Pre-Clinical Analysis of Fluvastatin As a Metastasis-Prevention Agent in Breast Cancer

Pre-clinical analysis of fluvastatin as a metastasis-prevention agent in breast cancer

Rosemary Yu

A thesis submitted in conformity with the requirements for the degree of Doctor of Philosophy Department of Medical Biophysics University of Toronto

Pre-clinical analysis of fluvastatin as a metastasis-prevention agent in breast cancer

Rosemary Yu

Doctor of Philosophy

Department of Medical Biophysics University of Toronto

2017 Abstract

Fluvastatin is a member of the statin family of drugs, widely prescribed to lower serum cholesterol. Accumulating evidence from epidemiological, pre-clinical, and clinical studies indicate that statins have pleiotropic anti-cancer activities. Previous work have suggested a mechanistic model where depletion of farnesyl diphosphate (FPP) and geranylgeranyl diphosphate (GGPP), and consequently inhibition of prenylation of RAS family members, underlies statin-induced cancer cell death. However, several lines of evidence suggest that this model is incomplete, yet alternative mechanisms have not been described. The purpose of this thesis is to (i) elucidate the molecular mechanism of fluvastatin sensitivity in breast cancer cells, and (ii) identify the point of therapeutic intervention in which fluvastatin treatment will provide clinical benefit in breast cancer patients. We first demonstrate that fluvastatin-induced cell death can be uncoupled from inhibition of protein prenylation, in breast cancer cells that have undergone epithelial-mesenchymal transition (EMT). In these cells, the mechanism of fluvastatin sensitivity is an increased dependency on the biosynthesis of dolichol, required for co- translational protein N-glycosylation. Fluvastatin treatment impairs the expression of tri- and tetra-antennary β1,6-branched complex type N-glycans associated with breast cancer cell EMT and metastasis. These results, along with supporting evidence from epidemiological studies, provided rationale to examine the efficacy of fluvastatin as a metastasis-prevention agent in ii breast cancer. Using a mouse model of post-surgical metastatic breast cancer that fully recapitulates the intravasation and dissemination process, we show that fluvastatin treatment in the adjuvant setting significantly delays breast cancer metastasis, reduces metastatic burden, and improves overall survival. Overall, this research provides novel insight into the mechanism of the anti-cancer effects of fluvastatin, and demonstrates efficacy of fluvastatin use in the adjuvant therapeutic setting to prevent breast cancer metastasis.

iii

Acknowledgments

To my supervisor, Dr. Linda Penn, thank you for your unwavering support throughout my time in your lab. I am sincerely grateful for the freedom and belief that you have given me to pursue my research ideas; your scientific guidance and intellectual support; and above all, the rich and engaging environment that you have built in which to learn, explore, and grow. For all this and a lot more, thank you.

To my supervisory committee members, Dr. Mark Minden and Dr. Brad Wouters, thank you both for your insight and advice. I have always felt more inspired about my research and its contribution to the field of cancer research after discussing my projects with you.

I have been privileged to collaborate with several outstanding scientists, and I am indebted to their knowledge and expertise. I wish to thank Dr. Wail Ba-Alawi, Dr. Eric Chen, Dr. Jim Dennis, Dr. Benjamin Haibe-Kains, Dr. Cunjie Zhang, and Dr. Wenjiang Zhang, who have all been instrumental in shaping this thesis.

To my lab mates, thank you for the unforgettable memories. I will miss working with such a fun and passionate team of people. To Joe and Jenna, thank you for the long hours in the animal house. To Manpreet and Diana, thank you for your support both in and out of the lab. To Ashley, Corey, Aaliya, and Jason, thank you for your encouragement and humour that kept me going. To all former Penn lab members, thank you for the good times.

To Broom-Broom, you have been a constant source of comfort and joy. To Mom and Dad, thank you for your never-ending love.

Table of Contents

Acknowledgments...... iv

Table of Contents ...... v

List of Figures ...... ix

List of Abbreviations ...... xi

Chapter 1 Introduction ...... 1

Introduction ...... 2

1.1 Treatment of cancer ...... 2

1.1.1 Treatment of breast cancer ...... 4

1.2 Drug repurposing ...... 7

1.2.1 Opportunities...... 8

1.2.2 Challenges ...... 11

1.3 Repurposing statins as anti-breast cancer agents ...... 15

1.3.1 Statins as cholesterol-lowering agents ...... 15

1.3.2 Evidence of anti-breast cancer effects ...... 18

1.4 The mevalonate pathway in cancer metabolism ...... 21

1.4.1 End products ...... 22

1.4.2 Deregulation ...... 26

1.5 Mechanism of statin-induced cancer cell apoptosis...... 29

1.6 Objectives and thesis outline ...... 33

1.7 References ...... 35

Chapter 2 Statin-induced cancer cell death can be mechanistically uncoupled from prenylation of RAS family proteins ...... 57

Statin-induced cancer cell death can be mechanistically uncoupled from prenylation of RAS family proteins ...... 58

2.1 Abstract ...... 58

2.2 Introduction ...... 58 v

2.3 Results ...... 60

2.3.1 HRASG12V and KRASG12V, but not other proteins in the RAS superfamily, sensitize MCF10A cells to fluvastatin ...... 60

2.3.2 Inhibition of RAS prenylation is uncoupled from fluvastatin-induced cell death ...... 61

2.3.3 The RAS-ZEB1-EMT signaling axis underlies increased sensitivity to fluvastatin ...... 65

2.3.4 Enrichment of EMT phenotype is associated with sensitivity to statins in a large panel of cancer cell lines ...... 75

2.4 Discussion ...... 78

2.5 Materials and Methods ...... 81

2.5.1 Reagents ...... 81

2.5.2 Cell culture ...... 81

2.5.3 MTT assay ...... 81

2.5.4 Immunoblotting...... 82

2.5.5 Soft agar colony formation ...... 82

2.5.6 Membrane fractionation ...... 82

2.5.7 Cell death assay...... 82

2.5.8 qRT-PCR...... 83

2.5.9 Small hairpin RNA-mediated gene silencing ...... 83

2.5.10 Pharmacogenomic analysis ...... 84

2.6 References ...... 85

Chapter 3 Fluvastatin inhibits EMT-associated protein N-glycosylation and delays breast cancer metastasis ...... 93

Fluvastatin inhibits EMT-associated protein N-glycosylation and delays breast cancer metastasis ...... 94

3.1 Abstract ...... 94

3.2 Introduction ...... 94

3.3 Results ...... 97 vi

3.3.1 EMT sensitizes breast cancer cells to fluvastatin and tunicamycin ...... 97

3.3.2 Fluvastatin inhibits dolichol-mediated protein N-glycosylation...... 103

3.3.3 Fluvastatin inhibits EMT-associated protein N-glycosylation ...... 110

3.3.4 Fluvastatin impairs protein N-glycosylation in vivo ...... 116

3.3.5 Post-surgical adjuvant fluvastatin treatment delays metastatic outgrowth and prolongs survival ...... 119

3.4 Discussion ...... 122

3.5 Materials and Methods ...... 126

3.5.1 Reagents ...... 126

3.5.2 Cell culture ...... 126

3.5.3 MTT assay ...... 127

3.5.4 Immunoblotting...... 127

3.5.5 Immunohistochemistry ...... 127

3.5.6 Cell death assay...... 128

3.5.7 qRT-PCR...... 128

3.5.8 N-glycan extraction ...... 128

3.5.9 Total glycan analysis by LC-MS/MS ...... 129

3.5.10 Animal models ...... 130

3.5.11 Fluvastatin quantification by HPLC-MS/MS ...... 130

3.6 References ...... 132

Chapter 4 Discussion ...... 140

Discussion ...... 141

4.1 Challenges in the treatment of cancer ...... 141

4.2 Mechanism of statin-induced cancer cell death ...... 142

4.3 Biomarkers of statin sensitivity ...... 143

4.4 Point of therapeutic intervention for statin use in breast cancer ...... 144

vii

4.5 Perspectives...... 146

4.5.1 Limitations of this work and future directions ...... 146

4.5.2 Anticipated impact ...... 150

4.6 References ...... 152

Appendix Positive feedback regulation of the mevalonate pathway ...... 159

Appendix – Positive feedback regulation of the mevalonate pathway ...... 160

5.1 Introduction ...... 160

5.2 Results and discussion ...... 163

5.3 Methods...... 169

5.3.1 Cell culture ...... 169

5.3.2 Cell death assay...... 169

5.3.3 qRT-PCR...... 169

5.4 References ...... 170

viii

List of Figures

Figure 1-1. Pharmacology of statins...... 17

Figure 1-2. The mevalonate pathway...... 23

Figure 1-3. Possible mechanisms of deregulation of the MVA pathway in cancer cells...... 27

Figure 1-4. Prenylation substrate is determined by the amino acid sequence in the CaaX domain...... 32

Figure 2-1. HRASG12V and KRASG12V, but not other prenylated proteins, sensitize MCF10A cells to fluvastatin...... 62

Figure 2-2. Inhibition of RAS prenylation is uncoupled from fluvastatin-induced cell death. .... 63

Figure 2-3. Overexpression of HRASG12V or myr-HRASG12V does not alter HMGCR or HMGCS1 expression...... 66

Figure 2-4. Fluvastatin cytotoxicity is not mediated by the BRAF, RALA, PI3K, or MYC downstream mediators of RAS signaling...... 67

Figure 2-5. Transforming oncogenes do not all sensitize MCF10As to fluvastatin...... 68

Figure 2-6. RAS induces EMT through ZEB1, and induction of EMT is sufficient for sensitizing cells to fluvastatin...... 70

Figure 2-7. TGF-β induced EMT and fluvastatin sensitivity are both reversible...... 72

Figure 2-8. shRNA knockdown of ZEB1 rescues the increased sensitivity in HRASG12V and myr- HRASG12V cells...... 73

Figure 2-9. Unimodal and bimodal gene expression in the CCLE database...... 76

Figure 2-10. Statin sensitivity is associated with cancer cell EMT...... 77

Figure 3-1. Induction of EMT by SNAIL overexpression increases cell sensitivity to inhibition of dolichol-dependent protein N-glycosylation by fluvastatin and tunicamycin...... 99

Figure 3-2. Induction of EMT increases cell sensitivity to fluvastatin and tunicamycin...... 102

Figure 3-3. Fluvastatin effect is independent from induction of ER stress ...... 105

Figure 3-4. Inhibition of dolichol synthesis underlies fluvastatin sensitivity in mesenchymal breast cancer cell lines...... 107

Figure 3-5. Exogenous addition of dolichol (dolichyl[C95]-PP) did not rescue viability of cells with fluvastatin treatment in DMEM supplemented with 10% FBS...... 109

Figure 3-6. Fluvastatin treatment decreases complex branched N-glycans associated with EMT...... 113

Figure 3-7. Fluvastatin treatment decreases complex branched N-glycans in vivo...... 117

Figure 3-8. Post-surgical adjuvant fluvastatin treatment delays metastasis and prolongs survival...... 120

Figure 3-9. Proposed mechanism of statin-mediated attenuation of breast cancer metastasis by targeting EMT-associated protein N-glycosylation...... 125

Figure 4-1. Pre-clinical models of the points of therapeutic intervention in breast cancer...... 145

Figure 5-1. Two-component mechanisms of feedback regulation...... 161

Figure 5-2. MVA and GGPP reverses fluvastatin-induced cell death but does not reverse upregulation of MVA pathway genes...... 164

Figure 5-3. HMGCS1 and HMGCR upregulation is sustained in the absence of negative feedback...... 166

Figure 5-4. HMGCS1 and HMGCR upregulation is dependent on SREBP2 in the absence of negative feedback...... 167

Figure 5-5. The MVA pathway may be regulated by a positive feedback loop...... 168

List of Abbreviations

μg microgram

μL microliter

μM micromolar

μm micrometer or micron

2-TTFA 2-thenoyltrifluoroacetone

5-FU 5-fluorouracil

ABI Applied Biosystems

ABL Abelson murine leukemia viral oncogene homolog

ACAT2 acetyl-CoA acetyltransferase

Acetyl-CoA acetyl coenzyme A

ACN acetonitrile

ADP adenosine diphosphate

AFP alpha-fetoprotein

AKT/PKB Protein kinase B

ALL acute lymphoblastic leukemia

AML acute myeloid leukemia

AMPK 5' adenosine monophosphate-activated protein kinase

ANOVA Analysis of variance

APL acute promyelocytic leukemia

ASCO American Society of Clinical Oncology xi

ASM aggressive systemic mastocytosis

ATP adenosine triphosphate

ATRA all-trans-retinoic acid

AUC area under the curve

B/R benefit/risk assessments

BCR breakpoint cluster region

BD Becton Dickinson Biosciences

BiP Binding immunoglobulin protein

BRAF v-Raf murine sarcoma viral oncogene homolog B

BRCA1/2 Breast Cancer susceptibility gene 1/2

BSA body surface area

BSA bovine serum albumin

CA-125 cancer antigen 125

CaaX cysteine-aliphatic-aliphatic-any amino acid

CAF cyclophosphamide, doxorubicin (Adriamycin), 5-fluorouracil

CAR chimeric antigen receptor

CCLE Cancer Cell Line Encyclopedia

CCND1 cyclin D1

CD4 cluster of differentiation 4

CDH1 E-cadherin

xii

CDH2 N-cadherin cDNA complementary DNA

CEA carcinoembryonic antigen

CEL chronic eosinophilic leukemia

CI concordance index

CMAP Connectivity Map

CML chronic myelogenous leukemia

CoQ coenzyme Q

COX-2 cyclooxygenase-2

CRBN Cereblon

CRC colorectal cancer

CST Cell Signaling Technologies

C-terminal carboxyl-terminal

CTLA-4 cytotoxic T lymphocyte-associated protein 4

CTRPv2 Cancer Therapeutics Response Portal version 2

CVD cardiovascular disease

CYP2C9 Cytochrome P450 family 2 subfamily C member 9

CYP3A4 Cytochrome P450 family 3 subfamily A member 4 d day

DBCO dibenzocyclooctyne group

xiii

DC dendritic cell

DFS disease-free survival

DFSP dermatofibrosarcoma protuberans

DNA deoxyribonucleic acid

DHDDS dehydrodolichyl diphosphate synthase dNTP Deoxynucleotide triphosphate

DOLK dolichol kinase

DOLPP1 dolichyl diphosphate phosphatase 1

DP dipyridamole

DTT Dithiothreitol

EDTA Ethylenediaminetetraacetic acid

EGF epidermal growth factor

EGFR epidermal growth factor receptor

EMT epithelial to mesenchymal transition

ER endoplasmic reticulum

ERdj4 Endoplasmic reticulum-localized DnaJ 4

ERK extracellular signal-regulated kinase

ERα estrogen receptor α

ESMO European Society for Medical Oncology

ESR1 estrogen receptor α

xiv

ETC electron transport chain

EtOH ethanol

FAS fatty acid synthesis

FDA Food and Drug Administration

FDPS farnesyl diphosphate synthase

FDR false discovery rate

FFPE formalin-fixed and paraffin-embedded fluva fluvastatin

FN1 fibronectin

FPKM fragments per kilobase of transcript per million mapped reads

FPP farnesyl diphosphate

FTI farnesyltransferase inhibitor

Fuc, F fucose fw forward

GAPDH Glyceraldehyde 3-phosphate dehydrogenase

GGPP geranylgeranyl diphosphate

GGPS1 geranylgeranyl diphosphate synthase 1

GGTI geranylgeranyl transferase inhibitor

GIST gastrointestinal stromal tumours

Glc glucose

GlcNAc, N N-acetylglucosamine

GLUT1 glucose transporter 1

GnT-III β1,4-N-acetylglucosaminyltransferase III

GnT-V β1,6-N-acetylglucosaminyltransferase V

GP130 glycoprotein 130

GPP geranyl diphosphate

H&E Hematoxylin and eosin

HCC hepatocellular carcinoma

HDL high density lipoprotein hEGFR human epidermal growth factor receptor

HEPES 2-[4-(2-hydroxyethyl)piperazin-1-yl]ethanesulfonic acid

HER2 human epidermal growth factor receptor 2

HES hypereosinophilic syndrome

Hex, H hexose

HIF-1a hypoxia-inducible factor 1-alpha

HIV human immunodeficiency virus

HMGCL 3-hydroxy-3-methylglutaryl-coenzyme A lyase

HMG-CoA 3-hydroxy-3-methylglutaryl-coenzyme A

HMGCR 3-hydroxy-3-methylglutaryl-coenzyme A reductase

HMGCS1 3-hydroxy-3-methylglutaryl-coenzyme A synthase 1

xvi

HPLC High performance liquid chromatography

HRAS Harvey rat sarcoma viral oncogene homolog

HTC high-throughput screening

IC50 half maximal inhibitory concentration

IFN interferon

IHC immunohistochemistry

INSIG insulin-induced gene

IPP isopentenyl diphosphate

Km Michaelis-Menton constant

KPC Pdx1-cre;LSL-KrasG12D;P53R172H/+

KRAS V-Ki-ras2 Kirsten rat sarcoma viral oncogene homolog

Lac lactose

LC liquid chromatography

LCA Lens culinaris agglutinin

LDL low density lipoprotein

LDLR low density lipoprotein receptor

Man, M mannose

MEF mouse embryonic fibroblast

MGAT3 β1,4-N-acetylglucosaminyltransferase III

MGAT5 β1,6-N-acetylglucosaminyltransferase V

xvii

MM multiple myeloma mM millimolar

MMTV mouse mammary tumour virus

MPD myeloproliferative diseases mRNA messenger ribonucleic acid

MS mass spectrometry

MTD maxim tolerated dose mTORC1 mammalian target of rapamycin complex 1

MTT 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium

MVA mevalonate

MVAC methotrexate, vincristine, doxorubicin, cisplatin

MVD mevalonate diphosphate decarboxylase

MVK mevalonate kinase

MYC avian myelocytomatosis viral oncogene homolog myr myristoylated

NAT neo-adjuvant therapy

NCCN National Comprehensive Cancer Network

NEB New England Biolabs ns not significant

NSCLC non-small cell lung cancer

xviii

NUS1 dehydrodolichyl diphosphate synthase regulatory subunit

OBD optimal biological dose

OS overall survival

PBS phosphate-buffered saline

PD-1 programmed cell-death protein 1

PDAC pancreatic ductal adenocarcinoma

PDE5 phosphodiesterase type 5

PDGF-R platelet-derived growth factor receptor

PDO patient-derived organoid

PDX patient-derived xenograft

PEP phosphoenolpyruvate

PGAM1 phosphoglycerate mutase

PGC graphitized carbon

P-gp P-glycoprotein

Ph+ Philadelphia chromosome positive

PHA-E Phytohaemagglutinin-erythroagglutinin

PHA-L Phytohaemagglutinin-leucoagglutinin

PI3K Phosphoinositide 3-kinase

PI3K-p110a phosphatidylinositol-4,5-bisphosphate 3-kinase, catalytic subunit alpha

PKM1/2 pyruvate kinase, isoform M1/M2

xix

PMVK phosphomevalonate kinase

POLR2A RNA polymerase II, subunit A

PP diphosphate

PR progesterone receptor

PRO patient-reported outcomes

PSA prostate-specific antigen

PTGS2 prostaglandin-endoperoxide synthase 2

PyMT polyoma middle T qRT-PCR real-time quantitative reverse transcription polymerase chain reaction

RAC1 Ras-related C3 botulinum toxin substrate 1

RALA Ras-related protein Ral-A

RAP1A Ras-related protein Rap-1A

RECIST Response Evaluation Criteria in Solid Tumors

RCC renal cell carcinoma

RDI relative dose-intensity

RHOA/B Ras homolog gene family member A/B

RNA ribonucleic acid

RNA-seq ribonucleic acid sequencing rv reverse

SCAP SREBP cleavage-activating protein

SCB Santa Cruz Biotechnology

SCID Severe combined immunodeficiency

SD standard deviation

SDS Sodium dodecyl sulfate shRNA small hairpin ribonucleic acid

SILAC Stable isotope labeling with amino acids in cell culture

SLC3A2 solute carrier family 3 member 2

SLCS small cell lung cancer

SLUG snail family transcriptional repressor 2

SNAIL snail family transcriptional repressor 1

SRD5A3 steroid 5-alpha-reductase 3

SREBP1/2 sterol regulatory element-binding protein 1/2

TAZ transcriptional co‑activator with PDZ-binding motif

TB tumour burden

TCA tricarboxylic acid

TCGA The Cancer Genome Atlas

TFA trifluoroacetic acid

TGF-β transforming growth factor beta

Th1/2 T helper 1/2 cell thap thapsigargin

xxi

TKI tyrosine kinase inhibitor

TNBC triple-negative breast cancer

Treg T regulatory cell

Tris 2-Amino-2-(hydroxymethyl)propane-1,3-diol tRNA transfer ribonucleic acid

TTP time to progression tuni tunicamycin

TWIST Twist Family BHLH Transcription Factor 1

UGT1A3 UDP glucuronosyltransferase family 1 member A3

US United States

VIM vimentin

VLDL very low density lipoprotein

YAP yes-associated protein 1 yr year

ZEB1 Zinc finger E-box-binding homeobox 1

xxii

Chapter 1 Introduction

1 2

Introduction 1.1 Treatment of cancer

Surgery and radiotherapy. Cancer deaths account for 14-15% of human deaths each year worldwide (1), approximately 20% in the US (2), and up to 30% in Canada (3). For most solid tumours presenting without metastases, the standard of care and often the cure, involves complete surgical resection of the tumour, surrounding tissue, and/or the axillary lymph node(s)

(4). Surgery is often accompanied by radiotherapy, where ionizing radiation is administered to the target tissue to destroy cancer cells by inducing DNA damage (5). In rare cases, radiotherapy alone is curative (6). In general, however, additional treatment is needed, in the form of cytotoxic chemotherapies, targeted chemotherapies, and/or immunotherapies.

Cytotoxic chemotherapy. Cancer has traditionally been treated with general cytotoxic chemotherapy targeted to rapidly dividing cells, with little tumour specificity (7). Many of these drugs still constitute the standard of care for many cancers, often in combination, such as CAF for breast cancer (cyclophosphamide, Adriamycin (doxorubicin), 5-FU (5-fluorouracil)) and

MVAC for bladder cancer (methotrexate, vincristine, doxorubicin, cisplatin) (7). Their typical mechanisms of action are by inducing DNA damage (alkylating agents), by inhibiting DNA synthesis (antimetabolites and topoisomerase inhibitors), or by inhibiting cell division (anti- microtubule agents, cytotoxic antibiotics) (7). However, they are associated with wide-ranging toxicities and reduced quality of life, leading to questions regarding their clinical benefit in many cancers (8).

Targeted chemotherapy. In the past two decades, growing understanding of the hallmarks of cancer (9) and the molecular alterations leading to their selection (10) has given rise to rationally designed inhibitors to target these drivers (11). These targeted therapeutics also represent the

3 growing effort to combat tumour heterogeneity, by stratifying patients based on predicted response (11). One strategy for targeted therapy is the use of monoclonal antibodies, such as the anti-CD20 antibody rituximab, approved by the US Food and Drug Administration (FDA) in

1997 for B-cell malignancies; and the anti-HER2 (human epidermal growth factor receptor 2) antibody trastuzumab, approved in 1998 for HER2+ breast cancer. Another class of targeted therapeutics is small molecule drugs, including a large family of tyrosine kinase inhibitors

(TKI’s), such as imatinib, approved by the FDA in 2001 for chronic myelogenous leukemia

(CML); and gefitinib, approved in 2003 for non-small cell lung cancer (NSCLC). Recent advances in targeted chemotherapies include the rapid development of drugs to target epigenetic regulators (eg. JQ-1 (12, 13)) and hypoxia (eg. TH-302 (14, 15)).

Targeted therapeutics have dramatically improved the survival rate of some cancers. For example, the 5-year survival rate for CML was 31% in the early 1990s, and was improved to

63% by 2011, due in part to the use of imatinib and related agents (16). Unfortunately, despite the best of intentions, it is now clear that targeted agents are generally more toxic than the older chemotherapeutics (17). Inhibition of EGFR (gefitinib and related agents) causes Grade 3/4 – severe to life-threatening – adverse events in 40% of NSCLC patients (18), and cardiotoxicity is a huge problem in patients treated with TKI’s (19). In addition to questions about their safety, the clinical utility of targeted agents is further challenged by their small margin of benefit and high cost. As many tumours quickly gain resistance against targeted therapeutics, many drugs are only able to extend survival by a few days to weeks, with significant declines in quality of life (QOL)

(20). For example, the addition of bevacizumab to carboplatin and paclitaxel, now the standard of care for metastatic NSCLC, extends progression-free survival (PFS) by 12-18 days, depending on the bevacizumab dose. Bevacizumab costs US$90,000 for one cycle. In pancreatic cancer patients, addition of erlotinib to gemcitabine extends median overall survival (OS) by 10 days,

4 and costs US$16,000 (20). Clearly, development of safe, effective, and inexpensive drugs for the treatment of cancer remains a gap that urgently needs to be filled.

Immunotherapy. Immunotherapies manipulate the interaction between the body’s immune system and the tumour. Recently there has been unprecedented interest in this field, owing in part to remarkable successes seen with the use of an anti-CTLA-4 (cytotoxic T lymphocyte- associated protein 4) antibody (ipilimumab) against melanoma (21), and anti-PD-1 (programmed cell-death protein 1) antibodies (nivolumab, pembrolizumab) for melanoma (22, 23) and NSCLC

(24). Anti-PD-1 and anti-CTLA-4 immunotherapies belong to a class of immunotherapy strategies called checkpoint blockade, where the mechanism(s) by which tumours suppress T cell activation – immune-checkpoint signaling – is inhibited (25). T cell function is re-activated as a result, and can eliminate malignant cells through immune surveillance (25). Although it is currently associated with severe immune-related adverse events (26) and staggering cost (27), immunotherapy holds the unique possibility of generating immunological memory that will be able to prevent the rise of therapy-resistant malignant clones, thereby preventing disease recurrence (28-31). Other modalities of immunotherapy, including adoptive cellular therapy, chimeric antigen receptor (CAR) T-cell therapy, and oncolytic viruses, are under intensive clinical investigation at present (25, 32).

1.1.1 Treatment of breast cancer

By stage and grade. Current prognosis and treatment of breast cancer depends on a variety of clinical and pathological factors, including patient age, tumour stage and grade, and receptor status (33). Tumour stage considers the size, axillary lymph node involvement, and metastatic status of breast cancer, combined to give a score between 0 and 4 (Stage 0 to IV) (16). Stage 0 breast cancers are non-invasive ductal carcinoma in situ (DCIS) or lobular carcinoma in situ

(LCIS), and are curable with surgery and/or radiation. Stage I-II are “early-stage”, and Stage III is “locally advanced”, breast cancers with increasing size and lymph node involvement. Stage I-

III breast cancers are generally treatable with a combination of surgery, radiotherapy, and chemotherapy. Early-stage breast cancers have a 5-year survival rate of 85-99% (16). Stage IV breast cancer are metastatic breast cancer, with secondary sites often in the lung, liver, brain and bones. Metastatic breast cancer has a 5-year survival rate of 25% (16). While less than 10% of patients are diagnosed with metastatic breast cancer, 20-30% of early-stage breast cancer patients go on to develop metastatic breast cancer, despite aggressive treatment (34).

Breast cancer grade (Grade I-III, or low-intermediate-high) is determined histologically, by scoring tumour cell differentiation (33). Grading determines how fast the cancer cells are likely to grow and spread, and helps determine the aggressiveness of therapy (35).

By subtype. In clinical practice, three biomarkers are routinely used to assess breast cancers:

ERα (estrogen receptor α), PR (progesterone receptor), and HER2 (16, 36). This gives rise to three major subtypes of breast cancers: ERα/PR positive (+) tumours have a more favourable prognosis, and respond well to hormone therapy (eg. tamoxifen, aromatase inhibitors); HER2+ tumours are more aggressive and more likely to metastasize and reoccur, but respond well to anti-HER2 therapies (eg. trastuzumab, lapatinib); and triple-negative breast cancers (TNBC) follow the most aggressive clinical course with early recurrence, and have limited therapeutic options other than surgery, radiotherapy, and cytotoxic chemotherapy (eg. CAF) (16, 36).

Virtually all TNBC relapse within 5 years of diagnosis, after which the median survival is only 6 months (37).

Breast cancer can also be classified by molecular profiling, historically through gene expression profiling and more recently by exome sequencing, DNA copy number, DNA methylation, and

6 many others (33, 38). These methods can subtype breast cancers into 20 or more subgroups, largely falling into luminal A/B, HER2+, claudin-low, normal-like, and basal-like categories

(33). A number of prognostic tests based on molecular profiling have been developed and incorporated in patient management, such as OncotypeDX (Genomic Health), PAM50

(Nanostring Technologies), and MammaPrint (Agendia) (38). Molecular profiling has also been successful in identifying novel treatment options for breast cancer, such as the use of poly(ADP- ribose) polymerase (PARP) inhibitors (eg. iniparib, olaparib) to treat breast cancers with mutations in BRCA1/2 (39).

Points of therapeutic intervention. Surgery is the most common breast cancer treatment given with curative intent, which involves breast conservation surgery or mastectomy, and removal of axillary lymph nodes if needed (40). Even in patients with evidence of metastases, surgery of either the primary tumour or at the metastatic site is still performed for palliative purposes (40).

Neo-adjuvant therapy (NAT), consisting of radiotherapy and/or chemotherapy, is given in the space between diagnosis and surgery, usually to locally advanced breast cancer patients, with the goal of reducing tumour size to facilitate surgery (41). NAT is also used in the attempt to (i) convert a previously inoperable tumour into an operable one; (ii) downstage a tumour to allow breast conservation surgery; and (iii) provide prognostic information and guide selection of appropriate follow-up therapies (41). As evidence accumulates that response to NAT can be correlated to long-term outcome (41, 42), there is increased interest to conduct clinical trials in the neoadjuvant setting (often called “window of opportunity” trials), as they require smaller sample sizes, are less expensive, and provide rapid assessment of efficacy (43). Importantly, most pre-clinical drug assessment techniques using 2D cell culture, as well as tumour xenografts, represent pre-clinical models of NAT.

Adjuvant therapies of either systemic or local-regional radiotherapy, chemotherapy (including targeted therapies against ERα or HER2), or both, is given after surgery. The duration of adjuvant therapies varies from a few weeks (chemo and/or radiotherapy) (44), to up to 1 year

(anti-HER2 therapy), to 5-10 years (anti-ERα therapy) (36). The goal of adjuvant therapies is to decrease the risk of breast cancer recurrence and prevent metastasis, but unfortunately 20-30% of tumours still recur (34).

In the 20-30% of early-stage patients that experience recurrence, treatment options are limited.

For local-regional recurrence, surgical removal of the recurrent tumour followed by radiotherapy, if radiotherapy has not previously been given, is recommended (34, 35). For metastatic recurrence, patients may be encouraged to enroll in clinical trials for novel therapeutics (34). As such, treatment of metastasis is the therapeutic setting that most oncology drugs are clinically tested in. As metastatic recurrences are highly aggressive and likely resistant to all current therapies, this may be contributing to the high failure rate of oncology drugs during clinical testing, discussed below.

1.2 Drug repurposing

As of 2011, the success rate of investigational drugs from clinical development to approval by the FDA is the lowest (6.7%) for oncology drugs, compared to drugs for all other indications

(12.1%) (45). This discrepancy alone reduces the probability of FDA approval for all investigational drugs from nearly one in eight to over one in ten (45). This high failure rate of investigational cancer drugs is alarming, considering the high cost of clinical trials, the long timeline for drug development (46), as well as the unmet needs from cancer patients, clinicians, and policy makers. In light of these challenges, repurposing drugs that are already approved for non-oncology use is an alternative strategy to bring more treatment options to cancer patients.

Drug repurposing has the potential to reduce the time, cost, and risk investments in phase I/II clinical trials, since the safety, dosage range, formulation, and side effects are already documented and approved. Additionally, many older drugs are now off-patent and manufactured as inexpensive generics. Drug repurposing is therefore attractive and timely, as the national healthcare budgets world-wide are unlikely to continue to support the exponential increase in the cost of new oncology drugs (20, 47).

1.2.1 Opportunities

Clinical and epidemiological observations. Association between use of a drug and an unexpected effect or outcome is primarily reported as clinical or epidemiological observations. A well-known example of clinical observations leading to drug repurposing is the case of sildenafil, originally developed to treat hypertension and angina (48), now widely marketed as Viagra due to patient reports of erection as an unexpected effect (49). Another example is thalidomide, originally developed and marketed as a sedative (50), then failed as a remedy for nausea in pregnant women due to severe fetal toxicity leading to birth defects (51, 52). Observations by clinicians that thalidomide relieved leprosy symptoms prompted a repurposing investigation

(53), leading to its approval for the treatment of leprosy complications in 1998 (50).

The stories of sildenafil and thalidomide do not end here. Sildenafil targets PDE5

(phosphodiesterase type 5) (48), and the observation that PDE5 is upregulated in pulmonary hypertensive lungs (54) prompted its current development for this indication. Thalidomide was found to have anti-angiogenic and anti-cancer activities (55, 56), leading to its approval for the treatment of multiple myeloma (MM) in 2006 (50, 57). These successes in drug repurposing were made possible by scientific progress in the understanding of disease biology and

9 vulnerabilities, highlighting the importance of basic and translational research to confirm and follow-up on opportunities of drug repurposing identified by clinical observations.

Epidemiological studies and post hoc analyses are additional avenues commonly used to examine the association between use of an already-approved drug and a specific outcome, usually from non-primary indications. Although rife with statistical bias and interpretation errors

(58, 59), epidemiological studies are nevertheless useful as a first-pass tool to identify drugs for repurposing (47). The power of epidemiological studies lies in the size of the population examined, which is particularly important for cases where effect sizes are small, or when the rate and magnitude of response are highly variable. The disadvantage is that, due to the need for a large sample size, conclusions from epidemiological studies are only meaningful when examining widely-prescribed drugs and prevalent outcomes (eg. risk of common cancers, as opposed to rare diseases). Some examples of epidemiologically reported association between drug use and response in oncology include metformin (60), currently used to treat diabetes; aspirin (61, 62), used as a painkiller and for prevention of stroke and coronary events; and statins

(63, 64), used to treat hypercholesterolemia and to prevent cardiovascular diseases (CVD). As indicated above, basic and translational research is needed prior to the successful repurposing of these drugs as oncology drugs. Pre-clinical studies of the mechanism and efficacy of fluvastatin, a member of the statin family of cholesterol-lowering drugs, for its application as an anti-breast cancer therapeutic, is the focus of this thesis.

Drug screening and data mining. Identification of anti-cancer activities in existing drugs can occur through high-throughput screening (HTS), which has been used for new drug discovery for decades (65), and are now being increasingly applied for drug repurposing (66-70). Libraries of approved small molecules ranging from ~100 drugs (69, 70) to >2,500 clinically approved and

10 bioactive compounds (66) have been assembled for this purpose. In light of growing evidence that multiple machineries are simultaneously derailed in most cancer cells (9), a particularly attractive application of the HTS method to identify anti-cancer drugs is the possibility to test for synergistic drug combinations. For economic reasons, development of novel oncology drugs focuses primarily on showing efficacy and/or superiority of the novel drugs when administered alone. In contrast, synergistic pairs of FDA-approved drugs have been successfully identified through chemical screening and validated in pre-clinical models, including chlorpromazine + pentamidine against lung cancer (71), and atorvastatin + dipyridamole against multiple myeloma

(MM) and acute myeloid leukemia (AML) (72).

Another emerging method to identify oncological targets and drug candidates is through in silico data mining (73), made possible by the public establishment and maintenance of large databases of genomics (eg. EntrezGene), cancer genomics (eg. The Cancer Genome Atlas; TCGA), proteomics (eg. UniProt), pharmaceuticals (eg. PubChem), cancer cell lines (eg. Cancer Cell

Line Encyclopedia; CCLE), and drug perturbations (eg. Connectivity Map; CMAP) (74). Major strategies for the selection of drugs for repurposing include searching for similarities based on gene expression profiles (75, 76), side-effect profiles (77, 78), and correlation between drug targets and patient survival (79). Although these computational approaches to assess the therapeutic potential of a known drug for repurposing is still highly experimental (80, 81), successes have been made such as the identification and pre-clinical validation of imipramine and clomipramine against small cell lung cancer (SCLC) (75), and pentamidine against renal cell carcinoma (RCC) (76).

1.2.2 Challenges

Targeted vs. non-targeted repurposing. The successful repurposing of sildenafil is partially attributable to its well-characterized mechanism of action as a PDE5 inhibitor, and the developing scientific understanding between PDE5 activity and penile erection. Similarly, all- trans-retinoic acid (ATRA) was originally marketed as a topical acne treatment (82). Later it was found that acute promyelocytic leukemia (APL) is driven by a fusion protein RARA-PML

(RARA, retinoic acid receptor α; PML, promyelocytic leukemia gene) (83), rendering the leukemic promyelocytes sensitive to ATRA-induced differentiation and subsequent apoptosis.

ATRA is currently the standard of care for APL patients, with concurrent use of arsenic trioxide

(84) or anthracyclines (85). Such are cases of “targeted” repurposing, and is well-represented in oncology as repurposing between cancer types. The best example is imatinib, a specific inhibitor of tyrosine kinases ABL (Abelson murine leukemia viral oncogene homolog), c-Kit, and PDGF-

R (platelet-derived growth factor receptor). Imatinib was first approved in 2001 for Philadelphia chromosome positive (Ph+) CML, harbouring the BCR-ABL fusion protein (BCR; breakpoint cluster region) (86). Since it is known to also inhibit c-Kit, imatinib was immediately tested for the treatment of gastrointestinal stromal tumours (GIST), which express c-Kit in up to 95% of cases examined, and was approved by 2002 (87). Today, imatinib is also FDA-approved to treat

Ph+ ALL (acute lymphoblastic leukemia); MPD (myeloproliferative diseases); ASM (aggressive systemic mastocytosis); HES (hypereosinophilic syndrome); CEL (chronic eosinophilic leukemia); and DFSP (dermatofibrosarcoma protuberans) (86), all via its on-target effects as a selective tyrosine kinase inhibitor.

In contrast to the above examples, some drugs have been observed to exert potential anti-cancer effects, but the mechanism of action is either unclear, or independent from known target(s). For example, the anti-diabetic drug metformin has been reported to reduce cancer mortality by up to

34% (88, 89), with validation in pre-clinical models of breast cancer (90, 91) and response in clinical trials (92-94). However, the molecular target of metformin leading to cancer cell death remains elusive, and it is unclear whether metformin directly induces cancer cell death, or indirectly through its systemic effects on blood sugar levels and insulin sensitivity (95).

Similarly, ample evidence exists to support the use of aspirin in post-surgical colorectal cancer

(CRC), including positive results from randomized trials (96-98), without fully elucidating the molecular target. Recently, however, several studies have implicated modulation of cyclooxygenase-2 (COX-2; officially known as prostaglandin-endoperoxide synthase 2, PTGS2) as the underlying mechanism of aspirin-induced CRC cell death (99-101). If validated, this would move aspirin into the “targeted” category, and aspirin treatment may be stratified based on

COX-2 activity (47).

Patient selection and the search for biomarkers. Whereas the attrition rate of oncological drugs as a whole is estimated to be 82-95% (102, 103), for a subset of these drugs – kinase inhibitors – the attrition rate is reduced to an astounding 53% (102). This is largely attributable to a careful selection of patients with an appropriate molecular drug target (104). In other words, knowledge of the molecular target and mechanism of action for a given drug is critical for its successful development. In drug repurposing, this is exemplified by the quick acceptance of ATRA by the medical community to treat APL, once it became clear that APL is driven by the fusion protein

RARA-PML, which can be targeted by ATRA (83).

In the case of thalidomide, its clinical development for MM between 1990s to 2000s (50, 57) is paralleled by intense pre-clinical research to identify its molecular target, cumulating in the identification of the Cereblon (CRBN) E3 ubiquitin ligase complex as its target for both

13 teratogenicity and immunomodulation in 2010-2012 (105, 106). CRBN was immediately developed as a predictive biomarker for thalidomide and thalidomide-analogues (107).

A drug is less likely to be accepted for repurposing without a definitive molecular target to aid patient selection (although it is not impossible, as seen with thalidomide). Aspirin has been reported to reduce CRC recurrence by 6.7-10% in multiple double-blinded, randomized, placebo- controlled phase II/III clinical trials (96-98), and yet is still not recommended by most clinical guidelines, including the most influential US National Comprehensive Cancer Network (NCCN) and European Society for Medical Oncology (ESMO). For the repurposing of statins, currently prescribed to patients with high cholesterol, the question of patient selection becomes even more challenging, with (i) conflicting data on the association between cholesterol levels and cancer risk (eg. positive association with breast cancer (108); negative association with breast cancer

(109)), (ii) conflicting data on cholesterol levels after treatment with standard of care (eg. positive association with aromatase inhibitors (110, 111); negative association with tamoxifen

(112)), and (iii) pre-clinical indication that the anti-cancer effects of statins are independent from its cholesterol-lowering effects altogether (113, 114). As a result, after two decades of pre- clinical research, statins have yet to reach the clinic as an oncology drug.

Understanding the target and mechanism of a drug for repurposing is also important for the search for molecular biomarkers of response. Currently, FDA-approval of oncology drugs relies on direct measurements of clinical benefit, such as survival (OS, overall survival; DFS, disease- free survival), quality of life (QOL; PRO, patient-reported outcomes), and safety (B/R, benefit/risk assessments) (115, 116). Endpoints often used in oncology trials, defined by

Response Evaluation Criteria in Solid Tumors (RECIST), are also objective measures of tumour response (progressive disease, stable disease, partial response, and complete response) (115,

116). In contrast, molecular biomarkers of response are routinely used in the evaluation of drugs for other diseases, such as cholesterol and triglyceride levels for atherosclerotic disease, and CD4

(cluster of differentiation 4) T lymphocyte count for HIV (human immunodeficiency virus) infection (116). Drug trials and drug repurposing trials could be faster, easier, and cheaper to conduct if a molecular biomarker of response is incorporated as a surrogate endpoint (116).

Partially contributing to this challenge is the criticism that biomarkers of response in oncology have not been reliable to date. For example, the utility of PSA (prostate-specific antigen) to monitor response in prostate cancer (117), CEA (carcinoembryonic antigen) for colon cancer

(118), and CA-125 (cancer antigen 125) for ovarian cancer (119), are all being debated. It is therefore recommended that clinical trials simultaneously assess both the direct and the surrogate endpoints (116). Thus the efficacy, the mechanism of action, and biomarker(s) of response all need to be pre-clinically determined, for the successful development or repurposing of a drug.

Dosing. Particularly relevant to the issue of drug repurposing is that it is often unknown whether the dose recommended for other indications is sufficient to achieve anti-cancer activity. For example, pre-clinical experiments with statins often require higher doses than the recommended cholesterol-lowering dose (72, 120), but lower than the maximum tolerated dose (MTD) in cancer patients (121, 122). The anti-cancer effects of metformin are often examined at millimolar

(mM) concentrations in cell culture, which is >100 fold higher than the achievable serum concentration in patients taking diabetic doses of metformin (123). Additionally, high variability in intra-individual pharmacokinetics are commonly observed for orally-administered drugs

(124), highlighting an additional aspect of personalized medicine that requires optimization.

Identification of the optimal dose is further challenged by the growing realization that (i) the optimal biological dose (OBD) of a non-cytotoxic drug is not always associated with the MTD

(124), (ii) the schedule of drug delivery may influence its cellular target (125-127), and (iii) novel dose-escalation strategies with optimized scheduling may be needed for effective execution of drug combinations (128, 129). Together, these factors mean that Phase I dose- finding/schedule-finding trials, and the associated safety assessments, are required in the (new) target population for repurposed drugs.

Financial difficulties. Without financial incentives, the task of repurposing a non-oncology drug depends on combined efforts from the scientific community and society. Mobilization of not-for- profit foundations, health insurance companies, and governments are needed to fund pre-clinical studies and clinical trials (91, 130). Additional responsibilities may need to be taken on by project initiators (often academics), from dedicated project management to realistic budgeting, efficient and transparent trial conduct, and targeted communication to interest groups and stakeholders. Influential clinical guideline authorities and regulating bodies should be consulted from the beginning.

1.3 Repurposing statins as anti-breast cancer agents 1.3.1 Statins as cholesterol-lowering agents

Pharmacology of statins. High levels of cholesterol in the blood, specifically cholesterol that is carried in low density lipoprotein (LDL) particles, is a major risk factor for cardiovascular diseases (CVD) (131). The statins family of drugs are profoundly successful in lowering cholesterol and reducing CVD (132-134), reducing all-cause mortality by 30% and coronary deaths by 42% in the landmark “4S study” (Scandinavian Simvastatin Survival Study) (132) while having no effect on non-cardiovascular deaths (133). Since the discovery of the first statin, mevastatin, in 1976 (135) and the first FDA-approval of a statin (lovastatin, also known as mevinolin) in 1987 (136), statins have become one of the most successful pharmaceuticals to

16 date. As of 2012, statins are prescribed to approximately 23% of US adults over 45 years of age

(137), and approximately 18% of Canadians at high to intermediate risk of CVD (138).

There are seven approved statins in North America, with differing pharmacological characteristics (139, 140) (Fig 1-1). In general, the lipophilic statins are more likely to be detected in extra-hepatic tissues, while hydrophilic statins are only consistently detected in the liver (141). Although simvastatin is the most frequently used statin for cardiovascular indications, its peak plasma concentration is one of the lowest, being 13-fold lower than fluvastatin (Fig 1-1). While most statins are metabolized by the liver CYP3A4 system commonly used to process many other drugs, fluvastatin, pitavastatin, and rosuvastatin are metabolized by alternative enzymes (139, 140), decreasing the chance of adverse pharmacokinetic interactions between these statins and other co-administered therapeutics. Lovastatin and simvastatin are lactone pro-drugs that require acidic activation, and both have bioactivities independent from inhibition of HMGCR, interacting directly with P-glycoprotein (P-gp) (142) and the proteasome

(143-145) (Fig 1-1).

Mechanism of action. Statins competitively inhibit 3-hydroxy-3-methylglutaryl-coenzyme A reductase (HMGCR), the rate-limiting enzyme of the MVA pathway, responsible for the biosynthesis of cholesterol as well as isoprenoid end products (135, 146) (reviewed below). The liver is the major site of cholesterol production in the human body (147). In liver cells, cellular cholesterol can either be synthesized via the MVA pathway, or be absorbed from the blood plasma through receptor-mediated uptake of LDL-cholesterol (148). Inhibition of HMGCR by statins lower the cholesterol biosynthetic activity in the cell, and as a response cells upregulate the LDL receptor (LDLR), leading to increased uptake of LDL-cholesterol from the blood plasma (147). As a result, plasma LDL levels are decreased, and risk of cardiovascular events is

Peak plasma Primary HMGCR- Trade Rate of Statin Solubility concentration liver independent Name use (%) (ng/ml) metabolism effects Fluvastatin Lescol Lipophilic <20 448 CYP2C9 No Atorvastatin Lipitor Lipophilic 20.2 66 CYP3A4 No Pitavastatin Livalo Lipophilic <20 41 UGT1A3 No Simvastatin Zocor Lipophilic 42.0 34 CYP3A4 Yes Lovastatin Mevacor Lipophilic 7.4 20 CYP3A4 Yes Pravastatin Pravachol Hydrophilic 11.2 55 CYP3A4 No Rosuvastatin Crestor Hydrophilic 8.2 37 CYP2C9 No

Adapted from references (139, 140)

Fluvastatin Atorvastatin Pitavastatin (Lescol) (Lipitor) (Livalo)

Simvastatin Lovastatin Pravastatin Rosuvastatin (Zocor) (Mevacor) (Crestor) (Pravachol)

Figure 1-1. Pharmacology of statins.

Pharmacodynamic/pharmacokinetic characteristics (top) and chemical structures (bottom) of seven members of the statin family of drugs that are available in North America.

18 lowered (131, 148).

Molecularly, this mechanism of action exploits the robust negative feedback regulation of MVA pathway genes, mediated transcriptionally by the SREBP (sterol regulatory element-binding protein) family of transcription factors (148). Of the three SREBP family members (SREBP1a,

1c, and 2), their strength of MVA pathway regulation is ranked as follows: SREBP2 >>

SREBP1a > SREBP1c (149-151). SREBP activity is controlled by the cholesterol sensor SCAP

(SREBP cleavage-activating protein) (152). When cells have adequate supply of cholesterol,

SCAP retains SREBP in the endoplasmic reticulum (ER) in an inactive state, through a three- way interaction with the ER-resident protein INSIG (insulin-induced gene). When cholesterol levels in the ER are low, such as when HMGCR is inhibited by statins, SCAP dissociates from

INSIG and escorts SREBP into the Golgi, where SREBP is activated by sequential cleavage by

Golgi-resident proteases S1P (site-1 protease) and S2P (site-2 protease) (151). Cleavage- activated SREBP then translocates into the nucleus to upregulate the transcription of MVA pathway genes, as well as LDLR, thus replenishing intracellular cholesterol stores (147, 148).

Although the main site of statin action as a cholesterol-lowering agent is in the liver, this pathway and its extraordinary feedback mechanism can be observed in liver cells as well as all normal cells of the body (153, 154).

1.3.2 Evidence of anti-breast cancer effects

Epidemiological studies. As a part of the Women’s Health Initiative, Cauley et al. (155) examined breast cancer incidence in 150,000 post-menopausal women, and reported that use of lipophilic statins is associated with an 18% reduction of breast cancer incidence. Nielson et al.

(64) assessed the mortality of the entire Danish population from 1995 to 2007 (n=300,000), and reported statin use to be associated with a 15% reduction in death from all cancers, and 13%

19 reduction in death from breast cancer. Additional epidemiological studies with smaller population cohorts (n=2,000-3,000) similarly reported reduced cancer incidence or death in statin users, with effect sizes as high as 55% (63, 156-158). Particularly interesting is the observation by Kumar et al. (156) that use of lipophilic statins for >1 year (yr) have a 37% reduction in

ERα/PR-negative tumour incidence, compared to never- or <1 yr-users. Consistently, breast cancer cases in the >1 yr-user cohort are significantly more likely to be low grade and stage

(156).

Not all studies agree on the association of statin use and reduced breast cancer risk and/or death described above (159-163). In a meta-analysis of 42 studies, Kuoppala et al. (164) reported that statin use is associated with reduced risk of stomach cancer, liver cancer, and lymphoma, but not of breast cancer. It should be noted that many of these studies do not report covariates and/or confounding factors, such as physical activity and diet, duration of follow-up times, and duration of statin use.

In addition to the mixed results described above, several epidemiological studies have examined the effect of statin use on breast cancer recurrence, after front-line treatment. Ahern et al. (165) reported with a cohort of 18,000 stage I-III breast cancer patients that any-time statin use is associated with 10% reduction in risk of recurrence. Studies with smaller cohorts (n=700-4,000) also report reductions in recurrence, with effect sizes as high as 60% (166-168). Furthermore, increasing duration of statin use after diagnosis is associated with decreasing risk of recurrence

(168). These epidemiology data suggest that statins may have different clinical utility depending on the breast cancer subtype, and the points of therapeutic intervention.

Pre-clinical studies. Pre-clinical evaluations of statins in breast cancer cell lines or tumour xenografts have demonstrated that some breast cancer cells are highly sensitive to statin-induced

20 apoptosis, while others are resistant (169-171). This indicates that repurposing statins into the breast cancer is dependent on identification of predictive markers of sensitivity, in order to inform clinical trial designs, targeted patient population, and co-development of diagnostics for assessment of therapeutic response. To this end, our group has previously assessed the statin sensitivity of a panel of 25 breast cancer cell lines, the largest to date, and reported an association of statin sensitivity with three factors (169): (i) relatively lower expression of HMGCR; (ii) and

ERα-negative, basal-like tumour subtype; and (iii) a 10-gene candidate signature. Other studies, using much smaller panels of breast cancer cell lines, have also reported statin sensitivity to be associated with ERα negativity (170, 171), activation of NFκB (170), activation of RAS family members (170, 172-174), and specific p53 mutations (175, 176).

Additional pre-clinical support for the potential use of statins in breast cancer include evidence where ectopic introduction of the catalytic domain of HMGCR (177) or exogenous addition of

MVA (178) is sufficient to promote breast tumour xenograft growth. These data, combined with the observation that high mRNA expression of HMGCR is correlated with poor prognosis of breast cancer patients (176, 177), suggest that HMGCR is a metabolic oncogene capable of driving breast cancer cell transformation. Therefore, inhibition of HMGCR by statins may have therapeutic activity.

Clinical trials. The clinical efficacy of various statins, at cholesterol-lowering doses or higher doses, administered with or without combination therapeutics, have been tested in a variety of solid and liquid cancer types (for examples see review (179)). In breast cancer, however, clinical evaluations of statins have thus far been restricted to small-cohort (n=40-50) window-of- opportunity trials (180-182). In the two trials published to date, both reported that some patients responded to statin treatment, while others did not, which mirrors results from pre-clinical

21 models of breast cancer (169). Efforts were therefore concentrated in identifying biomarkers of statin sensitivity.

Garwood et al. (180) treated 40 patients with a diagnosis of DCIS or stage I breast cancer with cholesterol-lowering doses of fluvastatin (20-80 mg/d) for 3-6 weeks, and reported that high grade tumours were more likely to respond. Bjarnadottir et al. (181) treated 50 patients with primary invasive breast cancer with a cholesterol-lowering dose of atorvastatin (80 mg/d) for 2 weeks, and reported that tumours with higher expression of HMGCR prior to treatment, assessed at the protein level by immunohistochemical (IHC) analysis of formalin-fixed and paraffin- embedded (FFPE) tissue samples, were more likely to respond. A companion study (182) performed mRNA analysis using materials harvested from Bjanadottir et al. (181), and reported that tumours with lower mRNA expression of CCND1 (cyclin D1) prior to treatment were more likely to respond, although this result does not reach statistical significance. A later follow-up publication (183) then reported that tumours with higher mRNA expression of HMGCR prior to treatment were less likely to respond to statin treatment, which directly contradicts results from the original study published by Bjarnadottir et al. (181). Thus the search for biomarkers of statin sensitivity continues.

1.4 The mevalonate pathway in cancer metabolism

Cancer cells reprogram their metabolism to ensure survival and sustain abnormal growth (184).

The mevalonate (MVA) pathway is an anabolic pathway fueled by acetyl-CoA to produce both steroid and isoprenoid end products critical for proliferation and cell signaling (151) (Fig 1-2).

Inhibition of MVA production by the statin family of drugs potently inhibits cancer cell growth in many cancer types, both in tissue culture (113, 185-187) and in tumour xenografts (72, 120,

170, 188), indicating that consumption of MVA pathway end products from the blood plasma is

22 insufficient to sustain cancer cell growth, and MVA pathway activity is indispensable. In particular, cell cycle progression at G1/S is dependent on production of isoprenoids by the MVA pathway (153, 189), while G2/M progression requires adequate supplies of cholesterol (190,

191). Recently it has also been shown that, in a p53-null setting, the conversion of MVA to isopentenyl diphosphate (IPP) is essential for maintaining dNTP levels and S phase progression, although the exact mechanism of this is currently unknown (192).

Evidence indicates that increased flux through the MVA pathway, experimentally induced either by ectopic expression of the catalytic domain of HMGCR (177), or by exogenous supplementation of MVA (178), can directly promote tumourigenesis. In primary patient samples, high expression of MVA pathway genes are correlated with poor prognosis in breast cancer (177), corroborating these experimental findings. In a cancer cell, the MVA pathway may be deregulated at multiple levels and through multiple mechanisms (152), although it is currently unknown whether this deregulation can act as an independent oncogenic signal, or simply creates a dependency in a fast-growing cell (reviewed below). In either case, tumours become uniquely vulnerable to the depletion of MVA pathway end products, which can be targeted for therapy.

1.4.1 End products

Cholesterol. The best-known end product of the MVA pathway is cholesterol (Fig 1-2), which is required for all animal life to build and maintain membranes. In addition to its requirement for mitosis, cancer cells often have an increased demand for cholesterol to maintain elevate levels of lipid rafts and caveolae, important for protein membrane trafficking and signal transduction

(193-195). These are sometimes exemplified in cancer patients that present with extremely low levels of serum cholesterol (196), likely due to high LDLR activity in the tumour (197).

Cholesterol also serves as a precursor for the biosynthesis of steroid hormones in the estrogen,

Acetyl-CoA ACAT2 HMGCS1 HMG-CoA HMGCR Statins Mevalonate (MVA) fluvastatin MVK PMVK MVD Isopentenyl-PP FDPS Geranyl-PP FDPS Cholesterol Farnesyl-PP Heme A (FPP) GGPS1

Geranylgeranyl-PP Coenzyme Q 4 (GGPP) DHDDS/NUS1 DOLPP1 SRD5A3 DOLK

Dolichol-PP 19-21

Figure 1-2. The mevalonate pathway.

Enzymes of the mevalonate (MVA) pathway are shown in blue. Chemical structures of fluvastatin and biologically important end products of the MVA pathway are shown. Red box represents the chemical group of statins that inhibit HMGCR. ACAT2, acetyl-CoA acetyltransferase; HMGCS1, 3-hydroxy-3-methylglutaryl-CoA synthase 1; HMGCR, 3-hydroxy- 3-methylglutaryl-CoA reductase; MVK, mevalonate kinase; PMVK, phosphomevalonate kinase; MVD, mevalonate diphosphate decarboxylase; FDPS, farnesyl diphosphate synthase; GGPS1, geranylgeranyl diphosphate synthase 1; DHDDS, dehydrodolichyl diphosphate synthase; NUS1, dehydrodolichyl diphosphate synthase regulatory subunit; DOLPP1, dolichyl diphosphate phosphatase 1; SRD5A3, steroid 5-alpha-reductase 3; DOLK, dolichol kinase; PP, diphosphate.

24 progestogen, and androgen families, which are important for the transformation and therapeutic resistance in some subtypes of breast and prostate cancers (198, 199).

Isopentenyl diphosphate. The MVA pathway is the sole de novo synthetic pathway for isoprenoid end products in all mammalian cells (200) (Fig 1-2). Isopentenyl diphosphate (IPP) is the smallest isoprenoid metabolite, and is used to modify tRNAs on adenosine-37 (A37) located in the anti-codon loop (201). The function of tRNA isopentenylation in human cells, both in normal and in cancerous cells, is an understudied area. One study (202) has reported that overexpression of the enzyme TRIT1, which catalyzes the isopentenylation of tRNAs, has tumour-suppressing activities in lung cancer cell lines. However, the yeast homologue of TRIT1 has been reported to have distinct cytoplasmic (ie. tRNA-isopentenylating) and nuclear functions

(203). It is currently unknown whether the apparent tumour suppressing activities of TRIT1 in lung cancer are due to its tRNA isopentenylation activity, or its nuclear function, if it is indeed located in the nucleus.

Farnesyl diphosphate and geranylgeranyl diphosphate. Sequential polymerization of IPP produces isoprenoid metabolites farnesyl diphosphate (FPP), composed of 3 isopentenyl moieties, and geranylgeranyl diphosphate (GGPP), composed of 4 isopentenyl moieties (151)

(Fig 1-2). FPP and GGPP are essential substrates for protein geranylgeranylation and farnesylation, respectively, together referred to as protein prenylation (204-206). These processes add a hydrophobic farnesyl or geranylgeranyl moiety to proteins containing a C-terminal CaaX domain, thus localizing them to cellular membranes (207, 208). A large body of work has implicated GGPP, and to a lesser extent FPP, as the limiting metabolic end products of the MVA pathway in cancer cells (209-214). However, despite clear evidence of reduced protein

25 prenylation when the MVA pathway is inhibited, the functional role of these small GTPases in conferring statin sensitivity have been conflicting (reviewed below).

Heme A and coenzyme Q. Two components of the electron transport chain (ETC) require isoprenoid metabolites produced by the MVA pathway. Heme A is an iron-chelating compound that participates in the reduction of dioxygen to water by cytochrome c oxidase during aerobic respiration (215). Heme A contains a hydroethylfarnesyl group derived from FPP (216) (Fig 1-

2), which is functionally important for the conservation of the energy of oxygen reduction (217).

Larger isoprenoids containing 6-10 IPP subunits are also used in the production of coenzyme Q

(CoQ) (Fig 1-2). The hydrophobic isoprenoid chain localizes CoQ to the inner membrane of the mitochondria, where the quinone group transfers electrons from complex I or II to complex III of the electron transport chain, thus enabling ATP (adenosine triphosphate) production (218).

Synthesis of isoprenoid end products via the MVA pathway is therefore critical for ATP production during oxidative phosphorylation.

Dolichol. Dolichol is a large isoprenoid end product of the MVA pathway, derived from 19-21

IPP molecules (Fig 1-2). Dolichol resides in the endoplasmic reticulum (ER) as the carrier of N- glycans such as N-acetylglucosamine (GlcNAc), mannose (Man), glucose (Glc), as well as the precursor glycan (GlcNAc2-Man9-Glc3) prior to its en bloc transfer onto nascent polypeptides

(219, 220). The cellular demand for dolichol synthesis is large and continuous, as the synthesis of each precursor glycan to be transferred to nascent peptides requires 8 dolichol molecules

(221). Recycling of dolichol is poorly characterized (222) and thought to be inefficient, as it accumulates with aging (223). Furthermore, N-glycosylation occurs on an estimated 50-70% of all proteins (224), many of which contain multiple sites for N-glycosylation (although not all

26 may be occupied) (225). Protein N-glycosylation is frequently altered in cancer and plays central roles in tumour formation, proliferation and metastasis (226).

1.4.2 Deregulation

LDL-cholesterol uptake. Deregulation of the MVA pathway can occur at multiple levels.

Atypical plasma lipid profiles are commonly observed in all cancer types, usually with lower levels of serum cholesterol and elevated levels of triglycerides (227). The lowered cholesterol levels are thought to be mediated by a 2- to 11-fold increased expression and activity of LDLR in malignant cells (Fig 1-3A), leading to removal of LDL-cholesterol from the serum (228, 229).

Cholesterol is then used for rapid proliferation (230) or to populate elevated levels of cholesterol- rich lipid rafts (231).

Genomic alterations and protein expression. Analysis of large cancer genomics datasets using cBioPortal (232, 233) shows that MVA pathway genes are frequently amplified (eg. HMGCS1,

GGPS1) or mutated (eg. SREBP2) in a variety of cancer types. These genomic alterations (Fig 1-

3B), combined with multiple lines of evidence indicating that cancer cells upregulate the MVA pathway enzymes (Fig 1-3F) for acquisition of MVA pathway end products (113, 178, 234, 235), indicate that deregulation of the MVA pathway can contribute to cell transformation.

Remarkably, overexpression of the catalytic domain of HMGCR could cooperate with activated

RAS to transform normal diploid mouse embryonic fibroblasts (MEFs) (177).

The MVA pathway could also be deregulated through altered SREBP2 processing (Fig 1-3D) or aberrant sterol feedback mechanism (Fig 1-3E). Our lab has previously reported that a panel of

17 multiple myeloma (MM) cell lines can be split into two categories: those that can and do mount a sterol feedback response after treatment with statins, and those that cannot or do not do so (113). Why or how this arises in cancer cells is currently unknown, but clearly indicates that

A Increased LDLR expression LDLR G cytoplasm nucleus Altered upstream Glucose, metabolic glutamine, SREBP2 target genes acetate B pathways HMGCS1 SRE Genomic acetyl-CoA HMGCR SRE alterations F LDLR acetoacetyl-CoA SRE C Deregulated HMGCS1 enzyme activity HMG-CoA Altered regulatory or turnover HMGCR signaling MVA pathways Active SREBP2 MVA end-products, SREBP2 esp. cholesterol D E cleavage Altered SREBP2 Deregulated processing sterol feedback Latent SREBP2

Figure 1-3. Possible mechanisms of deregulation of the MVA pathway in cancer cells.

A simplified MVA pathway is shown on the left in the cytoplasm of the schematic of a cell. The MVA pathway is regulated by a robust negative feedback mechanism involving cholesterol- mediated inhibition of cleavage of the transcription factor sterol regulatory element binding transcription factor 2 (SREBP2). Representative SREBP2-regulated genes is shown in the nucleus on the right. HMGCS1, 3-hydroxy-3-methylglutaryl-CoA synthase 1; HMGCR, 3- hydroxy-3-methylglutaryl-CoA reductase; LDLR, low density lipoprotein receptor. A-G, possible mechanisms of MVA pathway deregulation in cancer cells. Deregulation of the MVA pathway can contribute to cancer initiation and/or progression.

28 deregulation of the sterol feedback response is selected for in certain cancer patients.

Intersection with cancer cell signaling. Multiple oncogenic and tumour-suppressive pathways regulate the MVA pathway (Fig 1-3C), and can contribute to deregulation of the MVA pathway in cancer cells. The PI3K-AKT/PKB-mTORC1 signaling axis (PI3K, phosphoinositide 3-kinase;

AKT/PKB, protein kinase B; mTORC1, mammalian target of rapamycin complex 1) is the major regulator of cell proliferation in response to growth factor stimulation, and mediates its effects partly through increasing lipid and cholesterol production (236-238). This occurs through activation of SREBP by multiple mechanisms, resulting in upregulation of the FAS and MVA pathways (236-238). AMPK (5' adenosine monophosphate-activated protein kinase) is a negative regulator of cell growth in response to low intracellular ATP levels, and reduces the activity of both SREBP and HMGCR (239, 240). Several gain-of-function p53 mutations commonly found in cancer cells have been shown to functionally interact with nuclear SREBP2, and increase the transcription of MVA pathway genes (176). Additionally, metabolites produced by the MVA pathway play integral roles in multiple signaling pathways (Fig 1-3F). Signaling through the

YAP/TAZ (yes-associated protein 1/transcriptional co‑activator with PDZ-binding motif) pathway, a major pro-growth and anti-apoptosis signal in some cancer cells, requires MVA pathway activity to produce GGPP (175, 241). GGPP allows prenylation and membrane localization of RHOA (Ras homolog gene family member A), whose activity is necessary for

YAP/TAZ signaling transduction (175, 241). Cholesterol-derived end products of the mevalonate pathway, particularly steroid hormones, are also important drivers of initiation and progression of hormone-dependent cancers such as breast and prostate cancers (151).

Intersection with cancer cell metabolism. Recent studies in cancer metabolism indicate that cancer cells preferentially uptake glucose and alternative carbon sources from the external

29 milieu, including acetate (242, 243) and glutamine (244, 245). This leads to abnormally elevated rate of acetyl-CoA production, which is used for many cellular processes, including metabolic processes such as the TCA (tricarboxylic acid) cycle, FAS pathway, and MVA pathway (242).

The Appendix of this thesis presents preliminary data showing that the MVA pathway can be deregulated by metabolites upstream of MVA, such as acetyl-CoA produced from altered upstream metabolic pathways (Fig 1-3G).

Systemic contribution. MVA pathway activity in non-malignant tissue may have systemic activity to fuel cancer cell growth. For example, cholesterol synthesized from the liver may be preferentially used to fuel cancer cell growth due to upregulation of LDLR in cancer cells, as reviewed above. Steroid hormones required for hormone-dependent breast and prostate cancers may be produced by gonads and adrenal glands. On the other hand, MVA pathway products have been associated with activation of the immune system: cholesterol skews T cell programming to promote a T helper (Th)1-to-Th2 switch, and sustains an increase in Tregs in the spleen (246,

247). The isoprenoid metabolites of the MVA pathway, including IPP, FPP, and GGPP, can activate the innate-like γδ T cells, with the diphosphate moiety being recognized as a non-self mycobacterial antigen (248), thus promoting immune-surveillance of cancer cells with elevated flux through the MVA pathway.

1.5 Mechanism of statin-induced cancer cell apoptosis

Current model. We and others have shown that statin-induced cancer cell apoptosis is an on- target effect that can be rescued by co-administration with MVA (113, 187, 249), or can be rescued with FPP and GGPP (113, 214, 249-252). FPP and GGPP are essential substrates for protein geranylgeranylation and farnesylation, respectively, together referred to as protein prenylation (204-206). These processes add a hydrophobic farnesyl or geranylgeranyl moiety to

30 proteins containing a C-terminal CaaX domain (cysteine-aliphatic-aliphatic-any amino acid), thus localizing them to cellular membranes (207, 208). Prenylation-driven membrane localization is required for all proteins in the RAS GTPase superfamily (207, 208), prompting many groups to examine the effect of statin treatment on prenylation status of RAS family proteins. Statin treatment leads to a decrease in the prenylated and membrane-associated forms of RAS super family proteins, including those in the RAS, RHO, RAC, RAP, and RAB subfamilies (209-214). This inhibition of protein prenylation is reversed with FPP or GGPP co- treatment with statins (113, 174, 209-211, 213, 249, 250). On the other hand, cholesterol, CoQ, and dolichol cannot rescue cancer cell apoptosis when co-administered with statins (113, 114).

Based on these data, the current model of the mechanism statin-induced cancer cell apoptosis is via inhibition of FPP and GGPP synthesis, which becomes limiting for prenylation of proteins in the RAS superfamily.

In cells that are dependent on RAS, such as many cancer cells that harbor mutations in RAS proteins or upregulated RAS signaling pathways, inhibition of RAS prenylation may be therapeutic (253-255). This was the scientific basis for the development of farnesyltransferase inhibitors (FTIs) and geranylgeranyltransferase inhibitors (GGTIs), both of which have shown efficacy in pre-clinical models but have been limited in clinical application due to high general toxicity (256). A number of studies have therefore tested the use of low-dose GGTIs/FTIs in combination with statins, and have shown that the safety profiles can be improved (257, 258).

Discrepancies. The current model of the mechanism of action of statins described above predicts that cancer cells that are dependent on RAS or RAS family members would be more sensitive to statins, as they cannot survive without the requisite protein prenylation. However, this has not been consistently observed. For example, cancer cells with upregulated or hyperactivated RAS or

RHO are associated with increased statin sensitivity in some studies (209-211), but not others

(113, 174, 251). The majority of epidemiological studies and clinical trials report no association between response to statins and RAS mutations (259-264). In cell lines that were sensitive to statin-induced cell kill, rescuing RAS localization (113, 265) or RAF-MEK-ERK signaling (266) did not make cells resistant, and intrinsic sensitivity to statin kill was largely independent of RAS function (113, 169, 267). The literature on other prenylated proteins are also inconsistent with this model of statin kill.

The amino acid sequence in the CaaX domain dictates whether a given protein prefers to be prenylated using FPP, GGPP, or have no preference for either one over the other (Fig 1-4) (254,

255). Evidence from studies using GGTIs and/or FTIs indicate that many proteins can be alternatively prenylated when their preferred process is limited: for example, when farnesylation is inhibited, KRAS (CaaX sequence: CVIM) and NRAS (CaaX sequence: CVVM) can be alternatively prenylated by geranylgeranylation (Fig 1-4), and remain fully functional (268, 269).

Curiously, however, many studies have reported that while GGPP consistently rescues statin- induced cancer cell death (113, 210, 211, 213, 249-252), FPP does so less consistently, sometimes rescuing completely (249), sometimes partially (113, 210, 250, 251), and sometimes not at all (174, 210, 252). In general, these results were interpreted in two ways. The first interpretation is that protein geranylgeranylation plays a more important role than protein farnesylation in cancer cell survival. However, this has not been supported by studies using

GGTIs and FTIs to specifically block these processes (255, 268, 269). The second interpretation suggests that FPP is shunted towards cholesterol synthesis rather than being used for protein prenylation. Indeed, differential KM suggest that decreasing FPP concentrations are 1,000-fold more likely to impact flux through squalene synthase, the first step of sterol synthesis, than the activity of farnesyltransferase (270). Accordingly, it has been suggested that all observations of

Statins MVA pathway

FTase GGTase FPP GGPP

FTIs GGTIs

Protein-CaaQ Protein-CaaF ‡ Protein-CaaI

Protein-CaaS Protein-CKVL ‡ Protein-CaaL †

Protein-CaaC † can be alternatively prenylated if preferred process is inhibited Protein-CaaM † ‡ can be farnesylated or geranylgeranylated

Figure 1-4. Prenylation substrate is determined by the amino acid sequence in the CaaX domain.

The CaaX amino acid sequence determines requirement or preference for farnesylation or geranylgeranylation. The chemical structures of the farnesyl group (left) and geranylgeranyl group (right) are shown. FTase, farnesyltransferase; FTIs, farnesyltransferase inhibitors; FPP, farnesyl diphosphate; GGPP, geranylgeranyl diphosphate; GGTase, geranylgeranyltransferase; GGTIs, geranylgeranyltransferase inhibitors.

33 statins inhibiting protein prenylation are artefacts of supra-pharmacologic levels of statins used in tissue culture (271), and statins in fact have no effect on protein prenylation in vivo (254).

These contradicting data raise the possibility that inhibition of protein prenylation is likely not the sole contributor to statin sensitivity (254). This implicates not only an alternative mechanism of statin-induced cell death, but also the potential to develop better biomarkers for the identification of patients that will benefit from statin treatment.

1.6 Objectives and thesis outline

The overarching hypothesis of this thesis is that inhibition of the MVA pathway by statins can induce cell death in a subset of breast cancer cells, and improve clinical outcome. My objectives were (i) to understand the rationale behind the discrepancies in the current model of statin mechanism of action; (ii) to identify molecular biomarkers that could predict sensitivity to statins in breast cancer cell lines; and (iii) to identify the point of therapeutic intervention in which statin treatment will impact breast cancer patient care.

Chapter 2 of this thesis presents direct evidence that inhibition of protein prenylation is independent from statin-induced apoptosis. We showed that increased cellular demand for GGPP and FPP did not always sensitize breast cancer cells to fluvastatin, and when it did, the effect was independent from inhibition of protein prenylation. We then identified breast cancer cell EMT to be the underlying cause of increased sensitivity to statin-induced cell death. Five bi-modally distributed EMT genes was sufficient to accurately predict statin sensitivity in ~700 cancer cell lines, suggesting cancer cell EMT to be a universal biomarker of statin sensitivity.

In Chapter 3, we further delineated the mechanism of statin-induced breast cancer cell death, and identified the inhibition of dolichol biosynthesis to be the mechanism of statin sensitivity in breast cancer cells that have undergone EMT. Dolichol is a large isoprenoid metabolite produced

34 by the MVA pathway, and is important for N-linked glycosylation of nascent peptides.

Fluvastatin treatment impairs the presentation of the EMT-associated β1,6-branched, tri- and tetra-antennary, unfucosylated and singly fucosylated complex glycans, both in cell culture and in vivo. We then tested the efficacy of fluvastatin in a mouse model of post-surgical metastatic breast cancer, that closely mimics the treatments experienced by early-stage breast cancer patients. We showed that fluvastatin treatment in the adjuvant setting effectively delayed breast cancer metastasis and improved survival.

The Appendix of this thesis document observations suggesting a novel mechanism of MVA pathway (de)regulation, that may contribute to oncogenic transformation through intersection with cancer cell metabolism.

Overall, this thesis significantly advances our understanding of how statins induce breast cancer cell death. Two biomarkers of statin sensitivity are identified, one based on gene expression and one based on glycan presentation. We further show in a pre-clinical mouse model of breast cancer that fluvastatin treatment in the adjuvant therapeutic setting can effectively delay metastatic recurrence. By demonstrating efficacy, elucidating mechanism, and identifying biomarkers of fluvastatin sensitivity, this thesis provides evidence to support the immediate repurposing of this effective, safe, and inexpensive drug to target breast cancer metastasis.

1.7 References

1. World Health Organization: World Cancer Report. 2014.

2. Centers for Disease Control and Prevention: National Center for Health Statistics: Life Stages and Populations: Deaths. 2016.

3. Statistics Canada: Canadian Socio-Economic Information Management System: Leading causes of death, by sex (both sexes). 2015.

4. National Cancer Institute: Types of Cancer Treatment: Surgery. 2015.

5. Darzynkiewicz Z, Traganos F, Wlodkowic D: Impaired DNA damage response--an Achilles' heel sensitizing cancer to chemotherapy and radiotherapy. Eur J Pharmacol 2009, 625(1-3):143-150.

6. American Cancer Society: A Guide to Radiation Therapy. 2015.

7. Corrie PG: Cytotoxic chemotherapy: clinical aspects. Medicine 2008, 36(1):24-28.

8. Morgan G, Ward R, Barton M: The contribution of cytotoxic chemotherapy to 5-year survival in adult malignancies. Clin Oncol (R Coll Radiol) 2004, 16(8):549-560.

9. Hanahan D, Weinberg RA: Hallmarks of cancer: the next generation. Cell 2011, 144(5):646-674.

10. Vogelstein B, Papadopoulos N, Velculescu VE, Zhou S, Diaz LA, Jr., Kinzler KW: Cancer genome landscapes. Science 2013, 339(6127):1546-1558.

11. Gerber DE: Targeted therapies: a new generation of cancer treatments. Am Fam Physician 2008, 77(3):311-319.

12. Filippakopoulos P, Qi J, Picaud S, Shen Y, Smith WB, Fedorov O, Morse EM, Keates T, Hickman TT, Felletar I et al: Selective inhibition of BET bromodomains. Nature 2010, 468(7327):1067-1073.

13. Zuber J, Shi J, Wang E, Rappaport AR, Herrmann H, Sison EA, Magoon D, Qi J, Blatt K, Wunderlich M et al: RNAi screen identifies Brd4 as a therapeutic target in acute myeloid leukaemia. Nature 2011, 478(7370):524-528.

14. Duan JX, Jiao H, Kaizerman J, Stanton T, Evans JW, Lan L, Lorente G, Banica M, Jung D, Wang J et al: Potent and highly selective hypoxia-activated achiral phosphoramidate mustards as anticancer drugs. J Med Chem 2008, 51(8):2412-2420.

15. Hu J, Handisides DR, Van Valckenborgh E, De Raeve H, Menu E, Vande Broek I, Liu Q, Sun JD, Van Camp B, Hart CP et al: Targeting the multiple myeloma hypoxic niche with TH-302, a hypoxia-activated prodrug. Blood 2010, 116(9):1524-1527.

16. American Cancer Society: Cancer Facts and Figures 2016. 2016.

17. Niraula S, Seruga B, Ocana A, Shao T, Goldstein R, Tannock IF, Amir E: The price we pay for progress: a meta-analysis of harms of newly approved anticancer drugs. J Clin Oncol 2012, 30(24):3012-3019.

18. Ding PN, Lord SJ, Gebski V, Links M, Bray V, Gralla RJ, Yang JC, Lee CK: Risk of Treatment-Related Toxicities from EGFR Tyrosine Kinase Inhibitors: A Meta- analysis of Clinical Trials of Gefitinib, Erlotinib, and Afatinib in Advanced EGFR- Mutated Non-Small Cell Lung Cancer. J Thorac Oncol 2016.

19. Force T, Kerkela R: Cardiotoxicity of the new cancer therapeutics--mechanisms of, and approaches to, the problem. Drug Discov Today 2008, 13(17-18):778-784.

20. Fojo T, Grady C: How much is life worth: cetuximab, non-small cell lung cancer, and the $440 billion question. J Natl Cancer Inst 2009, 101(15):1044-1048.

21. Hodi FS, O'Day SJ, McDermott DF, Weber RW, Sosman JA, Haanen JB, Gonzalez R, Robert C, Schadendorf D, Hassel JC et al: Improved survival with ipilimumab in patients with metastatic melanoma. N Engl J Med 2010, 363(8):711-723.

22. Robert C, Schachter J, Long GV, Arance A, Grob JJ, Mortier L, Daud A, Carlino MS, McNeil C, Lotem M et al: Pembrolizumab versus Ipilimumab in Advanced Melanoma. N Engl J Med 2015, 372(26):2521-2532.

23. Robert C, Long GV, Brady B, Dutriaux C, Maio M, Mortier L, Hassel JC, Rutkowski P, McNeil C, Kalinka-Warzocha E et al: Nivolumab in previously untreated melanoma without BRAF mutation. N Engl J Med 2015, 372(4):320-330.

24. Rizvi NA, Mazieres J, Planchard D, Stinchcombe TE, Dy GK, Antonia SJ, Horn L, Lena H, Minenza E, Mennecier B et al: Activity and safety of nivolumab, an anti-PD-1 immune checkpoint inhibitor, for patients with advanced, refractory squamous non- small-cell lung cancer (CheckMate 063): a phase 2, single-arm trial. Lancet Oncol 2015, 16(3):257-265.

25. Khalil DN, Smith EL, Brentjens RJ, Wolchok JD: The future of cancer treatment: immunomodulation, CARs and combination immunotherapy. Nat Rev Clin Oncol 2016, 13(6):394.

26. Weber JS, Kahler KC, Hauschild A: Management of immune-related adverse events and kinetics of response with ipilimumab. J Clin Oncol 2012, 30(21):2691-2697.

27. Kantarjian H, Steensma D, Rius Sanjuan J, Elshaug A, Light D: High cancer drug prices in the United States: reasons and proposed solutions. J Oncol Pract 2014, 10(4):e208-211.

28. Chapman PB, D'Angelo SP, Wolchok JD: Rapid eradication of a bulky melanoma mass with one dose of immunotherapy. N Engl J Med 2015, 372(21):2073-2074.

29. Postow MA, Chesney J, Pavlick AC, Robert C, Grossmann K, McDermott D, Linette GP, Meyer N, Giguere JK, Agarwala SS et al: Nivolumab and ipilimumab versus ipilimumab in untreated melanoma. N Engl J Med 2015, 372(21):2006-2017.

30. Schadendorf D, Hodi FS, Robert C, Weber JS, Margolin K, Hamid O, Patt D, Chen TT, Berman DM, Wolchok JD: Pooled Analysis of Long-Term Survival Data From Phase II and Phase III Trials of Ipilimumab in Unresectable or Metastatic Melanoma. J Clin Oncol 2015, 33(17):1889-1894.

31. Pedicord VA, Montalvo W, Leiner IM, Allison JP: Single dose of anti-CTLA-4 enhances CD8+ T-cell memory formation, function, and maintenance. Proceedings of the National Academy of Sciences of the United States of America 2011, 108(1):266- 271.

32. Farkona S, Diamandis EP, Blasutig IM: Cancer immunotherapy: the beginning of the end of cancer? BMC Med 2016, 14:73.

33. Schnitt SJ: Classification and prognosis of invasive breast cancer: from morphology to molecular taxonomy. Mod Pathol 2010, 23 Suppl 2:S60-64.

34. Canadian Breast Cancer Foundation: Metastatic breast cancer. 2014.

35. Canadian Breast Cancer Foundation: Staging and Grading Breast Cancer. 2014.

36. ASCO Care and Treatment Recommendations for Patients: Hormonal Therapy for Early-Stage Hormone Receptor-Positive Breast Cancer. 2016.

37. Kennecke H, Yerushalmi R, Woods R, Cheang MC, Voduc D, Speers CH, Nielsen TO, Gelmon K: Metastatic behavior of breast cancer subtypes. J Clin Oncol 2010, 28(20):3271-3277.

38. Kittaneh M, Montero AJ, Gluck S: Molecular profiling for breast cancer: a comprehensive review. Biomark Cancer 2013, 5:61-70.

39. Livraghi L, Garber JE: PARP inhibitors in the management of breast cancer: current data and future prospects. BMC Med 2015, 13:188.

40. Rostas JW, Dyess DL: Current operative management of breast cancer: an age of smaller resections and bigger cures. Int J Breast Cancer 2012, 2012:516417.

41. Pennisi A, Kieber-Emmons T, Makhoul I, Hutchins L: Relevance of Pathological Complete Response after Neoadjuvant Therapy for Breast Cancer. Breast Cancer (Auckl) 2016, 10:103-106.

42. Rastogi P, Anderson SJ, Bear HD, Geyer CE, Kahlenberg MS, Robidoux A, Margolese RG, Hoehn JL, Vogel VG, Dakhil SR et al: Preoperative chemotherapy: updates of National Surgical Adjuvant Breast and Bowel Project Protocols B-18 and B-27. J Clin Oncol 2008, 26(5):778-785.

43. U.S. Food and Drug Administration: Pathological Complete Response in Neoadjuvant Treatment of High-Risk Early-Stage Breast Cancer: Use as an Endpoint to Support Accelerated Approval. Guidance for Industry 2014.

44. Colleoni M, Litman HJ, Castiglione-Gertsch M, Sauerbrei W, Gelber RD, Bonetti M, Coates AS, Schumacher M, Bastert G, Rudenstam CM et al: Duration of adjuvant chemotherapy for breast cancer: a joint analysis of two randomised trials investigating three versus six courses of CMF. Br J Cancer 2002, 86(11):1705-1714.

45. Hay M, Thomas DW, Craighead JL, Economides C, Rosenthal J: Clinical development success rates for investigational drugs. Nat Biotechnol 2014, 32(1):40-51.

46. Roses AD: Pharmacogenetics in drug discovery and development: a translational perspective. Nat Rev Drug Discov 2008, 7(10):807-817.

47. Bertolini F, Sukhatme VP, Bouche G: Drug repurposing in oncology--patient and health systems opportunities. Nat Rev Clin Oncol 2015, 12(12):732-742.

48. Corbin JD, Francis SH: Cyclic GMP phosphodiesterase-5: target of sildenafil. J Biol Chem 1999, 274(20):13729-13732.

49. Boolell M, Allen MJ, Ballard SA, Gepi-Attee S, Muirhead GJ, Naylor AM, Osterloh IH, Gingell C: Sildenafil: an orally active type 5 cyclic GMP-specific phosphodiesterase inhibitor for the treatment of penile erectile dysfunction. Int J Impot Res 1996, 8(2):47-52.

50. Franks ME, Macpherson GR, Figg WD: Thalidomide. Lancet 2004, 363(9423):1802- 1811.

51. Kelsey FO: Events after thalidomide. J Dent Res 1967, 46(6):1201-1205.

52. Webb JF: Canadian Thalidomide Experience. Can Med Assoc J 1963, 89:987-992.

53. Silverman WA: The schizophrenic career of a "monster drug". Pediatrics 2002, 110(2 Pt 1):404-406.

54. Sanchez LS, de la Monte SM, Filippov G, Jones RC, Zapol WM, Bloch KD: Cyclic- GMP-binding, cyclic-GMP-specific phosphodiesterase (PDE5) gene expression is regulated during rat pulmonary development. Pediatr Res 1998, 43(2):163-168.

55. Folkman J: Angiogenesis-dependent diseases. Semin Oncol 2001, 28(6):536-542.

56. D'Amato RJ, Loughnan MS, Flynn E, Folkman J: Thalidomide is an inhibitor of angiogenesis. Proceedings of the National Academy of Sciences of the United States of America 1994, 91(9):4082-4085.

57. Singhal S, Mehta J, Desikan R, Ayers D, Roberson P, Eddlemon P, Munshi N, Anaissie E, Wilson C, Dhodapkar M et al: Antitumor activity of thalidomide in refractory multiple myeloma. N Engl J Med 1999, 341(21):1565-1571.

58. Hernan MA, Hernandez-Diaz S, Robins JM: A structural approach to selection bias. Epidemiology 2004, 15(5):615-625.

59. Greenland S, Pearce N: Statistical foundations for model-based adjustments. Annu Rev Public Health 2015, 36:89-108.

60. Evans JM, Donnelly LA, Emslie-Smith AM, Alessi DR, Morris AD: Metformin and reduced risk of cancer in diabetic patients. BMJ 2005, 330(7503):1304-1305.

61. Thun MJ, Jacobs EJ, Patrono C: The role of aspirin in cancer prevention. Nat Rev Clin Oncol 2012, 9(5):259-267.

62. Algra AM, Rothwell PM: Effects of regular aspirin on long-term cancer incidence and metastasis: a systematic comparison of evidence from observational studies versus randomised trials. Lancet Oncol 2012, 13(5):518-527.

63. Poynter JN, Gruber SB, Higgins PD, Almog R, Bonner JD, Rennert HS, Low M, Greenson JK, Rennert G: Statins and the risk of colorectal cancer. N Engl J Med 2005, 352(21):2184-2192.

64. Nielsen SF, Nordestgaard BG, Bojesen SE: Statin use and reduced cancer-related mortality. N Engl J Med 2012, 367(19):1792-1802.

65. Brown D: Unfinished business: target-based drug discovery. Drug Discov Today 2007, 12(23-24):1007-1012.

66. Zhang L, He M, Zhang Y, Nilubol N, Shen M, Kebebew E: Quantitative high- throughput drug screening identifies novel classes of drugs with anticancer activity in thyroid cancer cells: opportunities for repurposing. J Clin Endocrinol Metab 2012, 97(3):E319-328.

67. Cao L, Weetall M, Bombard J, Qi H, Arasu T, Lennox W, Hedrick J, Sheedy J, Risher N, Brooks PC et al: Discovery of Novel Small Molecule Inhibitors of VEGF Expression in Tumor Cells Using a Cell-Based High Throughput Screening Platform. PLoS One 2016, 11(12):e0168366.

68. Shahar OD, Kalousi A, Eini L, Fisher B, Weiss A, Darr J, Mazina O, Bramson S, Kupiec M, Eden A et al: A high-throughput chemical screen with FDA approved drugs reveals that the antihypertensive drug Spironolactone impairs cancer cell survival by inhibiting homology directed repair. Nucleic Acids Res 2014, 42(9):5689-5701.

69. Holbeck SL, Collins JM, Doroshow JH: Analysis of Food and Drug Administration- approved anticancer agents in the NCI60 panel of human tumor cell lines. Mol Cancer Ther 2010, 9(5):1451-1460.

70. Sharmeen S, Skrtic M, Sukhai MA, Hurren R, Gronda M, Wang X, Fonseca SB, Sun H, Wood TE, Ward R et al: The antiparasitic agent ivermectin induces chloride- dependent membrane hyperpolarization and cell death in leukemia cells. Blood 2010, 116(18):3593-3603.

71. Borisy AA, Elliott PJ, Hurst NW, Lee MS, Lehar J, Price ER, Serbedzija G, Zimmermann GR, Foley MA, Stockwell BR et al: Systematic discovery of multicomponent therapeutics. Proceedings of the National Academy of Sciences of the United States of America 2003, 100(13):7977-7982.

72. Pandyra A, Mullen PJ, Kalkat M, Yu R, Pong JT, Li Z, Trudel S, Lang KS, Minden MD, Schimmer AD et al: Immediate utility of two approved agents to target both the metabolic mevalonate pathway and its restorative feedback loop. Cancer research 2014, 74(17):4772-4782.

73. Lamb J: The Connectivity Map: a new tool for biomedical research. Nat Rev Cancer 2007, 7(1):54-60.

74. Lamb J, Crawford ED, Peck D, Modell JW, Blat IC, Wrobel MJ, Lerner J, Brunet JP, Subramanian A, Ross KN et al: The Connectivity Map: using gene-expression signatures to connect small molecules, genes, and disease. Science 2006, 313(5795):1929-1935.

75. Jahchan NS, Dudley JT, Mazur PK, Flores N, Yang D, Palmerton A, Zmoos AF, Vaka D, Tran KQ, Zhou M et al: A drug repositioning approach identifies tricyclic antidepressants as inhibitors of small cell lung cancer and other neuroendocrine tumors. Cancer Discov 2013, 3(12):1364-1377.

76. Zerbini LF, Bhasin MK, de Vasconcellos JF, Paccez JD, Gu X, Kung AL, Libermann TA: Computational repositioning and preclinical validation of pentamidine for renal cell cancer. Mol Cancer Ther 2014, 13(7):1929-1941.

77. Campillos M, Kuhn M, Gavin AC, Jensen LJ, Bork P: Drug target identification using side-effect similarity. Science 2008, 321(5886):263-266.

78. Bresso E, Grisoni R, Marchetti G, Karaboga AS, Souchet M, Devignes MD, Smail- Tabbone M: Integrative relational machine-learning for understanding drug side- effect profiles. BMC Bioinformatics 2013, 14:207.

79. Amelio I, Gostev M, Knight RA, Willis AE, Melino G, Antonov AV: DRUGSURV: a resource for repositioning of approved and experimental drugs in oncology based on patient survival information. Cell Death Dis 2014, 5:e1051.

80. Lussier YA, Chen JL: The emergence of genome-based drug repositioning. Sci Transl Med 2011, 3(96):96ps35.

81. Jin G, Wong ST: Toward better drug repositioning: prioritizing and integrating existing methods into efficient pipelines. Drug Discov Today 2014, 19(5):637-644.

82. Schenk T, Stengel S, Zelent A: Unlocking the potential of retinoic acid in anticancer therapy. Br J Cancer 2014, 111(11):2039-2045.

83. Castaigne S, Chomienne C, Daniel MT, Ballerini P, Berger R, Fenaux P, Degos L: All- trans retinoic acid as a differentiation therapy for acute promyelocytic leukemia. I. Clinical results. Blood 1990, 76(9):1704-1709.

84. Lo-Coco F, Avvisati G, Vignetti M, Thiede C, Orlando SM, Iacobelli S, Ferrara F, Fazi P, Cicconi L, Di Bona E et al: Retinoic acid and arsenic trioxide for acute promyelocytic leukemia. N Engl J Med 2013, 369(2):111-121.

85. Ades L, Guerci A, Raffoux E, Sanz M, Chevallier P, Lapusan S, Recher C, Thomas X, Rayon C, Castaigne S et al: Very long-term outcome of acute promyelocytic leukemia after treatment with all-trans retinoic acid and chemotherapy: the European APL Group experience. Blood 2010, 115(9):1690-1696.

86. U.S. Food and Drug Administration: GLEEVEC (imatinib mesylate) tablets Label. 2008.

87. Dagher R, Cohen M, Williams G, Rothmann M, Gobburu J, Robbie G, Rahman A, Chen G, Staten A, Griebel D et al: Approval summary: imatinib mesylate in the treatment of metastatic and/or unresectable malignant gastrointestinal stromal tumors. Clin Cancer Res 2002, 8(10):3034-3038.

88. Yin M, Zhou J, Gorak EJ, Quddus F: Metformin is associated with survival benefit in cancer patients with concurrent type 2 diabetes: a systematic review and meta- analysis. Oncologist 2013, 18(12):1248-1255.

89. Gandini S, Puntoni M, Heckman-Stoddard BM, Dunn BK, Ford L, DeCensi A, Szabo E: Metformin and cancer risk and mortality: a systematic review and meta-analysis taking into account biases and confounders. Cancer Prev Res (Phila) 2014, 7(9):867- 885.

90. Orecchioni S, Reggiani F, Talarico G, Mancuso P, Calleri A, Gregato G, Labanca V, Noonan DM, Dallaglio K, Albini A et al: The biguanides metformin and phenformin inhibit angiogenesis, local and metastatic growth of breast cancer by targeting both neoplastic and microenvironment cells. Int J Cancer 2015, 136(6):E534-544.

91. Blandino G, Valerio M, Cioce M, Mori F, Casadei L, Pulito C, Sacconi A, Biagioni F, Cortese G, Galanti S et al: Metformin elicits anticancer effects through the sequential modulation of DICER and c-MYC. Nat Commun 2012, 3:865.

92. Hadad SM, Coates P, Jordan LB, Dowling RJ, Chang MC, Done SJ, Purdie CA, Goodwin PJ, Stambolic V, Moulder-Thompson S et al: Evidence for biological effects of metformin in operable breast cancer: biomarker analysis in a pre-operative window of opportunity randomized trial. Breast Cancer Res Treat 2015, 150(1):149- 155.

93. Hadad S, Iwamoto T, Jordan L, Purdie C, Bray S, Baker L, Jellema G, Deharo S, Hardie DG, Pusztai L et al: Evidence for biological effects of metformin in operable breast cancer: a pre-operative, window-of-opportunity, randomized trial. Breast Cancer Res Treat 2011, 128(3):783-794.

94. Goodwin PJ, Pritchard KI, Ennis M, Clemons M, Graham M, Fantus IG: Insulin- lowering effects of metformin in women with early breast cancer. Clin Breast Cancer 2008, 8(6):501-505.

95. Pernicova I, Korbonits M: Metformin--mode of action and clinical implications for diabetes and cancer. Nat Rev Endocrinol 2014, 10(3):143-156.

96. Ishikawa H, Mutoh M, Suzuki S, Tokudome S, Saida Y, Abe T, Okamura S, Tajika M, Joh T, Tanaka S et al: The preventive effects of low-dose enteric-coated aspirin tablets on the development of colorectal tumours in Asian patients: a randomised trial. Gut 2014, 63(11):1755-1759.

97. Sandler RS, Halabi S, Baron JA, Budinger S, Paskett E, Keresztes R, Petrelli N, Pipas JM, Karp DD, Loprinzi CL et al: A randomized trial of aspirin to prevent colorectal adenomas in patients with previous colorectal cancer. N Engl J Med 2003, 348(10):883-890.

98. Cole BF, Logan RF, Halabi S, Benamouzig R, Sandler RS, Grainge MJ, Chaussade S, Baron JA: Aspirin for the chemoprevention of colorectal adenomas: meta-analysis of the randomized trials. J Natl Cancer Inst 2009, 101(4):256-266.

99. Nan H, Hutter CM, Lin Y, Jacobs EJ, Ulrich CM, White E, Baron JA, Berndt SI, Brenner H, Butterbach K et al: Association of aspirin and NSAID use with risk of colorectal cancer according to genetic variants. JAMA 2015, 313(11):1133-1142.

100. Fink SP, Yamauchi M, Nishihara R, Jung S, Kuchiba A, Wu K, Cho E, Giovannucci E, Fuchs CS, Ogino S et al: Aspirin and the risk of colorectal cancer in relation to the expression of 15-hydroxyprostaglandin dehydrogenase (HPGD). Sci Transl Med 2014, 6(233):233re232.

101. Chan AT, Ogino S, Fuchs CS: Aspirin use and survival after diagnosis of colorectal cancer. JAMA 2009, 302(6):649-658.

102. Walker I, Newell H: Do molecularly targeted agents in oncology have reduced attrition rates? Nat Rev Drug Discov 2009, 8(1):15-16.

103. Kola I, Landis J: Can the pharmaceutical industry reduce attrition rates? Nat Rev Drug Discov 2004, 3(8):711-715.

104. Garrido-Laguna I, Hidalgo M, Kurzrock R: The inverted pyramid of biomarker-driven trials. Nat Rev Clin Oncol 2011, 8(9):562-566.

105. Lopez-Girona A, Mendy D, Ito T, Miller K, Gandhi AK, Kang J, Karasawa S, Carmel G, Jackson P, Abbasian M et al: Cereblon is a direct protein target for immunomodulatory and antiproliferative activities of lenalidomide and pomalidomide. Leukemia 2012, 26(11):2326-2335.

106. Ito T, Ando H, Suzuki T, Ogura T, Hotta K, Imamura Y, Yamaguchi Y, Handa H: Identification of a primary target of thalidomide teratogenicity. Science 2010, 327(5971):1345-1350.

107. Schuster SR, Kortuem KM, Zhu YX, Braggio E, Shi CX, Bruins LA, Schmidt JE, Ahmann G, Kumar S, Rajkumar SV et al: The clinical significance of cereblon expression in multiple myeloma. Leuk Res 2014, 38(1):23-28.

108. Delimaris I, Faviou E, Antonakos G, Stathopoulou E, Zachari A, Dionyssiou-Asteriou A: Oxidized LDL, serum oxidizability and serum lipid levels in patients with breast or ovarian cancer. Clin Biochem 2007, 40(15):1129-1134.

109. Saintot M, Astre C, Pujol H, Gerber M: Tumor progression and oxidant-antioxidant status. Carcinogenesis 1996, 17(6):1267-1271.

110. Coates AS, Keshaviah A, Thurlimann B, Mouridsen H, Mauriac L, Forbes JF, Paridaens R, Castiglione-Gertsch M, Gelber RD, Colleoni M et al: Five years of letrozole compared with tamoxifen as initial adjuvant therapy for postmenopausal women with endocrine-responsive early breast cancer: update of study BIG 1-98. J Clin Oncol 2007, 25(5):486-492.

111. Engan T, Krane J, Johannessen DC, Lonning PE, Kvinnsland S: Plasma changes in breast cancer patients during endocrine therapy--lipid measurements and nuclear magnetic resonance (NMR) spectroscopy. Breast Cancer Res Treat 1995, 36(3):287- 297.

112. Love RR, Wiebe DA, Newcomb PA, Cameron L, Leventhal H, Jordan VC, Feyzi J, DeMets DL: Effects of tamoxifen on cardiovascular risk factors in postmenopausal women. Ann Intern Med 1991, 115(11):860-864.

113. Wong WW, Clendening JW, Martirosyan A, Boutros PC, Bros C, Khosravi F, Jurisica I, Stewart AK, Bergsagel PL, Penn LZ: Determinants of sensitivity to lovastatin-induced apoptosis in multiple myeloma. Mol Cancer Ther 2007, 6(6):1886-1897.

114. Xia Z, Tan MM, Wong WW, Dimitroulakos J, Minden MD, Penn LZ: Blocking protein geranylgeranylation is essential for lovastatin-induced apoptosis of human acute myeloid leukemia cells. Leukemia 2001, 15(9):1398-1407.

115. Kane RC: FDA Oncology Drug Approval - Endpoints, Effectiveness, and Approval. US Food and Drug Administration Archived Documents 2006.

116. Sullivan EJ: Clinical Trial Endpoints. US Food and Drug Administration Archived Documents 2013.

117. Prensner JR, Rubin MA, Wei JT, Chinnaiyan AM: Beyond PSA: the next generation of prostate cancer biomarkers. Sci Transl Med 2012, 4(127):127rv123.

118. Lech G, Slotwinski R, Slodkowski M, Krasnodebski IW: Colorectal cancer tumour markers and biomarkers: Recent therapeutic advances. World J Gastroenterol 2016, 22(5):1745-1755.

119. U.S. Preventive Services Task Force: Screening for ovarian cancer: reaffirmation recommendation statement. Am Fam Physician 2013, 87(10):Online.

120. Clendening JW, Pandyra A, Li Z, Boutros PC, Martirosyan A, Lehner R, Jurisica I, Trudel S, Penn LZ: Exploiting the mevalonate pathway to distinguish statin-sensitive multiple myeloma. Blood 2010, 115(23):4787-4797.

121. Holstein SA, Knapp HR, Clamon GH, Murry DJ, Hohl RJ: Pharmacodynamic effects of high dose lovastatin in subjects with advanced malignancies. Cancer Chemother Pharmacol 2006, 57(2):155-164.

122. Lopez-Aguilar E, Sepulveda-Vildosola AC, Rivera-Marquez H, Cerecedo-Diaz F, Valdez-Sanchez M, Villasis-Keever MA: Security and maximal tolerated doses of fluvastatin in pediatric cancer patients. Arch Med Res 1999, 30(2):128-131.

123. Dowling RJ, Lam S, Bassi C, Mouaaz S, Aman A, Kiyota T, Al-Awar R, Goodwin PJ, Stambolic V: Metformin Pharmacokinetics in Mouse Tumors: Implications for Human Therapy. Cell Metab 2016, 23(4):567-568.

124. Mathijssen RH, Sparreboom A, Verweij J: Determining the optimal dose in the development of anticancer agents. Nat Rev Clin Oncol 2014, 11(5):272-281.

125. Scharovsky OG, Mainetti LE, Rozados VR: Metronomic chemotherapy: changing the paradigm that more is better. Curr Oncol 2009, 16(2):7-15.

126. Klement G, Baruchel S, Rak J, Man S, Clark K, Hicklin DJ, Bohlen P, Kerbel RS: Continuous low-dose therapy with vinblastine and VEGF receptor-2 antibody induces sustained tumor regression without overt toxicity. J Clin Invest 2000, 105(8):R15-24.

127. Browder T, Butterfield CE, Kraling BM, Shi B, Marshall B, O'Reilly MS, Folkman J: Antiangiogenic scheduling of chemotherapy improves efficacy against experimental drug-resistant cancer. Cancer Res 2000, 60(7):1878-1886.

128. Lopez JS, Banerji U: Combine and conquer: challenges for targeted therapy combinations in early phase trials. Nat Rev Clin Oncol 2017, 14(1):57-66.

129. Lohse I, Rasowski J, Cao P, Pintilie M, Do T, Tsao MS, Hill RP, Hedley DW: Targeting hypoxic microenvironment of pancreatic xenografts with the hypoxia-activated prodrug TH-302. Oncotarget 2016, 7(23):33571-33580.

130. Collins FS: Mining for therapeutic gold. Nat Rev Drug Discov 2011, 10(6):397.

131. Gofman JW, Delalla O, Glazier F, Freeman NK, Lindgren FT, Nichols AV, Strisower B, Tamplin AR: The serum lipoprotein transport system in health, metabolic disorders, atherosclerosis and coronary heart disease. J Clin Lipidol 2007, 1(2):104-141.

132. Scandinavian Simvastatin Survival Study Group: Randomised trial of cholesterol lowering in 4444 patients with coronary heart disease: the Scandinavian Simvastatin Survival Study (4S). Lancet 1994, 344(8934):1383-1389.

133. Pedersen TR, Berg K, Cook TJ, Faergeman O, Haghfelt T, Kjekshus J, Miettinen T, Musliner TA, Olsson AG, Pyorala K et al: Safety and tolerability of cholesterol lowering with simvastatin during 5 years in the Scandinavian Simvastatin Survival Study. Arch Intern Med 1996, 156(18):2085-2092.

134. Davidson MH: Safety profiles for the HMG-CoA reductase inhibitors: treatment and trust. Drugs 2001, 61(2):197-206.

135. Endo A, Kuroda M, Tanzawa K: Competitive inhibition of 3-hydroxy-3- methylglutaryl coenzyme A reductase by ML-236A and ML-236B fungal metabolites, having hypocholesterolemic activity. FEBS Lett 1976, 72(2):323-326.

136. Tobert JA: Lovastatin and beyond: the history of the HMG-CoA reductase inhibitors. Nat Rev Drug Discov 2003, 2(7):517-526.

137. Gu Q, Paulose-Ram R, Burt VL, Kit BK: Prescription Cholesterol-lowering Medication Use in Adults Aged 40 and Over: United States, 2003-2012. NCHS Data Brief 2014, 177.

138. Hennessy DA, Tanuseputro P, Tuna M, Bennett C, Perez R, Shields M, Ko DT, Tu J, Manuel DG: Population health impact of statin treatment in Canada. Health Rep 2016, 27(1):20-28.

139. Gazzerro P, Proto MC, Gangemi G, Malfitano AM, Ciaglia E, Pisanti S, Santoro A, Laezza C, Bifulco M: Pharmacological actions of statins: a critical appraisal in the management of cancer. Pharmacol Rev 2012, 64(1):102-146.

140. Bockorny B, Dasanu CA: HMG-CoA reductase inhibitors as adjuvant treatment for hematologic malignancies: what is the current evidence? Ann Hematol 2015, 94(1):1- 12.

141. Chen C, Lin J, Smolarek T, Tremaine L: P-glycoprotein has differential effects on the disposition of statin acid and lactone forms in mdr1a/b knockout and wild-type mice. Drug Metab Dispos 2007, 35(10):1725-1729.

142. Goard CA, Mather RG, Vinepal B, Clendening JW, Martirosyan A, Boutros PC, Sharom FJ, Penn LZ: Differential interactions between statins and P-glycoprotein: implications for exploiting statins as anticancer agents. Int J Cancer 2010, 127(12):2936-2948.

143. Rao S, Porter DC, Chen X, Herliczek T, Lowe M, Keyomarsi K: Lovastatin-mediated G1 arrest is through inhibition of the proteasome, independent of hydroxymethyl glutaryl-CoA reductase. Proceedings of the National Academy of Sciences of the United States of America 1999, 96(14):7797-7802.

144. Park IH, Kim JY, Choi JY, Han JY: Simvastatin enhances irinotecan-induced apoptosis in human non-small cell lung cancer cells by inhibition of proteasome activity. Invest New Drugs 2011, 29(5):883-890.

145. Murray SS, Tu KN, Young KL, Murray EJ: The effects of lovastatin on proteasome activities in highly purified rabbit 20 S proteasome preparations and mouse MC3T3-E1 osteoblastic cells. Metabolism 2002, 51(9):1153-1160.

146. Bucher NL, Overath P, Lynen F: beta-Hydroxy-beta-methyl-glutaryl coenzyme A reductase, cleavage and condensing enzymes in relation to cholesterol formation in rat liver. Biochim Biophys Acta 1960, 40:491-501.

147. Brown MS, Goldstein JL: A receptor-mediated pathway for cholesterol homeostasis. Science 1986, 232(4746):34-47.

148. Goldstein JL, Brown MS: A century of cholesterol and coronaries: from plaques to genes to statins. Cell 2015, 161(1):161-172.

149. Amemiya-Kudo M, Shimano H, Hasty AH, Yahagi N, Yoshikawa T, Matsuzaka T, Okazaki H, Tamura Y, Iizuka Y, Ohashi K et al: Transcriptional activities of nuclear SREBP-1a, -1c, and -2 to different target promoters of lipogenic and cholesterogenic genes. J Lipid Res 2002, 43(8):1220-1235.

150. Shimomura I, Shimano H, Horton JD, Goldstein JL, Brown MS: Differential expression of exons 1a and 1c in mRNAs for sterol regulatory element binding protein-1 in human and mouse organs and cultured cells. J Clin Invest 1997, 99(5):838-845.

151. Mullen PJ, Yu R, Longo J, Archer MC, Penn LZ: The interplay between cell signalling and the mevalonate pathway in cancer. Nat Rev Cancer 2016, 16(11):718-731.

152. Clendening JW, Penn LZ: Targeting tumor cell metabolism with statins. Oncogene 2012, 31(48):4967-4978.

153. Brown MS, Goldstein JL: Multivalent feedback regulation of HMG CoA reductase, a control mechanism coordinating isoprenoid synthesis and cell growth. J Lipid Res 1980, 21(5):505-517.

154. Pocathikorn A, Taylor RR, Mamotte CD: Atorvastatin increases expression of low- density lipoprotein receptor mRNA in human circulating mononuclear cells. Clin Exp Pharmacol Physiol 2010, 37(4):471-476.

155. Cauley JA, McTiernan A, Rodabough RJ, LaCroix A, Bauer DC, Margolis KL, Paskett ED, Vitolins MZ, Furberg CD, Chlebowski RT: Statin use and breast cancer:

prospective results from the Women's Health Initiative. J Natl Cancer Inst 2006, 98(10):700-707.

156. Kumar AS, Benz CC, Shim V, Minami CA, Moore DH, Esserman LJ: Estrogen receptor-negative breast cancer is less likely to arise among lipophilic statin users. Cancer Epidemiol Biomarkers Prev 2008, 17(5):1028-1033.

157. Graaf MR, Beiderbeck AB, Egberts AC, Richel DJ, Guchelaar HJ: The risk of cancer in users of statins. J Clin Oncol 2004, 22(12):2388-2394.

158. Khurana V, Bejjanki HR, Caldito G, Owens MW: Statins reduce the risk of lung cancer in humans: a large case-control study of US veterans. Chest 2007, 131(5):1282-1288.

159. Dale KM, Coleman CI, Henyan NN, Kluger J, White CM: Statins and cancer risk: a meta-analysis. JAMA 2006, 295(1):74-80.

160. Bonovas S, Filioussi K, Tsavaris N, Sitaras NM: Use of statins and breast cancer: a meta-analysis of seven randomized clinical trials and nine observational studies. J Clin Oncol 2005, 23(34):8606-8612.

161. Baigent C, Blackwell L, Emberson J, Holland LE, Reith C, Bhala N, Peto R, Barnes EH, Keech A, Simes J et al: Efficacy and safety of more intensive lowering of LDL cholesterol: a meta-analysis of data from 170,000 participants in 26 randomised trials. Lancet 2010, 376(9753):1670-1681.

162. Browning DR, Martin RM: Statins and risk of cancer: a systematic review and metaanalysis. Int J Cancer 2007, 120(4):833-843.

163. McDougall JA, Malone KE, Daling JR, Cushing-Haugen KL, Porter PL, Li CI: Long- term statin use and risk of ductal and lobular breast cancer among women 55 to 74 years of age. Cancer Epidemiol Biomarkers Prev 2013, 22(9):1529-1537.

164. Kuoppala J, Lamminpaa A, Pukkala E: Statins and cancer: A systematic review and meta-analysis. Eur J Cancer 2008, 44(15):2122-2132.

165. Ahern TP, Pedersen L, Tarp M, Cronin-Fenton DP, Garne JP, Silliman RA, Sorensen HT, Lash TL: Statin prescriptions and breast cancer recurrence risk: a Danish nationwide prospective cohort study. J Natl Cancer Inst 2011, 103(19):1461-1468.

166. Boudreau DM, Yu O, Chubak J, Wirtz HS, Bowles EJ, Fujii M, Buist DS: Comparative safety of cardiovascular medication use and breast cancer outcomes among women with early stage breast cancer. Breast Cancer Res Treat 2014, 144(2):405-416.

167. Chae YK, Valsecchi ME, Kim J, Bianchi AL, Khemasuwan D, Desai A, Tester W: Reduced risk of breast cancer recurrence in patients using ACE inhibitors, ARBs, and/or statins. Cancer Invest 2011, 29(9):585-593.

168. Kwan ML, Habel LA, Flick ED, Quesenberry CP, Caan B: Post-diagnosis statin use and breast cancer recurrence in a prospective cohort study of early stage breast cancer survivors. Breast Cancer Res Treat 2008, 109(3):573-579.

169. Goard CA, Chan-Seng-Yue M, Mullen PJ, Quiroga AD, Wasylishen AR, Clendening JW, Sendorek DH, Haider S, Lehner R, Boutros PC et al: Identifying molecular features that distinguish fluvastatin-sensitive breast tumor cells. Breast Cancer Res Treat 2014, 143(2):301-312.

170. Campbell MJ, Esserman LJ, Zhou Y, Shoemaker M, Lobo M, Borman E, Baehner F, Kumar AS, Adduci K, Marx C et al: Breast cancer growth prevention by statins. Cancer Res 2006, 66(17):8707-8714.

171. Mueck AO, Seeger H, Wallwiener D: Effect of statins combined with estradiol on the proliferation of human receptor-positive and receptor-negative breast cancer cells. Menopause 2003, 10(4):332-336.

172. Garnett DJ: Caveolae as a target to quench autoinduction of the metastatic phenotype in lung cancer. J Cancer Res Clin Oncol 2016, 142(3):611-618.

173. Kang S, Kim ES, Moon A: Simvastatin and lovastatin inhibit breast cell invasion induced by H-Ras. Oncol Rep 2009, 21(5):1317-1322.

174. Kusama T, Mukai M, Tatsuta M, Nakamura H, Inoue M: Inhibition of transendothelial migration and invasion of human breast cancer cells by preventing geranylgeranylation of Rho. Int J Oncol 2006, 29(1):217-223.

175. Sorrentino G, Ruggeri N, Specchia V, Cordenonsi M, Mano M, Dupont S, Manfrin A, Ingallina E, Sommaggio R, Piazza S et al: Metabolic control of YAP and TAZ by the mevalonate pathway. Nat Cell Biol 2014, 16(4):357-366.

176. Freed-Pastor WA, Mizuno H, Zhao X, Langerod A, Moon SH, Rodriguez-Barrueco R, Barsotti A, Chicas A, Li W, Polotskaia A et al: Mutant p53 disrupts mammary tissue architecture via the mevalonate pathway. Cell 2012, 148(1-2):244-258.

177. Clendening JW, Pandyra A, Boutros PC, El Ghamrasni S, Khosravi F, Trentin GA, Martirosyan A, Hakem A, Hakem R, Jurisica I et al: Dysregulation of the mevalonate pathway promotes transformation. Proceedings of the National Academy of Sciences of the United States of America 2010, 107(34):15051-15056.

178. Duncan RE, El-Sohemy A, Archer MC: Mevalonate promotes the growth of tumors derived from human cancer cells in vivo and stimulates proliferation in vitro with enhanced cyclin-dependent kinase-2 activity. J Biol Chem 2004, 279(32):33079- 33084.

179. Boudreau DM, Yu O, Johnson J: Statin use and cancer risk: a comprehensive review. Expert Opin Drug Saf 2010, 9(4):603-621.

180. Garwood ER, Kumar AS, Baehner FL, Moore DH, Au A, Hylton N, Flowers CI, Garber J, Lesnikoski BA, Hwang ES et al: Fluvastatin reduces proliferation and increases apoptosis in women with high grade breast cancer. Breast Cancer Res Treat 2010, 119(1):137-144.

181. Bjarnadottir O, Romero Q, Bendahl PO, Jirstrom K, Ryden L, Loman N, Uhlen M, Johannesson H, Rose C, Grabau D et al: Targeting HMG-CoA reductase with statins in a window-of-opportunity breast cancer trial. Breast Cancer Res Treat 2013, 138(2):499-508.

182. Feldt M, Bjarnadottir O, Kimbung S, Jirstrom K, Bendahl PO, Veerla S, Grabau D, Hedenfalk I, Borgquist S: Statin-induced anti-proliferative effects via cyclin D1 and p27 in a window-of-opportunity breast cancer trial. J Transl Med 2015, 13:133.

183. Kimbung S, Lettiero B, Feldt M, Bosch A, Borgquist S: High expression of cholesterol biosynthesis genes is associated with resistance to statin treatment and inferior survival in breast cancer. Oncotarget 2016, 7(37):59640-59651.

184. Boroughs LK, DeBerardinis RJ: Metabolic pathways promoting cancer cell survival and growth. Nat Cell Biol 2015, 17(4):351-359.

185. Koyuturk M, Ersoz M, Altiok N: Simvastatin induces apoptosis in human breast cancer cells: p53 and estrogen receptor independent pathway requiring signalling through JNK. Cancer Lett 2007, 250(2):220-228.

186. Shibata MA, Ito Y, Morimoto J, Otsuki Y: Lovastatin inhibits tumor growth and lung metastasis in mouse mammary carcinoma model: a p53-independent mitochondrial- mediated apoptotic mechanism. Carcinogenesis 2004, 25(10):1887-1898.

187. Wong WW, Tan MM, Xia Z, Dimitroulakos J, Minden MD, Penn LZ: Cerivastatin triggers tumor-specific apoptosis with higher efficacy than lovastatin. Clin Cancer Res 2001, 7(7):2067-2075.

188. Kobayashi Y, Kashima H, Wu RC, Jung JG, Kuan JC, Gu J, Xuan J, Sokoll L, Visvanathan K, Shih Ie M et al: Mevalonate Pathway Antagonist Suppresses Formation of Serous Tubal Intraepithelial Carcinoma and Ovarian Carcinoma in Mouse Models. Clin Cancer Res 2015, 21(20):4652-4662.

189. Rao S, Lowe M, Herliczek TW, Keyomarsi K: Lovastatin mediated G1 arrest in normal and tumor breast cells is through inhibition of CDK2 activity and redistribution of p21 and p27, independent of p53. Oncogene 1998, 17(18):2393- 2402.

190. Martinez-Botas J, Ferruelo AJ, Suarez Y, Fernandez C, Gomez-Coronado D, Lasuncion MA: Dose-dependent effects of lovastatin on cell cycle progression. Distinct requirement of cholesterol and non-sterol mevalonate derivatives. Biochim Biophys Acta 2001, 1532(3):185-194.

191. Fernandez C, Martin M, Gomez-Coronado D, Lasuncion MA: Effects of distal cholesterol biosynthesis inhibitors on cell proliferation and cell cycle progression. J Lipid Res 2005, 46(5):920-929.

192. Martin Sanchez C, Perez Martin JM, Jin JS, Davalos A, Zhang W, de la Pena G, Martinez-Botas J, Rodriguez-Acebes S, Suarez Y, Hazen MJ et al: Disruption of the mevalonate pathway induces dNTP depletion and DNA damage. Biochim Biophys Acta 2015, 1851(9):1240-1253.

193. Pike LJ: The challenge of lipid rafts. J Lipid Res 2009, 50 Suppl:S323-328.

194. Mollinedo F, Gajate C: Lipid rafts as major platforms for signaling regulation in cancer. Adv Biol Regul 2015, 57:130-146.

195. Ray S, Kassan A, Busija AR, Rangamani P, Patel HH: The plasma membrane as a capacitor for energy and metabolism. Am J Physiol Cell Physiol 2015:ajpcell 00087 02015.

196. Chao FC, Efron B, Wolf P: The possible prognostic usefulness of assessing serum proteins and cholesterol in malignancy. Cancer 1975, 35(4):1223-1229.

197. Peterson C, Vitols S, Rudling M, Blomgren H, Edsmyr F, Skoog L: Hypocholesterolemia in cancer patients may be caused by elevated LDL receptor activities in malignant cells. Med Oncol Tumor Pharmacother 1985, 2(3):143-147.

198. Mokarram P, Alizadeh J, Razban V, Barazeh M, Solomon C, Kavousipour S: Interconnection of Estrogen/ Testosterone Metabolism and Mevalonate Pathway in Breast and Prostate Cancers. Curr Mol Pharmacol 2016.

199. Ko YJ, Balk SP: Targeting steroid hormone receptor pathways in the treatment of hormone dependent cancers. Curr Pharm Biotechnol 2004, 5(5):459-470.

200. Miziorko HM: Enzymes of the mevalonate pathway of isoprenoid biosynthesis. Arch Biochem Biophys 2011, 505(2):131-143.

201. Litwack MD, Peterkofsky A: Transfer ribonucleic acid deficient in N6-(delta 2- isopentenyl)adenosine due to mevalonic acid limitation. Biochemistry 1971, 10(6):994-1001.

202. Spinola M, Galvan A, Pignatiello C, Conti B, Pastorino U, Nicander B, Paroni R, Dragani TA: Identification and functional characterization of the candidate tumor suppressor gene TRIT1 in human lung cancer. Oncogene 2005, 24(35):5502-5509.

203. Tolerico LH, Benko AL, Aris JP, Stanford DR, Martin NC, Hopper AK: Saccharomyces cerevisiae Mod5p-II contains sequences antagonistic for nuclear and cytosolic locations. Genetics 1999, 151(1):57-75.

204. Clarke S, Vogel JP, Deschenes RJ, Stock J: Posttranslational modification of the Ha- ras oncogene protein: evidence for a third class of protein carboxyl

methyltransferases. Proceedings of the National Academy of Sciences of the United States of America 1988, 85(13):4643-4647.

205. Boyartchuk VL, Ashby MN, Rine J: Modulation of Ras and a-factor function by carboxyl-terminal proteolysis. Science 1997, 275(5307):1796-1800.

206. Moores SL, Schaber MD, Mosser SD, Rands E, O'Hara MB, Garsky VM, Marshall MS, Pompliano DL, Gibbs JB: Sequence dependence of protein isoprenylation. J Biol Chem 1991, 266(22):14603-14610.

207. Willumsen BM, Christensen A, Hubbert NL, Papageorge AG, Lowy DR: The p21 ras C- terminus is required for transformation and membrane association. Nature 1984, 310(5978):583-586.

208. Hart KC, Donoghue DJ: Derivatives of activated H-ras lacking C-terminal lipid modifications retain transforming ability if targeted to the correct subcellular location. Oncogene 1997, 14(8):945-953.

209. Tsubaki M, Takeda T, Sakamoto K, Shimaoka H, Fujita A, Itoh T, Imano M, Mashimo K, Fujiwara D, Sakaguchi K et al: Bisphosphonates and statins inhibit expression and secretion of MIP-1alpha via suppression of Ras/MEK/ERK/AML-1A and Ras/PI3K/Akt/AML-1A pathways. Am J Cancer Res 2015, 5(1):168-179.

210. Elsayed M, Kobayashi D, Kubota T, Matsunaga N, Murata R, Yoshizawa Y, Watanabe N, Matsuura T, Tsurudome Y, Ogino T et al: Synergistic Antiproliferative Effects of Zoledronic Acid and Fluvastatin on Human Pancreatic Cancer Cell Lines: An in Vitro Study. Biol Pharm Bull 2016, 39(8):1238-1246.

211. Rigoni M, Riganti C, Vitale C, Griggio V, Campia I, Robino M, Foglietta M, Castella B, Sciancalepore P, Buondonno I et al: Simvastatin and downstream inhibitors circumvent constitutive and stromal cell-induced resistance to doxorubicin in IGHV unmutated CLL cells. Oncotarget 2015, 6(30):29833-29846.

212. Reilly JE, Zhou X, Tong H, Kuder CH, Wiemer DF, Hohl RJ: In vitro studies in a myelogenous leukemia cell line suggest an organized binding of geranylgeranyl diphosphate synthase inhibitors. Biochem Pharmacol 2015, 96(2):83-92.

213. Moriceau G, Roelofs AJ, Brion R, Redini F, Ebetion FH, Rogers MJ, Heymann D: Synergistic inhibitory effect of apomine and lovastatin on osteosarcoma cell growth. Cancer 2012, 118(3):750-760.

214. Tsubaki M, Mashimo K, Takeda T, Kino T, Fujita A, Itoh T, Imano M, Sakaguchi K, Satou T, Nishida S: Statins inhibited the MIP-1alpha expression via inhibition of Ras/ERK and Ras/Akt pathways in myeloma cells. Biomed Pharmacother 2016, 78:23-29.

215. Hederstedt L: Heme A biosynthesis. Biochim Biophys Acta 2012, 1817(6):920-927.

216. Caughey WS, Smythe GA, O'Keeffe DH, Maskasky JE, Smith MI: Heme A of cytochrome c oxicase. Structure and properties: comparisons with hemes B, C, and S and derivatives. J Biol Chem 1975, 250(19):7602-7622.

217. Shimokata K, Katayama Y, Murayama H, Suematsu M, Tsukihara T, Muramoto K, Aoyama H, Yoshikawa S, Shimada H: The proton pumping pathway of bovine heart cytochrome c oxidase. Proceedings of the National Academy of Sciences of the United States of America 2007, 104(10):4200-4205.

218. Ernster L, Dallner G: Biochemical, physiological and medical aspects of ubiquinone function. Biochim Biophys Acta 1995, 1271(1):195-204.

219. Chojnacki T, Dallner G: The biological role of dolichol. Biochem J 1988, 251(1):1-9.

220. Carlberg M, Dricu A, Blegen H, Wang M, Hjertman M, Zickert P, Hoog A, Larsson O: Mevalonic acid is limiting for N-linked glycosylation and translocation of the insulin-like growth factor-1 receptor to the cell surface. Evidence for a new link between 3-hydroxy-3-methylglutaryl-coenzyme a reductase and cell growth. J Biol Chem 1996, 271(29):17453-17462.

221. Eklund EA, Freeze HH: The congenital disorders of glycosylation: a multifaceted group of syndromes. NeuroRx 2006, 3(2):254-263.

222. Breitling J, Aebi M: N-linked protein glycosylation in the endoplasmic reticulum. Cold Spring Harb Perspect Biol 2013, 5(8):a013359.

223. Parentini I, Cavallini G, Donati A, Gori Z, Bergamini E: Accumulation of dolichol in older tissues satisfies the proposed criteria to be qualified a biomarker of aging. J Gerontol A Biol Sci Med Sci 2005, 60(1):39-43.

224. An HJ, Froehlich JW, Lebrilla CB: Determination of glycosylation sites and site- specific heterogeneity in glycoproteins. Curr Opin Chem Biol 2009, 13(4):421-426.

225. Lau KS, Partridge EA, Grigorian A, Silvescu CI, Reinhold VN, Demetriou M, Dennis JW: Complex N-glycan number and degree of branching cooperate to regulate cell proliferation and differentiation. Cell 2007, 129(1):123-134.

226. Pinho SS, Reis CA: Glycosylation in cancer: mechanisms and clinical implications. Nat Rev Cancer 2015, 15(9):540-555.

227. Cengiz O, Kocer B, Surmeli S, Santicky MJ, Soran A: Are pretreatment serum albumin and cholesterol levels prognostic tools in patients with colorectal carcinoma? Med Sci Monit 2006, 12(6):CR240-247.

228. Vitols S, Gunven P, Gruber A, Larsson O: Expression of the low-density lipoprotein receptor, HMG-CoA reductase, and multidrug resistance (Mdr1) genes in colorectal carcinomas. Biochem Pharmacol 1996, 52(1):127-131.

229. Rentzhog L, Wikstrom S: Pharmacological effects on gastric emptying following laparotomy in the rat. Acta Chir Scand 1975, 141(7):649-653.

230. Murtola TJ, Syvala H, Pennanen P, Blauer M, Solakivi T, Ylikomi T, Tammela TL: The importance of LDL and cholesterol metabolism for prostate epithelial cell growth. PLoS One 2012, 7(6):e39445.

231. Li YC, Park MJ, Ye SK, Kim CW, Kim YN: Elevated levels of cholesterol-rich lipid rafts in cancer cells are correlated with apoptosis sensitivity induced by cholesterol- depleting agents. Am J Pathol 2006, 168(4):1107-1118; quiz 1404-1105.

232. Gao J, Aksoy BA, Dogrusoz U, Dresdner G, Gross B, Sumer SO, Sun Y, Jacobsen A, Sinha R, Larsson E et al: Integrative analysis of complex cancer genomics and clinical profiles using the cBioPortal. Sci Signal 2013, 6(269):pl1.

233. Cerami E, Gao J, Dogrusoz U, Gross BE, Sumer SO, Aksoy BA, Jacobsen A, Byrne CJ, Heuer ML, Larsson E et al: The cBio cancer genomics portal: an open platform for exploring multidimensional cancer genomics data. Cancer Discov 2012, 2(5):401-404.

234. Mollinedo F, de la Iglesia-Vicente J, Gajate C, Estella-Hermoso de Mendoza A, Villa- Pulgarin JA, Campanero MA, Blanco-Prieto MJ: Lipid raft-targeted therapy in multiple myeloma. Oncogene 2010, 29(26):3748-3757.

235. Brusselmans K, Timmermans L, Van de Sande T, Van Veldhoven PP, Guan G, Shechter I, Claessens F, Verhoeven G, Swinnen JV: Squalene synthase, a determinant of Raft- associated cholesterol and modulator of cancer cell proliferation. J Biol Chem 2007, 282(26):18777-18785.

236. Ricoult SJ, Yecies JL, Ben-Sahra I, Manning BD: Oncogenic PI3K and K-Ras stimulate de novo lipid synthesis through mTORC1 and SREBP. Oncogene 2016, 35(10):1250-1260.

237. Yamauchi Y, Furukawa K, Hamamura K: Positive feedback loop between PI3K-Akt- mTORC1 signaling and the lipogenic pathway boosts Akt signaling: induction of the lipogenic pathway by a melanoma antigen. Cancer Res 2011, 71(14):4989-4997.

238. Calvisi DF, Wang C, Ho C, Ladu S, Lee SA, Mattu S, Destefanis G, Delogu S, Zimmermann A, Ericsson J et al: Increased lipogenesis, induced by AKT-mTORC1- RPS6 signaling, promotes development of human hepatocellular carcinoma. Gastroenterology 2011, 140(3):1071-1083.

239. Li Y, Xu S, Mihaylova MM, Zheng B, Hou X, Jiang B, Park O, Luo Z, Lefai E, Shyy JY et al: AMPK phosphorylates and inhibits SREBP activity to attenuate hepatic steatosis and atherosclerosis in diet-induced insulin-resistant mice. Cell Metab 2011, 13(4):376-388.

240. Clarke PR, Hardie DG: Regulation of HMG-CoA reductase: identification of the site phosphorylated by the AMP-activated protein kinase in vitro and in intact rat liver. EMBO J 1990, 9(8):2439-2446.

241. Wang Z, Wu Y, Wang H, Zhang Y, Mei L, Fang X, Zhang X, Zhang F, Chen H, Liu Y et al: Interplay of mevalonate and Hippo pathways regulates RHAMM transcription via YAP to modulate breast cancer cell motility. Proceedings of the National Academy of Sciences of the United States of America 2014, 111(1):E89-98.

242. Pietrocola F, Galluzzi L, Bravo-San Pedro JM, Madeo F, Kroemer G: Acetyl coenzyme A: a central metabolite and second messenger. Cell Metab 2015, 21(6):805-821.

243. Schug ZT, Peck B, Jones DT, Zhang Q, Grosskurth S, Alam IS, Goodwin LM, Smethurst E, Mason S, Blyth K et al: Acetyl-CoA synthetase 2 promotes acetate utilization and maintains cancer cell growth under metabolic stress. Cancer Cell 2015, 27(1):57-71.

244. Mashimo T, Pichumani K, Vemireddy V, Hatanpaa KJ, Singh DK, Sirasanagandla S, Nannepaga S, Piccirillo SG, Kovacs Z, Foong C et al: Acetate is a bioenergetic substrate for human glioblastoma and brain metastases. Cell 2014, 159(7):1603- 1614.

245. Comerford SA, Huang Z, Du X, Wang Y, Cai L, Witkiewicz AK, Walters H, Tantawy MN, Fu A, Manning HC et al: Acetate dependence of tumors. Cell 2014, 159(7):1591- 1602.

246. Maganto-Garcia E, Tarrio ML, Grabie N, Bu DX, Lichtman AH: Dynamic changes in regulatory T cells are linked to levels of diet-induced hypercholesterolemia. Circulation 2011, 124(2):185-195.

247. Zhou X, Paulsson G, Stemme S, Hansson GK: Hypercholesterolemia is associated with a T helper (Th) 1/Th2 switch of the autoimmune response in atherosclerotic apo E- knockout mice. J Clin Invest 1998, 101(8):1717-1725.

248. Gruenbacher G, Nussbaumer O, Gander H, Steiner B, Leonhartsberger N, Thurnher M: Stress-related and homeostatic cytokines regulate Vgamma9Vdelta2 T-cell surveillance of mevalonate metabolism. Oncoimmunology 2014, 3(8):e953410.

249. Kaneko R, Tsuji N, Asanuma K, Tanabe H, Kobayashi D, Watanabe N: Survivin downregulation plays a crucial role in 3-hydroxy-3-methylglutaryl coenzyme A reductase inhibitor-induced apoptosis in cancer. J Biol Chem 2007, 282(27):19273-19281.

250. Finlay GA, Malhowski AJ, Liu Y, Fanburg BL, Kwiatkowski DJ, Toksoz D: Selective inhibition of growth of tuberous sclerosis complex 2 null cells by atorvastatin is associated with impaired Rheb and Rho GTPase function and reduced mTOR/S6 kinase activity. Cancer Res 2007, 67(20):9878-9886.

251. Araki M, Maeda M, Motojima K: Hydrophobic statins induce autophagy and cell death in human rhabdomyosarcoma cells by depleting geranylgeranyl diphosphate. Eur J Pharmacol 2012, 674(2-3):95-103.

252. Taylor-Harding B, Orsulic S, Karlan BY, Li AJ: Fluvastatin and cisplatin demonstrate synergistic cytotoxicity in epithelial ovarian cancer cells. Gynecol Oncol 2010, 119(3):549-556.

253. Kohl NE, Omer CA, Conner MW, Anthony NJ, Davide JP, deSolms SJ, Giuliani EA, Gomez RP, Graham SL, Hamilton K et al: Inhibition of farnesyltransferase induces regression of mammary and salivary carcinomas in ras transgenic mice. Nat Med 1995, 1(8):792-797.

254. Cox AD, Der CJ, Philips MR: Targeting RAS Membrane Association: Back to the Future for Anti-RAS Drug Discovery? Clin Cancer Res 2015, 21(8):1819-1827.

255. Downward J: Targeting RAS signalling pathways in cancer therapy. Nat Rev Cancer 2003, 3(1):11-22.

256. Lobell RB, Omer CA, Abrams MT, Bhimnathwala HG, Brucker MJ, Buser CA, Davide JP, deSolms SJ, Dinsmore CJ, Ellis-Hutchings MS et al: Evaluation of farnesyl:protein transferase and geranylgeranyl:protein transferase inhibitor combinations in preclinical models. Cancer Res 2001, 61(24):8758-8768.

257. Wojtkowiak JW, Fouad F, LaLonde DT, Kleinman MD, Gibbs RA, Reiners JJ, Jr., Borch RF, Mattingly RR: Induction of apoptosis in neurofibromatosis type 1 malignant peripheral nerve sheath tumor cell lines by a combination of novel farnesyl transferase inhibitors and lovastatin. J Pharmacol Exp Ther 2008, 326(1):1-11.

258. Morgan MA, Sebil T, Aydilek E, Peest D, Ganser A, Reuter CW: Combining prenylation inhibitors causes synergistic cytotoxicity, apoptosis and disruption of RAS-to-MAP kinase signalling in multiple myeloma cells. Br J Haematol 2005, 130(6):912-925.

259. Krens LL, Simkens LH, Baas JM, Koomen ER, Gelderblom H, Punt CJ, Guchelaar HJ: Statin use is not associated with improved progression free survival in cetuximab treated KRAS mutant metastatic colorectal cancer patients: results from the CAIRO2 study. PLoS One 2014, 9(11):e112201.

260. Lee JE, Baba Y, Ng K, Giovannucci E, Fuchs CS, Ogino S, Chan AT: Statin use and colorectal cancer risk according to molecular subtypes in two large prospective cohort studies. Cancer Prev Res (Phila) 2011, 4(11):1808-1815.

261. Baas JM, Krens LL, ten Tije AJ, Erdkamp F, van Wezel T, Morreau H, Gelderblom H, Guchelaar HJ: Safety and efficacy of the addition of simvastatin to cetuximab in previously treated KRAS mutant metastatic colorectal cancer patients. Invest New Drugs 2015, 33(6):1242-1247.

262. Baas JM, Krens LL, Bos MM, Portielje JE, Batman E, van Wezel T, Morreau H, Guchelaar HJ, Gelderblom H: Safety and efficacy of the addition of simvastatin to panitumumab in previously treated KRAS mutant metastatic colorectal cancer patients. Anticancer Drugs 2015, 26(8):872-877.

263. Hong JY, Nam EM, Lee J, Park JO, Lee SC, Song SY, Choi SH, Heo JS, Park SH, Lim HY et al: Randomized double-blinded, placebo-controlled phase II trial of simvastatin and gemcitabine in advanced pancreatic cancer patients. Cancer Chemother Pharmacol 2014, 73(1):125-130.

264. Ng K, Ogino S, Meyerhardt JA, Chan JA, Chan AT, Niedzwiecki D, Hollis D, Saltz LB, Mayer RJ, Benson AB, 3rd et al: Relationship between statin use and colon cancer recurrence and survival: results from CALGB 89803. J Natl Cancer Inst 2011, 103(20):1540-1551.

265. DeClue JE, Vass WC, Papageorge AG, Lowy DR, Willumsen BM: Inhibition of cell growth by lovastatin is independent of ras function. Cancer Res 1991, 51(2):712-717.

266. Wu J, Wong WW, Khosravi F, Minden MD, Penn LZ: Blocking the Raf/MEK/ERK pathway sensitizes acute myelogenous leukemia cells to lovastatin-induced apoptosis. Cancer Res 2004, 64(18):6461-6468.

267. Kanugula AK, Dhople VM, Volker U, Ummanni R, Kotamraju S: Fluvastatin mediated breast cancer cell death: a proteomic approach to identify differentially regulated proteins in MDA-MB-231 cells. PLoS One 2014, 9(9):e108890.

268. Whyte DB, Kirschmeier P, Hockenberry TN, Nunez-Oliva I, James L, Catino JJ, Bishop WR, Pai JK: K- and N-Ras are geranylgeranylated in cells treated with farnesyl protein transferase inhibitors. J Biol Chem 1997, 272(22):14459-14464.

269. Rowell CA, Kowalczyk JJ, Lewis MD, Garcia AM: Direct demonstration of geranylgeranylation and farnesylation of Ki-Ras in vivo. J Biol Chem 1997, 272(22):14093-14097.

270. Sinensky M, Beck LA, Leonard S, Evans R: Differential inhibitory effects of lovastatin on protein isoprenylation and sterol synthesis. J Biol Chem 1990, 265(32):19937- 19941.

271. Cho KJ, Hill MM, Chigurupati S, Du G, Parton RG, Hancock JF: Therapeutic levels of the hydroxmethylglutaryl-coenzyme A reductase inhibitor lovastatin activate ras signaling via phospholipase D2. Mol Cell Biol 2011, 31(6):1110-1120.

Chapter 2 Statin-induced cancer cell death can be mechanistically uncoupled from prenylation of RAS family proteins

Contributions: The work presented in this chapter was performed by the author except for Fig 2-9 and Fig 2-10, where Dr. Wail Ba-Alawi and Dr. Benjamin Haibe-Kains performed datamining, bimodal index calculations, and pharmacogenomic analyses.

This project was completed with supervisory support from Dr. Linda Z. Penn.

57 58

Statin-induced cancer cell death can be mechanistically uncoupled from prenylation of RAS family proteins 2.1 Abstract

Statins inhibit the rate-limiting enzyme of the mevalonate pathway, responsible for the de novo production of farnesyl diphosphate (FPP) and geranylgeranyl diphosphate (GGPP). FPP and

GGPP are essential substrates for post-translational prenylation of RAS family proteins, and a long-standing open question has been whether inhibition of protein prenylation of RAS family members underlies statin-induced cancer cell death. Here, we demonstrate that statin-induced cancer cell death can be uncoupled from protein prenylation of RAS family members. Increased cellular demand for GGPP and FPP for prenylation does not uniformly sensitize cells to inhibition of GGPP and FPP synthesis, as only HRAS and KRAS, but not other family members, sensitized MCF10A cells to fluvastatin-mediated cell death. However, this sensitivity was independent of RAS prenylation, but dependent on the RAS-ZEB1-EMT signaling axis.

Induction of EMT was sufficient to phenocopy the increase in fluvastatin sensitivity in RAS- transformed cells, and knocking down ZEB1 rescued RAS-transformed cells from fluvastatin- induced cell death. By mining publicly available gene expression and statin sensitivity databases, we showed that enrichment of EMT features is associated with increased sensitivity to statins in a large panel of cancer cell lines, suggesting that this association can be generalized across multiple cancer types. Taken together, these results indicate that the anti-cancer effect of statins is independent from prenylation of RAS family proteins, even in RAS-transformed cancer cells.

2.2 Introduction

Statins are inhibitors of 3-hydroxy-3-methylglutaryl coenzyme A reductase (HMGCR) (Fig 2-

1A), and have been widely prescribed to lower cholesterol levels (1). Epidemiological evidence

(2-5) and clinical trials (6-10) indicate that statins have anti-cancer activities, particularly in breast, prostate, lung, and colorectal cancers. Pre-clinical data demonstrate that statin treatment induces cancer cells to undergo apoptosis (11-14), and lowers disease burden in tumour xenografts (15-18). Despite the promising potential to repurpose statins as anti-cancer agents, the molecular mechanism of how inhibition of HMGCR can specifically kill cancer cells remains unclear.

HMGCR catalyzes the conversion of HMG-CoA to mevalonate (MVA), the sole precursor for the de novo synthesis of sterols, geranylgeranyl diphosphate (GGPP), farnesyl diphosphate

(FPP), and several other metabolic end products (19, 20) (Fig 2-1A). Statin-induced apoptosis can be rescued by co-administration with MVA (13, 14, 21), demonstrating that this is an on- target effect, or with GGPP or FPP (14, 21-25). Sterols cannot rescue cancer cell apoptosis induced by statins (14, 26). These results have led to a model where statins induce apoptosis by inhibiting GGPP and FPP synthesis.

GGPP and FPP are essential substrates for protein geranylgeranylation and farnesylation, respectively, together referred to as protein prenylation (27-29). Prenylation with the hydrophobic geranylgeranyl or farnesyl moiety localizes proteins to the cell membrane (30, 31).

Prenylation-driven membrane localization is required for all proteins in the RAS GTPase superfamily, and several groups have shown that statin treatment decreases the prenylated and membrane-associated forms of RAS, RHO, RAC, RAP, and RAB subfamily proteins (22, 32-

36). However, evidence for the functional importance of these small GTPases in conferring statin sensitivity have been conflicting. For example, cancer cells with upregulated or hyperactivated

RAS or RHO are associated with increased statin sensitivity in some studies (32-34), but not

60 others (14, 24, 37). Understanding the mechanism driving these discrepancies has remained a challenge and has been an area of debate for many years (19, 38).

In this Chapter, we directly address these discrepancies and show that inhibition of RAS family protein prenylation can be uncoupled from statin-induced apoptosis. We systematically introduced several members of the RAS superfamily in the MCF10A basal breast cell line (39-

41), producing a panel of sublines with increased demand for GGPP and/or FPP. This did not uniformly sensitize cells to inhibition of GGPP and FPP synthesis, as only HRASG12V and

KRASG12V exhibited an increased sensitivity to fluvastatin. HRASG12V and myristoylated-

HRASG12V sensitized cells to statins to a similar extent, indicating that statin-induced cell death is independent of RAS prenylation, even in RAS-transformed cells. We then showed that overexpression of RAS induced epithelial-to-mesenchyme transition (EMT) in these cells, in part by upregulating the EMT driver ZEB1. Exogenous expression or knockdown of ZEB1 conferred or rescued statin sensitivity, respectively, suggesting that EMT was the critical feature that was functionally important for statin-induced cell death. Taking a computational pharmacogenomics approach, we demonstrated that EMT was associated with statin sensitivity across a large panel of cancer cell lines. Taken together, our results provide a rationale for why RAS-related oncogenes have been poor biomarkers of statin sensitivity, and suggest that a set of EMT- associated genes should be further evaluated in the pre-clinical and clinical setting.

2.3 Results 2.3.1 HRASG12V and KRASG12V, but not other proteins in the RAS superfamily, sensitize MCF10A cells to fluvastatin

Using the MCF10A breast epithelial cell line as a non-transformed, genomically stable cell background (39-41), we ectopically expressed representative proteins from the RAS, RHO,

RAC, and RAP subfamilies in their dominantly active forms (42) (Fig 2-1B). These mutants remain dependent on prenylation for activity, allowing us to simulate the increase in demand for

FPP and/or GGPP as a result of aberrant activation of these GTPases. The increase in demand for

FPP and/or GGPP did not universally sensitize cells to fluvastatin (Fig 2-1C), as only cells

G12V G12V overexpressing HRAS or KRAS had lowered fluvastatin IC50 values (13.6 μM and 13.2

μM, respectively) compared to the vector control (22.2 μM), which indicated an increased sensitivity to fluvastatin (Fig 2-1C). A colony formation assay in soft agar was used to test the effect of fluvastatin treatment on the transformation potential of cells overexpressing HRASG12V or KRASG12V (Fig 2-1D). In this and all subsequent colony formation assays, 20 μM fluvastatin was used to allow for adequate penetrance into the soft agar. Both colony count (Fig 2-1E) and colony size (Fig 2-1F) were significantly reduced compared to the untreated control.

2.3.2 Inhibition of RAS prenylation is uncoupled from fluvastatin-induced cell death

If inhibition of RAS localization was the mechanism of fluvastatin-induced cell death, cells overexpressing myristoylated HRASG12V (myr-HRAS), which localizes to the cell membrane independently of prenylation with FPP or GGPP, should remain insensitive to fluvastatin. Fig 2-

2A shows that overexpression of HRASG12V and myristoylated HRASG12V both activated downstream signaling to a similar extent, as seen by the increase in Erk and Akt phosphorylation. While membrane-bound HRASG12V began being mislocalized into the cytoplasmic compartment after treatment with 10 μM of fluvastatin for 24 h, myr-HRASG12V remained in the membrane fraction, confirming that myristoylation occurs independent of FPP and GGPP (Fig 2-2B). EGFR, HMGCS1, and actin were used as controls for membrane- localized, cytosol-localized, and total proteins, respectively (Fig 2-2B). Unexpectedly, the

G12V fluvastatin IC50 value of cells overexpressing myr-HRAS was lowered similarly to

A Acetyl-CoA B

Acetoacetyl-CoA Protein Prenylation Activating HMGCS1 substrate mutation HMG-CoA HRAS FPP G12V Fluvastatin HMGCR KRAS Prefers FPP G12V Mevalonate (MVA) RHOA GGPP G14V RHOB FPP or GGPP G14V Isopentenyl-PP Protein prenylation RAC1 Prefers GGPP G12V RAP1A GGPP G12V Farnesyl-PP Geranylgeranyl-PP Sterols (FPP) (GGPP)

Dolichol CoQ

C D EtOH fluva E vector HRAS KRAS

 (

50 20 * *

** ** vector

Fluvastatin IC Fluvastatin fluva - + - + - + 0

HRAS F

RAC1

HRAS KRAS

RHOA RHOB

vector RAP1A FLAG ** actin KRAS *

fluva - + - + - +

Figure 2-1. HRASG12V and KRASG12V, but not other prenylated proteins, sensitize MCF10A cells to fluvastatin.

A, a simplified schematic of the MVA pathway. B, representative RAS family proteins selected for ectopic expression in MCF10A cell lines. C, ectopic expression of HRASG12V and KRASG12V sensitized MCF10A cells to fluvastatin, as assessed by MTT assay following 72 h of treatment. Bars are mean + SD, n=3. **, p<0.01 (one-way ANOVA with a Dunnett post-test, comparing all columns vs. vector control column). D-F, treatment with fluvastatin decreased colony count and colony size of RAS-driven anchorage-independent growth in soft agar. Colonies were treated with 20 μM fluvastatin 2x weekly for 18 days. Bars are mean + SD, n=4. *, p<0.05; **, p<0.01 (unpaired, two-tailed t test, comparing fluvastatin-treated vs. no treatment control).

A B C

HRAS myr-HRAS 25 M)

HRAS EtOH fluva EtOH fluva -

 20

c m c m c m c m ( *

HRAS myr vector 15 ** FLAG FLAG 10 p-Erk EGFR

Erk HMGCS1 5 Fluvastatin IC Fluvastatin p-Akt actin 0

Akt HRAS

actin vector myr-HRAS

D EtOH fluva E myr- F myr- vector HRAS vector HRAS 800 100 80

r 600

60 vecto 400 * 40 *

200 20

colony countcolony

colony size (AU) size colony HRAS

- 0 0 vector myr-HRas vector myr-HRas fluva - + - +

myr fluva - + - +

vectorpMN TWISTRas myr G 50 vector H 50 HRAS I 50 myr-HRAS 60 vector 60 TWIST ** 4060 60 40 40

3040 40 30 30 *** 40 40

20 ** ** 20 20

% pre-G1 % pre-G1 % pre-G1 % pre-G1

20 % pre-G1 20

% cell death % cell % cell death % cell

10 death % cell 10 10 % pre-G1 20 % pre-G1 20

00 0 0 0 10µM fluva - + - - - - + - - - 0 - + + + FPP+ FPP FPP EtOH Fluva MVA FPP 0 EtOH FluvaMVA MVA FPP EtOH Fluva MVA 100µM MVA EtOH-- Fluva-+ + GGPPGGPP-- -- EtOH- Fluva+ + GGPP- GGPP- GGPP FPP 10µM GGPP EtOH- + MVA- + - - + MVA- + FPP- - Fluva- - GGPP+ - EtOH Fluva GGPP 10µM FPP -- -+SNAIL-- -- ++ - +SLUG- - +

60 SNAIL 60 SLUG Figure 2-602. Inhibition of RAS prenylation60 is uncoupled from fluvastatin-induced cell death. 40 40

A, overexpression40 of HRASG12V and40 myr-HRASG12V activated Erk phosphorylation and Akt % pre-G1 20 % pre-G1 20

phosphorylation to a similar extent. B, the proportion of HRASG12V in the cytoplasmic (c) % pre-G1 % pre-G1 20 20 fraction was0 increased, and the proportion0 in the membrane (m) proportion was decreased, after 10µM fluva - + - - - - + - - - G12V 0 MVA FPP 0 MVA FPP 24100µM h of MVAtreatmentEtOH- Fluva +with+ 10GGPP- μM- fluvastatin.EtOH- FluvaIn+ contrast,+ GGPP- the- localization of myr-HRAS was - + MVA- + FPP- - + MVA- + FPP- 10µM GGPP EtOH Fluva GGPP EtOH Fluva GGPPG12V G12V unaffected10µM FPP by- fluvastatin+ - -treatment.+ C, both- +HRAS- - and+ myr-HRAS sensitized MCF10As

64 to fluvastatin as assessed by MTT assay following 72 h of treatment. Bars are mean + SD, n=3. *, p<0.05; **, p<0.01 (one-way ANOVA with a Dunnett post-test, comparing all columns vs. vector control column). D-F, treatment with fluvastatin decreased colony count and colony size of myr-HRASG12V-driven anchorage-independent growth in soft agar. Colonies were treated with 20 μM fluvastatin 2x weekly for 18 days. Bars are mean + SD, n=4. *, p<0.05 (unpaired, two- tailed t test, comparing fluvastatin-treated vs. no treatment control). G-I, 10 μM fluvastatin induced cell death in MCF10A cells overexpressing HRASG12V and myr-HRASG12V, fully reversed by co-administration with MVA, GGPP, or FPP. Bars are mean + SD, n=3. **, p<0.01; ***, p<0.001 (one-way ANOVA with a Dunnett post-test, comparing all columns vs. no treatment control column).

HRASG12V (Fig 2-2C). Colony formation was also inhibited by fluvastatin treatment in cells overexpressing myr-HRASG12V (Fig 2-2D-F), to the same extent as cells overexpressing

HRASG12V (Fig 2-1D-F). Addition of MVA, GGPP, or FPP rescued the fluvastatin-induced cell death in both HRASG12V and myr-HRASG12V cells (Fig 2-2G-I). Together, these data uncouple statin-induced cell death from GGPP and FPP demand for prenylation of RAS family proteins, and implicate that events downstream of RAS signaling are responsible for the increase in statin sensitivity in this cell system.

Overexpression of HRASG12V or myr-HRASG12V in MCF10A cells did not affect the expression of MVA pathway genes HMGCR and HMGCS1, either basally or in response to fluvastatin exposure (Fig 2-3A-B). This rules out deregulation of the MVA pathway, or its sterol feedback response, as the mechanism of increased statin sensitivity in this model.

2.3.3 The RAS-ZEB1-EMT signaling axis underlies increased sensitivity to fluvastatin

Since RAS signaling, rather than RAS localization, was implicated as the driver of fluvastatin sensitivity, we overexpressed constitutively active forms of several classic effectors of RAS signaling, namely RALAG23V, BRAFV600E, PI3K-p110αE545K, PI3K-p110αH1047R, and MYCT58A, in MCF10As (Fig 2-4A-B). None of these sublines exhibited a lowered fluvastatin IC50; overexpression of PIK3CA-p110α led to significant increases in IC50 values (Fig 2-4C). Similar results were found in colony formation assays in soft agar (Fig 2-5A-C). Therefore, the observed increase in fluvastatin sensitivity in RAS-transformed cells was not mediated through these downstream mediators of RAS signaling.

The RAS sublines were phenotypically distinct from the MCF10A parental cells. Instead of an epithelial phenotype with a cobblestone-like appearance, the RAS sublines appeared more

myr- myr- A vector HRAS HRAS B vector HRAS HRAS 0.15 ** * * * 0.10 * *

0.05

relative HMGCS1 mRNA HMGCS1 relative 0.00 10μM fluva - + - + - + 10μM fluva - + - + - + pMN HRas

myr-HRas HMGCR HMGCS1 actin actin

Figure 2-3. Overexpression of HRASG12V or myr-HRASG12V does not alter HMGCR or HMGCS1 expression.

A-B, Cells were treated with 10 μM fluvastatin for 16 h and analyzed for HMGCR and HMGCS1, mRNA and protein expression. Bars are mean + SD, n=3. *, p<0.05; **, p<0.01 (unpaired, two-tailed t test, comparing fluvastatin-treated vs. no treatment control).

A B C

40 *** BRAF

vector *** RALA

Protein Activating vector M)

mutation BRAF  30

RALA ( 50 RALA G23V tubulin tubulin 20

BRAF V600E

EK HR

- - α

p110α E545K α 10

Fluvastatin IC Fluvastatin

vector MYC

p110 p110 H1047R vector 0

MYC T58A p110α MYC

-EK

-HR

MYC



RALA BRAF

actin actin vector p110 p110 Figure 2-4. Fluvastatin cytotoxicity is not mediated by the BRAF, RALA, PI3K, or MYC downstream mediators of RAS signaling.

A-B, ectopic expression of downstream mediators of RAS signaling. C, none of the oncoproteins indicated in A-B sensitized MCF10A cells to fluvastatin; p110α overexpression desensitized MCF10As to fluvastatin. Bars are mean + SD, n=3. ***, p<0.001 (one-way ANOVA with a Dunnett post-test, comparing all columns vs. vector control column).

A EtOH fluva B p100α p100α 400 vector EK HR MYC ns

300 ns vector 200 ns

* * EK - 100 ns

colonycount * α

0 p110 fluva - + - + - + - + MYC

vector HR

- C 15 p110a-EK p110a-HR α

p110 10

MYC colony size (AU) size colony

0 fluva - + - + - + - + -EK  MYC vector p110 p110a-HR Figure 2-5. Transforming oncogenes do not all sensitize MCF10As to fluvastatin.

A-C, Cells with ectopic expression of PI3K-p110αE545K, PI3K-p110αH1047R, and MYCT58A were assayed for colony formation in soft agar. Without fluvastatin treatment, all three oncogenes transformed MCF10As as assayed by soft agar colony formation assay. Bars are mean + SD, n=3. *, p<0.05 (one-way ANOVA with a Dunnett post-test, comparing all columns vs. vector control column). Fluvastatin treatment had no significant effect on the colony size or count. Colonies were treated with 20 μM fluvastatin 2x weekly for 18 days. Bars are mean + SD, n=3. ns, not significant (unpaired, two-tailed t test, comparing fluvastatin-treated vs. no treatment control).

69 mesenchymal, with an elongated and spindle-shaped morphology (Fig 2-6A). These cells had undergone epithelial-to-mesenchymal transition (EMT), with a dramatic loss of E-cadherin expression and a gain of vimentin expression (Fig 2-6B). To test whether induction of EMT confers sensitivity to fluvastatin, we treated MCF10A cells with transforming growth factor beta

(TGF-β) for 3 days, which induces EMT independently of RAS status (Fig 2-6C). This led to an increased sensitivity to fluvastatin, as indicated by a decrease in fluvastatin IC50 in TGF-β-treated cells (Fig 2-6D). After TGF-β treatment, removal of TGF-β from the culture media gradually reverses EMT (Fig 2-7). MCF10A cells fully reverted back to an epithelial phenotype after 7 days of TGF-β removal, and sensitivity to fluvastatin was restored to control levels (Fig 2-7).

These data indicate that a mesenchymal state is sufficient to confer sensitivity to fluvastatin.

RAS induces EMT by upregulating the EMT-driving transcription factor ZEB1 through several potential mechanisms (43, 44), and not by other EMT regulators such as SNAIL or TWIST (45,

46) in our MCF10A system (Fig 2-6E). ZEB1-overexpression in MCF10As led to a loss of E- cadherin and gain of vimentin expression independent of RAS status (Fig 2-6F), and decreased the fluvastatin IC50 value similarly to the RAS sublines (Fig 2-6G). ZEB1 overexpression had no effect on the IC50 value of GGTI-298 (Fig 2-6H), a specific inhibitor to geranylgeranyltransferase I, reinforcing the model that prenylation of RAS family proteins is uncoupled from fluvastatin-induced tumour cell death. We then knocked down ZEB1 in cells overexpressing HRASG12V and myr-HRASG12V using two independent shRNAs (Fig 2-8A).

ZEB1 knockdown could partially rescue the RAS-driven fluvastatin sensitivity, both by IC50 measurements (Fig 2-8B) and by colony formation in soft agar (Fig 2-8C-E).

A vector HRAS B

HRAS

KRAS myr

Vector HRAS

FLAG

KRAS myr-HRAS E-cadherin

Vimentin Tubulin

C D E TWIST SNAIL ZEB1 6 TGF-β - + 25

** M)

 20 **

E-Cadherin ( 4 50 15 * vimentin tubulin 10 2

5 Fluvastatin IC Fluvastatin 0 expression mRNA relative 0

TGF-β - + 

TGF

CTRL

HRAS HRAS HRAS

vector vector vector

myr-HRAS myr-HRAS myr-HRAS

F G 30 H 60

ector ector

ZEB1



 (

ZEB1 (

50 20 40 50 E-Cadherin ** 10 20

vimentin

GGTI-298 IC GGTI-298 Fluvastatin IC Fluvastatin

tubulin 0 0

ZEB1 ZEB1 vector vector Figure 2-6. RAS induces EMT through ZEB1, and induction of EMT is sufficient for sensitizing cells to fluvastatin.

A, RAS-overexpressing cells appear more mesenchymal than the vector control cells. B, RAS overexpression reduced expression of E-cadherin, an epithelial cell marker, and increased

71 expression of vimentin, a mesenchymal cell marker. C, treating MCF10A cells with 5 ng/mL TGF-β for 3 days induced EMT. D, induction of EMT by 5 ng/mL TGF-β treatment sensitized MCF10A cells to fluvastatin, as assessed by MTT assay following 72 h of treatment. Dashed line G12V represents IC50 of HRAS -overexpressing cells. Bars are mean + SD, n=3. *, p<0.05 (unpaired, two-tailed t test, comparing TGF-β-treated vs. no treatment control). E, HRASG12V and myr-HRASG12V cells upregulate the EMT transcription factor ZEB1. Bars are mean + SD, n=3. **, p<0.01 (one-way ANOVA with a Dunnett post-test, comparing all columns vs. vector control column). F, ectopic expression of ZEB1 induced EMT. G, ZEB1 overexpression sensitized MCF10A cells to fluvastatin, as assessed by MTT assay following 72 h of treatment. G12V Dashed line represents IC50 of HRAS -overexpressing cells. Bars are mean + SD, n=3. **, p<0.01 (unpaired, two-tailed t test, comparing TGF-β-treated vs. no treatment control). H, ZEB1 overexpression did not sensitize cells to GGTI-298. Bars are mean + SD, n=3.

20 M)

 15 **

( *** 50

5 Fluvastatin IC Fluvastatin 0 TGF-β on - 3d 3d 3d TGF-β off - - 3d 7d

E-Cadherin untreated TGFb3don

vimentin TGFb3don 7d off tubulin TGFb3don 3d off

Figure 2-7. TGF-β induced EMT and fluvastatin sensitivity are both reversible.

After 3 days of treatment with 5 ng/mL TGF-β, MCF10A cells undergo EMT and become more sensitive to fluvastatin, as assessed by MTT assays for 72 h. Dashed line represents IC50 of HRASG12V-overexpressing cells. Cells gradually reverted to epithelial with continued culturing after TGF-β removal; after 7 days, sensitivity to fluvastatin was restored to control levels. Bars are mean + SD, n=3. **, p<0.01; ***, p<0.001 (one-way ANOVA with a Dunnett post-test, comparing all columns vs. no treatment control column).

vector HRAS myr-HRAS A vector HRAS myr-HRAS B

shRNA - A B - A B - A B ** * * ZEB1 **

actin

shRNA - A B - A B - A B

C sh neg sh neg sh A sh A sh B sh B

EtOH fluva EtOH fluva EtOH fluva

vector

HRAS

- myr

D 8 vector HRAS myr-HRAS E 4 vector HRAS myr-HRAS

6 * 3

4 2 *

2 * 1

relative colony area colony relative

with fluva treatment fluva with

with fluva treatment fluva with relative colony count colony relative 0 0 shRNA - A B - A B - A B shRNA - A B - A B - A B scr A2 B7 scr A2 B7 scr A2 B7 scr A2 B7 scr A2 B7 scr A2 B7

Figure 2-8. shRNA knockdown of ZEB1 rescues the increased sensitivity in HRASG12V and myr-HRASG12V cells.

A, two independent shRNAs were used to knockdown ZEB1 in cells overexpressing HRASG12V G12V and myr-HRAS . B, knockdown of ZEB1 rescued the decreased fluvastatin IC50 observed in

HRASG12V and myr-HRASG12V cells. Bars are mean + SD, n=3. *, p<0.05; **, p<0.01 (one-way ANOVA with a Dunnett post-test, comparing all columns vs. shRNA control column). C-E, knockdown of ZEB1 in HRASG12V and myr-HRASG12V cells led to increased colony formation in soft agar under fluvastatin treatment. Colonies were treated with 20 μM fluvastatin 2x weekly for 18 days. Bars are mean + SD, n=4. *, p<0.05 (one-way ANOVA with a Dunnett post-test, comparing all columns vs. shRNA control column).

2.3.4 Enrichment of EMT phenotype is associated with sensitivity to statins in a large panel of cancer cell lines

The Cancer Cell Line Encyclopedia (CCLE) database (47) contains RNA-seq data on 927 cancer cell lines across >20 cancer types (Fig 2-9A). Mining this large database, we observed that while the expression pattern for most genes follow a unimodal (Gaussian) distribution, as exemplified by POLR2A (RNA polymerase II, subunit A; Fig 2-9B), some genes are bimodally expressed, such as ESR1 (estrogen receptor α; Fig 2-9C-D). ESR1 is known for a strong bimodal expression in the breast tissue (48-50), representing ERα-low (left peak) and ERα-high (right peak) cell lines (Fig 2-9D). We then examined the expression profile of 11 well-characterized EMT- associated genes, and found that their expression were strongly bimodal, with bimodality indices

(50, 51) higher than that of ESR1, our positive control (Fig 2-10A). The distribution of the top five bimodally expressed genes (VIM, CDH1, ZEB1, FN1, and CDH2) is shown in Fig 2-10B.

For each gene, we computed the cutoff optimally discriminating between the two modes of the expression distribution, and classified the cell lines with either low or high expression of the gene of interest (Fig 2-10B). Each cell line was therefore characterized by a binary vector representing the activation of the top five bimodally expressed EMT-associated genes.

Using the top five bimodally expressed EMT-associated genes as features, we built a binary classification rule that classified each solid tumour cell line as enriched with an EMT phenotype if at least one of the EMT-associated genes were activated (Fig 2-10B, right peak for VIM, ZEB1,

FN1, and CDH2; left peak for CDH1). We then mined the Cancer Therapeutics Response Portal version 2 (CTRPv2) database (52-54) using the PharmacoGx R/Bioconductor package (55), and found that the EMT-enriched cell lines were associated with significantly higher AUC (more sensitive) to all three statin family members that have been assessed in CTRPv2: fluvastatin (Fig

2-10C), lovastatin (Fig 2-10D), and simvastatin (Fig 2-10E). Thus, cancer cell EMT is associated

A CCLE database: total 927 cell lines

POLR2A, ESR1,

B RNA polymerase II subunit A C estrogen receptor α

Frequency Frequency

of expression of of expression of Gene expression Gene expression log2(FPKM+1) log2(FPKM+1)

D ESR1, estrogen receptor α breast tissue only (57 cell lines) ERα-low cell lines ERα-high cell lines

Frequency Frequency Cutoff of expression of Gene expression log (FPKM+1) 2 Figure 2-9. Unimodal and bimodal gene expression in the CCLE database.

A, tissue type distribution across the Cancer Cell Line Encyclopedia (CCLE) database. B, unimodal (Gaussian) distribution of POLR2A. C-D, bimodal distribution of ESR1. Gene expression values are log2(FPKM+1) with FPKM = Fragments Per Kilobase of transcript per Million mapped reads. The script and data used for the generation of these figures can be downloaded at https://github.com/bhklab/StatinEMT.

A B Gene Bimodality VIM, vimentin Protein name name index Low-expression cell lines High-expression cell lines RNA polymerase II, POLR2A 0.09 Cutoff

subunit A

Frequency Frequency of expression of

controls ESR1 Estrogen receptor α 0.6 VIM Vimentin 1.9 CDH1, E-cadherin ZEB1, ZEB1 CDH1 E-Cadherin 1.7 ZEB1 ZEB1/TCF8 1.6

FN1 Fibronectin 1.5

Frequency Frequency

Frequency Frequency of expression of CDH2 N-Cadherin 1.5 expression of

associated genes associated HMGA2 HMGA2 1.3 - FN1, fibronectin CDH2, N-cadherin ZEB2 ZEB2/SIP1 1.3 SNAI2 SNAIL2/SLUG 1.1

Known EMT Known TWIST1 TWIST1 1.1

Frequency Frequency

Frequency Frequency of expression of SNAI1 SNAIL1 1.0 expression of Gene expression Gene expression FOXC2 FOXC2 0.7 log (FPKM+1) log2(FPKM+1) 2

C fluvastatin D lovastatin E simvastatin 60 60 60 40 40

AUC (%) AUC AUC (%) AUC 20 (%) AUC 20 20

0 0 0 Enrichment - + Enrichment - + Enrichment - + for EMT for EMT for EMT

Figure 2-10. Statin sensitivity is associated with cancer cell EMT.

A, bimodality index of 11 known EMT-associated genes. B, bimodal expression of the five EMT-associated genes used to classify cell lines based on EMT activity. Gene expression values are log2(FPKM+1) with FPKM = Fragments Per Kilobase of transcript per Million mapped reads. C-E, sensitivity to three statin family members are significantly associated with cell lines enriched for EMT features. AUC, area under the curve (higher values represent higher drug sensitivity); CI, Concordance Index; P-value, Wilcoxon rank sum test, comparing statin response on EMT ‘enriched’ vs. ‘not-enriched’ cell lines; n, number of cell lines. The script and data used for the generation of these figures can be downloaded at https://github.com/bhklab/StatinEMT.

78 with sensitivity to multiple statin family members in a large panel of cancer cell lines, across multiple cancer types. It should be noted that most, but not all, cell lines in the CTRPv2 database were evaluated for sensitivity to all three statin members, which is why the number of cell lines

(n) is not identical between Fig 2-10C, D, and E.

2.4 Discussion

Previously, the increased cancer cell sensitivity to statins was thought to be mediated by inhibiting prenylation of proteins in the RAS superfamily. This model was built on three observations: i) statins inhibit prenylation of RAS family proteins (14, 21, 22, 32-37); ii) co- administration of GGPP or FPP with statins reverses the effect on protein prenylation (14, 21, 23,

32-34, 36, 37); and iii) co-administration of GGPP or FPP can rescue statin kill (14, 23-25, 33,

34, 36). However, most epidemiological studies and clinical trials do not support an association between response to statins and RAS mutations (56-61). Additionally, in cell lines that were sensitive to statins, rescuing RAS localization (14, 62) or RAF-MEK-ERK signaling (63) did not decrease statin sensitivity, and intrinsic sensitivity to statin kill was largely independent of RAS function (14, 64, 65). These contradicting data raise the possibility that inhibition of RAS family protein prenylation is not the sole contributor to statin sensitivity, implicating not only an alternative mechanism of statin-induced apoptosis, but also the potential to develop better biomarkers for the identification of patients that will benefit from statin treatment.

We show here that statin-induced cell death can indeed be uncoupled from inhibition of RAS family protein prenylation. First, increased cellular demand for GGPP and FPP for ectopic expression of RAS family proteins requiring prenylation for activity did not always sensitize

MCF10A cells to fluvastatin (Fig 2-1). Although RAS overexpression led to an increased sensitivity to fluvastatin, this effect was independent from RAS prenylation and localization to

79 the cell membrane (Fig 2-2). Instead, RAS induced EMT by upregulating ZEB1, which was the underlying cause of increased sensitivity to fluvastatin-induced cell death (Fig 2-6, Fig 2-8). This can, in part, explain why statins have been reported to be more effective in more aggressive

(invasive/metastatic) subtypes of breast (64, 66) and prostate cancers (67, 68), while mutations in

RAS family proteins are poorly associated with statin response (56-61).

Interestingly, when we assayed for the expression profile of a list of 11 EMT-associated genes known for a strong association with EMT (45, 69), we found that they followed a bimodal distribution (Fig 2-10). Bimodal distribution of VIM, CDH1, ZEB1, FN1, or CDH2 were used as features to classify all solid tumour cell lines in the Cancer Cell Line Encyclopedia (47) into high-expression and low-expression populations (Fig 5B). As a comparison, ESR1 (estrogen receptor α) is known to be bimodally expressed in breast cancer (48-50), which could be used as a classifier of estrogen receptor-positive and -negative breast tumours in large bioinformatics databases (50) (Fig 2-9D). Mining the Cancer Therapeutics Response Portal (52-54), we interrogated sensitivity to three statin family members in >500 solid tumour cell lines for any association with an EMT phenotype. Cell lines enriched with EMT features were associated with significantly higher AUC in response to statin treatment (Fig 2-10C-E), indicating that they were more sensitive to statin kill. We have thus expanded our original observation in the MCF10A model system to include >20 cancer types and three statins, suggesting that the association between EMT and sensitivity to statins can be generalized across all solid tumour cell lines.

Activation of EMT is proposed to be the critical initiating step in metastatic dissemination of late-stage cancers (70, 71). Although it is still debated whether this process is required for metastasis, as opposed to being a phenotype of aggressive/metastatic disease (72, 73), it is nevertheless known to be associated with cell de-differentiation, stem-like properties, and anti-

80 apoptotic signaling (74, 75). Importantly, activation of EMT is typically associated with therapeutic resistance (72-75). We show here that activation of EMT increased cell sensitivity to fluvastatin kill (Fig 2-6, Fig 2-7), consistent with previous reports (76-78). This suggests the intriguing possibility that statins may be used to target disseminated and/or dormant cancer cells

(that is, those that presumably have undergone EMT) that are responsible for therapeutic failure and refractory disease. Several epidemiological studies have reported supporting evidence, showing that statin use following front-line treatment was associated with better disease-free survival and overall survival in breast (4, 18, 79-82), prostate (83, 84), and renal cell carcinomas

(85). Further testing of statins as adjuvant therapeutics in the pre-clinical and clinical setting is warranted.

It is perhaps tempting to ask which prenylated protein(s), other than the ones selected for testing in this study, is/are responsible for the anti-cancer effects of statins in the context of EMT.

Indeed, we show that co-administration of GGPP or FPP could rescue fluvastatin kill in both

HRASG12V- and myr-HRASG12V-overexpressing cells (Fig 2-2G-I). However, the two isoprenoids did not rescue to the same extent: GGPP and MVA completely rescued cell death to control levels, but FPP was less effective (Fig 2-2G-I). This is reminiscent of previous studies, where GGPP consistently rescued statin effects (14, 21, 23-25, 33, 34, 36), while FPP did so less consistently, rescuing completely (21), partially (14, 23, 24, 33), or not at all (25, 33, 37).

Nonetheless, the observation that FPP did not rescue as well as GGPP in an HRASG12V- overexpressing system was unexpected, since HRAS prefers FPP over GGPP for prenylation (42,

86). Two explanations are possible for this observation. First, since FPP also acts as the precursor to sterols (19, 20), a portion of the supplemented FPP could be shunted towards cholesterol production, which does not play a role in statin-induced apoptosis (14, 26). However, the observation that MVA consistently rescues statin-induced cell death (13, 14, 21) (Fig 2-2G-I)

81 argues against this interpretation. Alternatively, GGPP also functions as the precursor for other isoprenoids such as dolichols and isoprenoid moieties on coenzyme Q (19, 20), and depletion of these larger isoprenoids could be contributing to statin sensitivity. Consistent with this is the observation that cells overexpressing ZEB1 were more sensitive to fluvastatin, but not to inhibition of geranylgeranylation itself (Fig 2-6G-H). Taken together, our data reinforce the new model presented here that, in the context of cancer cell EMT, fluvastatin-induced cell death is uncoupled from inhibition of RAS family protein prenylation. Why cells become dependent on the MVA pathway when undergoing EMT, and are therefore sensitive to fluvastatin inhibition, remains to be elucidated and will be an interesting area for future investigation that could lead to the identification of biomarkers of fluvastatin-responsive cancers.

2.5 Materials and Methods 2.5.1 Reagents

Fluvastatin was purchased from US Biologicals. TGF-β was purchased from PeproTech. All other chemicals were purchased from Sigma unless otherwise specified.

2.5.2 Cell culture

MCF10A cells were a kind gift of Dr. Senthil Muthuswamy and were cultured as previously described (87). Transgene expression was stably introduced into MCF10A cells using retroviral insertion with pBabePuro (88). Cells were imaged on the Leica MZ FLIII Stereomicroscope.

2.5.3 MTT assay

3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assays were performed as previously described (64). Briefly, MCF10A cells were seeded at 750 cells/well in 96-well plates overnight, then treated in triplicate with 0-200 μM fluvastatin for 72 h. IC50 values were computed using GraphPad Prism with a bottom constraint equal to 0.

2.5.4 Immunoblotting

Cell lysates were prepared by lysing directly in boiling SDS lysis buffer (1% SDS, 11% glycerol,

10% β-mercaptoethanol, 0.1 M Tris pH 6.8). The following antibodies were used: E-Cadherin

(CST 3195), vimentin (CST 5741), actin (Sigma A2066), tubulin (Millipore CP06), FLAG

(Sigma F1804), EGFR (CST 2232), ERK (CST 4695), p-ERK (CST 4370), AKT (CST 9272), p-

AKT (CST 9271), HMGCS1 (SCB sc-32423), RALA (BD 610221), BRAF (Sigma

HPA001328), MYC (9E10 in-house), ZEB1 (Sigma HPA027524), HMGCR (A9 in-house).

2.5.5 Soft agar colony formation

Anchorage-independent colony growth of MCF10A sublines in soft agar was evaluated using standard protocols (87). Colonies were imaged at 1.2x magnification on the Leica MZ FLIII

Stereomicroscope after 18 days. Colony number and size were quantified using ImageJ.

2.5.6 Membrane fractionation

Cell were seeded at 2x106/plate overnight and treated as indicated. Harvested cells were re- suspended in 1 mL HEPES buffer (0.25 M sucrose, 50 mM HEPES pH7.5, 5 mM NaF, 5 mM

EDTA, 2 mM DTT) and lysed by sonication. Homogenate was cleared at 2,000 xg for 20 min at

4oC, then ultracentrifuged at 115,000 xg for 70 min at 4 oC for membrane fractionation.

Membrane protein pellet was re-suspended in Triton buffer (1% Triton X-114, 50 mM Tris pH

7.5, 0.1 mM NaCl, 5 mM EDTA, 5 mM NaF, 2 mM DTT).

2.5.7 Cell death assay

Cells were seeded at 2.5x105/plate overnight and treated as indicated. After 72 h, cells were fixed in 70% ethanol for >24 h, stained with propidium iodide, and analyzed by flow cytometry for the

% sub-diploid DNA population as % cell death, as previously described (64).

2.5.8 qRT-PCR

Total RNA was harvested from subconfluent cells using TRIzol Reagent (Invitrogen). cDNA was synthesized from 500 ng of RNA using SuperScript III (Invitrogen). Real-time quantitative

RT-PCR was performed using TaqMan probes for HMGCR (ABI Hs00168352), HMGCS1 (ABI

Hs00266810), and GAPDH (ABI Hs99999905), and using SYBR Green for:

TWIST_fw 5'-CCGGAGACCTAGATGTCATTG-3'

TWIST_rv 5'-CCACGCCCTGTTTCTTTG-3'

SNAIL_fw 5'-CACTATGCCGCGCTCTTTC-3'

SNAIL_rv 3'-GGTCGTAGGGCTGCTGGAA-3'

ZEB1_fw 5'-GCCAATAAGCAAACGATTCTG-3'

ZEB1_rv 5'-TTTGGCTGGATCACTTTCAAG-3'

18S_rRNA_fw 5’-GTAACCCGTTGAACCCCATT-3’

18S_rRNA_fw 3’-CCATCCAATCGGTAGTAGCG-3’

2.5.9 Small hairpin RNA-mediated gene silencing

Small hairpin RNA were cloned into single-vector inducible shRNA construct pLKO-TetON, and lentiviruses generated as previously described (89). Expression of shRNAs were induced where indicated using 100 ng/mL doxycycline. Sense strand of shRNAs cloned were: shZeb1 A 5’-

CCGGTGTCTCCCATAAGTATCAATTCTCGAGAATTGATACTTATGGGAGACATTTTT

G-3’

84 shZeb1 B 5’-

CCGGCCTACCACTGGATGTAGTAAACTCGAGTTTACTACATCCAGTGGTAGGTTTTT

G-3’

2.5.10 Pharmacogenomic analysis

RNA-seq and drug sensitivity data were retrieved and curated from the Cancer Cell Line

Encyclopedia (CCLE) (47) and the Cancer Therapeutics Portal version 2 (CTRPv2) (52-54) databases, and were mined using the R/Bioconductor PharmacoGx package (55). To calculate the bimodality index (50, 51), a mixture of two Gaussian models was used to fit the RNA-seq expression values of each gene across all cell lines in CCLE, as implemented in the bimod function in the R/Bioconductor genefu package (version 2.6.0) (50). The cutoff was calculated by taking the average of the end-points where the two Gaussian models meet and was used to classify cell lines into either showing low- or high-expression of a gene. A binary classification rule was developed to determine whether a cell line is ‘enriched’ with EMT phenotype or not.

Cell lines from tumours of hematopoietic and lymphoid tissues, and those with unknown origin, were excluded from this analysis. If expression of any of VIM, ZEB1, FN1, or CDH2 was high, or if expression of CDH1 was low, in a cell line according to the bimodality cutoff, then it was classified as ‘enriched’ with EMT phenotype. Concordance index (CI) and p-value were calculated to measure the association between statin sensitivity, obtained from CTRPv2 database, and cell lines that were classified as either ‘enriched’ with EMT phenotype or ‘not enriched’. p-value was calculated using the non-parametric Wilcoxon rank sum test, comparing statins response on cell lines ‘enriched’ vs. ‘not enriched’ with EMT phenotype. The script and data used for the generation of these figures can be downloaded at https://github.com/bhklab/StatinEMT.

2.6 References

1. Goldstein JL, Brown MS: A Century of Cholesterol and Coronaries: From Plaques to Genes to Statins. Cell 2015, 161(1):161-172.

2. Poynter JN, Gruber SB, Higgins PD, Almog R, Bonner JD, Rennert HS, Low M, Greenson JK, Rennert G: Statins and the risk of colorectal cancer. The New England journal of medicine 2005, 352(21):2184-2192.

3. Nielsen SF, Nordestgaard BG, Bojesen SE: Statin use and reduced cancer-related mortality. The New England journal of medicine 2012, 367(19):1792-1802.

4. Ahern TP, Pedersen L, Tarp M: Statin prescriptions and breast cancer recurrence risk: a Danish nationwide prospective cohort study. J Natl Cancer Inst 2011, 103(19):1461-1468.

5. Hamilton RJ, Banez LL, Aronson WJ, Terris MK, Platz EA, Kane CJ, Presti JC, Jr., Amling CL, Freedland SJ: Statin medication use and the risk of biochemical recurrence after radical prostatectomy: results from the Shared Equal Access Regional Cancer Hospital (SEARCH) Database. Cancer 2010, 116(14):3389-3398.

6. Bjarnadottir O, Romero Q, Bendahl P-OO, Jirström K, Rydén L, Loman N, Uhlén M, Johannesson H, Rose C, Grabau D et al: Targeting HMG-CoA reductase with statins in a window-of-opportunity breast cancer trial. Breast cancer research and treatment 2013, 138(2):499-508.

7. Garwood ER, Kumar AS, Baehner FL, Moore DH, Au A, Hylton N, Flowers CI, Garber J, Lesnikoski B-AA, Hwang ES et al: Fluvastatin reduces proliferation and increases apoptosis in women with high grade breast cancer. Breast cancer research and treatment 2010, 119(1):137-144.

8. Lin JJ, Ezer N, Sigel K, Mhango G, Wisnivesky JP: The effect of statins on survival in patients with stage IV lung cancer. Lung Cancer 2016, 99:137-142.

9. Fiala O, Pesek M, Finek J, Minarik M, Benesova L, Bortlicek Z, Topolcan O: Statins augment efficacy of EGFR-TKIs in patients with advanced-stage non-small cell lung cancer harbouring KRAS mutation. Tumour Biol 2015, 36(8):5801-5805.

10. Lee J, Hong YS, Hong JY, Han SW, Kim TW, Kang HJ, Kim TY, Kim KP, Kim SH, Do IG et al: Effect of simvastatin plus cetuximab/irinotecan for KRAS mutant colorectal cancer and predictive value of the RAS signature for treatment response to cetuximab. Invest New Drugs 2014, 32(3):535-541.

11. Koyuturk M, Ersoz M, Altiok N: Simvastatin induces apoptosis in human breast cancer cells: p53 and estrogen receptor independent pathway requiring signalling through JNK. Cancer Lett 2007, 250(2):220-228.

12. Shibata MA, Ito Y, Morimoto J, Otsuki Y: Lovastatin inhibits tumor growth and lung metastasis in mouse mammary carcinoma model: a p53-independent mitochondrial- mediated apoptotic mechanism. Carcinogenesis 2004, 25(10):1887-1898.

13. Wong WW, Tan MM, Xia Z, Dimitroulakos J, Minden MD, Penn LZ: Cerivastatin triggers tumor-specific apoptosis with higher efficacy than lovastatin. Clin Cancer Res 2001, 7(7):2067-2075.

14. Wong WW, Clendening JW, Martirosyan A, Boutros PC, Bros C, Khosravi F, Jurisica I, Stewart AK, Bergsagel PL, Penn LZ: Determinants of sensitivity to lovastatin-induced apoptosis in multiple myeloma. Mol Cancer Ther 2007, 6(6):1886-1897.

15. Pandyra A, Mullen PJ, Kalkat M, Yu R, Pong JT, Li Z, Trudel S, Lang KS, Minden MD, Schimmer AD et al: Immediate utility of two approved agents to target both the metabolic mevalonate pathway and its restorative feedback loop. Cancer research 2014, 74(17):4772-4782.

16. Clendening JW, Pandyra A, Li Z, Boutros PC, Martirosyan A, Lehner R, Jurisica I, Trudel S, Penn LZ: Exploiting the mevalonate pathway to distinguish statin-sensitive multiple myeloma. Blood 2010, 115(23):4787-4797.

17. Kobayashi Y, Kashima H, Wu RC, Jung JG, Kuan JC, Gu J, Xuan J, Sokoll L, Visvanathan K, Shih Ie M et al: Mevalonate Pathway Antagonist Suppresses Formation of Serous Tubal Intraepithelial Carcinoma and Ovarian Carcinoma in Mouse Models. Clin Cancer Res 2015, 21(20):4652-4662.

18. Campbell MJ, Esserman LJ, Zhou Y, Shoemaker M, Lobo M, Borman E, Baehner F, Kumar AS, Adduci K, Marx C et al: Breast cancer growth prevention by statins. Cancer Res 2006, 66(17):8707-8714.

19. Mullen PJ, Yu R, Longo J, Archer MC, Penn LZ: The interplay between cell signalling and the mevalonate pathway in cancer. Nat Rev Cancer 2016, 16(11):718-731.

20. Goldstein JL, Brown MS: Regulation of the mevalonate pathway. Nature 1990, 343(6257):425-430.

21. Kaneko R, Tsuji N, Asanuma K, Tanabe H, Kobayashi D, Watanabe N: Survivin downregulation plays a crucial role in 3-hydroxy-3-methylglutaryl coenzyme A reductase inhibitor-induced apoptosis in cancer. J Biol Chem 2007, 282(27):19273-19281.

22. Tsubaki M, Mashimo K, Takeda T, Kino T, Fujita A, Itoh T, Imano M, Sakaguchi K, Satou T, Nishida S: Statins inhibited the MIP-1alpha expression via inhibition of Ras/ERK and Ras/Akt pathways in myeloma cells. Biomed Pharmacother 2016, 78:23-29.

23. Finlay GA, Malhowski AJ, Liu Y, Fanburg BL, Kwiatkowski DJ, Toksoz D: Selective inhibition of growth of tuberous sclerosis complex 2 null cells by atorvastatin is associated with impaired Rheb and Rho GTPase function and reduced mTOR/S6 kinase activity. Cancer Res 2007, 67(20):9878-9886.

24. Araki M, Maeda M, Motojima K: Hydrophobic statins induce autophagy and cell death in human rhabdomyosarcoma cells by depleting geranylgeranyl diphosphate. Eur J Pharmacol 2012, 674(2-3):95-103.

25. Taylor-Harding B, Orsulic S, Karlan BY, Li AJ: Fluvastatin and cisplatin demonstrate synergistic cytotoxicity in epithelial ovarian cancer cells. Gynecol Oncol 2010, 119(3):549-556.

26. Xia Z, Tan MM, Wong WW, Dimitroulakos J, Minden MD, Penn LZ: Blocking protein geranylgeranylation is essential for lovastatin-induced apoptosis of human acute myeloid leukemia cells. Leukemia 2001, 15(9):1398-1407.

27. Clarke S, Vogel JP, Deschenes RJ, Stock J: Posttranslational modification of the Ha- ras oncogene protein: evidence for a third class of protein carboxyl methyltransferases. Proc Natl Acad Sci U S A 1988, 85(13):4643-4647.

28. Boyartchuk VL, Ashby MN, Rine J: Modulation of Ras and a-factor function by carboxyl-terminal proteolysis. Science 1997, 275(5307):1796-1800.

29. Moores SL, Schaber MD, Mosser SD, Rands E, O'Hara MB, Garsky VM, Marshall MS, Pompliano DL, Gibbs JB: Sequence dependence of protein isoprenylation. J Biol Chem 1991, 266(22):14603-14610.

30. Willumsen BM, Christensen A, Hubbert NL, Papageorge AG, Lowy DR: The p21 ras C- terminus is required for transformation and membrane association. Nature 1984, 310(5978):583-586.

31. Hart KC, Donoghue DJ: Derivatives of activated H-ras lacking C-terminal lipid modifications retain transforming ability if targeted to the correct subcellular location. Oncogene 1997, 14(8):945-953.

32. Tsubaki M, Takeda T, Sakamoto K, Shimaoka H, Fujita A, Itoh T, Imano M, Mashimo K, Fujiwara D, Sakaguchi K et al: Bisphosphonates and statins inhibit expression and secretion of MIP-1alpha via suppression of Ras/MEK/ERK/AML-1A and Ras/PI3K/Akt/AML-1A pathways. Am J Cancer Res 2015, 5(1):168-179.

33. Elsayed M, Kobayashi D, Kubota T, Matsunaga N, Murata R, Yoshizawa Y, Watanabe N, Matsuura T, Tsurudome Y, Ogino T et al: Synergistic Antiproliferative Effects of Zoledronic Acid and Fluvastatin on Human Pancreatic Cancer Cell Lines: An in Vitro Study. Biol Pharm Bull 2016, 39(8):1238-1246.

34. Rigoni M, Riganti C, Vitale C, Griggio V, Campia I, Robino M, Foglietta M, Castella B, Sciancalepore P, Buondonno I et al: Simvastatin and downstream inhibitors circumvent constitutive and stromal cell-induced resistance to doxorubicin in IGHV unmutated CLL cells. Oncotarget 2015, 6(30):29833-29846.

35. Reilly JE, Zhou X, Tong H, Kuder CH, Wiemer DF, Hohl RJ: In vitro studies in a myelogenous leukemia cell line suggest an organized binding of geranylgeranyl diphosphate synthase inhibitors. Biochem Pharmacol 2015, 96(2):83-92.

36. Moriceau G, Roelofs AJ, Brion R, Redini F, Ebetion FH, Rogers MJ, Heymann D: Synergistic inhibitory effect of apomine and lovastatin on osteosarcoma cell growth. Cancer 2012, 118(3):750-760.

37. Kusama T, Mukai M, Tatsuta M, Nakamura H, Inoue M: Inhibition of transendothelial migration and invasion of human breast cancer cells by preventing geranylgeranylation of Rho. Int J Oncol 2006, 29(1):217-223.

38. Cox AD, Der CJ, Philips MR: Targeting RAS Membrane Association: Back to the Future for Anti-RAS Drug Discovery? Clin Cancer Res 2015, 21(8):1819-1827.

39. Soule HD, Maloney TM, Wolman SR, Peterson WD, Jr., Brenz R, McGrath CM, Russo J, Pauley RJ, Jones RF, Brooks SC: Isolation and characterization of a spontaneously immortalized human breast epithelial cell line, MCF-10. Cancer Res 1990, 50(18):6075-6086.

40. Yoon DS, Wersto RP, Zhou W, Chrest FJ, Garrett ES, Kwon TK, Gabrielson E: Variable levels of chromosomal instability and mitotic spindle checkpoint defects in breast cancer. Am J Pathol 2002, 161(2):391-397.

41. Yaswen P, Stampfer MR: Molecular changes accompanying senescence and immortalization of cultured human mammary epithelial cells. Int J Biochem Cell Biol 2002, 34(11):1382-1394.

42. Berndt N, Hamilton AD, Sebti SM: Targeting protein prenylation for cancer therapy. Nat Rev Cancer 2011, 11(11):775-791.

43. Liu Y, Lu X, Huang L, Wang W, Jiang G, Dean KC, Clem B, Telang S, Jenson AB, Cuatrecasas M et al: Different thresholds of ZEB1 are required for Ras-mediated tumour initiation and metastasis. Nat Commun 2014, 5:5660.

44. Liu Y, Sanchez-Tillo E, Lu X, Huang L, Clem B, Telang S, Jenson AB, Cuatrecasas M, Chesney J, Postigo A et al: The ZEB1 transcription factor acts in a negative feedback loop with miR200 downstream of Ras and Rb1 to regulate Bmi1 expression. J Biol Chem 2014, 289(7):4116-4125.

45. De Craene B, Berx G: Regulatory networks defining EMT during cancer initiation and progression. Nat Rev Cancer 2013, 13(2):97-110.

46. Singh A, Greninger P, Rhodes D, Koopman L, Violette S, Bardeesy N, Settleman J: A gene expression signature associated with "K-Ras addiction" reveals regulators of EMT and tumor cell survival. Cancer Cell 2009, 15(6):489-500.

47. Barretina J, Caponigro G, Stransky N, Venkatesan K, Margolin AA, Kim S, Wilson CJ, Lehar J, Kryukov GV, Sonkin D et al: The Cancer Cell Line Encyclopedia enables predictive modelling of anticancer drug sensitivity. Nature 2012, 483(7391):603-607.

48. Bessarabova M, Kirillov E, Shi W, Bugrim A, Nikolsky Y, Nikolskaya T: Bimodal gene expression patterns in breast cancer. BMC Genomics 2010, 11 Suppl 1:S8.

49. Muftah AA, Aleskandarany M, Sonbul SN, Nolan CC, Diez Rodriguez M, Caldas C, Ellis IO, Green AR, Rakha EA: Further evidence to support bimodality of oestrogen receptor expression in breast cancer. Histopathology 2017, 70(3):456-465.

50. Gendoo DM, Ratanasirigulchai N, Schroder MS, Pare L, Parker JS, Prat A, Haibe-Kains B: Genefu: an R/Bioconductor package for computation of gene expression-based signatures in breast cancer. Bioinformatics 2016, 32(7):1097-1099.

51. Wang J, Wen S, Symmans WF, Pusztai L, Coombes KR: The bimodality index: a criterion for discovering and ranking bimodal signatures from cancer gene expression profiling data. Cancer Inform 2009, 7:199-216.

52. Rees MG, Seashore-Ludlow B, Cheah JH, Adams DJ, Price EV, Gill S, Javaid S, Coletti ME, Jones VL, Bodycombe NE et al: Correlating chemical sensitivity and basal gene expression reveals mechanism of action. Nat Chem Biol 2016, 12(2):109-116.

53. Seashore-Ludlow B, Rees MG, Cheah JH, Cokol M, Price EV, Coletti ME, Jones V, Bodycombe NE, Soule CK, Gould J et al: Harnessing Connectivity in a Large-Scale Small-Molecule Sensitivity Dataset. Cancer Discov 2015, 5(11):1210-1223.

54. Basu A, Bodycombe NE, Cheah JH, Price EV, Liu K, Schaefer GI, Ebright RY, Stewart ML, Ito D, Wang S et al: An interactive resource to identify cancer genetic and lineage dependencies targeted by small molecules. Cell 2013, 154(5):1151-1161.

55. Smirnov P, Safikhani Z, El-Hachem N, Wang D, She A, Olsen C, Freeman M, Selby H, Gendoo DM, Grossmann P et al: PharmacoGx: an R package for analysis of large pharmacogenomic datasets. Bioinformatics 2016, 32(8):1244-1246.

56. Krens LL, Simkens LH, Baas JM, Koomen ER, Gelderblom H, Punt CJ, Guchelaar HJ: Statin use is not associated with improved progression free survival in cetuximab treated KRAS mutant metastatic colorectal cancer patients: results from the CAIRO2 study. PLoS One 2014, 9(11):e112201.

57. Lee JE, Baba Y, Ng K, Giovannucci E, Fuchs CS, Ogino S, Chan AT: Statin use and colorectal cancer risk according to molecular subtypes in two large prospective cohort studies. Cancer Prev Res (Phila) 2011, 4(11):1808-1815.

58. Baas JM, Krens LL, ten Tije AJ, Erdkamp F, van Wezel T, Morreau H, Gelderblom H, Guchelaar HJ: Safety and efficacy of the addition of simvastatin to cetuximab in previously treated KRAS mutant metastatic colorectal cancer patients. Invest New Drugs 2015, 33(6):1242-1247.

59. Baas JM, Krens LL, Bos MM, Portielje JE, Batman E, van Wezel T, Morreau H, Guchelaar HJ, Gelderblom H: Safety and efficacy of the addition of simvastatin to panitumumab in previously treated KRAS mutant metastatic colorectal cancer patients. Anticancer Drugs 2015, 26(8):872-877.

60. Hong JY, Nam EM, Lee J, Park JO, Lee SC, Song SY, Choi SH, Heo JS, Park SH, Lim HY et al: Randomized double-blinded, placebo-controlled phase II trial of

simvastatin and gemcitabine in advanced pancreatic cancer patients. Cancer Chemother Pharmacol 2014, 73(1):125-130.

61. Ng K, Ogino S, Meyerhardt JA, Chan JA, Chan AT, Niedzwiecki D, Hollis D, Saltz LB, Mayer RJ, Benson AB, 3rd et al: Relationship between statin use and colon cancer recurrence and survival: results from CALGB 89803. J Natl Cancer Inst 2011, 103(20):1540-1551.

62. DeClue JE, Vass WC, Papageorge AG, Lowy DR, Willumsen BM: Inhibition of cell growth by lovastatin is independent of ras function. Cancer Res 1991, 51(2):712-717.

63. Wu J, Wong WW, Khosravi F, Minden MD, Penn LZ: Blocking the Raf/MEK/ERK pathway sensitizes acute myelogenous leukemia cells to lovastatin-induced apoptosis. Cancer Res 2004, 64(18):6461-6468.

64. Goard CA, Chan-Seng-Yue M, Mullen PJ, Quiroga AD, Wasylishen AR, Clendening JW, Sendorek DH, Haider S, Lehner R, Boutros PC et al: Identifying molecular features that distinguish fluvastatin-sensitive breast tumor cells. Breast Cancer Res Treat 2014, 143(2):301-312.

65. Kanugula AK, Dhople VM, Volker U, Ummanni R, Kotamraju S: Fluvastatin mediated breast cancer cell death: a proteomic approach to identify differentially regulated proteins in MDA-MB-231 cells. PLoS One 2014, 9(9):e108890.

66. Kumar AS, Benz CC, Shim V, Minami CA, Moore DH, Esserman LJ: Estrogen receptor-negative breast cancer is less likely to arise among lipophilic statin users. Cancer Epidemiol Biomarkers Prev 2008, 17(5):1028-1033.

67. Loeb S, Kan D, Helfand BT, Nadler RB, Catalona WJ: Is statin use associated with prostate cancer aggressiveness? BJU Int 2010, 105(9):1222-1225.

68. Allott EH, Farnan L, Steck SE, Arab L, Su LJ, Mishel M, Fontham ET, Mohler JL, Bensen JT: Statin Use and Prostate Cancer Aggressiveness: Results from the Population-Based North Carolina-Louisiana Prostate Cancer Project. Cancer Epidemiol Biomarkers Prev 2016, 25(4):670-677.

69. Peinado H, Olmeda D, Cano A: Snail, Zeb and bHLH factors in tumour progression: an alliance against the epithelial phenotype? Nat Rev Cancer 2007, 7(6):415-428.

70. Thiery JP: Epithelial-mesenchymal transitions in tumour progression. Nat Rev Cancer 2002, 2(6):442-454.

71. Fidler IJ, Poste G: The "seed and soil" hypothesis revisited. Lancet Oncol 2008, 9(8):808.

72. Fischer KR, Durrans A, Lee S, Sheng J, Li F, Wong STC, Choi H, Rayes T, Ryu S, Troeger J et al: Epithelial-to-mesenchymal transition is not required for lung metastasis but contributes to chemoresistance. Nature 2015, 527(7579):472-476.

73. Zheng X, Carstens JL, Kim J, Scheible M, Kaye J, Sugimoto H, Wu C-C, LeBleu VS, Kalluri R: Epithelial-to-mesenchymal transition is dispensable for metastasis but induces chemoresistance in pancreatic cancer. Nature 2015, 527(7579):525-530.

74. Pattabiraman DR, Weinberg RA: Targeting the Epithelial-to-Mesenchymal Transition: The Case for Differentiation-Based Therapy. Cold Spring Harb Symp Quant Biol 2017.

75. Lamouille S, Xu J, Derynck R: Molecular mechanisms of epithelial-mesenchymal transition. Nat Rev Mol Cell Biol 2014, 15(3):178-196.

76. Kang S, Kim ES, Moon A: Simvastatin and lovastatin inhibit breast cell invasion induced by H-Ras. Oncol Rep 2009, 21(5):1317-1322.

77. Warita K, Warita T, Beckwitt CH, Schurdak ME, Vazquez A, Wells A, Oltvai ZN: Statin-induced mevalonate pathway inhibition attenuates the growth of mesenchymal-like cancer cells that lack functional E-cadherin mediated cell cohesion. Sci Rep 2014, 4:7593.

78. Fan Z, Jiang H, Wang Z, Qu J: Atorvastatin partially inhibits the epithelial- mesenchymal transition in A549 cells induced by TGF-beta1 by attenuating the upregulation of SphK1. Oncol Rep 2016, 36(2):1016-1022.

79. Ahern TP, Lash TL, Damkier P, Christiansen PM, Cronin-Fenton DP: Statins and breast cancer prognosis: evidence and opportunities. Lancet Oncol 2014, 15(10):e461-468.

80. Boudreau DM, Yu O, Chubak J, Wirtz HS, Bowles EJ, Fujii M, Buist DS: Comparative safety of cardiovascular medication use and breast cancer outcomes among women with early stage breast cancer. Breast Cancer Res Treat 2014, 144(2):405-416.

81. Chae YK, Valsecchi ME, Kim J, Bianchi AL, Khemasuwan D, Desai A, Tester W: Reduced risk of breast cancer recurrence in patients using ACE inhibitors, ARBs, and/or statins. Cancer Invest 2011, 29(9):585-593.

82. Kwan ML, Habel LA, Flick ED, Quesenberry CP, Caan B: Post-diagnosis statin use and breast cancer recurrence in a prospective cohort study of early stage breast cancer survivors. Breast Cancer Res Treat 2008, 109(3):573-579.

83. Allott EH, Howard LE, Cooperberg MR, Kane CJ, Aronson WJ, Terris MK, Amling CL, Freedland SJ: Postoperative statin use and risk of biochemical recurrence following radical prostatectomy: results from the Shared Equal Access Regional Cancer Hospital (SEARCH) database. BJU Int 2014, 114(5):661-666.

84. Gutt R, Tonlaar N, Kunnavakkam R, Karrison T, Weichselbaum RR, Liauw SL: Statin use and risk of prostate cancer recurrence in men treated with radiation therapy. J Clin Oncol 2010, 28(16):2653-2659.

85. Hamilton RJ, Morilla D, Cabrera F, Leapman M, Chen LY, Bernstein M, Hakimi AA, Reuter VE, Russo P: The association between statin medication and progression after surgery for localized renal cell carcinoma. J Urol 2014, 191(4):914-919.

86. Whyte DB, Kirschmeier P, Hockenberry TN, Nunez-Oliva I, James L, Catino JJ, Bishop WR, Pai JK: K- and N-Ras are geranylgeranylated in cells treated with farnesyl protein transferase inhibitors. J Biol Chem 1997, 272(22):14459-14464.

87. Wasylishen AR, Stojanova A, Oliveri S, Rust AC, Schimmer AD, Penn LZ: New model systems provide insights into Myc-induced transformation. Oncogene 2011, 30(34):3727-3734.

88. Morgenstern JP, Land H: Advanced mammalian gene transfer: high titre retroviral vectors with multiple drug selection markers and a complementary helper-free packaging cell line. Nucleic Acids Res 1990, 18(12):3587-3596.

89. Pandyra AA, Mullen PJ, Goard CA, Ericson E, Sharma P, Kalkat M, Yu R, Pong JT, Brown KR, Hart T et al: Genome-wide RNAi analysis reveals that simultaneous inhibition of specific mevalonate pathway genes potentiates tumor cell death. Oncotarget 2015, 6(29):26909-26921.

Chapter 3 Fluvastatin inhibits EMT-associated protein N-glycosylation and delays breast cancer metastasis

Contributions: The work presented in this chapter was performed by the author except for the following:

Table 3-I and Fig 3-6A-E, Dr. Cunjie Zhang and Dr. Jim Dennis performed N-glycan profiling by LC-MS/MS.

Fig 3-7B-C, Dr. Wenjiang Zhang and Dr. Eric Chen performed HPLC-MS/MS analysis of fluvastatin in the mouse serum and tumour xenograft.

This project was completed with supervisory support from Dr. Linda Z. Penn.

Fluvastatin inhibits EMT-associated protein N- glycosylation and delays breast cancer metastasis 3.1 Abstract

Metastatic recurrence accounts for the majority of breast cancer-related deaths. Epidemiological studies indicate that the risk of metastatic recurrence in breast cancer patients is lowered by adjuvant statin use. Statins are routinely prescribed to lower serum cholesterol, but the efficacy and molecular mechanism of their anti-breast cancer activity as an adjuvant therapeutic are unclear. We show that fluvastatin induces cell death in breast cancer cells that have undergone epithelial-to-mesenchymal transition (EMT) by inhibiting the biosynthesis of dolichol, required for protein N-glycosylation. Total glycome analysis indicates that fluvastatin treatment impairs the expression of complex type N-glycans that are upregulated during EMT and associated with breast cancer metastasis, both in cultured cells and in vivo. In a mouse model of post-surgical metastatic breast cancer, adjuvant fluvastatin treatment delays metastasis, reduces metastatic burden, and improves overall survival. Our data support the immediate repurposing of fluvastatin as an adjuvant therapeutic to target metastatic recurrence in breast cancer.

3.2 Introduction

The front-line therapy for early-stage breast cancer entails surgical removal of the tumour followed by adjuvant therapies (1, 2). Despite aggressive treatment, 20-30% of these patients experience recurrence, often as distant metastases (2, 3). Novel and effective therapeutics to prevent metastatic recurrence will greatly impact breast cancer patient survival. Several retrospective epidemiological studies have indicated that the risk of post-surgical breast cancer recurrence is reduced by 30-60% in patients who are taking statins (4-7), a class of FDA- approved medications that are commonly prescribed to lower serum cholesterol levels.

Importantly, increasing duration of adjuvant statin use is associated with decreasing risk of recurrence (7), suggesting that long-term intake of statins in the adjuvant setting can prolong patient survival. This potential therapeutic option is attractive, as statins are inexpensive, their safety profiles are well understood, and they can be taken for many years without serious adverse events. These properties make statins well-suited to be repurposed for use in the adjuvant setting, which can extend for >10 years in breast cancer patients (1, 2).

Pre-clinical studies have shown that statins can induce apoptosis in breast cancer cells in vitro (8-

10), inhibit growth of the primary breast tumour (11, 12), and prolong survival in mouse models of frank metastases, involving intravenous or intracardiac injections of a large bolus of cancer cells (13-15). These pre-clinical models have been well-established for studying neo-adjuvant and salvage therapies, but are poor models for evaluating the efficacy of adjuvant therapeutics

(16). As such, the utility of statins in the adjuvant setting remains unclear. The molecular mechanism(s) through which statins may prevent metastatic recurrence of breast cancer also remains to be elucidated. Statins inhibit the conversion of 3-hydroxy-3-methylglutaryl coenzyme

A (HMG-CoA) to mevalonate (MVA), the rate-limiting step of the MVA pathway (Fig 3-1A)

(17-19). The MVA pathway synthesizes cholesterol and isoprenoid metabolites including farnesyl diphosphate (FPP) and geranylgeranyl diphosphate (GGPP), required for posttranslational prenylation of proteins such as RAS (20-22). Previous work has suggested that depletion of FPP and/or GGPP, and consequently inhibition of protein prenylation of RAS family members, underlies statin-induced breast cancer cell death (23-26). However, mutations in RAS family proteins could not predict response to statins in several prospective clinical trials

(27-29), suggesting that alternative or additional mechanisms are at play. Understanding these mechanisms will be crucial to guide the design of clinical trials, identify biomarkers of statin

96 response, and identify novel anti-cancer pathways for the development of next-generation therapeutics to prevent metastatic breast cancer.

Here we sought to interrogate the mechanism of statin sensitivity in cell line and in vivo models that are reflective of the adjuvant therapeutic space. The endpoint of adjuvant therapies is metastatic outgrowth, which arises from disseminated primary tumour cells (1-3). Activation of epithelial-to-mesenchymal transition (EMT) in cancer cells has been proposed to be the critical initiating step for dissemination, which needs to be reversed at the secondary site to give rise to metastatic outgrowth (30). Although it is still debated whether these processes are required for metastasis, the observation that disseminated cancer cells often gain mesenchymal characteristics while losing epithelial ones (30) suggests that therapeutic targeting of breast cancer cells that have undergone EMT will reflect utility in the adjuvant setting, after cells have disseminated from the primary tumour but prior to metastatic reactivation. Thus, in this Chapter, we first report the use of an EMT model system in breast cancer cell lines to evaluate the mechanism of statin sensitivity in vitro. We show that breast cancer cells that have undergone EMT are sensitized to fluvastatin-induced cell death, and identify the molecular mechanism of statin sensitivity to be an increased dependency on the biosynthesis of dolichol via the MVA pathway

(Fig 3-1A). Dolichol is a group of long-chain isoprenoids that comprises the lipid component of lipid-linked oligosaccharides (LLO), essential for N-linked glycosylation of nascent peptides

(17-19). Previous studies have shown that statin treatment can block N-glycosylation on individual proteins such as P-gp (31), IGFR (32), EpoR (33), and FLT3 (34). Building on these studies, we show at a global level that fluvastatin treatment impairs the expression of complex type N-glycans associated with breast cancer EMT and metastasis (35-38), both in cell lines and in vivo. Finally, using an in vivo model of post-surgical metastasis that closely follows the course of human breast cancer progression and treatment (16, 39), we show that post-surgical adjuvant

97 fluvastatin treatment delays breast cancer metastasis, reduces metastatic load, and improves survival. Taken together, our results warrant the immediate clinical testing of fluvastatin as a safe and effective adjuvant therapeutic to prevent breast cancer metastasis, and suggest the development of novel therapeutics to combat metastatic recurrence in breast cancer by inhibiting aberrant protein N-glycosylation.

3.3 Results 3.3.1 EMT sensitizes breast cancer cells to fluvastatin and tunicamycin

We used MCF10A breast epithelial cells as our model system as they possess a highly stable genome, which allows for the evaluation of EMT in the absence of gross genetic instability (40).

Fluvastatin was chosen for our studies based on its favorable pharmacokinetic properties and promising anti-breast cancer activities in pre-clinical and clinical settings (8). Ectopic expression of the EMT-inducing transcription factor SNAIL triggered EMT in MCF10A cells, as shown by downregulation of E-cadherin and upregulation of fibronectin (Fig 3-1B). Treatment with fluvastatin readily induced cell death in SNAIL-overexpressing cells, but not vector control cells, as assessed by quantification of DNA content by fixed propidium iodide staining followed by flow cytometry (Fig 3-1C). Fluvastatin-induced cell death in SNAIL-overexpressing cells was fully rescued by co-administration with MVA or GGPP, but interestingly not FPP (Fig 3-1C), a metabolic intermediate between MVA and GGPP (Fig 3-1A) (17-19). This has also been reported in several other cell lines (9, 41-45), together suggesting that exogenous FPP is shunted towards cholesterol synthesis and away from GGPP synthesis, while disruption of biological processes downstream of GGPP is critical for statin-induced cell death.

GGPP is used for three biological processes: protein prenylation; synthesis of coenzyme Q

(CoQ) used in the electron transport chain (ETC); and synthesis of dolichol used for protein N-

98 glycosylation (Fig 3-1A) (17-19). We tested whether inhibiting any of these pathways individually using specific inhibitors (Fig 3-1A) could phenocopy statins and preferentially kill cells that have undergone EMT. EMT sensitized cells to fluvastatin, as indicated by a lowered

IC50 value in SNAIL-overexpressing cells (Fig 3-1D). To our surprise, EMT did not sensitize cells to GGTI-298 or GGTI-2133 (Fig 3-1D), suggesting that fluvastatin-induced cell death in the context of EMT is independent from inhibition of protein prenylation. Instead, inhibition of

LLO assembly downstream of dolichol synthesis by tunicamycin (46, 47) phenocopied fluvastatin by decreasing the IC50 in SNAIL-overexpressing cells (Fig 3-1D). The IC50 for 2- thenoyltrifluoroacetone (2-TTFA) and rotenone, two inhibitors of the ETC (48), were similar between both cell lines (Fig 3-1D), showing that EMT does not sensitize cells to inhibition of the

ETC.

These observations were validated in MCF10A cells overexpressing additional EMT-inducing transcription factors SLUG, TWIST or ZEB1 (49-51), as well as two independent breast cancer cell lines, MDA-MB-231 and MCF-7 (Fig 3-2A-D). Overexpressing of TWIST or ZEB1 induced

EMT in MCF10A cells, as indicated by downregulation of E-cadherin and upregulation of fibronectin, while overexpression of SLUG in the MCF10A cell system was unable to induce

EMT (Fig 3-2A). Consistently, the mesenchymal TWIST- and ZEB1-overexpressing cells became more sensitive to fluvastatin and tunicamycin compared to the vector control and the

SLUG-overexpressing cells, which remained epithelial (Fig 3-2B). The IC50 for GGTIs and ETC inhibitors were unaffected by EMT (Fig 3-2B). Similarly, immunoblotting for E-cadherin and vimentin, another marker of mesenchymal cells, indicated that MCF-7 cells were epithelial and

MDA-MB-231 cells were mesenchymal (Fig 3-2C). These two cell lines exhibited a 50-fold difference in sensitivity to both fluvastatin and tunicamycin, which could not be phenocopied by

GGTI-298, GGTI-2133, 2-TTFA, or rotenone (Fig 3-2D). Together, these data indicate that

vector A Acetyl-CoA B C 60 HMG-CoA Fluvastatin Myc tag 40 MVA E-cadherin 20 FPP Cholesterol % cell death

Protein prenylation fibronectin 0

SNAILMVA FPP GGTI-298; tubulin 60 EtOH Fluva GGPP GGTI-2133 * GGPP ETC 40 ** ** CoQ 2-TTFA; Dolichol 20 Rotenone % cell death Tunicamycin LLO 0

10 μM Fluva + MVA+ FPP+ + EtOH Fluva GGPP 100 μM MVA + Protein 10 μM FPP + N-glycosylation 10 μM GGPP +

D fluvastatin tunicamycin GGTI-298 GGTI-2133 2-TTFA rotenone 25 12 40 60 40 0.4

10 * 20 30 30 0.3

8 40

M) M)

15 M)



g/ml)

 

* 

 (

( (

** 6 ( 20 20 0.2

(

(ng/ml)

50 50

10 50 50

50 4 20

IC IC IC IC IC 10 10 0.1 5 2

0 0 0 0 0 0.0

vector SNAIL vector SNAIL vector SNAIL vector SNAIL vector SNAIL vector SNAIL

E vectorvector SNAILSNAIL TimeTime (h) (h) TimeTime (h) (h) 0 0 2424 4848 7272 - - - - 0 0 2424 4848 7272 - - - - 2020 µM µM fluva fluva ------0 0 2424 ------0 0 2424 1010 ng/ml ng/ml tuni tuni

180kDa180kDa GlycosylatedGlycosylated EGFREGFR 135kDa135kDa UnderUnder-glycosylated-glycosylated 135kDa135kDa GP130GP130 GlycosylatedGlycosylated 100kDa100kDa UnderUnder-glycosylated-glycosylated 100kDa100kDa GlycosylatedGlycosylated SLC3A2SLC3A2 75kDa75kDa 63kDa63kDa UnderUnder-glycosylated-glycosylated

Ku80Ku80 75kDa75kDa

Figure 3-1. Induction of EMT by SNAIL overexpression increases cell sensitivity to inhibition of dolichol-dependent protein N-glycosylation by fluvastatin and tunicamycin.

100

A, a simplified schematic of the mevalonate (MVA) pathway. Inhibitors of specific components of the pathway are represented in red. CoA, coenzyme A; HMG-CoA, 3-hydroxy-3- methylglutaryl coenzyme A; MVA, mevalonate; FPP, farnesyl diphosphate; GGPP, geranylgeranyl diphosphate; LLO, lipid-linked oligosaccharides; GGTI, geranylgeranyltransferase inhibitor; CoQ, coenzyme Q; ETC, electron transport chain; 2-TTFA, 2-thenoyltrifluoroacetone. B, immunoblot (IB) of E-cadherin, an epithelial cell marker, and fibronectin, a mesenchymal cell marker, revealed that overexpression of SNAIL induced EMT in MCF10A cells. Tubulin is used as loading control. C, Flow cytometric quantification of % dead cells (% population in the pre-G1 cell cycle) with propidium iodide DNA staining after fixation. Fluvastatin treatment for 72 h induced cell death in MCF10A cells overexpressing SNAIL, but not in vector control cells. Fluvastatin-induced cell death was fully rescued by co-administration with MVA or GGPP, but not FPP, at the indicated doses. Bars are mean + SD, n=3. *, p<0.05; **, p<0.01 (one-way ANOVA with a Dunnett post-test, comparing all columns vs. fluvastatin column). D, SNAIL overexpression sensitized cells to fluvastatin and tunicamycin, but not inhibitors of other components of the MVA pathway. IC50 values as calculated based on MTT assays after cells were treated with 8 doses of each drug for 72 h. Bars are mean + SD, n=3-4. *, p<0.05; **, p<0.01; (unpaired, two-tailed t test, comparing SNAIL vs. vector columns). E, IB for EGFR, GP130, and SLC3A2 for glycosylation status indicates that fluvastatin treatment for 48- 72 h led to under-glycosylation of EGFR, GP130, and SLC3A2 in both vector control and SNAIL-overexpressing cells as indicated by the appearance of lower-molecular weight bands. Tunicamycin treatment for 24 h was a positive control for protein under-glycosylation. Ku80 is used as a loading control. Representative images are shown, n=3-4.

101

A B fluvastatin tunicamycin 25 15

20 10 Myc tag M) 15

 * *

( (ng/ml) 50 10 *

** 50 5 IC E-cadherin 5 IC

0 0 b nec n ZEB1 ZEB1 vector SLUG TWIST vector SLUG TWIST b n GGTI-298 GGTI-2133 60 60

40 40

M) M)

  (

Myc tag (

50 50

20 20

IC IC E-cadherin 0 0

vimentin ZEB1 ZEB1 vector SLUG TWIST vector SLUG TWIST tubulin 2-TTFA rotenone 50 0.4

40 0.3

30 M)



g/ml) 

ZEB1 ( 0.2 (

20 50

50 IC

IC 0.1 E-cadherin 10 0 0.0

ZEB1 ZEB1 fibronectin vector SLUG TWIST vector SLUG TWIST tubulin

C D fluvastatin tunicamycin GGTI-298 150 1500 20

100 1000 15

M) M) 

 50 500 * (

( 10

(ng/ml) 50

50 4 40

IC IC

IC 5 2 ** 20 ** E-cadherin 0 0 0 en n MCF-7 MCF-7 MCF-7 b n MDA-MB-231 MDA-MB-231 MDA-MB-231 GGTI-2133 2-TTFA rotenone 40 150 1.0

0.8 30

100 M)

M) 0.6

 

g/ml)

 (

( 20

( 50 50 ** 0.4

50 50

IC IC

10 IC 0.2

0 0 0.0

MCF-7 MCF-7 MCF-7 MDA-MB-231 MDA-MB-231 MDA-MB-231 102

Figure 3-2. Induction of EMT increases cell sensitivity to fluvastatin and tunicamycin.

A, IB of E-cadherin, an epithelial cell marker, and fibronectin, a mesenchymal cell marker, revealed that MCF10A vector control and SLUG-overexpressing cells remained epithelial, while TWIST- and ZEB1-overexpressing cells underwent EMT. Tubulin is used as a loading control. B, Induction of EMT by TWIST or ZEB1 overexpression sensitized cells to fluvastatin and tunicamycin, but not inhibitors of other components of the MVA pathway. IC50 values as calculated based on MTT assays after cells were treated with 8 doses of each drug for 72 h. Bars are mean + SD, n=3-4. *, p<0.05; **, p<0.01 (one-way ANOVA with a Dunnett post-test, comparing all columns vs. vector column). C, IB of E-cadherin, an epithelial cell marker, and vimentin, a mesenchymal cell marker, revealed that MCF-7 cells are more epithelial, while MBA-MB-231 cells are more mesenchymal. Tubulin is used as a loading control. D, Fluvastatin and tunicamycin preferentially killed the more mesenchymal MDA-MB-231 cells, compared to the more epithelial MCF-7 cells, by approximately 50-fold. This difference could not be phenocopied by GGTI-298, GGTI-2133, 2-TTFA, or rotenone. IC50 values are calculated based on MTT assays after cells were treated with 8 doses of each drug for 72 h. Bars are mean + SD, n=3. *, p<0.05; **, p<0.01 (unpaired, two-tailed t test, comparing MDA-MB-231 and MCF-7 columns).

103 breast cancer cells with mesenchymal phenotypes are more sensitive to inhibition of dolichol synthesis and function by fluvastatin and tunicamycin.

3.3.2 Fluvastatin inhibits dolichol-mediated protein N-glycosylation

Dolichol is a group of large hydrophobic molecules that constitutes the lipid component of

LLOs, playing an essential role in the assembly of precursor N-glycans used for co-translational protein N-glycosylation (Fig 3-1A) (52, 53). We therefore tested whether fluvastatin treatment can inhibit protein N-glycosylation. EGFR, GP130, and SLC3A2 are three membrane proteins with multiple glycosylation sites. Under-glycosylated proteins migrate faster on polyacrylamide gels and can be visualized as lower molecular weight bands by immunoblotting (54, 55). All three membrane proteins became under-glycosylated after 48-72 h of fluvastatin treatment, in both vector and SNAIL-overexpressing cells (Fig 3-1E). As a positive control, tunicamycin- mediated inhibition of protein glycosylation was observed after 24 h (Fig 3-1E).

As tunicamycin treatment is known to elicit ER stress, we tested whether the observed effect on protein glycosylation is specific to the inhibition of dolichol biosynthesis and/or function, or a general feature of ER stress. To this end, we treated cells with thapsigargin, another known inducer of ER stress that acts through a dolichol-independent mechanism (56, 57). First we confirmed that treatment with tunicamycin or thapsigargin both induced ER stress, as indicated by upregulation of ER stress markers ERdj4 and BiP (58) in both vector and SNAIL- overexpressing cells, within 24 h of treatment (Fig 3-3A). In contrast, treatment with fluvastatin had little effect on the expression of ERdj4 or BiP for up to 72 h (Fig 3-3A), suggesting that defects in protein glycosylation caused by fluvastatin treatment (Fig 3-1E) can be substantially uncoupled from induction of ER stress. Consistently, treatment with thapsigargin for up to 72 h did not result in under-glycosylation of EGFR, GP130, or SLC3A2 in either vector or SNAIL-

104

A ERdj4 8 *** * ** 6 pLPC

SNAIL 4

expression 2 relative mRNA

0 BiP

sol sol 10 Tuni Thap Tuni Thap F20 24F20 48F20 72 * F20 24F20 48F20 72 8 * * * pLPC 6 SNAIL

4 expression

relative mRNA 2

sol sol Tuni Thap Tuni Thap F20 24 F20 48 F20 72 F20 24F20 48F20 72

B vectorvector SNAILSNAIL TimeTime (h) (h) TimeTime (h) (h) 00 2424 4848 7272 -- -- 00 2424 4848 7272 -- -- 1010 nM nMthapthap ------00 2424 ------00 2424 1010 ng/ml ng/ml tuni tuni 180kDa EGFR GlycosylatedGlycosylated 135kDa UnderUnder--glycosylatedglycosylated 135kDa GlycosylatedGlycosylated GP130 100kDa UnderUnder--glycosylatedglycosylated 100kDa GlycosylatedGlycosylated 75kDa SLC3A2 UnderUnder--glycosylatedglycosylated 63kDa

Ku80 75kDa

C thapsigargin thapsigargin 8 15

6 **

10 (nM)

4 (nM)

50 50

5 IC 2 IC

0 0

vector SNAIL MCF-7

MDA-MB-231

105

Figure 3-3. Fluvastatin effect is independent from induction of ER stress

A, qRT-PCR of ER stress markers ERdj4 and BiP revealed that fluvastatin treatment for up to 72 h did not induce ER stress, while tunicamycin and thapsigargin treatment both induced ER stress after 24 h of treatment. qRT-PCR of 18S rRNA was used as a housekeeping control. Bars are mean + SD, n=3. *, p<0.05; **, p<0.01; ***, p<0.001 (one-way ANOVA with a Dunnett post- test, comparing all columns vs. solvent control). B, IB for EGFR, GP130, and SLC3A2 for glycosylation status indicates that thapsigargin treatment for up to 72 h did not affect glycosylation status of these proteins in both vector control and SNAIL-overexpressing cells. Tunicamycin treatment for 24 h was a positive control for protein under-glycosylation. Ku80 is used as a loading control. Representative images are shown, n=3-4. C, IC50 values as calculated based on MTT assays after cells were treated with 8 doses of thapsigargin for 72 h, showed that sensitivity to thapsigargin was not modulated by EMT. Bars are mean + SD, n=3. **, p<0.01 (unpaired, two-tailed t test, comparing MDA-MB-231 and MCF-7 columns).

106 overexpressing cells (Fig 3-3B). Additionally, whereas overexpression of SNAIL sensitized cells to fluvastatin and tunicamycin (Fig 3-1D), their sensitivity to thapsigargin remained the same as vector control cells (Fig 3-3C). Comparing between MDA-MB-231 and MCF-7 cell lines, the mesenchymal MDA-MB-231 cells were in fact more resistant to thapsigargin than the epithelial

MCF-7 cells (Fig 3-3C), further supporting that mesenchymal breast cancer cells are sensitized to fluvastatin and tunicamycin treatment through a dolichol-dependent mechanism.

As the presence of serum prohibits the uptake of exogenous dolichol (59), we used a panel of 5 breast cancer cell lines that can proliferate in serum-reduced media (Opti-MEM, Thermo Fisher) to test whether exogenous dolichol can functionally rescue statin-induced cell death. Immunoblot of E-cadherin and vimentin showed that MCF-7 and HCC1937 were epithelial, while MDA-MB-

231, HCC1806, and BT549s acquired mesenchymal characteristics (Fig 3-4A). The mesenchymal cell lines were more sensitive to fluvastatin treatment, while the epithelial cell lines were more resistant, as indicated by the fluvastatin IC50 in each cell line (Fig 3-4B).

Consistently, fluvastatin treatment readily reduced cell viability in the mesenchymal breast cancer cell lines, but had no effect on the epithelial cell lines (Fig 3-4C). Exogenous addition of

MVA, GGPP, or dolichol (dolichyl[C95]-PP) fully rescued the cell viability in the presence of fluvastatin (Fig 3-4C). The presence of serum abolished the ability of dolichol to rescue fluvastatin-induced reduction in cell viability (Fig 3-5A-B), which likely explains previous contradicting results (60-65). Taken together, these data suggest that breast cancer cells with mesenchymal characteristics are more sensitive to fluvastatin-induced cell death due to an increased dependency on dolichol-dependent protein N-glycosylation.

107

A B fluvastatin 60

50 M)

 40

( 10 50

E-cadherin IC 5

vimentin 0 Ku80 MCF-7 BT549 HCC1937 HCC1806 MDA-MB-231

C ld_1937 bar ld_MCF7MCF bar-7 HCC1937 150 150

100 100

% cell viability cell % % cell viability cell %

0 0 3 μM Fluva sol F3 3 μM Fluva sol F3+ + + + + + + + MVA200 DOL 1.5 DOL 1.5 200 μM MVA GGPP 10 200 μM MVA MVA200+GGPP 10 + 10 μM GGPP + 10 μM GGPP + 1.5 μM Dol + 1.5 μM Dol +

MDA-MB-231 HCC1806 BT549 231 bar 1806 bar BT549 bar 150 * 150 * 150 * * * * ** ** ** * * ** 100 100 100

50 50 50

% cell viability cell %

% cell viability cell % % cell viability cell %

0 0 0

0.5 μM Fluva sol + + + + 2.5 μM Fluva sol + + + + 2.5 μM Fluva sol + + + + F0.5 F2.5 F2.5 DOL 1.5 GGPP 5DOL 1.5 200 μM MVA MVA200 +GGPP 10 200 μM MVA MVA200+GGPP 5 DOL 1.5 200 μM MVA MVA200 + 10 μM GGPP + 5 μM GGPP + 5 μM GGPP + 1.5 μM Dol + 1.5 μM Dol + 1.5 μM Dol +

Figure 3-4. Inhibition of dolichol synthesis underlies fluvastatin sensitivity in mesenchymal breast cancer cell lines.

A, IB of E-cadherin, an epithelial cell marker, and vimentin, a mesenchymal cell marker, indicates that MCF-7 and HCC1937 cells are more epithelial, while MBA-MB-231, HCC1806, and BT549 cells are more mesenchymal. Ku80 is used as a loading control. B, fluvastatin

108 preferentially killed the more mesenchymal MDA-MB-231, HCC1806, and BT549 cells, compared to the more epithelial MCF-7 and HCC1937 cells. IC50 values are calculated based on MTT assays after cells were treated with 8 doses of each drug for 72 h in Opti-MEM serum reduced media. Bars are mean + SD, n=3. C, MTT assays reveal that fluvastatin treatment for 72 h in Opti-MEM serum reduced media reduced cell viability in mesenchymal cells, but not in epithelial cell lines. Fluvastatin-induced reduction in cell viability was fully rescued by co- administration with MVA, GGPP, or dolichol (dolichyl[C95]-PP), at the indicated doses. Bars are mean + SD, n=3-4. *, p<0.05; **, p<0.01 (one-way ANOVA with a Dunnett post-test, comparing all columns vs. fluvastatin column).

109

A MDA231-MB-231 150 no DOL 1.5 M DOL 100

50 % MTT activity % MTT

0 0.01 0.1 1 10 fluvastatin (M) B MCF7MCF-7 150 no DOL 1.5 M DOL 100

50 % MTT activity % MTT

0 1 10 100 fluvastatin (M)

Figure 3-5. Exogenous addition of dolichol (dolichyl[C95]-PP) did not rescue viability of cells with fluvastatin treatment in DMEM supplemented with 10% FBS.

Cell viability of MDA-MB-231 cells (A) or MCF-7 cells (B) were calculated based on MTT assays after cells were treated with 7-8 doses of fluvastatin for 72 h. Data points are mean ± SD, n=3.

110

3.3.3 Fluvastatin inhibits EMT-associated protein N-glycosylation

Using EGFR, GP130, and SLC3A2 as reporters of glycosylation status, we showed that proteins became under-glycosylated with fluvastatin or tunicamycin treatment, in both the vector control cells and the SNAIL-overexpressing EMT cells (Fig 3-1E). Why, then, are SNAIL- overexpressing cells more sensitive to fluvastatin- and tunicamycin-induced cell death? Since several studies have previously reported an aberrant expression of N-glycans during cancer cell

EMT (66-70), we hypothesized that EMT leads to the upregulation of a subset of N-glycans, and are dependent on their expression for survival. To test this hypothesis, we quantified the total glycome of membrane proteins (71) in control and SNAIL-overexpressing cells by LC-MS/MS, both basally and after treatment with fluvastatin for 48 h (Table 3-I).

N-glycans share a common core region (Fig 3-6A, highlighted in box) and are classified based on the oligosaccharide composition in the antennae. The MCF10A membrane protein glycome consisted of 32% high mannose type N-glycans, containing exclusively mannose (M, Man) residues in the antennae (Fig 3-6A). Complex type N-glycans, characterized by alternating units of galactose (H, Gal) and N-acetylglucosamine (N, GlcNAc), represented a total of 59% of the

MCF10A membrane protein glycome, which can be further subdivided by the degree of modification by the monosaccharide fucose (F, Fuc) (Fig 3-6A). In MCF10A cells, complex type

N-glycans were commonly expressed in the unfucosylated and singly fucosylated (core) forms, with a small amount of doubly fucosylated (core and antennae) structures (Fig 3-6A). With induction of EMT, the expression of 12 N-glycans were significantly upregulated, all of which belonged to the complex type subgroup; and 15 N-glycans were downregulated, which included all 9 of the doubly fucosylated complex structures detected (Fig 3-6B, Table 3-I). This specific loss of doubly fucosylated complex type N-glycan structures in EMT has not been previously reported, and its underlying mechanism is unknown.

111

Table 3-I. Total membrane protein N-glycans in vector and SNAIL-overexpressing MCF10A cells with fluvastatin treatment.

Glycan Charge / ‘m/z’ Vector Vector SNAIL SNAIL (retention time) EtOH Fluva EtOH Fluva N2M3+N2H2 2+/821.3034 (10-13) 7.945 8.570 8.357 7.860 N2FM3+N2H2 2+/894.3335 (11.2-15) 16.122 18.598 18.002 16.848 N2FM3+N2H2F 2+/967.36245 (10.5-14) 1.232 1.303 0.151 0.209 N2M3+N3H3 2+/1003.872 (10.5-14) 3.058 2.732 4.385 3.589 N2FM3+N3H3 2+/1076.8981 (12-15.5) 7.993 7.871 10.469 8.711 N2FM3+N3H3F 2+/1149.9352 (12-15.9) 0.991 1.065 0.147 0.225 N2FM3+N3H3F 3+/766.9568 (12-15.9) 0.289 0.326 0.023 0.047 N2M3+N4H4 2+/1186.4468 (12.3-15.5) 1.689 1.292 2.531 1.793 N2M3+N4H4 3+/791.6279 (12.3-15.5) 1.103 0.868 1.376 1.088 N2FM3+N4H4 2+/1259.464 (13.8-16.5) 4.514 4.148 6.472 4.369 N2FM3+N4H4 3+/839.9795 (13.8-16.5) 5.107 4.306 5.848 4.193 N2FM3+N4H4F 2+/1332.5011(13.5-16.5) 0.511 0.518 0.046 0.071 N2FM3+N4H4F 3+/888.9998(13.5-16.5) 0.961 0.927 0.096 0.147 N2M3+N5H5 2+/1369.0154 (13.5-16.5) 0.216 0.191 0.245 0.202 N2M3+N5H5 3+/913.0102 (13.5-16.5) 0.663 0.586 0.706 0.636 N2FM3+N5H5 2+/1442.0326 (14.2-17.5) 0.463 0.423 0.593 0.484 N2FM3+N5H5 3+/961.6884 (14.2-17.5) 1.739 1.596 2.097 1.712 N2FM3+N5H5F 2+/1515.0697 (15-17.5) 0.110 0.116 0.004 0.007 N2FM3+N5H5F 3+/1010.3798 (15-17.5) 0.578 0.564 0.035 0.070 N2M3+N6H6 2+/1551.584 (15-17) 0.033 0.029 0.042 0.030 N2M3+N6H6 3+/1034.7227 (15-17) 0.310 0.279 0.414 0.335 N2FM3+N6H6 2+/1624.6012 (16-17.5) 0.099 0.085 0.135 0.106 N2FM3+N6H6 3+/1083.4008 (16-17.5) 0.928 0.812 1.261 0.995 N2FM3+N6H6F 3+/1132.0922 (16-17.5) 0.392 0.375 0.031 0.048 N2M3+N7H7 3+/1156.435 (15.8-17.5) 0.100 0.090 0.168 0.121 N2FM3+N7H7 3+/1205.1132 (16-17.5) 0.310 0.242 0.448 0.316 N2FM3+N7H7F 3+/1253.8046 (16-17.4) 0.184 0.168 0.012 0.016 N2M3+N8H8 3+/1278.14746 (16-17.3) 0.035 0.030 0.051 0.042 N2FM3+N8H8 3+/1326.7827 (17-17.5) 0.168 0.127 0.257 0.229 N2FM3+N8H8F 3+/1375.517 (17-17.5) 0.101 0.087 0.012 0.015 N2M3+N9H9 3+/1399.8598 (17-17.5) 0.020 0.019 0.028 0.025 N2FM3+N9H9 3+/1448.538 (17-17.5) 0.097 0.076 0.161 0.153 N2FM3+N9H9 3+/1497.2294 (16.2) 0.054 0.045 0.007 0.010 N2M3+N10H10 3+/1521.5722 0.008 0.009 0.014 0.013 N2FM3+N10H10 3+/1570.2504 (16.2) 0.055 0.043 0.087 0.083 N2FM3+N10H10 4+/1178.4269 (17.2-17.5) 0.147 0.133 0.191 0.228 N2FM3+N10H10F 4+/1214.4596 0.047 0.034 0.006 0.009 N2FM3+N11H11 3+/1691.9628 0.028 0.023 0.040 0.038

112

N2FM3+N12H12 4+/1360.5064 0.054 0.043 0.071 0.067 N2FM3+N14H14 4+/1543.075 0.023 0.019 0.027 0.023 N2FM3+N15H15 4+/1634.358 0.015 0.014 0.018 0.015 total complex 58.491 58.781 65.060 55.178 N2M3+H1(M4) 1+/1073.3895 (8-11.5) 1.459 3.093 1.273 1.995 N2M3+H2(M5) 2+/618.2244 (8.2-11.5) 2.368 2.607 1.920 2.558 N2M3+H2(M5) 1+/1235.4411 (8.2-11.5) 1.268 2.425 1.105 1.861 N2M3+H3(M6) 2+/699.2515 (8.2-10.5) 2.834 3.043 2.467 3.121 N2M3+H3(M6+NH4) 2+/707.768 (8.2-10.5) 1.528 1.528 1.228 1.735 N2M3+H3(M6 mod2) 2+/718.2252 (8.5-10.5) 0.144 0.155 0.104 0.136 N2M3+H3(M6) 1+/1397.4939 (8.2-10.5) 1.654 1.840 1.329 1.888 N2M3+H4(M7) 2+/780.2879 (8.2-10.5) 3.703 3.786 3.332 3.872 N2M3+H4(M7+NH4) 2+/788.7886 (8.2-10.5) 1.544 1.821 1.448 1.776 N2M3+H4(M7) 1+/1559.5467 (8.2-10.5) 0.449 0.646 0.422 0.504 N2M3+H5(M8) 2+/861.3047 (8-10.8) 4.751 3.782 4.226 4.545 N2M3+H5(M8+NH4) 2+/869.8157 (8-10.8) 2.631 2.324 2.633 3.192 N2M3+H6(M9) 2+/942.3285 (8-10.5) 4.740 2.735 3.415 2.057 N2M3+H6(M9+NH4) 2+/950.8421 (8-10.5) 2.564 1.511 1.940 1.163 N2M3+H7(M10) 2+/1023.356 (8.6-10.6) 0.684 0.282 0.361 0.192 N2M3+H7(M10+NH4) 2+/1031.869 (8.6-10.6) 0.349 0.143 0.204 0.105 N2M3+H6(M9 dimer) 3+/1256.106 (8-9.6) 0.096 0.043 0.048 0.023 N2M3+H6(M9 M8 3+/1202.087 (8-9.5) 0.248 0.144 0.155 0.125 dimer) total high mannose 33.014 31.908 27.612 30.846 N2M3+N1H2 2+/719.7605 (8-12.8) 1.616 1.669 1.597 1.969 N2FM3+N1H2 2+/792.791 (8.5-12.5) 0.714 0.893 0.538 0.988 N2M3+N1H3 2+/800.79645 (9-12) 2.033 1.792 1.720 2.108 N2FM3+N1H3 2+/873.8254 (10-13) 0.404 0.658 0.268 0.569 N2M3+N1H4 2+/881.8194 (8.6-11.5) 0.268 0.227 0.244 0.206 N2M3+N2H3 2+/902.3314 (10.4-14.5) 0.285 0.239 0.367 0.326 N2FM3+N2H3 2+/975.3618 (12-14.5) 0.270 0.328 0.276 0.295 N2M3+N2H4 2+/983.36245 (11.5-14) 0.064 0.050 0.094 0.108 N2FM3+N3H4 2+/1157.9311(13.5-16.5) 0.087 0.091 0.119 0.091 N2FM3+N6H8 3+/1191.4360 (16.3-17.6) 0.014 0.013 0.028 0.020 total hybrid 5.754 5.961 5.251 6.681 N2FM3+N1H1 2+/711.7674 (11-16) 0.676 0.722 0.323 0.670 N2M3+N2 2+/732.2813 (10.2-13.5) 0.761 1.042 0.505 2.903 N2M3+N3 2+/833.821 (10.1-13.8) 0.132 0.178 0.128 0.883 N2M3+N2H1 2+/813.3077 (8.01-13.01) 0.729 0.958 0.608 1.945 N2M3F+N3H2 2+/995.8738 (9.2-14.5) 0.294 0.317 0.324 0.670 N2M3F+N4H3 2+/1178.4399 (10.5-14.5) 0.149 0.133 0.190 0.224 total incomplete 2.741 3.350 2.077 7.295 GRAND TOTAL 100.000 100.000 100.000 100.000

113

A F, fucose complex high D E H, glu or gal mannose M, mannose N2FM3+N3H3 N2FM3+N4H4 N, GlcNAc 15 * * 15 * * ns ns

Core 10 10 N2(F)M3

5 5

9% vector vector % total glycans complex - unfucosylated % total glycans SNAIL SNAIL 15% * * complex - singly fucosylated (core) 0 0 32% complex - doubly fucosylated * * 38% high mannose * * EtOH Fluva EtOH Fluva EtOH Fluva EtOH Fluva other * 6% * * B C * * F G * * EMT *Fluva/EtOH* glycan SNAIL/vector vector* SNAIL* N2M3+H1(M4) N2M3+N2H2 * * GnT-V N2M3+N3H3N2M3+H2(M5) * * N2M3+H3(M6) GnT-III N2M3+N4H4 * * * MGAT5 MGAT3 N2M3+H4(M7) GnT-V GnT-III N2M3+N5H5 * * actin N2M3+N6H6N2M3+H5(M8) * *

N2M3+N7H7N2M3+H6(M9)

complex

complex high mannose high high mannose high * *

N2M3+H6(M9d) PHA-L PHA-E unfucosylated unfucosylated N2M3+N8H8 N2M3+N9H9N2M3+H6(M9/8d) N2M3+H7(M10) N2M3+N10H10 promotes represses N2M3+N2H2 N2M3+N11H11 * * metastasis metastasis N2M3+N3H3 N2FM3+N2H2 * * * N2M3+N4H4 N2FM3+N3H3 * * N2M3+N5H5 H N2FM3+N4H4 * * vector SNAIL N2M3+N6H6

N2FM3+N5H5 * * complex

complex N2M3+N7H7 20 μM fluva - + - - - + - - 20uM fluva unfucosylated unfucosylated N2FM3+N6H6 * * N2M3+N8H8 10 ng/ml tuni - - + - - - + - 10ng/ml tuni N2FM3+N7H7 * N2M3+N9H9 10 nM thap - - - + - - - + 10nM thap

N2FM3+N8H8 complex complex N2M3+N10H10 N2FM3+N9H9 N2FM3+N2H2

N2FM3+N10H10 (core) fucosylated singly singly fucosylated (core) fucosylated singly * N2FM3+N3H3 N2FM3+N11H11 PHA-L N2FM3+N4H4 PHA-L N2FM3+N12H12 N2FM3+N5H5 N2FM3+N13H13 N2FM3+N6H6 N2FM3+N2H2F * Fold N2FM3+N7H7 N2FM3+N3H3F * change

N2FM3+N8H8 complex complex N2FM3+N4H4F 2.5 N2FM3+N9H9 * N2FM3+N5H5F * PHA-E

N2FM3+N10H10 PHA-E singly fucosylated (core) fucosylated singly singly fucosylated (core) fucosylated singly N2FM3+N6H6F

N2FM3+N11H11 * complex complex N2FM3+N7H7F

N2FM3+N12H12 * doubly fucosylated doubly doubly fucosylated doubly N2FM3+N8H8F * N2FM3+N14H14 actin actin N2FM3+N9H9F N2FM3+N15H15 * N2FM3+N10H10F 0.01 1 2 3 4 5 6 7 8 N2FM3+N2H2F * N2FM3+N3H3F N2FM3+N4H4F N2FM3+N5H5F

FigureN2 3F-M3+N6H66. FluvastatinF treatment decreases complex branched N-glycans associated with complex

complex N2FM3+N7H7F doubly fucosylated doubly EMT. fucosylated doubly N2FM3+N8H8F N2FM3+N9H9F N2FM3+N10H10F A, schematic representation of high mannose type and complex type N-glycans, the two major classes of N-glycans (top), and distribution of major classes of N-glycans in the total cell surface glycome in MCF10A cells quantified by LC-MS/MS (bottom). B, heatmap of the expression of complex N-glycans following SNAIL-induced EMT. Data presented are the mean of 3 biological

114 replicates with 1-2 technical replicates each. *, q<0.0301 (unpaired, 2-tailed t test with Benjamini-Hochberg FDR correction). C, heatmap of the expression of complex N-glycans following treatment with 20 µM fluvastatin in vector cells (left column) and SNAIL- overexpressing cells (right column). Black arrowheads indicate glycan species that are significantly upregulated in EMT (B) and downregulated by fluvastatin treatment in SNAIL- overexpressing cells (C, right column), but not affected by fluvastatin treatment in control cells (C, left column). Data presented are the mean of 3 biological replicates with 1-2 technical replicates each. *, q<0.0108 (unpaired, 2-tailed t test with Benjamini-Hochberg FDR correction). D-E, LC-MS/MS quantification of the β1,6-branched, tri-antennary (D) and tetra-antennary (E) N-glycans indicate that these N-glycan species are upregulated in EMT, which is inhibited by 20 μM fluvastatin treatment for 48 h. Bars are mean + SD, n=3. ns, not significant; *, p<0.05 (one- way ANOVA with a Bonferroni post-test, comparing selected pairs of columns). F, schematic representation of two mutually exclusive N-glycan branching pathways, catalyzed by different enzymes, recognized by different lectins, and with opposing effects on metastasis. G, IB of GnT- III (MGAT3) and GnT-V (MGAT5) showed that SNAIL-overexpressing cells upregulated GnT-V while GnT-III expression was unchanged, compared to vector control cells. Actin was used as a loading control. Representative images are shown, n=3. H, lectin blotting revealed that PHA-L ligand was upregulated in SNAIL-overexpressing cells, and was decreased by fluvastatin after 48 h. PHA-E ligand was expressed at comparable levels between the two cell lines and was not affected by fluvastatin treatment. Treatment with tunicamycin for 48 h was used as a positive control that decreased expression of both PHA-L and PHA-E ligands in both cell lines. Treatment with thapsigargin for 48 h was used as a negative control. Actin was used as a loading control. Representative images are shown, n=4.

115

We then examined the effect of fluvastatin treatment on N-glycan profiles. Our hypothesis predicts that fluvastatin treatment will inhibit the expression of N-glycans that are upregulated by

EMT. Indeed, 6 of the 12 N-glycans upregulated following induction of EMT were specifically inhibited by fluvastatin treatment in SNAIL-overexpressing cells, but not control cells (Fig 3-6C, black arrowheads), suggesting that these N-glycans were important mediators of fluvastatin- induced cell death. Of these, the singly fucosylated tri-antennary (N2FM3+N3H3) and singly fucosylated tetra-antennary (N2FM3+N4H4) structures, each representing ~10% of the total surface glycome, remained unchanged with fluvastatin treatment in the vector control cells (Fig

3-6D-E). SNAIL-overexpressing cells upregulated the expression of these N-glycans, and this upregulation is blocked when cells were exposed to fluvastatin (Fig 3-6D-E).

The tri- and tetra-antennary complex type N-glycans contain a β1,6 glycosidic linkage

(β1,6GlcNAc-branching) produced by the enzyme β1,6-N-acetylglucosaminyltransferase V

(GnT-V; encoded by MGAT5) and recognized by the lectin phytohaemagglutinin-L (PHA-L; Fig

3-6F) (67, 72). β1,6GlcNAc-branched tri- and tetra-antennary N-glycans (PHA-L ligands) are known to promote metastasis (67, 72). As a control, we also examined the expression of β1,4-N- acetylglucosaminyltransferase III (GnT-III; encoded by MGAT3), which adds a bisecting

GlcNAc to the trimannosyl core (PHA-E ligands), and has been suggested to have metastasis- suppressing activities (Fig 3-6F) (73). Consistent with the increase in the tri- and tetra-antennary complex type N-glycans with induction of EMT (Fig 3-6D-E), we showed that induction of EMT by SNAIL overexpression led to upregulation of GnT-V, while GnT-III expression remained unchanged (Fig 3-6G). SNAIL-overexpressing cells showed an upregulation of PHA-L ligands

(74, 75) compared to the vector control (Fig 3-6H, compare lane 5 to lane 1). Following fluvastatin treatment, PHA-L binding is unaffected in vector cells (compare lane 2 to lane 1), but is decreased in SNAIL-overexpressing cells (compare lane 6 to lane 5). In contrast, PHA-E

116 binding (74, 75) was similar between SNAIL-overexpressing cells and vector control cells, and there was minimal effect of fluvastatin on PHA-E substrates (Fig 3-6H). Tunicamycin and thapsigargin were used as positive and negative controls, respectively, for inhibition of protein glycosylation (Fig 3-6H).

3.3.4 Fluvastatin impairs protein N-glycosylation in vivo

Our in vitro data indicate that fluvastatin treatment impairs the biosynthesis of dolichol, which is needed to support the expression of N-glycans associated with EMT and metastasis. To test this mechanism of fluvastatin action in vivo, we treated a xenograft model of the aggressive LM2-4 breast cancer cell line, a derivative of MDA-MB-231 cells that preferentially metastasizes to the mouse lung (16, 39), with PBS or 50 mg/kg/d fluvastatin. Since fluvastatin treatment delayed growth of the primary tumour (Fig 3-7A), we extended treatment time in the fluvastatin group so that the tumours harvested for further analyses were size-matched (Fig 3-7A). Fluvastatin was readily detected by HPLC-MS/MS in both the mouse serum (4.9 ± 0.3 μg/ml or 11.9 ± 0.9 μM) and the xenograft tissue (0.59 ± 0.08 μg/g tissue) (Fig 3-7B-C). To the best of our knowledge, this is the first report of fluvastatin detection in extrahepatic tumour tissue to support the model that statins can directly induce cancer cell death in vivo, independent from their systemic cholesterol-lowering and/or immunomodulatory activities.

To test whether fluvastatin can inhibit EMT-associated protein N-glycosylation in vivo, we performed lectin histochemistry (74, 76) using PHA-L, and showed that fluvastatin decreased the number of cells staining positive for PHA-L in Ki67-positive regions of tumour sections (Fig 3-

7D-E). We confirmed that LM2-4 xenografts will readily metastasize to the mouse lung by identifying lung lesions that stained positive for human EGFR (hEGFR) (Fig 3-7F). In the PBS treated group, 3 out of 6 mice (50%) had one or more lung lesions (Fig 3-7G). In the fluvastatin

sc serum 8000

6000

4000

2000 fluvastatin (ng/ml) fluvastatin

0 117

PBS fluva

A B C sc serum sc tumour 1000 PBS 50 mg/kg/d fluva serum tumor

) 8000 1000 3 800 800 6000 600 600 4000 400 400 200 2000

200

fluvastatin (ng/ml) fluvastatin

tumor volume (mm volume tumor fluvastatin (ng/g tissue) fluvastatin 0 0 0 0 2 4 6 8 10 12 14

days of treatment PBS fluva PBS fluva

PHA-L positivity D PBS Fluva E 60 sc tumour 1000 40 800 ** PHA-L 600 20 400

# of PHA-L-positive cells PHA-L-positive of # 200

Ki67 0 fluvastatin (ng/g tissue) fluvastatin 0 PBS fluva

PBS fluva F Mouse lung G H&E hEGFR Number of mice with metastasis (%) PBS 3/6 (50%) Fluva 1/5 (20%)

Figure 3-7. Fluvastatin treatment decreases complex branched N-glycans in vivo.

A, fluvastatin delayed growth of the primary tumour. One million LM2-4 cells were subcutaneously implanted in SCID mice and treated with PBS or 50 mg/kg/d fluvastatin by oral gavage. Mice were sacrificed 6 days (PBS) or 12 days after treatment (fluvastatin) to ensure tumours from the two groups were size-matched for downstream analyses. Data points are mean ± SD, n=5-6. B-C, at time of sacrifice, mouse serum and a piece of tumour xenograft were flash- frozen in liquid N2, and fluvastatin was extracted and quantified by HPLC-MS/MS. Data points are mean ± SD, n=4-6. D-E, lectin histochemistry of tumour xenografts indicated that PHA-L ligand in Ki67-positive tissue areas was decreased with fluvastatin treatment. Representative

118 images are shown, n=5. Scale bars = 50 μm. Data points are mean ± SD, n=5. **, p<0.01 (unpaired, two-tailed t test, comparing fluvastatin treatment vs. PBS control). F-G, fluvastatin treatment decreased proportion of mice with metastatic lesions in the lungs at the time of sacrifice. At time of sacrifice, mouse lungs were resected and FFPE. Two sequential slices were obtained every 200 μm for three depths containing all five lobes, and stained for H&E or hEGFR (F) Scale bars = 100 μm. Metastatic colonies were identified by hEGFR and confirmed by H&E. Each lung slice was independently reviewed by two personnel. Representative images are shown, n=5-6.

119 treated group, 1 out of 5 mice (20%) had hEGFR-positive metastatic lesions, providing the first indication that fluvastatin treatment can prevent breast cancer metastasis in this model (Fig 3-

7G). Thus, fluvastatin treatment inhibits the expression of N-glycans associated with EMT and metastasis in vivo.

3.3.5 Post-surgical adjuvant fluvastatin treatment delays metastatic outgrowth and prolongs survival

To test the efficacy of fluvastatin treatment in the post-surgical adjuvant setting, we allowed

LM2-4 xenografts to reach ~500 mm3, then surgically removed the primary tumours to mimic front-line treatment in breast cancer patients (16, 39). This experimental design also serves to uncouple the effect of fluvastatin on metastatic outgrowth from delay of metastasis due to inhibition of primary tumour growth. After surgery, mice were randomly assigned to receive

PBS or 50 mg/kg/d fluvastatin orally (Fig 3-8A), mimicking a p.o./q.d. (per os/quaque die, by mouth/once a day) prescription. Adjuvant fluvastatin treatment significantly prolonged overall survival in this mouse model (Fig 3-8B).

To evaluate the potential anti-metastatic activity of fluvastatin, we analyzed lung samples at three time points during the course of the experiment: (i) at time of surgery; (ii) at 8-9 days post- surgery; and (iii) at endpoint (Fig 3-8A). At time of surgery, we confirmed that this model accurately represented early-stage breast cancer, as the majority of mice did not have any observable metastases, and 2/8 mice had very small lung lesions (Fig. 3-8C). At 8-9 days post- surgery, analysis of a small cohort of animals showed that adjuvant fluvastatin treatment effectively inhibited metastatic outgrowth from disseminated breast cancer cells after the primary tumour was removed (Fig 3-8D). Finally, at endpoint, we showed that fluvastatin treatment decreased the proportion of mice with heavy (>50 colonies per slice) or medium metastatic load

(5-50 colonies per slice), while increasing the proportion of mice with light metastatic load

120

adjuvant survival adjuvant survivalSurvival A (i) (ii) (iii) B 100 Median survival 100 PBS PBS 28 days50mg/kg fluvastatin 80 Fluva 37 days * 80 50mg/kg fluvastatin Primary tumour Surgical resection of implantation primary tumour 60 60 PBS 40

40 Percentsurvival

Primary 20 Percentsurvival tumour grown 50mg/kg/d 20 until 500mm3 fluvastatin 0 0 0 10 20 30 40 50 0 10 20 30days after40 surgery50 days after surgery 3 C (i) At tumour volume = 500mm , D (ii) Post-surgery d8-9 i.e. at time of surgery PBS Fluva Lung mets: 2/8 mice Lung mets: 1/2 mice Lung mets: 1/3 mice

E (iii) Endpoint(iii) Endpoint F <5 colonies<5 colonies per slice per slice5-50 5colonies-50 colonies per slice per slice>50 colonies>50 colonies per slice per slice 120 120 <5 colonies<5 colonies per slice per slice 100 100 5-50 colonies5-50 colonies per slice per slice >50 colonies>50 colonies per slice per slice 80 80

60 60

PBS PBS 40 40

% population % population 20 20

0 0 (3/10(3/10 mice) mice) (4/10(4/10 mice) mice) (3/10(3/10 mice) mice) PBS PBSFluvaFluva G60 60

40 40

FluvaFluva

20 20

% hEGFR positivity % hEGFR positivity % hEGFR

0 0 (6/9 mice)(6/9 mice) (1/9 mice)(1/9 mice) PBS PBSFluva Fluva (2/9 mice)(2/9 mice)

Figure 3-8. Post-surgical adjuvant fluvastatin treatment delays metastasis and prolongs survival.

A, schematic of the mouse model and the time points where mice were sacrificed. B, fluvastatin treatment at 50 mg/kg/d orally significantly prolonged survival of mice with post-surgical metastatic breast cancer. *, p<0.05 (Log-rank test, n=12). C-E, At the indicated time point, mice were sacrificed and lungs were resected for FFPE. Two sequential slices were obtained every 200 μm for three depths containing all five lobes, and stained for H&E or hEGFR. Metastatic foci were identified by hEGFR staining and confirmed by H&E. At time of surgery, mouse lungs

121 were clear of metastatic foci or had very small lesions (C). Scale bar = 20 μm. At 8-9 days post- surgery, mice receiving fluvastatin treatment had less metastatic tumour load than mice receiving PBS control (D). At endpoint, fluvastatin treatment decreased the proportion of mice with heavy (>50 colonies per slice) or intermediate metastatic load (5-50 colonies per slice). The proportion of mice with light metastatic load (<5 colonies per slice) were increased (E). Each lung slice was independently reviewed by two personnel. Representative images are shown. F-G, quantification of metastatic load by colony count (F) or by hEGFR positivity (G) both showed lowered metastatic load in fluvastatin-treated mice.

122

(<5 colonies per slice; Fig 3-8E-G). We have thus demonstrated using a post-surgical metastatic breast cancer model that closely follows the course of human disease, that adjuvant fluvastatin use can delay metastatic spread and prolong survival.

3.4 Discussion

Metastatic recurrence is the main cause of breast cancer deaths (1-3). Accumulating evidence supports a role for statins in reducing the risk of post-surgical breast cancer recurrence. A major obstacle to the repurposing of statins as anti-cancer drugs has been an incomplete understanding of their molecular mechanism of action. Statins were previously thought to induce apoptosis in cancer cells by inhibiting FPP and GGPP synthesis, thereby disrupting the membrane localization of RAS family members (23-26). However, mutations in RAS family members have been poor biomarkers of statin response (27-29), suggesting not only that an alternative mechanism is at play, but also that better biomarkers need to be developed to identify patients that will benefit from statin treatment. Here, we address this important gap by demonstrating that sensitivity to fluvastatin in the context of breast cancer cell EMT is mediated by depletion of dolichol, a class of metabolites downstream of FPP and GGPP, which functions as the essential lipid carrier of glycans prior to protein N-glycosylation (52, 53). Fluvastatin treatment impaired the upregulation of β1,6GlcNAc-branched tri- and tetra-antennary N-glycans associated with breast cancer EMT and metastasis (66-70), both in cell culture and in vivo. Finally, using a clinically relevant model of metastatic recurrence, we demonstrate that adjuvant use of fluvastatin delayed breast cancer metastasis and prolonged survival. Our results support that the statin family of inexpensive and well-tolerated drugs could be repurposed as adjuvant therapeutics, and suggest the development of novel therapeutics to prevent metastatic recurrence in breast cancer by targeting aberrant protein N-glycosylation.

123

Activation of EMT is proposed to be a key component in cancer cell dissemination and metastatic dormancy (30). Altered protein N-glycosylation, and notably the upregulation of GnT-

V (MGAT5), is pivotal to EMT (70, 77, 78). The tri- and tetra-antennary β1,6GlcNAc-branched

N-glycans produced by GnT-V are potent modulators of metastatic potential (35, 36).

Consistently, high levels of tri- and tetra-antennary complex N-glycans are associated with disease progression and poor prognosis in breast and colon cancer patients (37, 38). Importantly, reducing the glycan structures produced by GnT-V is known to suppress cancer metastasis and

EMT. In polyomavirus middle T transgenic mice (MMTV-PyMT), an Mgat5-deficient background markedly suppressed breast cancer growth and metastasis (36). Mgat5-deficient

PyMT tumour cells displayed a block to EMT and reduced sensitivity to growth factors that were both rescued by transfection with Mgat5 (70). Here we demonstrated that the aberrant upregulation of N-glycan structures caused by increased expression of GnT-V during EMT is dependent on dolichol availability to support protein N-glycosylation in the ER. The assembly of dolichol-GlcNAc2Man9Glc3, the donor substrate for N-glycosylation, occurs in two phases (79): phase I takes place in the cytoplasmic face of the ER, producing dolichol-GlcNAc2Man5 using dolichol, UDP-GlcNAc, and GDP-Man; and phase II occurs in the ER lumen and requires dolichol-Man (4 molecules) and dolichol-Glc (3 molecules) as substrates. Thus, the assembly of each molecule of dolichol-GlcNAc2Man9Glc3 requires 8 dolichol molecules, and the observation that dolichol cannot be efficiently recycled (80, 81) indicates that cells must continuously synthesize dolichol through the MVA pathway to meet this demand. Indeed, dolichol-

GlcNAc2Man9Glc3 levels can become limiting during nutrient deprivation (82), or conversely in rapidly dividing cancer cells. Additionally, approximately 30% of the transcriptome in vertebrates are N-glycosylated at one or more sites in the ER, before transiting into the Golgi where the N-glycans are remodelled by the branching pathway (79). The N-glycan branching

124 enzymes GnT-I to GnT-V (MGAT1 to MGAT5) all compete for available substrates, UDP-

GlcNAc (69) and N-glycan structures that are to be remodeled (79), with a ~300-fold stepwise decline in substrate affinity (69) moving from GnT-I to GnT-V. Consistently, we observed a proportional reduction in bi-, tri-, and tetra-antennary glycan structures (7.3%, 16.8%, and

30.0%; Fig 3-6C, Table 3-I) with fluvastatin treatment in SNAIL-overexpressing cells. Thus, our results here suggest that breast cancer cells that have undergone EMT display increased tri- and tetra-antennary glycan structures mediated by GnT-V, that must be supported by synthesis of dolichol and assembly of dolichol-GlcNAc2Man9Glc3 in the ER. This vulnerability is exploited at the level of MVA pathway inhibition by fluvastatin treatment, in turn impacting the Golgi N- glycan branching pathway (Fig 3-9).

Two prospective window-of-opportunity clinical trials have been completed to evaluate the impact of statin use as a neo-adjuvant therapeutic in breast cancer patients (83, 84). Comparing tumour samples at biopsy and at surgery, both studies reported a decrease in Ki67 staining after atorvastatin (83) or fluvastatin (84) treatment, indicating that statins decreased tumour cell proliferation, and supporting that statins should be repurposed to treat breast cancer. Here, we have carefully chosen a mouse model of post-surgical metastatic breast cancer that closely mimics front-line treatment and disease progression (16, 39), to test the efficacy of fluvastatin when used in the adjuvant setting to prevent metastasis, where long-term use of this safe and inexpensive drug will likely have clinical benefit. Fluvastatin treatment in this therapeutic space effectively delayed metastasis and prolonged survival at a daily dose of 50 mg/kg in the mouse, equivalent to a well-tolerated daily dose of 4 mg/kg in human patients for the management of primary hypercholesterolemia (85). This study was also the first to measure the concentration of fluvastatin in the mouse serum and xenograft tissue, with a treatment regimen and route closely mimicking what would be prescribed to human patients (p.o./q.d.). A serum concentration

125

Primary Metastatic tumor EMT recurrence

Fluvastatin Mevalonate pathway

Dolichyl [C95-105]-PP

Figure 3-9. Proposed mechanism of statin-mediated attenuation of breast cancer metastasis by targeting EMT-associated protein N-glycosylation.

Breast cancer cells disseminate from the primary tumour by undergoing EMT, whereby they lose cell-cell interactions, degrade and invade through the basement membrane, enter the circulatory system, and ultimately give rise to metastatic outgrowth in a secondary site. Breast cancer cells that have undergone EMT are characterized by an upregulation of tri- and tetra-antennary complex type N-glycans, whose expression is dependent on dolichol synthesis via the MVA pathway, making them more sensitive to fluvastatin-induced cell death. Thus fluvastatin may have clinical impact if prescribed in the adjuvant therapeutic space to prevent metastatic recurrence of breast cancer.

126 of 4.9 μg/ml (11.9 μM) fluvastatin was achieved in our mouse experiments (Fig 3-7B). In comparison, in human volunteers, up to 7 μM fluvastatin was achieved when treated at 1 mg/kg/d (86). and up to 12.3 μM lovastatin, another statin family member, was achievable when prescribed at 10 mg/kg/d (87). Additionally, a fluvastatin concentration of 586.2 ng/g tissue was measured in the tumour xenograft (Fig 3-7C). Assuming a tissue density of 1 g/ml, this is equivalent to a fluvastatin concentration of approximately 1.4 μM, proving that fluvastatin can reach extrahepatic tumour tissues to exert direct anti-cancer effects. Together these results warrant the immediate clinical evaluation of fluvastatin at this well-tolerated dose in the adjuvant setting in breast cancer patients.

3.5 Materials and Methods 3.5.1 Reagents

Fluvastatin was purchased from US Biologicals (F5277-76). TGF-β was purchased from

PeproTech (100-21). PNGase F was purchased from NEB (P0704). cOmplete protease inhibitor was purchased from Roche (11697498001). RapiGest SF was purchased from Waters

(186001861). Sialidase was purchased from Glyko (GK80040). All other chemicals were purchased from Sigma unless otherwise specified.

3.5.2 Cell culture

MCF10A cells were a kind gift of Dr. Senthil Muthuswamy. MDA-MB-231 and LM 2-4 cells were a kind gift of Dr. Robert Kerbel. BT549, HCC1806, HCC1937, MCF-7, MCF10A, MDA-

MB-231, LM2-4, and 293Tv cells were cultured as previously described (8, 10, 16, 39, 88).

Transgene expression was stably introduced into MCF10A cells using retroviral insertion with pLPC, a kind gift of Dr. Roberta Maestro, or pBabePuro (89). Cells were imaged on a Leica

Stereomicroscope (Leica MZ FLIII).

127

3.5.3 MTT assay

3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assays were performed as previously described [6]. Cells were seeded at 750-5,000 cells/well in 96-well plates and treated in triplicate with 8 doses of drugs or the solvent control for 72 h, either in growth media or in

Opti-MEM serum reduced media as indicated. IC50 values were computed using GraphPad Prism with a bottom constraint equal to 0.

3.5.4 Immunoblotting

Lysates were prepared by lysing directly in boiling SDS lysis buffer (1% SDS, 11% glycerol,

10% β-mercaptoethanol, 0.1 M Tris pH 6.8). The following antibodies were used for detection: c-MYC (MAb 9E10, in-house), E-Cadherin (CST 3195), vimentin (CST 5741), fibronectin

(Abcam ab32419), actin (Sigma A2066), tubulin (Millipore CP06), GP130 (SCB sc-655), EGFR

(CST 2232), SLC3A2 (SCB sc-7095), Ku80 (CST 2180), GnT-III (Thermo PA5-12156), GnT-V

(Thermo MA5-24325), PHA-L (EY Labs H-1801-1), PHA-E (EY labs H-1802-1).

3.5.5 Immunohistochemistry

Tissue samples were formalin-fixed and paraffin-embedded by standard protocols. For tumours, three sequential slices were stained for H&E, Ki67 (Novus NB110-90592), or biotin-PHA-L (EY

Labs BA-1801-2). The number of PHA-L positive cells were counted at 20x magnification in

Ki67-positive regions with a minimum of four views per slice. For lungs, two sequential slices were obtained every 200 μm for three depths containing all five lobes, and stained for H&E or hEGFR (Zymed 28005). Metastatic colonies were identified by hEGFR staining and confirmed by H&E. Each lobe was individually outlined and total hEGFR positivity was computed using

ImageScope.

128

3.5.6 Cell death assay

Cells were seeded at 250,000/plate overnight, then treated with as indicated for 72 h. Cells were fixed in 70% ethanol overnight, stained with propidium iodide (Sigma), and analyzed for the sub-diploid DNA (“pre-G1”) population as previously described (8).

3.5.7 qRT-PCR

Total RNA was harvested from subconfluent cells using TIRzol Reagent (Invitrogen). cDNA was synthesized from 500 ng of RNA using SuperScript III (Invitrogen). Real-time quantitative

RT-PCR was performed using SYBR Green (Applied Biosystems) with the following primers:

BiP_fw 3’- TGACATTGAAGACTTCAAAGCT-5’

BiP_rv 3’- CTGCTGTATCCTCTTCACCAGT-5’

ERdj4_fw 3’- AAAATAAGAGCCCGGATGCT-5’

ERdj4_rv 3’- CGCTTCTTGGATCCAGTGTT-5’

18S_rRNA_fw 5’-GTAACCCGTTGAACCCCATT-3’

18S_rRNA_fw 3’-CCATCCAATCGGTAGTAGCG-3’

3.5.8 N-glycan extraction

A total of 15 million cells were seeded overnight and treated as indicated for 48 h. Cells were harvested, suspended 1 mL of HEPES homogenization buffer (0.25 M sucrose, 50 mM HEPES pH 7.5, 5 mM NaF, 5 mM EDTA, 2 mM DTT, cOmplete protease inhibitor), and lysed using a probe sonicator. Homogenate was cleared at 2,000 xg for 20 min at 4oC, then ultracentrifuged at

115,000 xg for 70 min at 4oC. The pellet was vigorously suspended in 650 μL Tris buffer (0.8%

Triton X-114, 50 mM Tris pH 7.5, 0.1 mM NaCl, 5 mM EDTA, 5 mM NaF, 2 mM DTT,

129 cOmplete protease inhibitor). The homogenate was chilled on ice for 10 min, incubated at 37oC for 20 min, then phase partitioned at 1,950 xg for 2 min at room temperature. The upper phase was discarded. Membrane proteins in the lower phase was precipitated with 1 mL acetone at -

20oC overnight.

Precipitated proteins were suspended in 60 μL of suspension buffer (0.25% RapiGest SF, 50 mM ammonium bicarbonate, 5 mM DTT). The completely dissolved solution was heated for 3 min at

85oC. Approximately 30 μg proteins was mixed with 0.5 μL of PNGase F, 0.7 μL of sialidase, and 20 μL of 50 mM ammonium bicarbonate, and incubated at 42oC for 2 h followed by 37oC overnight. Released N-glycans were extracted with 4-5 volumes of 100% ethanol at -80oC for 2 hours. The supernatant containing released N-glycans was speed vacuumed to dry.

Home-made porous graphitized carbon (PGC) microtips containing 10 mg PGC in a bed volume of 50 μL was washed with 500 μL of ddH2O, 500 μL of 80% acetonitrile (ACN), and equilibrated with 500 μL 0.1% trifluoroacetic acid (TFA). N-glycan pellets were dissolved in 50

μL of 0.1% TFA and slowly loaded into microtips at a flow rate of ~100 μL/min. Microtips were washed with 500 μL 0.1% TFA. N-glycans were eluted several times with 500 μL of elution buffer (0.05% TFA, 40% ACN). The eluted N-glycans were speed vacuumed to dry.

3.5.9 Total glycan analysis by LC-MS/MS

Analysis of the eluted N-glycans was modified from a previous method [34]. Total glycan samples were applied to a nano-HPLC Chip using a Agilent 1200 series microwell-plate autosampler, and interface with a Agilent 6550 Q-TOF MS (Agilent Technologies, Inc., Santa

Clara, CA). The HPLC Chip (glycan Chip) had a 40 nL enrichment column and a 75 μm x 43mm separation column packed with 5 μm graphitized carbon as stationary phase. The mobile phase was 0.1% formic acid in water (v/v) as solvent A, and 0.1% formic acid in ACN (v/v) as solvent

130

B. The flow rate at 0.3 μL/min with gradient schedule; 5% B (0−1 min); 5−20% B (1−15 min);

20−70% B(15−16 min); 70% B (16−19 min) and 70-5% B (19-20 min). PNGase F released free glycan was identified by Agilent Masshunter Quanlititive Analysis software by the presence of hexose and N-acetylhexosamine. Glycan structures were predict by online GlycoMod

(http://web.expasy.org/glycomod/). Agilent Masshunter Quantitative Analysis software was used to quantify the extracted glycan peaks.

3.5.10 Animal models

Animal work was carried out with the approval of the Princess Margaret Hospital ethics review board in accordance to the regulations of the Canadian Council on Animal Care. LM2-4 cells (1 million cells in 50 μL) were implanted subcutaneously in female SCID mice (6-8 wks), obtained in-house from the University Health Network animal colony. Primary tumours were measured every two days and calculated by (width x width x length)/2. After surgical removal of the primary tumours, animals were monitored daily for endpoint, including signs of metastatic load in the lung (laboured breathing). Treatment was given daily orally with PBS or 50 mg/kg/d fluvastatin. Necropsy was performed at endpoint where any tissues with evidence of metastatic disease is rapidly excised and fixed in formalin for histopathology.

3.5.11 Fluvastatin quantification by HPLC-MS/MS

Fluvastatin concentrations were determined by a modified HPLC-MS/MS method with atorvastatin as the internal standard (IS). Mouse serum or xenograft tissue were incubated with methyl tert-butyl ether for 30 min followed by centrifugation at 3000 rpm for 30 min. The supernatant was separated, dried at room temperature, and reconstituted in methanol/water (1:1).

The HPLC system consisted of a Shimadzu LC-20AD pump and a Shimadzu LC-20 AC autosampler (Shimadzu Corporation, Columbia, MD). The column used was a Phenomex

131 hyperclone BDS C18 column (50 × 2.0mm, 5 µm, Torrance, CA). The binary mobile phase consisted of mobile phase A: 5 mM ammonium acetate in water and mobile phase B: 5 mM ammonium acetate in acetonitrile. The gradient conditions for the mobile phase were as follows:

0.0-1.0 min, 20-100% B; 1.0-3.0 min, 100% B; 3.0-3.2 min, 100-10% B; 3.2-6.0 min, 20% B.

The flow rate was 0.5 ml/min. The HPLC system was interfaced to an Applied Biosystem MDS

Sciex triple quadrupole mass spectrometer (API 3200) (Applied Biosystems, Foster City, CA) operating in the negative electrospray ionization mode. For multiple-reaction monitoring, the transitions monitored were m/z 410.3 to 209.9 for fluvastatin, and m/z 557.0 to 278.1 for the IS

(atorvastatin). Data collection, peak integration, and calculation were performed using Applied

Biosystem MDS Analyst 1.4.2 software.

132

3.6 References

1. Chaffer CL, Weinberg RA: A perspective on cancer cell metastasis. Science 2011, 331(6024):1559-1564.

2. Weigelt B, Peterse JL, van 't Veer LJ: Breast cancer metastasis: markers and models. Nat Rev Cancer 2005, 5(8):591-602.

3. Bonadonna G, Valagussa P: Dose-response effect of adjuvant chemotherapy in breast cancer. N Engl J Med 1981, 304(1):10-15.

4. Ahern TP, Pedersen L, Tarp M, Cronin-Fenton DP, Garne JP, Silliman RA, Sorensen HT, Lash TL: Statin prescriptions and breast cancer recurrence risk: a Danish nationwide prospective cohort study. J Natl Cancer Inst 2011, 103(19):1461-1468.

5. Boudreau DM, Yu O, Chubak J, Wirtz HS, Bowles EJ, Fujii M, Buist DS: Comparative safety of cardiovascular medication use and breast cancer outcomes among women with early stage breast cancer. Breast Cancer Res Treat 2014, 144(2):405-416.

6. Chae YK, Valsecchi ME, Kim J, Bianchi AL, Khemasuwan D, Desai A, Tester W: Reduced risk of breast cancer recurrence in patients using ACE inhibitors, ARBs, and/or statins. Cancer Invest 2011, 29(9):585-593.

7. Kwan ML, Habel LA, Flick ED, Quesenberry CP, Caan B: Post-diagnosis statin use and breast cancer recurrence in a prospective cohort study of early stage breast cancer survivors. Breast Cancer Res Treat 2008, 109(3):573-579.

8. Goard CA, Chan-Seng-Yue M, Mullen PJ, Quiroga AD, Wasylishen AR, Clendening JW, Sendorek DH, Haider S, Lehner R, Boutros PC et al: Identifying molecular features that distinguish fluvastatin-sensitive breast tumor cells. Breast Cancer Res Treat 2014, 143(2):301-312.

9. Kusama T, Mukai M, Tatsuta M, Nakamura H, Inoue M: Inhibition of transendothelial migration and invasion of human breast cancer cells by preventing geranylgeranylation of Rho. Int J Oncol 2006, 29(1):217-223.

10. Pandyra AA, Mullen PJ, Goard CA, Ericson E, Sharma P, Kalkat M, Yu R, Pong JT, Brown KR, Hart T et al: Genome-wide RNAi analysis reveals that simultaneous inhibition of specific mevalonate pathway genes potentiates tumor cell death. Oncotarget 2015, 6(29):26909-26921.

11. Campbell MJ, Esserman LJ, Zhou Y, Shoemaker M, Lobo M, Borman E, Baehner F, Kumar AS, Adduci K, Marx C et al: Breast cancer growth prevention by statins. Cancer Res 2006, 66(17):8707-8714.

12. Li J, Liu J, Liang Z, He F, Yang L, Li P, Jiang Y, Wang B, Zhou C, Wang Y et al: Simvastatin and Atorvastatin inhibit DNA replication licensing factor MCM7 and effectively suppress RB-deficient tumors growth. Cell Death Dis 2017, 8(3):e2673.

133

13. Wolfe AR, Debeb BG, Lacerda L, Larson R, Bambhroliya A, Huang X, Bertucci F, Finetti P, Birnbaum D, Van Laere S et al: Simvastatin prevents triple-negative breast cancer metastasis in pre-clinical models through regulation of FOXO3a. Breast Cancer Res Treat 2015, 154(3):495-508.

14. Vintonenko N, Jais JP, Kassis N, Abdelkarim M, Perret GY, Lecouvey M, Crepin M, Di Benedetto M: Transcriptome analysis and in vivo activity of fluvastatin versus zoledronic acid in a murine breast cancer metastasis model. Mol Pharmacol 2012, 82(3):521-528.

15. Mandal CC, Ghosh-Choudhury N, Yoneda T, Choudhury GG: Simvastatin prevents skeletal metastasis of breast cancer by an antagonistic interplay between p53 and CD44. J Biol Chem 2011, 286(13):11314-11327.

16. Guerin E, Man S, Xu P, Kerbel RS: A model of postsurgical advanced metastatic breast cancer more accurately replicates the clinical efficacy of antiangiogenic drugs. Cancer Res 2013, 73(9):2743-2748.

17.. Goldstein JL, Brown MS: Regulation of the mevalonate pathway. Nature 1990, 343(6257):425-430.

18. Goldstein JL, Brown MS: A century of cholesterol and coronaries: from plaques to genes to statins. Cell 2015, 161(1):161-172.

19. Mullen PJ, Yu R, Longo J, Archer MC, Penn LZ: The interplay between cell signalling and the mevalonate pathway in cancer. Nat Rev Cancer 2016, 16(11):718-731.

20. Boyartchuk VL, Ashby MN, Rine J: Modulation of Ras and a-factor function by carboxyl-terminal proteolysis. Science 1997, 275(5307):1796-1800.

21. Clarke S, Vogel JP, Deschenes RJ, Stock J: Posttranslational modification of the Ha- ras oncogene protein: evidence for a third class of protein carboxyl methyltransferases. Proc Natl Acad Sci U S A 1988, 85(13):4643-4647.

22. Moores SL, Schaber MD, Mosser SD, Rands E, O'Hara MB, Garsky VM, Marshall MS, Pompliano DL, Gibbs JB: Sequence dependence of protein isoprenylation. J Biol Chem 1991, 266(22):14603-14610.

23. Elsayed M, Kobayashi D, Kubota T, Matsunaga N, Murata R, Yoshizawa Y, Watanabe N, Matsuura T, Tsurudome Y, Ogino T et al: Synergistic Antiproliferative Effects of Zoledronic Acid and Fluvastatin on Human Pancreatic Cancer Cell Lines: An in Vitro Study. Biol Pharm Bull 2016, 39(8):1238-1246.

24. Tsubaki M, Mashimo K, Takeda T, Kino T, Fujita A, Itoh T, Imano M, Sakaguchi K, Satou T, Nishida S: Statins inhibited the MIP-1alpha expression via inhibition of Ras/ERK and Ras/Akt pathways in myeloma cells. Biomed Pharmacother 2016, 78:23-29.

134

25. Rigoni M, Riganti C, Vitale C, Griggio V, Campia I, Robino M, Foglietta M, Castella B, Sciancalepore P, Buondonno I et al: Simvastatin and downstream inhibitors circumvent constitutive and stromal cell-induced resistance to doxorubicin in IGHV unmutated CLL cells. Oncotarget 2015, 6(30):29833-29846.

26. Tsubaki M, Takeda T, Sakamoto K, Shimaoka H, Fujita A, Itoh T, Imano M, Mashimo K, Fujiwara D, Sakaguchi K et al: Bisphosphonates and statins inhibit expression and secretion of MIP-1alpha via suppression of Ras/MEK/ERK/AML-1A and Ras/PI3K/Akt/AML-1A pathways. Am J Cancer Res 2015, 5(1):168-179.

27. Baas JM, Krens LL, ten Tije AJ, Erdkamp F, van Wezel T, Morreau H, Gelderblom H, Guchelaar HJ: Safety and efficacy of the addition of simvastatin to cetuximab in previously treated KRAS mutant metastatic colorectal cancer patients. Invest New Drugs 2015, 33(6):1242-1247.

28. Baas JM, Krens LL, Bos MM, Portielje JE, Batman E, van Wezel T, Morreau H, Guchelaar HJ, Gelderblom H: Safety and efficacy of the addition of simvastatin to panitumumab in previously treated KRAS mutant metastatic colorectal cancer patients. Anticancer Drugs 2015, 26(8):872-877.

29. Hong JY, Nam EM, Lee J, Park JO, Lee SC, Song SY, Choi SH, Heo JS, Park SH, Lim HY et al: Randomized double-blinded, placebo-controlled phase II trial of simvastatin and gemcitabine in advanced pancreatic cancer patients. Cancer Chemother Pharmacol 2014, 73(1):125-130.

30. Giancotti FG: Mechanisms governing metastatic dormancy and reactivation. Cell 2013, 155(4):750-764.

31. Atil B, Berger-Sieczkowski E, Bardy J, Werner M, Hohenegger M: In vitro and in vivo downregulation of the ATP binding cassette transporter B1 by the HMG-CoA reductase inhibitor simvastatin. Naunyn Schmiedebergs Arch Pharmacol 2016, 389(1):17-32.

32. Dricu A, Wang M, Hjertman M, Malec M, Blegen H, Wejde J, Carlberg M, Larsson O: Mevalonate-regulated mechanisms in cell growth control: role of dolichyl phosphate in expression of the insulin-like growth factor-1 receptor (IGF-1R) in comparison to Ras prenylation and expression of c-myc. Glycobiology 1997, 7(5):625-633.

33. Hamadmad SN, Hohl RJ: Lovastatin suppresses erythropoietin receptor surface expression through dual inhibition of glycosylation and geranylgeranylation. Biochem Pharmacol 2007, 74(4):590-600.

34. Williams AB, Li L, Nguyen B, Brown P, Levis M, Small D: Fluvastatin inhibits FLT3 glycosylation in human and murine cells and prolongs survival of mice with FLT3/ITD leukemia. Blood 2012, 120(15):3069-3079.

35. Cheung P, Dennis JW: Mgat5 and Pten interact to regulate cell growth and polarity. Glycobiology 2007, 17(7):767-773.

135

36. Granovsky M, Fata J, Pawling J, Muller WJ, Khokha R, Dennis JW: Suppression of tumor growth and metastasis in Mgat5-deficient mice. Nat Med 2000, 6(3):306-312.

37. Fernandes B, Sagman U, Auger M, Demetrio M, Dennis JW: Beta 1-6 branched oligosaccharides as a marker of tumor progression in human breast and colon neoplasia. Cancer Res 1991, 51(2):718-723.

38. Seelentag WK, Li WP, Schmitz SF, Metzger U, Aeberhard P, Heitz PU, Roth J: Prognostic value of beta1,6-branched oligosaccharides in human colorectal carcinoma. Cancer Res 1998, 58(23):5559-5564.

39. Kerbel RS, Guerin E, Francia G, Xu P, Lee CR, Ebos JM, Man S: Preclinical recapitulation of antiangiogenic drug clinical efficacies using models of early or late stage breast cancer metastatis. Breast 2013, 22 Suppl 2:S57-65.

40. Soule HD, Maloney TM, Wolman SR, Peterson WD, Jr., Brenz R, McGrath CM, Russo J, Pauley RJ, Jones RF, Brooks SC: Isolation and characterization of a spontaneously immortalized human breast epithelial cell line, MCF-10. Cancer Res 1990, 50(18):6075-6086.

41. Wong WW, Clendening JW, Martirosyan A, Boutros PC, Bros C, Khosravi F, Jurisica I, Stewart AK, Bergsagel PL, Penn LZ: Determinants of sensitivity to lovastatin-induced apoptosis in multiple myeloma. Mol Cancer Ther 2007, 6(6):1886-1897.

42. Taylor-Harding B, Orsulic S, Karlan BY, Li AJ: Fluvastatin and cisplatin demonstrate synergistic cytotoxicity in epithelial ovarian cancer cells. Gynecol Oncol 2010, 119(3):549-556.

43. Araki M, Maeda M, Motojima K: Hydrophobic statins induce autophagy and cell death in human rhabdomyosarcoma cells by depleting geranylgeranyl diphosphate. Eur J Pharmacol 2012, 674(2-3):95-103.

44. Finlay GA, Malhowski AJ, Liu Y, Fanburg BL, Kwiatkowski DJ, Toksoz D: Selective inhibition of growth of tuberous sclerosis complex 2 null cells by atorvastatin is associated with impaired Rheb and Rho GTPase function and reduced mTOR/S6 kinase activity. Cancer Res 2007, 67(20):9878-9886.

45. Yoshimura M, Nishikawa A, Ihara Y, Taniguchi S, Taniguchi N: Suppression of lung metastasis of B16 mouse melanoma by N-acetylglucosaminyltransferase III gene transfection. Proc Natl Acad Sci U S A 1995, 92(19):8754-8758.

46. Schwarz RT: Manipulation of the biosynthesis of protein-modifying glycoconjugates by the use of specific inhibitors. Behring Inst Mitt 1991(89):198-208.

47. Lehle L, Tanner W: The specific site of tunicamycin inhibition in the formation of dolichol-bound N-acetylglucosamine derivatives. FEBS Lett 1976, 72(1):167-170.

136

48. Chen Y, McMillan-Ward E, Kong J, Israels SJ, Gibson SB: Mitochondrial electron- transport-chain inhibitors of complexes I and II induce autophagic cell death mediated by reactive oxygen species. J Cell Sci 2007, 120(Pt 23):4155-4166.

49. De Craene B, Berx G: Regulatory networks defining EMT during cancer initiation and progression. Nat Rev Cancer 2013, 13(2):97-110.

50. Yang J, Weinberg RA: Epithelial-mesenchymal transition: at the crossroads of development and tumor metastasis. Dev Cell 2008, 14(6):818-829.

51. Thiery JP: Epithelial-mesenchymal transitions in tumour progression. Nat Rev Cancer 2002, 2(6):442-454.

52. Carlberg M, Dricu A, Blegen H, Wang M, Hjertman M, Zickert P, Hoog A, Larsson O: Mevalonic acid is limiting for N-linked glycosylation and translocation of the insulin-like growth factor-1 receptor to the cell surface. Evidence for a new link between 3-hydroxy-3-methylglutaryl-coenzyme a reductase and cell growth. J Biol Chem 1996, 271(29):17453-17462.

53. Chojnacki T, Dallner G: The biological role of dolichol. Biochem J 1988, 251(1):1-9.

54. Chesnokov V, Gong B, Sun C, Itakura K: Anti-cancer activity of glucosamine through inhibition of N-linked glycosylation. Cancer Cell Int 2014, 14:45.

55. Powlesland AS, Hitchen PG, Parry S, Graham SA, Barrio MM, Elola MT, Mordoh J, Dell A, Drickamer K, Taylor ME: Targeted glycoproteomic identification of cancer cell glycosylation. Glycobiology 2009, 19(8):899-909.

56. Zhang X, Yuan Y, Jiang L, Zhang J, Gao J, Shen Z, Zheng Y, Deng T, Yan H, Li W et al: Endoplasmic reticulum stress induced by tunicamycin and thapsigargin protects against transient ischemic brain injury: Involvement of PARK2-dependent mitophagy. Autophagy 2014, 10(10):1801-1813.

57. Oslowski CM, Urano F: Measuring ER stress and the unfolded protein response using mammalian tissue culture system. Methods Enzymol 2011, 490:71-92.

58. Fritz JM, Dong M, Apsley KS, Martin EP, Na CL, Sitaraman S, Weaver TE: Deficiency of the BiP cochaperone ERdj4 causes constitutive endoplasmic reticulum stress and metabolic defects. Mol Biol Cell 2014, 25(4):431-440.

59. Palamarczyk G, Butters TD: Uptake of dolichol into cultured cells. FEBS Lett 1982, 143(2):241-246.

60. Hamadmad SN, Hohl RJ: Lovastatin suppresses erythropoietin receptor surface expression through dual inhibition of glycosylation and geranylgeranylation. Biochem Pharmacol 2007, 74(4):590-600.

137

61. Xia Z, Tan MM, Wong WW, Dimitroulakos J, Minden MD, Penn LZ: Blocking protein geranylgeranylation is essential for lovastatin-induced apoptosis of human acute myeloid leukemia cells. Leukemia 2001, 15(9):1398-1407.

62. Bokoch GM, Prossnitz V: Isoprenoid metabolism is required for stimulation of the respiratory burst oxidase of HL-60 cells. J Clin Invest 1992, 89(2):402-408.

63. Dunnington DJ, Prichett W, Greig R: Stimulation of anchorage independent proliferation of human adrenocortical carcinoma cells by inhibition of cholesterol biosynthesis. Biochem Biophys Res Commun 1989, 165(1):219-225.

64. Cutts JL, Scallen TJ, Watson J, Bankhurst AD: Role of mevalonic acid in the regulation of natural killer cell cytotoxicity. J Cell Physiol 1989, 139(3):550-557.

65. Maltese WA, Sheridan KM: Differentiation of neuroblastoma cells induced by an inhibitor of mevalonate synthesis: relation of neurite outgrowth and acetylcholinesterase activity to changes in cell proliferation and blocked isoprenoid synthesis. J Cell Physiol 1985, 125(3):540-558.

66. Li X, Wang X, Tan Z, Chen S, Guan F: Role of Glycans in Cancer Cells Undergoing Epithelial-Mesenchymal Transition. Front Oncol 2016, 6:33.

67. Lange T, Samatov TR, Tonevitsky AG, Schumacher U: Importance of altered glycoprotein-bound N- and O-glycans for epithelial-to-mesenchymal transition and adhesion of cancer cells. Carbohydr Res 2014, 389:39-45.

68. Pinho SS, Reis CA: Glycosylation in cancer: mechanisms and clinical implications. Nat Rev Cancer 2015, 15(9):540-555.

69. Lau KS, Partridge EA, Grigorian A, Silvescu CI, Reinhold VN, Demetriou M, Dennis JW: Complex N-glycan number and degree of branching cooperate to regulate cell proliferation and differentiation. Cell 2007, 129(1):123-134.

70. Partridge EA, Le Roy C, Di Guglielmo GM, Pawling J, Cheung P, Granovsky M, Nabi IR, Wrana JL, Dennis JW: Regulation of cytokine receptors by Golgi N-glycan processing and endocytosis. Science 2004, 306(5693):120-124.

71. Abdel Rahman AM, Ryczko M, Nakano M, Pawling J, Rodrigues T, Johswich A, Taniguchi N, Dennis JW: Golgi N-glycan branching N- acetylglucosaminyltransferases I, V and VI promote nutrient uptake and metabolism. Glycobiology 2015, 25(2):225-240.

72. Cummings RD, Kornfeld S: Characterization of the structural determinants required for the high affinity interaction of asparagine-linked oligosaccharides with immobilized Phaseolus vulgaris leukoagglutinating and erythroagglutinating lectins. J Biol Chem 1982, 257(19):11230-11234.

138

73. Yoshimura M, Nishikawa A, Ihara Y, Taniguchi S, Taniguchi N: Suppression of lung metastasis of B16 mouse melanoma by N-acetylglucosaminyltransferase III gene transfection. Proc Natl Acad Sci U S A 1995, 92(19):8754-8758.

74. Carvalho S, Catarino TA, Dias AM, Kato M, Almeida A, Hessling B, Figueiredo J, Gartner F, Sanches JM, Ruppert T et al: Preventing E-cadherin aberrant N- glycosylation at Asn-554 improves its critical function in gastric cancer. Oncogene 2016, 35(13):1619-1631.

75. Zhao Y, Nakagawa T, Itoh S, Inamori K, Isaji T, Kariya Y, Kondo A, Miyoshi E, Miyazaki K, Kawasaki N et al: N-acetylglucosaminyltransferase III antagonizes the effect of N-acetylglucosaminyltransferase V on alpha3beta1 integrin-mediated cell migration. J Biol Chem 2006, 281(43):32122-32130.

76. Korourian S, Siegel E, Kieber-Emmons T, Monzavi-Karbassi B: Expression analysis of carbohydrate antigens in ductal carcinoma in situ of the breast by lectin histochemistry. BMC Cancer 2008, 8:136.

77. Terao M, Ishikawa A, Nakahara S, Kimura A, Kato A, Moriwaki K, Kamada Y, Murota H, Taniguchi N, Katayama I et al: Enhanced epithelial-mesenchymal transition-like phenotype in N-acetylglucosaminyltransferase V transgenic mouse skin promotes wound healing. J Biol Chem 2011, 286(32):28303-28311.

78. Miyoshi E, Nishikawa A, Ihara Y, Saito H, Uozumi N, Hayashi N, Fusamoto H, Kamada T, Taniguchi N: Transforming growth factor beta up-regulates expression of the N- acetylglucosaminyltransferase V gene in mouse melanoma cells. J Biol Chem 1995, 270(11):6216-6220.

79. Varki A, Cummings RD, Esko JD, Freeze HH, Stanley P, Bertozzi CR, Hart GW, Etzler ME: Essentials of Glycobiology, 2010/03/20 edn; 2009.

80. Breitling J, Aebi M: N-linked protein glycosylation in the endoplasmic reticulum. Cold Spring Harb Perspect Biol 2013, 5(8):a013359.

81. Parentini I, Cavallini G, Donati A, Gori Z, Bergamini E: Accumulation of dolichol in older tissues satisfies the proposed criteria to be qualified a biomarker of aging. J Gerontol A Biol Sci Med Sci 2005, 60(1):39-43.

82. Shang J, Gao N, Kaufman RJ, Ron D, Harding HP, Lehrman MA: Translation attenuation by PERK balances ER glycoprotein synthesis with lipid-linked oligosaccharide flux. J Cell Biol 2007, 176(5):605-616.

83. Bjarnadottir O, Romero Q, Bendahl PO, Jirstrom K, Ryden L, Loman N, Uhlen M, Johannesson H, Rose C, Grabau D et al: Targeting HMG-CoA reductase with statins in a window-of-opportunity breast cancer trial. Breast Cancer Res Treat 2013, 138(2):499-508.

84. Garwood ER, Kumar AS, Baehner FL, Moore DH, Au A, Hylton N, Flowers CI, Garber J, Lesnikoski BA, Hwang ES et al: Fluvastatin reduces proliferation and increases

139

apoptosis in women with high grade breast cancer. Breast Cancer Res Treat 2010, 119(1):137-144.

85. Sabia H, Prasad P, Smith HT, Stoltz RR, Rothenberg P: Safety, tolerability, and pharmacokinetics of an extended-release formulation of fluvastatin administered once daily to patients with primary hypercholesterolemia. J Cardiovasc Pharmacol 2001, 37(5):502-511.

86. Siekmeier R, Lattke P, Mix C, Park JW, Jaross W: Dose dependency of fluvastatin pharmacokinetics in serum determined by reversed phase HPLC. J Cardiovasc Pharmacol Ther 2001, 6(2):137-145.

87. Holstein SA, Knapp HR, Clamon GH, Murry DJ, Hohl RJ: Pharmacodynamic effects of high dose lovastatin in subjects with advanced malignancies. Cancer Chemother Pharmacol 2006, 57(2):155-164.

88. Wasylishen AR, Stojanova A, Oliveri S, Rust AC, Schimmer AD, Penn LZ: New model systems provide insights into Myc-induced transformation. Oncogene 2011, 30(34):3727-3734.

89. Morgenstern JP, Land H: Advanced mammalian gene transfer: high titre retroviral vectors with multiple drug selection markers and a complementary helper-free packaging cell line. Nucleic Acids Res 1990, 18(12):3587-3596.

140

Chapter 4 Discussion

141

Discussion 4.1 Challenges in the treatment of cancer

The ultimate goal of non-palliative cancer treatment strategies is to completely eradicate tumour cells. Several challenges exist that limit the chemotherapy treatment modality from achieving this goal, with or without concomitant surgery and/or radiotherapy. Major clinical challenges include intrinsic and acquired resistance, excessive toxicities, and issues with dosing. Resistance to targeted chemotherapies arises from complex intra-tumoural heterogeneity (1, 2), and an unanticipated large number of mutations capable of participating in oncogenic transformation (3,

4). As the majority of cancer-initiating mutations occur in regulators of core cellular processes

(cell fate, cell survival, and genome maintenance), the use of targeted chemotherapeutics is often associated with unexpected tissue-specific toxicities (5, 6). Additionally, effectiveness of chemotherapies can be greatly compromised by underdosing, an issue that is often overlooked.

Despite consistent evidence demonstrating that most elderly patients can both tolerate and benefit from standard chemotherapy regimens (7-10), fear of excessive toxicity in elderly patients remains a major contributor to dose reduction by >15% relative dose-intensity (RDI), in up to 40% of cancer patients (11, 12). The second population that is consistently undertreated is obese patients. This is because the dose of chemotherapeutics is calculated based on body surface area (BSA), but many oncologists either use “ideal”, rather than the actual body weight of the patient, to calculate BSA, or cap BSA at 2.0cm2 (13, 14). This leads to systematic underdosing of chemotherapy among overweight and obese patients (13), which could be contributing to the poorer outcomes often observed in this patient population (15, 16).

In addition to clinical challenges discussed above, several administrative difficulties could also undermine the effectiveness of chemotherapy, including access to care (eg. for intravenously

142 administered therapeutics), nonadherence (eg. due to poor patient education), and prohibitive cost (17). These clinical and administrative challenges are further compounded by aggressive screening and treatment practices (18), leading to an estimated 25% of mammographically detected breast cancers to be over-diagnosis (19), and an estimated 60% breast cancer cases to be over-treated with chemotherapy (20-22).

To address these challenges, significant effort has been put toward advancement of precision medicine, to account for individual variability in genomics, lifestyle, and environment of each person (23, 24). From the therapeutics perspective, precision medicine requires in-depth understanding of the efficacy, safety, mechanism of action, and biomarkers of response, for each drug used alone or in combination with others. In other words, to maximize therapeutic efficacy and minimize overtreatment, it is imperative to understand what tumours to treat with what drug(s), at what time, and why. In line with these overarching questions, this thesis provides novel insights to further our understanding of the anti-breast cancer effects of the statin family of drugs, described below.

4.2 Mechanism of statin-induced cancer cell death

The previous model of statin-induced cancer cell death is that statins inhibit the biosynthesis of

FPP and GGPP, thus limiting protein prenylation of the RAS superfamily (25, 26). In Chapter 2 of this thesis, we addressed the on-going discrepancies in this prenylation model of statin sensitivity, by building an experimental system that directly tests the association between demand for GGPP and FPP for protein prenylation, and sensitivity to fluvastatin. We showed that fluvastatin induces cancer cell death through inhibition of GGPP synthesis, but independently from inhibition of protein prenylation, thus suggesting that an alternative process downstream of GGPP underlines the mechanism of statin-induced cancer cell death.

143

In Chapter 3, we further delineated this GGPP-involving alternative process to be the biosynthesis of dolichol, a large isoprenoid end product of the MVA pathway that is functionally important for the co-translational N-glycosylation of nascent peptides. In particular, breast cancer cells that have undergone EMT become dependent on the dolichol-mediated expression of a class of EMT-associated N-glycans, leading to a vulnerability that can be exploited by statin treatment. Together, these data provide an explanation to why RAS superfamily proteins have been poor biomarkers of statin sensitivity in epidemiological, pre-clinical, and clinical studies, and suggest that glycan and glycoproteins biomarkers may better predict statin response, discussed below.

4.3 Biomarkers of statin sensitivity

Increased sensitivity to statin-induced cancer cell death has previously been associated with ERα status (27-29), activation of RAS family members (27, 30-32), p53 mutations (33, 34), MVA pathway gene expression (28, 35), and a 10-gene candidate signature derived from a panel of breast cancer cell lines with a heterogeneous response to statin treatment (28). In Chapter 2 of this thesis, we identified a gene-expression based biomarker of statin sensitivity, which exploits the bi-modal distribution of five core EMT-associated genes, namely CDH1, CDH2, FN1, VIM, and ZEB1. The selective use of bi-modally distributed genes allowed us to calculate biologically meaningful cut-off values for the binning of disease subsets (in this case, statin-sensitive and - insensitive cancer cell lines), which is superior to the traditional use of expression median as an arbitrary cut-off in genes with Gaussian expression (36, 37). However, the rationale behind EMT markers as predictors of statin sensitivity remains poorly understood, and will be an intriguing area of research to be further explored.

144

In Chapter 3, sensitivity to fluvastatin is shown to be associated with the EMT-associated β1,6- branched, tri- and tetra-antennary, unfucosylated and singly fucosylated complex glycans. These biomarkers can be collectively recognized by the lectin PHA-L, which we have shown to be readily adaptable for histochemical assays. This is particularly important, as histochemistry is within the capability of most cancer diagnostic labs, as opposed to assays based on gene expression and/or sequencing. Thus, once validated, histochemical PHA-L staining of a biopsy specimen may be rapidly advanced into the clinic. Identification of the glycoprotein biomarker may further improve the sensitivity and specificity of these assays.

4.4 Point of therapeutic intervention for statin use in breast cancer

Current pre-clinical models of breast cancer fall into four basic categories: (i) cell lines and traditional xenografts; (ii) patient-derived xenograft and organoids (PDXs and PDOs); (iii) transgenic animal models; and (iv) infusion models of metastatic breast cancer (Fig 4-1). Pre- clinical testing of drug sensitivities, including statins, are therefore limited to the points of therapeutic intervention represented by these models: primarily neo-adjuvant, preventative, and metastatic settings (Fig 4-1). However, our data on the association of breast cancer cell EMT and statin sensitivity indicate that statin use will be effective as an adjuvant therapeutic. We have thus carefully chosen a post-surgical model of breast cancer metastasis that mimics the patient treatment experience, and showed that fluvastatin treatment effectively delayed breast cancer metastasis and improved survival. Statins are particularly suitable for use in the adjuvant setting, due to their superior safety profiles and cost-effectiveness, thus providing the greatest clinical benefit for the treatment of breast cancer.

Additional considerations for the pre-clinical to clinical translation of statins lie in the question

145

Surgery Metastasis/ (90%) recurrence Diagnosis (20%)

Preventative Adjuvant (many years) Salvage

Neo-adjuvant (4-6 weeks)

Point of therapeutic Pre-clinical model intervention

Cell lines and traditional xenografts Neo-adjuvant

Patient-derived xenografts and Neo-adjuvant organoids (PDXs and PDOs) Preventative; Transgenic animals Neo-adjuvant

Infusion models of metastasis Salvage

Neo-adjuvant; Post-surgical model of metastasis Adjuvant

Figure 4-1. Pre-clinical models of the points of therapeutic intervention in breast cancer.

146 of which statin to prescribe to patients, and at what dose. Studies consistently show that lipophilic statins outperform hydrophilic statins in their anti-cancer activities, due to increased systemic circulation and decreased hepatoselectiveness (40). We and others have shown that some statins have additional bioactivities independent from inhibition of HMGCR (41, 42, 79).

This thesis has used exclusively fluvastatin for all biological studies, which was chosen based on superior pharmacological characteristics including high plasma concentration, and metabolism through a CYP3A4-independent mechanism (43, 44). In Chapter 2, our analysis of >700 cell lines for their sensitivity to three statin family members indicate that our observations with fluvastatin sensitivity can be extended to at least two other lipophilic statins, lovastatin and simvastatin. In Chapter 3, we further characterized the association between fluvastatin dose, serum concentration, and concentration in the xenografts tissue. We demonstrated that an equivalent dose of 4 mg/kg/d in humans is correlated to a plasma concentration of 4.9 μg/ml

(11.9 μM) fluvastatin, and a concentration of 586.2 ng/g reaching the tumour xenograft. Pending clinical validation, these data suggest that fluvastatin should be used at the maximally tolerated dose of 4 mg/kg/d (45), rather than the cholesterol-lowering dose (approximately 1 mg/kg/d), to effectively delay breast cancer metastasis and impact patient care. Alternatively, in a truly personalized treatment model, fluvastatin could be dose-escalated on an individual-by-individual basis until a stable plasma concentration of 11.9 μM is reached without major safety concerns.

4.5 Perspectives 4.5.1 Limitations of this work and future directions

The role of EMT in cancer progression. The bulk of the cell culture experiments performed in

Chapters 2 and 3 of this thesis use EMT as a surrogate indicator of the metastatic capability of breast cancer cells. However, the role of EMT in cancer progression, and particularly in cancer

147 metastasis, is hotly debated. EMT was proposed to be a prerequisite for cancer cell metastasis in order to explain why, in carcinomas bound by epithelial characteristics such as tight junctions and desmosomes, increasingly invasive cells could dissolve underlying basement membranes, invade adjacent stromal compartments, and eventually intravasate into the draining lymph nodes and/or systemic circulation (46). However, to date accumulating data disagree on whether the acquisition of these malignant traits is a required series of events for tumour progression, or an accidental consequence thereof (46-51). Moreover, metastatic outgrowths tend to exhibit epithelial phenotypes (46, 51). Proponents of the EMT model of cancer metastasis suggest that cells that have escaped the primary tumour site must have a mechanism to regain epithelial characteristics, a process termed mesenchymal-to-epithelial transition (MET). However, experimental evidence of MET following EMT has been limited (46, 51).

Cancer cell EMT has also been shown to share characteristics with cancer cell hypoxia, therapeutic resistance, dormancy, and various definitions of stemness (46, 51-53). Again, it is unknown if EMT is a prerequisite or a byproduct of these aggressive phenotypes, or if all of these phenotypes (including EMT) are functionally equivalent, and if so, to what degree (52-55).

In Chapter 3, we complemented our cell culture approach with a clinically relevant in vivo model of post-surgical breast cancer metastasis, and showed that fluvastatin can delay breast cancer metastasis when used in the adjuvant setting. The use of additional in vivo models of breast cancer metastasis independent of EMT will be necessary to address whether the anti-breast cancer effects of statins are specific to cancer cell EMT, or to cancer metastasis in general.

Examples of such a model include the transgenic breast cancer lineage-tracing models developed by Fischer et al. (49), which crosses the highly metastatic MMTV-PyMT or MMTV-Neu with

Vim-Cre or Fsp1-Cre mesenchymal-specific reporters, or the model developed by Zheng et al.

(50), where EMT-inducers Twist1 or Snail1 were deleted in the KPC transgenic mouse model of

148

PDAC. In both models, cancer cell EMT was not required for metastasis, but mediated resistance to common chemotherapeutics (49, 50). Characterizing the fluvastatin sensitivity of these cells will significantly impact the translation of statins into the cancer clinic.

Glycans and glycoproteins as cancer biomarkers. Glycosylation of proteins occur on an estimated 50% of total proteins, and up to 80% in some glycoprotein-rich fluids (56). As most of the FDA-approved cancer biomarkers have been proteins derived from the serum (57), it is not surprising that the majority of these biomarkers are glycosylated. Importantly, for many of these protein biomarkers, different glycoforms are expressed in normal and in tumour cells, which could provide a valuable biochemical tool for diagnosis. For example, the AFP-L3% assay exploits the specific expression of the L3 glycoform of AFP in HCC, in order to differentiate between patients with HCC or chronic liver disease, which specifically express the L1 glycoform

(38, 39). Similarly, glycosylation of PSA secreted by the prostate tumour cell line LNCaP is significantly different from PSA present in the normal seminal plasma (58), and glycosylation of human pancreatic ribonuclease 1 (RNase 1) isolated from healthy pancreas and pancreatic cancer cell lines are distinct (59).

In Chapter 3 of this thesis, we assessed global glycosylation changes in breast cancer cells after fluvastatin treatment using MS-based assays and PHA-L binding assays, both in tissue culture and in tumour xenografts. Although this was immensely useful and allowed the identification of

β1,6-branched, tri- and tetra-antennary, unfucosylated and singly fucosylated complex N-glycans as biomarkers of fluvastatin response, the drawback of these approaches is that the glycan cannot be traced back to the glycoprotein and the glycosylation site of origin. Unfortunately, identification of glycopeptides is challenged by several technical difficulties, such as identification of monosaccharides with the same mass, low efficiencies in glycopeptide

149 fragmentation, and poor algorithms to identify glycopeptide spectra (56, 60). Nevertheless, progress is being made to address all of these limitations, and several work-arounds exist to obtain glycosylation information from intact proteins (top-down approaches), and to identify peptides after enrichment for a glycan moiety (bottom-up approaches) (56, 61-63).

An immediate next step to follow up on the identification of glycoproteins affected by fluvastatin treatment is to identify the site(s) of glycosylation where occupancy is reduced by fluvastatin treatment. This analysis involves releasing N-glycans using PNGase F, which deaminates the glycosylated asparagine residues (N) to aspartic acid (D), leading to a difference in mass by

0.984Da that can be distinguished by MS (64). More sophisticated methods, if needed, include glycan-tagging through click chemistry reactions between azide-labeled monosaccharides and dibenzocyclooctyne group (DBCO)-linked tags (usually biotin), which allows for both identification of glycosylated peptides, and their quantification if combined with SILAC (stable isotope labeling with amino acids in cell culture) (65, 66). Validation of the identified glycoprotein biomarkers of fluvastatin response will be prioritized by cross-referencing with known breast cancer-associated antigens that are reactive to PHA-L (67), and performed by immunoblotting and IHC on cell lysates and FFPE samples of tumour xenografts. Knockdown and overexpression of validated biomarkers may further provide an indication of their functional relevance to statin-induced cancer cell death.

Combination strategies. Prescribing cancer chemotherapeutics in rationally identified combination partners could have multiple benefits, from improving therapeutic efficacy to potentially preventing drug resistance. Statins has been combined with several currently available chemotherapies and targeted agents, demonstrating both synergism (41, 68-71) and improved safety profiles in vivo (70, 72-75). As statins are unlikely to replace the current

150 standard of care, a major limitation in the studies described in Chapter 3 of this thesis is that the anti-breast cancer effect of fluvastatin treatment was not tested in combination with chemotherapeutics currently prescribed in the adjuvant setting. We are immediately equipped to address this question using the same model of post-surgical metastatic breast cancer, or using the transgenic breast cancer models described above. Importantly, it should be noted that in both the post-surgical model and the transgenic models, classic chemotherapeutics treatment strategies were ineffective at reducing tumour burden or prolonging survival (49, 76-78), although metronomic delivery of these chemotherapeutics, alone or in combination with each other, were effective in the post-surgical model (76-78). This is in line with the epidemiology of breast cancers, where approximately 20% of patients receiving surgery followed by adjuvant chemotherapy and/or radiotherapy have experienced recurrence, despite aggressive treatment.

Demonstrating the efficacy and safety of adjuvant statin use in combination with current standard-of-care chemotherapeutics will have immediate impact on patient care.

4.5.2 Anticipated impact

In this thesis, we have provided strong pre-clinical evidence for the repurposing of the statin family of drugs as anti-metastatic breast cancer chemotherapeutics. We have resolved a key discrepancy in the mechanism of action for the anti-cancer effects of statins, which is critical for the interpretation of epidemiological and pre-clinical studies where efficacy of statins was poorly associated with previously proposed biomarker(s). We have identified two new biomarkers of statin sensitivity, one based on gene expression and one based on glycan presentation, and demonstrated that statin use as an adjuvant therapeutic effectively delayed breast cancer metastasis. Statins are readily available, well-tolerated, and affordable drugs that can be rapidly repurposed into the breast cancer clinic. This body of work provides new insights on the efficacy, mechanism of action, and biomarkers of statin sensitivity for this purpose. In the era of

151 personalized medicine, these insights will facilitate the design of clinical trials to evaluate statins as anti-breast cancer agents, and ultimately to impact patient care.

152

4.6 References

1. Yap TA, Gerlinger M, Futreal PA, Pusztai L, Swanton C: Intratumor heterogeneity: seeing the wood for the trees. Sci Transl Med 2012, 4(127):127ps110.

2. Gerlinger M, Rowan AJ, Horswell S, Larkin J, Endesfelder D, Gronroos E, Martinez P, Matthews N, Stewart A, Tarpey P et al: Intratumor heterogeneity and branched evolution revealed by multiregion sequencing. N Engl J Med 2012, 366(10):883-892.

3. Vogelstein B, Papadopoulos N, Velculescu VE, Zhou S, Diaz LA, Jr., Kinzler KW: Cancer genome landscapes. Science 2013, 339(6127):1546-1558.

4. Dienstmann R, Jang IS, Bot B, Friend S, Guinney J: Database of genomic biomarkers for cancer drugs and clinical targetability in solid tumors. Cancer Discov 2015, 5(2):118-123.

5. Ding PN, Lord SJ, Gebski V, Links M, Bray V, Gralla RJ, Yang JC, Lee CK: Risk of Treatment-Related Toxicities from EGFR Tyrosine Kinase Inhibitors: A Meta- analysis of Clinical Trials of Gefitinib, Erlotinib, and Afatinib in Advanced EGFR- Mutated Non-Small Cell Lung Cancer. J Thorac Oncol 2016.

6. Force T, Kerkela R: Cardiotoxicity of the new cancer therapeutics--mechanisms of, and approaches to, the problem. Drug Discov Today 2008, 13(17-18):778-784.

7. Shayne M, Culakova E, Poniewierski MS, Wolff D, Dale DC, Crawford J, Lyman GH: Dose intensity and hematologic toxicity in older cancer patients receiving systemic chemotherapy. Cancer 2007, 110(7):1611-1620.

8. Kuderer NM, Dale DC, Crawford J, Lyman GH: Impact of primary prophylaxis with granulocyte colony-stimulating factor on febrile neutropenia and mortality in adult cancer patients receiving chemotherapy: a systematic review. J Clin Oncol 2007, 25(21):3158-3167.

9. Smith TJ, Khatcheressian J, Lyman GH, Ozer H, Armitage JO, Balducci L, Bennett CL, Cantor SB, Crawford J, Cross SJ et al: 2006 update of recommendations for the use of white blood cell growth factors: an evidence-based clinical practice guideline. J Clin Oncol 2006, 24(19):3187-3205.

10. Balducci L, Hardy CL, Lyman GH: Hemopoiesis and aging. Cancer Treat Res 2005, 124:109-134.

11. Lyman GH, Dale DC, Friedberg J, Crawford J, Fisher RI: Incidence and predictors of low chemotherapy dose-intensity in aggressive non-Hodgkin's lymphoma: a nationwide study. J Clin Oncol 2004, 22(21):4302-4311.

12. Lyman GH, Dale DC, Crawford J: Incidence and predictors of low dose-intensity in adjuvant breast cancer chemotherapy: a nationwide study of community practices. J Clin Oncol 2003, 21(24):4524-4531.

153

13. Field KM, Kosmider S, Jefford M, Michael M, Jennens R, Green M, Gibbs P: Chemotherapy dosing strategies in the obese, elderly, and thin patient: results of a nationwide survey. J Oncol Pract 2008, 4(3):108-113.

14. Sparreboom A, Wolff AC, Mathijssen RH, Chatelut E, Rowinsky EK, Verweij J, Baker SD: Evaluation of alternate size descriptors for dose calculation of anticancer drugs in the obese. J Clin Oncol 2007, 25(30):4707-4713.

15. Colleoni M, Li S, Gelber RD, Price KN, Coates AS, Castiglione-Gertsch M, Goldhirsch A: Relation between chemotherapy dose, oestrogen receptor expression, and body- mass index. Lancet 2005, 366(9491):1108-1110.

16. Rosner GL, Hargis JB, Hollis DR, Budman DR, Weiss RB, Henderson IC, Schilsky RL: Relationship between toxicity and obesity in women receiving adjuvant chemotherapy for breast cancer: results from cancer and leukemia group B study 8541. J Clin Oncol 1996, 14(11):3000-3008.

17. McCue DA, Lohr LK, Pick AM: Improving adherence to oral cancer therapy in clinical practice. Pharmacotherapy 2014, 34(5):481-494.

18. Esserman LJ, Thompson IM, Reid B, Nelson P, Ransohoff DF, Welch HG, Hwang S, Berry DA, Kinzler KW, Black WC et al: Addressing overdiagnosis and overtreatment in cancer: a prescription for change. Lancet Oncol 2014, 15(6):e234-242.

19. Welch HG, Black WC: Overdiagnosis in cancer. J Natl Cancer Inst 2010, 102(9):605- 613.

20. Gokmen-Polar Y, Badve S: Molecular profiling assays in breast cancer: are we ready for prime time? Oncology (Williston Park) 2012, 26(4):350-357, 361.

21. Piccart-Gebhart MJ: New developments in hormone receptor-positive disease. Oncologist 2011, 16 Suppl 1:40-50.

22. Straehle C, Cardoso F, Azambuja E, Dolci S, Meirsman L, Vantongelen K, Ignatiadis M, Sotiriou C, Piccart-Gebhart MJ: Better translation from bench to bedside: breakthroughs in the individualized treatment of cancer. Crit Care Med 2009, 37(1 Suppl):S22-29.

23. Lu YF, Goldstein DB, Angrist M, Cavalleri G: Personalized medicine and human genetic diversity. Cold Spring Harb Perspect Med 2014, 4(9):a008581.

24. Garay JP, Gray JW: Omics and therapy - a basis for precision medicine. Mol Oncol 2012, 6(2):128-139.

25. Wong WW, Clendening JW, Martirosyan A, Boutros PC, Bros C, Khosravi F, Jurisica I, Stewart AK, Bergsagel PL, Penn LZ: Determinants of sensitivity to lovastatin-induced apoptosis in multiple myeloma. Mol Cancer Ther 2007, 6(6):1886-1897.

154

27. Campbell MJ, Esserman LJ, Zhou Y, Shoemaker M, Lobo M, Borman E, Baehner F, Kumar AS, Adduci K, Marx C et al: Breast cancer growth prevention by statins. Cancer Res 2006, 66(17):8707-8714.

28. Goard CA, Chan-Seng-Yue M, Mullen PJ, Quiroga AD, Wasylishen AR, Clendening JW, Sendorek DH, Haider S, Lehner R, Boutros PC et al: Identifying molecular features that distinguish fluvastatin-sensitive breast tumor cells. Breast Cancer Res Treat 2014, 143(2):301-312.

29. Mueck AO, Seeger H, Wallwiener D: Effect of statins combined with estradiol on the proliferation of human receptor-positive and receptor-negative breast cancer cells. Menopause 2003, 10(4):332-336.

30. Garnett DJ: Caveolae as a target to quench autoinduction of the metastatic phenotype in lung cancer. J Cancer Res Clin Oncol 2016, 142(3):611-618.

31. Kang S, Kim ES, Moon A: Simvastatin and lovastatin inhibit breast cell invasion induced by H-Ras. Oncol Rep 2009, 21(5):1317-1322.

32. Kusama T, Mukai M, Tatsuta M, Nakamura H, Inoue M: Inhibition of transendothelial migration and invasion of human breast cancer cells by preventing geranylgeranylation of Rho. Int J Oncol 2006, 29(1):217-223.

33. Freed-Pastor WA, Mizuno H, Zhao X, Langerod A, Moon SH, Rodriguez-Barrueco R, Barsotti A, Chicas A, Li W, Polotskaia A et al: Mutant p53 disrupts mammary tissue architecture via the mevalonate pathway. Cell 2012, 148(1-2):244-258.

34. Sorrentino G, Ruggeri N, Specchia V, Cordenonsi M, Mano M, Dupont S, Manfrin A, Ingallina E, Sommaggio R, Piazza S et al: Metabolic control of YAP and TAZ by the mevalonate pathway. Nat Cell Biol 2014, 16(4):357-366.

35. Bjarnadottir O, Romero Q, Bendahl PO, Jirstrom K, Ryden L, Loman N, Uhlen M, Johannesson H, Rose C, Grabau D et al: Targeting HMG-CoA reductase with statins in a window-of-opportunity breast cancer trial. Breast Cancer Res Treat 2013, 138(2):499-508.

36. Santarpia L, Iwamoto T, Di Leo A, Hayashi N, Bottai G, Stampfer M, Andre F, Turner NC, Symmans WF, Hortobagyi GN et al: DNA repair gene patterns as prognostic and predictive factors in molecular breast cancer subtypes. Oncologist 2013, 18(10):1063- 1073.

37. Kernagis DN, Hall AH, Datto MB: Genes with bimodal expression are robust diagnostic targets that define distinct subtypes of epithelial ovarian cancer with different overall survival. J Mol Diagn 2012, 14(3):214-222.

155

38. Kamoto T, Satomura S, Yoshiki T, Okada Y, Henmi F, Nishiyama H, Kobayashi T, Terai A, Habuchi T, Ogawa O: Lectin-reactive alpha-fetoprotein (AFP-L3%) curability and prediction of clinical course after treatment of non-seminomatous germ cell tumors. Jpn J Clin Oncol 2002, 32(11):472-476.

39. Li D, Mallory T, Satomura S: AFP-L3: a new generation of tumor marker for hepatocellular carcinoma. Clin Chim Acta 2001, 313(1-2):15-19.

40. Chen C, Lin J, Smolarek T, Tremaine L: P-glycoprotein has differential effects on the disposition of statin acid and lactone forms in mdr1a/b knockout and wild-type mice. Drug Metab Dispos 2007, 35(10):1725-1729.

41. Goard CA, Mather RG, Vinepal B, Clendening JW, Martirosyan A, Boutros PC, Sharom FJ, Penn LZ: Differential interactions between statins and P-glycoprotein: implications for exploiting statins as anticancer agents. Int J Cancer 2010, 127(12):2936-2948.

42. Rao S, Porter DC, Chen X, Herliczek T, Lowe M, Keyomarsi K: Lovastatin-mediated G1 arrest is through inhibition of the proteasome, independent of hydroxymethyl glutaryl-CoA reductase. Proceedings of the National Academy of Sciences of the United States of America 1999, 96(14):7797-7802.

43. Bockorny B, Dasanu CA: HMG-CoA reductase inhibitors as adjuvant treatment for hematologic malignancies: what is the current evidence? Ann Hematol 2015, 94(1):1- 12.

44. Gazzerro P, Proto MC, Gangemi G, Malfitano AM, Ciaglia E, Pisanti S, Santoro A, Laezza C, Bifulco M: Pharmacological actions of statins: a critical appraisal in the management of cancer. Pharmacol Rev 2012, 64(1):102-146.

45. Sabia H, Prasad P, Smith HT, Stoltz RR, Rothenberg P: Safety, tolerability, and pharmacokinetics of an extended-release formulation of fluvastatin administered once daily to patients with primary hypercholesterolemia. J Cardiovasc Pharmacol 2001, 37(5):502-511.

46. Chaffer CL, Weinberg RA: A perspective on cancer cell metastasis. Science 2011, 331(6024):1559-1564.

47. Shamir ER, Pappalardo E, Jorgens DM, Coutinho K, Tsai WT, Aziz K, Auer M, Tran PT, Bader JS, Ewald AJ: Twist1-induced dissemination preserves epithelial identity and requires E-cadherin. J Cell Biol 2014, 204(5):839-856.

48. Mani SA, Guo W, Liao MJ, Eaton EN, Ayyanan A, Zhou AY, Brooks M, Reinhard F, Zhang CC, Shipitsin M et al: The epithelial-mesenchymal transition generates cells with properties of stem cells. Cell 2008, 133(4):704-715.

49. Fischer KR, Durrans A, Lee S, Sheng J, Li F, Wong ST, Choi H, El Rayes T, Ryu S, Troeger J et al: Epithelial-to-mesenchymal transition is not required for lung metastasis but contributes to chemoresistance. Nature 2015, 527(7579):472-476.

156

50. Zheng X, Carstens JL, Kim J, Scheible M, Kaye J, Sugimoto H, Wu CC, LeBleu VS, Kalluri R: Epithelial-to-mesenchymal transition is dispensable for metastasis but induces chemoresistance in pancreatic cancer. Nature 2015, 527(7579):525-530.

51. Heerboth S, Housman G, Leary M, Longacre M, Byler S, Lapinska K, Willbanks A, Sarkar S: EMT and tumor metastasis. Clin Transl Med 2015, 4:6.

52. Ibrahim AA, Schmithals C, Kowarz E, Koberle V, Kakoschky B, Pleli T, Kollmar O, Nitsch S, Waidmann O, Finkelmeier F et al: Hypoxia causes down-regulation of Dicer in hepatocellular carcinoma, which is required for up-regulation of hypoxia inducible factor 1alpha and epithelial-mesenchymal transition. Clin Cancer Res 2017.

53. van den Beucken T, Koch E, Chu K, Rupaimoole R, Prickaerts P, Adriaens M, Voncken JW, Harris AL, Buffa FM, Haider S et al: Hypoxia promotes stem cell phenotypes and poor prognosis through epigenetic regulation of DICER. Nat Commun 2014, 5:5203.

54. Monteiro J, Fodde R: Cancer stemness and metastasis: therapeutic consequences and perspectives. Eur J Cancer 2010, 46(7):1198-1203.

55. Patel P, Chen EI: Cancer stem cells, tumor dormancy, and metastasis. Front Endocrinol (Lausanne) 2012, 3:125.

56. Kailemia MJ, Park D, Lebrilla CB: Glycans and glycoproteins as specific biomarkers for cancer. Anal Bioanal Chem 2017, 409(2):395-410.

57. Ludwig JA, Weinstein JN: Biomarkers in cancer staging, prognosis and treatment selection. Nat Rev Cancer 2005, 5(11):845-856.

58. Peracaula R, Tabares G, Royle L, Harvey DJ, Dwek RA, Rudd PM, de Llorens R: Altered glycosylation pattern allows the distinction between prostate-specific antigen (PSA) from normal and tumor origins. Glycobiology 2003, 13(6):457-470.

59. Peracaula R, Royle L, Tabares G, Mallorqui-Fernandez G, Barrabes S, Harvey DJ, Dwek RA, Rudd PM, de Llorens R: Glycosylation of human pancreatic ribonuclease: differences between normal and tumor states. Glycobiology 2003, 13(4):227-244.

60. Strum JS, Nwosu CC, Hua S, Kronewitter SR, Seipert RR, Bachelor RJ, An HJ, Lebrilla CB: Automated assignments of N- and O-site specific glycosylation with extensive glycan heterogeneity of glycoprotein mixtures. Anal Chem 2013, 85(12):5666-5675.

61. Ito S, Hayama K, Hirabayashi J: Enrichment strategies for glycopeptides. Methods Mol Biol 2009, 534:195-203.

62. Siuti N, Kelleher NL: Decoding protein modifications using top-down mass spectrometry. Nat Methods 2007, 4(10):817-821.

157

63. Reid GE, Stephenson JL, Jr., McLuckey SA: Tandem mass spectrometry of ribonuclease A and B: N-linked glycosylation site analysis of whole protein ions. Anal Chem 2002, 74(3):577-583.

64. Zhu Z, Go EP, Desaire H: Absolute quantitation of glycosylation site occupancy using isotopically labeled standards and LC-MS. J Am Soc Mass Spectrom 2014, 25(6):1012-1017.

65. Xiao H, Tang GX, Wu R: Site-Specific Quantification of Surface N-Glycoproteins in Statin-Treated Liver Cells. Anal Chem 2016, 88(6):3324-3332.

66. Dehnert KW, Beahm BJ, Huynh TT, Baskin JM, Laughlin ST, Wang W, Wu P, Amacher SL, Bertozzi CR: Metabolic labeling of fucosylated glycans in developing zebrafish. ACS Chem Biol 2011, 6(6):547-552.

67. Abbott KL, Aoki K, Lim JM, Porterfield M, Johnson R, O'Regan RM, Wells L, Tiemeyer M, Pierce M: Targeted glycoproteomic identification of biomarkers for human breast carcinoma. J Proteome Res 2008, 7(4):1470-1480.

68. Roudier E, Mistafa O, Stenius U: Statins induce mammalian target of rapamycin (mTOR)-mediated inhibition of Akt signaling and sensitize p53-deficient cells to cytostatic drugs. Mol Cancer Ther 2006, 5(11):2706-2715.

69. Feleszko W, Mlynarczuk I, Olszewska D, Jalili A, Grzela T, Lasek W, Hoser G, Korczak-Kowalska G, Jakobisiak M: Lovastatin potentiates antitumor activity of doxorubicin in murine melanoma via an apoptosis-dependent mechanism. Int J Cancer 2002, 100(1):111-118.

70. Feleszko W, Mlynarczuk I, Balkowiec-Iskra EZ, Czajka A, Switaj T, Stoklosa T, Giermasz A, Jakobisiak M: Lovastatin potentiates antitumor activity and attenuates cardiotoxicity of doxorubicin in three tumor models in mice. Clin Cancer Res 2000, 6(5):2044-2052.

71. Pandyra A, Mullen PJ, Kalkat M, Yu R, Pong JT, Li Z, Trudel S, Lang KS, Minden MD, Schimmer AD et al: Immediate utility of two approved agents to target both the metabolic mevalonate pathway and its restorative feedback loop. Cancer research 2014, 74(17):4772-4782.

72. Damrot J, Nubel T, Epe B, Roos WP, Kaina B, Fritz G: Lovastatin protects human endothelial cells from the genotoxic and cytotoxic effects of the anticancer drugs doxorubicin and etoposide. Br J Pharmacol 2006, 149(8):988-997.

73. Arlt A, Vorndamm J, Breitenbroich M, Folsch UR, Kalthoff H, Schmidt WE, Schafer H: Inhibition of NF-kappaB sensitizes human pancreatic carcinoma cells to apoptosis induced by etoposide (VP16) or doxorubicin. Oncogene 2001, 20(7):859-868.

74. Morgan MA, Sebil T, Aydilek E, Peest D, Ganser A, Reuter CW: Combining prenylation inhibitors causes synergistic cytotoxicity, apoptosis and disruption of

158

RAS-to-MAP kinase signalling in multiple myeloma cells. Br J Haematol 2005, 130(6):912-925.

75. Wojtkowiak JW, Fouad F, LaLonde DT, Kleinman MD, Gibbs RA, Reiners JJ, Jr., Borch RF, Mattingly RR: Induction of apoptosis in neurofibromatosis type 1 malignant peripheral nerve sheath tumor cell lines by a combination of novel farnesyl transferase inhibitors and lovastatin. J Pharmacol Exp Ther 2008, 326(1):1-11.

76. Di Desidero T, Xu P, Man S, Bocci G, Kerbel RS: Potent efficacy of metronomic topotecan and pazopanib combination therapy in preclinical models of primary or late stage metastatic triple-negative breast cancer. Oncotarget 2015, 6(40):42396- 42410.

77. Kerbel RS, Guerin E, Francia G, Xu P, Lee CR, Ebos JM, Man S: Preclinical recapitulation of antiangiogenic drug clinical efficacies using models of early or late stage breast cancer metastatis. Breast 2013, 22 Suppl 2:S57-65.

78. Munoz R, Man S, Shaked Y, Lee CR, Wong J, Francia G, Kerbel RS: Highly efficacious nontoxic preclinical treatment for advanced metastatic breast cancer using combination oral UFT-cyclophosphamide metronomic chemotherapy. Cancer Res 2006, 66(7):3386-3391.

79. Martirosyan A, Clendening JW, Goard CA, Penn LZ: Lovastatin induces apoptosis of ovarian cancer cells and synergizes with doxorubicin: potential therapeutic relevance. BMC Cancer 2010, 10:103.

159

Appendix Positive feedback regulation of the mevalonate pathway

Contributions: The work presented in this chapter was performed by the author.

This project was completed with supervisory support from Dr. Linda Z. Penn.

160

Appendix – Positive feedback regulation of the mevalonate pathway 5.1 Introduction

The mevalonate (MVA) pathway is regulated by a robust negative feedback system (Fig 5-1A) that was first described in 1933 (1). In this landmark study, mice placed in sealed bottles did not synthesize cholesterol when their diet contained cholesterol, but began synthesizing it de novo when their diet was cholesterol-free (1). Later it was elucidated that, when cholesterol levels are low, cells not only increase the expression of MVA pathway enzymes (eg. HMGCR and

HMGCS1) to increase cholesterol synthesis (Fig 5-1A), but also upregulate LDLR to increase cholesterol uptake from the extracellular milieu (2-4). This negative feedback loop is exploited by the statins family of drugs, which blocks de novo cholesterol synthesis by inhibiting HMGCR

(5, 6). In response to a falling rate of cholesterol synthesis, cells increase the expression of

LDLR, leading to a decrease in LDL-cholesterol in the plasma (2, 7). This prevents build-up of atherosclerotic plaques in the arteries, a major risk factor for cardiovascular diseases (8).

Negative feedback loops are widely used by biological systems to maintain homeostasis (9, 10).

For example, each non-essential amino acid can allosterically inhibit an enzyme in its biosynthetic pathway, often the rate-limiting one, at high concentration (11). When the concentration of an amino acid is low, this inhibition is lifted, and cells begin synthesizing more of the amino acid (11). Negative feedback can also occur at the level of transcription, such as increased expression of System A transporters during amino acid starvation (12, 13), and of

GLUT1 when glucose concentrations are low (14). Similarly, the negative feedback loop of the

MVA pathway is also regulated at the level of gene transcription, mediated by the transcription factor SREBP2 (7) (Fig 5-1A). When cholesterol levels are high, SREBP2 is held in a latent form in the ER. When cholesterol levels are low, SREBP2 is activated by protein cleavage,

161

A B C

A A A

B B B

Acetyl-CoA lac Glucose repressor

Acetoacetyl-CoA 3PG HMGCS1 PGAM1 lac HMG-CoA transporter HMGCR PEP Isoprenoids MVA PKM2 SREBP2 Pyruvate Cholesterol

Figure 5-1. Two-component mechanisms of feedback regulation.

A, schematic of a negative feedback loop. The MVA pathway is regulated by a negative feedback loop. B, schematic of a positive feedback loop consisting of double-negative regulation. The bacterial lac operon is regulated by this mechanism. C, schematic of a true positive feedback loop. Cancer cells are known to regulate glycolysis through positive feedback.

162 translocates into the nucleus, and induces transcription of MVA pathway genes as well as LDLR

(2, 7).

In contrast to negative feedback, positive feedback occurs when a deviation from homeostasis is further amplified by the control system (Fig 5-1B, C). Positive feedback commonly functions as a cascading mechanism such as in cell signaling (15, 16), and as a transcription network motif where a transcription factor upregulates its own expression (17). Positive feedback loops are uncommon in metabolism under normal conditions, but an example can be found in the bacterial lac operon, where availability of lactose induces the expression of the lactose transporter lacY, as well as enzymes to metabolize it (lacZ and lacA) (18). This positive feedback is mediated through a double-negative feedback loop (Fig 5-1B): the lac repressor lacI blocks the expression of lacY under normal circumstances, but activity of lacY (that is, transport of lactose) blocks the activity of lacI, thus leading to amplification of lacY expression (18).

In cancer cells, however, several examples of positive feedback with mutual activators (Fig 5-

1C) has been reported in glycolysis. Cancer cells have been reported to exclusively express the

M2 isoform of pyruvate kinase (PKM2), whose enzymatic activity is only 25-50% that of the M1 isoform (PKM1) expressed in normal differentiated tissues (19). This decreased PK activity leads to an accumulation of its substrate, phosphoenolpyruvate (PEP). PEP directly activates phosphoglycerate mutase (PGAM1) by donating a phosphate group for phosphorylation of this enzyme, leading to increased flux through PGAM1 and increased rate of glycolysis (19) (Fig 5-

1C). Additionally, by donating the phosphate group to PGAM1, PEP is still converted to pyruvate, but without generating ATP as it would have been in the typical pathway, where the phosphate group is donated to ADP (19). Failure to generate ATP may further stimulate glycolysis through allosteric regulation of phosphofructokinase (20). A third positive feedback

163 loop involving PKM2 in cancer cells is in hypoxic conditions, where PKM2 translocates into the nucleus, interacts directly with HIF-1α, and promotes HIF1-dependent transcription, including transcription of PKM2 itself (21). These positive feedback loops act together to contribute to the

Warburg effect and contribute to tumour development (22, 23).

Here, we describe preliminary data suggesting that, in addition to the well-characterized negative feedback loop, the MVA pathway also participates in a positive feedback loop that has not been previously described. Implications of this positive feedback in the context of metabolic deregulation in cancer is discussed.

5.2 Results and discussion

We and others have previously shown that statin-induced cancer cell death can be reversed by co-administration with MVA, GGPP, or FPP (Fig 5-2A), indicating that depletion of these metabolites underlies cell death. Due to the negative feedback loop, statin treatment leads to upregulation of MVA pathway genes including HMGCR and HMGCS1. We thus hypothesized that co-administration of MVA, GGPP, or FPP will also reverse this change in gene expression.

Surprisingly, this was not the case. As shown in Fig 5-2B-C, co-administration of MVA, GGPP, and FPP with fluvastatin did not reverse the increase in expression of HMGCR and HMGCS1.

This was observed in all three MCF10A sublines tested, namely those with ectopic expression of

TWIST, SNAIL, SLUG, and the vector control. Although they have differing sensitivities to fluvastatin-induced cell death after 72 h of treatment (Fig 5-2A), this experiment was performed at 16 h after treatment, which should precede cell death. Nevertheless, we lowered the dose of fluvastatin to 1 μM in the second experiment, which does not lead to cell death in any of the three sublines, even after 72 h treatment. We’ve also extended our analysis to include samples at

24 h, 48 h, and 72 h after treatment. As shown in Fig 5-3, changing these parameters did not

vector TWIST 60 vector 60 TWIST 60 60

40 40

40 40 % pre-G1

20 % pre-G1 20 % pre-G1 20 % pre-G1 20 0 0 164 10µM fluva - + +- +- +- - + +- +- +- 0 MVA FPP 0 MVA FPP 100µM MVA EtOH- Fluva+- + GGPP- - EtOH- Fluva+- + GGPP- - FPP 10µM GGPP EtOH- +- MVA- + - - +- MVA- + FPP- Fluva GGPP EtOH Fluva GGPP vector TWIST 10µM FPP - +-SNAIL- - + - +-SLUG- - + A 60 vector 60 TWIST 60 SNAIL 60 SLUG pLPC pLPC TWIST TWIST 60 60 pLPC 60 pLPCvectorSNAIL 60 SNAILTWIST 10 5 10 5 5 5 40 40 4060 10 vector 406010 TWIST 4 4 4 4

40 40 8 8 4060 8 40608

% pre-G1

% pre-G1 % pre-G1

20 20 6 3 6 20403 6 % pre-G1 204063 3

% % Celldeath

% pre-G1

% pre-G1 % pre-G1 20 % pre-G1 20 2040 2040

4 2 4 2 4 42 2

% pre-G1

% pre-G1 % pre-G1 0 0 200 % pre-G1 200 1 1 1 1 10µM fluva - + - - - 2 - + - - 10µM- fluva2 -2 + - - - 2 - + - - -

- + + + + - + + + + % pre-G1 - + + + + - + + + +

% pre-G1 % pre-G1

0 20 % pre-G1 relative HMGCR mRNA HMGCR relative

MVA FPP 0 MVA FPP 20 mRNA HMGCR relative 0 MVA FPP 200 MVA FPP relative HMGCS1 mRNA HMGCS1 relative EtOH- Fluva+ + GGPP- - mRNA HMGCS1 relative EtOH- Fluva+ + GGPP- - EtOH Fluva GGPP EtOH Fluva GGPP 100µM MVA - - + - - 0 - - 0 + - 100µM- MVA0 0 -0 +- + - - 0 - +- + - 0 - FPP 0 FPP 0 FPP 10µM GGPP EtOH- +- MVA- + - - +- MVA- E + 10FPPµM- GGPP EtOH- E +- MVA- + - EtOH-E +- MVA- + - E Fluva GGPP EtOH E Fluva GGPP E FluvaE GGPP E Fluva GGPP 10µM FPP - +- - - + - +-F20 - F20- 10µM++ fluva - F20 F20+- F20+- +- +- - F20F20+- +- +- +- F20 SNAIL SLUG FPP1010µM FPPFPP100 - + + FPP10+ FPP10FPP+FPP10 0 - + + + FPP10FPP10FPP+ FPP10 MVA200 MVA200 EtOH FluvaMVA200MVA200MVAMVA200 FPP 0 EtOH FluvaMVA200MVA200MVA FPP MVA200 GGPP10100µMGGPP10 MVA EtOH- Fluva+- GGPP10+ GGPP10GGPPGGPP10- - EtOH- Fluva+- +GGPP10GGPP10GGPP- - GGPP10 60 60 - - + - - - - + - - SNAIL SLUG - + MVA- + FPP- - + - + FPP- vector TWIST 10µM GGPP EtOH- Fluva+- MVA- + - EtOH- Fluva+- MVA- + FPP- Fluva GGPP EtOH Fluva GGPP 60 vector 60 TWIST 10µM FPP - +- - - + - +- - - + pLPCB 60 vectorpLPCvector 60 TWISTpLPCTWISTTWISTSNAIL SNAIL10µM FPPTWISTTWISTSNAILSNAILSNAILSLUG SLUGSLUG SLUG 5 406060105 vector 1060 4060105 10 TWIST5 5 10 60105 10 SNAIL 60105 SLUG 5 vector TWIST pLPC SNAILpLPC TWISTSLUG TWIST vector TWIST 406040 40 4060 60 60 4 48 8 6084 8 45 4 8 60845 8 60845 45

60 vector 60 TWIST % pre-G1 204040 40% pre-G1 2040 40 40 6 6 6 6 6 6

3 3 3 6 34 3 60634 6034 34

% pre-G1

% pre-G1 % pre-G1 % pre-G1 204020 20 2040 40 40

2 24 4 42 4 23 2 4 40423 4 40423 23 % pre-G1

20 % pre-G1

% pre-G1

% pre-G1 % pre-G1

200 20 200 20 % pre-G1 20

% pre-G1

% pre-G1 % pre-G1 20 20 20 % pre-G1 20 10µM fluva02 - + - - - 20 2 2 - + 12 - - - 2 40212 2 4021 12

1 % pre-G1 1 1 1 % pre-G1 - + + + + - + + + + % pre-G1

200 % pre-G1 200 20 20

% pre-G1 % pre-G1 FPP FPP % pre-G1 20 MVA % pre-G1 20 MVA 20 20

EtOH Fluva GGPP EtOH Fluva GGPP relative HMGCS1 mRNA HMGCS1 relative

relative HMGCS1 mRNA HMGCS1 relative - + + - - - + + - - relative HMGCR mRNA HMGCR relative

relative HMGCR mRNA HMGCR relative 100µM MVA - + - - - - + - - - % pre-G1 10µM fluva - +- ++ + - + - - - + -+ ++ -+ - % pre-G1 0 FPP 0 FPP 20 20 00 EtOH Fluva MVA 0 EtOH0 0Fluva MVA01 0 0001 0 001 1 0 100µM MVA00 -- ++- MVA+- GGPP- + -FPP- 0- - + +- + MVAGGPP-- +- 0FPP- 0 00 0 10µM GGPP EtOH- Fluva- + GGPP- - -EtOH - Fluva+ - GGPP-

E E E E % pre-G1 E E E E

% pre-G1 relative HMGCR mRNA HMGCR relative E E FPP E E E 20 mRNA HMGCR relative E 20 E E 1010µMµM GGPP fluva EtOH-- F20+-+- MVA-+- ++- - +- - - F20+- F20MVA+- -F20 +-+ FPP+-- 10µM++- fluva -F20 F20+ F20+- +- +- - F20+ +- +- +- F20 10µM10µM fluvaFPP0 - FluvaF20+ +- GGPP+- FPP+- EtOH0 Fluva- F20+ GGPP+- F20+- 10µMFPP+- fluvaF20 0 - F20+ +- +- FPP+- 0 - F20+ +- +- FPP+- F20 0 EtOH Fluva MVA FPP10FPP 0 EtOH Fluva 0MVA FPP10 FPP 00EtOH Fluva MVA FPP 00 EtOH Fluva MVA 0FPP EtOH- +-+ MVA-+ GGPP- - + - -EtOH- +- +- MVA+- GGPP-+ FPP10FPP10- FPP10EtOH- + MVA+ GGPPFPP10- FPP10FPP10- FPP10 EtOH- + MVA+ GGPP- FPP10- 100µM10µM MVA FPPFPP10 - Fluva+SNAIL- MVA200+GGPP10GGPP- FPP10- - SLUGFluvaMVA200+- MVA200GGPP10MVA200+GGPP10GGPPGGPP10- 100µMFPP10- MVA0 - MVA200Fluva+- MVA200FPP10GGPP10+MVA200GGPP10GGPP-GGPP10- 0 - Fluva+- MVA200+GGPP10GGPP- FPP10- FPP10 MVA200100µMGGPP10 MVA - - MVA200+GGPP10- - - - MVA200+GGPP10E - MVA200100µM- GGPP10 MVAMVA200-GGPP10E - MVA200+ GGPP10- - -E - MVA200+GGPP10- - E MVA200GGPP10 - +- MVA- + FPP- - +- MVA- +F20 FPP- - F20+ MVA- + FPP- - F20+ MVA- + FPP- F20 10µM GGPP EtOH- Fluva+ MVA- GGPP+ FPP- EtOH- Fluva+ - +10FPPµM- GGPP EtOH- Fluva+- MVA- GGPP+ FPP- EtOH- Fluva+- MVA- GGPP+ FPP- 10µM GGPP60 EtOH- FluvaSNAIL- - + - 60 EtOH- SLUGFluva- MVA- GGPP+ 1010µMµM- GGPP fluva EtOH- Fluva+- +- +- +- EtOH- Fluva+- +- +- +- GGPP GGPP FPP100 GGPP FPP10FPP 0 GGPP FPP10FPP FPP10 10µM FPP - +- - - + - +- - - MVA20010µM++ GGPP10 FPP EtOH- +- MVA200MVA- GGPP10- + EtOH- +- MVA200MVA- GGPP10- + MVA200GGPP10 10µM FPP - +-SNAIL- - + - +-SLUG- - 100µM10µM++ MVAFPP - Fluva+- +- GGPP- +- - Fluva+- +- GGPP- +- SNAIL 60 TWISTSNAIL 60 TWISTSLUGSLUG SLUG SLUG 60 vector TWIST - + MVA- + FPP- - + - + FPP- SNAIL 60 SLUG 10µM GGPP EtOH- Fluva- - + - EtOH- Fluva- MVA- + - 106040 SNAIL 1040 60 10 SLUG 10 GGPP GGPP 5 605 60 5 510µM FPP - +- - - + - +- - - + pLPC 60 vectorpLPCvector 60 TWISTTWISTTWIST SNAIL TWIST SNAILSNAIL SLUGSLUG SLUG C 60840 840 60 8 8 45 606045 vector 60 6045 5 45 605 605 5

40 40 TWIST SNAIL SLUG % pre-G1 4020 % pre-G1 20 40 60640 640 60 6 606 60

34 4034 4034 4 34 4 4 4 % pre-G1 % pre-G1 4020 20 40 40 40

404 4 40 4 404 40 % pre-G1

23 20203 0% pre-G1 2023 3 23 3 3 3 % pre-G1

20 % pre-G1 20 % pre-G1 20 % pre-G1 20

10µM fluva40 - + - - - 40- + - - - 40 40 % pre-G1 % pre-G1 2 - + + + + 2 - 2 + + + + 2 2 2002 FPP 0 202 2 FPP 2 2 2 2

MVA % pre-G1 % pre-G1 MVA

% pre-G1 EtOH EtOH

20 Fluva % pre-G1 20 Fluva

1 100µM MVA201 - + + GGPP- - 20 1- + + GGPP- - 1

% pre-G1

% pre-G1 % pre-G1 20 - - + - - 20 - - + - - 20 % pre-G1 20 10µM GGPP00 - +- MVA- + FPP- 0 0- 0 +- MVA- + FPP- 0 01 001 EtOH Fluva GGPP 0 EtOH01 Fluva 1 GGPP 1 1 1 1

% pre-G1 0 0

% pre-G1 % pre-G1 20 % pre-G1 20 20

10µM10µM fluva FPP E-- +-+ -+- - +- + +- 20E- - +- E +- +-- +-+ +- E relative HMGCR mRNA HMGCR relative relative HMGCR mRNA HMGCR relative 10µM fluva0 - + - - - 0- + - - - E 10µM fluva0 -E F20++ MVA++- ++- +FPP+- - F20-E + ++F20MVA+-+ +-+ FPP+- E F20 0 EtOH Fluva MVA GGPP FPPFPP 0 0 EtOH FluvaMVA GGPPFPP FPP 0 F20 100µM MVA0 EtOH-- FluvaFluvaF20++- MVA++ GGPP- - FPP10- - EtOH0-EtOHFluva- +FluvaF20+- +0 MVAGGPP+- FPP10-- 0 FPP10- F20 0 FPP10 0 0 100µM MVA0 - -+MVA200++ GGPP- - - - 0- - -MVA200++ MVA200+- GGPP-- - 0 MVA200 0 100µM MVAFPP10 - - +GGPP10- FPP10- - - GGPP10+ GGPP10- FPP10- FPP10 GGPP10 - + MVA- + FPP- - + MVA- E + FPP- E E 10MVA200µMGGPP10GGPP EtOH--E Fluva+--MVA200MVA--GGPP10++ FPP- - -EtOH-E +- FluvaMVA- MVA200- -+GGPP10FPP+- - E MVA200GGPP10 E E 1010µMµM GGPP fluva EtOH- Fluva+ - GGPPGGPP- FPP- EtOH Fluva- + - GGPPF20- 10µMFPP- fluva - F20+ - - - - + - - - F2010µM GGPP0 EtOH- FluvaF20+- MVA+- + +- EtOH- FluvaF20+- MVAGGPP+- + +- F20 0 - + + + + 0 - F20+ + + + F20 - +- MVA- GGPP- FPP+ 0 - +- MVA- GGPP- FPP++ MVA FPP MVA FPP 10µM10µM FPPFPPFPP10EtOH- Fluva+-+SNAIL-+ GGPP- - FPP10+ - -EtOH- +-SLUGFluva+- +- GGPP-+ FPP10- FPP10EtOH Fluva FPP10 GGPP FPP10 EtOH Fluva GGPP FPP10 FPP10 MVA200100µM10µMGGPP10 MVAFPP - +- MVA200+- GGPP10- +- - +- MVA200+- GGPP10- MVA200100µM+-+GGPP10 MVAMVA200-GGPP10+- MVA200+ GGPP10- - - +- MVA200+GGPP10- - MVA200GGPP10 FPP FPP FPP 10µM GGPP60 EtOH- +- MVA- + - 60 - +- MVA- + 10FPPµM- GGPP EtOH- +- MVA- + - EtOH- +- MVA- + - FluvaSNAIL GGPP EtOH SLUGFluva GGPP Fluva GGPP Fluva GGPP 10µM FPP - +-SNAIL- - + - +-SLUG- - 10µM++ FPP - +- - - + - +- - - + SNAIL 60 SNAIL 60 SLUG SLUG 60 60 5 405 SNAIL 40 5 SLUG 5 6040 40 60

4 Figure 5-2. 4MVA and GGPP reverses4 fluvastatin4 -induced cell death but does not reverse % pre-G1 4020 % pre-G1 20 40

3 3 3 3 % pre-G1 upregulation% pre-G1 4020 of MVA pathway20 40 genes.

2 20 0 2 2 % pre-G1 20 % pre-G1 20 10µM fluva - + +- +- +- - + +- +- +- 0 FPP 0 FPP

MVA % pre-G1 1 % pre-G1 201 EtOH Fluva 20EtOH Fluva MVA 100µM MVA - +- + GGPP- - 1- +- + GGPP- - 1 FPP FPP A, 1010 μMµM GGPP fluvastatinEtOH- Fluva+- MVA- induced+ - cellEtOH- deathFluva+- MVA- in+ MCF10A- cells overexpressing TWIST, SNAIL, and 0 0 GGPP 0 GGPP 10µM FPP - +- - - + 0- +- - - + 0 E 10µM fluva0 -E + +- +- +- 0 -E + +- +- +- E F20 F20 MVA FPP MVA FPP EtOH- Fluva+ + GGPP- - EtOH- FluvaF20+ + GGPP- - F20 SLUG,100µM fully MVAFPP10 reversed- - + by- FPP10co--administration- - + with- FPP10- MVA, GGPP,FPP10 but not FPP. Cells were treated as MVA200GGPP10 MVA200GGPP10 MVA200GGPP10 MVA200GGPP10 - +- MVA- + FPP- - +- MVA- + FPP- 10µM GGPP EtOH Fluva GGPP EtOH Fluva GGPP indicated10µM forFPP 72- h.+- B--C, -10 +μM fluvastatin- +- - induced- ++ upregulation of HMGCS1 and HMGCR mRNA, which was not reversed by co-administration with MVA, GGPP, or FPP. Cells were treated as indicated for 16 h. Bars are mean + SD, n=2.

165 affect our previous finding, and we observed a sustained HMGCR and HMGCS1 upregulation with MVA co-treatment with fluvastatin, a condition where negative feedback control should be absent. This is a first indication that the MVA pathway may be regulated by a positive feedback loop, where excess accumulation of HMG-CoA leads to further activation of HMGCS1 and

HMGCR.

The negative feedback loop that regulates MVA pathway genes is mediated through the transcription factor SREBP2, which can be inhibited by the drug dipyridamole (DP) (24).

Specifically, DP blocks the cleavage activation of SREBP2 in response to sterol depletion, retains SREBP2 in the inactive form in the ER, and blocks the upregulation of HMGCR and

HMGCS1 in response to statin treatment (24). We thus tested whether positive feedback loop is also dependent on SREBP2 activation. As shown in Fig 5-4, DP treatment inhibited the upregulation of HMGCR and HMGCS1 in both the negative feedback scenario (fluva vs. fluva+DP), and the positive feedback scenario (fluva+MVA vs. fluva+MVA+DP). This suggests that SREBP2 activity mediates both the negative feedback loop (Fig 5-1A) and the positive feedback loop proposed by these experiments (Fig 5-5).

Our immediate plans to validate this model of MVA pathway regulation is to (i) assess SREBP2 activation in the context of HMGCS1 overexpression and (ii) knockdown. Since HMG-CoA cannot cross the cell membrane, future experiments will be focused on genetically manipulating cells so that they will accumulate HMG-CoA, including (i) knocking down HMGCR with independent shRNAs; (ii) knocking down HMGCL, which catabolizes HMG-CoA independently of HMGCR; and (iii) feeding cells with acetate as a carbon source, in conjunction with overexpression of HMGCS1 and/or upstream enzymes. If validated, this model could represent

166

A 24h 48h 72h B 24h 48h 72h 10 5

8 4

6 3

4 2

2 1

relative HMGCR mRNA HMGCR relative relative HMGCS1 mRNA HMGCS1 relative 0 0

E E E E E E 1μM fluva - F1+ + - F1+ + - F1+ + - F1+ + - F1+ + - F1+ + 200μM MVA - - + - - + - - + - - + - - + - - + F1 M200 F1 M200 F1 M200 F1 M200 F1 M200 F1 M200

10 5 Figure 5-3. HMGCS1 and HMGCR upregulation is sustained in the absence of negative 8 4 feedback. 6 3

A-B, HMGCR4 and HMGCS1 mRNA is consistently upregulated2 for 72h with high-dose MVA

co-treatment with2 low-dose fluvastatin. Bars are mean + SD,1 n=2.

relative HMGCR mRNA HMGCR relative relative HMGCS1 mRNA HMGCS1 relative 0 0 E E F1 F1 M200 M200 F1 M200 F1 M200

10 5

8 4

6 3

4 2

2 1

relative HMGCR mRNA HMGCR relative relative HMGCS1 mRNA HMGCS1 relative 0 0

E E E E E E F1 F1 F1 F1 F1 F1 167

F1 M200 F1 M200 F1 M200 F1 M200 F1 M200 F1 M200

A B 6 5

4 4 3

2 2

relative HMGCR mRNA HMGCR relative relative HMGCS1 mRNA HMGCS1 relative 0 0 E E 1μM fluva - - + F1 + - - + + - - +F1 + - - + + 200μM MVA - - - - M200+ + + + - - - - M200+ + + + F1 M200 F1 M200 5μM DP - + - + - + - + - + - + - + - +

Figure 5-4. HMGCS1 and HMGCR upregulation is dependent on SREBP2 in the absence of negative feedback.

A-B, upregulation of HMGCR and HMGCS1 mRNA with high-dose MVA co-treatment with low-dose fluvastatin is inhibited by DP, an inhibitor of SREBP2 processing. Cells were treated as indicated for 24 h. Bars are mean + SD, n=2.

168

Acetyl-CoA

Acetoacetyl-CoA HMGCS1 SREBP2 HMG-CoA HMGCR

Isoprenoids MVA

Cholesterol

Figure 5-5. The MVA pathway may be regulated by a positive feedback loop.

169 another mode of MVA pathway deregulation (Fig 1-3) that may be exploited by cancer cells to promote transformation (25).

5.3 Methods 5.3.1 Cell culture

MCF10A cells were a kind gift of Dr. Senthil Muthuswamy and were cultured as previously described. Transgene expression was stably introduced into MCF10A cells using retroviral insertion with pLPC, a kind gift of Dr. Roberta Maestro.

5.3.2 Cell death assay

Cells were seeded at 2.5x105/plate overnight, then treated as indicated. After 72 h of treatment, cells were fixed in 70% ethanol for >24 h, stained with propidium iodide, and analyzed by flow cytometry for the % sub-diploid DNA population as % cell death, as previously described.

5.3.3 qRT-PCR

Total RNA was harvested from subconfluent cells using TRIzol Reagent (Invitrogen). cDNA was synthesized from 500 ng of RNA using SuperScript II (Invitrogen). Real-time quantitative

RT-PCR was performed using TaqMan probes for HMGCR (ABI Hs00168352), HMGCS1 (ABI

Hs00266810), and GAPDH (ABI Hs99999905).

170

5.4 References

1. Schoenheimer R, Breusch F: Synthesis and destruction of cholesterol in the organism. Journal of Biological Chemistry 1933.

2. Brown MS, Goldstein JL: A receptor-mediated pathway for cholesterol homeostasis. Science 1986, 232(4746):34-47.

3. Brown MS, Faust JR, Goldstein JL, Kaneko I, Endo A: Induction of 3-hydroxy-3- methylglutaryl coenzyme A reductase activity in human fibroblasts incubated with compactin (ML-236B), a competitive inhibitor of the reductase. J Biol Chem 1978, 253(4):1121-1128.

4. Kovanen PT, Bilheimer DW, Goldstein JL, Jaramillo JJ, Brown MS: Regulatory role for hepatic low density lipoprotein receptors in vivo in the dog. Proceedings of the National Academy of Sciences of the United States of America 1981, 78(2):1194-1198.

5. Endo A, Kuroda M, Tanzawa K: Competitive inhibition of 3-hydroxy-3- methylglutaryl coenzyme A reductase by ML-236A and ML-236B fungal metabolites, having hypocholesterolemic activity. FEBS Lett 1976, 72(2):323-326.

6. Bucher NL, Overath P, Lynen F: beta-Hydroxy-beta-methyl-glutaryl coenzyme A reductase, cleavage and condensing enzymes in relation to cholesterol formation in rat liver. Biochim Biophys Acta 1960, 40:491-501.

7. Goldstein JL, Brown MS: A century of cholesterol and coronaries: from plaques to genes to statins. Cell 2015, 161(1):161-172.

8. Gofman JW, Delalla O, Glazier F, Freeman NK, Lindgren FT, Nichols AV, Strisower B, Tamplin AR: The serum lipoprotein transport system in health, metabolic disorders, atherosclerosis and coronary heart disease. J Clin Lipidol 2007, 1(2):104-141.

9. Freeman M: Feedback control of intercellular signalling in development. Nature 2000, 408(6810):313-319.

10. Mitrophanov AY, Groisman EA: Positive feedback in cellular control systems. Bioessays 2008, 30(6):542-555.

11. Umbarger HE: Amino acid biosynthesis and its regulation. Annu Rev Biochem 1978, 47:532-606.

12. Guidotti GG, Borghetti AF, Gazzola GC: The regulation of amino acid transport in animal cells. Biochim Biophys Acta 1978, 515(4):329-366.

13. Christie GR, Hyde R, Hundal HS: Regulation of amino acid transporters by amino acid availability. Curr Opin Clin Nutr Metab Care 2001, 4(5):425-431.

171

14. Klip A, Tsakiridis T, Marette A, Ortiz PA: Regulation of expression of glucose transporters by glucose: a review of studies in vivo and in cell cultures. FASEB J 1994, 8(1):43-53.

15. Kochanczyk M, Kocieniewski P, Kozlowska E, Jaruszewicz-Blonska J, Sparta B, Pargett M, Albeck JG, Hlavacek WS, Lipniacki T: Relaxation oscillations and hierarchy of feedbacks in MAPK signaling. Sci Rep 2017, 7:38244.

16. Eissing T, Conzelmann H, Gilles ED, Allgower F, Bullinger E, Scheurich P: Bistability analyses of a caspase activation model for receptor-induced apoptosis. J Biol Chem 2004, 279(35):36892-36897.

17. Bateman E: Autoregulation of eukaryotic transcription factors. Prog Nucleic Acid Res Mol Biol 1998, 60:133-168.

18. Jacob F, Monod J: Genetic regulatory mechanisms in the synthesis of proteins. J Mol Biol 1961, 3:318-356.

19. Vander Heiden MG, Locasale JW, Swanson KD, Sharfi H, Heffron GJ, Amador-Noguez D, Christofk HR, Wagner G, Rabinowitz JD, Asara JM et al: Evidence for an alternative glycolytic pathway in rapidly proliferating cells. Science 2010, 329(5998):1492-1499.

20. Berg JM, Tymoczko JL, Stryer L: Section 16.2, The Glycolytic Pathway Is Tightly Controlled. In: Biochemistry, 5th edition. New York: Freeman, W. H.; 2002.

21. Luo W, Hu H, Chang R, Zhong J, Knabel M, O'Meally R, Cole RN, Pandey A, Semenza GL: Pyruvate kinase M2 is a PHD3-stimulated coactivator for hypoxia-inducible factor 1. Cell 2011, 145(5):732-744.

22. Vander Heiden MG, Cantley LC, Thompson CB: Understanding the Warburg effect: the metabolic requirements of cell proliferation. Science 2009, 324(5930):1029-1033.

23. Kim JW, Dang CV: Cancer's molecular sweet tooth and the Warburg effect. Cancer Res 2006, 66(18):8927-8930.

24. Pandyra A, Mullen PJ, Kalkat M, Yu R, Pong JT, Li Z, Trudel S, Lang KS, Minden MD, Schimmer AD et al: Immediate utility of two approved agents to target both the metabolic mevalonate pathway and its restorative feedback loop. Cancer research 2014, 74(17):4772-4782.

25. Clendening JW, Pandyra A, Boutros PC, El Ghamrasni S, Khosravi F, Trentin GA, Martirosyan A, Hakem A, Hakem R, Jurisica I et al: Dysregulation of the mevalonate pathway promotes transformation. Proceedings of the National Academy of Sciences of the United States of America 2010, 107(34):15051-15056.