diameter with

lengths that can reach up to several cell diameters and connect cells up to 100-200 µm

apart (Desir et al., 2016; Gerdes et al., 2007; Rustom et al., 2004).

TNTs are upregulated under conditions of metabolic and environmental stress,

such as ATT. Chemoresistant ovarian cancer cells show significantly increased TNT

formation after being placed in hypoxic conditions when compared

to chemosensitive cells, suggesting that TNT formation might be advantageous

for chemoresistance (Desir et al., 2016). Interestingly, chemosensitive cells produce

significantly more TNTs when co-cultured with chemoresistant cells, resulting in TNT

formation between the two cell populations (Desir et al., 2016). While it seems that

Introduction 40

transfer of cellular components via TNTs could provide cancer cells with alternative

mechanisms of survival, little is known about TNTs in the context of lipid uptake and

whether they contribute to the increased levels of intracellular lipid accumulation that

are characteristic of advanced PCa cells.

1.3.9 Alternative mechanisms of lipid scavenging

Selective uptake of lipids and other nutrient sources via receptor-mediated

activity is well-described, however increasing attention is being placed on alternative

scavenging mechanisms as routes of nutrient acquisition. For example, albumin, the

most abundant plasma protein (Merlot, Kalinowski, & Richardson, 2014), can be

bound to several fatty acids, and receptor-mediated albumin uptake releases those

exogenous fatty acids intracellularly (Finicle et al., 2018). Macropinocytosis is a non-

selective form of nutrient uptake, and has also gained recent attention given that it

serves as a source of a range of macromolecules from the extracellular matrix,

including lipids (Finicle et al., 2018). This form of endocytosis acquires

macromolecules through the generation of large endocytic vesicles called

macropinosomes, which have been shown to take up exosomes and necrotic cell debris

(Kerr & Teasdale, 2009). Macropinosomes are then trafficked towards lysosomes for

degradation, and components are released to the cell for use in anabolic biomass

production. It has recently been shown that macropinocytosis contributes 15-25% of

amino acids for protein synthesis in prostate cancer cells grown in complete growth

medium, and this increases to 35-71% of amino acids under nutrient stress conditions

(Kim et al., 2018). Thus, macropinocytosis serves as a substantial source of nutrients

that is upregulated in times of metabolic stress. These additional mechanisms of lipid

scavenging add to the complexity of understanding lipid metabolic pathways in both

healthy and disease states.

Introduction 41

1.4 LIPID METABOLISM IN PROSTATE CANCER

1.4.1 Aberrant lipid metabolism in prostate cancer

Obesity, a disorder of increased body fat mass featuring increased circulating

lipid content (Fruh, 2017), has been associated with changes in progression of several

cancer types, leading to higher-grade disease and poorer patient outcomes (Balaban et

al., 2015). In PCa specifically, there is a strong link between obesity and more

aggressive disease after diagnosis, as well as reduced time to recurrence (Balaban et

al., 2015; Butler et al., 2016). Enhanced lipogenesis, regardless of nutritional lipid

supply, is now acknowledged as a metabolic hallmark of cancer and is an early

metabolic switch observed in the development of PCa. It is maintained throughout the

progression of PCa and associated with poor prognosis and aggressiveness of disease

(Deep & Schlaepfer, 2016; Flavin et al., 2011; Fritz et al., 2010; Menendez & Lupu,

2007; Swinnen et al., 2006). Increased de novo lipogenesis enhances membrane

phospholipid saturation and this may help to protect cancer cells from

chemotherapeutics and oxidative stress-induced cell death (Rysman et al., 2010).

Because mammalian cells lack the delta-12 desaturase to generate polyunsaturated

acyl chains (described in section 1.3.1), de novo synthesised lipids are primarily made

of saturated fatty acid chains. These fatty acid chains pack densely together in cell

membranes, making them more impermeable and protecting against

chemotherapeutics, free radical damage and cell death (Butler et al., 2016; Rysman et

al., 2010). Interestingly, dietary PUFAs have recently been explored for their potential

as an adjuvant cancer therapy and have been shown to have promising

chemosensitising effects in several cancer types (reviewed in (Corsetto, Colombo,

Kopecka, Rizzo, & Riganti, 2017)). The rationale behind the lipogenic switch remains

unclear, and the hypothesis that PCa cells increase DNL in order to attain the protective

benefits of a plasma membrane enriched in saturated FA content requires further

investigation.

Overexpression of the lipid remodelling enzymes stearoyl-CoA desaturase 1

(SCD1) (Fritz et al., 2010; Peck et al., 2016) and fatty acid elongase 7 (ELOVL7)

(Tamura et al., 2009) have also been found in PCa cells. SCD1 inhibition results in

decreased lipid synthesis and proliferation of androgen-sensitive and -resistant PCa

cells and decreased growth of PCa xenografts in mice (Fritz et al., 2010). Similarly,

ELOVL7 knockdown attenuated PCa cell growth in vitro, while overexpression of

Introduction 42

ELOVL7 in mice fed a high-fat diet significantly promoted tumour growth, compared

to ELOVL7-mock treated mice fed a high fat diet, which exhibited no tumour growth

effect (Tamura et al., 2009). In addition to increased de novo lipogenesis and lipid

remodelling, increased lipolysis via overexpression of the lipolytic enzyme

monoacylglycerol lipase (MGL) is also found in several aggressive cancer types

including PCa (Nomura et al., 2010).

1.4.2 Androgen regulation of lipid metabolism

In PCa, androgens stimulate expression of FASN via activation of sterol

regulatory element-binding proteins (SREBPs) (Butler et al., 2016; Heemers et al.,

2004). The SREBP chaperone (SCAP) escorts SREBP precursors from the

endoplasmic reticulum to Golgi bodies, where they are cleaved into their active forms,

allowing mature SREBP binding to sterol response elements within the promoter of

target genes (Heemers et al., 2004). Androgen-induced expression of SCAP results in

significantly increased activation of lipogenic enzymes (Brusselmans & Swinnen,

2009; Heemers et al., 2004).

Figure 1.10 Androgen regulation of lipid metabolism

Androgens stimulate the androgen receptor (AR), resulting in activation of the sterol

regulatory element-binding proteins (SREBPs). SREBPs then activate both de novo

lipogenesis and fatty acid uptake through membrane receptors. Thesis fatty acids are used

as cellular membrane components, stored in lipid droplets, or oxidized for energy. Figure

from Butler et al. (2016).

Introduction 43

Because SREBPs regulate the expression of many lipogenic and lipid uptake

enzymes, they are considered a master regulator of lipid homeostasis in cells (Butler

et al., 2016). Our laboratory has shown in a PCa LNCaP xenograft tumour model that

the expression of SREBP-1 and SREBP-2, along with activating protein SCAP, are

upregulated in PCa (Ettinger et al., 2004). Furthermore, expression of these transcripts

decreases initially after androgen withdrawal by surgical castration, but increases

significantly throughout the development to CRPC to levels far greater than pre-

castrate levels (Ettinger et al., 2004). These data suggest that dysregulated lipid

metabolism may be an adaptive response throughout the development to CRPC. In

fact, a relatively new non-toxic small molecule Silibinin has proven to inhibit neutral

lipid, cholesterol and citrate levels in both LNCaP and DU145 PCa cells by inhibiting

the proteolytic activation of SREBP, thus blocking the transcriptional activation of

several target genes involved in lipid biosynthesis (Nambiar, Deep, Singh, Agarwal,

& Agarwal, 2014). This presents a promising new therapeutic avenue for targeting

cancer cells that are dependent on de novo lipid synthesis.

In addition to increased de novo lipogenesis, enhanced uptake of exogenous

lipids can also drive the proliferation of cancer cells (Aritro Nath & Chan, 2016a;

Nieman, Romero, Van Houten, & Lengyel, 2013). It has been shown that AR

signalling may help to regulate cellular uptake of exogenous lipids in PCa cells, and

this can lead to increased proliferation (Butler et al., 2016). Androgens play an

important systemic role in fat distribution in humans by stimulating lipolysis of fatty

acids from adipocytes (O’Reilly, House, & Tomlinson, 2014) and inducing expression

of cell surface proteins that help to regulate exogenous lipid uptake (Butler et al.,

2016). Androgen-sensitive AR-positive LNCaP cells treated with androgens have been

shown to exhibit increased accumulation of LDs within the cytoplasm, as well as

increased synthesis of cholesterol and fatty acids, and this response was absent in AR-

negative PCa cells (Butler et al., 2016; Swinnen, Van Veldhoven, Esquenet, Heyns, &

Verhoeven, 1996b). Furthermore, the lipogenic enzymes ACACA and ACLY and

cholesterol synthesis enzymes are known to be induced by androgens (Fritz et al.,

2010; Swinnen et al., 1996b). Cistromic data of AR binding sites in PCa cells and

prostate tumours show that AR binding sites occur within several genes involved in

cellular and lipid metabolism (Barfeld, Itkonen, Urbanucci, & Mills, 2014), thus AR

may directly regulate their expression.

Introduction 44

While there is a clear increase in anabolic metabolism in PCa cells, the

advantage of this adaptation remains unclear. Targeting de novo synthesis pathways

remains an attractive therapeutic target, but this alone has not shown promising clinical

results as the tumour cells deficit in synthesised lipids can be rescued by the uptake of

lipids from exogenous sources. Thus, for better efficacy, inhibiting de novo synthesis

should be co-targeted by limiting the availability of exogenous lipids to cancer cells.

The complex link and reliance between androgen signalling, lipid metabolism and

cancer cell physiology make this an interesting area of exploration for uncovering

potential new therapeutic targets or strategies.

1.4.3 Metabolic adaptations to AR-regulated pathways in ATT

The prostate and PCa cells are highly lipogenic. ATT initially suppresses the

lipogenic pathways associated with PCa with reduced expression of SREBP, FASN,

and HMGCR. However, these key mediators of lipogenesis are reactivated in CRPC

tumours (Ettinger et al., 2004; Locke et al., 2010). The reactivation of lipogenic

pathways may be due to the upregulation of systemic regulators of metabolism, as

ongoing work in our laboratory shows an increase in insulin, ghrelin and leptin levels

in response to ATT (Gunter, Lubik, McKenzie, Pollak, & Nelson, 2012; Locke et al.,

2010; Seim et al., 2013).

Recent evidence shows that cholesterol uptake and synthesis play a major role

in PCa progression. LNCaP cells were found to have a significant increase in

cholesterol ester (CE) levels during progression to androgen-independence (Yue et al.,

2014). This accumulation was found to be a result of loss of Phosphatase and tensin

homolog (PTEN) and activation of the P13K/AKT/mTOR/SREBP pathway.

Inhibiting de novo cholesterol synthesis with simvastatin did not affect CE

accumulation, suggesting that cholesterol uptake was the main source of accumulating

cholesterol esters (Yue et al., 2014). Exogenous cholesterol is primarily taken up by

cells via the LDLR and SCARB1, both of which are found to be highly expressed in

metastatic PCa (Schörghofer et al., 2015a; Thysell et al., 2010), and increased

cholesterol levels have previously been associated with PCa (Thysell et al., 2010).

Cholesterol serves as an upstream precursor in the steroidogenic pathway for de novo

androgen synthesis within tumour cells (Twiddy, Cox, & Wasan, 2012; Twiddy, Leon,

& Wasan, 2011). Interfering with cholesterol metabolism pathways is therefore a

promising new avenue for targeting the progression to CRPC.

Introduction 45

Previous data from a longitudinal xenograft study developed in our laboratory

show alterations in lipid metabolism during the development to CRPC (Locke et al.,

2010), including increased essential fatty acid content (Fig 1.11), suggesting that

exogenous fatty acid uptake may serve as a significant and previously

underappreciated supply route for lipids that helps facilitate progression to CRPC.

However, lipid uptake and the lipid transporter landscape in PCa are poorly

characterised. This has led us to investigate the role of exogenous lipid uptake in the

progression of PCa.

Figure 1.11 LNCaP longitudinal xenograft study shows increased FA content

in tumours throughout progression to CRPC

Tumours were collected from mice bearing androgen-dependent (AD) or Nadir (N)

tumours and castrate-resistant PCa (CRPC) tumours and analysed for fatty acid

content. Figure from Locke et al. (2010).

Introduction 46

1.5 THESIS OUTLINE

1.5.1 Rationale

PCa remains the second most commonly diagnosed cancer among men and is

the third leading cause of cancer related mortalities in men worldwide (Litwin & Tan,

2017). Despite initial disease regression following AR-targeted therapies, almost all

PCa patients develop recurrent disease and progress to incurable and lethal castrate-

resistant PCa (CRPC) (Heinlein & Chang, 2004; Kirby et al., 2011) due to the

activation of adaptive response pathways that lead to treatment resistance. Therefore,

it is imperative to identify and target these adaptive response pathways in order to fight

disease progression.

Increased de novo lipogenesis (DNL) is a well-characterised AR-regulated

pathway and metabolic hallmark of PCa, however therapeutic targeting of DNL has

had only limited clinical success. This can be explained by the observation that

inhibiting lipogenesis can be rescued by the addition of exogenous lipids (Griffiths et

al., 2013; Kuemmerle et al., 2011), highlighting lipid uptake as a mechanism of clinical

resistance to lipogenesis inhibitors and that lipid uptake capacity is sufficient to

substitute for the loss of lipogenesis. It is becoming evident that enhanced lipogenesis

in PCa development and progression is not the sole deregulated lipid supply pathway,

and that lipid uptake might play an important role in biochemical recurrence of prostate

cancer. Yet, the contribution and identity of lipid uptake pathways as a supply route of

exogenous lipids and their roles in disease development and progression remain

largely unknown.

Little is known about therapy induced metabolic alterations that help to

facilitate cancer cell survival and drive disease progression, however the activation of

metabolic networks beyond DNL has recently emerged as a mechanism of drug

resistance in several cancer types (Hangauer et al., 2017; Lue et al., 2017;

Vijayaraghavalu et al., 2012). This work will investigate the

metabolic rewiring induced by current anti-cancer treatments in prostate cancer.

Furthermore, it will explore novel therapies to use in combination with current anti-

cancer therapies to fight the emergence of drug resistance.

Introduction 47

1.5.2 Hypothesis

ATTs induce rewiring of lipid metabolic networks in PCa cells that help to facilitate

cell survival and the progression to CRPC.

1.5.3 Aims

Aim 1: To investigate the androgen regulation of lipid uptake and to

characterise the lipid transporter landscape in PCa (Chapter 3).

Aim 2:

Aim 2.1: To integrate multiple ‘omics platforms to investigate changes

in metabolic pathways induced by ATTs in a long-term in vitro ATT

model (Chapter 4).

Aim 2.2: To functionally assess various lipid supply routes in PCa cells

undergoing ATT (Chapter 4).

Aim 3: To characterise the role of key mediators of lipid remodelling and their

therapeutic potential in PCa (Chapter 5).

Materials and methods 48

Materials and Methods

2.1 CELL CULTURE

The following cell lines were acquired from the American Type Culture

Collection (ATCC): LNCaP (CVCL_0395), C4-2B (CVCL_4784), VCaP

(CVCL_2235), PC-3 (CVCL_0035), LAPC-4 (CVCL_4744), BPH1 (CVCL_1091),

and RWPE (CVCL_3791). Fibroblast-free DuCaP cells were a generous gift from M.

Ness (VTT Technical Research Centre of Finland). LNCaP, C4-2B, PC-3, BPH-1, and

RWPE cells were cultured in Roswell Park Memorial Institute (RPMI) medium

(Thermo Fisher Scientific) supplemented with 5% Fetal Bovine Serum (FBS). DuCaP

and VCaP cells were cultured in RPMI supplemented with 10% FBS. LAPC4 cells

were cultured in Dulbecco’s Modified Eagle Medium (DMEM) (Thermo Fisher

Scientific) supplemented with 5% FBS. Media were changed every 3-4 days. All cell

lines were incubated at 37°C in 5% CO2. Cells were passaged at approximately 80%

confluency by trypsinization. Cell lines were genotyped in March 2018 by Genomics

Research Centre (Brisbane) and routinely tested for mycoplasma infection.

For androgen and anti-androgen treatments, cells were seeded in regular growth

medium for 72 hours before media were replaced with RPMI supplemented with 5%

CSS (Sigma-Aldrich). After 48 hours, media were replaced with fresh 5% CSS RPMI

and cells were treated with dihydroxytestosterone (DHT, 10 nM), synthetic androgen

R1881 (1 nM), or vehicle control (0.1% DMSO unless otherwise indicated) for 48

hours to activate AR signalling. The AR-antagonist Enzalutamide or Bicalutamide

(Selleck Chemicals, Houston, TX, USA) was used at 10 µM. For long-term ATT

studies, cells were treated with Enzalutamide (10 µM) or 0.1 % DMSO for up to 21

days. Medium was changed every 3-4 days.

2.2 RNA EXTRACTION AND QUANTITATIVE REAL-TIME

POLYMERASE CHAIN REACTION (PCR)

Cells were seeded at a density of 9.0 x 104 (LNCaP and C4-2B) or 1.2x 105

(DuCaP and VCaP) cells/well in 6 well plates (ThermoScientific). Following


completion of treatment, total RNA was isolated using the RNEasy mini kit (Qiagen)

following the manufacturer’s instructions. Before elution, RNA was treated with

DNase I (Qiagen) to improve RNA purity. RNA concentration was measured using a

NanoDrop ND-1000 Spectrophotometer (ThermoScientific) and frozen at -80°C until

further use.

Up to 2 µg of total RNA was used to prepare cDNA with SensiFast cDNA

synthesis kit (Bioline) according to the manufacturer’s instructions and diluted 1:5. To

each well, 4 µl 2X SYBR-Green Master Mix (Invitrogen), 2 µl of each forward and

reverse primer (final concentration of 0.2 µM) and 2 µl cDNA were added for a total

volume of 8 µl per well in 384 well optical reaction plates. qRT-PCR was performed

with SYBR-Green Master Mix (Thermo Fisher Scientific) using the ViiA-7 Real-Time

PCR system (Applied Biosystems). Determination of relative mRNA levels was

calculated using the comparative Ct method (Schmittgen & Livak, 2008), where

expression levels were normalised relative to that of the housekeeping gene receptor-

like protein 32 (RPL32) for each treatment and calculated as fold change relative to

the vehicle control treatment. All experiments were performed in technical duplicate

and biological triplicate. RT-PCR primer sequences are listed in Table 2.1.


Table 2.1 Table of forward and reverse primer sequences used for qRT-PCR

Forward 5’ 3’ Reverse 5’ 3’

AR CTGGACACGACAACAACCAG

PSA AGTGCGAGAAGCATTCCCAAC

RPL32N GCACCAGTCAGACCGATATG

SLC27A1 AGGTGGTTCAGTACATCGGG

SLC27A2 GCACATTGCTGATTACCTACC

SLC27A3 TTTCTTCAGGAGGTGAACG

SLC27A4 GTCTTGGAGAAGGAACTGC

SLC27A5 CCCATTTCATCCGCATCC

SLC27A6 CTGAGGTTGCTGATGTTATTGG

LDLR GTTGACTCCAAACTTCACTCC

VLDLR TCAGTGTATCCCAGTGTCC

GOT2 GATCCGTCCCATGTATTCC

SCARB1 CCTTGTTTCTCTCCCATCC

PLIN ACCCCCCTGAAAAGATTGCTT

LRP8 GCAAATGAAGACAGTAAGATGG

SREBF2 TCCGCCTGTTCCGATGTAC

FASN CGCTCGGCATGGCTATCT

ACAT1 AATGAACAGAGGATCAACACC

ACAT2 GCCTTCCATTATGGGAATAGG

HMGCS TTCACCATGCCTGGATCACTT

HMGCR GGATGACTCGTGGCCCAGT

ACLY AAACTTGGTCTCGTTGGG

ELOVL4 TTTCAGATGCGTCTAGTGC

ELOVL5 GGTGGTTTGTGATGAACTGG

ELOVL6 GAAGCCATTAGTGCTCTGG

ELOVL7 GAGCCAGTAAATCCTGTGG

SCD1 CCAGCTGTCAAAGAGAAGG

FADS1 ATGAACTCTCTCCTGATTGG

FADS2 CCCGGCACAACTTACACA

PLA2G2A CTCAGTTATGGCTTCTACGG

PLA2G4 GAGCATGAAGAAACTCTTGGG

PLA2G15 CTCCAAGAAGACCGAAAGC

PLA2G7 GCAATACATAAATCCTGTTGCC

PLA2G12A AGCCTTTCCCACGTTATGG

AR CAGATCAGGGGCGAAGTAGA

PSA CCAGCAAGATCACGCTTTTGTT

RPL32N ACTGGGCAGCATGTGCTTTG

SLC27A1 AGAACTCCCCGATTTGGC

SLC27A2 GATGACAGCAGGGTTAAAGC

SLC27A3 GGTGTAGAGCTGCATAAGG

SLC27A4 CAATAGCCGGGTCAAAGC

SLC27A5 GTTGTCCAGTACAAACAGAGG

SLC27A6 CATCTTCCACCAACTGATGC

LDLR GCTTCGTTGATGATATCTGTCC

VLDLR ATACAAAGTTCCTGGAGATGC

GOT2 CCATGACTTTCACTTCTTGC

SCARB1 TTCACAGAGCAGTTCATGG

PLIN GATGGGAACGCTGATGCTGTT

LRP8 GTTTCTCCAGATCAGGTATCC

SREBF2 TGCACATTCAGCCAGGTTCA

FASN CTCGTTGAAGAACGCATCCA

ACAT1 GTGCAATATTCAGCTTCTTTGC

ACAT2 CTATTGCAGCAGAGACAGC

HMGCS ATCTCAAGGGCAACAATTCCC

HMGCR TCGAGCCAGGCTTTCACTTC

ACLY TCGATCAGAAAGTTCTTGAGG

ELOVL4 CCACACTCTGGCAAATATAGC

ELOVL5 TGTACTTCTTCCACCAGAGG

ELOVL6 ACAAACTGACTGCTTCAGG

ELOVL7 CTAGGAGGATGGTTTGTGG

SCD1 AAATACCAGGGCACAAGC

FADS1 AGGAAGAAGACATGGTTGG

FADS2 CCATGCTTGGCACATAGACACTT

PLA2G2A GAAATTTGGTGCCACATCC

PLA2G4 CGTATAATGCCTTCATCACACC

PLA2G15 CACACCATCAGGAAACTGG

PLA2G7 TGTACAACCAACGGAATAAGG

PLA2G12A ATAGCACCTGTCGTGTTGG


2.3 DETECTION OF LIPID CONTENT USING QUANTITATIVE

FLUORESCENT MICROSCOPY (QFM)

Prior to seeding, optical imaging plates (IBIDI) were coated with 150 µl Poly-l-

ornithine (Sigma-Aldrich) and washed with phosphate buffered saline (PBS) to

increase cell attachment as described previously (Liberio, Sadowski, Soekmadji,

Davis, & Nelson, 2014). LNCaP and C4-2B cells were seeded at a density of 4,000

and 3,000 cells/well, respectively, in Roswell Park Memorial Institute (RPMI) medium

(Thermo Fisher Scientific) supplemented with either 5% fetal bovine serum (FBS)

(Thermo Fisher Scientific) (+/- 10 µM Enzalutamide or DMSO control) or 5% RPMI

charcoal-stripped fetal bovine serum (CSS) (Thermo Fisher Scientific). DuCaP and

VCaP cells were seeded in 10% RPMI FBS (+/- 10 µM Enzalutamide or 0.1% DMSO

control) or 10% RPMI CSS at a density of 15,000 cells/well. After treatment as

indicated, media were removed, cells were washed with PBS once, fixed with 4%

paraformaldehyde (Electron Microscopy Sciences, Thermo Fisher Scientific) for 20

minutes at room temperature, and any remaining aldehyde reacted with 30 mM glycine

in PBS for an additional 30 minutes. Nuclear DNA was then stained with 1 µg/ml 4’,6-

diamidino-2-phenylindole (DAPI, Thermo Fisher Scientific) and lipids were stained

with 0.1 µg/ml Nile Red (Sigma-Aldrich) overnight at 4°C as described previously

(Levrier, Sadowski, Nelson, Healy, & Davis, 2015). Alternatively, free cholesterol was

stained with 50 µg/ml Filipin (Sigma-Aldrich) for 40 minutes at room temperature.

>500 cells/well were imaged using the InCell 2200 automated fluorescence

microscope system (GE Healthcare Life Sciences). Quantitative analysis of 1500

cells/treatment (3 wells) was performed in at least two independent experiments with

Cell Profiler Software (Software from Broad Institute, (Kamentsky et al., 2011)).

2.4 MEASUREMENT OF LIPID UPTAKE USING QUANTITATIVE

FLUORESCENT MICROSCOPY (QFM)

Cells were seeded as described above. For quantifying C16:0 fatty acid uptake,

media were replaced with 65 µl/well of 0.2% BSA (lipid-free, Sigma-Aldrich) serum-

free RPMI media supplemented with 5 µM Bodipy-C16:0 (Thermo-Fisher) and

incubated at 37°C for one hour. Cellular uptake of cholesterol was measured as

described recently (Egbewande et al., 2018). Briefly, media were replaced with serum-

free 0.2% BSA RPMI media supplemented with 15 µM NBD cholesterol (22-(N-(7-


Nitrobenz-2-Oxa-1,3-Diazol-4-yl) Amino-23,24-Bisnor-5-Cholen-3β-Ol) (Thermo-

Fisher Scientific) and cells were incubated at 37°C for 2 hours. For quantifying

lipoprotein complex uptake, serum-free 0.2% BSA RPMI media were supplemented

with 15µg/ml 1,1'-Dioctadecyl-3,3,3',3'-Tetramethylindocarbocyanine (DiI)-labelled

acetylated-LDL (Thermo Fischer Scientific) or 15µg/ml DiI-labelled LDL (Thermo

Fisher Scientific) and incubated at 37°C for 2 hours. Phosphoethanolamine uptake was

measured as described above using 5 µM NBD-PE (22-(N-(7-Nitrobenz-2-Oxa-1,3-

Diazol-4-yl) -1, 2-Dihexadecanoyl-sn-Glycero-3-Phosphoethanolamine,

Triethylammonium Salt)) (Thermo Fischer Scientific) with 1-hour incubation at

37°C. After incubation, cells were washed and fixed as described above. Cellular DNA

and F-actin were counterstained with 1 µg/ml DAPI and 1 µg/ml Alexa Fluor 647

Phalloidin (Thermo Fisher Scientific), respectively, at 37°C. Image acquisition and

quantitative analysis were performed as described in Section 2.3. The cell mask based

on Phalloidin staining was used to calculate morphometric features of cells (cellular

area, perimeter, major axis length).

2.5 MEASURMENT OF GLUCOSE UPTAKE

Cells were seeded in 96-well IBIDI imaging plates as described previously

(Section 2.3). Media were removed and cells were stained with 65 µL/well solution of

glucose-free media with the addition of 50 µM 2-NBDG (2-(N-(7_Nitrobenz-2-oxa-

1,3-diazol-4-yl)Amino)-2-Deoxyglucose) (Thermo Fisher Scientific), and incubated at

37°C for one hour. After incubation, cells were washed and fixed as described above.

Cellular DNA and F-actin was counterstained with 1 µg/ml DAPI and 1 µg/ml Alexa

Fluor 647 Phalloidin (Thermo Fisher Scientific), respectively, at 37°C. Image

acquisition and quantitative analysis were performed as described in Section 2.3.

2.6 CELL VIABILITY, LIVE/DEAD STAINING AND LIVE-CELL

IMAGING ASSAYS

Cell viability as a function of metabolic activity was measured by a PrestoBlue

end point assay (Thermo Fischer Scientific) according to manufacturer’s instructions.

Briefly, cells were seeded into 96-well tissue culture plates (Corning, Corning, NY,

USA) at densities described in Section 2.3. After 48 hours, cells were treated with the

indicated compounds dissolved in DMSO or the equivalent dose of DMSO (final


concentration 0.2%) or 0.5 M H2O2 as a positive control to induce cell death. After 48

hours of treatment at 37°C, 12 µL/well Presto Blue reagent was added to each well

and cells were incubated at 37°C for one hour. Following incubation, fluorescence was

measured at 560nm/590nm (excitation/emission). For live/dead staining assays, cells

were seeded and treated as described in Section 2.3. Following 48 hours of treatment

at 37°C, the following solution was added to cells: Propidium Iodide (1 mg/ml final

concentration), a non-permeant dye that penetrates the membrane of dead cells;

Fluorescein Diacetate (10 mg/ml final concentration), a fluorescein analogue which

will pass through the membrane and be cleaved exclusively by living, viable cells; and

Hoescht (10 mg/mL final concentration), which stains the nuclei of both live and dead

cells. Cells were incubated at 37°C for 30 minutes and >500 cells/well were imaged

using the InCell 2200 automated fluorescence microscope system (GE Healthcare Life

Sciences) as described in Section 2.3. Cell Profiler software (Broad Institute) was used

to calculate number of live and dead cells per well.

For live-cell imaging, cells were seeded and treated as described in Section 2.3

in 96-well Essen ImageLockTM plates (Essen BioScience, Ann Arbor, Michigan,

USA). Cell proliferation as a function of percent confluency was measured by live

imaging microscopy with the IncuCyte FLR system (Essen BioScience, Ann Arbor,

Michigan, USA). Images were acquired at 2 hour intervals with a 10x objective for up

to 7 days. Graphpad Prism was used to calculate IC50s using the nonlinear fit of

log(inhibitor) vs. normalised response. All experiments were performed in technical

duplicate and biological triplicate.

2.7 PROTEIN EXTRACTION AND WESTERN BLOT ANALYSIS

Proteins for western blotting were isolated by lysing cells in

radioimmunoprecipitation buffer [RIPA, 25 mM Tris, HCl pH 7.6, 150 mM NaCl, 1%

NP-40, 1% sodium deoxycholate, 0.1% SDS, one cOmpleteTM EDTA-free Protease

Inhibitor Cocktail tablet (Roche) per 10 ml, phosphatase inhibitors 30 µM NaF, 20 µM

Sodium Pyrophosphate, 10 µM β-glycerophosphate, and 1 µM Na Vanadate]. With

cells on ice, media were carefully removed, and cells were washed with PBS. RIPA

lysis buffer was added, and cells were incubated for 5 minutes on ice before collection

of protein lysates. Samples were repeatedly pipetted up and down to shear DNA before

being centrifuged (2,000 x g) for 10 minutes at 4°C. Protein concentration was


measured using Pierce BCA Protein Assay kit according to manufacturer’s

instructions (Thermo Fisher Scientific).

20 µg of total protein/lane were separated by SDS-polyacrylamide gel

electrophoresis (SDS-PAGE) using NuPAGETM 4-12% Bis-Tris SDS-PAGE Protein

Gels (Thermo Fisher Scientific), and Western blot was completed using the Bolt Mini

Blot Module (Thermo Fisher Scientific) according to the manufacturer’s instructions.

Membranes were incubated overnight at 4ºC with primary antibodies raised against

LDLR (Abcam, ab52818), SCARB1 (Abcam, ab217318), and PLA2G2A (Abcam,

ab23705), and loading controls gamma Tubulin (Abcam, ab11316) and Vinculin (Sant

Cruz Biotechnology, sc-73614) at a dilution of 1:1000 followed by probing with the

appropriate Odyssey fluorophore-labelled secondary antibody and visualisation on the

LiCor® Odyssey imaging system (LI-COR® Biotechnology, NE, USA). Protein

expression levels were quantified using Image Studio Lite (LI-COR® Biotechnology,

NE, USA), normalised relative to the indicated housekeeping protein, and expressed

as fold-changes relative to the vehicle control treatment.

2.8 IMMUNOFLUORESCENCE STAINING

72 hours after seeding in 96-well IBIDI imaging plates as described in Section

2.3, cells were fixed with 4% paraformaldehyde for 20 minutes, washed once with

PBS and replaced with TBS-0.1% Triton in PBS to permeabilise cells. Each well was

then treated with 2% BSA to block non-specific antibody binding for 5-10 minutes at

room temperature protected from light. BSA was removed and primary antibodies

(LDLR (Abcam, ab52818) and SCARB1 (Abcam, ab217318)) were added at a 1:100

dilution in 2% BSA (60 µL/well). After 24 hours at 4°C, primary antibody was

removed, and cells were washed with TBS-0.1% Triton 3 times for 5 minutes each.

Secondary antibody (anti-rabbit, fluorophore 568 nm wavelength) was added to cells

at a 1:250 dilution in 2% BSA (60 µL /well). After one hour incubation at room

temperature, secondary antibody was removed and cells were washed with TBS-0.1%

Triton 2 times for 5 minutes each, protected from light, and treated with a counterstain

of 1 µL/ml 4’,6-diamidino-2-phenylindole (DAPI) (Invitrogen) and Phalloidin. After

24 hours at 4°C, the counterstain was removed and replaced with 200 µL PBS. Cells

were imaged using the InCell/Cytell live imaging system.


2.9 MEMBRANE FRACTION PROTEIN MASS SPECTROMETRY

For protein mass spectrometry analysis, cells were seeded in 6 cm dishes and

incubated for 48 hours at 37°C before protein collection. Samples were separated into

membrane and cytosolic fractions using the Mem-PERTM Membrane Protein

Extraction kit (Thermo Fisher Scientific) according to the manufacturer’s instructions.

Briefly, on ice, cells were washed twice with PBS and Permeabilisation Buffer was

added for 1 hour at 4°C. Lysates were collected and centrifuged for 15 minutes (16,000

x g) at 4°C. The supernatant containing cytosolic proteins was collected. The

remaining pellet was resuspended in 0.5 mL Solubilisation Buffer and mixed

thoroughly before being centrifuged at 16,000x g for 15 minutes at 4°C. The

supernatant containing membrane proteins was collected. The remaining pellet

containing nuclear proteins, along with previously collected supernatants, were stored

at -80°C.

Proteins samples were precipitated using the methanol/chloroform technique

(Engholm-Keller et al., 2012). Dried pellets were resuspended in 100 µL 50 mM

triethylammonium buffer (TEAB, pH 8.5) and samples were digested in trypsin

overnight at -20°C at a ratio of 50:1 (100 µg protein to 2 µg Trypsin). Formic acid (1%

final concentration) was added to acidify peptides. After isolation of peptides, salts

and buffers were removed using reversed phase resins on C18 matrix spin columns

(PierceTM Detergent Removal Spin Column, Thermo Fisher Scientific) at room

temperature. Following C18 clean-up, samples were dried and resuspended to a final

concentration of 0.5 µg/µl trifluoroacetic acid (TFA). 2 µg of total protein was injected

for analysis on an HPLC CHIP QTOF 6530 mass spectrometer (Agilent). Data were

analysed using Spectrum Mill analysis software (Agilent).

2.10 ISOBARIC MASS TAGGING PROTEIN MASS SPECTROMETRY

LNCaP cells were treated with 0.1% DMSO or 10 µM Enzalutamide for 21 days

as described in Section 2.1. Cell lysates from 3 independent experiments were

prepared and peptides were labelled according to manufacturer’s protocol

(ThermoFisher, cat. 90064). Briefly, proteins were precipitated using 6 volumes of

acetone overnight at -20°C. Dried pellets were resuspended in 50 mM TEAB (pH 8.5)


and samples were digested in trypsin overnight. TMT label reagent was added to each

sample and incubated for 1 hour at room temperature, after which 5% hydroxylamine

was added to quench the reaction and equal amounts of each sample were combined.

Following C18 clean-up and sample resuspension (described in Section 2.9), 2 µg of

total protein was injected for analysis using an electrospray ionisation

(ESI) QExactive HCD mass spectrometer. Data were analysed

using Proteome DiscovererTM software (Thermo Fisher Scientific).

2.11 CISTROME ANALYSIS OF AR CHIPSEQ PEAKS

AR ChIPseq analysis used BED files (hg38) downloaded from Cistrome (Mei

et al., 2017) for the LNCaP +/- Bicalutamide or 0.1% ethanol controls for the ChIPseq

dataset, GSE49832 (Ramos‐Montoya et al., 2014). The bedtools software tool (version

2.27.0) was used to identify AR ChIPseq peaks enriched in regions 5kb upstream and

also in a 25kb window around Gencode transcripts (total number of genes=60,609;

version 21).

2.12 RNA SEQUENCING ANALYSIS

For mRNAseq, total cellular RNA was extracted using the Norgen RNA

Purification PLUS kit #48400 (Norgen Biotek Corp., Thorold, Canada) according to

the manufacturer's instructions, including DNase treatment. RNA quality and quantity

were determined on an Agilent 2100 Bioanalyser (Agilent Technologies, Santa Clara,

USA) and Qubit®. 2.0 Fluorometer (Thermo Fisher Scientific Inc, Waltham, USA).

Library preparation and sequencing was done at the Kinghorn Centre for Clinical

Genomics (KCCG, Garvan Institute, Sydney) using the Illumina TruSeq Stranded

mRNA Sample Prep Kit with an input of 1 µg total RNA (RIN>8), followed by paired-

end sequencing (125 bp) on an Illumina HiSeq2500 v4.0 (Illumina, San Diego, USA),

multiplexing 6 samples per lane and yielding about 30M reads per sample.

Raw data were processed through a custom designed pipeline. Raw reads were

trimmed using 'TRIMGALORE' (Krueger, 2012), followed by parallel alignments to

the genome (hg38) and transcriptome (Ensembl v77 / Gencode v21) using the 'STAR'

(Dobin et al., 2013) aligner and read quantification with 'RSEM' (B. Li & Dewey,

2011). Differential transcript expression between two conditions was calculated after

between sample TMM normalization (Robinson & Oshlack, 2010) using 'edgeR'

https://wiki.qut.edu.au/download/attachments/182357064/Norgen_total_rna_purification_plus_kit_protocol_48400.pdf?version=1&modificationDate=1456895502384&api=v2

https://wiki.qut.edu.au/download/attachments/182357064/Norgen_total_rna_purification_plus_kit_protocol_48400.pdf?version=1&modificationDate=1456895502384&api=v2


(Robinson, McCarthy, & Smyth, 2010) (no replicates: Fisher Exact Test; replicates:

General Linear Model) and was defined by an absolute fold change of >1.5 and a false

discovery rate (FDR) corrected p-value <0.05. Quality control of raw data included

sequential mapping to the External RNA Controls Consortium (ERCC) spike-in

controls, rRNA and a comprehensive set of pathogen genomes as well as detection and

quantification of 3'bias. Heatmaps were generated with a hierarchical clustering

algorithm (heatpmap.2) using completed linkage and Euclidean distance measures

through the Shiny Application developed by Dr. Chenwei Wang. All analyses were

performed by biostatisticians at the APCRC-Q.

2.13 MICROARRAY GENE EXPRESSION PROFILING USING THE 180K

VPC CUSTOM ARRAYS

For gene expression profiling, LNCaP cells were treated for up to 21 days in

FBS + 10 µM Enzalutamide or 0.1% DMSO control. Triplicates of samples were

analysed on a custom 180k Agilent oligo microarray (VPCv3, ID032034,

GEO:GPL16604) (Sieh et al., 2012). This array contains probes mapping to human

protein-coding as well as non-coding loci; with probes targeting exons, 3’UTRs,

5’UTRs, intronic and intergenic regions. RNA was isolated using the RNeasy Mini

Kit (Qiagen, Hilden, Germany) according to the manufacturer’s protocol including an

on-column DNAse treatment step. RNA purity and quality were evaluated on a

NanoDrop1000 (Thermo Fisher Scientific Inc, Waltham, USA) and Agilent 2100

Bioanalyser (Agilent Technologies, Santa Clara, USA). 150 ng RNA of each sample

were amplified and labelled using the Agilent ‘Low Input Quick Amp Labelling Kit’

for One-Color Microarray-Based Gene Expression Analysis. In brief, the input RNA

was reverse transcribed into cDNA, using an oligo-dT/T7-promoter hybrid primer

which introduces a T7 promoter region into the newly synthesised cDNA. The

subsequent in vitro transcription uses a T7 RNA polymerase, which simultaneously

amplifies the target material and incorporates cyanine 3-labeled CTP. cDNA synthesis

and in vitro transcription were performed at 40°C for 2 h. The labelled cRNA was

purified using the Qiagen RNeasy Mini Kit and quantified on a NanoDrop1000.

Finally, 1650 ng cRNA of each sample was hybridised at 65°C for 17 h, washed for 1

minute in Wash Buffer 1 (Agilent #5188-5327) at room temperature, then washed for

1 minute in Wash Buffer 2 (Agilent #5188-5327) at 37°C, and the arrays subsequently

http://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GPL16604


scanned on an Agilent C-type Microarray Scanner G2565CA. Microarray sample

processing performed by Dr. Anja Rockstroh.

2.14 MICROARRAY DATA ANALYSIS

The microarray raw data were processed using the Agilent Feature Extraction

Software (v10.7). A quantile between array normalization was applied and differential

expression was determined using the Baysian adjusted t-statistic linear model of the

‘Linear Models for Microarray Data’ (LIMMA) (Smyth, 2005) package in R. The p-

values were corrected for a false discovery rate (Benjamini & Hochberg, 1995) of 5%

and the gene expression levels were presented as log2 transformed intensity values.

Normalised gene expression data of the experiment are ‘Minimum Information About

a Microarray Experiment’ (MIAME) compliant and will be deposited in NCBI's Gene

Expression Omnibus (GEO Edgar et al., 2002). Probes that were significantly

differently labelled between two groups were identified with an adjusted p-value of

<=0.05 and mean absolute fold change of >=1.5. For functional annotation and gene

network analysis, filtered gene lists were examined using QIAGEN’s Ingenuity®

Pathway Analysis (IPA®, QIAGEN, Redwood City, www.qiagen.com/ingenuity) and

‘Gene Set Enrichment Analysis’ (GSEA) (Subramanian et al., 2005), ‘Gene

Ontology enRIchment anaLysis and visualisation tool’ (GOrilla) (Eden, Navon,

Steinfeld, Lipson, & Yakhini, 2009), and GOsummaries (Kolde & Vilo, 2015).

2.15 LIPID EXTRACTION

All extractions were performed in 2 mL glass vials. Methanol (220 uL) was

added to cell pellets of approximately 2 million cells and vortexed for 2 minutes at

room temperature. Internal standard (40 µL SPLASH, Table 2.2) was

added and briefly vortexed. 770 µL methyl tert-butyl ether (MTBE) was added and

mixture was incubated for 1 h at room temperature in a shaker (400 rpm). Phase

separation was induced by adding 200 µL 150 mM NH4CH3CO2. After vortexing for

20 sec, the sample was centrifuged at 2,000 g for 5 min at room temperature. The upper

(organic) phase was collected and stored at -80°C, then diluted into

2:1 MeOH:ChCl3 with 7.5 mM NH4CH3CO2 for MS analysis.

http://www.qiagen.com/ingenuity


Table 2.2 Internal standards for lipid quantitation

Target lipid class, ion detected, internal standard used and amount (nmol) per sample

used are shown.1

Lipid Class Ion Internal Standard nmol/sample

PC [M+H] + PC 15:0_18:1(d7) 5

PE [M+H] + PE 15:0_18:1(d7) 1

Lyso PC [M+H] + Lyso PC 18:1(d7) 5

Lyso PE [M+H] + Lyso PE 18:1(d7) 1

CE [M+NH4] + CE 18:1(d7) 0.5

TAG [M+NH4] + TAG 15:0_18:1(d7) _15:0 0.1

2.16 LIPIDOMICS ANALYSIS

Tandem mass spectrometry of the intact lipids was performed using a triple

quadrupole mass spectrometer (6500 Qtrap, Sciex, ON, Canada). The lipid extracts (as

described in Section 2.15) were diluted 40-fold prior to analysis. Samples were infused

using a loop injection method, where 100 uL of sample was loaded into a sample loop

using an autosampler and subsequently infused into the mass spectrometer by

electrospray ionisation at a flow rate of 20 uL/min. Lipid classes were targeted using

either precursor ion or neutral loss scans. For quantification, 40 µL SPLASH Lipid-o-

mix deuterated internal standard (Avanti Polar Lipids, Alabaster, AL) was added

to cells prior to lipid extraction. Tandem MS data was processed

using LipidView (version 1.3beta; Sciex) using predefined target lists annotated by

Dr. Blanskby and Dr. Poad.

The fatty acid methyl ester (FAME) extracts were analysed with a gas

chromatograph coupled to a mass spectrometer (GCMS – TQ8040; Shimadzu, Kyoto,

Japan). The separation was carried out on an RTX-2330 capillary column

(cyanopropyl stationary phase, 60 m x 0.25 mm, film thickness 0.20 μM; Restek,

Bellefonte, PA, USA) and the electron ionisation was set at 70 eV. Conditions for the

1 PC, phosphatidylcholine; PE, phosphatidylethanolamine; CE, cholesteryl ester; TAG,

triacylglycerol.


analysis of FAMEs were as follows: carrier gas, He: 2.6 mL/min; 22:1 split ratio,

injection volume 1 μL; injector temperature 220oC; thermal gradient 150oC to 170oC

at 10oC/min, then 170oC to 200oC at 2oC/min, then 200oC to 211oC at 1.3oC/min and

temperature held for 5 minutes. The data were acquired with Q3 scan mode

from m/z 50 – 650. For data collection, MS spectra were recorded from 4 min to 30.5

minutes.

2.17 METABOLOMICS

LNCaP cells were treated with 0.1% DMSO or 10 µM Enzalutamide for up to

21 days as described in Section 2.1. 72 hours prior to harvesting, media were replaced

with complete growth media supplemented with uniformly labelled 500 µM 13C-

acetate (Sigma). Metabolites were extracted using 80% methanol in water extraction

buffer (Sigma) supplemented with 2 µM uniformly deuterated myristic acid (Sigma)

as an internal control. BCA protein quantification was performed on the protein

precipitates using a Pierce BCA Protein Assay kit according to manufacturer’s

instructions (Thermo Fisher Scientific).

GC-MS analyses were performed using an Agilent 7890A GC equipped with a

DB-35MS (30 m - 0.25 mm i.d. - 0.25 μm) capillary column (Agilent Technologies),

interfaced with a triple quadruple tandem mass spectrometer (Agilent 7000B, Agilent

Technologies) operating under ionization by electron impact at 70 eV. The injection

port, interface and ion source temperatures were maintained at 230 °C. The

temperature of the quadrupoles was maintained at 150°C. The injection volume was

1 μl, and samples were injected at a 1:25 split ratio. Helium flow was kept constant at

1 ml/min.

The GC oven temperature was maintained at 60 °C for 1 minute and increased

to 300 °C at 10.0 °C / minute. The post-run temperature was 325 °C for 5 min. The

mass spectrometer operated in SIM mode and cholesterol-tms derivative was

determined from the fragment at m/z 458,4 (C30H54OSi) as M+0, also m/z fragments

representing M+1 up to M+16 were detected in order to determine the 13C-Acetate

incorporation into cholesterol.


2.18 PHOSPHOLIPASE A2 ACTIVITY

PLA2G2A activity was measured using 5 µM Red/Green BODIPY® PC-A2

(A10072) (1-O-(6-BODIPY® 558/568-aminohexyl)-2-BODIPY® FL C5-sn-glycero-

3-phosphocholine) (Thermo Fischer Scientific) according to the manufacturer’s

instructions. Briefly, cells were plated as described in Section 2.3. Media were

replaced with PC-A2 activity assay buffer and PC-A2 substrate (Thermo Fischer

Scientific) with the addition (alone and in combination) of 100 ng sPLA2 inhibitor

(Sigma-Aldrich; cat # S3319) and human recombinant PLA2G2A protein (R&D

systems; cat # 5374-PL-010). Extracellular enzymatic activity was measured at 3-

minute intervals for 1 hour at room temperature using a fluorescence microplate

reader. To detect cellular uptake of cleaved PCs following the incubation period, the

assay buffer was removed, and cells were fixed with 4% PFA. Cellular DNA and F-

actin was then counterstained with DAPI and Alexa Fluor 647 Phalloidin (Thermo

Fisher Scientific) as described in Section 2.4. Image acquisition and quantitative

analysis were performed as described in Section 2.3. PLA2 activity was calculated as

the ratio of green (cleaved PC) vs red (uncleaved PC) signal.

2.19 ENZYME-LINKED IMMUNOSORBENT ASSAY (ELISA)

PCa cells were treated for up to 14 days with 10 µM Enzalutamide or 0.1%

DMSO control. Approximately 10 mL of conditioned media were collected and

concentrated at room temperature using Amicon Ultra-15 Centrifugal Filter Units

(Sigma, UFC9003) with a molecular weight cut-off of 3 KDa. Levels of PLA2G2A

were measured using an ELISA kit (Cat. Number 501380, Cayman

Chemical, AnnArbour, MI, USA) following the manufacturer’s instructions.

2.20 STATISTICAL ANALYSIS

Statistical analyses were performed with Graphpad Prism 7.0 (Graphpad

Software, San Diego, CA) and R Studio (RStudio, Boston, MA). Data and statistical

tests are included in figure legends.

Androgen regulation of lipid uptake 63

Androgen regulation of lipid uptake

3.1 INTRODUCTION

The role of lipid metabolism in the incidence and progression of PCa and several

other cancer types has gained notable attention in an attempt to develop new

therapeutic interventions. Lipids represent a diverse group of compounds derived from

fatty acids and cholesterol that serve an essential role in many physiological and

biochemical processes. Lipids function in energy generation and storage as well as

intracellular signalling, protein modification, and precursor for steroid hormone

synthesis. Additionally, fatty acids serve as the main building blocks for phospholipids

that are incorporated together with free cholesterol into membranes and are critical for

membrane function, cell signalling and proliferation.

As a source of lipid supply, uptake of circulating exogenous lipids is sufficient

for the requirements of most normal cells, and following development, lipogenic

enzymes remain expressed at relatively low levels apart from a few specific biological

processes (surfactant production in the lungs, production of fatty acids for milk lipids

during lactation, and steroidogenic activity in tissues including prostate) (Menendez

& Lupu, 2007). However, lipogenic pathways, i.e. de novo lipogenesis (DNL) of fatty

acids and cholesterol, are reactivated or upregulated in many solid cancer types

including prostate cancer (Brusselmans & Swinnen, 2009). Enhanced lipogenesis is

now acknowledged as a metabolic hallmark of cancer and is an early metabolic switch

in the development of prostate cancer. This phenotype is maintained throughout the

progression of PCa and associated with poor prognosis and aggressiveness of disease.

(Deep & Schlaepfer, 2016; Flavin et al., 2011; Fritz et al., 2010; Menendez & Lupu,

2007; Swinnen et al., 2006). Yet, the contribution and identity of lipid uptake pathways

as a supply route of exogenous lipids and their role in disease development and

progression remain largely unknown (Pinthus et al., 2007; Nath et al., 2016b).

Several lipogenic enzymes including fatty acid synthase (FASN) are found to be

overexpressed in PCa [reviewed in (Swinnen et al., 2006; Galbraith, Leung & Ahmad,

2018)]. Because increased FASN gene copy number, transcriptional activation or

protein expression are common characteristics of prostate cancer (Swinnen et al.,

2006), fatty acid and cholesterol synthesis have become an attractive therapeutic


target. However the antineoplastic effects observed by inhibiting lipogenesis can be

rescued by the addition of exogenous lipids (Griffiths et al., 2013; Kuemmerle et al.,

2011), highlighting lipid uptake as a mechanism of clinical resistance to lipogenesis

inhibitors and that lipid uptake capacity is sufficient to substitute for the loss of

lipogenesis. Indeed, it was recently reported that lung cancer cells expressing a strong

lipogenic phenotype generated up to 70% of their cellular lipid carbon biomass from

exogenous fatty acids and only 30% from de novo synthesis supplied by glucose and

glutamine as carbon sources (Hosios et al., 2016).

While altered cellular lipid metabolism is a hallmark of malignant phenotype,

PCa is unique in that it is characterised by a relatively low uptake of glucose and

glycolytic rate compared to many solid tumours subscribing to the “Warburg effect”

phenotype (Effert et al., 1996; Zadra et al., 2013). Concordantly, PCa cells showed a

dominant uptake of fatty acids over glucose, with the uptake of palmitic acid measured

at ~20 times higher than uptake of glucose in both malignant and benign PCa cells

(Liu, Zuckier, & Ghesani, 2010a). Furthermore, exogenous fatty acids are readily

oxidized by PCa, reducing glucose uptake (Schlaepfer et al., 2015). Together, these

studies demonstrate that exogenous uptake is a significant and previously

underappreciated supply route of lipids in cancer cells with a lipogenic phenotype.

Both healthy and malignant prostate cells rely on androgens for a variety of

physiological processes, including several metabolic signalling pathways. Androgens,

through binding to the androgen receptor (AR), transcriptionally regulate a multitude

of pathways, including proliferation, differentiation and cell survival of PCa, with

approximately equal numbers of genes activated and suppressed by androgen-

activated AR (Lonergan & Tindall, 2011). Targeting the AR signalling axis is the

mainstay treatment strategy for advanced prostate cancer. While initially effective in

suppressing tumour growth, patients inevitably develop castrate-resistant PCa (CRPC)

which remains incurable (Kirby et al., 2011). Importantly, during progression to

CRPC, survival and growth of PCa cells remain dependent on AR activity, as

demonstrated by treatment resistance mechanisms involving AR mutation,

amplification and intratumoural steroidogenesis [reviewed in (Dutt & Gao, 2009)].

Thus, identifying critical pathways regulated by AR might provide novel therapeutic

strategies to combat development of CRPC.

Lipogenesis is a well-described AR-regulated metabolic pathway that supports

PCa cell growth by providing fuel, membrane material and steroid hormone precursor


(cholesterol). Androgens stimulate expression of FASN via activation of sterol

regulatory element-binding proteins (SREBPs) (Heemers et al., 2004), lipogenic

enzymes ACACA and ACLY, and cholesterol synthesis enzymes HMGCS1 and

HMGCR (Fritz et al., 2010; Swinnen et al., 1996b). In contrast, the role and expression

of lipid transporters and their regulation by AR in PCa remain largely uncharacterised

(Liu et al., 2010a; Pinthus et al., 2007)

Our current understanding of lipid uptake mostly derived from studies in non-

malignant cells and tissues (Doege & Stahl, 2006; Sahoo et al., 2014). The

hydrophobic properties of lipids allow for passive, non-specific uptake via diffusion

into the cell. Selective, protein-mediated lipid uptake involves receptor-mediated

endocytosis of lipid transporters and their cognate lipoprotein cargo which contains

various lipid components (phospholipids, cholesterol esters, triacylglycerol) that can

be internalised via lipoprotein receptors (LDLR, VLDLR) or scavenger receptors

(SCARB1, SCARB2) (Doege & Stahl, 2006; Sahoo et al., 2014). Various scavenger

receptors have also been shown to be associated with uptake of modified (acetylated

or oxidized) LDL particles including SCARF1, SCARF2 and CXCL16 (Miller et al.,

2010; Tamura et al., 2004). Free fatty acids can be taken up by a family of six fatty

acid transport proteins (SLC27A1-6) as well as fatty acid translocase (FAT/CD36) and

mitochondrial aspartate aminotransferase (GOT2) (Pinthus et al., 2007).

Taken together, it is becoming evident that enhanced lipogenesis in PCa

development and progression is not the sole deregulated lipid supply pathway, and that

lipid uptake might play an important role in biochemical recurrence of prostate cancer.

This warranted a comprehensive investigation and delineation of lipid uptake and the

lipid transporter landscape in PCa as well as the regulation of lipid uptake by AR.

3.2 RESULTS

3.2.1 Androgens strongly increased cellular lipid content in AR-positive PCa

cells

Previous analysis by cellular Oil Red O staining and lipid chromatography of

cellular extracts have demonstrated that androgens strongly enhance lipogenesis and

cellular lipid content in PCa cells, predominantly that of neutral lipids (triacylglycerols

and cholesterol-esters) stored in lipid droplets and phospholipids and free cholesterol


present in membranes (Swinnen, Esquenet, Goossens, Heyns, & Verhoeven, 1997;

Swinnen et al., 1996b). Consistent with these findings, quantitative fluorescence

microscopy (qFM) assays (Levrier et al., 2015) of Nile Red-stained AR-positive PCa

cell lines LNCaP, C4-2B, VCaP and DuCaP showed that synthetic androgen R1881

significantly increased cellular phospholipid and neutral lipid content across all four

cell lines, as well as lipid droplet number in LNCaP, C4-2B and VCaP cells (Fig.

3.1A). This stimulatory effect of R1881 was also observed with DHT and mibolerone

and blocked in the presence of Enzalutamide (personal communication, Dr. Martin

Sadowski). Furthermore, qFM of filipin-stained LNCaP cells confirmed that both 1881

and DHT also increased cellular levels of free, unesterified cholesterol (Fig. 3.1B),

which was also blocked by Enzalutamide.


Figure 3.1 Androgens increase lipid content of AR-positive PCa cell lines

(A) LNCaP, C4-2B, VCaP and DuCaP cells were grown in charcoal-dextran stripped

serum (CSS) for 48 hours and treated with 1 nM R1881 or 0.1% ethanol vehicle (Ctrl)

for 48 hours. Fixed cells were stained with fluorescent lipid stain Nile Red, and cellular

mean fluorescent intensities (MFI) of phospholipid content (top left panel) and neutral

lipid content (top right panel) as well as cellular number of lipid droplets (bottom left

panel) and total cellular area of lipid droplets (bottom right panel) were measured by

quantitative fluorescence microscopy (qFM) (n~3000 cells, graphs show mean±SD,

One-way ANOVA with Dunnett’s multiple comparisons test relative to cell line

specific 0.1% ethanol vehicle (Ctrl): ns=not significant, ***p<0.001, or ethanol-treated

LNCaP cells: #p<0.001, representative of 3 independent experiments). Representative

40x images of LNCaP cell are shown (blue=DNA, yellow=lipid droplets containing

neutral lipids, scale bar=20 µm). Data in Fig. 3.1 A had contribution from Dr.

Sadowski. (B) LNCaP cells were grown as described in (A) and treated with the

indicated androgens in the presence or absence of Enzalutamide (Enz, 10 µM). Fixed

cells were stained with Filipin to label free, unesterified cholesterol, and MFI of

cellular free cholesterol was measured by qFM. (n~3000 cell, graphs show mean±SD,

One-way ANOVA with Dunnett’s multiple comparisons test relative to 0.1% ethanol

vehicle (Ctrl): ns=not significant, ***p<0.001, or ethanol treated LNCaP cells:

#p<0.001, representative of 3 independent experiments). Representative 40x images of

LNCaP cell are shown (blue=DNA, green=free cholesterol, scale bar=10 µm). Images

in Fig 3.1A-B provided by Dr. Sadowski.


3.2.2 Fatty acid, cholesterol and lipoprotein uptake are increased by

androgens

While androgen-enhanced lipogenesis is a well characterised fuel source for

increased cellular lipid content, the role of lipid uptake in this process is still poorly

understood. To directly measure the stimulatory effect of androgens on lipid uptake, a

series of lipid uptake assays (Fig. 3.2) was used based on qFM of fluorophore labelled

lipid probes (Bodipy-C16:0, NBD-cholesterol, DiI-LDL, and DiI-acetylated LDL). As

shown in Figure 3.2A, androgen treatment (R1881) of four AR positive PCa cell lines

(LNCaP, C4-2B, DuCaP, and VCaP) for 48 hours significantly increased uptake of

Bodipy-C16:0. This effect was significantly blocked when cells were co-treated with

Enzalutamide. Similar to fatty acid uptake, DHT also significantly increased uptake of

NBD-cholesterol in AR-positive PCa cells (Fig. 3.2B), and Enzalutamide significantly

suppressed this effect. Representative images of LNCaP cells show that Bodipy-C16:0

and NBD-cholesterol were incorporated into lipid droplets (Fig. 3.2A and 3.2B).

The majority of serum lipids are transported as lipoprotein particles (chylomicrons,

VLDL, LDL, HDL), containing a complex mixture of apolipoproteins, phospholipids,

cholesterol and triacylglycerols which are taken up into cells by receptor-mediated

endocytosis through cognate lipoprotein receptors such as the LDL receptor (LDLR)

for LDL and scavenger receptor SCARB1 for acetylated LDL/HDL (Miller et al.,

2010; Tamura et al., 2004). Notably, in contrast to the covalent Bodipy and NBD

fluorophore tags on C16:0 and cholesterol, the DiI label is a non-covalently bound dye

infused into the lipoprotein particles that dissociates after cellular uptake and

lysosomal processing. As shown in Figure 3.2C, R1881 and DHT significantly

enhanced uptake of DiI-complexed LDL and acetylated LDL in LNCaP cells in a dose-

dependent manner, indicating a potential role for their cognate receptors LDLR and

SCARB1.



Figure 3.2 Androgens strongly increase lipid uptake

(A) To measure fatty acid uptake, indicated cell lines were grown in CSS for 48 hours and

treated with 1 nM R1881 in the presence or absence of Enz (10 µM) or 0.1% ethanol vehicle

(Ctrl) for 48 hours. Before fixation, cells were incubated with Bodipy-C16:0 for one hour

and lipid uptake was measured by qFM (n~3000 cells from 3 wells, mean±SD, One-way

ANOVA with Dunnett’s multiple comparisons test relative to cell line specific 0.1% ethanol

control (Ctrl, grey), ns=not significant, ***p<0.001, representative of 2 independent

experiments). Representative 40x images of DuCaP cell are shown (blue=DNA, red=F-

actin, green=lipid droplets containing C16:0-Bodipy, scale bar=20 µm). (B) To measure

cholesterol uptake, LNCaP cells were grown in CSS for 48 hours and treated with either 1

nM R1881 or 10 nM DHT in the presence or absence of Enz (10 µM). Before fixation, cells

were incubated with NBD-cholesterol for 2 hours and cellular levels were measured by

qFM (n~3000 cells from 3 wells, mean±SD, One-way ANOVA with Dunnett’s multiple

comparisons test relative to 0.1% ethanol vehicle (Ctrl), or unpaired t-test between androgen

treatment alone or in combination with Enzalutamide, ns=not significant, ****p<0.0001,

representative of 2 independent experiments). Representative 40x images of LNCaP cell

are shown (blue=DNA, red=F-actin, green=lipid droplets containing NBD-cholesterol,

scale bar=20 µm). Data analysis had contribution from Dr. Sadowski. (C) To measure

lipoprotein uptake, LNCaP cells were grown in CSS for 48 h and treated with increasing

concentrations of DHT or 1 nM R1881. Before fixation, cells were incubated with DiI-LDL

or DiI-acLDL for 2 hours and lipoprotein uptake was measured by qFM (n~3000 cells from

3 wells, mean±SD, One-way ANOVA with Dunnett’s multiple comparisons test relative to

0.1% ethanol control (Ctrl) ****p<0.0001, representative of 3 independent experiments.

Images (Fig. 3.2A-B) provided by Dr. Sadowski.


3.2.3 Androgen-enhanced lipid uptake is independent of cell cycle progression

and proliferation

Androgen-mediated activation of AR promotes G0/G1 to S phase progression of the

cell cycle and proliferation in PCa cells [reviewed in (Balk & Knudsen, 2008; Heinlein

& Chang, 2004; Lonergan & Tindall, 2011)]. Because cell proliferation requires

substantial membrane biogenesis for daughter cell generation, it was possible that

androgen-enhanced lipid uptake was not mediated directly through AR signalling but

indirectly as a result of androgen-stimulated proliferation. To address this possibility,

LNCaP cells were synchronised in G0/G1 (>95% of cell population, Fig. 3.3A) by

incubation in CSS medium for 48 hours and treated for another 24 hours with three

different cell cycle inhibitors, which upon androgen (DHT) treatment induced re-entry

into the cell cycle and caused arrest in G0/G1 (Tunicamycin), S phase (Hydroxyurea)

or G2/M (Nocodazole) (Fig. 3.3A). As shown in Figure 3E-F, lipid uptake of Bodipy-

C16:0 and NBD-cholesterol was significantly and to a similar magnitude enhanced by

androgen in the presence of all three cell cycle inhibitors when compared to 0.1%

DMSO control. Flow cytometry of DNA content confirmed cell cycle arrest (Fig.

3.3B) and decreased proliferation (Fig. 3.3C) of the inhibitors, respectively. Thus,

androgen regulation of lipid uptake is directly mediated by AR throughout the cell

cycle and is independent of cell cycle progression and proliferation. Notably, a time

course experiment of DHT-treated G0/G1 synchronised LNCaP cells in the presence

of Tunicamycin (Fig. 3.3B) confirmed that the androgen-enhanced expression of

classical AR-regulated genes KLK3/PSA (Fig. 3.3D), TMPRSS2 and FKBP5 (data not

shown) (Jin, Kim, & Yu, 2013) remained unaffected under the experimental

conditions.



Figure 3.3 Androgen-enhanced lipid uptake is independent of cell cycle progression

and proliferation

(A) LNCaP cells were synchronised in G0/G1 by androgen deprivation (CSS for 48 h)

followed by treatment with Tunicamycin (1 mg/mL), Hydroxyurea (1 M), or Nocodazole

(25 µg/mL) for another 24 h, placing cell cycle blocks in G0/G1, S phase and mitosis,

respectively. Cell cycle re-entry and progression to the respective cell cycle block was

stimulated by DHT (10 nM) for 24 h in all samples but CSS which was treated with DMSO

vehicle. Cells were harvested and fixed, and DNA content and cell cycle distribution were

analysed by flow cytometry and quantitation with ModFit LT software, respectively (n=1).

(B) LNCaP cells were synchronised in G0/G1 as described above (CSS) and treated with

vehicle or Tunicamycin (1 mg/mL, CSS+Tun) for 24 h. Cells were then treated with DHT

in the absence (CSS+DHT) or presence of Tunicamycin for an additional 24 h, and samples

were taken at the indicated times (3h – 24h) and analysed as above (gray bars=G0/G1, red

bars=S phase, blue bars=G2/M; n=1) (C) LNCaP cells were treated with cell cycle inhibitors

Tunicamycin (1 mg/mL), Hydroxyurea (1 M), or Nocodazole (25 µg/mL) in the presence or

absence of DHT (10 nM) and confluence was measured every 2 h for 96 h using the live cell

IncuCyte FLR imaging system (n=3 independent experiments, mean±SD). (D) Transcript

expression of PSA was measured by qRT-PCR in LNCaP cells synchronised in G0/G1 as

described above followed by treatment as described in B (n=3 independent experiments,

mean±SD) After 24 h, (E) cholesterol (NBD-Cholesterol) and (F) fatty acid uptake (Bodipy-

C16:0) was measured by qFM. (n~3000 cells from 3 wells, mean±SD, One-way ANOVA

with Dunnett’s multiple comparisons test relative to 0.1% ethanol vehicle (Ctrl), or unpaired

t-test between androgen treatment alone or in combination with cell cycle inhibitor, ns=not

significant, ****p<0.0001, **<0.01, representative of 2 independent experiments. Cell cycle

analysis had contribution from Dr. Sadowski.


3.2.4 Delineating the lipid transporter landscape in PCa

While the role of proteins involved in de novo lipogenesis (e.g. ACLY, ACACA,

FASN, HMGCR) are well described in PCa and their overexpression is associated with

tumour development, disease progression, aggressiveness, and poor prognosis,

(reviewed in (Menendez & Lupu, 2007; Swinnen et al., 2006, Galbraith et al., 2018),

very little is known about the expression and functional importance of lipid

transporters in prostate cancer, their regulation by androgens and their clinical

relevance. To delineate the lipid transporter landscape in prostate cancer, a panel of 44

candidate genes encoding lipid transporters was generated based on previous work

describing lipid transport function in various human tissues (Anderson & Stahl, 2013;

Doege & Stahl, 2006; Go & Mani, 2012; Goldstein, Anderson, & Brown, 1982;

Kennedy, Charman, & Karten, 2012; Nath & Chan, 2016b; Wang et al., 2007).

Transcriptomic analysis by RNAseq revealed that 41 candidate lipid transporter

genes were expressed in five PCa cell lines (LNCaP, DuCaP, VCaP, PC-3 and Du145)

under normal culture conditions (a selection of 36 candidates are shown in Fig. 3.4A).

Importantly, lipid transporters LDLR, SCARB1, SCARB2, and GOT2/ FABPpm; were

robustly expressed at levels comparable to lipogenic genes HMGCR and FASN (Fig.

3.4A), whereas seven transporter genes, including CD36 and SLC27A6 displayed

negligible fragments per kilobase million (FPKM) values in the majority of cell lines.

In addition, transcripts encoding 41 lipid transporters was detected using integrated

transcriptomics and proteomics in LNCaP and Du145 cells and six additional PCa cell

lines (CWR22RV1, EF1, H660, LASCPC-01, NB120914 and NE1_3) (Lee et al.,

2018), verifying expression of these transporters in a total of nine PCa cell lines.

Comparison of this list with the recently delineated plasma membrane proteome of

eight PCa cell lines, including LNCaP, Du145 and CWR22Rv1 (Lee et al., 2018), and

previous work in LNCaP and CWR22Rv1 cells (Pinthus et al., 2007), confirmed the

protein surface expression of LDLR, GOT2, LRPAP1, LRP8 and SCARB2. In

addition, our proteomics analysis confirmed the exclusive expression of SCARB1,

SCARB2, LRPAP1, SLC27A1 and SLC27A2 in the membrane fraction of LNCaP

cells, while GOT2 was also present in the soluble fraction (Fig. 3.4B), which is

consistent with its mitochondrial function (Bradbury, Stump, Guarnieri, & Berk,

2011). Western blot analysis detected LDLR and SCARB1 in cell lysates from 7

malignant and 2 non-malignant prostate cell lines (Fig. 3.4C).


Subsequently, PCa patient samples and clinical relevance was investigated by

analysing published tumour transcriptome datasets with Oncomine (Rhodes et al.,

2004). Comparison of primary, localised PCa versus normal prostate gland revealed

that transcript levels of only a few lipid transporter genes were significantly (p<0.05)

upregulated in primary PCa and no lipid transporter was significantly downregulated

(Grasso et al., 2012; data not shown). However data mining of the reported proteome

analysis of primary PCa versus neighboring non-malignant tissue revealed that levels

of 21 lipid transporters were lower in primary PCa, whereas protein levels of both de

novo lipogenesis enzymes FASN and HMGCR were increased by several magnitudes

compared to non-malignant tissue (Iglesias-Gato et al., 2016). Although a measurable

degree of discordance between transcript and protein levels has been previously noted

in integrated transcriptome and proteome studies of PCa, the proteomics data

suggested that lipid uptake is reduced and DNL is increased in primary PCa when

compared to normal prostate gland (Iglesias-Gato et al., 2016; Latonen et al., 2018).

In contrast, transcripts of several lipid transporter genes were significantly upregulated

in metastatic PCa tumour samples compared to primary site in the Grasso dataset

(Grasso et al., 2012), including SLC27A1, SLC27A3, SCARB1 and LDLR (Fig. 3.4D).

Concordantly, analysis of the proteome comparison of localised PCa versus bone

metastasis (Iglesias-Gato et al., 2018) demonstrated that expression of 16 lipid

transporters and FASN was higher in bone metastases (Fig. 3.4E), suggesting that

tumour lipid supply from both uptake and DNL was increased. The lipoprotein

transporters LDLR and SCARB1 were further investigated across other PCa patient

cohorts in Oncomine, including the Varambally (Varambally et al., 2005) and La

Tulippe (LaTulippe et al., 2002) data sets. Both lipid transporter mRNAs were found

to be significantly upregulated in samples from PCa metastases when compared to

primary tumours (Fig. 3.4F). Together, independently published data and the analyses

described here confirmed the mRNA, protein and plasma membrane expression of

several lipid transporters in PCa cell lines and patient-derived tumour samples.

Importantly, this analysis demonstrated that this route of lipid supply is clinically

significant during disease progression and is associated with metastasis to the bone.



Not detected


Figure 3.4 Delineation of the lipid transporter landscape in PCa

(A) Transcript (mean FPKM= fragments per kilobase million, n=2 independent

replicates) of the indicated candidate lipid transporter genes and 2 lipogenic genes

(FASN and HMGCR) were measured by RNAseq in the five indicated PCa cell lines

grown in their respective maintenance media. Data analysis has contribution from Dr.

Sadowski. (B) Mass spectrometry analysis of LNCaP cells shows total ion peak

intensity of lipid transporters based on Gene ontology accession numbers in membrane

and cytosolic fractions. (C) Western blot confirmed detection of LDLR and SCARB1

in the seven indicated PCa cell lines and in two non-malignant prostate cell lines

(RWPE-1, BPH-1) grown in maintenance media. Bar chart shows densitometric

analysis of each transporter normalised to Vinculin and expressed as fold change

relative to LNCaP. A representative blot of two independent experiments is shown;

full blot shown in Appendix FigA3. (D) Differential expression analysis of candidate

lipid transporter gene transcripts measured by microarray (Grasso et al., 2012)

comparing transcript levels of indicated genes in localised, primary PCa versus

metastatic PCa. Over-expression (red) fold change is designated with a positive

number; under-expression (blue) is designated with a negative number. Student’s t-

test used to generate p-value for fold change. (E) Protein analysis of indicated 18 lipid

transporters and two lipogenesis enzymes (FASN and HMGCR) in paired patient

samples of localised primary tumour and (blue) and bone metastasis (red) in the

Iglesias-Gato proteome dataset (Iglesias-Gato et al., 2018). Graph shows peak ion

intensity measured by mass spectrometry. Analysis generated by Dr. Sadowski. (F)

Gene expression of LDLR (left) and SCARB1 (right) was compared in primary (P) vs

metastatic (M) PCa in Grasso (n=59 P, 35 M) (Grasso et al., 2012), Varambally (n=7

P, 6 M) (Varambally et al., 2005), and LaTulippe (n=23 P, 9 M) (LaTulippe et al.,

2002) cohorts (graph shows log2 median-centered ratio; Student’s t-test used to

generate p-value; ns=not significant, ****p<0.0001, ***p<0.001, **p<0.01,

*p<0.05).


3.2.5 Androgens regulate the expression of several lipid transporters

As shown in Section 3.2.2, androgens strongly enhance lipid uptake in AR-positive

PCa cell lines. However, our current understanding of the androgen receptor regulation

of lipid transporters is very limited (Pinthus et al., 2007; Galbraith et al., 2018). To

address this, a comprehensive analysis of androgen-regulated lipid transporter genes

was initiated by searching for AR binding sites within a 25 kb window of the gene

sequence and a 5 kb window upstream of the first ATG codon of 48 candidate lipid

metabolism genes in the reported AR ChIPseq data set of LNCaP cells treated with

AR-antagonist Bicalutamide (Ramos‐Montoya et al., 2014). As shown in Figure 3.5A,

19 and 27 lipid transporter genes showed enrichment of AR ChIPseq peaks in the 5 kb

and 25 kb windows, respectively, which was reduced in the presence of Bicalutamide.

Consistent with the reported androgen-regulation (Pinthus et al., 2007), AR ChIPseq

peaks were detected in the 25 kb window of the GOT2 gene which were absent after

Bicalutamide treatment. For comparison, the AR-regulated lipogenesis genes ACACA,

FASN and HMGCR (Swinnen, Ulrix, Heyns, & Verhoeven, 1997) also showed less

enrichment of AR ChIP peaks with Bicalutamide.

Next, transcript levels of 42 candidate lipid transporter genes were measured by

RNAseq in three AR-positive, androgen-sensitive PCa cell lines (LNCaP, DuCaP,

VCaP) under conditions identical to the lipid content and uptake studies shown above

(androgen deprivation in CSS for 48 hours and treatment with either 0.1% ethanol or

DHT (10 nM) for 48 hours). As a control, AR function was blocked with Enzalutamide

in the presence and absence of DHT. As shown in Figure 3.5B, RNAseq analysis

demonstrated that expression of 36 lipid transporter genes was altered by androgen

treatment in LNCaP cells. Cholesterol efflux pump ABCA1 and scavenger receptor

SCARF1 transcripts were reduced by DHT, a response that was antagonised by

Enzalutamide. In contrast, DHT increased the expression of fatty acid transporter

genes (GOT2, SLC27A3, SLC27A4, SLC27A5, CD36) and lipoprotein transporters

(LDLR, LRP8, SCARB1) which was also prevented by Enzalutamide treatment.

Receptor-mediated endocytosis of lipoprotein particles through LDLR, VLDLR,

SCARB1, SCARB2 and LDL receptor related proteins (LRP1-12, LRPAP1)

converges in lysosomes for lipolysis and release of free cholesterol and fatty acids into

the cytoplasm through their respective efflux pumps (Schneider, 2016). Consistent


with this, transcript expression of lysosomal cholesterol efflux transporter NPC1 was

also increased by DHT compared to vehicle-treated cells. Similar effects of DHT

regulation of lipid transporter gene expression was observed in DuCaP and VCaP cells,

with the exception of SLC25A, LRP8 and SCARB1 which were repressed by DHT (Fig.

A2 in Appendices).


Figure 3.5 AR binding sites and the androgen regulated expression of lipid

transporters

(A) AR ChIPseq peak enrichment analysis of 42 lipid transporter genes and six lipogenesis

genes (ACLY, ACSS2, ACACA, FASN, HMGCS1, HMGCR) in the Ramos-Montoya data set

of LNCaP cells treated with Bicalutamide (BIC) compared to vehicle control (VEH)

(Ramos‐Montoya et al., 2014). The number of peaks detected using the Cistrome Analysis

Pipeline within the analysis window is highlighted by the bubble size and the enrichment

score by the grey scale. Maximum score represents the peak with the maximum Cistrome

Analysis Pipeline enrichment score within the window. (B) LNCaP cells were grown in CSS

for 48 hours and treated with 10 nM DHT in the absence or presence of Enz (10 µM) or

0.1% ethanol (Ctrl) for 48 hours. Indicated lipid transporter gene expression was analysed

by RNAseq. Heatmaps were generated with a hierarchical clustering algorithm using

completed linkage and Euclidean distance measures and scaled by row z score (red=positive

z score, blue=negative z score (heatmap.2, Section 2.13). AR ChIPseq analysis generated by

Dr. Lehman.


qRT-PCR showed significantly increased transcript expression of the lipoprotein

transport receptor genes LDLR (p<0.0001) and VLDLR (p<0.0001) in LNCaP cells

treated with 1 nM R1881 (Fig. 3.6A top panel), however no significant change in

expression was detected for SCARB1 and SLC27A4 (Fig. 3.6A bottom panel). Co-

treatment with Enzalutamide (10 µM) blocked the increase in lipid transporter

transcript expression (Fig. 3.6A). Western blot analysis demonstrated that LNCaP cells

exposed to 10nM DHT showed an almost 2-fold increase in LDLR protein detection,

which was suppressed to levels similar to 0.1% ethanol control when co-treated with

Enzalutamide (Fig. 3.6B, left panel). No significant increase in SCARB1 was detected

in cells treated with DHT (Fig. 3.6B, right panel). Cellular localisation of LDLR

protein in response to R1881 (1 nM) using immunofluorescent microscopy showed

that androgen treatment resulted in significantly increased detection of LDLR at the

cellular periphery (plasma membrane, Fig. 3.6C), which was blocked by Enzalutamide

(10 µM), confirming that AR signalling enhanced the detection of LDLR protein at

the cell surface. Finally, analysis of our previously reported longitudinal LNCaP

xenograft study (Locke et al., 2008) revealed that transcript levels of LDLR, VLDLR,

SCARB1, SLC27A5 and SLC27A6 were reduced in tumours seven days after castration

(nadir) when compared to tumours from mice that had not been castrated (intact) (Fig

3.6D), which is consistent with the positive AR-regulation of expression in LNCaP

cells in vitro shown in Fig 3.5.



3.3 DISCUSSION

Increased activation of de novo lipogenesis is a well-established metabolic

phenotype in PCa and other types of solid cancer, however therapeutic inhibition of

DNL alone has so far had only limited clinical success as therapy against neoplastic

disease (Schcolnik-Cabrera et al., 2018). Targeting DNL in pre-clinical cancer models,

including work done by our laboratory in prostate cancer, demonstrated that inhibition

of DNL leading to lipid starvation can be efficiently rescued by exogenous lipids

(Sadowski et al., 2014). Furthermore, obesity has been associated with more

aggressive disease at diagnosis and higher rate of recurrence in PCa patients [reviewed

in (Balaban et al., 2015; Taylor, Lo, Ascui, & Watt, 2015)]. Thus, exogenous lipids may

Figure 3.6 Androgens regulated the transcript and protein expression of lipid

transporters in vitro and in vivo

(A) Transcript expression of indicated lipid transporter genes was measured by qRT-

PCR in LNCaP cells grown for 48 hours in CSS followed by treatment with 1 nM R1881

in the presence or absence of Enz (10 µM) for an additional 48 hours (n=3 independent

experiments, mean±SD, One-way ANOVA with Dunnett’s multiple comparisons test

relative to 0.1% DMSO CSS (Ctrl ns=not significant, ****p<0.0001). (B) LNCaP cells

were grown in CSS for 48 hours and treated with 10 nM DHT in the presence or absence

of Enz (10 µM) or 0.1% DMSO control. Protein was measured by Western blot analysis

(top panel). LDLR and SCARB1 were normalised by loading control (gamma tubulin),

and fold changes were calculated relative to CSS control (bottom panel) (n=3,

mean±SD, One-way ANOVA with Dunnett’s multiple comparisons test relative to

vehicle control (CSS). A representative blot of three independent experiments is shown;

full blot shown in Appendix FigA4). (C) Cells were treated as described in (B). After

fixation, cells were incubated with LDLR primary antibody for 24 hours and

counterstained with anti-rabbit secondary antibody. Protein expression was measured by

qFM. (D) Analysis of indicated lipid transporter genes and lipogenic genes in paired

LNCaP tumour xenografts before (intact) and seven days after castration (nadir) of our

previously reported longitudinal LNCaP tumour progression dataset (n=10, mean±SD,

unpaired t-test; ns=not significant, *p<0.05, **p<0.01, ***p<0.001) (Locke et al.,

2008).

Figure 3.6 Androgens regulated the mRNA and protein expression of lipid

transporters in vitro and in vivo

(A) Transcript expression of indicated lipid transporter genes was measured by qRT-

PCR in LNCaP cells grown for 48 hours in CSS followed by treatment with 1 nM R1881

in the presence or absence of Enz (10 µM) for an additional 48 hours (n=3 independent

experiments, mean±SD, One-way ANOVA with Dunnett’s multiple comparisons test

relative to 0.1% DMSO vehicle (Ctrl ns=not significant, ****p<0.0001). (B) LNCaP

cells were grown in CSS for 48 hours and treated with 10 nM DHT in the presence or

absence of Enz (10 µM) or 0.1% DMSO control. Protein was measured by Western blot

analysis (top panel) and quantitated by densitometry analysis (bottom panel), and total

protein levels were normalised to loading control (gamma tubulin) (mean±SD, One-way

ANOVA with Dunnett’s multiple comparisons test relative to vehicle control (CSS). A

representative blot of three independent experiments is shown). (C) Cells were treated


play a much more significant role in PCa and other types of cancer than previously

acknowledged (Swinnen et al., 2006). Indeed, recent estimates derived from studies in

lung cancer cells with a similar lipogenic phenotype as PCa cells suggested that 70%

of lipid carbon biomass is derived from exogenous lipids and only 30% from DNL

(Hosios et al., 2016). While androgens are known to activate DNL in PCa (Swinnen,

Van Veldhoven, Esquenet, Heyns, & Verhoeven, 1996a, Swinnen et al., 2006), little is known

about the androgen regulation of lipid uptake.

In this study, the effect of androgen treatment on lipid content (free cholesterol,

neutral and phospholipids and lipid droplets) and lipid uptake of several lipid probes

(C5:0, C12:0 and C16:0 fatty acids, cholesterol, LDL and acetylated LDL) in a panel

of PCa cells was evaluated. By applying cutting-edge automated quantitative

fluorescence microscopy and image analysis, this work provides the functionally most

comprehensive analysis of lipid uptake in PCa cells to date. This work demonstrates

that R1881 and DHT significantly enhanced cellular uptake of LDL particles as well

as free fatty acids and cholesterol and their subcellular storage in lipid droplets (Fig.

3.2A-C). Consistent with this was a concordant increase in cellular phospholipids

(membrane), neutral lipids (cholesterol-FA esters and TAGs stored in lipid droplets)

and free cholesterol (Fig. 3.1A-B), which is a major component of cell membranes and

essential for membrane structure and functional organization as well as a precursor for

steroidogenesis, as reviewed in (Subczynski, Pasenkiewicz-Gierula, Widomska, Mainali, &

Raguz, 2017). While our work did not delineate the relative contributions of various

anabolic and catabolic lipid metabolism processes to the net increase in cellular lipid

content in response to androgen treatment, e.g. enhanced lipid uptake and lipogenesis

(Swinnen et al., 1996a) versus fatty acid oxidation, phospholipid degradation,

steroidogenesis and lipid efflux, we nevertheless demonstrated that R1881 and DHT

caused a strong increase in lipid uptake across various lipid species (fatty acids,

cholesterol, and lipoprotein complexes). Ongoing work suggests that lipid uptake has

a higher supply capacity than DNL in PCa which is consistent with the ability of

exogenous lipids to efficiently rescue DNL inhibition (Sadowski et al., 2014) and

recent work estimating that 70% of carbon lipid biomass is derived from exogenous

lipids in lung cancer cells showing the lipogenic phenotype (Hosios et al., 2016).

Critically, it was demonstrated here that androgen-enhanced lipid uptake is directly

mediated by AR signalling and independent of its stimulatory effect on cell cycle

progression and proliferation, (Heinlein & Chang, 2004; Lonergan & Tindall, 2011),


i.e. androgen-enhanced fatty acid and cholesterol uptake remained unaffected in PCa

cells arrested in G0/G1, S phase or G2/M and in the absence of cell proliferation. This

suggests that AR-regulated lipid uptake is maintained throughout the cell cycle, is not

part of a cell cycle specific AR subnetwork (McNair et al., 2017) and is not indirectly

caused by lipid biomass demand of daughter cell generation.

Importantly, this work for the first time comprehensively elucidated the lipid

transporter gene and protein landscape in prostate cancer. Recent integrative omics

studies of PCa patient samples highlighted a measurable degree of discordance

between genomics, epigenetics, transcriptomics and proteomics, i.e., that gene copy

number, DNA methylation and transcript levels did not reliably predict proteomic

changes (Iglesias-Gato et al., 2016; Latonen et al., 2018). In addition, the plasma

membrane localisation of most candidate lipid transporters remains to be confirmed in

PCa (Pinthus et al., 2007; Stump, Zhou, & Berk, 1993) despite recent progress in

overcoming technical limitations challenging the comprehensive delineation of the

surface proteome of PCa cells (Lee et al., 2018). By comparing transcriptomic and

proteomic analyses of cell lines, tumour xenografts and patient samples, this work has

conclusively demonstrated robust transcript levels of 34 lipid transporter genes in

multiple PCa cell lines and expression of six lipid transporter proteins in the membrane

fraction of LNCaP cells, of which plasma membrane expression was independently

confirmed for LDLR, GOT2, LRPAP1, LRP8 and SCARB2 in eight PCa cell lines

(Lee et al., 2018; Pinthus et al., 2007). Our analysis of androgen-regulated lipid

transporter genes through AR binding sites suggested direct AR-regulation of several

lipid transporters. Notably, additional lipid transporter genes might contain AR ChIP

peaks outside the cut-off of 25 kb. Alternatively, the absence of AR ChIP peaks might

indicate that they are indirectly regulated by androgen-activated transcription factors,

e.g. sterol element binding proteins 1 and 2 (SREPB1/2). Indeed, the LDLR gene lacks

AR ChIP peaks but contains flanking sterol regulatory elements and is positively

regulated by SREBP1/2 (Streicher et al., 1996; Yokoyama et al., 1993).

Data mining of previously reported PCa tumour proteomes (normal gland vs

primary PCa and primary PCa vs bone metastasis, (Iglesias-Gato et al., 2016; Iglesias-

Gato et al., 2018) demonstrated that the expression of the lipid transporter landscape

substantially changes during PCa progression from localised disease (21 lipid

transporters downregulated=low lipid uptake) to bone metastatic disease (16 lipid

transporters upregulated=high lipid uptake). For comparison, the enhanced expression


of lipogenic enzymes suggested that lipid synthesis was upregulated throughout PCa

progression from primary to metastatic disease. Oncomine analysis revealed a similar

increase in the transcript levels of lipid transporters LDLR and SCARB1 in three PCa

patient sample cohorts reported previously (Grasso 2012, Varambally 2005, La

Tulippe 2002) (Fig. 3.4D). If, and to what extent, the extremely lipid-rich environment

of the bone marrow (50-70% adiposity in adult men (Blebea et al., 2007)) is associated

with enhanced lipid uptake in PCa bone metastases remains to be investigated,

including the possibility that the increased incidence of PCa metastases to bone is linked

to high levels of adiposity and specific lipid species within bone marrow which provide

increased stimulus for more aggressive growth and pro-tumourgenic lipid signalling

of metastatic prostate cancer. Of the 22 bone metastasis proteomes that were analysed,

16 were from patients after long-term ADT and classified as CRPC, with one short-

term ADT and five hormone-naïve cases, yet all shared the same general features,

including enhanced lipid transport and fatty acid oxidation (Iglesias-Gato et al., 2018).

This suggests that castration-resistant PCa bone metastases rely on similar

mechanisms for growth as hormone-naïve metastatic bone tumours. Contrary to above

reports comparing metastatic and localised tumours (Fig. 3.4E), an integrated

transcriptomics and lipidomics study highlighted increased transcript levels of

SCARB1, GOT2 and SLC27A2, SLC27A4 and SLC27A5 as well as polyunsaturated

fatty acid (PUFA) accumulation in 20 paired localised tumours compared to matched

adjacent non-malignant prostate tissues (Li et al., 2016). While PUFA synthesis from

essential fatty acids α -linolenic acid and linoleic acids remained transcriptionally

unchanged between localised tumours and non-malignant prostate tissue, Li et al.

(2016) proposed that increased phospholipid uptake through SCARB1 caused

intratumoural PUFA enrichment in localised prostate cancer; however, this hypothesis

still awaits experimental confirmation. The reason for discordance between the

transcriptomic and proteomic analyses regarding lipid uptake in localised PCa is

unclear, but it is noteworthy that the activity of lipid transporters is also regulated

through changes in their subcellular localisation, highlighting the need for an

integrated analysis of the cell surface proteome and tumour lipidome in prostate

cancer. From the present study, it can be concluded that LDLR, GOT2, LRPAP1,

LRP8, SCARB1 and SCARB2 are high confidence lipid transporters (Fig. 3.4A-E)

that are associated with PCa disease progression and bone metastasis, but further work


is needed to fully delineate the lipid transporter proteome at the plasma membrane in

prostate cancer.

The present study provides the most comprehensive functional analysis of lipid

uptake in PCa cells to date and demonstrates that androgens strongly enhanced lipid

uptake of fatty acids, cholesterol and lipoprotein particles LDL and acetylated LDL in

AR-positive PCa cell lines (summarised in Fig. 3.7). Previous work indicated that

protein expression of GOT2/FABPpm is enhanced by androgens and increases the

cellular uptake of medium and long chain fatty acids in LNCaP and CWR22Rv1 PCa

cells (Pinthus et al., 2007). The comprehensive analyses of AR binding sites (ChIPseq

peaks), RNAseq (of three DHT-treated AR-positive PCa cell lines), qRT-PCR,

Western blot and cDNA microarray of LNCaP tumour xenografts (Locke et al., 2008)

revealed different lipid transporters are either activated and suppressed by androgens.

AR-negative malignant and non-malignant prostate cell lines (PC-3, Du145 and BPH-1) also

Figure 3.7 Androgen receptor regulates lipid uptake and lipogenesis

Schematic representation of cellular supply pathways of cholesterol and fatty acids in

PCa cells (transporter-mediated uptake, lipogenesis, passive diffusion, tunneling

nanotubes). Lipid transporters and lipogenic enzymes whose expression is increased

or decreased by androgens are highlighted in red and blue, respectively. Lipid

transporters without confirmed surface expression in PCa are marked by lighter shades

of red and blue. Only lipid transporters with confirmed transcript and protein

expression in cell lines and patient samples are shown. Schematic generated in

collaboration with Dr. Sadowski.


show expression of lipid transporters (Fig. 3.4A, C). Thus, it is likely that signalling pathways

other than the AR must regulate lipid supply in these cell lines. Furthermore, after using the

independently confirmed plasma membrane protein analysis performed by Lee et al.

(2018) as a high confidence filter, it can be concluded that LRPAP1 and SCARB2 are

androgen-suppressed and LRP8, SCARB1, LDLR and GOT2 are androgen-enhanced

surface lipid transporters in PCa cells. Interestingly, GOT2 (mitochondrial aspartate

aminotransferase) is better known for its role in amino acid metabolism, the

cytoplasm-mitochondria malate-aspartate shuttle, and the urea and tricarboxylic acid

cycles (Bradbury et al., 2011). This suggests additional functions of metabolic

enzymes in other subcellular compartments (Boukouris, Zervopoulos, & Michelakis,

2016), such as the plasma membrane (Pinthus et al., 2007; Stump et al., 1993) and

strikingly, with additional substrate specificities and catalytic activity (Bradbury et al.,

2011). Thus, there is a possibility that future studies will discover additional proteins

involved in lipid uptake due to their plasma membrane localisation.

Targeting cholesterol homeostasis in PCa as a therapeutic strategy to delay

development of CRPC has recently received increasing attention (Allott et al., 2018;

Gordon et al., 2016; Patel et al., 2018). Cholesterol is a precursor of steroid hormone

synthesis, and we previously showed that progression to CRPC is associated with

increased intratumoural steroidogenesis of androgens (Locke et al., 2008).

Hypercholesterolemia has been reported to enhance LNCaP tumour xenograft growth

and intratumoural androgen synthesis (Mostaghel, Solomon, Pelton, Freeman, &

Montgomery, 2012). Furthermore, targeting dietary cholesterol adsorption in the

intestine with ezetimibe (Allott et al., 2018) or de novo cholesterol synthesis with

simvastatin reduced LNCaP tumour xenograft growth and delayed development of

CRPC (Gordon et al., 2016). Targeting cholesterol uptake via SCARB1 antagonism

with ITX5061 reduced HDL (but not LDL) uptake in LNCaP, VCaP and CRW22Rv1

cells and sensitised CWR22Rv1 tumour orthografts to ADT (Patel et al., 2018).

Comparatively, the same study showed that ITX5061 conferred stronger growth

inhibition than simvastatin in LNCaP and CWR22Rv1 cells under hormone-deprived

conditions (Patel et al., 2018), suggesting that cholesterol uptake via SCARB1 is a

significant supply route in these PCa cell lines. Due to the detection of multiple

lipoprotein transporters (LDLR, VLDLR, SCARB1, LRP1-12) in conjunction with

increased cholesterol synthesis in PCa (Swinnen et al., 2006), novel co-targeting

strategies antagonising this cholesterol supply redundancy might have profound


synergies in extending the efficacy of ADT and delaying the development of CRPC.

Such co-treatment strategies could include simvastatin in combination with specific

inhibitors of lipid processing in the lysosome, which is a critical hub for lipid uptake

through endocytosis, including phagocytosis, pinocytosis and receptor-mediated

endocytosis (Schneider, 2016). The latter pathway is used by all major lipoprotein

receptors, including LDLR, VLDLR, SCARB1 and the LRPs 1-12 (Schneider, 2016),

and focus of a recently started Phase I/II clinical trial (NCT03513211). Strategies of

co-targeting lipid uptake and synthesis with repurposed drugs are currently under

investigation by our group and show very promising and potent anti-neoplastic

synergies in pre-clinical models of advanced prostate cancer.

Enhanced lipid uptake fuels the extensive remodelling of the PCa lipidome in response to androgen-targeted

therapies 91

Enhanced lipid uptake fuels the extensive

remodelling of the PCa lipidome

in response to androgen-targeted

therapies

4.1 INTRODUCTION

It is well-established that androgen signalling is fundamental to PCa growth,

and that suppressing the androgen axis results in tumour regression. Consequently,

AR-targeted therapies (ATT) remain the mainstay treatment for patients with advanced

PCa. Unfortunately, like many cancer types, acquired treatment resistance by PCa cells

ultimately results in relapse of the disease and its progression (described in section

1.2.5). While metabolic reprogramming is a well-described hallmark of cancer, less is

known about therapy induced metabolic alterations that help to facilitate cancer cell

survival and drive disease progression (Hanahan & Weinberg, 2011; Hendrich &

Michalak, 2003).

Previous data from a longitudinal PCa xenograft study generated by our

laboratory showed that several metabolic pathways are altered throughout tumour

progression to CRPC (Locke et al., 2009). These include arachidonic acid metabolism,

de novo steroidogenesis, and lipid metabolism, and tumours contained increased levels

of essential and non-essential fatty acids (Locke et al., 2009). Given that essential fatty

acids cannot be synthesized de novo and therefore must be acquired from uptake of

circulating lipids, this has led to the investigation of the role of exogenous lipids and

associated metabolic pathways as an early adaptive response to ATTs.

Interestingly, recent studies have shown that anti-cancer treatments activate

lipid metabolic networks that contribute to drug resistance in renal cell carcinoma (Lue

et al., 2017) and breast cancer (Hangauer et al., 2017; Vijayaraghavalu et al., 2012)

models. Furthermore, despite reduced cellular proliferation, pathways such as

phospholipid metabolism, lipid droplet formation and mitochondrial respiration (Lue

et al., 2017; Vijayaraghavalu et al., 2012) have also been shown to be increased in

cancer cells as an adaptive response to oncogenic signalling inhibition. While there is

no uniform pattern of lipid alteration across different types of cancer, increased


therapies 92

cholesterol and phospholipid levels have been reported in multidrug resistant-

compared to -sensitive leukemic, breast, and hepatoma tumour cells (Hendrich &

Michalak, 2003).

Given the critical role of lipids in cellular membrane remodelling (described in

section 1.3.2), it is plausible that the increased lipid content and associated membrane

remodelling contribute to drug resistance by altering membrane fluidity (thus altering

drug penetration), and cell signalling cascades (Corsetto et al., 2017; Hendrich &

Michalak, 2003). However, little is known about molecular mechanisms of therapy-

induced lipid remodelling and its contribution to therapy resistance. In the present

study, it was hypothesised that blocking growth and proliferative signals via inhibition

of the AR-axis induces rewiring of lipid metabolic networks in PCa cells that allow

for the progression to CRPC.

4.2 RESULTS

4.2.1 Characterisation of a long-term in vitro model of ATT

To investigate the early metabolic mechanisms of acquired treatment resistance

in PCa, an in vitro model to represent long-term ATT was developed. This study was

aimed at investigating the metabolic adaptations that occur beyond the acute response

to ATTs (48 hrs), however still preceding the complete development of Enz resistance,

which occurs in PCa cells following several months of Enz treatment when cells once

again begin to proliferate (Lu, Tsai, & Tsai, 1999; Xu et al., 2010; Yu et al., 2017).

PCa cells were grown for up to 21 days in either Charcoal Stripped Serum (CSS) to

represent androgen-deprivation therapy or in the presence of AR-antagonist

Enzalutamide (Enz, 10 µM) to isolate AR-regulated effects on the lipid metabolic

pathways assessed. The observed increase in AR transcript levels with increasing time

on ATT (Fig 4.1A) is consistent with the described progression to CRPC, whereby AR

protein expression and activity is reactivated in order to compensate for the lack of

systemic androgens (Heinlein & Chang, 2004; Risbridger et al., 2010). However, PSA

expression remained down throughout the time-course study, suggesting that AR-

regulated pathways are not reactivated. Additionally, cellular viability as measured by

reducing power, ATP production and mitochondrial membrane potential were

significantly reduced with Enz treatment (Fig 4.1A). Graphs for individual assays used

to generate Fig 4.1A can be found in the appendices (Fig. A1). Using the IncuCyte


therapies 93

live-cell imaging system, it was observed that LNCaP cells halt proliferation with ATT

(Fig 4.1B), albeit cell death increased only modestly (Fig 4.1A). As seen in Fig 4.1B,

ATT treated cells underwent vast morphological changes, characterised by elongation

and the formation of long, thin cell protrusions. qFM revealed no significant change

in overall cell area, however Enz treated cells had a significantly larger average

perimeter and major axis length (Fig 4.1C). Notably, all of the described Enz

responses were also observed in CSS treated cells, and the changes were often more

robust than those seen following Enz treatment. Charcoal stripping of serum depletes

a wide range of lipophilic molecules like steroid hormones (Sikora, Johnson, Lee, &

Oesterreich, 2016), making it a useful tool for studying the effects of androgen

deprivation. However, CSS also strips out numerous other small molecules such as

growth factors, hormones, and cytokines (Cao et al., 2009). While the initial

characterisation of this model utilised CSS as a valuable validation tool for observed

Enz responses, subsequent experiments were performed primarily in FBS+Enz

conditions to isolate AR-mediated effects.


therapies 94


therapies 95

4.2.2 Metabolic characterisation of the transcriptome in PCa cells in

response to ATTS

In order to investigate global changes to the PCa transcriptome in response to

ATTs, LNCaP cells were collected after up to 21 days Enz treatment and analysed

using a custom made 180K probe Agilent microarray (Agilent-027516 VPC Human

180K v2; GPL14873) (Sieh et al 2012). Gene set variation analysis (GSVA) showed

that the majority of unique differential expression changes (75.1%), identified with an

absolute fold change of >=1.5 and an adjusted p-value of <=0.05, occurred within 7

days of treatment, with a smaller proportion of additional changes (8.1%) occurring

between days 7 and 14 (Fig 4.2A). After day 14, no unique differentially expressed

genes were detected, suggesting that the altered phenotype induced by ATTs is

established within 7-14 days.

Gene set enrichment analysis (GSEA) performed by Dr. Ati Fard showed that

prolonged treatment with Enz induced changes in transcript levels of genes within

several lipid metabolic pathways in LNCaP cells (Fig 4.2B). This model revealed a

temporally dynamic response regarding the immediate and delayed responses, in

which three unique patterns emerged in response to ATTs: first, transcript that were

Figure 4.1 Long-term in vitro model to study the adaptive response of LNCaP cells

to treatment with ATTs

(A) LNCaP cells were treated for up to 21 days in FBS supplemented with Enz (10 µM)

or vehicle control (0.1% DMSO in FBS). qRT-PCR for AR and PSA transcript

expression, ATP production (Cell Titer-Glo®), cell viability (PrestoBlue) and

mitochondrial activity assays (qFM) were used to characterise the androgen response and

metabolic phenotype induced by long-term ATT, calculated as percent change relative

to vehicle control (n=3 independent experiments, mean±SD *p<0.05 **p<0.01

**p<0.001 ****p<0.0001). (B) Following 21 days of treatment with Enz or 0.1% DMSO

(vehicle), LNCaP cells were seeded in 96-well plates. IncuCyte live-cell confluence

imaging was used to assess proliferation rates and morphological changes for an

additional 96 hours (images representative of 3 independent experiments; scale bar=300

µm). (C) qFM was used to measure surface area (left), perimeter (middle) and major axis

length (right) of cells treated with Enz for 21 days or FBS+DMSO control (Student’s t-

test, n>1000 cells from two independent experiments, mean±SD, ****p<0.0001).

Figure 4.2 Transcriptomic profiling of LNCaP cells undergoing ATT

(A) LNCaP cells were treated for up to 21 days in FBS supplemented with AR antagonist

Enz (10 µM). Transcriptome profiling was performed using a custom made 180K probe

microarray (n=3 biological replicates). Gene set variation analysis shows the number of

unique differentially expressed genes between FBS+DMSO vs day 7 Enz, day 7 Enz vs

day 14 Enz, and day 14 Enz vs day 21 Enz. (B) Gene set enrichment analysis heatmap

showing pathway enrichment with increasing time ATT. Analysis and heatmap

generated by Dr. Fard.

Figure 4.2 Long-term in vitro model to study the adaptive response of

LNCaP cells to treatment with ATTs

(A) LNCaP cells were treated for up to 21 days in FBS supplemented with Enz (10 µM)

or vehicle control (0.1% DMSO in FBS). qRT-PCR for AR and PSA transcript

expression, ATP production (Cell Titer-Glo®), cell viability (PrestoBlue) and

mitochondrial activity assays (qFM) were used to characterise the androgen response and

metabolic phenotype induced by long-term ATT, calculated as percent change relative


therapies 96

downregulated and stayed downregulated; second, transcripts that were upregulated

and stayed upregulated; and third, transcripts that were up- or down-regulated initially

and then returned to pre-ATT expression levels over time. Genes involved in

lipogenesis, cell cycle, mitochondrial activity and oxidative phosphorylation showed

a consistent decrease in enrichment upon Enz treatment, validating the decrease in

proliferation and mitochondrial activity described in Fig 4.1A. However, there was a

consistent increased enrichment in genes involved in lipid transport activity,

lipoprotein metabolism and lipid remodelling pathways (Fig 4.2B). Cholesterol

homeostasis and lipid storage were initially reduced but showed increased enrichment

by day 21 Enz treatment. These temporal differences could be used to predict which

pathways contribute to PCa cells eventually overcoming Enz treatment and the

emergence of CRPC.


therapies 97

Figure 4.2 Transcriptomic profiling of LNCaP cells undergoing ATT

(A) LNCaP cells were treated for up to 21 days in FBS supplemented with AR

antagonist Enz (10 µM). Transcriptome profiling was performed using a custom

made 180K probe microarray (n=3 biological replicates). Gene set variation

analysis shows the number of unique differentially expressed genes between

FBS+DMSO vs day 7 Enz, day 7 Enz vs day 14 Enz, and day 14 Enz vs day 21

Enz. (B) Gene set enrichment analysis heatmap showing enrichment of genes

involved in indicated pathways with increasing time ATT. Samples are row-scale

normalised.


therapies 98

4.2.3 Investigation of the LNCaP lipidome in response to ATTs

Given the vast transcriptional changes in genes whose products participate in

lipid metabolic pathways that were observed with Enz treatment, ATT induced

alterations to the PCa lipidome were further investigated. Fatty acyl methyl ester

(FAME) extracts were analysed with a gas chromatograph coupled to a mass

spectrometer (GCMS). MetaboAnalyst (Chong & Xia, 2018) was used to perform

Sparse Partial Least Squares-Discriminant Analysis (sPLS-DA) in order to cluster

samples (Fig 4.3A) and to identify the top deregulated lipid species induced by Enz

treatment (Fig 4.3B). The sPLS-DA plot is a classification model that enables the

selection of the most discriminative features in the dataset to help group the samples

(Lê Cao, Boitard, & Besse, 2011). In Fig 4.3A, sPLS-DA showed clear clustering

between sample groups, where each of the Enz-treated time point groups cluster

distinctly from the FBS control group. Of the top 50 deregulated lipid species, 86% of

these were found to have increased abundance with increasing time of Enz treatment.

The 12 fatty acids detected by GCMS FAME analysis are shown in Table 4.1,

including the abundance of indicated fatty acids following 21 days Enzalutamide

treatment compared to vehicle control (0.1% DMSO). Included in these are essential

fatty acids and their metabolites, linoleic and arachidonic acid, further supporting the

hypothesis of increased lipid uptake in response to Enz. Arachidonic acid has

previously been implicated in PCa progression for its contribution to intratumoural

steroid synthesis and inflammatory pathways (Chaudry, Wahle, McClinton, & Moffat,

1994; Locke et al., 2010; Yang et al., 2012), however less is known about the other

detected fatty acid species in the context of ATT.


therapies 99

Figure 4.3 ATTs induce vast lipid remodelling in PCa cells

(A) LNCaP cells were grown for up to 21 days in Enz (10 µM) or 0.1% DMSO vehicle

control. Lipids were extracted and intact lipid content was quantified by LCMS. sPLS-

DA plot was generated with MetaboAnalyst (Chong & Xia, 2018) and shows clear

separation of Enz treated cells from FBS control. Red triangle=FBS; green cross=Enz d7;

blue x=Enz d14; teal diamond=Enz d21. (B) MetaboAnalyst was used to perform

hierarchical clustering analysis considering similarity measure (Pearson’s correlation),

and complete linkage clustering algorithm (Chong et al., 2018). Clustering analysis

helped to identify top 50 deregulated lipid species in Enz treated cells compared to

FBS+DMSO control (blue=decreased, red=increased; n=2 biological and 3 technical

replicates for lipidomics analysis, n=3 biological replicates for transcriptomic analysis;

clustering calculated by distance measure ‘correlation’ and clustering algorithm

‘complete’ (Chong et al., 2018)). Lipid analysis collected with the help of Dr. Poad and

Dr. Gupta.


therapies 100

4.2.4 Integrated ‘omics analysis of the early adaptive response to ATTs

Integration of the FAME-GCMS identified fatty acids and transcriptomics data

predicted significant enrichment in metabolic pathways that have been previously

associated with the progression to CRPC, such as arachidonic acid metabolism and

steroid biosynthesis (Locke et al., 2008; Locke et al., 2010). This analysis predicted

previously unexplored metabolic pathways in PCa, such as branched-chain amino acid

degradation, essential fatty acid metabolism, and sphingolipid metabolism (Fig 4.4)

(Xia, Mandal, Sinelnikov, Broadhurst, & Wishart, 2012). Given the complexity of

intact lipids, only free fatty acids identified by GMCS, which provide a metabolite ID,

were used to run the current pathway analyses. Intact lipids identified through LCMS

require further structural evaluation to identify specific IDs and were therefore not

included in this pathway analysis.

Table 4.1 Fatty acids detected by GCMS FAME in LNCaP cells

following Enzlutamide treatment













therapies 101

Isobaric mass tagging protein mass spectrometry was used to investigate the PCa

proteome in response to long-term Enz treatment. Here, LNCaP cells were treated for

21 days with Enz or DMSO control. Proteins from three biological replicates were

extracted, reduced and digested overnight before each sample was labeled with a

unique tandem mass tagging (TMT) isobaric label reagent. The six samples (three

replicates and two time points) were then combined and loaded in a HPLC CHIP

QTOF 6530 mass spectrometer for relative quantitative analysis.

Using the top 500 most abundant proteins identified here, a sPLS-DA plot was

generated, which shows notable separation of Enz treated cells compared to vehicle

control (Fig 4.5A). A volcano plot was then generated to show the top significant

proteins (p<0.05) with more than 2-fold change between D21 Enz vs vehicle control

groups (Fig 4.5B). These proteins were further divided into top 25 upregulated and top

25 downregulated proteins as shown in Fig 4.5C. Once the top deregulated proteins

were identified, this list was integrated with the top deregulated fatty acid list to

generate a new joint pathway enrichment analysis (Fig 4.5D). While the addition of

Figure 4.4 Integrated analysis of LNCaP transcriptome and lipidome

(A) Joint pathway analysis of integrated lipidomics and transcriptomics was performed

using MetaboAnalyst, which predicted enrichment of several metabolic pathways in Enz

treated cells compared to FBS control (n=2 biological and 3 technical replicates for

lipidomics analysis, n=3 biological replicates for transcriptomic analysis, p-values were

calculated by two-way ANOVA (Xia et al., 2012).


therapies 102

proteomics data revealed possible enrichment of additional metabolic pathways, the

integrated analysis confirmed enrichment in branched-chain amino acid degradation

as well as pathways involved in fatty acid metabolism. However, some pathways

shown in Fig 4.4 through integrated transcriptomics and lipidomics analysis were not

identified in the proteomics analysis, (i.e. retinal metabolism and arachidonic acid

metabolism). This could be due to lower abundance of proteins in these pathways, as

described previously, or a discordance between transcript and protein levels that this

work was not able to address. Despite these limitations, the integration of

transcriptomic, proteomic and lipidomic analysis highlighted the importance of lipid

remodelling as an early adaptive response to ATTs and suggested additional metabolic

pathways associated with PCa.


therapies 103


therapies 104

Figure 4.5 Proteomic analysis of Enz treated LNCaP cells

LNCaP cells were grown for 21 days in Enz (10 µM) or FBS+DMSO control. Proteins

were collected and labelled using TMT isobaric labelling reagents and samples were

analysed using an HPLC CHIP QTOF 6530 mass spectrometer. (A) sPLS-DA shows clear

separation of Enz treated cells compared to DMSO control. (red triangle=D0; green

cross=D21 Enz) (B) Volcano plot showing proteins with a fold change of >2 or <0.5 in

Enz treated cells compared to FBS control. Dark grey: p<0.05, light grey: p>0.5, red: fold

change >2 or <0.5. (C) Hierarchical clustering (described in Fig 4.3) showing top 25

significantly (p<0.05) upregulated and downregulated protein IDs (UniProt). For

clustering, red=D0, green=D21 Enz; for heatmap, blue=down, red=up. List of top 100

deregulated protein names can be found in Appendix AFig5. (D) Integration of proteomics

and transcriptomics using MetaboAnalyst pathway analysis function (described in Fig 4.4)

to identify top enriched pathways in Enz treated cells compared to DMSO control.

Pathways highlighted in blue are shared among lipidomics integrated analysis while

orange represents a unique pathway identified by proteomics (n=3 biological replicates).

Proteomics analysis and graphs generated with the help of Dr. Zang.

Figure 4.6 Increased lipid content is an adaptive response to ATT

(A) LNCaP cells were cells grown for up to 21 days in Enz (10 µM) or FBS+DMSO

control. Fixed cells were stained for 24 h with fluorescent lipid stain Nile Red (0.1 µg/ml)

and cellular mean fluorescent intensities (MFI) of neutral lipid content (left) and

phospholipid content (right) were measured by quantitative fluorescence microscopy

(qFM) (n>1000 cells from 3 independent experiments mean±SD, ****p<0.0001, One-

way ANOVA followed by Dunnett’s multiple comparisons test compared to FBS control).

(B) Images representative of phospholipids in (A). Blue=DAPI, Red=Nile Red

(phospholipids). (C) LNCaP cells were treated and stained as described above. Cellular

lipid droplet number (left) and mean total cellular area of lipid droplets (middle) were

measured by qFM. Perilipin was measured using qRT-PCR and shown as fold change

relative to control (right) (n>1000 cells from 3 independent experiments, mean±SD,

**p<0.01 ****p<0.0001, One-way ANOVA followed by Dunnett’s multiple comparisons

test compared to control.) (D) Following either Enz (10 µM) or CSS treatment for up to

21 days, cells were fixed as described above and stained with Filipin (50 µg/ml). Free

cholesterol was measured by qFM. (n>1000 cells from 2 independent experiments,

mean±SD, ****p<0.0001, One-way ANOVA followed by Dunnett’s multiple

comparisons test compared to FBS control).


therapies 105

4.2.5 Increased lipid content is an adaptive response to ATTs

Previous studies, including the work described above, have demonstrated that

R1881 and DHT strongly enhanced cellular lipid content of PCa cells through

enhanced lipogenesis and lipid uptake (Swinnen, Esquenet, et al., 1997; Swinnen et

al., 1996a; Tousignant et al., 2019). Here, quantitative fluorescent microscopy (qFM)

assays of two different fluorescent lipophilic dyes, Nile Red and Filipin (Sadowski et

al., 2014), were used to investigate changes in lipid content in the context of ATT.

Nile Red is able to distinguish between the phospholipids and neutral lipids

based on the degree of hydrophobicity of its environment, where Nile Red bound to

polar lipids, such as phospholipids found in membranes, have a red shifted

fluorescence, whereas Nile Red bound to non-polar neutral lipids stored in lipid

droplets have a strong yellow-orange shift (Diaz, Melis, Batetta, Angius, & Falchi,

2008; Greenspan, 1985). Using this approach, it was found that LNCaP cells treated

with Enz (10 µM) had significantly increased neutral lipid and phospholipid content

with increasing time of ATT (Fig 4.6A-B), measured as mean fluorescent intensity of

each lipid probe per cell. Further investigation of lipid droplet morphometry based on

Nile Red staining and measurement of neutral lipids revealed the lipid droplet number

and sum of area of all lipid droplets per cell also increased by 28% and 24%,

respectively, by day 21 of Enz treatment compared to FBS control (Fig 4.6C left and

middle). The perilipin (PLIN1-5) genes encode for a family of proteins known to

associate with the surface of lipid droplets, reviewed in (Sztalryd & Brasaemle, 2017).

qRT-PCR was used to show PLIN1 transcript levels increased up to 8.20-fold by day

21 of Enz treatment (Fig 4.6C right) compared to DMSO control, further supporting

the increased lipid droplet formation induced by ATT.


therapies 106


therapies 107

Next, free cholesterol content in LNCaP cells was measured using the

fluorescent probe Filipin (Fig 4.6D), which does not bind to cholesterol esters stored

in lipid droplets but selectively stains free cholesterol embedded in membranes

(Maxfield & Wüstner, 2012). LNCaP cells had significantly increased free,

unesterified cholesterol following both Enz (18% by day 21) or CSS (19% by day 21)

treatment, as compared to D0 control cells.

To obtain a more comprehensive understanding of the lipid profile of LNCaP

cells in response to ATTs, intact lipid species were further investigated using LCMS

lipidomics. Consistent with the Nile Red staining of phospholipids, it was found that

Enz treated cells had significantly increased levels of sphingomyelin,

phosphatidylcholine, phosphatidylethanolamine, phosphatidylserine and

phosphatidylglycerol with increasing time of ATT as compared to DMSO control (Fig

4.7A). However, inconsistent with the Nile Red neutral lipid staining that suggested



control. Fixed cells were stained for 24 h with fluorescent lipid stain Nile Red (0.1

µg/ml) and cellular mean fluorescent intensities (MFI) of neutral lipid content (left) and


(qFM) (n>1000 cells from 3 independent experiments mean±SD, ****p<0.0001, One-

way ANOVA followed by Dunnett’s multiple comparisons test compared to FBS

control). (B) Images representative of phospholipids in (A). Blue=DAPI, Red=Nile Red

(phospholipids). Scale bar=100 µm. (C) LNCaP cells were treated and stained as

described above. Cellular lipid droplet number (left) and total cellular area of lipid

droplets (middle) were measured by qFM. Perilipin was measured using qRT-PCR and

shown as fold change relative to D0 control (right) (n>1000 cells from 3 independent

experiments, mean±SD, **p<0.01 ****p<0.0001, One-way ANOVA followed by

Dunnett’s multiple comparisons test compared to control.) (D) Following treatment

with Enz (10 µM), CSS, or 0.1% DMSO for up to 21 days, cells were fixed as described

above and stained with Filipin (50 µg/ml). Free cholesterol was measured by qFM.

(n>1000 cells from 2 independent experiments, mean±SD, ****p<0.0001, One-way

ANOVA followed by Dunnett’s multiple comparisons test compared to FBS control).

Figure 4.7 Lipidomics analysis

Following up to 21 days Enz (10 µM) treatment, lipids were extracted, and lipid content

was quantified using LCMS. (A) Total combined lipid classes and (B) individual

cholesterol ester and triacylglycerol species are shown as fold change relative to DMSO

control (n=2 biological and 3 technical replicates, mean±SD, * p<0.05 **p<0.01

***p<0.01 ****p<0.0001, Two-way ANOVA followed by Tukey’s multiple

comparisons test). SM: sphingomyelin; PC: phosphatidylcholine; PE:

phosphatidylethanolamine; PS: phosphatidylserine; PG: phosphatidylglycerol; CE:

cholesteryl ester; TAG: triacylglycerol. LCMS data and analysis generated with the

help of Dr. Poad.



control. Fixed cells were stained for 24 h with fluorescent lipid stain Nile Red (0.1

µg/ml) and cellular mean fluorescent intensities (MFI) of neutral lipid content (left) and



therapies 108

increased lipid droplets, both cholesterol ester and triacylglycerol stores were

significantly depleted by Enz treatment in LNCaP cells (Fig 4.7B). This suggests that

additional lipid species that were not measured by LCMS were contributing to the

increased lipid droplet content detected by Nile Red. Overall, this data showed that

Enz treatment resulted in an increased abundance of major lipid classes in LNCaP

cells.

Figure 4.7 Lipidomics analysis

Following up to 21 days Enz (10 µM) treatment, lipids were extracted, and lipid content

was quantified using LCMS. (A) Total combined lipid classes and (B) individual cholesterol

ester and triacylglycerol species are shown as fold change relative to DMSO control (n=2

biological and 3 technical replicates, mean±SD, * p<0.05 **p<0.01 ***p<0.01

****p<0.0001, Two-way ANOVA followed by Tukey’s multiple comparisons test). SM:

sphingomyelin; PC: phosphatidylcholine; PE: phosphatidylethanolamine; PS:

phosphatidylserine; PG: phosphatidylglycerol; CE: cholesteryl ester; TAG: triacylglycerol.

LCMS data and analysis generated with the help of Dr. Poad.


therapies 109

4.2.6 Increased lipid uptake fuels the lipid rich phenotype induced by ATTs

The observed increased lipid content in the present ATT model could come from

two sources: de novo synthesis or exogenous uptake. Given the relatively poor

understanding of exogenous uptake in PCa, both supply routes were investigated. First,

lipid composition of cell culture media (100% FBS and CSS) was analysed for fatty

acid content by GCMS FAME to confirm the availability of fatty acids in culture

conditions (Fig 4.8A), and cholesterol and TAG compositions were acquired from the

serum supplier (Fig 4.8B).

Figure 4.8 Analysis of lipid composition in cell culture media

(A) GCMS FAME analysis of fatty acid content in FBS and CSS. Analysis provided

by Dr. Gupta and graph generated by Dr. Sadowski. (B) Extract from Sigma-Aldrich

analysis showing levels of cholesterol and triacylglycerides (TAGs) in FBS and CSS.


therapies 110

To measure if lipid uptake was increased by Enz, a series of lipid uptake assays

was used based on qFM of fluorophore labelled lipid probes (Bodipy-C16:0, NBD-

cholesterol, DiI-LDL, and DiI-acetylated LDL) coupled with quantitative image

analysis. Interestingly, fatty acid uptake decreased by 55% at day 21 of Enz treatment

and 58% at day 21 of CSS treatment compared to FBS control in LNCaP cells (Fig

4.9A). This was surprising, given the significant increase in total lipid content

described in Sections 4.2.3 and 4.2.5. There is vast diversity of lipids in circulation;

free fatty acids make up roughly 4% of the fatty acid pool, while the majority are

contained in more complex lipids such as phospholipids, lysophospholipids,

triacylglycerols and lipoprotein complexes (Doege & Stahl, 2006). Therefore, the

uptake of more complex lipid species was investigated in the context of ATT. qFM of

LNCaP cells incubated with NBD-cholesterol showed significantly increased free

cholesterol uptake up to 57% higher than baseline levels (D0) by day 21 of Enz

treatment (Fig 4.9B), supporting the increased free cholesterol content of cells shown

in Fig 4.6D.

Another major source of exogenous lipid supply is the uptake of lipoprotein

particles that deliver a complex combination of apolipoproteins, cholesterol,

triacylglycerols, and phospholipids to cells via receptor-mediated endocytosis by

lipoprotein receptors such as the low-density lipoprotein receptor (LDLR) and

scavenger receptor SCARB1 (Goldstein et al., 1982; Ikonen, 2008). qFM of LDL or

acetylated LDL (acLDL) complexed with 1,1-Dioctadecyl-3,3,3’,3’-

tertramethylindocarbocyanine perchlorate (DiI) showed a 15% increase from baseline

levels in LDL uptake and a 51% increase from baseline levels in acLDL uptake by day

21 of Enz treatment compared to FBS control cells (Fig 4.9B). Cellular uptake of

NBD-PE was also higher in LNCaP cells treated with Enz for 14 days compared to

control cells (Fig 4.9C). One caveat of the fluorophore labelling methodologies

described here is that there are technical limitations to the number of lipid species that

can be investigated. There is likely differential efficiency in uptake of lipid species

across different cell types, and this is an area the demands more attention. Nonetheless,

the lipids investigated by this approach were complemented with other analysis

methods (lipid mass spectrometry). Additionally, media were replaced every 2-3 days

in order to avoid effects of spent medium on cells.


therapies 111

These data suggest that PCa cells were acquiring fatty acids via lipoprotein

particles rather than through free fatty acids, contributing to the total increased lipid

pool. The GCMS FAME analysis shows the total fatty acid content of LNCaP cells

undergoing Enz treatment. Here, a significant increase in the total cellular C16:0 and

C18:0 fatty acid content as well as increased essential fatty acids and their metabolites

(Fig 4.9D, 4.9E) was observed. Because essential fatty acids cannot be synthesised

endogenously and must come from exogenous dietary sources (section 1.3.1), these

data indicate that increased lipid uptake is a cellular adaptive response to ATT.


therapies 112


therapies 113

Genes involved in de novo lipogenesis are well characterised in many cancers,

including prostate, and their overexpression is associated with tumour development

and disease progression (reviewed in (Menendez & Lupu, 2007, Galbraith et al.,

(2018)). Furthermore, our work, and previous reports, have shown that androgens

regulate lipogenesis in PCa cells, but the functional role of lipid transporters in the

context of ATT remains poorly characterised (Swinnen, et al., 1997). To further

interrogate the lipid transporter landscape in response to Enz treatment, qRT-PCR

analysis of several well described lipid transporter genes (section 1.3.3) was

performed. Here it was found that several lipid transporters have increased transcript

levels in LNCaP (Fig 4.10A) and C42B (data not shown) cells undergoing Enz

treatment. This was validated in an LNCaP tumour xenograft model of CRPC, in

which majority of these transporter genes are minimally expressed in tumour samples

one week post castration (PostCX) but have increased expression in recurring tumours

Figure 4.9 Enhanced lipid uptake in response to ATT

(A) LNCaP cells were treated for up to 21 days with Enz (10 µM), CSS, or FBS+DMSO

control. Before fixation, cells were incubated with Bodipy-C16:0 for one hour and lipid

uptake was measured as mean fluorescent intensity (MFI; calculated as intensity per

pixel/cellular area) by qFM (n>1000 cells, representative of 3 independent experiments,

mean±SD, ***p<0.001, One-way ANOVA followed by Dunnett’s multiple

comparisons test compared to DMSO control). (B) Before fixation, NBD-PE was added

to conditioned media for one hour and lipid uptake was measured by qFM. (n>3000

cells from 3 wells, representative of 2 independent experiments, mean±SD,

****p<0.0001) (C) Before fixation, media were removed and cells were incubated with

NBD-cholesterol, DiI-LDL or acetylated-LDL for two hours and lipid uptake was

measured by qFM. (n>1000 cells representative of 3 independent experiments *p<0.05

***p<0.001, One-way ANOVA followed by Dunnett’s multiple comparisons test

compared to DMSO control). (D) Quantitative lipid profiling of free fatty acid content,

including essential fatty acids (E), of Enz treated LNCaP cells was measured using

GCMS (Values represent fold change relative to FBS control, n=2 biological replicates

and 3 technical replicates, *p<0.05 **p<0.01 ***p<0.001Two-way ANOVA followed

by Tukey’s multiple comparisons test). GCMS FAME analysis generated with the help

of Dr. Gupta.

Figure 4.9 Enhanced lipid uptake in response to ATT

(A) LNCaP cells were treated for up to 21 days with Enz (10 µM), CSS, or FBS+DMSO

control. Before fixation, cells were incubated with Bodipy-C16:0 for one hour and lipid

uptake was measured as mean fluorescent intensity (MFI) by qFM (n>1000 cells,

representative of 3 independent experiments, mean±SD, ***p<0.001, One-way

ANOVA followed by Dunnett’s multiple comparisons test compared to DMSO control).

(B) Before fixation, NBD-PE was added to conditioned media for one hour and lipid

uptake was measured by qFM. (n>3000 cells from 3 wells, representative of 2

independent experiments, mean±SD, ****p<0.0001) (C) Before fixation, media were

removed and cells were incubated with NBD-cholesterol, DiI-LDL or acetylated-LDL

for two hours and lipid uptake was measured by qFM. (n>1000 cells representative of

3 independent experiments *p<0.05 ***p<0.001, One-way ANOVA followed by

Dunnett’s multiple comparisons test compared to DMSO control). (D) Quantitative

lipid profiling of free fatty acid content, including essential fatty acids (E), of Enz treated

LNCaP cells was measured using GCMS (Values represent fold change relative to FBS


therapies 114

treated with Enz and those that had developed to CRPC (Fig 4.10B). In this model,

surgical castration of mice (ADT) initially results in tumour regression until it reaches

nadir, the point of lowest PSA or smallest tumour volume before tumours begin to

recur. Tumours were then collected following biochemical recurrence (CRPC), or

mice were treated with Enzalutamide and tumours were collected once the ethical

endpoint was reached.

Driven by the observed increased uptake of lipoproteins in the present ATT model

(Fig. 4.9B), levels of lipoprotein transporters LDLR and SCARB1 were investigated,

both of which have been previously associated with PCa before, but with conflicting

conclusions regarding their pro- or anti- tumourigenic activity (Furuya et al., 2016;

Schörghofer et al., 2015b). LDLR and SCARB1 protein levels were increased by 35%

and 9.5%, respectively, at day 21 Enz treatment compared to FBS control (Fig 4.10C),

suggesting that this route of protein mediated lipid uptake may contribute to the

increased cellular lipid pool observed following long-term ATT treatment.


therapies 115

Figure 4.10 Enz-induced upregulation of transcripts encoding lipid transporters

(A) Expression of lipid transporters was measured by qRT-PCR (n=3, mean±SD, *p<0.05,

One-way ANOVA followed by Dunnett’s multiple comparisons test compared to D0 control;

heatmap represents fold change relative to D0). (B) Transcript levels of indicated lipid

transporters in an LNCaP xenograft model of CRPC progression was analysed by RNAseq and

heatmaps were generated with a hierarchical clustering algorithm using completed linkage and

Euclidean distance measures and scaled by row z-score (red=positive z score, blue=negative z

score). PostCX=one-week post castration; CRPC=tumours collected following biochemical

recurrence; Enz=following castration, mice were treated with Enz and tumours were collected

once the ethical endpoint was reached. (C) Protein expression of selected lipid transporters

LDLR and SCARB1 was measured using immunofluorescence microscopy (Section 2.8) in

Enz (10 µM) treated cells (neg control: secondary antibody only to control for background;

n>1000 cells representative of 2 independent experiments, mean±SD, ****p<0.0001, One-way

ANOVA followed by Dunnett’s multiple comparisons test compared to D0 control).


therapies 116

4.2.7 Lipogenesis is downregulated by Enzalutamide

To directly address the contribution of de novo lipogenesis, qRT-PCR and

protein mass spectrometry analysis were used to investigate changes in major DNL

enzymes in the long-term ATT model. FASN, the main enzyme involved in the

synthesis of C16:0 (palmitic acid) and a well-characterised enzyme in PCa (Swinnen

et al., 2002a; Swinnen et al., 2000), had a significant decrease in transcript expression

in LNCaP cells treated for 21 days with Enz compared to DMSO control (Fig 4.11A).

This was accompanied by a decrease in transcript levels of HMGCS1, a major enzyme

involved in the biosynthesis of cholesterol. Protein levels of SREBF2, a major

regulator in cholesterol synthesis, and ACLY, as well as transcript expression of

lipogenic enzymes ACACA and ACAT2 were also reduced with Enz treatment, while

expression of ACAT1 was increased (Fig 4.11A).

Figure 4.11 De novo lipogenesis decreases in the early adaptive response to ATT

(A) Transcript levels (grey bars) of selected DNL genes was measured using a custom

made 180K microarray. Protein (blue bars) was measured by mass spectrometry. Values

are shown as fold change at day 21 Enz treatment relative to FBS+DMSO control. (B)

Following 14 days Enz treatment, LNCaP and C42B cells were incubated for 2 hours

with Di-2DG and cellular glucose uptake was measured by qFM (n>1000 cells

representative of 2 independent experiments, mean±SD, ***p<0.001, One-way

ANOVA followed by Dunnett’s multiple comparisons test compared to FBS control).

(C) Following up to 21 days Enz (10 µM) treatment, cells were incubated with 13C-

acetate for 72 h. Metabolites were extracted and 13C-acetate incorporation into

cholesterol was measured by GCMS. (Values represent fold change relative to FBS

control, n=3, ****p<0.0001 One-way ANOVA followed by Dunnett’s multiple

comparisons test).

Figure 4.12 Fatty acid remodelling contributes to the adaptive response of PCa

cells to ATTs


therapies 117

Furthermore, qFM showed that glucose uptake, which is the predominant source of

carbon for DNL (Brusselmans & Swinnen, 2009), decreased in both LNCaP and C42B

cells following 14 days of Enz treatment (Fig 4.11B). To investigate the contribution

of de novo synthesized cholesterol to the total cellular cholesterol pool, LNCaP cells

were incubated with 13C-acetate for the last 72 hours of ATT treatment and

incorporation into endogenous cholesterol was measured by mass spectrometry.

Metabolomic analysis confirmed that de novo cholesterol synthesis significantly

decreased by 71% compared to FBS vehicle control cells by day 21 Enz treatment (Fig

4.11C). This approach was also used to investigate endogenous fatty acid synthesis,

however the incorporation of 13C-acetate into palmitate, oleate and stearate was

negligible, and no significant incorporation of 13C-glucose into lipids was measured

(data not shown). Taken together, these data suggest that uptake of exogenous lipids,

rather than DNL, is the major contributing source to the increased cellular lipid pool

measured in response to Enz treatment, thus highlighting a significant role for lipid

uptake in PCa progression.

4.2.8 Enzalutamide treatment induces lipid remodelling including fatty acid

elongation and desaturation

Integrative lipidomic and transcriptomic analysis of the in vitro ATT model

revealed enrichment of lipid membrane remodelling pathways in Enz treated cells (Fig

4.4A). While total cellular PE levels were significantly increased by 21 days

Enzalutamide treatment (Fig 4.7A), there was considerable variation between the lipid

species examined within that group. Shorter chain PE 34:2 and PE 36:2 lipids were

significantly decreased in response to ATTs, while PE 38:3, 38:4, 38:5 and 38:6 were

significantly increased by day 21 Enz treatment compared to FBS+DMSO control cells

(Fig 4.12A). A similar increase in longer chain polyunsaturated fatty acids was also

seen in phosphatidylglycerol, phosphatidylserine, phosphatidylcholine, and

sphingomyelin (PG, PS, PC, and SM) lipid classes (Fig 4.12B). Given this phenotypic

observation, gene expression of elongase and desaturase enzymes in Enz treated cells

was investigated (Fig 4.12C). While ELOVL4, DEGS2 and SCD5 had increased gene

expression, the majority of elongase and desaturase genes had decreased expression,

supporting the hypothesis that Enz treated cells are acquiring these longer chain

PUFAs predominantly from uptake of exogenous lipids.


therapies 118


therapies 119

Figure 4.12 Fatty acid remodelling contributes to the adaptive response of PCa cells to

ATTs

(A) Following 21 days Enz (10 µM) treatment, lipids were extracted, and lipid content was

quantified using GMCS. PE species were analysed using LipidView (Values represent fold

change relative to FBS control, n=2 biological and 3 technical replicates, ****p<0.0001,

Two-way ANOVA followed by Tukey’s multiple comparisons test). (B) PC, PS, PG, and

SM data were collected and analysed as described in (A). (C) Microarray analysis was used

to determine changes in transcript levels for key lipid remodelling enzymes following up to

21 days Enz treatment (bar represents average fold change relative to FBS+DMSO control

from 3 replicates). (D) Expression levels of major desaturase (left) and elongase (right)

enzymes in an LNCaP tumour xenograft progression model (reg=regressing, nad=nadir,

CRPC=castrate resistant PCa, Rec=recurring) was measured by microarray. Heatmaps were

generated with a hierarchical clustering algorithm (heatmap.2, Section 2.13) using

completed linkage and Euclidean distance; color refers to normalised row z score

(red=positive z score, blue=negative z score). (E) Expression levels of major desaturases and

elongases in primary (P) vs metastatic (M) patient samples (Grasso et al., 2012) (unpaired t-

test between primary and metastatic tissue; *p<0.01; ****p<0.0001)

Figure 4.12 Fatty acid remodelling contributes to the adaptive response of PCa cells to

ATTs


therapies 120

To validate lipid remodelling effects in vivo, the expression of desaturases (Fig

4.12D left) and elongases (Fig 4.12D right) in an LNCaP tumour xenograft model of

PCa progression (Locke et al., 2010) was investigated. In both gene sets, regressing

and nadir tumour samples had lower levels of desaturase and elongase genes compared

to recurring and CRPC tumours. This is consistent with the present long-term ATT

model described in this study in that cells in a state of quiescence, prior to tumour

recurrence, likely acquired lipids through uptake of exogenous sources rather than

through the remodelling of endogenous lipids. Further evidence of lipid remodelling

was observed in the analysis of the publicly available Grasso patient dataset (Grasso

et al., 2012) in which it was found that transcript expression of several elongases and

desaturases are highly upregulated in metastatic vs primary tumours (Fig. 4.12E).


therapies 121

4.2.9 ATT-induced lipid remodelling makes PCa cells more susceptible to

lipid peroxidation

Lipid peroxidation is a process in which oxidants such as free radicals or reactive

oxygen species (ROS) attack carbon-carbon double bonds in lipids (reviewed in

(Ayala, Muñoz, & Argüelles, 2014)), resulting in damage to the lipid-rich cell

membrane. PUFAs are especially susceptible to lipid peroxidation due to their multiple

double bonds. It was hypothesised that the increased PUFA content in Enz treated cells

would make them more susceptible to oxidative damage by ROS. To address this, qFM

of a fluorescent ratio-probe C11-Bodipy (581/591) was used in which unoxidised

lipids fluoresce in the red channel and oxidised lipids fluoresce in the green channel

(Drummen, van Liebergen, Op den Kamp, & Post, 2002). Indeed, qFM analysis

revealed that baseline levels of lipid peroxidation were significantly higher in both

LNCaP and C42B cells following 14 days Enz treatment as compared to FBS+DMSO

control (Fig 4.13A). Supporting this, glutathione peroxidase 4 (GPX4) and glutamate

cysteine ligase (GCL), genes encoding two enzymes involved in antioxidant defence

systems, were increased in Enz treated cells (Fig 4.13B). To investigate the sensitivity

of ATT cells to ROS, PCa cells were pre-treated for 14 days with Enz and then treated

for 24 hrs with the ROS inducer RSL3 (1 µM). A significant increase in cell death

measured by propidium iodide staining was observed in Enz pre-treated cells

compared to control cells following RSL3 treatment (Fig 4.13C).


therapies 122

Figure 4.13 RSL3 sensitivity

Following 14 days Enz treatment, levels of lipid peroxidation in LNCaP and C42B

cells were measured using a C11-Bodipy fluorophore probe and analysed by qFM

(mean±SD of n>1000 cells representative of 2 independent experiments,

***p<0.001, One-way ANOVA followed by Dunnett’s multiple comparisons test

compared to FBS+DMSO control). (B) Microarray analysis was used to analyse

GPX4 and GCL transcript levels following Enz treatment (n=3, **p<0.01,

****p<0.0001). Graph shows average fold change relative to FBS+DMSO control.

(C) Percent cell death following 24 hr RSL3 (1 µM) or DMSO (0.1%) treatment was

measured using propidium iodide staining in LNCaP and C42B cells following 14

days Enz treatment (n>1000 cells representative of 2 independent experiments,

mean±SD, ***p<0.001, One-way ANOVA followed by Dunnett’s multiple

comparisons test compared to FBS control).

Figure 4.13 RSL3 sensitivity

Following 14 days Enz treatment, levels of lipid peroxidation in LNCaP and C42B


therapies 123

4.2.10 PLA2G2A expression is a major contributor to ATT-induced

lipid remodelling in PCa cells

Microarray analysis of Enz treated LNCaP cells predicted changes to phospholipid

metabolism pathways, including an increased expression of phospholipase group IIA

(PLA2G2A) transcript, an enzyme that hydrolyses the sn-2 ester bond in phospholipids

found in lipoproteins and cell membranes (reviewed in (Brglez, Lambeau, & Petan,

2014; Makoto Murakami & Lambeau, 2013). The microarray data were validated by

qRT-PCR across 5 independent PCa cell lines in which 14 days Enz treatment

consistently upregulated PLA2G2A expression levels by between 3-200 fold

compared to FBS controls (Fig 4.14A). Surprisingly, Western blot analysis of

corresponding LNCaP cell lysates showed that PLA2G2A protein decreased with Enz

treatment (Fig 4.14B). Given that the enzymatic activity of PLA2G2A occurs in the

extracellular space (Mounier et al., 2004; Murakami et al., 1998), PLA2G2A protein

levels were investigated by ELISA (Cayman Chemical) of the conditioned media from

Enz treated cells. A significant increase in PLA2G2A protein was measured in the

conditioned media of Enz treated cells compared to media corresponding to FBS

control across all 5 PCa cell lines tested (Fig 4.14C). Together with decreased

PLA2G2A in cell lysates, Enz treatment may increase PLA2G2A transcript levels and

protein production and activate PLA2G2A secretion. Further investigation into the

secretory mechanisms of PLA2G2A would be of extreme value to help in our

understanding of the androgen regulation of this process.

PLA2G2A is a lipolytic enzyme that releases free FA (primarily arachidonic acid)

and lysophospholipids by catalysing hydrolysis of the sn-2 ester bond (Murakami &

Lambeau, 2013). The increase in PLA2G2A transcript levels and protein expression by

LNCaP cells was accompanied by a significant increase in cellular arachidonic acid as

measured by GCMS (Fig 4.14D). RNA sequencing of LNCaP xenograft tumours show

a significant increase in PLA2G2A transcript levels in castrated tumours treated with

Enz compared to sham-castrated controls (Fig 4.14E). In this model, intact tumours

were collected from mice that were not surgically castrated (ADT), while Enz tumours

were collected from mice that were castrated and had tumour regression to nadir

(lowest detected PSA levels), followed by tumour recurrence (tumour volume and PSA

recurrence) and Enz treatment. The increase in PLA2G2A in Enz treated tumours


therapies 124

compared to nadir supported the observed in vitro activation of PLA2G2A with ATTs

(Fig 4.14A).


therapies 125

Figure 4.14 Enz induces expression of PLA2G2A in PCa cells

Following 14 days Enz treatment, PLA2G2A was measured by qRT-PCR in 5 PCa cell lines

and shown as fold change relative to FBS control (n=3, mean±SD, *p<0.05 ***p<0.001

****p<0.0001, One-way ANOVA followed by Dunnett’s multiple comparisons test). (B)

PLA2G2A was measured by Western Blot in LNCaP cell lysates following up to 21 days Enz

treatment. Image representative of 3 independent experiments. (C) Secreted PLA2G2A in

conditioned media from 5 PCa cell lines treated for 14 days with Enz or FBS+DMSO control

was measured with an ELISA assay (Cayman Chemical; n=3, mean±SD, *p<0.05, One-way

ANOVA followed by Dunnett’s multiple comparisons test). (D) Schematic showing

PLA2G2A activity. LCMS and GCMS were used to measure lysophospholipid and AA

content, respectively, in LNCaP cells after 21 days Enz treatment. Values represent fold

change relative to FBS+DMSO control (n=2, **p<0.01 One-way ANOVA followed by

Dunnett’s multiple comparisons test). (E) PLA2G2A was measured by RNA-sequencing in

intact (sham-castrated) vs Enz treated LNCaP xenograft tumours (n=9 intact, n=26 Enz,

*p<0.05 unpaired t-test).


therapies 126

4.2.11 Phospholipase PLA2G2A facilitates lipid uptake and remodelling

in PCa cells

Given the significant increase in PLA2G2A transcript levels and protein secretion

by PCa cells in response to Enz, it was important to determine the contribution of the

enzymatic activity to therapy-induced phenotypic changes observed in the long-term

adaptive response model described in this study. It was hypothesised that PCa cells

secrete PLA2G2A into the extracellular space to hydrolyse phospholipids and provide

lysophospholipids and PUFAs for uptake by cells. For this experiment, two fluorescent

PLA2G2A substrates were used. First, a nitrobenzoxadiazole (NBD)-fluorophore was

employed that is bound to the headgroup of a PE and fluoresces in the green channel

(463/536 nm) (Fig 4.15A). In this case, intact PEs and hydrolysed lyso-PEs taken up

by the cells could be detected by qFM. The second probe, Red/Green BODPIY® PC-

A2, allows for dual emission fluorescence ratio detection where cleavage of the

BODIPY® FL pentanoic acid at the sn-2 position results in decreased quenching to the

BODIPY 558/568 dye attached to the sn-1 position (Fig 4.15B), therefore increasing

fluorescence in the green channel (530 nm) with a reciprocal decrease in the red

channel (590 nm) upon cleaving by sPLA2s. In this method, red signal detects

uncleaved PCs, whereas green signal detects sPLA2-cleaved lyso-PCs.

Figure 4.15 Schematic of fluorescent PLA2G2A substrates

(A) Following incubation with (22-(N-(7-Nitrobenz-2-Oxa-1,3-Diazol-4-yl) Amino-23,24-

Bisnor-5-Cholen- 3β-Ol)-phosphatidylethanolamine (NBD-PE), cellular uptake can be

measured by qFM. (B) When uncleaved, BODIPY C5-Sn-Glycero-3-Phosphocholine

(PC-A2) taken up by the cells will fluoresce in the red channel (590 nm). However,

cleavage by PLA2G2A results in a quenching of the red fluorescence and reciprocal

increase in the green channel (530 nm)

Figure 4.15 Schematic of fluorescent PLA2G2A substrates

(A) (22-(N-(7-Nitrobenz-2-Oxa-1,3-Diazol-4-yl) Amino-23,24-Bisnor-5-Cholen- 3β-Ol)-

phosphatidylethanolamine (NBD-PE) and (B) BODIPY C5-Sn-Glycero-3-

Phosphocholine (PC-A2) were employed to investigate PLA2G2A activity in PCa cells.


therapies 127

In both LNCaP and C42B cell lines, cells incubated in 96-well plates (roughly 1000

cells/well) with human recombinant PLA2G2A (50 ng) (hrPLA2G2) for one hour had

significantly higher NBD-PE uptake than control cells, and this effect could be blocked

using 30 µM KH064, a potent inhibitor of PLA2G2A (Reid, 2005) (Fig 4.16A-B). To

test that lysolipid uptake is mediated by PLA2G2A activity and enhanced cleavage of

phospholipids, a second fluorescent probe (PC-A2) was utilised, which undergoes a

shift in fluorescent emission upon cleavage of fatty acids at the sn-2 position of the PC

substrate, thus releasing lyso-PCs. Using a fluorometric analysis assay, a significant

increase in cleavage of PC-A2 was observed in the media with the addition of 50 ng

hrPLA2G2A protein (Fig 4.16C). Notably, there was a significantly higher increase in

fluorescent intensity following hrPLA2G2A addition in Enz treated cells compared to

DMSO controls, suggesting that the previously described increased PLA2G2A levels

in media also contributed to the accumulation of PLA2G2A products observed in Fig.

4.16C. Following the one-hour incubation with hrPLA2G2A and PC-A2, media were

removed, and cells were imaged in order to measure uptake of cleaved phospholipids.

The increased PC cleavage described above resulted in increased cellular detection of

these lyso-PCs as baseline levels of lyso-PC uptake were higher in Enz treated LNCaP

cells and significantly higher in Enz treated C42B cells compared to DMSO controls

(Fig 4.16D). 30 µM KH064 reduced PLA2G2A mediated PC cleavage and subsequent

cellular lyso-PC uptake in both LNCaP and C42B cells (Fig 4.16E). Importantly, the

higher baseline lyso-PC uptake in Enz treated cells compared to FBS+DMSO controls

in both cell lines tested (Fig. 4.16D) suggests they have undergone the transcriptional

changes to facilitate lysolipid uptake. These include increased transcript levels of

genes encoding lysolipid transporters such as the P4-ATPases. This was investigated

by transcriptomic analysis of Enz treated LNCaP cells (Fig 4.16F).


therapies 128


therapies 129

Figure 4.16 Exogenous PLA2G2A promotes lysolipid uptake in Enz treated PCa cells

(A) Following 14 days treatment with Enz or FBS+DMSO control, media were removed and

cells were incubated with fluorogenic phosphoethanolamine (NBD-PE) with the addition of 30

µM KH064 and 50 ng hrPLA2G2A, alone or in combination, or 0.1% DMSO for one hour and

lipid uptake was measured by qFM (mean±SD of n>1000 cells from 2 independent experiments,

** p<0.01 ****p<0.0001, One-way ANOVA followed by Dunnett’s multiple comparisons test

compared to FBS control). Staining procedure described in Section 2.4 and 2.18. (B) Images

representative of (A). (C) Enz-treated cells were prepared as in (A) with the addition of

fluorogenic PLA2 substrate PC-A2. Fluorescent intensity was measured using the Pherostar

fluorescent plate reader (n=2 biological and 3 technical replicates, ****p<0.0001, Two-way

ANOVA followed by Dunnett’s multiple comparisons test comparing effect of hrPLA2G2A in

FBS vs 14-day Enz treated cells). (D) LNCaP and C4-2B cells were treated with Enz or

FBS+DMSO for 14 days. NBD-PC was added to media for one hour and baseline levels of

NBD-PC uptake were measured using qFM. (mean±SD, shown as average ratio

cleaved/uncleaved PC/well, n=3 wells representing >1000 cells per well, ** p<0.01

****p<0.0001, One-way ANOVA followed by Dunnett’s multiple comparisons test compared

to FBS control). (E) PC-A2 uptake of LNCaP and C42B cells treated with hrPLA2G2A (50 ng)

and KH064 (30 µM), alone or in combination, was measured using qFM (mean±SD), shown as

average ratio cleaved/uncleaved PC/well, n=3 wells representing >1000 cells, ** p<0.01

****p<0.0001, One-way ANOVA followed by Dunnett’s multiple comparisons test compared

to FBS control). (F) mRNA expression of phospho/lyso-phospholipid transporters in LNCaP

cells following 21 days Enz treatment was measured by microarray. Values shown as fold

change rel to FBS control (n=3, mean, * p<0.05 ** p<0.01 ***p<0.001 ****p<0.0001, Student’s

t-test of Day 21 Enz compared to FBS control).


therapies 130

In the experiments shown in Fig 4.16A and 4.16D, growth media were removed

prior to replacement with fluorophore label. There was no significant increase in

baseline levels of NBD-PE uptake (Fig 4.16A) or lyso-PC uptake (Fig 4.16D) in Enz

treated LNCaP cells compared to controls. Given the previously described evidence

that Enz treated LNCaP cells upregulate PLA2G2A secretion into the media, these

experiments were repeated, but lipid probes were added directly to the conditioned

media rather than removing the media prior to the addition of the probe. As shown in

Fig 4.17A-B, uptake of both lyso-PCs and PEs (PEs also shown in Fig 4.9C) was

significantly increased in LNCaP cells following 14 days Enz treatment, reaching

levels comparable to those seen following the addition of 50 ng PLA2G2A (Fig

4.17A).

Figure 4.17 Lipid uptake in conditioned media

Following 14 days treatment with Enz or FBS+DMSO control, PLA2G2A substrates

PC-A2 (A) or (B) NBD-PE were added to the conditioned media. hrPLA2G2A (50

ng) was added as a positive control. Following one hour incubation, media were

removed and lipid uptake was measured by qFM (A) shown as average

cleaved/uncleaved PC ratio per well (n=6 wells), (B) and shown as quantitation of

single cells; mean±SD for n>1000 cells, representative of 2 independent experiments,

**p<0.01 ****p<0.0001, Unpaired t-test comparing D14 Enz treated cells to

FBS+DMSO control).

Figure 4.17 Lipid uptake in conditioned media

Following 14 days treatment with Enz or FBS+DMSO control, PLA2G2A substrates

PC-A2 (A) or (B) NBD-PE were added to the conditioned media. hrPLA2G2A (50


therapies 131

4.2.12 Summary of ATT-induced lipid reprogramming in PCa cells

Using a combination of expression microarray analysis, qRT-PCR, protein mass

spectrometry and Western blot analysis, this study revealed time-dependent rewiring

of lipid and energy metabolism networks in PCa cells in response to ATTs. Among

these networks were fatty acid remodelling, lipogenesis, phospholipase activity, lipid

transport, phospholipid metabolism and lipid storage. Changes in expression of key

players in each of these pathways are summarised in Fig 4.18. These results suggest

that the upregulation of lipid transport, storage, and remodelling pathways and

accompanying downregulation of lipogenesis are early adaptive responses to ATTs,

revealing novel pathways which could be exploited in the fight against acquired drug

resistance in PCa.


therapies 132

4.18 Summary of lipid metabolic pathways altered by Enzalutamide

Regulation of lipid metabolic pathways induced by Enz was measured by protein

mass spectrometry, Western blot, ELISA, microarray and qRT-PCR. Bar

colours represent method of detection: black=ELISA; dark grey=protein mass

spectrometry; light grey= Western blot analysis; dark blue=microarray; light

blue=qRT-PCR.

DEGS2

SCD5

ELO

VL4

FADS2

ELO

VL5

ELO

VL6

FADS1

SCD1

ELO

VL7

-10

-5

0

5

10

15

Fatty acid remodelling

Fold

change r

el to

FB

S c

ontr

ol

PLA

2G2A

PLD

6

PLA

2G16

PLA

2G12

A

PLA

2G7

PLA

2G4C

PLA

2G3

PLA

2G5

-10

0

10

20

30

Phospholipase activity

Fold

change r

el to

FB

S c

ontr

ol

LPCAT4

LPCAT3

AGPAT2

DAGLB

AGPAT1

PTD

SS1

CHKA

-4

-2

0

2

4

Phospholipid metabolism

Fold

change r

el to

FB

S c

ontr

ol

HM

GCR

ACAT1

HM

GCS

ACACA

ACAT2

FASN

ACLY

SREBF2

-3

-2

-1

0

1

2

3

De novo lipogenesis

Fold

change r

el to

FB

S c

ontr

ol

ATP

8A1

ATP

10D

ATP

11A

SLC

27A1

ABCG1

SLC

27A4

SLC

27A2

ATP

11C

LDLR

ATP

9A

ATP

8B2

ATP

11B

TMEM

30A

FABP3

ATP

8B3

SCARF1

SCARB1

FFAR2

SLC

27A5

SLC

27A3

-5

0

5

10

15

Lipid transportF

old

change r

el to

FB

S c

ontr

ol

ATP

8A1

ATP

10D

ATP

11A

PLI

N

PPPDC1A

DGAT2

DGAT1

-5

0

5

10

15

Lipid storage

Fold

change r

el to

FB

S c

ontr

ol


therapies 133

4.3 DISCUSSION

Despite initial tumour regression following ATT, acquired resistance and

subsequent disease progression remains a major obstacle in fighting CRPC, which

remains incurable. The totality and complexity of mechanisms driving Enzalutamide

resistance remain unclear. This study for the first time provides a mechanism directly

linking PLA2G2A activity to enhanced lipid uptake and lipid accumulation that

promotes cell survival and drug resistance in PCa.

Metabolically and proliferation-wise, LNCaP cells responded to extended Enz

treatment by entering into a state of cellular quiescence, i.e. reduced proliferation, ATP

production and mitochondrial activity, as summarised in Fig 4.1A. Transcriptional

data supports this quiescent state, with the downregulation of pathways involved in

cell cycle, mitochondrial activity, and oxidative phosphorylation (Fig 4.2B). This is

consistent with previous studies that show subpopulations of cancer cells enter a

quiescent, “persister” state in response to anti-cancer treatments (Hangauer et al.,

Figure 4.19 ATTs induce rewiring of metabolic networks in PCa cells to fuel

survival

Schematic representation of cellular lipid supply and lipid remodelling pathways in PCa

cells. Integration of expression microarray, proteomics, lipid mass spectrometry and

metabolomics analyses were used to identify altered metabolic pathways in our LNCaP

ATT model. Lipid transporters, lipogenic enzymes, and enzymes involved in

remodelling pathways that are increased or decreased with ATT treatment are

highlighted in red and blue, respectively. Schematic generated by K. Tousignant and Dr.

Sadowski.


therapies 134

2017; Ramirez et al., 2016), reviewed by Vallette and colleagues (Vallette et al., 2018).

Interestingly, this quiescence was accompanied by increased lipid uptake and lipid

content in ATT treated cells. In order to further investigate the role of quiescence in

the development of therapy resistance in PCa cells, it would be beneficial to first

validate these pathways using functional assays, however, that was not a focus of the

present study.

In this study, these findings were validated using novel automated quantitative

fluorescent microscopy assays and image analysis to gain a comprehensive

understanding of alterations to the lipid landscape of PCa cells in response to ATTs.

The increase in lipid droplet number, neutral and phospholipid content, and increased

Filipin staining of free, unesterified cholesterol all support the enrichment in pathways

identified in the transcriptomic analysis. Shotgun lipidomics, which provides a

comprehensive and unbiased understanding of the PCa cell lipidome, validated the

lipid probe-based assays in which a significant increase in lipid content was observed

in response to ATTs. While lipid droplets are most well characterised for their

triacylglyceride and cholesteryl ester storage, both lipid species were depleted in Enz

treated cells regardless of the increased lipid droplet number and size. Driven by this

observation, other pathways of lipid droplet biogenesis were investigated. It was

recently shown that ceramide is converted to acylceramide, via DGAT2, and stored in

lipid droplets (Senkal et al., 2017). While DGAT1 and DGAT2 are both involved in

triacylglyceride synthesis, DGAT2 also drives acylceramide synthesis and subsequent

storage in lipid droplets. DGAT2 expression is upregulated in Enz treated cells while

DGAT1 is downregulated (Fig 4.18), suggesting that increased acylceramide synthesis

may be contributing to the increased lipid droplet accumulation induced by ATTs.

However, given the heterogeneous nature of lipid droplets within cell populations

(Herms et al., 2013), further investigation into lipid droplet formation and content in

drug-induced disease progression would be of extreme value. In agreement with the

data described here, lipid droplet accumulation has recently been described as a

mechanism of drug resistance in colorectal cancer (Cotte et al., 2018), renal cell

carcinoma (Lue et al., 2017), and breast cancer (Hultsch et al., 2018) cell lines,

suggesting that this phenotype may be a general characteristic of cancer treatment

resistance.


therapies 135

It was found that Enzalutamide and CSS treatment induce a significant increase

in the uptake of LDL and ac-LDL particles. Notably, charcoal-stripping of serum has

little effect on lipid content, suggesting that availability of lipids is not a contributing

factor to alterations in lipid metabolism in response to CSS treatment but rather that

the ATT-environment that promotes lipid accumulation. The increase in the bulk

transport of lipids through lipoprotein particles, which include fatty acids, could

explain the increase in total fatty acid content, regardless of the reduction in C16:0

uptake observed. The increase in linoleic acid (LA; 18:2, -6) and essential fatty acid

metabolites arachidonic acid (AA; C20:4 -6) and docosahexaenoic acid (DHT; 22:6

-3) measured by GCMS validate that exogenous lipid uptake is increased as an

adaptive response to ATT. Only recently has attention been given to unveiling the lipid

transporter protein landscape in cancer cells (Iglesias-Gato et al., 2018; Iglesias-Gato

et al., 2016), including our own work in PCa. The observed increase in protein

expression of both LDLR and SCARB1 in the long-term in vitro ATT model suggests

that this receptor mediated transport contributes to the increased lipid uptake observed

in Fig 4.9B, and that LDLR and SCARB1 are lipid supply pathways involved in PCa

progression. However, the net contribution of any one transport pathway to the total

lipid pool remains unclear and requires further investigation. One caveat of the lipid

uptake assays performed in this study is the redundancy of lipid transporters and the

difficulty in isolating the contribution of individual transporters (Doege et al., 2006;

Schneider et al., 2016). This redundancy also makes it difficult to therapeutically target

individual lipid transporters. Regardless, LDLR and SCARB1, along with several other

lipid transporters, were validated in an LNCaP tumour xenograft progression model

(Locke et al., 2008) in which recurring and CRPC tumours expressed higher transcript

and/or protein levels of several lipid transporters. While both LDLR and SCARB1

have previously been associated with PCa, there have been conflicting conclusions as

to their role in tumour progression (Furuya et al., 2016; Schörghofer et al., 2015b;

Stopsack et al., 2017). This study suggests that lipid scavenging from exogenous

sources, including via LDLR and SCARB1, is a response to ATTs that helps fuel

survival. However, the limited understanding of lipid transporters in cancer and their

therapeutic potential requires further attention. siRNA or small molecule inhibition of

individual transporters could help to address which transporters are actually

responsible for uptake and could be exploited therapeutically.


therapies 136

Enhanced lipogenesis is a well-established metabolic phenotype of PCa,

therefore the effect of ATT on the de novo contribution to the cellular lipid pool was

investigated. De novo cholesterol synthesis from C13-labelled acetate was significantly

decreased throughout Enz treatment, which was surprising given the markedly

increased cellular cholesterol content, suggesting that the increase in cellular

cholesterol content throughout ATT treatment was coming from the uptake of

exogenous cholesterol sources. However, these data only show that lipogenesis from

acetate is reduced by Enz. It is plausible that other carbon sources are used for the de

novo synthesis of lipids, however delineating the contribution of various carbon

sources to lipid synthesis remains a major challenge in the field. Nevertheless,

increased cholesterol uptake was validated using a fluorescently labelled cholesterol

analogue. It has been shown by our group and others that cholesterol serves as a

precursor for the steroidogenic pathway in which testosterone can be endogenously

synthesized in prostate cells, and that activation of the steroidogenic pathway is an

adaptive response contributing to the development of CRPC (Dillard et al., 2008;

Ghayee & Auchus, 2007; Leon et al., 2010). Additionally, membrane cholesterol is

shown to affect lipid raft composition which may influence oncogenic signalling

(Zhuang et al., 2005), further highlighting the dependence of PCa cells on cholesterol

for growth and survival. Cholesterol has gained notable interest in PCa given that it

has been shown to play a central role in de novo androgen synthesis in CRPC (Dillard,

Lin, & Khan, 2008; Locke et al., 2008). The increased cholesterol content shown in

this study provides important validation that the long-term in vitro ATT model

recapitulates pathways observed xenograft models and patient samples. Targeting

cholesterol as a therapeutic strategy has solely focused on the inhibition of de novo

cholesterol synthesis in PCa, i.e. inhibition of HGMCR with Statin treatment (Gordon

et al., 2016). These data suggest for the first time that cholesterol uptake rather than

synthesis is a major contributor to the cholesterol accumulation characteristic of CRPC

and gives a rationale for the reconsideration on how cholesterol metabolism is targeted

in PCa patients.

Interestingly, endogenous fatty acid synthesis was also investigated using C13

metabolomics, however the incorporation of 13C-acetate into palmitate, oleate and

stearate was negligible. Preliminary data collected during the optimisation of this

metabolomics experiment showed that incorporation of 13C -glucose, which is widely


therapies 137

accepted as the precursor for de novo synthesis of lipids, was also negligible. This

result supports the notion that glucose is not a major contributing cellular carbon

source and suggests that alternative sources of carbon for lipid biomass production are

utilised by PCa cells. Support from this comes from recent studies in which authors

show that majority of lipid-derived carbon comes from exogenous lipids (Balaban et

al., 2019; Hosios et al., 2016). It would be of major value to investigate the contribution

of different carbon sources to total cellular biomass, especially in order to identify the

most relevant substrate for carbon tracing metabolomics experiments such as the one

described here. Once this is achieved, a more accurate delineation of the relative

contribution of de novo lipogenesis and exogenous uptake of lipids may be

accomplished.

Gene set enrichment analysis identified previously identified ATT-induced

pathways associated with resistance and development of CRPC such as arachidonic

acid metabolism and steroid hormone biosynthesis (Dillard et al., 2008; Locke et al.,

2010). This serves as validation that this model recapitulates what occurs in vivo.

Additionally, extensive lipid membrane remodelling in response to Enz treatment was

identified, including significantly increased fatty acid elongation and desaturation,

both of which have previously been associated with PCa incidence and aggressiveness

(Peck et al., 2016; Tamura et al., 2009). Long-chain fatty acids found in membrane

phospholipids, along with cholesterol molecules, are critical for membrane

stabilisation and are involved in lipid raft formation where oncogenic growth

signalling occurs (Tamura et al., 2009; Zhuang et al., 2005). The incorporation of long

chain polyunsaturated fatty acids (PUFAs) into lipid rafts can have a significant impact

on the content and function of transmembrane proteins, many of which are involved

in cancer cell signalling, suggesting one rationale for the shift to this lipid phenotype.

Additionally, the effect of PUFAs on membrane fluidity is thought to play a role in

sensitivity to chemotherapy reagents (Corsetto et al., 2017). Increased PUFA content

does, however, make cells more susceptible to oxidative damage via reactive oxygen

species (ROS) (Ayala et al., 2014). The increased PUFA content observed in this ATT

model would presumably make cells more sensitive to ferroptosis via GPX4 inhibition,

as has been shown in drug-resistant persister cells across several cancer types

(Hangauer et al., 2017). Consistent with this, it was found that ATT treated PCa cells

were significantly more sensitive to selective GPX4 inhibitor, RSL3. This metabolic


therapies 138

vulnerability has yet to be exploited in PCa, and treatment with ROS inducers could

serve as a novel therapeutic strategy in the treatment of PCa patients. Together, these

data support the hypothesis that lipid remodelling plays a role in PCa progression.

In addition to fatty acid elongation and desaturation, an integrated ‘omics

analysis strategy identified upregulation of novel pathways previously not implicated

in ATT-resistance. Multiplex labelling for protein mass spectrometry is powerful in its

ability to quantitate six different samples in one analysis, therefore reducing technical

variation between samples as well as run time on the instrument. However, one

limitation of this approach is the reduced robustness of detectable proteins; because

the final loading protein concentration is a combination of six samples, less abundant

proteins would have easily been under the detectable threshold. To account for this in

future experiments, samples could be fractionated (via selective solubilisation, pH

gradient fractionation, etc.) prior to labelling. With this approach, detection of less

abundant proteins could be enhanced in certain fractions, i.e. lipid transporters in

membrane fractions. Nevertheless, our integrated ‘omics analysis was able to identify

metabolic pathways to investigate in the context of ATT. These include pathways of

lipid remodelling, specifically via phospholipase PLA2G2A. Further investigation

confirmed that ATT induced significant upregulation of PLA2G2A mRNA and protein

levels across 5 PCa cell lines, which contributed to enhanced enzymatic activity and

lysophospholipid uptake in the 2 cell lines tested. Importantly, it is shown that

PLA2G2A activity occurs extracellularly, releasing lyso-PLs and arachidonic acid into

the medium for subsequent uptake by PCa cells. The observed increase in lysolipid

uptake observed in Fig. 4.17A-B is further evidence that PLA2G2A-mediated cleavage

of lipid substrates for subsequent cellular uptake is an adaptive response to ATTs in

LNCaP cells. In future experiments, previously described uptake assays (NBD-Ch,

LDL, acLDL, C16:0 BODPIY) should be repeated in conditioned media of Enz treated

cells to investigate additional effects on the uptake of various lipid substrates.

Given that AA serves as a precursor for pro-inflammatory lipid mediators

including prostaglandin E2 (PGE2), which works paradoxically in both the activation

and suppression of immune responses (Kalinski, 2012), PLA2G2A mediated release

of AA may be advantageous to cancer cells by providing an inflammatory environment

to enhance immunosuppressive activity. However, in addition to its role in immune

signalling, these data suggest that PLA2G2A mediated lysophospholipid uptake is also


therapies 139

critical to PCa cells undergoing ATT. This is supported by a recent study showing

secreted PLA2G2A activity provides unsaturated fatty acids and contributes to lipid

droplet accumulation to protect breast cancer cells from nutrient stress and PUFA

lipotoxicity (Jarc et al., 2018).

It has recently been shown that lysophospholipids are a more accessible

nutrient source than serum phospholipids and that lysophospholipid scavenging helps

cancer cells to survive by providing fatty acids to meet their metabolic demands in

periods of nutrient stress, (Kamphorst et al., 2013). Lysophospholipids can also serve

as signalling molecules and contribute to membrane structure and fluidity, both of

which are important in cancer cell survival (Zalba & Ten Hagen, 2017). Additional

evidence shows lysophospholipids can alter mitochondrial activity (Hollie et al., 2014)

and suppress fatty acid oxidation (Labonté et al., 2010), both of which are consistent

with the PCa long-term ATT model. PLA2G2A has previously been suggested as a

biomarker for PCa (Dong et al., 2010; Leslie et al., 2012) and breast cancer (Qu et al.,

2018), however no functional role in PCa progression has been provided thus far.

Importantly, increased PLA2G2A expression is also seen in renal cell carcinoma (Lue

et al., 2017) and breast cancer (Hangauer et al., 2017) models in response to anti-cancer

treatments, suggesting again that the lipid remodelling cascade induced by anti-cancer

therapies is a more global drug resistance phenotype. This study provides a novel

mechanism behind therapy induced PLA2G2A upregulation by showing that

PLA2G2A is directly linked to enhanced cellular lipid uptake in cancer cells by

providing lyso-lipids to meet metabolic demands during the early phase of acquired

drug resistance. This discovery could have major implications on therapeutic targeting

of lipid metabolic pathways to be used in combination with ATT to fight the

progression to CRPC.

PLA2G2A is a novel target to fight the development of therapy resistance in PCa cells 140

PLA2G2A is a novel target to fight the

development of therapy

resistance in PCa cells

5.1 INTRODUCTION

Despite the complex and dynamic role of lipids in many biological processes,

the involvement of lipid remodelling to therapy resistance, cancer cell survival, and

disease progression has only recently emerged as an area of investigation (Zalba &

Ten Hagen, 2017). Phospholipases are of particular interest due to their dual role in

the recycling process of phospholipids (PLs) through the Lands Cycle (lipid

remodelling), and the production of inflammatory lipid mediators that feed into

eicosanoid production (Murakami & Lambeau, 2013). PLs can be generated through

two major pathways: The Lands Cycle and Kennedy pathway (Hishikawa et al., 2013).

The Lands Cycle describes the recycling of PLs in which in which phospholipases A2

(sPLA2s) cleave PLs at the sn-2 position and are later re-esterified into PLs via the

actions of lysophospholipid acyltransferases (LPCATs) (Hishikawa et al., 2013).

Alternatively, PLs can be synthesised de novo using acyl-CoA donors via glycerol-3-

phosphate (G3P) in the Kennedy pathway, (Hishikawa, Eto, Shimizu, Harayama, &

Shindou, 2013). The interplay between these pathways results in highly diverse fatty

acid compositions making up cellular and organelle membranes (Fig 5.1).

Figure 5.1 Phospholipid metabolism via Lands Cycle and Kennedy Pathway

Acyl-CoA donors are converted to lysophospholipids by glycerol-3-phosphate

(G3P), which can then be incorporated into diacylglycerols (DAGs) and

triacylglycerols (TAGs). Figure from Hishikawa et al. (2013).

Figure 5.2 The multifunctional role of sPLA2 in cancer


Additional functions of phospholipases in mechanisms of cancer progression

have recently gained notable attention, as outlined in Fig 5.2 (Brglez et al., 2014).

Secreted phospholipase A2 (sPLA2) belongs to a group of lipolytic enzymes that

release free FA and lysophospholipids from phospholipids by catalysing hydrolysis of

the sn-2 ester bond in the extracellular space (reviewed in (Murakami & Lambeau,

2013)). PLA2G2A in particular has a high affinity for phosphatidylserine,

phosphatidylethanolamine and phosphatidylglycerol (Murakami & Lambeau, 2013).

The enzymatic activity can act on phospholipids of cell membranes and other

phospholipid substrates such as lipoproteins, making phospholipase activity an

important mediator in cellular lipid homeostasis. Under normal physiological

conditions, PLA2G2A is produced primarily by the pancreas and is subsequently

secreted to aid in the degradation of dietary components or foreign microbial

phospholipids. Additionally, it is found in sites of inflammation where it is known to

release bioactive fatty acids including, but not limited to, arachidonic acid,

docosahexaenoic acid, and eicosapentaenoic acid for the production of eicosanoids

(Murakami & Lambeau, 2013; Murakami et al., 1998). Because of this, it is often

referred to as Bacteroidal sPLA2 or Inflammatory sPLA2 (Murakami et al., 1998).

Figure 5.2 The multifunctional role of sPLA2 in cancer

The enzymatic and ligand activity of secreted PLA2s is thought to drive cancer cells

by providing lipid substrates and bioactive signaling molecules. Figure from Brglez

et al. (2014).


Recently, however, research has found that altered expression of sPLA2 is associated

with various cancers including prostate, breast, colon, gastric, and lung, but with

controversial roles (Juan, Long, Yu, & Yoshikazhu, 2017). Increased PLA2G2A

expression correlated with a poor therapeutic response in rectal (Hong‐Lin et al., 2015)

and lung cancer (Wang, Hao, Wang, & Xiao, 2014), but conversely was found to play

a protective role in gastric adenocarcinoma (Leung et al., 2002), suggesting that the

pro-or anti-tumourigenic role of PLA2G2A activity is dependent on cancer location

and microenvironment. The current hypothesis surrounding sPLA2 activity in cancer

is that lipid mediators released by sPLA2 can promote tumourigenesis by stimulating

proliferation, increasing local inflammation and promoting angiogenesis through

signalling of bioactive lipids via binding of their cognate receptors (Brglez et al.,

2014).

While recent evidence suggests PLA2G2A as a potential biomarker for PCa

(Dong et al., 2010; Leslie et al., 2012; Li et al., 2016), the functional role of PLA2G2A

in PCa cell survival and growth remains unknown. Previous chapters highlight the

importance of lipid acquisition in PCa cells undergoing ATT, as well as increased

PLA2G2A transcript expression, protein secretion and activity in response to Enz

treatment. This chapter presents results from the continued investigation of the

mechanistic action of PLA2G2A as well as its potential as a novel therapeutic target

in PCa cells undergoing ATT.

5.2 RESULTS

5.2.1 PLA2G2A is upregulated in PCa patients

Serum levels of PLA2G2A have recently been shown to be elevated in men with

PCa compared to healthy controls (Dong et al., 2010; Leslie et al., 2012; Li et al.,

2016), and thus have been suggested as a potential biomarker for disease. To validate

the clinical relevance of PLA2G2A in PCa, transcript expression was assessed in

clinical patient samples within the normal prostate gland vs primary tumour (Fig

5.3A). PLA2G2A transcripts were found to be significantly upregulated in PCa tumour

across 4 independent publicly available datasets (Holzbeierlein et al., 2004; Lapointe

et al., 2004; Singh et al., 2002; Tomlins et al., 2006). Furthermore, PLA2G2A levels

were found to increase with increasing Gleason grade, the recognised system for

scoring disease severity in PCa (Fig 5.3B) (Tomlins et al., 2006).


Previous studies reported average serum PLA2G2A concentrations of 1.11 ng/mL

in healthy men and 1.30 ng/mL in patients with benign prostatic hyperplasia (BPH)

(Menschikowski et al., 2012; Menschikowski et al., 2013). Strikingly, baseline serum

PLA2G2A levels of PCa patients ranged from 10.836 ng/mL to 37.102 ng/mL, with

an average of 18.575 ng/mL, nearly 18-fold higher than levels reported in healthy

controls (Fig 5.3C) (Menschikowski et al., 2013; Menschikowski et al., 2012). Serum

PLA2G2A positively correlated with PSA levels, with a correlation coefficient of

0.5802 (pval 0.0377). However, serum PLA2G2A was not significantly correlated

with patient testosterone levels. To investigate PLA2G2A expression in the context of

ATT, serum from men with advanced, metastatic PCa was collected, and the serum

level of PLA2G2A protein was measured by ELISA. In this study, serum was collected

from patients before and after receiving Eligard, a type of hormone therapy (ADT)

drug, for a period of 12 weeks (HREC/14/QPAH/135) (Rhee et al., manuscript in

preparation). As shown in Fig 5.3E, there was no significant difference in PLA2G2A

levels between pre- and post- ATT.


Figure 5.3 PLA2G2A is upregulated in PCa and is associated with higher Gleason

score

(A) Oncomine analysis of 4 independent PCa datasets comparing PLA2G2A expression

in prostate gland vs tumour (graph shows log2 median-centered ratio; Student’s t-test used

to generate p-value (B) PLA2G2A transcript levels were measured in Tomlins prostate

dataset (Tomlins et al., 2006) with increasing Gleason score (*p<0.05, **p<0.01, One-

way ANOVA followed by Dunnett’s multiple comparisons test). (C-D) Baseline levels of

serum PLA2G2A, PSA and testosterone in PCa patient serum samples was measured by

ELISA. Graphs show correlation analysis of (C) serum PSA and (D) testosterone with

serum PLA2G2A in patient samples (n=13 (PSA); n=11 (testosterone), Pearson’s

Correlation Coefficient). (E) PCa patient serum was collected prior to and following 12

weeks of androgen-deprivation therapy. Serum PLA2G2A was measured by ELISA

(n=14) (Rhee et al., manuscript in preparation).

Figure 5.3 PLA2G2A is upregulated in PCa and is associated with higher Gleason

score


5.2.2 Androgens supress PLA2G2A expression

Given the significant increase in PLA2G2A transcript expression, protein secretion

and enzymatic activity induced by Enz treatment and androgen depleted (CSS), the

role of androgens and the AR axis on PLA2G2A regulation was further investigated.

RNA-sequencing of 7 PCa cell lines revealed that the highest transcript levels were

amongst AR-positive, androgen-dependent LNCaP, DuCaP, VCaP and LAPC4 cells,

and androgen-responsive C4-2B cells, as compared to AR-negative DU145 and PC3

cells that exhibited similar levels to the prostatic stromal myofibroblast WPMY1 cell

line (Fig 5.4A). Notably, all cell lines were grown in androgen replete media before

sample preparation. Baseline expression of PLA2G2A was validated by qRT-PCR in

selected cell lines in addition to non-malignant RWPE1 and BPH1 cells (Fig 5.4B).

Baseline PLA2G2A was almost negligible in PC3, DU145, RWPE1 and BPH1 cells,

while AR-positive LNCaP and VCaP cells had 350-fold and 315-fold higher

expression, respectively, compared to BPH1 cells.

To directly interrogate the influence of the AR-axis on PLA2G2A, LNCaP cells

were grown in androgen-depleted media (CSS) for 48 hours and were then treated with

R1881 or DHT in the presence or absence of Enz for an additional 48 hours. As shown

in Fig 5.4C, treatment with both R1881 and DHT resulted in a significant reduction in

PLA2G2A transcript levels. This response could be moderately reduced by co-

treatment with Enz. This data suggests that androgens strongly repress PLA2G2A

transcript levels in PCa cell lines in vitro. This was further validated in DuCaP and

VCaP cells treated with DHT in the presence or absence of Enz by RNA-sequencing

analysis. In agreement with the negative androgen-regulation observed in LNCaP cells

in Fig 5.4C, PLA2G2A was almost negligible in both DuCaP and VCaP cells treated

with DHT (Fig 5.4D).


Figure 5.4 PLA2G2A is androgen repressed in PCa cells

(A) Transcript levels of PLA2G2A across one prostatic stromal myofibroblast and 7 PCa cell

lines was analysed by RNA-sequencing. (B) Baseline levels of PLA2G2A were validated in

PCa cell lines in addition to non-malignant RWPE1 and BPH1 cells by qRT-PCR (n=3, values

shown as fold change relative to BPH1, mean±SD, One-way ANOVA with Dunnett’s

multiple comparisons test relative to BPH1, ****p<0.0001). (C) Transcript levels of indicated

phospholipases was measured by qRT-PCR in LNCaP cells grown for 48 hours in CSS

followed by treatment with 0.1% ethanol (Ctl), 1 nM R1881 (blue) or 10 nM DHT (purple)

in the presence or absence of Enz (10 µM, red) for an additional 48 hours (n=3, mean±SD,

One-way ANOVA with Dunnett’s multiple comparisons test relative to ethanol control, or

Students t-test comparing DHT or R1881 +/-Enz (10 µM) *p<0.05 ***p<0.001

****p<0.0001). (D) Androgen regulation of PLA2G2A was validated in DuCaP and VCaP

cell lines by RNA-seq analysis. Fpkm=fragments per million kilobase reads. (n=2, mean±SD,

One-way ANOVA with Dunnett’s multiple comparisons test relative to vehicle control

*p<0.05).

Figure 5.4 PLA2G2A is androgen repressed in PCa cells


5.2.3 Investigation of selected PLA2 family members in PCa

The phospholipase 2 family consists of 15 distinct groups further divided into 4

main categories: the secreted sPLA2, cytosolic cPLA2, calcium-independent iPLA2,

and platelet activating factor (PAF) (Burke & Dennis, 2009; Schaloske & Dennis,

2006). These proteins differ by their substrate specificity and cellular distribution,

however all PLA2 enzymes share the ability to hydrolyse the fatty acid from the sn-2

position of phospholipids. To address if the lipid remodelling response observed in the

long-term ATT model was specific to PLA2G2A and its extracellular activity, or rather

occurred because of a general demand of ATT-treated PCa cells for PLA2 products,

i.e, arachidonic acid and lysophospholipids, PLA2 family members that were

detectable by RNA-sequencing were investigated for androgen regulation in PCa cells.

As seen in Fig 5.5A, transcript levels of other sPLA2s was much lower than that of

PLA2G2A in LNCaP, DuCaP, and VCaP cell lines. Furthermore, the androgen-

suppression observed for PLA2G2A was absent for all other PLA2 family members

except for PLA2G12A in LNCaP cells (Fig 5.5B). These results were validated in

DuCaP and VCaP cells in which no androgen regulation was observed for any

additional PLA2 family members (Fig 5.5C).

Next, transcript levels of PLA2s was explored in the context of long-term ATT (7-

21 days). Here, it was found that PLA2G7 and PLA2G12 transcript levels (Fig 5.5D)

were significantly increased by chronic Enz treatment (10 µM), but not to the

magnitude of the roughly 14-fold increase observed with PLA2G2A (Fig 4.13). In

agreement with this, transcriptional profiling of an LNCaP tumour xenograft model of

CRPC progression in mice revealed modest changes in expression of several genes

encoding phospholipases in tumours that had progressed to CRPC after castration and

in CRPC tumours treated with Enz (Fig 5.5E). However, significantly increased levels

were only observed for PLA2G2A. Together, these data show that PLA2G2A is highly

expressed in AR-positive PCa cell lines relative to AR-negative PCa and non-

malignant prostate cells and that PLA2G2A transcript levels are androgen-suppressed.

While PLA2G2A is not the sole contributor to the lipid remodelling phenotype

described in the present model, it clearly shows the strongest androgenic regulation

and response to ATTs across the tested PCa cell lines and in vivo models investigated.




5.2.4 Targeting PLA2G2A in PCa cells

The data shown in this study highlight the importance of PLA2G2A in PCa patients

as well as in vivo and in vitro models of ATT, where androgen regulation often plays

a major contributing role in treatment response and resulting disease progression. Here

the functional importance of PLA2G2A for PCa cell survival was explored. Gene

silencing of PLA2G2A using 3 unique small interfering ribonucleic acid (siRNA)

sequences resulted in significant growth inhibition in LNCaP cells compared to a non-

targeting siRNA sequence (SiCtl, Fig 5.6A-B). LNCaP cells were transfected with

each of the 3 siRNA sequences for 48 hours and PLA2G2A expression was analysed

by qRT-PCR. Sequence #1 was chosen to take forward into subsequent experiments

as it provided the most efficient knockdown of PLA2G2A (Fig 5.6A-B) and growth

suppression in LNCaP cells (Fig 5.6C). A growth inhibitory response was absent from

Figure 5.5 Transcript levels of PLA2 family in PCa cells

(A) Transcript levels of PLA2 family members across LNCaP, DuCaP and VCaP cell

lines was analysed by RNA-sequencing. (B) Transcript levels of indicated PLA2

family members was measured by qRT-PCR in LNCaP cells grown for 48 hours in

CSS followed by treatment with 1 nM R1881 or 10 nM DHT in the presence or

absence of Enz (10 µM) for an additional 48 hours. PSA measurements are shown as

a control for androgen response (n=3, mean±SD, One-way ANOVA with Dunnett’s

multiple comparisons test relative to 0.1% ethanol control, *p<0.05 ****p<0.0001).

(C) DuCaP and VCaP cells were treated as in (B). Fpkm (fragments per kilobase

million reads) values of indicated genes were measured by RNA-sequencing. (D)

Transcript levels of PLA2 family members in LNCaP cells treated for up to 21 days

with Enz (10 µM) or FBS+DMSO control was measured by qRT-PCR (n=3,


FBS+DMSO control, *p<0.05, **p<0.01, ***p<0.001). (E) Heatmap showing fpkm

(fragments per kilobase million reads) values measured by RNA-sequencing of

selected phospholipase genes in an LNCaP xenograft tumour progression model of

CRPC progression. (Intact=sham-castrated; Post Cx=one week following castration;

CRPC=tumours following biochemical recurrence; Enz=following castration, mice

were treated with Enz and tumours were collected once the ethical endpoint was

reached).

Figure 5.5 Transcript levels of PLA2 family members in PCa cells

(A) Transcript levels of PLA2 family members across LNCaP, DuCaP and VCaP cell

lines was analysed by RNA-sequencing. (B) Transcript levels of indicated PLA2

family members was measured by qRT-PCR in LNCaP cells grown for 48 hours in

CSS followed by treatment with 1 nM R1881 or 10 nM DHT in the presence or

absence of Enz (10 µM) for an additional 48 hours. PSA measurements are shown as

a control for androgen response (n=3, mean±SD, One-way ANOVA with Dunnett’s

multiple comparisons test relative to 0.1% ethanol control, *p<0.05 ****p<0.0001).

(C) DuCaP and VCaP cells were treated as in (B). Fpkm (fragments per kilobase

million reads) values of indicated genes were measured by RNA-sequencing.

Figure 5.5 Transcript levels of PLA2 family members in PCa cells


AR-negative DU145 and non-malignant BPH1 cells transfected with siPLA2G2A

sequence #1 (Fig 5.6D), which is not surprising given the essentially negligible

PLA2G2A transcript levels in these cells compared to AR-positive cell lines, as was

shown in Fig 5.4B. Next, efficiency of knockdown at the protein level was analysed

by Western blot. Surprisingly, the band for PLA2G2A was significantly increased in

size with increasing time of siRNA knockdown of PLA2G2A (Fig 5.6E). It could be

speculated that this antibody cross-reacts with other PLA2 family members which

shared the same molecular weight as PLA2G2A, and that other PLA2 family members

may increase with the knockdown of PLA2G2A as a compensatory mechanism in order

to sustain the availability of lysophospholipids and arachidonic acid to PCa cells.

Indeed, transcript levels of secreted PLA2G7, PLA2G15, and cytosolic PLA2G4 were

significantly increased in siPLA2G2A transfected cells compared to the siCtl cells (Fig

5.6F), although this was not confirmed at a protein level. To determine the effect of

PLA2G2A silencing on lipid content of LNCaP cells grown in androgen replete

medium, after 48 h of siPLA2G2A and siCtl treatment cells were fixed, stained with

Nile Red, and analysed by qFM (Fig 5.6G). While there was no effect on Nile Red

staining for phospholipid detection, PLA2G2A silencing significantly lowered neutral

lipid detection when compared to siCtl cells.



Figure 5.6 Targeting PLA2G2A in PCa cells by gene silencing

(A) LNCaP cells grown in androgen replete medium were transfected with 3 different

siPLA2G2A sequences (siPLA #1-3) for 48 hrs. Transcript levels of PLA2G2A were measured

by qRT-PCR and the most efficient knockdown (B) was chosen to use in subsequent experiments

(n=3, mean±SD, One-way ANOVA with Dunnett’s multiple comparisons test relative to siCtl

control ****p<0.0001). (C) LNCaP cells were transfected as in (A) Confluence was measured

every 2 hours using the IncuCyte live-cell imaging system (n=3), mean±SD, One-way ANOVA

with Dunnett’s multiple comparisons test relative so siCtl control, ****p<0.0001. (D) DU145

and BPH1 cells were transfected 5-20 nM siPLA2G2A #1 for 48 hrs and confluence was

measured as in (C). (E) Effect of PLA2G2A siRNA on its protein level in transfected LNCaP

cells was analysed by Western blot. Data shown are a representative of 2 independent

experiments. (F) Transcript levels of PLA2 family members following 48 hrs siPLA2G2A #1

transfection in LNCaP cells (androgen replete medium) was measured by qRT-PCR (n=3,

mean±SD, Two-way ANOVA with Dunnett’s multiple comparisons test relative to 10 nM siCtl

control, **p<0.01 ****p<0.0001. (G) Before fixation, LNCaP cells were transfected with 10

nM siPLA2G2A #1 for 48 hours. Lipid staining was measured by qFM (n~3000 cells from 3

wells, individual values shown as mean fluorescence intensity (MFI) per cell; error bars show

mean±SD, Unpaired t-test comparing 10 nM siPLA2G2A to 10 nM siCtl transfected cells,

****p<0.0001).


Next, three selective inhibitors of PLA2G2A were tested for their effect on cell

proliferation in LNCaP cells. KH064 is a D-tyrosine derived compound (Hansford et

al., 2003) that has been shown to fully inhibit human PLA2G2A and successfully

reduce inflammatory responses in both arthritis (Hansford et al., 2003) and

ovariectomy induced bone loss (Gregory, Kelly, Reid, Fairlie, & Forwood, 2006)

models with a reported IC50 of 29 nM (MW 487.27 g/mol) (Hansford et al., 2003).

LY311727 is a N-benzyl derivative with a reported IC50 of 60 nM (MW 430.4 g/mol)

(Schevitz et al., 1995), and the indole analogue Varespladib is claimed to be the most

potent PLA2G2A inhibitor, with an IC50 of 9 nM (MW 380.4 g/mol) (Snyder et al.,

1999). As shown in Fig 5.7A, KH064 inhibited cell proliferation at a concentration of

40 µM and above, with a calculated IC50 of 31 µM (Fig 5.7B). Cellular reducing

power was significantly decreased from 10 µM KH064 and higher as measured by

Presto Blue (Fig 5.7C). LY311727 and Varespladib treatment showed negligible

growth inhibition up to the maximum concentration tested (100 µM). Literature

suggests that LY311727 and Varespladib are more potent inhibitors of PLA2G2A than

KH064 (Reid, 2005), however both LY311727 and Varespladib failed to inhibit

growth of PCa cells (Fig 5.7D), as was seen in LNCaP cells following treatment with

KH064. LNCaP cells treated with either KH064 or LY311727 had decreased

PLA2G2A transcript levels, with a much stronger effect observed in KH064 treated

cells (Fig 5.7E).


Figure 5.7 Characterisation of small molecule inhibitors of PLA2G2A in vitro

(A) LNCaP cells were treated with KH064 up to 100 µM. Confluence was measured

every 2 hours using the IncuCyte live-cell imaging system (n=3, mean±SD, One-way

ANOVA with Dunnett’s multiple comparisons test relative to 0.1% DMSO control,

****p<0.0001). (B) Half-maximal inhibitory concentration (IC50) was calculated

using nonlinear fit of normalised data. (C) LNCaP cells were treated as in (A) and cell

viability was measured using Presto Blue (n=3, mean±SD, One-way ANOVA with

Dunnett’s multiple comparisons test relative to 0.1% DMSO control, *p<0.05,

***p<0.001, ****p<0.0001). (D) LNCaP cells were treated with LY311727 (left) and

Varespladib (right) up to 100 µM and confluence was measured as in (A). (E) LNCaP

cells were treated with 30 µM KH064 or LY311727 for 48 hrs. PLA2G2A was

measured by qRT-PCR and shown as fold change relative to 0.1% DMSO control

(n=1).

Figure 5.7 Characterisation of small molecule inhibitors of PLA2G2A in vitro

(A) LNCaP cells were treated with KH064 up to 100 µM. Confluence was measured



****p<0.0001). (B) IC50 was calculated using nonlinear fit of normalized data. (C)

LNCaP cells were treated as in (A) and cell viability was measured using Presto

Blue (n=3, mean±SD, One-way ANOVA with Dunnett’s multiple comparisons test


PLA2G2A purified from bee venom is often used as a positive control for

measuring PLA2G2A activity and inhibition (Kolde & Vilo, 2015). In order to directly

measure the enzymatic inhibition of PLA2G2A using the above 3 inhibitors, bee

PLA2G2A was added alone or in combination with each of the 3 inhibitors at 50 and

100 µM and enzyme activity as detected by absorbance was measured over 60 minutes

using a plate reader. No significant reduction in bee PLA2G2A activity was observed

following KH064, Varespladib, or LY311727 treatment (Fig 5.8A). Early literature

characterising PLA2G2A inhibitors suggests KH064 to be specific for human

PLA2G2A (Gregory et al., 2006). For this reason, human recombinant (hr) PLA2G2A

was then used to characterise the inhibitors. Given the role of PLA2G2A in

Figure 5.8 Small molecule inhibition of PLA2G2A activity

(A) PLA2G2A protein purified from bee venom was added to LNCaP cells in the

presence or absence of 30 µM KH064, Varespladib or LY311727. PLA2G2A activity

as a function of increasing absorbance was measured for one hour every 6 minutes

using a plate reader (n=2 biological and 3 technical replicates, mean±SD, One-way

ANOVA with Dunnett’s multiple comparisons test relative to bee PLA2G2A

control). (B) LNCaP cells were incubated for one hour with 50 ng hrPLA2G2A alone

or in the presence of 30 µM each KH064, Varespladib or LY311727, in addition to

NBD-PE. After fixation, cellular PE staining was measured by qFM (values show

mean fluorescence intensity (MFI) PE per cell, mean±SD of n~3000 cells from 3

wells, One-way ANOVA with Dunnett’s multiple comparisons test relative to DMSO

control (Ctrl), ****p<0.0001).


(A) PLA2G2A protein purified from bee venom was added to LNCaP cells in the

presence or absence of KH064, Varespladib or LY311727. PLA2G2A activity as a

function of increasing absorbance was measured for one hour every 6 minutes using

a plate reader (n=2 biological and 3 technical replicates, mean±SD, One-way

ANOVA with Dunnett’s multiple comparisons test relative to bee PLA2G2A

control). (B) LNCaP cells were incubated for one hour with 50 ng hrPLA2G2A alone

or in the presence of KH064, Varespladib or LY311727, in addition to NBD-PE. After

fixation, lipid uptake was measured by qFM (n~3000 cells from 3 wells, mean±SD,

One-way ANOVA with Dunnett’s multiple comparisons test relative to DMSO

control (Ctrl), ****p<0.0001).



lysophospholipid uptake described in Chapter 4, this assay was repeated in LNCaP

cells treated with hrPLA2G2A in the presence or absence of KH064, LY311727, and

Varespladib at 30 µM. This concentration was chosen based on the IC50 calculation

for KH064 obtained from cell-based assays (Fig. 5.7A-B). Addition of hrPLA2G2A

to LNCaP cells alone resulted in significantly increased cellular uptake of fluorophore

(NBD)-labelled PE (NBD-PE) (Fig. 5.8B). This suggested that hrPLA2G2A-mediated

hydrolysis of the sn-2 acyl ester and release of fluorophore-labelled lyso-PE resulted

in a lipid substrate that was more efficiently taken up by cells than the intact PE probe.

This stimulatory effect of hrPLA2G2A was significantly reduced by the addition of 30

µM KH064 and entirely blocked with 30 µM both LY311727 and Varespladib (Fig

5.8B). These findings confirm the inhibitory action of each selected PLA2G2A

inhibitor and their subsequent effect on lyso-PE uptake in LNCaP cells in vitro.

5.2.5 Growth inhibition with KH064 cannot be rescued by arachidonic acid

As mentioned previously, PLA2G2A acts on phospholipids to release

lysophospholipids and bioactive PUFAs including arachidonic acid for subsequent

cellular uptake or activation of signalling pathways. Arachidonic acid has previously

been shown to promote PCa growth (Chaudry et al., 1994) and has gained considerable

attention for its role in the generation of eicosanoids that serve as key mediators of the

inflammatory response (reviewed in (Yang et al., 2012)). To investigate if the growth

inhibition observed with 30 µM KH064 treatment over 96 hours was caused by the

loss of cellular supply of arachidonic acid, rescue experiments with the addition of

exogenous arachidonic acid were carried out. LNCaP cells were treated with 30 µM

KH064 and 1 µM or 10 µM arachidonic acid individually or in combination, and cell

growth as a function of confluence was measured using the IncuCyte live-cell imaging

system (Fig 5.9). Arachidonic acid alone had no significant effect on growth rates in

LNCaP cells at 1 µM and 10 µM. However, arachidonic acid at the concentrations

tested was unable to rescue the growth inhibition induced by KH064 (30 µM).


5.2.6 Investigation of PLA2G2A inhibition as a therapy against ATT-induced

resistance

Given the evidence that increased PLA2G2A transcript levels (Section 4.2.10) and

enzymatic activity (Section 4.2.11) are an early adaptive response to ATTs which

facilitates enhanced lipid uptake, it was speculated that KH064 would sensitise AR-

positive PCa cells to Enz treatment, and that this approach could ultimately be used as

a therapeutic strategy to fight ATT resistance and delay progression to CRPC. Tc

address this hypothesis, LNCaP cells were treated simultaneously with 30 µM KH064

and 10 µM Enz in androgen-replete medium or pre-treated with 30 µM KH064 for 48

hours before 10 µM Enz was added (indicated by red arrow in Fig 5.10A). Indeed,

simultaneous and consecutive co-treatments with 30 µM KH064 10 µM Enz

Figure 5.9 KH064 and arachidonic acid combination treatment

(A) LNCaP cells were treated with 30 µM KH064 alone or in combination with

arachidonic acid (1 µM and 10 µM). Confluence was measured every 2 hours using

the IncuCyte live-cell imaging system (n=2 biological and 3 technical replicates,


vehicle control, ****p<0.0001).




the IncuCyte live-cell imaging system (n=2, mean±SD, One-way ANOVA with

Dunnett’s multiple comparisons test relative to vehicle control, ****p<0.0001).




the IncuCyte live-cell imaging system (n=2, mean±SD, One-way ANOVA with

Dunnett’s multiple comparisons test relative to vehicle control, ****p<0.0001).


significantly decreased the proliferation of LNCaP cells (Fig 5.10A). Interestingly,

cotreatment induced morphological changes including flattening of LNCaP cells

compared to 0.1% DMSO following 120 hours treatment (Fig 5.10B). These

observations were absent in cotreatments with Enz and LY311727 (Fig 5.10C). Given

the effect on NBD-PE uptake (Fig. 5.8B) and the similar growth-suppressing action

when compared to siRNA gene silencing (Fig. 5.6C), KH064 was chosen from the

three PLA2G2A inhibitors to progress to pre-clinical in vivo models of ATT.


Figure 5.10 Co-targeting AR and PLA2G2A in PCa cells

(A) LNCaP cells were treated with Enz (10 µM) and KH064 (30 µM) alone or in

combination for 120 hours in androgen replete growth medium. Confluence was measured



****p<0.0001). Treatment groups are arranged in the order presented in the figure key. (B)

Images representative of (A). Scale bars=300 µm. (C) LNCaP cells were treated with Enz

(10 µM) and LY311727 (30 µM) alone or in combination for 120 hours. Confluence was

measured every 2 hours using the IncuCyte live-cell imaging system (n=3, mean±SD, One-

way ANOVA with Dunnett’s multiple comparisons test relative to vehicle control,

****p<0.0001).

Figure 5.10 Co-targeting AR and PLA2G2A in PCa cells

(A) LNCaP cells were treated with Enz (10 µM) and KH064 (30 µM) alone or in

combination for 120 hours in androgen replete growth medium. Confluence was measured


ANOVA with Dunnett’s multiple comparisons test relative to vehicle control,


5.2.7 Targeting PLA2G2A in an LNCaP tumour xenograft model of CRPC

progression

The above-mentioned in vitro results provide a strong rationale for investigating the

effect of PLA2G2A inhibition in an LNCaP tumour xenograft model of CRPC

progression. In the pilot study, 24x 4-week old NOD/SCID mice were inoculated with

2 million LNCaP cells by subcutaneous injection. Blood was collected (100 µl/week)

and PSA was measured once weekly. Tumours were allowed to grow until PSA had

reached 50 ng/mL, at which point mice were castrated. One week post-castration, mice

were be randomised into vehicle control (0.5% Carboxymethyl Cellulose + 2.5%

Tween-80) and KH064 treatment groups. KH064 was administered at 5 mg/kg/day by

intraperitoneal injection. Mice remained on treatment until the ethical endpoint was

reached, and tumours were collected along with serum for analysis. This work is

outlined in Fig 5.11.

Figure 5.11 Targeting PLA2G2A in an LNCaP tumour xenograft model of

CRPC progression

Figure 5.9 The role of lipid rafts in cell signalling. Figure from (Zalba & Ten

Hagen, 2017).Figure 5.11 Targeting PLA2G2A in an LNCaP tumour

xenograft model of CRPC progression

Figure 5.10 The role of lipid rafts in cell signalling. Figure from (Zalba & Ten

Hagen, 2017).Figure 5.11 Targeting PLA2G2A in an LNCaP tumour


Serum PLA2G2A from mice pre-castration, one-week following castration (nadir)

and at terminal endpoint was measured by ELISA. Serum PLA2G2A was below the

detectable limit (30 pg/mL) prior to castration but was significantly increased in final

serum collections (when tumour size reached approximately 1000mm3) (Fig 5.12A).

Furthermore, mice in the KH064 treatment group had significantly longer survival

times compared to mice in the vehicle group (Fig 5.12B).

Figure 5.12 Co-Targeting PLA2G2A and AR in vivo

(A) NOD/SCID mice were inoculated with LNCaP cells and tumours were allowed to grow

until serum PSA reached 50 ng/mL. Serum was collected on the day prior to castration

(PreCx), one week post-castration (nadir) and when tumours reached an ethical endpoint of

1000mm3 (Endpoint). Serum PLA2G2A was measured by ELISA (n=14, mean±SD, One-

way ANOVA with Dunnett’s multiple comparisons test relative to PreCx levels,

****p<0.0001). (B) One week post-castration, mice were randomised into KH064

(5mgs/kg/day, intraperitoneal injection; n=7) or vehicle control (100 µL 0.5%

Carboxymethyl Cellulose + 2.5% Tween-80 per day, intraperitoneal injection; n=7). Graph

shows days survival post-castration (mean±SD, Unpaired t-test, *p<0.05).


5.3 DISCUSSION

Dysregulation of lipid metabolic pathways is an emerging phenotype of therapy

resistance of several cancer types (Hangauer et al., 2017; Lue et al., 2017;

Vijayaraghavalu et al., 2012). Chapter 4 illustrates the role of phospholipase

PLA2G2A in the adaptive response of PCa cells to ATTs. A key observation in

Chapter 4 was that ATT induces a substantial remodelling of all major lipid classes

and enhances cellular lipid content, including phospholipids. Furthermore, it was

shown that enhanced lipid uptake fuelled lipid remodelling and that PLA2G2A’s

extracellular activity provided lysophospholipids which are the preferred substrate of

ATT-treated PCa cells for enhanced PL uptake. Given these observations, it was

hypothesised that PLA2G2A expression and activity was critical for the survival of

PCa cells undergoing ATTs by providing lysophospholipids for cellular uptake. While

PLA2G2A has recently been suggested as a serum biomarker for predicting PCa (Dong

et al., 2010; Leslie et al., 2012), its functional role in PCa as well as in the context of

ATTs and development to CRPC has yet to be explored.

Firstly, the clinical relevance of PLA2G2A in PCa was validated using 4

independent patient PCa datasets, in which higher PLA2G2A transcript levels were

observed in prostate tumours compared to normal prostate, and increased expression

correlated with increasing Gleason score. Next, serum from advanced PCa patients

was collected and used to measure PLA2G2A protein levels in men prior to and

following 12 weeks of ATT. No significant difference was observed in serum

PLA2G2A levels before and after ATT, however, baseline levels of serum PLA2G2A

in these patients was higher than reported levels in healthy patients. Notably, these

patients presented at clinic with treatment naïve metastatic PCa prior to commencing

ATT treatment and had a generally very high disease burden (Rhee et al., manuscript

in preparation). Serum PLA2G2A levels have been reported to be significantly higher

in metastatic PCa compared to localised tumours, with levels approaching 6 ng/mL in

metastatic patients (Menschikowski et al., 2012; Menschikowski et al., 2013). Given

the mean PLA2G2A level of 18.575 ng/mL in the patient cohort analysed in this study,

it can be speculated that levels would not be further increased by ATTs. A major caveat

of this study is that it did not include healthy controls for direct comparison of serum

PLA2G2A, resulting in the reliance on previously reported levels for control

comparisons (Menschikowski et al., 2012; Menschikowski et al., 2013). It would be


of extreme value to design a similar study, but including men with localised, advanced

PCa initially going on ATT in order to more accurately interrogate the effect of ADT

on serum PLA2G2A.

Next, further interrogation of the AR-axis revealed that androgens significantly

repress PLA2G2A levels in PCa cell lines in vitro. This is consistent with the observed

upregulation of PLA2G2A transcript and protein levels upon Enz or CSS treatment.

Given that the PLA2 family consists of several structurally distinct family members

with similar enzymatic activities (Murakami et al., 1998), it could be speculated that

it is the general phospholipase A2 activity, i.e. hydrolysis of phospholipids to release

bioactive fatty acids and lysophospholipids, that PCa cells become reliant on during

the early adaptive response to ATT. However, baseline expression of PLA2G2A is

higher than any other PLA2 family member across 3 different PCa cell lines. It is

possible that the list of phospholipases described here is limited, and additional

phospholipases may serve an important role in the lipid-rich phenotype observed in

the present model. However, the group of PLA2 members investigated in this study

was generated based on what could be detected by microarray and RNA-sequencing,

suggesting other family members are expressed at negligible levels.

In addition to already low baseline expression levels, no other PLA2 was found

to be androgen regulated. Furthermore, only one family member, PLA2G7, was

significantly upregulated in the 21-day ATT treatment, while no other PLA2 was

significantly altered in the LNCaP xenograft tumour model of CRPC. Together, it can

be concluded from this data that PLA2G2A is unique in its androgen regulation and

association with PCa, especially regarding its preferential upregulation in the adaptive

response to ATT. Given this conclusion, PLA2G2A was chosen to be further

characterised as a potential therapeutic target in PCa.

Silencing of PLA2G2A had a significant growth inhibitory effect in AR-positive

LNCaP cells but little to no effect on AR-negative DU145 or benign BPH1 cells,

further exemplifying its AR associated role in PCa. Intriguingly, PLA2G2A transcript

and PLA2G2A protein expression rarely showed similar results. This could be due to

the broad antigen-specificity of the PLA2 antibody recognising at least two more

PLA2s (PLA2G7 and PLA2G15), but this requires further investigation to confirm.

Indeed, upon siRNA silencing of PLA2G2A, LNCaP cells upregulated the transcript

levels of several other PLA2 members, presumably as a compensatory mechanism to

ensure sufficient access to lysophospholipid and FA substrates. The phospholipases


that the upregulated transcripts encode for, PLA2G7, PLA2G4, and PLA2G15, have

molecular weights of roughly 45, 47, and 85 kDa, respectively, meaning these are not

responsible for the increasing band observed at 14 kDa where PLA2G2A would be

expected. Each of the secreted sPLA2 group members shares this 14 kDa molecular

weight (Reid, 2005; Six & Dennis, 2000), leading to the speculation that expression of

another secreted enzyme is induced upon PLA2G2A silencing. However, as mentioned

previously, the upregulated PLA2s could not be detected in the PCa cell lines presented

here. This observation also begs the question of why PLA2G2A is specifically selected

by PCa cells if other sPLA2 members are able to provide similar products to fuel cell

survival. Future experiments using protein mass spectrometry could help to identify

low-abundance proteins and peptides and help to delineate the contribution of specific

sPLA2 family members in the context of PCa. Ultimately, this study is unable to

provide a conclusive explanation for the discrepancy between PLA2G2A transcript and

protein levels upon gene silencing. It is possible that transcript levels directly affect

protein processing and secretion, leading to an accumulation of cellular protein levels

upon gene silencing. If this were the case, it could be expected that overexpressing

PLA2G2A in cell lines with low baseline PLA2G2A levels would result in lower

cellular PLA2G2A protein levels due to increased secretion of PLA2G2A. This

experiment, along with further knockdown studies of individual PLA2 family members

are warranted to address this question. Other aspects of PLA2G2A function currently

unknown are its mechanism of secretion, the existence of a signal peptide and function

of its posttranslational modifications.

While to the best of our knowledge targeted inhibition of PLA2G2A has yet to

be explored as a PCa therapeutic, several selective inhibitors have been developed and

investigated in a variety of physiological conditions including arthritis, bone loss, and

cardiac fibrosis, amongst others [reviewed in (Reid, 2005)]. The previously confirmed

safety and tolerance of these drugs in vivo make them an attractive tool for repurposing

in PCa models; therefore, three selective inhibitors were chosen to characterise for

their potential use in an in vivo model of ATT. It is important to note that the IC50

values reported in literature (Reid, 2005) differ from the IC50 value measured in

KH064 in this study, however the 30 µM value calculated here was the concentration

required to cause 50% growth inhibition in a cell-based assay; enzymatic inhibition is

likely to occur at much lower concentrations (Reid, 2005). Additionally, each of the

three inhibitors is known to be selective for human PLA2G2A (Reid, 2005), however


it is likely that they also target other secreted PLA2 enzymes. Given that PLA2G2A

was the main enzyme of interest in this study and that other PLA2s did not seem to

play a role in the context of androgen replete growth conditions and ATT, any

additional PLA2 enzymatic inhibition was assumed irrelevant. Varespladib has

progressed into human clinical trials (Reid, 2005; Ferri, Ricci & Corsini, 2015),

including Phase 3 multi-institutional testing. It proved to be ineffective in treating

patients with severe sepsis, however it had acceptable safety profile and was able to

ablate PLA2G2A enzymatic activity (Reid, 2005). In the in vitro characterisation

described in this study, KH064 was the only inhibitor to have a growth inhibitory effect

in PCa cells. Furthermore, neither LY311727 nor Varespladib achieved an additive

growth inhibition when combined with Enz treatment, as seen with KH064. Given that

PLA2G2A possesses both enzymatic and enzyme-independent ligand binding

activities (Brglez et al., 2014), it could be possible that these inhibitors target each of

these pathways differently, and only KH064 affects growth and survival of PCa cells

due to its effect on both pathways. Consistent with this notion, KH064 reduced

transcript levels of PLA2G2A. The effect of these drugs on PLA2G2A gene regulation,

as well as their ability to penetrate the cell membrane versus solely inhibiting

extracellular enzymatic activity, could contribute to their differential growth effects.

Future studies are required in order to further characterise the difference in

mechanisms between the three inhibitors. This study showed that the growth inhibition

in LNCaP cells following 30 µM KH064 treatment could not be rescued by exogenous

arachidonic acid. These results suggest that cellular supply with arachidonic acid

through phospholipid hydrolysis is not the major role of PLA2G2A activity in PCa

cells. Thus, lysophospholipid production by PLA2G2A activity may provide a more

important survival benefit. Additionally, it is has been suggested that cPLA2 enzymes

are predominantly responsible for arachidonic acid release within the cell given their

preference for arachidonyl lipids, whereas PLA2G2A shows less specificity for the

type of fatty acyl chain present at the sn-2 position (Mayer & Marshall, 1993; Reid,

2005), providing further support that PLA2G2A upregulation is not due to increased

demand for arachidonic acid. This hypothesis could be further tested with future rescue

experiments using exogenous lysophospholipid species in combination with

PLA2G2A inhibition.

Given its strong growth inhibitory action and additive effect with Enz (Fig.

5.10A), KH064 was chosen to test in an in vivo LNCaP xenograft model of CRPC


progression after castration to investigate its potential as a therapeutic target in the

context of fighting ATT resistance and development to CRPC.

Data from our pilot in vivo study suggested that androgen-deprivation therapy,

i.e., surgical castration, significantly increased serum PLA2G2A in mice bearing

LNCaP tumours compared to pre-castrate levels. Furthermore, PLA2G2A inhibition

by KH064 (5mgs/kg/day) resulted in longer overall survival times compared to

vehicle treated mice. Ongoing analyses include tumour analysis for PCa markers

(AR and PSA), transcript levels of PLA2 family members and key players involved

in previously described lipid remodelling pathways, and lipid mass spectrometry of

tumours from vehicle and KH064 treatment groups. Nevertheless, the increased

survival time in KH064 treated mice suggest that PLA2G2A may serve as a novel

therapeutic strategy to use in combination with ATT in order to delay disease

progression.

Overall discussion and future directions 168

Overall discussion and future directions

6.1 DELINEATION OF THE LIPID TRANSPORTER LANDSCAPE AND

ANDROGEN REGULATION OF LIPID UPTAKE IN PCA

The first aim of the present study was to investigate the androgen regulation of

exogenous lipid uptake, as well as to delineate the lipid transporter landscape in PCa

cells. A major limitation in the targeting of PCa lipid supply is the incomplete

understanding of the contribution of different carbon sources to total cellular biomass.

It is unclear how much relative contribution comes from lipid uptake (e.g. lipid

transporters, TNTs, macropinocytosis, passive diffusion) versus lipogenesis, and this

knowledge gap has been due to technical challenges, complexity of lipid cargos (e.g.

lipoprotein particles which carry several different lipid classes), and system

redundancies (e.g. FA uptake via multiple routes as listed above).

However recent advancements in the field have shown that up to 70% of cellular

lipid carbon biomass in lung cancer (Hosios et al., 2016) and 83% of lipid carbon

biomass in prostate cancer (Balaban et al., 2019) was generated from exogenous fatty

acids, prompting the further investigation of lipid uptake in PCa shown here. Notably,

both studies estimated the biomass contribution of exogenous lipids in comparison to

synthesis by tracing the lipid biomass incorporation of just two fatty acids (palmitate

and oleate) provided as labelled free FAs. Although palmitate and oleate make up 70%

of fatty acids distributed across multiple lipid species in serum, only ~4% are present

as free fatty acids. Thus, considering the complexity of exogenous lipid cargos and

that multiple cargo-selective and non-selective lipid uptake mechanisms are utilised

by PCa cells, the actual contribution of exogenous FAs might be even higher than 70-

80%. In addition, these observations warrant similar studies to address the biomass

contribution of exogenous cholesterol uptake compared to synthesis.

De novo lipogenesis is a well-described AR-regulated pathway contributing to

PCa incidence and progression (Brusselmans & Swinnen, 2009; Swinnen et al., 2004;

Swinnen et al., 2002b; Swinnen et al., 2006; Zadra et al., 2013), however therapeutic

targeting of lipogenic enzymes such as FASN has had limited clinical success.

Furthermore, the contribution of uptake of exogenous lipids in PCa has been widely


underappreciated. The work described in this study revealed a strong AR-regulation

of uptake of several lipid substrates, i.e. free fatty acids, cholesterol and lipoprotein

complexes. Androgens also regulated gene expression of numerous lipid transporters.

Clinical validation of lipid transporters in PCa patients revealed increased lipid

transporter mRNA and protein expression in patients with bone metastatic disease

compared to localised cancer (Fig. 3.4). This study highlights for the first time the

previously underappreciated importance of lipid uptake to PCa proliferation and

survival. The demand for increased lipid accumulation in PCa cells represents a

metabolic vulnerability that should continue to be exploited as a therapeutic target,

however targeting a single transporter is unlikely to be curative due to system

redundancies and lipid supply plasticity.

While androgen regulation of lipogenesis is widely accepted (Swinnen et al.,

2004), this study, supported by the work described by Hosios et al. (2016) and Balaban

et al. (2019), suggest that lipid uptake is a highly utilised lipid supply pathway in PCa

cells and that this is an androgen-regulated process. Furthermore, by demonstrating

the importance of lipid uptake in PCa progression, this novel contribution to the field

provides a strong rationale to better understand exogenous lipid supply routes and to

re-evaluate the way lipid metabolism is currently targeted in PCa. This new knowledge

is likely to provide insight into lipid uptake in other types of cancer. While this study

provides the most comprehensive evaluation of lipid transporters in PCa to date,

further work is required to fully understand the relative contribution and androgen-

regulation of various lipid supply routes. One way in which this could be addressed

would be through an isotopic labelling approach with a complex lipid mixture.

Lipidome isotope labelling of yeast (LILY) is a recently reported, elegant approach to

generate a complex spectrum of various lipid species to measure the uptake and

incorporation of extracellular lipids in biomass both in vitro and in vivo (Rampler et

al., 2017). By supplementing growth media with a 13C labelled lipidome in

combination with 14C lactate, acetate, glucose, or glutamine, the relative contribution

of exogenous lipid uptake versus de novo synthesis from alternate carbon sources

could be delineated. This knowledge would help in the development of therapeutic

strategies targeting lipid metabolism in PCa.


6.2 LIPID REMODELLING IS A NOVEL ADAPTIVE PHENOTYPE IN

RESPONSE TO ATTS

The transcriptional characterisation of the long-term in vitro ATT model

described in the present study highlighted the therapy-induced dynamic rewiring of

metabolic networks by PCa cells. This model of therapy stress captures both the

immediate and delayed response, which showed that the response to ATTs is not static

but rather a complex process of dynamically changing pathways contributing to the

adaptation, survival and ultimate emergence of drug-resistance, i.e. when cells begin

growing again despite the presence of ATTs. These kinetic differences could help

guide treatments by highlighting when to target specific metabolic vulnerabilities

(therapeutic window) in order to most efficiently fight the emergence of CRPC.

Through the integration of multiple ‘omic platforms, lipid accumulation and

lipid remodelling were identified as significantly upregulated pathways induced by

ATTs. Lipid accumulation has recently been shown to protect cancer cells against

environmental stress such as hypoxia and nutrient depletion (Cabodevilla et al., 2013;

Koizume & Miyagi, 2016; Petan et al., 2018), as well as anti-cancer treatments in

several cancer types (Lue et al., 2017; Vijayaraghavalu et al., 2012), suggesting this

phenotype may extend beyond ATTs in PCa and may instead be a more global adaptive

response to anti-cancer therapies. There are several plausible explanations for the

increased lipid accumulation observed both here and in other cancer types. First, the

protective function of LDs may be due to their role in preventing lipotoxicity, ROS

damage and ER stress (Chitraju et al., 2017; Listenberger et al., 2003). Indeed, the

present study showed increased lipid peroxidation as a response to long-term ATT,

which could result in the demand for increased lipid droplet formation in order to

reduce the availability of exposed PUFAs to ROS damage, given that the location and

composition of PUFAs is critical for their role in ferroptosis (Ayala et al., 2014).

Second, altered lipid membrane composition affects membrane permeability and

fluidity (van Meer et al., 2008). In this case, the increased PUFA content described in

the present ATT model would likely result in a more fluid membrane, which could

play a role in membrane permeability as well as in migration and invasion of cancer

cells. Thirdly, LDs could serve as an energy reserve for cells, however this is not the

most likely function given that characterisation of the long-term ATT model revealed

decreased mitochondrial activity, oxidative phosphorylation and ATP production.


One intriguing observation was the depletion of TAGs and CEs in LDs. This

raises a major question in the present study: what is the nature of the lipid species

responsible for enhanced lipid content in LDs? Traditionally, LDs are described to be

primarily composed of TAGs and CEs, however both species are depleted in the

present ATT model. This led to the investigation of additional LD constituents.

Recently it was shown that free ceramides can be acylated and stored away in LDs

(Senkal et al., 2017), and this pathway was transcriptionally upregulated in the present

ATT model. If not stored in LDs, free ceramides can pack tightly with free cholesterol

to form lipid rafts that serve as major signalling domains. Ceramide-enriched lipid rafts

have been shown to initiate apoptosis signalling (Verheij et al., 1996; Zalba & Ten

Hagen, 2017), whereas lipid rafts enriched in cholesterol and other sphingolipids

promote cell proliferation and survival (Fig 6.1).

Here it is proposed that ceramides are acylated and stored in LDs in order to reduce

apoptotic signalling, resulting in increased LD accumulation and increased free

cholesterol and sphingolipids in the cell membrane. The ATT study described here

was limited by the detectable lipid species in the present LCMS analysis and warrants

further investigation into the lipid profile of subcellular organelle fractions. With

Figure 6.1 The role of lipid rafts in cell signalling

(A) Cholesterol and sphingolipid rich domains (green) promote proliferation, survival

and angiogenesis, while (B) ceramide enriched lipid rafts (red) promote apoptosis.

Figure from Zalba & Ten (2017).


(A) Cholesterol and sphingolipid rich domains promote proliferation, survival and

angiogenesis, while (B) ceramide enriched lipid rafts promote apoptosis. Figure from

Zalba & Ten (2017).


(A) Cholesterol and sphingolipid rich domains promote proliferation, survival and

angiogenesis, while (B) ceramide enriched lipid rafts promote apoptosis. Figure from

Zalba & Ten (2017).


emerging technical advances in the field of lipidomics, an extensive investigation of

LD composition in the context of anti-cancer therapy would be of extreme value.

Emerging evidence suggests that lipid remodelling in response to cancer

therapies is not confined to PCa but is also in other cancer types including breast cancer

and renal cell carcinoma (Lue et al., 2017; Vijayaraghavalu et al., 2012). Furthermore,

our ongoing collaboration revealed a similar PUFA-rich phenotype accompanied by

increased lipid peroxidation levels in both melanoma and lung cancer cell lines

following targeted anti-cancer therapies, and this phenotype was largely reversed once

cell became resistant (Dr. Helmut Schaider, personal communication). The in vitro

PCa model described here captures the early adaptive response to ATTs, however what

remains unknown is the activity of these pathways once PCa cells start growing again,

i.e. CRPC. Multiple studies demonstrated that treatment of various AR-positive PCa

cell lines for several months with ATTs generated resistant, proliferating cell lines (Lu

et al., 1999; Xu et al., 2010; Yu et al., 2017), but these studies have not characterised

the early or late changes to the lipidome associated with low and high proliferation,

respectively. By extending our treatment protocol to several months and monitoring

when cells resume proliferation, longitudinal analyses of lipid metabolism pathways

(DNL, uptake, and lipid remodelling) could further inform the dynamic nature of

therapy-induced lipid remodelling and the therapeutic window for novel co-

treatments. Here it is hypothesised that lipid membrane remodelling during this

adaptive response phase allows for altered signalling pathways via changes in

membrane fluidity and composition, which could eventually lead to disease

progression and metastasis. If this early response is a survival mechanism to evade the

initial oncogenic signalling blockade, then this pathway could be a potential target in

order to prevent disease progression. The discovery that therapy-induced lipid-

remodelling extends beyond PCa and may be a more general response to anti-cancer

therapies (Dr. Helmut Schaider, personal communication is a significant finding with

major implications in the field of therapy resistance.


6.3 LIPID UPTAKE IS THE MAJOR CONTRIBUTOR TO THE

INCREASED LIPID ACCUMULATION INDUCED BY ATT-

TREATMENT

Decades of research have focused on DNL in PCa incidence and progression,

while the contribution of exogenous lipids remains largely underappreciated (Heemers

et al., 2001; Swinnen et al., 2004; Swinnen et al., 2002a; Swinnen et al., 2000). Here

it was shown that DNL was downregulated in the adaptive response to ATTs, while

enhanced lipid uptake served as a major lipid supply route that provided PCa cells with

fuel, membrane material and signalling molecules. Previous work from our laboratory

and others showed that the anti-tumourigenic effect of DNL inhibition can be rescued

by the addition of exogenous lipids (Dr. Sadowski, unpublished), and this presents a

limitation in current treatment strategies. Only in the past few years have researchers

started to investigate lipid uptake in PCa (Balaban et al., 2019; Gaston et al., 2017;

Tousignant et al., 2019), and never in the context of ATTs, making this the first study

of its kind. To develop effective therapeutics targeting lipid metabolism in PCa, the

crosstalk between lipid uptake and synthesis must be acknowledged. This work for the

first time challenges the dogma around lipid supply primarily by lipogenesis in PCa

(Swinnen et al., 2000; Swinnen et al., 2004) and provides a strong rationale to

investigate lipid uptake as a therapeutic co-target in combination with lipogenesis

inhibitors.

A major limitation in targeting lipid metabolism in cancer cells is the lack of

understanding of lipid supply plasticity, i.e. the ability to cancer cells to dynamically

engage different lipid supply pathways (synthesis and uptake via lipid transporters,

macropinocytosis and tunnelling nanotubes) under nutrient stress. The androgen-

regulation of lipid transporters in PCa has been described in this study, however little

is known about alternative regulatory and sensory mechanisms that coordinate the

interplay between lipid uptake and de novo synthesis. However, this work highlights

the importance of acknowledging both supply pathways in order to efficiently block

cellular lipid accumulation. Preliminary data acquired in our laboratory showed that

co-targeting de novo synthesis of cholesterol with Simvastatin (HMGCR) and access

to exogenous cholesterol by inhibition of lysosomal cholesterol efflux pump NPC1

with U18666A resulted in significant synergistic growth inhibition in PCa cells in vitro

(Dr. Sadowski, unpublished). Interestingly, individual inhibition of HMGCR or NPC1


induced very similar transcriptional responses, e.g. upregulation of genes encoding

lipid transporters (SCARB1 and LDLR) and synthesis enzymes (HMGCR, HMGCS),

suggesting that depletion of cellular cholesterol levels triggers a coordinated increase

of both supply pathways. These observations warrant for future investigations of lipid

supply plasticity to design novel co-treatment strategies and their testing in pre-clinical

in vivo models of PCa and CRPC progression.

6.4 MECHANISTIC INSIGHTS INTO THE ROLE OF SECRETED

PHOSPHOLIPASE PLA2G2A IN PROSTATE CANCER

Through investigation of the transcriptome and proteome, it was discovered that

PLA2G2A plays a major role in the adaptive response to ATTs. PLA2G2A is unique

in its androgen suppression and induction by ATTs compared to other members of

both the secreted PLA2 and cytosolic PLA2 family, representing a novel therapeutic

target enhanced by ATTs in PCa. It was shown here that PLA2G2A contributed to

ATT-induced lipid remodelling by providing lysophospholipids and bioactive PUFAs

to potentially serve as a substrate for membrane lipid synthesis or to initiate one of

several key pathways induced by lipid mediators derived from PLA2G2A products.

However, there are additional hypotheses that could be explored regarding the role of

PLA2G2A in PCa. In the present study, the role of PLA2G2A on migration and

invasion was not explored but would be one avenue worth investigating. Previous

studies have shown that leptin and PLA2G2A act together to induce migration in

astrocytoma cells (Martín, Cordova, Gutiérrez, Hernández, & Nieto, 2017). Like

PLA2G2A, leptin levels are strongly upregulated by ATTs (Basaria, Muller, Carducci,

Egan, & Dobs, 2006). Additional work has started to highlight the contribution of

arachidonic acid and poly-unsaturated lysophospholipids to increased membrane

fluidity and resulting migration of cancer cells (Raynor et al., 2015; Tallima & El Ridi,

2018). It is plausible that through PLA2G2A-induced membrane remodelling, cells

acquire enhanced membrane fluidity, which in turn could facilitate migration and

invasion. Notably, serum PLA2G2A is dramatically increased in patients with

metastatic PCa compared to localised or benign tumours (Dong et al., 2010; Leslie et

al., 2012) as well as in the patient serum samples measured in the present study

(Chapter 5 Fig 1). This study suggested that PLA2G2A provides PUFAs for cellular

uptake and incorporation into membranes, which, as mentioned previously, could


confer higher membrane fluidity and likely enhanced migratory capabilities. Due to

time limitations, this hypothesis has not yet been tested, but is a promising avenue for

future investigations.

Lastly, PLA2G2A is largely known for its role in initiating inflammatory

pathways, reviewed in (Brglez et al., 2014). Briefly, the arachidonic acid and other

bioactive PUFAs released by PLA2G2A serve as precursors for a number of lipid

mediators that could be involved in both pro- and anti-inflammatory responses. In this

study, a protocol to measure lipid mediators in serum via GCMS was tested.

Unfortunately, the extraction protocol failed to produce detectable levels of lipid

mediators. As a result, this study was unable to comment on the contribution of

PLA2G2A to lipid mediator pathways induced by ATTs. Further optimisation of the

lipid mediator extraction and GCMS protocols would be of interest for future studies.

However, cytosolic cPLA2 rather than sPLA2 is believed to be the major contributor

to the production of arachidonic acid metabolites such as eicosanoids (Yedgar,

Lichtenberg, & Schnitzer, 2000), and because cPLA2 expression was only modestly

affected by ATTs, it is unlikely that this is the main role of ATT-induced PLA2G2A

upregulation. PLA2G2A can also function independently of its enzymatic activity by

serving as a ligand for its cognate receptor, PLA2R (Brglez et al., 2014). However,

like cPLA2, PLA2R expression was only modestly affected by ATTs, further

supporting that enzymatic release of lipid substrates is the main function of PLA2G2A

in the described model.

Taken together, the data shown in the present study delineated a functional role

for enhanced PLA2G2A activity in ATT-treated cells, i.e. enhanced lysophospholipid

uptake, and mechanistically linked PLA2G2A to ATT-induced lipid remodelling.

However, lipid remodelling and lysophospholipid uptake should be further

investigated in order to fully elucidate the role of PLA2G2A in the early adaptive

response to ATTs. Furthermore, the mechanism of cellular PUFA uptake after

PLA2G2A catalysis also warrants clarification given that palmitate uptake (Bodipy-

C16:0) was strongly reduced by Enz treatment.


6.5 PLA2G2A REPRESENTS A NOVEL THERAPEUTIC TARGET TO

COMBAT ATT-INDUCED LIPID REMODELLING AND DELAY

PROGRESSION TO CRPC

Identifying and targeting the early adaptive response pathways activated by

ATTs is critical in fighting the development of drug resistance and progression to

CRPC. The present study presents PLA2G2A as a major contributor to the metabolic

rewiring induced by ATTs and suggests that exploiting this pathway in combination

with ATTs represents a novel therapeutic strategy. There are currently several

clinically approved drugs, as described in Chapter 5, which make it an attractive

therapeutic option. Furthermore, the model described in this study suggested that

ATTs induced the production and secretion of PLA2G2A into the extracellular space,

which subsequently provided ATT-challenged PCa cells with lipid substrates to fuel

lipid remodelling and support survival. Each of the three drugs described in this study

(KH064, LY311727, and Varespladib) have been shown to ablate serum PLA2G2A

activity in patient samples (summarised in (Reid, 2005)), and have been characterised

in in vitro models described in Chapter 5 of the present study. Indeed, inhibitors that

do not enter the cell are proving most effective at ablating PLA2G2A activity (Yedgar

et al., 2000, Reid, 2005), which removes the challenge of selective uptake by cancer

cells. While PLA2G2A is highly expressed in humans in the gastrointestinal tract,

prostate epithelium, and many cells of the immune system including neutrophils, mast

cells and macrophages, selective inhibitors of PLA2G2A have proven to be well-

tolerated both in mice and in pre-clinical studies of inflammatory diseases (Reid,

2005). The ongoing in vivo study described in Chapter 5 should help to address the

therapeutic potential of PLA2G2A inhibition on PCa cell growth in ATT-treated mice.

In addition to serving as a therapeutic target, future studies should investigate

serum PLA2G2A as a biomarker of PCa. To do this, a large patient cohort consisting

of healthy controls, PCa patients of various disease grades, and patients pre- and post-

ATT would be required. This could help characterise the sensitivity and specificity of

PLA2G2A as a novel non-invasive biomarker of PCa disease severity and response to

ATTs. One potential caveat of using PLA2G2A as a PCa biomarker arises from the

role of PLA2G2A in the early inflammatory response. Both pro- and anti-

inflammatory actions of PLA2G2A have been described across various cancers, and

prostatic inflammation is thought to be associated with increased risk for PCa


incidence (reviewed in (Sfanos & De Marzo, 2012)). This would make it difficult to

attribute the increased PLA2G2A in PCa patient serum to chronic inflammation or to

the tumour itself. This could be addressed by measuring C-reactive protein (CRP)

levels in men with PCa to quantitate inflammation, followed by a correlation analysis

between CRP and PLA2G2A levels. Furthermore, in addition to serum levels,

PLA2G2A transcript in the tumour could be measured in PCa patients and compared

to disease grade in order to distinguish between tumour specific PLA2G2A and

systemic inflammatory mechanisms.

A unique therapeutic approach worth exploring would be to exploit the enhanced

PLA2G2A activity rather than inhibit it. This approach would involve the development

of a synthetic phospholipid analogue which, when released by PLA2G2A activity,

releases the active lyso-PL drug. This concept is based upon a current compound,

Edelfosine, which is a synthetic alkyl-lysophospholipid that has been shown to be

selectively toxic to cancer cells following the uptake and incorporation of the drug into

cell membranes (Czyz et al., 2013). It is proposed that Edelfosine disrupts membrane

structure and signalling as well as induces ER stress (Udayakumar et al., 2016), both

of which result in increased apoptosis. In this scenario, the increased PLA2G2A

activity induced by ATTs could hypothetically be taken advantage of by increasing the

release and uptake of the lyso-PL drug, thereby selectively killing PCa cells. For this

to be successful, it would be critical to identify the lipid species in which PLA2G2A

has the highest specificity towards in order to design a prodrug that most efficiently

exploits enhanced PLA2G2A activity therapeutically.

6.6 SUMMARY

In summary, the long-term in vitro ATT model revealed lipid metabolic rewiring

as an early adaptive response. These dynamic pathways could be exploited as novel

therapeutic targets to complement current ATTs and delay progression to CRPC.

Through the integration of multiple ‘omics platforms, an extensive analysis of both

passive and active lipid metabolism via quantitative single cell imaging, and cell

proliferation and viability assays, the following key findings are presented:

1) Lipid uptake is an androgen-regulated process that contributes to PCa

progression


2) ATTs induce rewiring of lipid metabolic networks which include enhanced

lipid uptake and lipid remodelling.

3) ATTs induce expression and secretion of PLA2G2A, which in turn cleaves

phospholipids for subsequent uptake by PCa cells. This pathway can be

exploited as a novel therapeutic target in the fight against the adaptive

response to ATTs.

Collectively, this study expanded our current understanding of lipid

metabolism in PCa and provided novel insights into mechanisms involved in lipid

supply routes and the lipid accumulation preceding CRPC. The described lipid

remodelling helps to facilitate cell survival in ATT-treated PCa cells, representing a

major metabolic vulnerability to fight the emergence of drug resistance and

progression to CRPC. Furthermore, this metabolic vulnerability may extend beyond

prostate cancer and represent a mechanism of therapy resistance in several cancer

types. Indeed, through an ongoing collaboration with Dr. Schaider and colleagues, it

was found that the increased PUFA content and subsequent enhanced sensitivity to

GPX4 inhibition in ATT cells (described in Chapter 4) was also observed in lung

cancer and melanoma cell lines treated with targeted therapies. One of the major

challenges in cancer treatment today stems not from the lack of effective anticancer

agents, but rather the phenomenon of acquired therapy resistance that ultimately results

in relapse (Hendrich et al., 2003; Ramirez et al., 2016; Hultsch et al., 2018). Here, the

identification of enhanced dependence on GPX4 during the acquired resistance to

anticancer therapies provides a novel area of therapeutic intervention in order to

overcome this challenge and delay disease progression.

Appendices 179

Appendices

Appendix A: Characterisation of long-term ATT model in LNCaP cells

Appendices 180

Fig A1 Characterisation of the effects of chronic Enz treatment on LNCaP cells

LNCaP cells were grown for up to 21 days in FBS supplemented with Enz (10 µM) or

0.1% DMSO (D0 Enz), or in CSS. (A) AR mRNA expression was measured by qRT-

PCR. (B) PSA was measured by microarray analysis. (C) ATP production was

measured using CellTiter-Glo® Cell Viability Assay. (D) Prior to fixation, cells were

incubated with MitoTracker for one hour and images were analysed by qFM (n>1000

cells, image representative of 3 independent experiments). (E) Cells were stained with

propidium iodide and Hoescht and percent cell death was measured by qFM. (F)

Metabolic activity of cells was measured by PrestoBlue Cell Viability Reagent. (G)

Confluence was measured every 2 hours using the IncuCyte live-cell imaging system

following 10 µM Enz (left) or CSS (right) treatment. (n=3 (with the exception of (D)),


control cells, *p<0.05 **<0.01 ***<0.001 ****p<0.0001)








propidium iodide and Hoescht and percent cell death was measured by qFM. (F)

Metabolic activity of cells was measured by PrestoBlue Cell Viability Reagent. (G)

Confluence was measured every 2 hours using the IncuCyte live-cell imaging system

following 10 µM Enz (left) or CSS (right) treatment. (n=3 (with the exception of (D)),


control cells, *p<0.05 **<0.01 ***<0.001 ****p<0.0001)








Appendices 181

Appendices 182

Fig A2 Androgen regulation of lipid transporter transcript levels in DuCaP and

VCaP cells

(B) DuCaP (top) and VCaP (bottom) cells were grown in CSS for 48 hours and treated

with 10 nM DHT in the absence or presence of Enz (10 µM) or 0.1% ethanol vehicle

(Ctrl) for 48 hours. Indicated lipid transporter gene expression was analysed by

RNAseq and heatmaps were generated with a hierarchical clustering algorithm using

completed linkage and Euclidean distance measures and were scaled by row z score

(red=positive z score, blue=negative z score).

Appendices 183

Fig A3 LDLR and SCARB1 in PCa cells

Full Western blot described in Fig 3.4C. Image representative

of three individual experiments.

Fig A4 Androgen regulation of LDLR and SCARB1 in LNCaP

cells

Full Western blot described in Fig 3.6B. Image shows 3 independent

experiments.

Appendices 184

Protein ID Protein symbol

P46821 MAP1B

Q8NFW8 NEUA

O60220 TIM8A

Q16555 DPYL2

P30044 PRDX5

Q7KZF4 SND1

P26583 HMGB2

P35270 SPRE

P25398 RS12

O60832 DKC1

P52597 HNRPF

Q14444 CAPR1

O60506 HNRPQ

Q13554 KCC2B

P15531 NDKA

Q9Y646 CBPQ

O43143 DHX15

P17050 NAGAB

Q15365 PCBP1

P05387 RLA2

P07686 HEXB

Q09028 RBBP4

P04066 FUCO

Q8NC51 PAIRB

Q9UM54 MYO6

P34913 HYES

O14773 TPP1

O60869 EDF1

P16949 STMN1

P07858 CATB

Q96BJ3 AIDA

Q14247 SRC8

Q9HAT2 SIAE

P62851 RS25

P55209 NP1L1

Q9H0E2 TOLIP

P62269 RS18

P20290 BTF3

Q08380 LG3BP

Q6P5R6 RL22L

P10253 LYAG

P24534 EF1B

P05386 RLA1

Q13247 SRSF6

Q9UHL4 DPP2

P04350 TBB4A

P05783 K1C18

Q13509 TBB3

P05787 K2C8

P10155 RO60

P39748 FEN1

P62829 RL23

P10619 PPGB

P61758 PFD3

P18510 IL1RA

P22392 NDKB

P09104 ENOG

P35580 MYH10

Q15555 MARE2

Q13510 ASAH1

P16403 H12

P07339 CATD

P49321 NASP

P23919 KTHY

O43583 DENR

Q9P016 THYN1

Q99471 PFD5

Q9BQ69 MACD1

P31948 STIP1

Q9Y6E2 BZW2

P17900 SAP3

Q8NHP8 PLBL2

P16401 H15

P62314 SMD1

P33316 DUT

Q8N0W3 FCSK

Q9UFN0 NPS3A

P52758 RIDA

Q15020 SART3

Q13310 PABP4

Q13442 HAP28

Q08211 DHX9

E9PAV3 NACAM

P58107 EPIPL

Q96Q11 TRNT1

Q13185 CBX3

P63241 IF5A1

P33991 MCM4

P07305 H10

Q01459 DIAC

Q96CW1 AP2M1

P43487 RANG

Q9Y224 RTRAF

O00571 DDX3X

P62249 RS16

P15586 GNS

P08236 BGLR

P62277 RS13

P06865 HEXA

Q00577 PURA

Fig A5 Top 100 deregulated protein IDs measured by mass spectrometry

Protein IDs shown in Fig. 4.5C with corresponding protein symbol were identified

using UniProt.

Appendices 185

Fig A6 PLA2G2A in LNCaP cells following up to 21 days Enz treatment

Full Western blot described in Fig 4.14B. Image shows 3 independent experiments.

Appendices 186

Appendix B: Resources and funding

The majority of equipment and specialist facilities required for this project were

accessed at Translational Research Institute (TRI). Outside resources that were utilized

include lipid mass spectrometry (Steve Blanksby, QUT) and radio-isotope labelling

(Johan Swinnen, University of Leuven, Belgium). Funding for the project was

provided by the Movember Revolutionary Team Award “Adaptive Response to

Androgen Targeted Therapies, An Offensive to Resistance”.

Appendix C: Coursework

IFN001 AIRS has been completed. REIS quiz 1 and 2 have been completed and

submitted to QUT at the time of Stage 2 submission. No additional coursework is

necessary.

Appendix D: Collaborative Arrangements

I have worked primarily with members of APCRC-Q. We also have collaborative

arrangements with Prof Steve Blanksby for his expertise in lipid mass spectrometry,

and Johannes Swinnen for his expertise in radio-isotope labelling of lipids derived

from lipogenesis. All other work was completed at TRI.

Appendix E: Intellectual Property

IP forms have been completed and submitted to QUT at the time of Stage 2 submission.

Acknowledgements

We would like to thank Prof Dr Bart Ghesquière, Mr. Abel Acosta Sanchez and the

Metabolomics Expertise Center from the department of Oncology (KU Leuven) and

the Center for Cancer Biology (CCB, VIB) for the metabolomics analyses.

Bibliography 187

Bibliography

Agmon, E., & Stockwell, B. R. (2017). Lipid homeostasis and regulated cell death.

Current opinion in chemical biology, 39, 83-89.

doi:10.1016/j.cbpa.2017.06.002

Allott, E. H., Masko, E. M., Freedland, A. R., Macias, E., Pelton, K., Solomon, K. R.,

Mostaghel, E. A., Thomas, G. V., Pizzo, S. V., Freeman, M. R., & Freedland,

S. J. (2018). Serum cholesterol levels and tumor growth in a PTEN-null

transgenic mouse model of prostate cancer. Prostate Cancer Prostatic Dis,

21(2), 196-203. doi:10.1038/s41391-018-0045-x

Andersen, J. P., Vestergaard, A. L., Mikkelsen, S. A., Mogensen, L. S., Chalat, M., &

Molday, R. S. (2016). P4-ATPases as Phospholipid Flippases—Structure,

Function, and Enigmas. Frontiers in Physiology, 7(275).

doi:10.3389/fphys.2016.00275

Anderson, C. M., & Stahl, A. (2013). SLC27 fatty acid transport proteins. Molecular

aspects of medicine, 34(2-3), 516-528. doi:10.1016/j.mam.2012.07.010

Armstrong, C. L., Marquardt, D., Dies, H., Kučerka, N., Yamani, Z., Harroun, T. A.,

Katsaras, J., Shi, A.-C., & Rheinstädter, M. C. (2013). The Observation of

Highly Ordered Domains in Membranes with Cholesterol. PLOS ONE, 8(6),

e66162. doi:10.1371/journal.pone.0066162

Ayala, A., Muñoz, M. F., & Argüelles, S. (2014). Lipid peroxidation: production,

metabolism, and signaling mechanisms of malondialdehyde and 4-hydroxy-2-

nonenal. Oxidative medicine and cellular longevity, 2014, 360438-360438.

doi:10.1155/2014/360438

Balaban, S., Lee, L. S., Schreuder, M., & Hoy, A. J. (2015). Obesity and Cancer

Progression: Is There a Role of Fatty Acid Metabolism? BioMed Research

International, 2015, 274585. doi:10.1155/2015/274585

Balaban, S., Nassar, Z. D., Zhang, A. Y., Hosseini-Beheshti, E., Centenera, M. M.,

Schreuder, M., Lin, H. M., Aishah, A., Varney, B., Liu-Fu, F., Lee, L. S.,

Nagarajan, S. R., Shearer, R. F., Hardie, R. A., Raftopulos, N. L., Kakani, M.

S., Saunders, D. N., Holst, J., Horvath, L. G., Butler, L. M., & Hoy, A. J.

(2019). Extracellular Fatty Acids Are the Major Contributor to Lipid Synthesis

in Prostate Cancer. Mol Cancer Res. doi:10.1158/1541-7786.MCR-18-0347

Balaban, S., Shearer, R. F., Lee, L. S., van Geldermalsen, M., Schreuder, M., Shtein,

H. C., Cairns, R., Thomas, K. C., Fazakerley, D. J., Grewal, T., Holst, J.,

Saunders, D. N., & Hoy, A. J. (2017). Adipocyte lipolysis links obesity to

breast cancer growth: adipocyte-derived fatty acids drive breast cancer cell

proliferation and migration. Cancer Metab, 5, 1. doi:10.1186/s40170-016-

0163-7

Balk, S. P., & Knudsen, K. E. (2008). AR, the cell cycle, and prostate cancer. Nuclear

Receptor Signaling, 6, e001. doi:10.1621/nrs.06001

Bibliography 188

Barfeld, S. J., Itkonen, H. M., Urbanucci, A., & Mills, I. G. (2014). Androgen-

regulated metabolism and biosynthesis in prostate cancer. Endocr Relat

Cancer, 21(4), T57-66. doi:10.1530/ERC-13-0515

Basaria, S., Muller, D. C., Carducci, M. A., Egan, J., & Dobs, A. S. (2006).

Hyperglycemia and insulin resistance in men with prostate carcinoma who

receive androgen-deprivation therapy. Cancer, 106(3), 581-588.

doi:10.1002/cncr.21642

Beloribi-Djefaflia, S., Vasseur, S., & Guillaumond, F. (2016). Lipid metabolic

reprogramming in cancer cells. Oncogenesis, 5, e189.

doi:10.1038/oncsis.2015.49

Benjamini, Y., & Hochberg, Y. (1995). Controlling the False Discovery Rate: A

Practical and Powerful Approach to Multiple Testing. Journal of the Royal

Statistical Society. Series B (Methodological), 57(1), 289-300.

Blebea, J. S., Houseni, M., Torigian, D. A., Fan, C., Mavi, A., Zhuge, Y., Iwanaga, T.,

Mishra, S., Udupa, J., Zhuang, J., Gopal, R., & Alavi, A. (2007). Structural and

Functional Imaging of Normal Bone Marrow and Evaluation of Its Age-

Related Changes. Seminars in Nuclear Medicine, 37(3), 185-194.

doi:https://doi.org/10.1053/j.semnuclmed.2007.01.002

Boukouris, A. E., Zervopoulos, S. D., & Michelakis, E. D. (2016). Metabolic Enzymes

Moonlighting in the Nucleus: Metabolic Regulation of Gene Transcription.

Trends in Biochemical Sciences, 41(8), 712-730.

doi:https://doi.org/10.1016/j.tibs.2016.05.013

Bradbury, M. W., Stump, D., Guarnieri, F., & Berk, P. D. (2011). Molecular modeling

and functional confirmation of a predicted fatty acid binding site of

mitochondrial aspartate aminotransferase. Journal of molecular biology,

412(3), 412-422. doi:10.1016/j.jmb.2011.07.034

Bray, F., Ferlay, J., Soerjomataram, I., Siegel, R. L., Torre, L. A., & Jemal, A. (2018).

Global cancer statistics 2018: GLOBOCAN estimates of incidence and

mortality worldwide for 36 cancers in 185 countries. CA: A Cancer Journal for

Clinicians, 68(6), 394-424. doi:doi:10.3322/caac.21492

Brglez, V., Lambeau, G., & Petan, T. (2014). Secreted phospholipases A2 in cancer:

Diverse mechanisms of action. Biochimie, 107, 114-123.

doi:https://doi.org/10.1016/j.biochi.2014.09.023

Bria, E., Cuppone, F., Giannarelli, D., Milella, M., Ruggeri, E. M., Sperduti, I.,

Pinnarò, P., Terzoli, E., Cognetti, F., & Carlini, P. (2009). Does hormone

treatment added to radiotherapy improve outcome in locally advanced prostate

cancer? Cancer, 115(15), 3446-3456. doi:doi:10.1002/cncr.24392

Brusselmans, K., & Swinnen, J. V. (2009). The Lipogenic Switch in Cancer. In

Mitochondria and Cancer (pp. 39-59). New York, NY: Springer New York.

Burdge, G. C., & Calder, P. C. (2015). Introduction to fatty acids and lipids. World

Rev Nutr Diet, 112, 1-16. doi:10.1159/000365423

Burke, J. E., & Dennis, E. A. (2009). Phospholipase A2 structure/function,

mechanism, and signaling. J Lipid Res, 50 Suppl, S237-242.

doi:10.1194/jlr.R800033-JLR200

Butler, L. M., Centenera, M. M., & Swinnen, J. V. (2016). Androgen control of lipid

metabolism in prostate cancer: novel insights and future applications. Endocr

Relat Cancer, 23(5), R219-227. doi:10.1530/ERC-15-0556

Cabodevilla, A. G., Sánchez-Caballero, L., Nintou, E., Boiadjieva, V. G., Picatoste,

F., Gubern, A., & Claro, E. (2013). Cell survival during complete nutrient

deprivation depends on lipid droplet-fueled β-oxidation of fatty acids. The

https://doi.org/10.1053/j.semnuclmed.2007.01.002

https://doi.org/10.1016/j.tibs.2016.05.013

https://doi.org/10.1016/j.biochi.2014.09.023

Bibliography 189

Journal of Biological Chemistry, 288(39), 27777-27788.

doi:10.1074/jbc.M113.466656

Calder, P. C. (2017). Omega-3 fatty acids and inflammatory processes: from molecules

to man. Biochemical Society Transactions, 45(5), 1105-1115.

doi:10.1042/bst20160474

Cao, Z., West, C., Norton-Wenzel C., Rej, R., Davis, F., Davis, P., & Rej,

R. (2009) Effects of Resin or Charcoal Treatment on Fetal Bovine Serum and

Bovine Calf Serum. Endocrine Research, 34:4, 101-

108, DOI: 10.3109/07435800903204082

Cases, S., Stone, S. J., Zhou, P., Yen, E., Tow, B., Lardizabal, K. D., Voelker, T., &

Farese, R. V. (2001). Cloning of DGAT2, a Second Mammalian

Diacylglycerol Acyltransferase, and Related Family Members. Journal of

Biological Chemistry, 276(42), 38870-38876. doi:10.1074/jbc.M106219200

Chaudry, A. A., Wahle, K. W. J., McClinton, S., & Moffat, L. E. F. (1994).

Arachidonic acid metabolism in benign and malignant prostatic tissue in vitro:

Effects of fatty acids and cyclooxygenase inhibitors. Int J Cancer, 57(2), 176-

180. doi:doi:10.1002/ijc.2910570208

Chilton, F. H., Dutta, R., Reynolds, L. M., Sergeant, S., Mathias, R. A., & Seeds, M.

C. (2017). Precision Nutrition and Omega-3 Polyunsaturated Fatty Acids: A

Case for Personalized Supplementation Approaches for the Prevention and

Management of Human Diseases. Nutrients, 9(11), 1165.

doi:10.3390/nu9111165

Chitraju, C., Mejhert, N., Haas, J. T., Diaz-Ramirez, L. G., Grueter, C. A., Imbriglio,

J. E., Pinto, S., Koliwad, S. K., Walther, T. C., & Farese, R. V., Jr. (2017).

Triglyceride Synthesis by DGAT1 Protects Adipocytes from Lipid-Induced

ER Stress during Lipolysis. Cell metabolism, 26(2), 407-418.e403.

doi:10.1016/j.cmet.2017.07.012

Chong, J., & Xia, J. (2018). MetaboAnalystR: an R package for flexible and

reproducible analysis of metabolomics data. Bioinformatics, 34(24), 4313-

4314. doi:10.1093/bioinformatics/bty528 %J Bioinformatics

Coleman, R. E. (2001). Metastatic bone disease: clinical features, pathophysiology and

treatment strategies. Cancer Treatment Reviews, 27(3), 165-176.

doi:http://dx.doi.org/10.1053/ctrv.2000.0210

Cook, T., & Sheridan, W. P. (2000). Development of GnRH antagonists for prostate

cancer: new approaches to treatment. Oncologist, 5(2), 162-168.

Corsetto, P., Colombo, I., Kopecka, J., Rizzo, A., & Riganti, C. (2017). ω-3 Long

Chain Polyunsaturated Fatty Acids as Sensitizing Agents and Multidrug

Resistance Revertants in Cancer Therapy. International Journal of Molecular

Sciences, 18(12), 2770.

Cotte, A. K., Aires, V., Fredon, M., Limagne, E., Derangère, V., Thibaudin, M.,

Humblin, E., Scagliarini, A., de Barros, J.-P. P., Hillon, P., Ghiringhelli, F., &

Delmas, D. (2018). Lysophosphatidylcholine acyltransferase 2-mediated lipid

droplet production supports colorectal cancer chemoresistance. Nature

Communications, 9(1), 322-322. doi:10.1038/s41467-017-02732-5

Currie, E., Schulze, A., Zechner, R., Walther, T. C., & Farese, R. V. (2013). Cellular

Fatty Acid Metabolism and Cancer. Cell metabolism, 18(2), 153-161.

doi:10.1016/j.cmet.2013.05.017

Czyz, O., Bitew, T., Cuesta-Marbán, A., McMaster, C. R., Mollinedo, F., &

Zaremberg, V. (2013). Alteration of plasma membrane organization by an

anticancer lysophosphatidylcholine analogue induces intracellular

https://doi.org/10.3109/07435800903204082

http://dx.doi.org/10.1053/ctrv.2000.0210

Bibliography 190

acidification and internalization of plasma membrane transporters in yeast. The

Journal of Biological Chemistry, 288(12), 8419-8432.

doi:10.1074/jbc.M112.425744

Daniëls, V. W., Smans, K., Royaux, I., Chypre, M., Swinnen, J. V., & Zaidi, N. (2014).

Cancer Cells Differentially Activate and Thrive on De Novo Lipid Synthesis

Pathways in a Low-Lipid Environment. PLOS ONE, 9(9), e106913.

doi:10.1371/journal.pone.0106913

Deep, G., & Schlaepfer, I. (2016). Aberrant Lipid Metabolism Promotes Prostate

Cancer: Role in Cell Survival under Hypoxia and Extracellular Vesicles

Biogenesis. International Journal of Molecular Sciences, 17(7), 1061.

Desir, S., Dickson, E. L., Vogel, R. I., Thayanithy, V., Wong, P., Teoh, D., Geller, M.

A., Steer, C. J., Subramanian, S., & Lou, E. (2016). Tunneling nanotube

formation is stimulated by hypoxia in ovarian cancer cells. Oncotarget, 7(28),

43150-43161. doi:10.18632/oncotarget.9504

Diaz, G., Melis, M., Batetta, B., Angius, F., & Falchi, A. M. (2008). Hydrophobic

characterization of intracellular lipids in situ by Nile Red red/yellow emission

ratio. Micron, 39(7), 819-824.

doi:http://dx.doi.org/10.1016/j.micron.2008.01.001

Diedrich, J. D., Herroon, M. K., Rajagurubandara, E., & Podgorski, I. (2018). The

Lipid Side of Bone Marrow Adipocytes: How Tumor Cells Adapt and Survive

in Bone. Current Osteoporosis Reports, 16(4), 443-457. doi:10.1007/s11914-

018-0453-9

Dillard, P. R., Lin, & Khan, S. A. (2008). Androgen-independent prostate cancer cells

acquire the complete steroidogenic potential of synthesizing testosterone from

cholesterol. Molecular and Cellular Endocrinology, 295(1–2), 115-120.

doi:https://doi.org/10.1016/j.mce.2008.08.013

Dobin, A., Davis, C. A., Schlesinger, F., Drenkow, J., Zaleski, C., Jha, S., Batut, P.,

Chaisson, M., & Gingeras, T. R. (2013). STAR: ultrafast universal RNA-seq

aligner. Bioinformatics, 29(1), 15-21. doi:10.1093/bioinformatics/bts635

Doege, H., & Stahl, A. (2006). Protein-Mediated Fatty Acid Uptake: Novel Insights

from In Vivo Models. Physiology, 21(4), 259-268.

doi:10.1152/physiol.00014.2006

Dong, Z., Liu, Y., Scott, K. F., Levin, L., Gaitonde, K., Bracken, R. B., Burke, B.,

Zhai, Q. J., Wang, J., Oleksowicz, L., & Lu, S. (2010). Secretory phospholipase

A2-IIa is involved in prostate cancer progression and may potentially serve as

a biomarker for prostate cancer. Carcinogenesis, 31(11), 1948-1955.

doi:10.1093/carcin/bgq188

Drummen, G. P. C., van Liebergen, L. C. M., Op den Kamp, J. A. F., & Post, J. A.

(2002). C11-BODIPY581/591, an oxidation-sensitive fluorescent lipid

peroxidation probe: (micro)spectroscopic characterization and validation of

methodology. Free Radical Biology and Medicine, 33(4), 473-490.

doi:https://doi.org/10.1016/S0891-5849(02)00848-1

Dutt, S. S., & Gao, A. C. (2009). Molecular mechanisms of castration-resistant prostate

cancer progression. Future oncology (London, England), 5(9), 1403-1413.

doi:10.2217/fon.09.117

Eden, E., Navon, R., Steinfeld, I., Lipson, D., & Yakhini, Z. (2009). GOrilla: a tool for

discovery and visualization of enriched GO terms in ranked gene lists. BMC

Bioinformatics, 10, 48. doi:10.1186/1471-2105-10-48

Effert, P. J., Bares, R., Handt, S., Wolff, J. M., Bull, U., & Jakse, G. (1996). Metabolic

Imaging of Untreated Prostate Cancer by Positron Emission Tomography with

http://dx.doi.org/10.1016/j.micron.2008.01.001

https://doi.org/10.1016/j.mce.2008.08.013

https://doi.org/10.1016/S0891-5849(02)00848-1

Bibliography 191

sup 18 Fluorine-Labeled Deoxyglucose. The Journal of Urology, 155(3), 994-

998. doi:http://dx.doi.org/10.1016/S0022-5347(01)66366-3

Egbewande, F. A., Sadowski, M. C., Levrier, C., Tousignant, K. D., White, J. M.,

Coster, M. J., Nelson, C. C., & Davis, R. A. (2018). Identification of

Gibberellic Acid Derivatives That Deregulate Cholesterol Metabolism in

Prostate Cancer Cells. Journal of Natural Products, 81(4), 838-845.

doi:10.1021/acs.jnatprod.7b00929

Elzanaty, S., Rezanezhad, B., & Dohle, G. (2017). Association between Serum

Testosterone and PSA Levels in Middle-Aged Healthy Men from the General

Population. Current Urology, 10(1), 40-44. doi:10.1159/000447149

Engholm-Keller, K., Birck, P., Størling, J., Pociot, F., Mandrup-Poulsen, T., & Larsen,

M. R. (2012). TiSH — a robust and sensitive global phosphoproteomics

strategy employing a combination of TiO2, SIMAC, and HILIC. Journal of

Proteomics, 75(18), 5749-5761.

doi:http://dx.doi.org/10.1016/j.jprot.2012.08.007

Epand, R. M. (2015). Introduction to Membrane Lipids. In D. M. Owen (Ed.), Methods

in Membrane Lipids (pp. 1-6). New York, NY: Springer New York.

Ettinger, S. L., Sobel, R., Whitmore, T. G., Akbari, M., Bradley, D. R., Gleave, M. E.,

& Nelson, C. C. (2004). Dysregulation of Sterol Response Element-Binding

Proteins and Downstream Effectors in Prostate Cancer during Progression to

Androgen Independence. Cancer Research, 64(6), 2212-2221.

doi:10.1158/0008-5472.can-2148-2

Fahy, E., Cotter, D., Sud, M., & Subramaniam, S. (2011). Lipid classification,

structures and tools. Biochim Biophys Acta, 1811(11), 637-647.

doi:10.1016/j.bbalip.2011.06.009

Fahy, E., Subramaniam, S., Brown, H. A., Glass, C. K., Merrill, A. H., Jr., Murphy, R.

C., Raetz, C. R., Russell, D. W., Seyama, Y., Shaw, W., Shimizu, T., Spener,

F., van Meer, G., VanNieuwenhze, M. S., White, S. H., Witztum, J. L., &

Dennis, E. A. (2005). A comprehensive classification system for lipids. J Lipid

Res, 46(5), 839-861. doi:10.1194/jlr.E400004-JLR200

Ferri, N., Ricci, C., & Corsini, A. 2015. Pharmacological Inhibition of Phospholipase

A2: Results from Phase 3 Clinical Trials with Darapladib and Varespladib in

patients with cardiovascular disease. Cardiovascular Pharmacology 4(2).

doi:10.4172/2329-6607.10000137

Finicle, B. T., Jayashankar, V., & Edinger, A. L. (2018). Nutrient scavenging in

cancer. Nature Reviews Cancer, 18(10), 619-633. doi:10.1038/s41568-018-

0048-x

Flavin, R., Zadra, G., & Loda, M. (2011). Metabolic alterations and targeted therapies

in prostate cancer. The Journal of pathology, 223(2), 283-294.

doi:10.1002/path.2809

Foletta, V. C., Palmieri, M., Kloehn, J., Mason, S., Previs, S. F., McConville, M. J.,

Sieber, O. M., Bruce, C. R., & Kowalski, G. M. (2016). Analysis of

Mammalian Cell Proliferation and Macromolecule Synthesis Using Deuterated

Water and Gas Chromatography-Mass Spectrometry. Metabolites, 6(4), 34.

doi:10.3390/metabo6040034

Fritz, V., Benfodda, Z., Rodier, G., Henriquet, C., Iborra, F., Avancès, C., Allory, Y.,

de la Taille, A., Culine, S., Blancou, H., Cristol, J. P., Michel, F., Sardet, C., &

Fajas, L. (2010). Abrogation of De novo Lipogenesis by Stearoyl-CoA

Desaturase 1 Inhibition Interferes with Oncogenic Signaling and Blocks

http://dx.doi.org/10.1016/S0022-5347(01)66366-3

http://dx.doi.org/10.1016/j.jprot.2012.08.007

Bibliography 192

Prostate Cancer Progression in Mice. Molecular Cancer Therapeutics, 9(6),

1740-1754. doi:10.1158/1535-7163.mct-09-1064

Fruh, S. M. (2017). Obesity: Risk factors, complications, and strategies for sustainable

long-term weight management. J Am Assoc Nurse Pract, 29(S1), S3-S14.

doi:10.1002/2327-6924.12510

Furuya, Y., Sekine, Y., Kato, H., Miyazawa, Y., Koike, H., & Suzuki, K. (2016). Low-

density lipoprotein receptors play an important role in the inhibition of prostate

cancer cell proliferation by statins. Prostate International, 4(2), 56-60.

doi:10.1016/j.prnil.2016.02.003

Galbraith L., Leung, HY., Ahmad, I. (2018). Lipid pathway deregulation in advanced

prostate cancer. Pharmacol Res (131), 177-184.

doi:10.1016/j.phrs.2018.02.022

Gartrell, B. A., Coleman, R., Efstathiou, E., Fizazi, K., Logothetis, C. J., Smith, M. R.,

Sonpavde, G., Sartor, O., & Saad, F. (2015). Metastatic Prostate Cancer and

the Bone: Significance and Therapeutic Options. European Urology, 68(5),

850-858. doi:https://doi.org/10.1016/j.eururo.2015.06.039

Gaston, S. M., Kearns, J., Adams, G. W., Rais-Bahrami, S., Nix, J., Kolettis, P. N.,

Bryant, J. E., Otali, D., & Grizzle, W. E. (2017). Satisfying the fatty acid

demand of prostate cancer: De novo synthesis versus uptake as alternative and

potentially cooperative prostate cancer phenotypes. Journal of Clinical

Oncology 35, 107-107. doi:10.1200/JCO.2017.35.6_suppl.107

Gerdes, H.-H., Bukoreshtliev, N. V., & Barroso, J. F. V. (2007). Tunneling nanotubes:

A new route for the exchange of components between animal cells. FEBS

Letters, 581(11), 2194-2201. doi:10.1016/j.febslet.2007.03.071

Ghayee, H. K., & Auchus, R. J. (2007). Basic concepts and recent developments in

human steroid hormone biosynthesis. Reviews in Endocrine and Metabolic

Disorders, 8(4), 289-300. doi:10.1007/s11154-007-9052-2

Gleason, D. (1966). Classification of Prostatic Carcinomas. Cancer Chemotherapy

Reports, 50, 125-128.

Go, G.-w., & Mani, A. (2012). Low-Density Lipoprotein Receptor (LDLR) Family

Orchestrates Cholesterol Homeostasis. The Yale Journal of Biology and

Medicine, 85(1), 19-28.

Goldstein, J. L., Anderson, R. G., & Brown, M. S. (1982). Receptor-mediated

endocytosis and the cellular uptake of low density lipoprotein. Ciba Found

Symp(92), 77-95.

Gordon, J. A., Midha, A., Szeitz, A., Ghaffari, M., Adomat, H. H., Guo, Y., Klassen,

T. L., Guns, E. S., Wasan, K. M., & Cox, M. E. (2016). Oral simvastatin

administration delays castration-resistant progression and reduces intratumoral

steroidogenesis of LNCaP prostate cancer xenografts. Prostate Cancer

Prostatic Dis, 19(1), 21-27. doi:10.1038/pcan.2015.37

Gossett, R. E., Frolov, A. A., Roths, J. B., Behnke, W. D., Kier, A. B., & Schroeder,

F. (1996). Acyl-CoA binding proteins: Multiplicity and function. Lipids, 31(9),

895-918. doi:10.1007/bf02522684

Grasso, C. S., Wu, Y. M., Robinson, D. R., Cao, X., Dhanasekaran, S. M., & Khan, A.

P. (2012). The mutational landscape of lethal castration-resistant prostate

cancer. Nature, 487. doi:10.1038/nature11125

Greenspan P, M. E., Fowler S. (1985). Nile red: a selective fluorescent stain for

intracellular lipid droplets. The Journal of Cell Biology, 100(3), 965-973.

Gregory, L. S., Kelly, W. L., Reid, R. C., Fairlie, D. P., & Forwood, M. R. (2006).

Inhibitors of cyclo-oxygenase-2 and secretory phospholipase A2 preserve bone

https://doi.org/10.1016/j.eururo.2015.06.039

Bibliography 193

architecture following ovariectomy in adult rats. Bone, 39(1), 134-142.

doi:10.1016/j.bone.2005.12.017

Griffiths, B., Lewis, C. A., Bensaad, K., Ros, S., Zhang, Q., Ferber, E. C., Konisti, S.,

Peck, B., Miess, H., East, P., Wakelam, M., Harris, A. L., & Schulze, A.

(2013). Sterol regulatory element binding protein-dependent regulation of lipid

synthesis supports cell survival and tumor growth. Cancer & Metabolism, 1,

3-3. doi:10.1186/2049-3002-1-3

Grouleff, J., Irudayam, S. J., Skeby, K. K., & Schiøtt, B. (2015). The influence of

cholesterol on membrane protein structure, function, and dynamics studied by

molecular dynamics simulations. Biochimica et Biophysica Acta (BBA) -

Biomembranes, 1848(9), 1783-1795.

doi:https://doi.org/10.1016/j.bbamem.2015.03.029

Guillou, H., Zadravec, D., Martin, P. G. P., & Jacobsson, A. (2010). The key roles of

elongases and desaturases in mammalian fatty acid metabolism: Insights from

transgenic mice. Progress in Lipid Research, 49(2), 186-199.

doi:https://doi.org/10.1016/j.plipres.2009.12.002

Gunter, J. H., Lubik, A. A., McKenzie, I., Pollak, M., & Nelson, C. C. (2012). The

Interactions between Insulin and Androgens in Progression to Castrate-

Resistant Prostate Cancer. Advances in Urology, 2012, 248607.

doi:10.1155/2012/248607

Gunter, J. H., Sarkar, P. L., Lubik, A. A., & Nelson, C. C. (2013). New Players for

Advanced Prostate Cancer and the Rationalisation of Insulin-Sensitising

Medication. International Journal of Cell Biology, 2013, 834684.

doi:10.1155/2013/834684

Gurr, M. I., Frayn, K. N., & Harwood, J. L. (2002). Lipid biochemistry (5th ed.).

Oxford Malden, MA: Blackwell Science.

Hanahan, D., & Weinberg, Robert A. (2011). Hallmarks of Cancer: The Next

Generation. Cell, 144(5), 646-674.

doi:https://doi.org/10.1016/j.cell.2011.02.013

Hangauer, M. J., Viswanathan, V. S., Ryan, M. J., Bole, D., Eaton, J. K., Matov, A.,

Galeas, J., Dhruv, H. D., Berens, M. E., Schreiber, S. L., McCormick, F., &

McManus, M. T. (2017). Drug-tolerant persister cancer cells are vulnerable to

GPX4 inhibition. Nature, 551, 247. doi:10.1038/nature24297

https://www.nature.com/articles/nature24297#supplementary-information

Hansford, K. A., Reid, R. C., Clark, C. I., Tyndall, J. D., Whitehouse, M. W., Guthrie,

T., McGeary, R. P., Schafer, K., Martin, J. L., & Fairlie, D. P. (2003). D-

Tyrosine as a chiral precusor to potent inhibitors of human nonpancreatic

secretory phospholipase A2 (IIa) with antiinflammatory activity.

Chembiochem, 4(2-3), 181-185. doi:10.1002/cbic.200390029

Harnden, P., Shelley, M. D., Coles, B., Staffurth, J., & Mason, M. D. (2007). Should

the Gleason grading system for prostate cancer be modified to account for high-

grade tertiary components? A systematic review and meta-analysis. The Lancet

Oncology, 8(5), 411-419. doi:https://doi.org/10.1016/S1470-2045(07)70136-5

Heemers, H., Maes, B., Foufelle, F., Heyns, W., Verhoeven, G., & Swinnen, J. V.

(2001). Androgens stimulate lipogenic gene expression in prostate cancer cells

by activation of the sterol regulatory element-binding protein cleavage

activating protein/sterol regulatory element-binding protein pathway. Mol

Endocrinol, 15(10), 1817-1828. doi:10.1210/mend.15.10.0703

Heemers, H., Verrijdt, G., Organe, S., Claessens, F., Heyns, W., Verhoeven, G., &

Swinnen, J. V. (2004). Identification of an androgen response element in intron

https://doi.org/10.1016/j.bbamem.2015.03.029

https://doi.org/10.1016/j.plipres.2009.12.002

https://doi.org/10.1016/j.cell.2011.02.013

https://www.nature.com/articles/nature24297#supplementary-information

https://doi.org/10.1016/S1470-2045(07)70136-5

Bibliography 194

8 of the sterol regulatory element-binding protein cleavage-activating protein

gene allowing direct regulation by the androgen receptor. J Biol Chem,

279(29), 30880-30887. doi:10.1074/jbc.M401615200

Heidenreich, A., Aus, G., Bolla, M., Joniau, S., Matveev, V. B., Schmid, H. P., &

Zattoni, F. (2008). EAU Guidelines on Prostate Cancer. European Urology,

53(1), 68-80. doi:http://dx.doi.org/10.1016/j.eururo.2007.09.002

Heidenreich, A., Bellmunt, J., Bolla, M., Joniau, S., Mason, M., Matveev, V., Mottet,

N., Schmid, H.-P., van der Kwast, T., Wiegel, T., & Zattoni, F. (2011). EAU

Guidelines on Prostate Cancer. Part 1: Screening, Diagnosis, and Treatment of

Clinically Localised Disease. European Urology, 59(1), 61-71.

doi:https://doi.org/10.1016/j.eururo.2010.10.039

Heinlein, C. A., & Chang, C. (2004). Androgen Receptor in Prostate Cancer.

Endocrine Reviews, 25(2), 276-308. doi:10.1210/er.2002-0032

Hendrich, A. B., & Michalak, K. (2003). Lipids as a Target for Drugs Modulating

Multidrug Resistance of Cancer Cells. Current Drug Targets, 4(1), 23-30.

doi:http://dx.doi.org/10.2174/1389450033347172

Herms, A., Bosch, M., Ariotti, N., Reddy, Babu J. N., Fajardo, A., Fernández-Vidal,

A., Alvarez-Guaita, A., Fernández-Rojo, Manuel A., Rentero, C., Tebar, F.,

Enrich, C., Geli, M.-I., Parton, Robert G., Gross, Steven P., & Pol, A. (2013).

Cell-to-Cell Heterogeneity in Lipid Droplets Suggests a Mechanism to Reduce

Lipotoxicity. Current Biology, 23(15), 1489-1496.

doi:http://dx.doi.org/10.1016/j.cub.2013.06.032

Hishikawa, D., Eto, M., Shimizu, T., Harayama, T., & Shindou, H. (2013). Generation

of membrane diversity by lysophospholipid acyltransferases. The Journal of

Biochemistry, 154(1), 21-28.

Hollie, N. I., Cash, J. G., Matlib, M. A., Wortman, M., Basford, J. E., Abplanalp, W.,

& Hui, D. Y. (2014). Micromolar changes in lysophosphatidylcholine

concentration cause minor effects on mitochondrial permeability but major

alterations in function. Biochimica et biophysica acta, 1841(6), 888-895.

doi:10.1016/j.bbalip.2013.11.013

Holzbeierlein, J., Lal, P., LaTulippe, E., Smith, A., Satagopan, J., Zhang, L., Ryan, C.,

Smith, S., Scher, H., Scardino, P., Reuter, V., & Gerald, W. L. (2004). Gene

expression analysis of human prostate carcinoma during hormonal therapy

identifies androgen-responsive genes and mechanisms of therapy resistance.

The American Journal of Pathology, 164(1), 217-227. doi:10.1016/S0002-

9440(10)63112-4

Hong‐Lin, H., Ying‐En, L., Yow‐Ling, S., Sung‐Wei, L., Li‐Ching, L., Tzu‐Ju, C.,

Ting‐Feng, W., & Chien‐Feng, L. (2015). PLA2G2A overexpression is

associated with poor therapeutic response and inferior outcome in rectal cancer

patients receiving neoadjuvant concurrent chemoradiotherapy.

Histopathology, 66(7), 991-1002. doi:doi:10.1111/his.12613

Hosios, Aaron M., Hecht, Vivian C., Danai, Laura V., Johnson, Marc O., Rathmell,

Jeffrey C., Steinhauser, Matthew L., Manalis, Scott R., & Vander Heiden,

Matthew G. (2016). Amino Acids Rather than Glucose Account for the

Majority of Cell Mass in Proliferating Mammalian Cells. Developmental Cell,

36(5), 540-549. doi:http://dx.doi.org/10.1016/j.devcel.2016.02.012

Hultsch, S., Kankainen, M., Paavolainen, L., Kovanen, R.-M., Ikonen, E.,

Kangaspeska, S., Pietiäinen, V., & Kallioniemi, O. (2018). Association of

tamoxifen resistance and lipid reprogramming in breast cancer. BMC Cancer,

18(1), 850-850. doi:10.1186/s12885-018-4757-z

http://dx.doi.org/10.1016/j.eururo.2007.09.002


http://dx.doi.org/10.2174/1389450033347172

http://dx.doi.org/10.1016/j.cub.2013.06.032

http://dx.doi.org/10.1016/j.devcel.2016.02.012

Bibliography 195

Iglesias-Gato, D., Thysell, E., Tyanova, S., Crnalic, S., Santos, A., Lima, T. S., Geiger,

T., Cox, J., Widmark, A., Bergh, A., Mann, M., Flores-Morales, A., &

Wikström, P. (2018). The proteome of prostate cancer bone metastasis reveals

heterogeneity with prognostic implications. Clinical Cancer Research.

doi:10.1158/1078-0432.ccr-18-1229

Iglesias-Gato, D., Wikström, P., Tyanova, S., Lavallee, C., Thysell, E., Carlsson, J.,

Hägglöf, C., Cox, J., Andrén, O., Stattin, P., Egevad, L., Widmark, A., Bjartell,

A., Collins, C. C., Bergh, A., Geiger, T., Mann, M., & Flores-Morales, A.

(2016). The Proteome of Primary Prostate Cancer. European Urology, 69(5),

942-952. doi:https://doi.org/10.1016/j.eururo.2015.10.053

Ikonen, E. (2008). Cellular cholesterol trafficking and compartmentalization. Nat Rev

Mol Cell Biol, 9(2), 125-138.

Ingrosso, G., Detti, B., Scartoni, D., Lancia, A., Giacomelli, I., Baki, M., Carta, G.,

Livi, L., & Santoni, R. (2018). Current therapeutic options in metastatic

castration-resistant prostate cancer. Semin Oncol, 45(5-6), 303-315.

doi:10.1053/j.seminoncol.2018.10.001

James, M. J., Gibson, R. A., & Cleland, L. G. (2000). Dietary polyunsaturated fatty

acids and inflammatory mediator production. The American Journal of

Clinical Nutrition, 71(1), 343s-348s. doi:10.1093/ajcn/71.1.343s

Jarc, E., Kump, A., Malavašič, P., Eichmann, T. O., Zimmermann, R., & Petan, T.

(2018). Lipid droplets induced by secreted phospholipase A2 and unsaturated

fatty acids protect breast cancer cells from nutrient and lipotoxic stress.

Biochimica et Biophysica Acta (BBA) - Molecular and Cell Biology of Lipids,

1863(3), 247-265. doi:https://doi.org/10.1016/j.bbalip.2017.12.006

Jin, H.-J., Kim, J., & Yu, J. (2013). Androgen receptor genomic regulation.

Translational andrology and urology, 2(3), 157-177. doi:10.3978/j.issn.2223-

4683.2013.09.01

Juan, C., Long, Y., Yu, S., & Yoshikazhu, T. (2017). A Concise Update on the

Relevance of Secretory Phospholipase A2 Group IIA and its Inhibitors with

Cancer. Medicinal Chemistry, 13(7), 606-615.

doi:http://dx.doi.org/10.2174/1573406413666170209121317

Kalinski, P. (2012). Regulation of immune responses by prostaglandin E2. Journal of

immunology (Baltimore, Md. : 1950), 188(1), 21-28.

doi:10.4049/jimmunol.1101029

Kamentsky, L., Jones, T. R., Fraser, A., Bray, M.-A., Logan, D. J., Madden, K. L.,

Ljosa, V., Rueden, C., Eliceiri, K. W., & Carpenter, A. E. (2011). Improved

structure, function and compatibility for CellProfiler: modular high-throughput

image analysis software. Bioinformatics, 27(8), 1179-1180.

doi:10.1093/bioinformatics/btr095

Kamphorst, J. J., Cross, J. R., Fan, J., de Stanchina, E., Mathew, R., White, E. P.,

Thompson, C. B., & Rabinowitz, J. D. (2013). Hypoxic and Ras-transformed

cells support growth by scavenging unsaturated fatty acids from

lysophospholipids. Proceedings of the National Academy of Sciences of the

United States of America, 110(22), 8882-8887. doi:10.1073/pnas.1307237110

Kennedy, B. E., Charman, M., & Karten, B. (2012). Niemann-Pick Type C2 protein

contributes to the transport of endosomal cholesterol to mitochondria without

interacting with NPC1. Journal of Lipid Research, 53(12), 2632-2642.

doi:10.1194/jlr.M029942

Kerr, M. C., & Teasdale, R. D. (2009). Defining Macropinocytosis. 10(4), 364-371.

doi:10.1111/j.1600-0854.2009.00878.x


https://doi.org/10.1016/j.bbalip.2017.12.006

http://dx.doi.org/10.2174/1573406413666170209121317

Bibliography 196

Kim, S. M., Nguyen, T. T., Ravi, A., Kubiniok, P., Finicle, B. T., Jayashankar, V.,

Malacrida, L., Hou, J., Robertson, J., Gao, D., Chernoff, J., Digman, M. A.,

Potma, E. O., Tromberg, B. J., Thibault, P., & Edinger, A. L. (2018). PTEN

Deficiency and AMPK Activation Promote Nutrient Scavenging and

Anabolism in Prostate Cancer Cells. Cancer Discovery, 8(7), 866-883.

doi:10.1158/2159-8290.CD-17-1215

Kinoshita, M., Suzuki, K. G. N., Murata, M., & Matsumori, N. (2018). Evidence of

lipid rafts based on the partition and dynamic behavior of sphingomyelins.

Chemistry and Physics of Lipids, 215, 84-95.

doi:https://doi.org/10.1016/j.chemphyslip.2018.07.002

Kirby, M., Hirst, C., & Crawford, E. D. (2011). Characterising the castration-resistant

prostate cancer population: a systematic review. International Journal of

Clinical Practice, 65(11), 1180-1192. doi:10.1111/j.1742-1241.2011.02799.x

Koizume, S., & Miyagi, Y. (2016). Lipid Droplets: A Key Cellular Organelle

Associated with Cancer Cell Survival under Normoxia and Hypoxia.

International Journal of Molecular Sciences, 17(9), 1430.

doi:10.3390/ijms17091430

Kolde, R., & Vilo, J. (2015). GOsummaries: an R Package for Visual Functional

Annotation of Experimental Data. F1000Res, 4, 574.

doi:10.12688/f1000research.6925.1

Krueger, F. (2012). Trim Galore. Retrieved from

http://www.bioinformatics.babraham.ac.uk/projects/trim_galore/

Kuemmerle, N. B., Rysman, E., Lombardo, P. S., Flanagan, A. J., Lipe, B. C., Wells,

W. A., Pettus, J. R., Froehlich, H. M., Memoli, V. A., Morganelli, P. M.,

Swinnen, J. V., Timmerman, L. A., Chaychi, L., Fricano, C. J., Eisenberg, B.

L., Coleman, W. B., & Kinlaw, W. B. (2011). Lipoprotein lipase links dietary

fat to solid tumor cell proliferation. Mol Cancer Ther, 10(3), 427-436.

doi:10.1158/1535-7163.MCT-10-0802

Kuhajda, F. P. (2000). Fatty-acid synthase and human cancer: new perspectives on its

role in tumor biology. Nutrition, 16(3), 202-208.

doi:https://doi.org/10.1016/S0899-9007(99)00266-X

Labonté, E. D., Pfluger, P. T., Cash, J. G., Kuhel, D. G., Roja, J. C., Magness, D. P.,

Jandacek, R. J., Tschöp, M. H., & Hui, D. Y. (2010). Postprandial

lysophospholipid suppresses hepatic fatty acid oxidation: the molecular link

between group 1B phospholipase A2 and diet-induced obesity. FASEB journal

: official publication of the Federation of American Societies for Experimental

Biology, 24(7), 2516-2524. doi:10.1096/fj.09-144436

Lagor, W, Millar J. (2009). Overview of the LDL receptor: relevance to cholesterol

metabolism and future approaches for the treatment of coronary heart disease.

Journal of Receptor, Ligand and Channel Research, 3, 1-14.

Lapointe, J., Li, C., Higgins, J. P., Rijn, M., Bair, E., & Montgomery, K. (2004). Gene

expression profiling identifies clinically relevant subtypes of prostate cancer.

Proc Natl Acad Sci U S A, 101. doi:10.1073/pnas.0304146101

Latonen, L., Afyounian, E., Jylha, A., Nattinen, J., Aapola, U., Annala, M.,

Kivinummi, K. K., Tammela, T. T. L., Beuerman, R. W., Uusitalo, H., Nykter,

M., & Visakorpi, T. (2018). Integrative proteomics in prostate cancer uncovers

robustness against genomic and transcriptomic aberrations during disease

progression. Nat Commun, 9(1), 1176. doi:10.1038/s41467-018-03573-6

LaTulippe, E., Satagopan, J., Smith, A., Scher, H., Scardino, P., Reuter, V., & Gerald,

W. L. (2002). Comprehensive gene expression analysis of prostate cancer

https://doi.org/10.1016/j.chemphyslip.2018.07.002

http://www.bioinformatics.babraham.ac.uk/projects/trim_galore/

https://doi.org/10.1016/S0899-9007(99)00266-X

Bibliography 197

reveals distinct transcriptional programs associated with metastatic disease.

Cancer Res, 62(15), 4499-4506.

Lê Cao, K.-A., Boitard, S., & Besse, P. (2011). Sparse PLS discriminant analysis:

biologically relevant feature selection and graphical displays for multiclass

problems. BMC Bioinformatics, 12(1), 253. doi:10.1186/1471-2105-12-253

Lee, J. K., Bangayan, N. J., Chai, T., Smith, B. A., Pariva, T. E., Yun, S., Vashisht,

A., Zhang, Q., Park, J. W., Corey, E., Huang, J., Graeber, T. G., Wohlschlegel,

J., & Witte, O. N. (2018). Systemic surfaceome profiling identifies target

antigens for immune-based therapy in subtypes of advanced prostate cancer.

Proceedings of the National Academy of Sciences.

doi:10.1073/pnas.1802354115

Leon, C. G., Locke, J. A., Adomat, H. H., Etinger, S. L., Twiddy, A. L., Neumann, R.

D., Nelson, C. C., Guns, E. S., & Wasan, K. M. (2010). Alterations in

cholesterol regulation contribute to the production of intratumoral androgens

during progression to castration-resistant prostate cancer in a mouse xenograft

model. Prostate, 70(4), 390-400. doi:10.1002/pros.21072

Leslie, O., Yin, L., Bruce, B. R., Krishnanath, G., Barbara, B., Paul, S., Linda, L.,

Zhongyun, D., & Shan, L. (2012). Secretory phospholipase A2‐IIa is a target

gene of the HER/HER2‐elicited pathway and a potential plasma biomarker for

poor prognosis of prostate cancer. The Prostate, 72(10), 1140-1149.

doi:doi:10.1002/pros.22463

Leung, S. Y., Chen, X., Chu, K. M., Yuen, S. T., Mathy, J., Ji, J., Chan, A. S. Y., Li,

R., Law, S., Troyanskaya, O. G., Tu, I. P., Wong, J., So, S., Botstein, D., &

Brown, P. O. (2002). Phospholipase A2 group IIA expression in gastric

adenocarcinoma is associated with prolonged survival and less frequent

metastasis. Proceedings of the National Academy of Sciences of the United

States of America, 99(25), 16203-16208. doi:10.1073/pnas.212646299

Levina, E., Ji, H., Chen, M., Baig, M., Oliver, D., Ohouo, P., Lim, C.-u., Schools, G.,

Carmack, S., Ding, Y., Broude, E. V., Roninson, I. B., Buttyan, R., &

Shtutman, M. (2015). Identification of novel genes that regulate androgen

receptor signaling and growth of androgen-deprived prostate cancer cells.

Oncotarget, 6(15), 13088-13104.

Levrier, C., Sadowski, M. C., Nelson, C. C., Healy, P. C., & Davis, R. A. (2015).

Denhaminols A-H, dihydro-beta-agarofurans from the endemic Australian

rainforest plant Denhamia celastroides. J Nat Prod, 78(1), 111-119.

doi:10.1021/np500740f

Li, B., & Dewey, C. N. (2011). RSEM: accurate transcript quantification from RNA-

Seq data with or without a reference genome. BMC Bioinformatics, 12, 323.

doi:10.1186/1471-2105-12-323

Li, J., Ren, S., Piao, H.-l., Wang, F., Yin, P., Xu, C., Lu, X., Ye, G., Shao, Y., Yan,

M., Zhao, X., Sun, Y., & Xu, G. (2016). Integration of lipidomics and

transcriptomics unravels aberrant lipid metabolism and defines cholesteryl

oleate as potential biomarker of prostate cancer. Scientific Reports, 6, 20984.

doi:10.1038/srep20984

Liberio, M. S., Sadowski, M. C., Soekmadji, C., Davis, R. A., & Nelson, C. C. (2014).

Differential effects of tissue culture coating substrates on prostate cancer cell

adherence, morphology and behavior. PLOS ONE, 9(11), e112122.


Listenberger, L. L., Han, X., Lewis, S. E., Cases, S., Farese, R. V., Jr., Ory, D. S., &

Schaffer, J. E. (2003). Triglyceride accumulation protects against fatty acid-

Bibliography 198

induced lipotoxicity. Proc Natl Acad Sci U S A, 100(6), 3077-3082.

doi:10.1073/pnas.0630588100

Litwin, M. S., & Tan, H. (2017). The diagnosis and treatment of prostate cancer: A

review. JAMA, 317(24), 2532-2542. doi:10.1001/jama.2017.7248

Liu, Zuckier, & Ghesani. (2010a). Dominant uptake of fatty acid over glucose by

prostate cells: a potential new diagnostic and therapeutic approach. Anticancer

Res, 30(2), 369-374.

Liu, Zuckier, & Ghesani. (2010b). Dominant Uptake of Fatty Acid over Glucose by

Prostate Cells: A Potential New Diagnostic and Therapeutic Approach.

Anticancer Research, 30(2), 369-374.

Locke, J. A., Guns, E. S., Lehman, M. L., Ettinger, S., Zoubeidi, A., Lubik, A.,

Margiotti, K., Fazli, L., Adomat, H., Wasan, K. M., Gleave, M. E., & Nelson,

C. C. (2010). Arachidonic acid activation of intratumoral steroid synthesis

during prostate cancer progression to castration resistance. Prostate, 70(3),

239-251. doi:10.1002/pros.21057

Locke, J. A., Guns, E. S., Lubik, A. A., Adomat, H. H., Hendy, S. C., Wood, C. A.,

Ettinger, S. L., Gleave, M. E., & Nelson, C. C. (2008). Androgen levels

increase by intratumoral de novo steroidogenesis during progression of

castration-resistant prostate cancer. Cancer Res, 68(15), 6407-6415.

doi:10.1158/0008-5472.CAN-07-5997

Locke, J. A., Nelson, C. C., Adomat, H. H., Hendy, S. C., Gleave, M. E., & Guns, E.

S. (2009). Steroidogenesis inhibitors alter but do not eliminate androgen

synthesis mechanisms during progression to castration-resistance in LNCaP

prostate xenografts. J Steroid Biochem Mol Biol, 115(3-5), 126-136.

doi:10.1016/j.jsbmb.2009.03.011

Lonergan, P. E., & Tindall, D. J. (2011). Androgen receptor signaling in prostate

cancer development and progression. Journal of Carcinogenesis, 10, 20.

doi:10.4103/1477-3163.83937

Lopez-Marques, R. L., Theorin, L., Palmgren, M. G., & Pomorski, T. G. (2014). P4-

ATPases: lipid flippases in cell membranes. Pflugers Archiv : European

journal of physiology, 466(7), 1227-1240. doi:10.1007/s00424-013-1363-4

Lu, S., Tsai, S. Y., & Tsai, M.-J. (1999). Molecular Mechanisms of Androgen-

Independent Growth of Human Prostate Cancer LNCaP-AI Cells1.

Endocrinology, 140(11), 5054-5059. doi:10.1210/endo.140.11.7086

Lue, H. W., Podolak, J., Kolahi, K., Cheng, L., Rao, S., Garg, D., Xue, C. H., Rantala,

J. K., Tyner, J. W., Thornburg, K. L., Martinez-Acevedo, A., Liu, J. J., Amling,

C. L., Truillet, C., Louie, S. M., Anderson, K. E., Evans, M. J., O'Donnell, V.

B., Nomura, D. K., Drake, J. M., Ritz, A., & Thomas, G. V. (2017). Metabolic

reprogramming ensures cancer cell survival despite oncogenic signaling

blockade. Genes Dev, 31(20), 2067-2084. doi:10.1101/gad.305292.117

Martín, R., Cordova, C., Gutiérrez, B., Hernández, M., & Nieto, M. L. (2017). A

dangerous liaison: Leptin and sPLA2-IIA join forces to induce proliferation

and migration of astrocytoma cells. PLOS ONE, 12(3), e0170675-e0170675.


Maxfield, F. R., & Wüstner, D. (2012). Analysis of cholesterol trafficking with

fluorescent probes. Methods in cell biology, 108, 367-393. doi:10.1016/B978-

0-12-386487-1.00017-1

Mayer, R. J., & Marshall, L. A. (1993). New insights on mammalian phospholipase

A2(s); comparison of arachidonoyl-selective and -nonselective enzymes. 7(2),

339-348. doi:10.1096/fasebj.7.2.8440410

Bibliography 199

McNair, C., Urbanucci, A., Comstock, C. E. S., Augello, M. A., Goodwin, J. F.,

Launchbury, R., Zhao, S., Schiewer, M. J., Ertel, A., Karnes, J., Davicioni, E.,

Wang, L., Wang, Q., Mills, I. G., Feng, F. Y., Li, W., Carroll, J. S., & Knudsen,

K. E. (2017). Cell-cycle coupled expansion of AR activity promotes cancer

progression. Oncogene, 36(12), 1655-1668. doi:10.1038/onc.2016.334

Mei, S., Qin, Q., Wu, Q., Sun, H., Zheng, R., Zang, C., Zhu, M., Wu, J., Shi, X., Taing,

L., Liu, T., Brown, M., Meyer, C. A., & Liu, X. S. (2017). Cistrome Data

Browser: a data portal for ChIP-Seq and chromatin accessibility data in human

and mouse. Nucleic Acids Res, 45(D1), D658-D662. doi:10.1093/nar/gkw983

Menendez, J. A., & Lupu, R. (2007). Fatty acid synthase and the lipogenic phenotype

in cancer pathogenesis. Nat Rev Cancer, 7(10), 763-777.

Menschikowski, M., Hagelgans, A., Fuessel, S., Mareninova, O. A., Asatryan, L.,

Wirth, M. P., & Siegert, G. J. I. R. (2013). Serum amyloid A, phospholipase

A2-IIA and C-reactive protein as inflammatory biomarkers for prostate

diseases. 62(12), 1063-1072. doi:10.1007/s00011-013-0665-5

Menschikowski, M., Hagelgans, A., Fuessel, S., Mareninova, O. A., Neumeister, V.,

Wirth, M. P., & Siegert, G. J. I. (2012). Serum Levels of Secreted Group IIA

Phospholipase A2 in Benign Prostatic Hyperplasia and Prostate Cancer: A

Biomarker for Inflammation or Neoplasia? Inflammation, 35(3), 1113-1118.

doi:10.1007/s10753-011-9418-1

Merlot, A. M., Kalinowski, D. S., & Richardson, D. R. (2014). Unraveling the

mysteries of serum albumin-more than just a serum protein. Frontiers in

Physiology, 5, 299-299. doi:10.3389/fphys.2014.00299

Merrill, A. H., Sullards, M. C., Allegood, J. C., Kelly, S., & Wang, E. (2005).

Sphingolipidomics: High-throughput, structure-specific, and quantitative

analysis of sphingolipids by liquid chromatography tandem mass spectrometry.

Methods, 36(2), 207-224. doi:https://doi.org/10.1016/j.ymeth.2005.01.009

Miller, Y. I., Choi, S.-H., Fang, L., & Tsimikas, S. (2010). Lipoprotein Modification

and Macrophage Uptake: Role of Pathologic Cholesterol Transport in

Atherogenesis. In J. R. Harris (Ed.), Cholesterol Binding and Cholesterol

Transport Proteins: Structure and Function in Health and Disease (pp. 229-

251). Dordrecht: Springer Netherlands.

Mostaghel, E. A., Solomon, K. R., Pelton, K., Freeman, M. R., & Montgomery, R. B.

(2012). Impact of circulating cholesterol levels on growth and intratumoral

androgen concentration of prostate tumors. PLOS ONE, 7(1), e30062.


Mounier, C. M., Ghomashchi, F., Lindsay, M. R., James, S., Singer, A. G., Parton, R.

G., & Gelb, M. H. (2004). Arachidonic acid release from mammalian cells

transfected with human groups IIA and X secreted phospholipase A(2) occurs

predominantly during the secretory process and with the involvement of

cytosolic phospholipase A(2)-alpha. J Biol Chem, 279(24), 25024-25038.

doi:10.1074/jbc.M313019200

Murakami, M., & Lambeau, G. (2013). Emerging roles of secreted phospholipase A2

enzymes: An update. Biochimie, 95(1), 43-50.

doi:https://doi.org/10.1016/j.biochi.2012.09.007

Murakami, M., Shimbara, S., Kambe, T., Kuwata, H., Winstead, M. V., Tischfield, J.

A., & Kudo, I. (1998). The functions of five distinct mammalian phospholipase

A2S in regulating arachidonic acid release. Type IIa and type V secretory

phospholipase A2S are functionally redundant and act in concert with cytosolic

phospholipase A2. J Biol Chem, 273(23), 14411-14423.

https://doi.org/10.1016/j.ymeth.2005.01.009

https://doi.org/10.1016/j.biochi.2012.09.007

Bibliography 200

Nambiar, D. K., Deep, G., Singh, R. P., Agarwal, C., & Agarwal, R. (2014). Silibinin

inhibits aberrant lipid metabolism, proliferation and emergence of androgen-

independence in prostate cancer cells via primarily targeting the sterol response

element binding protein 1. Oncotarget, 5(20), 10017-10033.

Nath, A., & Chan, C. (2016a). Genetic alterations in fatty acid transport and

metabolism genes are associated with metastatic progression and poor

prognosis of human cancers. Scientific Reports, 6, 18669.

doi:10.1038/srep18669

Nath, A., & Chan, C. (2016b). Genetic alterations in fatty acid transport and

metabolism genes are associated with metastatic progression and poor

prognosis of human cancers. Scientific Reports, 6, 18669.

doi:10.1038/srep18669

Nieman, K. M., Romero, I. L., Van Houten, B., & Lengyel, E. (2013). Adipose tissue

and adipocytes supports tumorigenesis and metastasis. Biochimica et

biophysica acta, 1831(10), 1533-1541. doi:10.1016/j.bbalip.2013.02.010

Nomura, D. K., Long, J. Z., Niessen, S., Hoover, H. S., Ng, S.-W., & Cravatt, B. F.

(2010). Monoacylglycerol Lipase Regulates a Fatty Acid Network that

Promotes Cancer Pathogenesis. Cell, 140(1), 49-61.

doi:http://dx.doi.org/10.1016/j.cell.2009.11.027

Nouri, M., Ratther, E., Stylianou, N., Nelson, C. C., Hollier, B. G., & Williams, E. D.

(2014). Androgen-Targeted Therapy-Induced Epithelial Mesenchymal

Plasticity and Neuroendocrine Transdifferentiation in Prostate Cancer: An

Opportunity for Intervention. Frontiers in Oncology, 4, 370.

doi:10.3389/fonc.2014.00370

O’Reilly, M. W., House, P. J., & Tomlinson, J. W. (2014). Understanding androgen

action in adipose tissue. The Journal of Steroid Biochemistry and Molecular

Biology, 143, 277-284. doi:http://dx.doi.org/10.1016/j.jsbmb.2014.04.008

Patel, R., Fleming, J., Mui, E., Loveridge, C., Repiscak, P., Blomme, A., Harle, V.,

Salji, M., Ahmad, I., Teo, K., Hamdy, F. C., Hedley, A., van den Broek, N.,

Mackay, G., Edwards, J., Sansom, O. J., & Leung, H. Y. (2018). Sprouty2 loss-

induced IL6 drives castration-resistant prostate cancer through scavenger

receptor B1. EMBO Mol Med, 10(4). doi:10.15252/emmm.201708347

PCF. (2018a). Prostate Cancer Overview. Retrieved from https://www.pcf.org/about-

prostate-cancer/what-is-prostate-cancer/

PCF. (2018b). Prostate Cancer Survival Rates. Retrieved from

https://www.pcf.org/about-prostate-cancer/what-is-prostate-cancer/prostate-

cancer-survival-rates/

Peck, B., Schug, Z. T., Zhang, Q., Dankworth, B., Jones, D. T., Smethurst, E., Patel,

R., Mason, S., Jiang, M., Saunders, R., Howell, M., Mitter, R., Spencer-Dene,

B., Stamp, G., McGarry, L., James, D., Shanks, E., Aboagye, E. O., Critchlow,

S. E., Leung, H. Y., Harris, A. L., Wakelam, M. J. O., Gottlieb, E., & Schulze,

A. (2016). Inhibition of fatty acid desaturation is detrimental to cancer cell

survival in metabolically compromised environments. Cancer & Metabolism,

4, 6. doi:10.1186/s40170-016-0146-8

Petan, T., Jarc, E., & Jusović, M. (2018). Lipid Droplets in Cancer: Guardians of Fat

in a Stressful World. Molecules (Basel, Switzerland), 23(8), 1941.

doi:10.3390/molecules23081941

Pezaro, C., Woo, H. H., & Davis, I. D. (2014a). Prostate cancer: measuring PSA.

Internal Medicine Journal, 44(5), 433-440. doi:doi:10.1111/imj.12407

http://dx.doi.org/10.1016/j.cell.2009.11.027

http://dx.doi.org/10.1016/j.jsbmb.2014.04.008

https://www.pcf.org/about-prostate-cancer/what-is-prostate-cancer/

https://www.pcf.org/about-prostate-cancer/what-is-prostate-cancer/

https://www.pcf.org/about-prostate-cancer/what-is-prostate-cancer/prostate-cancer-survival-rates/

https://www.pcf.org/about-prostate-cancer/what-is-prostate-cancer/prostate-cancer-survival-rates/

Bibliography 201

Pinthus, J. H., Lu, J. P., Bidaisee, L. A., Lin, H., Bryskine, I., Gupta, R. S., & Singh,

G. (2007). Androgen-dependent regulation of medium and long chain fatty

acids uptake in prostate cancer. Prostate, 67(12), 1330-1338.

doi:10.1002/pros.20609

Qu, J., Zhao, X., Wang, J., Liu, C., Sun, Y., Cai, H., & Liu, J. (2018). Plasma

phospholipase A2 activity may serve as a novel diagnostic biomarker for the

diagnosis of breast cancer. Oncol Lett, 15(4), 5236-5242.

doi:10.3892/ol.2018.7915

Quinn, P. J., & Wolf, C. (2009). The liquid-ordered phase in membranes. Biochimica

et Biophysica Acta (BBA) - Biomembranes, 1788(1), 33-46.

doi:https://doi.org/10.1016/j.bbamem.2008.08.005

Ramirez, M., Rajaram, S., Steininger, R. J., Osipchuk, D., Roth, M. A., Morinishi, L.

S., Evans, L., Ji, W., Hsu, C.-H., Thurley, K., Wei, S., Zhou, A., Koduru, P.

R., Posner, B. A., Wu, L. F., & Altschuler, S. J. (2016). Diverse drug-resistance

mechanisms can emerge from drug-tolerant cancer persister cells. Nature

Communications, 7, 10690-10690. doi:10.1038/ncomms10690

Ramos‐Montoya, A., Lamb, A. D., Russell, R., Carroll, T., Jurmeister, S., Galeano‐

Dalmau, N., Massie, C. E., Boren, J., Bon, H., Theodorou, V., Vias, M., Shaw,

G. L., Sharma, N. L., Ross‐Adams, H., Scott, H. E., Vowler, S. L., Howat, W.

J., Warren, A. Y., Wooster, R. F., Mills, I. G., & Neal, D. E. (2014). HES6

drives a critical AR transcriptional programme to induce castration‐resistant

prostate cancer through activation of an E2F1‐mediated cell cycle network.

EMBO Molecular Medicine, 6(5), 651-661. doi:10.1002/emmm.201303581

Rampler, E., Coman, C., Hermann, G., Sickmann, A., Ahrends, R., & Koellensperger,

G. (2017). LILY-lipidome isotope labeling of yeast: in vivo synthesis of (13)C

labeled reference lipids for quantification by mass spectrometry. Analyst,

142(11), 1891-1899. doi:10.1039/c7an00107j

Raynor, A., Jantscheff, P., Ross, T., Schlesinger, M., Wilde, M., Haasis, S.,

Dreckmann, T., Bendas, G., & Massing, U. (2015). Saturated and mono-

unsaturated lysophosphatidylcholine metabolism in tumour cells: a potential

therapeutic target for preventing metastases. Lipids in health and disease, 14,

69-69. doi:10.1186/s12944-015-0070-x

Reid, R. C. (2005). Inhibitors of secretory phospholipase A2 group IIA. Curr Med

Chem, 12(25), 3011-3026.

Rhodes, D., Yu, J., Shanker, K., Deshpande, N., Varambally, R., Ghosh, D., Barrette,

T., Pandey, A., Chinnaiyan, A. (2004) ONCOMINE: A cancer microarray

database and integrated data-mining platform. Neoplasia, 6(1), 1-6. doi:

10.1016/s1476-5586(04)80047-2

Richieri, G. V., & Kleinfeld, A. M. (1995). Unbound free fatty acid levels in human

serum. Journal of Lipid Research, 36(2), 229-240.

Risbridger, G. P., Davis, I. D., Birrell, S. N., & Tilley, W. D. (2010). Breast and

prostate cancer: more similar than different. Nat Rev Cancer, 10(3), 205-212.

Robinson, M. D., McCarthy, D. J., & Smyth, G. K. (2010). edgeR: a Bioconductor

package for differential expression analysis of digital gene expression data.

Bioinformatics, 26(1), 139-140. doi:10.1093/bioinformatics/btp616

Robinson, M. D., & Oshlack, A. (2010). A scaling normalization method for

differential expression analysis of RNA-seq data. Genome Biol, 11(3), R25.

doi:10.1186/gb-2010-11-3-r25

Rolin, J., & Maghazachi, A. A. (2011). Effects of lysophospholipids on tumor

microenvironment. Cancer microenvironment : official journal of the

https://doi.org/10.1016/j.bbamem.2008.08.005

Bibliography 202

International Cancer Microenvironment Society, 4(3), 393-403.

doi:10.1007/s12307-011-0088-1

Rubenstein, J. L., Smith, B. A., & McConnell, H. M. (1979). Lateral diffusion in binary

mixtures of cholesterol and phosphatidylcholines. Proceedings of the National

Academy of Sciences, 76(1), 15-18. doi:10.1073/pnas.76.1.15

Rustom, A., Saffrich, R., Markovic, I., Walther, P., & Gerdes, H.-H. (2004).

Nanotubular Highways for Intercellular Organelle Transport. Science,

303(5660), 1007-1010. doi:10.1126/science.1093133

Rysman, E., Brusselmans, K., Scheys, K., Timmermans, L., Derua, R., Munck, S., Van

Veldhoven, P. P., Waltregny, D., Daniels, V. W., Machiels, J., Vanderhoydonc,

F., Smans, K., Waelkens, E., Verhoeven, G., & Swinnen, J. V. (2010). De novo

lipogenesis protects cancer cells from free radicals and chemotherapeutics by

promoting membrane lipid saturation. Cancer Res, 70(20), 8117-8126.

doi:10.1158/0008-5472.CAN-09-3871

Saad, F. (2013). Evidence for the efficacy of enzalutamide in postchemotherapy

metastatic castrate-resistant prostate cancer. Therapeutic advances in urology,

5(4), 201-210. doi:10.1177/1756287213490054

Sadowski, M. C., Pouwer, R. H., Gunter, J. H., Lubik, A. A., Quinn, R. J., & Nelson,

C. C. (2014). The fatty acid synthase inhibitor triclosan: repurposing an anti-

microbial agent for targeting prostate cancer. Oncotarget, 5(19), 9362-9381.

doi:10.18632/oncotarget.2433

Sahoo, S., Aurich, M. K., Jonsson, J. J., & Thiele, I. (2014). Membrane transporters in

a human genome-scale metabolic knowledgebase and their implications for

disease. Frontiers in Physiology, 5, 91. doi:10.3389/fphys.2014.00091

Schaloske, R. H., & Dennis, E. A. (2006). The phospholipase A2 superfamily and its

group numbering system. Biochimica et Biophysica Acta (BBA) - Molecular

and Cell Biology of Lipids, 1761(11), 1246-1259.

doi:https://doi.org/10.1016/j.bbalip.2006.07.011

Schcolnik-Cabrera, A., Chávez-Blanco, A., Domínguez-Gómez, G., Taja-Chayeb, L.,

Morales-Barcenas, R., Trejo-Becerril, C., Perez-Cardenas, E., Gonzalez-

Fierro, A., & Dueñas-González, A. (2018) Orlistat as a FASN inhibitor and

multitargeted agent for cancer therapy. Expert Opinion on Investigational

Drugs, 27:5, 475-489, DOI: 10.1080/13543784.2018.1471132

Schevitz, R. W., Bach, N. J., Carlson, D. G., Chirgadze, N. Y., Clawson, D. K., Dillard,

R. D., Draheim, S. E., Hartley, L. W., Jones, N. D., Mihelich, E. D., & et al.

(1995). Structure-based design of the first potent and selective inhibitor of

human non-pancreatic secretory phospholipase A2. Nat Struct Biol, 2(6), 458-

465.

Schlaepfer, I. R., Glodé, L. M., Hitz, C. A., Pac, C. T., Boyle, K. E., Maroni, P., Deep,

G., Agarwal, R., Lucia, S. M., Cramer, S. D., Serkova, N. J., & Eckel, R. H.

(2015). Inhibition of Lipid Oxidation Increases Glucose Metabolism and

Enhances 2-Deoxy-2-[(18)F]Fluoro-D-Glucose Uptake in Prostate Cancer

Mouse Xenografts. Molecular imaging and biology, 17(4), 529-538.

doi:10.1007/s11307-014-0814-4

Schmittgen, T. D., & Livak, K. J. (2008). Analyzing real-time PCR data by the

comparative C(T) method. Nat Protoc, 3(6), 1101-1108.

Schmitz, G., & Ecker, J. (2008). The opposing effects of n−3 and n−6 fatty acids.

Progress in Lipid Research, 47(2), 147-155.

doi:https://doi.org/10.1016/j.plipres.2007.12.004

https://doi.org/10.1016/j.bbalip.2006.07.011

https://doi.org/10.1080/13543784.2018.1471132

https://doi.org/10.1016/j.plipres.2007.12.004

Bibliography 203

Schneider, W. J. (2016). Chapter 17 - Lipoprotein Receptors. In N. D. Ridgway & R.

S. McLeod (Eds.), Biochemistry of Lipids, Lipoproteins and Membranes (Sixth

Edition) (pp. 489-518). Boston: Elsevier.

Schörghofer, D., Kinslechner, K., Preitschopf, A., Schütz, B., Röhrl, C.,

Hengstschläger, M., Stangl, H., & Mikula, M. (2015a). The HDL receptor SR-

BI is associated with human prostate cancer progression and plays a possible

role in establishing androgen independence. Reproductive Biology and

Endocrinology, 13(1), 88. doi:10.1186/s12958-015-0087-z

Schörghofer, D., Kinslechner, K., Preitschopf, A., Schütz, B., Röhrl, C.,

Hengstschläger, M., Stangl, H., & Mikula, M. (2015b). The HDL receptor SR-

BI is associated with human prostate cancer progression and plays a possible

role in establishing androgen independence. Reproductive Biology and

Endocrinology : RB&E, 13, 88. doi:10.1186/s12958-015-0087-z

Seim, I., Lubik, A. A., Lehman, M. L., Tomlinson, N., Whiteside, E. J., Herington, A.

C., Nelson, C. C., & Chopin, L. K. (2013). Cloning of a novel insulin-regulated

ghrelin transcript in prostate cancer. J Mol Endocrinol, 50(2), 179-191.

doi:10.1530/jme-12-0150

Senkal, C. E., Salama, M. F., Snider, A. J., Allopenna, J. J., Rana, N. A., Koller, A.,

Hannun, Y. A., & Obeid, L. M. (2017). Ceramide Is Metabolized to

Acylceramide and Stored in Lipid Droplets. Cell metabolism, 25(3), 686-697.

doi:10.1016/j.cmet.2017.02.010

Sfanos, K. S., & De Marzo, A. M. (2012). Prostate cancer and inflammation: the

evidence. Histopathology, 60(1), 199-215. doi:10.1111/j.1365-

2559.2011.04033.x

Shah, R. B., & Zhou, M. (2016). Recent advances in prostate cancer pathology:

Gleason grading and beyond. Pathology International, 66(5), 260-272.

doi:doi:10.1111/pin.12398

Sieh, S., Taubenberger, A. V., Rizzi, S. C., Sadowski, M., Lehman, M. L., Rockstroh,

A., An, J., Clements, J. A., Nelson, C. C., & Hutmacher, D. W. (2012).

Phenotypic Characterization of Prostate Cancer LNCaP Cells Cultured within

a Bioengineered Microenvironment. PLOS ONE, 7(9), e40217.


Sikora, M. J., Johnson, M. D., Lee, A. V., & Oesterreich, S. (2016). Endocrine

Response Phenotypes Are Altered by Charcoal-Stripped Serum Variability.

Endocrinology, 157(10), 3760-3766. doi:10.1210/en.2016-1297

Singh, D., Febbo, P. G., Ross, K., Jackson, D. G., Manola, J., Ladd, C., Tamayo, P.,

Renshaw, A. A., D'Amico, A. V., Richie, J. P., Lander, E. S., Loda, M.,

Kantoff, P. W., Golub, T. R., & Sellers, W. R. (2002). Gene expression

correlates of clinical prostate cancer behavior. Cancer Cell, 1(2), 203-209.


Six, D. A., & Dennis, E. A. (2000). The expanding superfamily of phospholipase A2

enzymes: classification and characterization. Biochimica et Biophysica Acta

(BBA) - Molecular and Cell Biology of Lipids, 1488(1), 1-19.


Smyth, G. K. (2005). Limma: linear models for microarray data. Bioinformatics and

Computational Biology Solutions Using R and Bioconductor. 397-420.

Snyder, D. W., Bach, N. J., Dillard, R. D., Draheim, S. E., Carlson, D. G., Fox, N.,

Roehm, N. W., Armstrong, C. T., Chang, C. H., Hartley, L. W., Johnson, L.

M., Roman, C. R., Smith, A. C., Song, M., & Fleisch, J. H. (1999).

Pharmacology of LY315920/S-5920, [[3-(aminooxoacetyl)-2-ethyl-1-

https://doi.org/10.1016/S1535-6108(02)00030-2

https://doi.org/10.1016/S1388-1981(00)00105-0

Bibliography 204

(phenylmethyl)-1H-indol-4-yl]oxy] acetate, a potent and selective secretory

phospholipase A2 inhibitor: A new class of anti-inflammatory drugs, SPI. J

Pharmacol Exp Ther, 288(3), 1117-1124.

Soekmadji, C., Russell, P., & Nelson, C. (2013). Exosomes in Prostate Cancer: Putting

Together the Pieces of a Puzzle. Cancers, 5(4), 1522.

Stone, S. J., Myers, H. M., Watkins, S. M., Brown, B. E., Feingold, K. R., Elias, P. M.,

& Farese, R. V. (2004). Lipopenia and Skin Barrier Abnormalities in DGAT2-

deficient Mice. Journal of Biological Chemistry, 279(12), 11767-11776.

doi:10.1074/jbc.M311000200

Stopsack, K. H., Gerke, T. A., Andrén, O., Andersson, S.-O., Giovannucci, E. L.,

Mucci, L. A., & Rider, J. R. (2017). Cholesterol uptake and regulation in high-

grade and lethal prostate cancers. Carcinogenesis, 38(8), 806-811.

doi:10.1093/carcin/bgx058

Streicher, R., Kotzka, J., Muller-Wieland, D., Siemeister, G., Munck, M., Avci, H., &

Krone, W. (1996). SREBP-1 mediates activation of the low density lipoprotein

receptor promoter by insulin and insulin-like growth factor-I. J Biol Chem,

271(12), 7128-7133.

Stump, D. D., Zhou, S. L., & Berk, P. D. (1993). Comparison of plasma membrane

FABP and mitochondrial isoform of aspartate aminotransferase from rat liver.

American Journal of Physiology-Gastrointestinal and Liver Physiology,

265(5), G894-G902. doi:10.1152/ajpgi.1993.265.5.G894

Subczynski, W. K., Pasenkiewicz-Gierula, M., Widomska, J., Mainali, L., & Raguz,

M. (2017). High Cholesterol/Low Cholesterol: Effects in Biological

Membranes: A Review. Cell Biochemistry and Biophysics, 75(3), 369-385.

doi:10.1007/s12013-017-0792-7

Subczynski, W. K., & Wisniewska, A. (2000). Physical properties of lipid bilayer

membranes: relevance to membrane biological functions. Acta Biochim Pol,

47(3), 613-625.

Subramanian, A., Tamayo, P., Mootha, V. K., Mukherjee, S., Ebert, B. L., Gillette, M.

A., Paulovich, A., Pomeroy, S. L., Golub, T. R., Lander, E. S., & Mesirov, J.

P. (2005). Gene set enrichment analysis: a knowledge-based approach for

interpreting genome-wide expression profiles. Proc Natl Acad Sci U S A,

102(43), 15545-15550. doi:10.1073/pnas.0506580102

Suburu, J., & Chen, Y. Q. (2012). Lipids and Prostate Cancer. Prostaglandins & other

lipid mediators, 98(0), 1-10. doi:10.1016/j.prostaglandins.2012.03.003

Sun, Y., Wang, B.-E., Leong, K. G., Yue, P., Li, L., Jhunjhunwala, S., Chen, D., Seo,

K., Modrusan, Z., Gao, W.-Q., Settleman, J., & Johnson, L. (2012). Androgen

Deprivation Causes Epithelial–Mesenchymal Transition in the Prostate:

Implications for Androgen-Deprivation Therapy. Cancer Research, 72(2),

527-536. doi:10.1158/0008-5472.can-11-3004

Swinnen, Heemers, H., van de Sande, T., de Schrijver, E., Brusselmans, K., Heyns,

W., & Verhoeven, G. (2004). Androgens, lipogenesis and prostate cancer. J

Steroid Biochem Mol Biol, 92(4), 273-279. doi:10.1016/j.jsbmb.2004.10.013

Swinnen, Roskams, T., Joniau, S., Van Poppel, H., Oyen, R., Baert, L., Heyns, W., &

Verhoeven, G. (2002a). Overexpression of fatty acid synthase is an early and

common event in the development of prostate cancer. Int J Cancer, 98(1), 19-

22.

Swinnen, Vanderhoydonc, F., Elgamal, A. A., Eelen, M., Vercaeren, I., Joniau, S., Van

Poppel, H., Baert, L., Goossens, K., Heyns, W., & Verhoeven, G. (2000).

Bibliography 205

Selective activation of the fatty acid synthesis pathway in human prostate

cancer. Int J Cancer, 88(2), 176-179.

Swinnen, J. V., Brusselmans, K., & Verhoeven, G. (2006). Increased lipogenesis in

cancer cells: new players, novel targets. Curr Opin Clin Nutr Metab Care, 9(4),

358-365. doi:10.1097/01.mco.0000232894.28674.30

Swinnen, J. V., Esquenet, M., Goossens, K., Heyns, W., & Verhoeven, G. (1997).

Androgens stimulate fatty acid synthase in the human prostate cancer cell line

LNCaP. Cancer Res, 57(6), 1086-1090.

Swinnen, J. V., Ulrix, W., Heyns, W., & Verhoeven, G. (1997). Coordinate regulation

of lipogenic gene expression by androgens: evidence for a cascade mechanism

involving sterol regulatory element binding proteins. Proc Natl Acad Sci U S

A, 94(24), 12975-12980.

Swinnen, J. V., Van Veldhoven, P. P., Esquenet, M., Heyns, W., & Verhoeven, G.

(1996a). Androgens markedly stimulate the accumulation of neutral lipids in

the human prostatic adenocarcinoma cell line LNCaP. Endocrinology, 137(10),

4468-4474. doi:10.1210/endo.137.10.8828509

Swinnen, J. V., Van Veldhoven, P. P., Esquenet, M., Heyns, W., & Verhoeven, G.

(1996b). Androgens markedly stimulate the accumulation of neutral lipids in

the human prostatic adenocarcinoma cell line LNCaP. Endocrinology, 137(10),

4468-4474. doi:10.1210/endo.137.10.8828509

Swinnen, J. V., Van Veldhoven, P. P., Timmermans, L., De Schrijver, E.,

Brusselmans, K., Vanderhoydonc, F., Van de Sande, T., Heemers, H., Heyns,

W., & Verhoeven, G. (2003). Fatty acid synthase drives the synthesis of

phospholipids partitioning into detergent-resistant membrane microdomains.

Biochemical and Biophysical Research Communications, 302(4), 898-903.

doi:http://dx.doi.org/10.1016/S0006-291X(03)00265-1

Sztalryd, C., & Brasaemle, D. L. (2017). The perilipin family of lipid droplet proteins:

Gatekeepers of intracellular lipolysis. Biochimica et biophysica acta.

Molecular and cell biology of lipids, 1862(10 Pt B), 1221-1232.

doi:10.1016/j.bbalip.2017.07.009

Tallima, H., & El Ridi, R. (2018). Arachidonic acid: Physiological roles and potential

health benefits – A review. Journal of Advanced Research, 11, 33-41.

doi:https://doi.org/10.1016/j.jare.2017.11.004

Tamura, K., Makino, A., Hullin-Matsuda, F., Kobayashi, T., Furihata, M., Chung, S.,

Ashida, S., Miki, T., Fujioka, T., Shuin, T., Nakamura, Y., & Nakagawa, H.

(2009). Novel Lipogenic Enzyme ELOVL7 Is Involved in Prostate Cancer

Growth through Saturated Long-Chain Fatty Acid Metabolism. Cancer

Research, 69(20), 8133-8140. doi:10.1158/0008-5472.can-09-0775

Tamura, Y., Osuga, J., Adachi, H., Tozawa, R., Takanezawa, Y., Ohashi, K., Yahagi,

N., Sekiya, M., Okazaki, H., Tomita, S., Iizuka, Y., Koizumi, H., Inaba, T.,

Yagyu, H., Kamada, N., Suzuki, H., Shimano, H., Kadowaki, T., Tsujimoto,

M., Arai, H., Yamada, N., & Ishibashi, S. (2004). Scavenger receptor expressed

by endothelial cells I (SREC-I) mediates the uptake of acetylated low density

lipoproteins by macrophages stimulated with lipopolysaccharide. J Biol Chem,

279(30), 30938-30944. doi:10.1074/jbc.M313088200

Taylor, R. A., Lo, J., Ascui, N., & Watt, M. J. (2015). Linking obesogenic

dysregulation to prostate cancer progression. Endocrine Connections, 4(4),

R68-R80. doi:10.1530/EC-15-0080

http://dx.doi.org/10.1016/S0006-291X(03)00265-1

https://doi.org/10.1016/j.jare.2017.11.004

Bibliography 206

Tekpli, X., Holme, J. A., Sergent, O., & Lagadic-Gossmann, D. (2013). Role for

membrane remodeling in cell death: Implication for health and disease.

Toxicology, 304, 141-157. doi:https://doi.org/10.1016/j.tox.2012.12.014

Thysell, E., Surowiec, I., Hörnberg, E., Crnalic, S., Widmark, A., Johansson, A. I.,

Stattin, P., Bergh, A., Moritz, T., Antti, H., & Wikström, P. (2010).

Metabolomic Characterization of Human Prostate Cancer Bone Metastases

Reveals Increased Levels of Cholesterol. PLOS ONE, 5(12), e14175.


Tomlins, S. A., Mehra, R., Rhodes, D. R., Cao, X., Wang, L., Dhanasekaran, S. M.,

Kalyana-Sundaram, S., Wei, J. T., Rubin, M. A., Pienta, K. J., Shah, R. B., &

Chinnaiyan, A. M. (2006). Integrative molecular concept modeling of prostate

cancer progression. Nature Genetics, 39, 41. doi:10.1038/ng1935

https://www.nature.com/articles/ng1935#supplementary-information

Tousignant, K. D., Rockstroh, A., Taherian Fard, A., Lehman, M. L., Wang, C.,

McPherson, S. J., Philp, L. K., Bartonicek, N., Dinger, M. E., Nelson, C. C., &

Sadowski, M. C. (2019). Lipid uptake is an androgen-enhanced lipid supply

pathway associated with prostate cancer disease progression and bone

metastasis. Molecular Cancer Research. 1147.2018. doi:10.1158/1541-

7786.MCR-18-1147

Tran, C., Ouk, S., Clegg, N. J., Chen, Y., Watson, P. A., Arora, V., Wongvipat, J.,

Smith-Jones, P. M., Yoo, D., Kwon, A., Wasielewska, T., Welsbie, D., Chen,

C. D., Higano, C. S., Beer, T. M., Hung, D. T., Scher, H. I., Jung, M. E., &

Sawyers, C. L. (2009). Development of a second-generation antiandrogen for

treatment of advanced prostate cancer. Science (New York, N.Y.), 324(5928),

787-790. doi:10.1126/science.1168175

Twiddy, A. L., Cox, M. E., & Wasan, K. M. (2012). Knockdown of scavenger receptor

class B type I reduces prostate specific antigen secretion and viability of

prostate cancer cells. Prostate, 72. doi:10.1002/pros.21499

Twiddy, A. L., Leon, C. G., & Wasan, K. M. (2011). Cholesterol as a Potential Target

for Castration-Resistant Prostate Cancer. Pharmaceutical Research, 28(3),

423-437. doi:10.1007/s11095-010-0210-y

Udayakumar, T. S., Stoyanova, R., Shareef, M. M., Mu, Z., Philip, S., Burnstein, K.

L., & Pollack, A. (2016). Edelfosine Promotes Apoptosis in Androgen-

Deprived Prostate Tumors by Increasing ATF3 and Inhibiting Androgen

Receptor Activity. Molecular Cancer Therapeutics, 15(6), 1353-1363.

doi:10.1158/1535-7163.MCT-15-0332

Vallette, F. M., Olivier, C., Lézot, F., Oliver, L., Cochonneau, D., Lalier, L., Cartron,

P.-F., & Heymann, D. (2018). Dormant, quiescent, tolerant and persister cells:

four synonyms for the same target in cancer. Biochemical Pharmacology.

doi:https://doi.org/10.1016/j.bcp.2018.11.004

van Meer, G., Voelker, D. R., & Feigenson, G. W. (2008). Membrane lipids: where

they are and how they behave. Nature reviews. Molecular cell biology, 9(2),

112-124. doi:10.1038/nrm2330

Vander Heiden, M. G., Cantley, L. C., & Thompson, C. B. (2009). Understanding the

Warburg effect: the metabolic requirements of cell proliferation. Science,

324(5930), 1029-1033. doi:10.1126/science.1160809

Varambally, S., Yu, J., Laxman, B., Rhodes, D. R., Mehra, R., Tomlins, S. A., Shah,

R. B., Chandran, U., Monzon, F. A., Becich, M. J., Wei, J. T., Pienta, K. J.,

Ghosh, D., Rubin, M. A., & Chinnaiyan, A. M. (2005). Integrative genomic

https://doi.org/10.1016/j.tox.2012.12.014

https://doi.org/10.1016/j.bcp.2018.11.004

Bibliography 207

and proteomic analysis of prostate cancer reveals signatures of metastatic

progression. Cancer Cell, 8(5), 393-406. doi:10.1016/j.ccr.2005.10.001

Verheij, M., Bose, R., Hua Lin, X., Yao, B., Jarvis, W. D., Grant, S., Birrer, M. J.,

Szabo, E., Zon, L. I., Kyriakis, J. M., Haimovitz-Friedman, A., Fuks, Z., &

Kolesnick, R. N. (1996). Requirement for ceramide-initiated SAPK/JNK

signalling in stress-induced apoptosis. Nature, 380(6569), 75-79.

doi:10.1038/380075a0

Vijayaraghavalu, S., Peetla, C., Lu, S., & Labhasetwar, V. (2012). Epigenetic

modulation of the biophysical properties of drug-resistant cell lipids to restore

drug transport and endocytic functions. Molecular pharmaceutics, 9(9), 2730-

2742. doi:10.1021/mp300281t

Wang, M.-D., Kiss, R. S., Franklin, V., McBride, H. M., Whitman, S. C., & Marcel,

Y. L. (2007). Different cellular traffic of LDL-cholesterol and acetylated LDL-

cholesterol leads to distinct reverse cholesterol transport pathways. Journal of

Lipid Research, 48(3), 633-645. doi:10.1194/jlr.M600470-JLR200

Wang, M., Hao, F. Y., Wang, J. G., & Xiao, W. (2014). Group IIa secretory

phospholipase A2 (sPLA2IIa) and progression in patients with lung cancer.

Eur Rev Med Pharmacol Sci, 18(18), 2648-2654.

Warburg, O., Wind, F., & Negelein, E. (1927). The Metabolism of Tumors in the

Body. J Gen Physiol, 8(6), 519-530.

Wilfling, F., Haas, J. T., Walther, T. C., & Jr, R. V. F. (2014). Lipid droplet biogenesis.

Current Opinion in Cell Biology, 29, 39-45.

doi:https://doi.org/10.1016/j.ceb.2014.03.008

Wolf, A. M., Wender, R. C., Etzioni, R. B., Thompson, I. M., D'Amico, A. V., Volk,

R. J., Brooks, D. D., Dash, C., Guessous, I., Andrews, K., DeSantis, C., Smith,

R. A., & American Cancer Society Prostate Cancer Advisory, C. (2010).

American Cancer Society guideline for the early detection of prostate cancer:

update 2010. CA Cancer J Clin, 60(2), 70-98. doi:10.3322/caac.20066

Wong, M. C. S., Goggins, W. B., Wang, H. H. X., Fung, F. D. H., Leung, C., Wong,

S. Y. S., Ng, C. F., & Sung, J. J. Y. (2016). Global Incidence and Mortality for

Prostate Cancer: Analysis of Temporal Patterns and Trends in 36 Countries.

European Urology, 70(5), 862-874.

doi:https://doi.org/10.1016/j.eururo.2016.05.043

Xia, J., Mandal, R., Sinelnikov, I., Broadhurst, D., & Wishart, D. (2012).

MetaboAnalyst 2.0—a comprehensive server for metabolomic data

analysis, Nucleic Acids Research, 40(1), W127–

W133, https://doi.org/10.1093/nar/gks374

Xu, G., Wu, J., Zhou, L., Chen, B., Sun, Z., Zhao, F., & Tao, Z. (2010).

Characterization of the Small RNA Transcriptomes of Androgen Dependent

and Independent Prostate Cancer Cell Line by Deep Sequencing. PLOS ONE,

5(11), e15519. doi:10.1371/journal.pone.0015519

Yang, P., Cartwright, C. A., Li, J., Wen, S., Prokhorova, I. N., Shureiqi, I., Troncoso,

P., Navone, N. M., Newman, R. A., & Kim, J. (2012). Arachidonic acid

metabolism in human prostate cancer. International journal of oncology, 41(4),

1495-1503. doi:10.3892/ijo.2012.1588

Yedgar, S., Lichtenberg, D., & Schnitzer, E. (2000). Inhibition of phospholipase A2

as a therapeutic target. Biochimica et Biophysica Acta (BBA) - Molecular and

Cell Biology of Lipids, 1488(1), 182-187. doi:https://doi.org/10.1016/S1388-

1981(00)00120-7

https://doi.org/10.1016/j.ceb.2014.03.008


https://doi.org/10.1093/nar/gks374

https://doi.org/10.1016/S1388-1981(00)00120-7

https://doi.org/10.1016/S1388-1981(00)00120-7

Bibliography 208

Yen, C.-L. E., Monetti, M., Burri, B. J., & Farese, R. V. (2005). The triacylglycerol

synthesis enzyme DGAT1 also catalyzes the synthesis of diacylglycerols,

waxes, and retinyl esters. Journal of Lipid Research, 46(7), 1502-1511.

doi:10.1194/jlr.M500036-JLR200

Yen, C.-L. E., Stone, S. J., Koliwad, S., Harris, C., & Farese, R. V., Jr. (2008). DGAT

enzymes and triacylglycerol biosynthesis. Journal of Lipid Research, 49(11),

2283-2301. doi:10.1194/jlr.R800018-JLR200

Yokoyama, C., Wang, X., Briggs, M. R., Admon, A., Wu, J., Hua, X., Goldstein, J. L.,

& Brown, M. S. (1993). SREBP-1, a basic-helix-loop-helix-leucine zipper

protein that controls transcription of the low density lipoprotein receptor gene.

Cell, 75(1), 187-197.

Yu, P., Duan, X., Cheng, Y., Liu, C., Chen, Y., Liu, W., Yin, B., Wang, X., & Tao, Z.

(2017). Androgen-independent LNCaP cells are a subline of LNCaP cells with

a more aggressive phenotype and androgen suppresses their growth by

inducing cell cycle arrest at the G1 phase. International journal of molecular

medicine, 40(5), 1426-1434. doi:10.3892/ijmm.2017.3125

Yue, S., Li, J., Lee, S.-Y., Lee, Hyeon J., Shao, T., Song, B., Cheng, L., Masterson,

Timothy A., Liu, X., Ratliff, Timothy L., & Cheng, J.-X. (2014). Cholesteryl

Ester Accumulation Induced by PTEN Loss and PI3K/AKT Activation

Underlies Human Prostate Cancer Aggressiveness. Cell metabolism, 19(3),

393-406. doi:http://dx.doi.org/10.1016/j.cmet.2014.01.019

Zadra, G., Photopoulos, C., & Loda, M. (2013). The fat side of Prostate Cancer.

Biochimica et biophysica acta, 1831(10), 1518-1532.

doi:10.1016/j.bbalip.2013.03.010

Zalba, S., & Ten Hagen, T. L. M. (2017). Cell membrane modulation as adjuvant in

cancer therapy. Cancer Treatment Reviews, 52, 48-57.

doi:10.1016/j.ctrv.2016.10.008

Zhuang, L., Kim, J., Adam, R. M., Solomon, K. R., & Freeman, M. R. (2005).

Cholesterol targeting alters lipid raft composition and cell survival in prostate

cancer cells and xenografts. Journal of Clinical Investigation, 115(4), 959-968.

doi:10.1172/JCI200519935

http://dx.doi.org/10.1016/j.cmet.2014.01.019