The Role of Microrna-515-5P in Breast Cancer

The role of microRNA-515-5p

in breast cancer

FILIPA GOMES DE PINHO

A thesis submitted for the degree of DOCTOR OF PHILOSOPHY

at Imperial College London

Department of Oncology, 1st floor ICTEM. Imperial College London. Hammersmith Hospital, Du Cane Road. London, W12 0NN

Supervisor: Professor Justin Stebbing Co-supervisor: Leandro Castellano

Dedicated to my grandparents Gil and Estela

And to Rui, my One and only Love

STATEMENT OF ORIGINALITY

All the experiments presented in this thesis were performed by myself, aside from the following:

. Figures 18 and 33: RNA isolation and cDNA library was performed by myself. RNA-

sequencing was carried out by Genomic Laboratories, MRC Clinical Sciences Centre,

London UK. RNA-sequencing and relative bioinformatic analyses were conducted by

Dr Leandro Castellano.

. Figure 27: Re-analysis of ChIP-Seq data was performed by Dr Leandro Castellano.

. Figures 32, 36, and 37: Random migration experiments were conducted by myself

and the tracking analysis was performed by Catriona Munro.

. Figure 38: Human breast cancer MDA-MB-231 cells isolated from metastatic loci in

immunocompromised mice were kindly provided by Prof Harikrishna Nakshatri. Total

RNA extraction and RT-PCRs were conducted by myself.

. Figure 39: Primary breast cancer and lymph node metastatic samples were acquired

from surgery undertaken by members of the Breast Surgical Department at Imperial

College London, specifically Miss Jacqueline Lewis. Total RNA extraction from these

tissue samples was performed by Dr Jonathan Krell. RT-PCRs were conducted by

myself.

In the chapter 3, figures and part of the text have been published with myself as the first author on a research article entitled “Downregulation of microRNA-515-5p by the estrogen receptor modulates sphingosine kinase 1 and breast cancer cell proliferation.”,

Cancer Research, 2013 Oct 1, 73(19):5936-48.

I am appreciative of the aforementioned help.

Filipa Pinho

The copyright of this thesis rests with the author and is made available under a Creative

Commons Attribution Non-Commercial No Derivatives licence. Researchers are free to copy, distribute or transmit the thesis on the condition that they attribute it, that they do not use it for commercial purposes and that they do not alter, transform or build upon it.

For any reuse or redistribution, researchers must make clear to others the licence terms of this work.

ACKNOWLEDGEMENTS

First of all, I would like to express my immense gratitude to Dr Dmitry Pchejetski and

Professor Justin Stebbing for giving me the opportunity to study and work in such high standards over the last 1366 days of my life. Particularly, I am extremely grateful to

Professor Justin Stebbing for his guidance and continuous support. His generosity has been inspiring, his constant curiosity and enthusiasm for research has been contagious and his unlimited positive energy made my PhD experience very productive and stimulating.

My sincere gratitude is also reserved for Dr Leandro Castellano for his guidance, immense knowledge and expertise and to Professor Jonathan Waxman for his invaluable support and encouragement.

No work was possible without my dear lab and office mates. Your jokes, advises, smiles…were the reason why I have enjoyed every minute of my PhD. A big thank you to those of you whose knowledge and uncountable advises were crucial for the success of my PhD journey. A particular thank you to Dr Adam Frampton and Dr Catriona Munro for their invaluable insights and suggestions and to Dr Lysann Sauer, Dr Joao Nunes, Dr

Heba Alshaker, Dr Jimmy Jacob, Dr Loredana Pellegrino and Dr John Krell, who, with big doses of patience and humour, have taught me all the lab techniques, tricks and tips.

A very special thank you to those of you who I will always carry in my heart: Dr Heba

Alshaker and Dr Joao Nunes for being the best lab family I could ever have, a sister and brother always there no matter what; Dr Nair Bonito, Richard Schlegel and Dr Catriona

Munro for being my true friends.

Finally, I would like to thank my love, Rui, for his constant love and infinite patience to accompany all the PhD dramas, my dear friends for always cheering me up in all my down moments and my amazing family for always believing in me.

ABSTRACT

MicroRNAs are a class of small non-coding RNAs which modulate a wide range of physiological and pathological processes by negatively regulating gene expression at the post-transcriptional level. MicroRNAs have been found to be important modulators of the tumorigenesis and metastasis formation of different human cancers, including breast cancer.

Here we prove that miR-515-5p belongs to a particular subset of microRNAs which mediate estradiol (E2) action in the carcinogenesis of estrogen receptor-positive breast cancers, in which proliferation is highly dependent on E2. We show that miR-515-5p downregulation is the main responsible for the positive effect of E2 in the expression of

SK1, an oncogenic enzyme required for E2-dependent breast cancer tumorigenesis.

Upon E2 stimulation, estrogen receptor α (ERα) directly represses the transcription of miR-515-5p which leads to the upregulation of SK1 expression as a result of its reduced availability to target SK1.

By conducting the first functional studies of miR-515-5p, here we also demonstrate that miR-515-5p plays a tumour suppressive role in breast cancer. miR-515-5p was found to inhibit breast cancer proliferation by inducing caspase-dependent apoptosis via SK1 and to repress breast cancer cell motility mainly by targeting MARK4, an enzyme implicated in the regulation of cell cytoskeleton dynamics.

Overall, we identify a new signaling pathway involving ERα, SK1 and miR-515-5p and a new role for miR-515-5p which highlights its therapeutic potential for the treatment of breast cancer.

TABLE OF CONTENTS

List of figures ...... 12

List of tables ...... 14

Glossary ...... 15

Chapter 1: Introduction ...... 20

1.1 Cancer ...... 21 1.2 Breast cancer ...... 22 1.2.1 Breast cancer epidemiology ...... 22 1.2.2 Breast cancer classification ...... 22

1.3 Metastatic breast cancer ...... 25 1.3.1 Metastatic breast cancer disease ...... 25 1.3.2 The formation of breast cancer metastasis ...... 25 1.3.3 Cell migration in metastatic cancer ...... 27 The molecular basis of cancer cell migration ...... 27 MARK4 and cell migration ...... 30

1.4 ER-positive breast cancer ...... 32 1.4.1 Clinical relevance of ERα status ...... 32 1.4.2 E2/ER-signalling in breast cancer ...... 33 Estrogens ...... 33 Estrogen receptor α ...... 33

1.4.3 Estrogen receptor α and Sphingosine kinase 1 ...... 35 Sphingolipids ...... 35 Sphingosine kinase isoforms ...... 36 Sphingosine kinase 1 signalling ...... 38 Sphingosine kinase 1 regulation ...... 39 Sphingosine kinase 1 in cancer ...... 40

Sphingosine kinase 1 and ERα in cancer ...... 41

1.5 MicroRNAs and breast cancer ...... 42 1.5.1 microRNAs definition ...... 42 1.5.2 microRNAs historical overview ...... 42 1.5.3 The biogenesis of microRNAs ...... 43 1.5.4 The mechanism of action of microRNAs ...... 45 1.5.5 Identification of microRNA targets ...... 49 1.5.6 The role of microRNAs in breast cancer ...... 51 microRNA deregulation in cancer ...... 51 Onco-microRNAs ...... 52 microRNAs and metastasis ...... 54 microRNAs and ER2/ERα-signaling ...... 56 Clinical applications of microRNAs in cancer ...... 57 miR-515-5p and breast cancer ...... 60

1.6 Aims...... 61

Chapter 2: Material and methods ...... 62

2.1 Material ...... 63 2.1.1 Mammalian cell culture ...... 63 2.1.2 Plasmids ...... 63 2.1.3 siRNA oligonucleotides ...... 64 2.1.4 Primers ...... 64 Primers used in plasmid construction ...... 64 Primers used for qRT-PCR ...... 65 2.1.5 Buffers and reagents ...... 65 Buffers used in western blots ...... 65 Buffers and reagents used in SK1 activity assays ...... 66 Buffers used in immunoprecipitation ...... 67 Buffers used in cell proliferation assays ...... 68 2.1.6 Antibodies ...... 68 Antibodies used in western blots ...... 68 Antibodies used in immunoprecipitation...... 68

2.1.7 Probes used in immunofluorescence ...... 69 2.1.8 Clinicopathological details of patients ...... 69 Patient samples used in Chapter 3 ...... 69 Patient samples used in Chapter 4 ...... 71

2.2 Methods ...... 73 2.2.1 Mammalian cell culture ...... 73 Cell lines origin, growth and passage ...... 73 Estrogen and anti-estrogen treatments ...... 74 2.2.2 Luciferase assays ...... 74 SK1 3’UTR luciferase reporter system ...... 74 miR-515-5p’s promoter luciferase reporter system ...... 75 NRAS, MARK4 and PI3KC2B 3’UTR luciferase reporter systems ...... 75 2.2.3 Transfections ...... 76 miRNA and siRNA transfections ...... 76 Plasmid transfections ...... 77 2.2.4 Quantitative real-time Reverse Transcription-PCR ...... 77 RNA extraction ...... 77 mRNA reverse transcription and qRT-PCR ...... 78 microRNAs reverse transcription and qRT-PCR ...... 78 2.2.5 Western blot ...... 79 2.2.6 Immunoprecipitation assay ...... 80 2.2.7 SK1 activity assay ...... 81 2.2.8 Cell proliferation assays ...... 82 MTT assays ...... 82 Trypan blue assays ...... 83 Cell apoptosis assays...... 83 2.2.9 Immunofluorescence ...... 84 2.2.10 Cell migration assays ...... 84 Cell tracking assay ...... 84 Boyden chamber assay ...... 84 2.2.11 RNA-Seq and ChIP-Seq analysis ...... 85 2.2.12 Tumour tissues ...... 86 2.2.13 Statistical analysis ...... 86

Chapter 3: Downregulation of miR-515-5p by estrogen receptor leads to tumorigenesis via SK1 ...... 88

3.1 Overview ...... 89

3.2 Results ...... 91 3.2.1 SK1 mRNA directly interacts with Ago2 ...... 91 3.2.2 miR-515-5p directly regulate SK1 expression ...... 92 3.2.3 miR-515-5p inhibits breast cancer cell proliferation...... 96 3.2.4 Silencing SK1 partially rescues the effect of miR-515-5p inhibition on breast cancer cell proliferation ...... 98 3.2.5 miR-515-5p targets many oncogenes with an enrichment of genes belonging to the Wnt Pathway ...... 100 3.2.6 Estradiol downregulates miR-515-5p expression and upregulates SK1 activity ...... 102 3.2.7 ERα and SK1 expression are positively correlated ...... 104 3.2.8 SK1 activity downregulation by TAM requires miR-515-5p ...... 105 3.2.9 miR-515-5p levels are not altered in ERα-negative breast cancer cells after estradiol (E2) and TAM treatment ...... 108 3.2.10 ERα mediates miR-515-5p downregulation by estradiol (E2) by directly binding to its promoter region ...... 109

3.3 Discussion ...... 112

Chapter 4: miR-515-5p inhibits breast cancer cell migration via MARK4 ...... 116

4.1 Overview ...... 117

4.2 Results ...... 119 4.2.1 miR-515-5p changes MCF7 and MDA-MB-231 cell morphology ...... 119 4.2.2 miR-515-5p inhibits breast cancer cell migration ...... 120 4.2.3 miR-515-5p directly regulates NRAS, MARK4 and PIK3C2B expression . 121 4.2.4 Silencing of MARK4 mimics the effect of miR-515-5p on breast cancer migration ...... 124

4.2.5 miR-515-5p inhibits cell migration through MARK4 downregulation in breast cancer ...... 125 4.2.6 miR-515-5p levels are inversely correlated with metastasis in breast cancer ...... 126 4.2.7 miR-515-5p and MARK4 as prognostic markers for metastatic cancer patients ...... 128

4.3 Discussion ...... 130

Chapter 5: Discussion and conclusions ...... 135

5.1 miR-515-5p as a member of the C19MC miRNA cluster, the largest miRNA cluster discovered in the human genome so far ...... 136 miR-515-5p is a context-dependent C19MC microRNA ...... 136 The expression of miR-515-5p’s cluster neighbours is altered upon E2 stimulation ...... 137

5.2 The role of miR-515-5p in ER-positive breast cancer tumorigenesis ...... 139 miR-515-5p downregulation is responsible for the positive correlation between E2 and SK1 ...... 139 miR-515-5p inhibits breast cancer cell proliferation by inducing caspase- dependent cell apoptosis ...... 139

5.3 The role of miR-515-5p in cell motility in breast cancer ...... 141 miR-515-5p inhibits breast cancer cell migration by targeting MARK4 ...... 141 MARK4 mRNA contains miR-515-5p’s putative binding sites in both 3’UTR and 5’UTR regions ...... 141

5.4 Conclusions and future directions ...... 143

References ...... 145

Appendices ...... 168

List of figures

Figure 1: Formation of cancer metastasis...... 26 Figure 2: Cell migration step by step ...... 30 Figure 3: Scheme of aromatase reaction...... 32 Figure 4: Blockage of E2 production and action to treat ER-positive breast cancer ...... 34 Figure 5: Sphingolipids interconversion and localisation in cell...... 36 Figure 6: Schematic representation of S1P receptors and their respective pathways and functions...... 39 Figure 7: MicroRNAs biogenesis...... 45 Figure 8: The different mechanisms through which microRNA negatively regulate gene expression at the post-transcriptional level...... 47 Figure 9: MicroRNA targeting network in breast cancer...... 53 Figure 10: MicroRNAs and ERα...... 57 Figure 11: SK1 mRNA expression is higher in Ago2 immunoprecipitates...... 91 Figure 12: Identification of microRNAs that directly interact with SK1 3’UTR...... 94 Figure 13: miR-515-5p and miR-206 regulate SK1 expression...... 95 Figure 14: miRNA-515-5p directly interacts with SK1 3’UTR...... 96 Figure 15: miR-515-5p inhibits cell proliferation in MCF7...... 97 Figure 16: miR-515-5p reduces cell growth and induces cell apoptosis in MCF7 and ZR75-1...... 98 Figure 17: SK1 silencing partially rescues the enhanced cell proliferation after miR-515- 5p inhibition...... 99 Figure 18: miR-515-5p regulates the expression of genes involved in wnt signaling and cell apoptosis...... 101 Figure 19: ERα downregulates miR-515-5p expression in MCF7 and ZR75...... 102 Figure 20: SK1 activity upregulation by estradiol is rescued upon miR-515-5p overexpression...... 103 Figure 21: ERα and SK1 expression are positively correlated in ERα positive cell lines or tissues...... 105 Figure 22: Reduction in SK1 activity by TAM requires miR-515-5p...... 106 Figure 23: miR-515-5p expression in Ago2 immunoprecipitates is upregulated by tamoxifen...... 107 Figure 24: miR-515-5p and SK1 activity levels in MDA-MB-231 treated with estradiol or tamoxifen...... 108

Figure 25: ERα directly binds to the miR-515-5p promoter region...... 109 Figure 26: miR-515-5p transcription is repressed by ERα...... 110 Figure 27: miR-515-5p levels are upregulated in ERα-negative breast tumours compared to ERα-positive tumours...... 111 Figure 28: miR-515-5p downregulation by ERα leads to an increase in cell proliferation via SK1...... 114 Figure 29: miR-515-5p changes MCF7 and MDA-MB-231 cell morphology...... 119 Figure 30: miR-515-5p inhibits MDA-MB-231 cell migration...... 120 Figure 31: RNA-seq analysis of MCF7 and MDA-MB-231 treated with miR-515-5p. .. 121 Figure 32: NRAS, FZD4, CDC42BPA, PIK3C2B and MARK4 mRNA levels in MCF7 and MDA-MB-231 upon miRs or sponges overexpression...... 122 Figure 33: miR-515-5p directly interacts with NRAS, MARK4 and PI3KC2B’s 3’UTR. 123 Figure 34: MARK4 knockdown inhibits MDA-MB-241 cell migration...... 124 Figure 35: miR-515-5p inhibits breast cancer cell migration by downregulating MARK4...... 125 Figure 36: miR-515-5p and MARK4 expression of primary tumour and metastatic cells in vivo...... 127 Figure 37: miR-515-5p expression of primary tumour and metastatic cells of breast cancer patients...... 128 Figure 38: MARK4 and miR-515-5p as prognostic molecular markers in ER-negative metastatic breast cancer...... 129 Figure 39: miR-515-5p downregulation leads to an increase in cell motility via MARK4...... 134 Figure 40: Scheme of the C19MC microRNA cluster...... 137 Figure 41: Maturation of some microRNAs located in the C19MC cluster ...... 138 Figure 42: Proposed mechanism underlying the effect of E2 on cell apoptosis via SK1 ...... 140 Figure 43: Proposed mechanism underlying the inhibitory effect of miR-515-5p on cell migration...... 142

List of tables

Table 1: Classification systems for breast cancer diseases...... 23 Table 2: The major differences between SK1 and SK2...... 37 Table 3: MicroRNA non-canonical mechanisms of action...... 48 Table 4: Bioinformatic tools for the prediction of microRNA targets...... 49 Table 5: Approaches for high throughput identification of microRNA targets...... 51 Table 6: MicroRNAs implicated in the formation of breast cancer metastasis...... 55 Table 7: Examples of clinical trials analysing the potential application of microRNAs as biomarkers for cancer diagnosis and prognosis...... 60 Table 8: Human breast cancer cell lines and respective growth media...... 63 Table 9: Backbone and insert of the HA-MARK4 plasmid...... 63 Table 10: Sequence of the miR-515-5p promoter region cloned into the pGl2-basic plasmid...... 64 Table 11: Sequence of MARK4 siRNAs...... 64 Table 12: Primers used for the SK1 3'UTR amplification...... 64 Table 13: List of primers used in real-time qPCR...... 65 Table 14: Western Blotting buffers...... 66 Table 15: SPHK1 buffer components...... 66 Table 16: Reagents used in SPHK1 activity assay...... 67 Table 17: Immunoprecipitation buffers and reagents...... 67 Table 18: Cell proliferation assay buffers and reagents...... 68 Table 19: Western blot antibodies...... 68 Table 20: Immunoprecipitation antibodies...... 68 Table 21: Immunofluorescence probes...... 69 Table 22: Clinico-pathological information of patients whose tumours were analysed in chapter 3...... 70 Table 23: Clinico-pathological information of patients whose tumours and lymph nodes were analysed in chapter 4...... 71 Table 24: Bibliographic references of the microRNAs predicted to target the SK1 3’UTR...... 93

Glossary

ABC Adenoside triphosphate-binding cassettes

AGO Argonaute protein

AI Aromatase inhibitor aIF6E Anti-association eukaryotic translation initiation 6E factor

ANGPTL4 Angiopoietin-related protein 4

AP1 Activator protein 1

AP2 Activator protein 2

ARHGEF2 Rho guanine nucleotide exchange factor 2

APS Ammonium persulfate

ATP Adenoside triphosphate

BCL2 B-cell lymphoma 2

BCL9 B-cell CLL/lymphoma 9

BSA Bovin serum albumin

CAGRs Cancer-associated genomic regions

CCND1 Cyclin D1

CCR4 C-C chemokine receptor type 4

CDC42BPA CDC42-binding protein A

CDK Cyclin-dependent kinase

CDR1 Cerebellar degeneration-related protein 1

CLIP Cross-linking immunoprecipitation

CSDC2 Cold shock domain-containing protein C2

DGCR8 Core microprocessor component DiGeorge syndrome critical region 8

DIXDC1 DIX domain containing 1

DM Directional migration

DMEM Dulbecco’s modified eagle’s medium dsRNA Double stranded-RNA

E2 17β-estradiol

EGF Epidermal growth factor

EGFR Epidermal growth factor receptor eIF4E Eukaryotic translation initiation factor 4E eIF6E Eukaryotic translation initiation factor 6E

Ena-VASP Ena- vasodilator-stimulated phosphoprotein

ER Estrogen receptor

ERE Estrogen responsive element

ERK Extracellular signal-regulated kinase

ESR1 Estrogen receptor 1

FAK Focal adhesion kinase

FCS Fetal calf serum

FFPE Formalin-fixed paraffin embedded

FGFR2 Fibroblast growth factor receptor 2

FL Firefly luciferase

FRAT2 Frequently rearranged in advanced T-cell lymphomas 2

FXR1 Fragile X mental retardation syndrome-related protein 1

FZD4 Frizzled 4

GAPDH Glyceraldehyde 3-phosphate deHydrogenase

GDP Guanosine diphosphate

GEF Guanine nucleotide exchange factor

GRIP1 Glutamate-receptor-interacting protein 1

GTP Guanosine diphosphate h Hour/s

H3 Histone 3

HER2 human epidermal growth factor receptor 2

HIF Hypoxia-inducible factor

HMGA2 High-mobility group AT-hook 2

HOXD10 Homeobox D10

HR Hazard risk

HRAS Harvey rat sarcoma viral oncogene homolog

HRE Hypoxia-responsive element

HRP Horseradish peroxidase

IDC Invasive ductal carcinoma

IL Interleukin

IL6R Interleukin-6 receptor

ILC Invasive lobular carcinoma

IMC Invasive mixed carcinoma

IP Immunoprecipitation

ITGA5 Integrin alpha-5

LOX Lysyl oxidase

MAP Mitogen activated protein

MAPK Mitogen activated protein kinase

MRCK Myotonic dystrophy kinase-related Cdc42-binding kinase

MARK4 Microtubule affinity-regulating kinase

MCS Multiple cloning sites mDia2 Mouse Diaphanous 2 min Minute/s miR microRNA miRNA microRNA

MMP Matrix metalloproteinase

MTT 3-(4,5-diMethylThiazol-2-yl)-2,5-diphenylTetrazolium bromide

MUT Mutated

NGF Neural growth factor

NRAS Neuroblastoma Rat Sarcoma viral oncogene homolog

NSCLC Non-Small Lung Cancer

PAGE PolyAcrylamide Gel Electrophoresis

PAR Partitioning-defective protein

PBS Phosphate buffered saline solution

PBS-T Phosphate buffered saline-tween solution

PDCD4 Programmed cell death protein 4

PFA Paraformaldehyde

PGC-1α Peroxisome proliferator-activated receptor gamma coactivator 1-alpha

PGR Progestrone receptor

PI3K Phosphoinositol-3-kinase

PI3KC2B Phosphatidylinositol 3 kinase- domain-containing beta polypeptide

PKC Protein kinase C

PLC Phospholipase C

PMSF Phenylmethylsulfonyl fluoride

Pol RNA Polymerase II

PP2A Protein phosphatase 2

PPT1 palmitoyl-protein thioesterase 1

PR Progestrone receptor

Pre-microRNA Precursor-microRNA

Pri-microRNA Primary-microRNA

PTEN Phosphatase and tensin homolog

PTHRP Parathyroid hormone-related protein

PVDF Polyvinylidene fluoride

R Spearman coefficient

RAKE RNA-primed array-based Klenow enzyme

RAS Rat sarcoma

RASD1 Rat sarcoma dexamethasone-induced 1

RDX Radixin

RGNEF Rho guanine nucleotide exchange factor

RISC RNA-induced silencing complex

RL Renilla luciferase

RM Random migration

RT-PCR Real time polymerase chain reaction

S1P Sphingosine 1-phosphate

S1P1 Sphingosine 1-phosphate receptor 1

S1P2 Sphingosine 1-phosphate receptor 2

S1P3 Sphingosine 1-phosphate receptor 3

S1P4 Sphingosine 1-phosphate receptor 4

S1P5 Sphingosine 1-phosphate receptor 5

SCR Sarcoma

SDF-1 Stromal cell-derived factor 1

SDS Sodium dodecyl sulfate

Ser Serine

SERM Selective estrogen receptor modulator

SILAC Stable isotope labeling by amino acids in cell culture siRNA Small interfering RNA

SK Sphingosine kinase

SOX SRY-related HMG-box

SP1 Specificity protein 1

SPHK Sphingosine kinase

SRA Steroid receptor RNA activator ssRNA Single stranded-RNA

T Tetosterone

TAM Tamoxifen

TCF7L1 Transcription factor 7-like 1

T-DMR Tissue-dependent differentially methylated region

TEMED Tetramethylethylenediamine

TG Tris-Glicine

TGF Transforming growth factor

Thr Threonine

TJP1 Tight junction protein 1

TNF Tumour necrosis factor

TP Trypan blue

TRAF4 TNF receptor-associated factor 4

TRAP220 Thyroid receptor-associated protein-220

TU Transcriptional unit

UK United Kingdom

USA United States of America

UTR Untranslated region

VEGF Vascular endothelial growth factor

WASP Wiskott–Aldrich Syndrome protein

WAVE WASP-family verprolin-homologous protein

XNR1 Xenopus nodal related-1

ZBTB10 Zinc finger and BTB domain-containing 10

ZEB Zinc finger E-box-binding homeobox

Chapter 1 Introduction

Introduction

1.1 Cancer

Cancer is a term that refers to all diseases in which there is an uncontrolled growth of abnormal cells that can invade other locations of the body. A tumour is formed when an uncontrolled cell growth is promoted by the deregulated expression of genes involved in cell survival and/or proliferation. When a fraction of the tumour cells acquire the capacity of migrating and invading the surrounding tissues and other organs, the tumour is then considered to be malignant or simply named as cancer.

The term cancer has been continuously being adapted to the advances in cancer biology. At the present time, it is widely recognised that cancer cells are defined by ten different biological capabilities: proliferation signalling sustainment; energy metabolism reprogramming; growth suppressors evasion; cell death resistance; neo-angiogenesis stimulation; replicative immortality, tumour-promoting immune system stimulation; invasion and metastasis activation; genome instability and mutations; and immune surveillance evasion [1].

1.2 Breast cancer

1.2.1 Breast cancer epidemiology

Breast cancer is the most common cancer disease among women, being estimated that over 1 million of women are diagnosed with breast cancer annually. In UK, 45,000 new breast cancer cases were diagnosed in 2010 which represent approximately 30% of all female cancers. Breast cancer incidence rate has increased in UK by 72% since

1970s, possibly a consequence of the increasing number of national programs to improve the diagnosis of this cancer disease. Despite the progressive increase of breast cancer incidence, breast cancer mortality has been continuously decreasing in UK since

1970s. Although there could be other factors, the national programs to promote an earlier diagnosis and the development of more efficient therapies to treat breast cancer are believed to be the major contributors to this improvement [2].

1.2.2 Breast cancer classification

Breast cancer is a highly heterogeneous disease, considering the immense variety of morphologic, biologic and molecular features encountered in different breast cancers.

Notably, this heterogeneity is noted in the clinical response of breast cancers to therapies which complicates the clinicians’ task of identifying the most appropriate treatment for each breast cancer case. In order to overcome this barrier, several breast cancer classification systems have been created to group breast cancers in classes with more homogenous features. These classification systems permit the realization of more directed treatments which significantly improve the prognosis of the patients. Table 1

22 indicates the classes and criteria of the most widely used breast cancer classification systems [3-6].

CLASSIFICATION SYSTEMS FOR BREAST CANCER DISEASES

Histopathology Molecular status Differentiation Disease progression (histological subtypes) (molecular subtypes) (grades) (stages) CRITERIA

 Invasive ductal carcinomas  Grade 1:  Stage I: of no specific type  Luminal A: Tumour with less than 2cm ER-positive, Well- differentiated and possible presence of  Invasive lobular carcinomas HER2-negative cancer in nearby lymph and low Ki67 tumour nodes  Mucinous carcinomas  Luminal B:  Grade 2:  Stage II:  Medullary carcinoma ER-positive, Moderately Tumour with less than 2cm HER2-negative differentiated and presence of cancer in 1  Invasive papillary and high Ki67 tumour to 3 lymph nodes in the carcinoma or ER-positive armpit or near the and HER2- breastbone  Invasive cribriform positive  Grade 3: OR carcinoma Poorly Tumour between 2 and 5cm and possible presence  HER-enriched: differentiated of cancer in nearby lymph  Metaplastic carcinoma ER-negative, tumour nodes PR-negative

 Tubular carcinomas and HER-  Stage III: positive  Adenoid cystic carcinoma Tumour of any size and presence of 4 to 9 lymph CLASSES  Basal-like: glands under the arm or in  Secretory carcinoma ER-negative, the lymph glands near the MAIN PR-negative, breastbone  Aprocrine carcinoma HER-negative, OR EGFR-positive Tumour with more than  Neuroendocrine tumours and CDK5- 5cm and presence of positive cancer in lymph nodes near  Glycogen-rich clear-cell the tumour, in the armpit carcinoma or/and near the breastbone  Normal breast: or collarbone ER-negative,  Lipid-rich carcinoma OR PR-negative, Tumour has spread to the HER-negative,  Invasive micropapillary breast skin EGFR-negative carcinoma and CDK5-  Stage IV: negative  Acinic cell carcinoma Cancer has spread beyond the breast, in the majority of  Oncocytic carcinoma  Claudin-low: the cases to brain, bone, HER2-negative liver and lungs.  Sebaceous carcinoma and low claudin

Table 1: Classification systems for breast cancer diseases. The different criteria and main classes of the most commonly used systems to classify breast cancer disease [3-6]. ER: estrogen receptor; PR: progesterone receptor; HER2: human epidermal growth factor receptor 2; EFGR: epidermal growth factor receptor; CDK5: cyclin-dependent kinase 5.

Taking in consideration the main scope of this thesis, two particular types of breast cancer will be now examined in more detail:

. Metastatic breast cancer, the last stage of breast cancer in respect to cancer

progression, in most cases incurable;

. And ER-positive breast cancer, a cancer type that represents 70-80% of all

breast cancers and in which the disease progression is dependent on the

estrogens levels.

1.3 Metastatic breast cancer

1.3.1 Metastatic breast cancer disease

As breast is not a vital organ, breast cancer typically becomes life-threatening when the disease has spread to distant locations or organs. In this case, breast cancer is commonly referred to as metastatic breast cancer or invasive breast cancer. Although there are treatments to prolong life, delay cancer progression and relieve cancer-related symptoms, at this stage breast cancer is rarely curable. In UK, metastatic breast cancer is the second cause of cancer death among females, being estimated to be responsible for the death of over 10,000 women every year [2].

1.3.2 The formation of breast cancer metastasis

Metastasis is the process of dissemination of cells from the primary tumour to distant locations and/or organs. Primary tumours harbour distinct cell sub-populations which have been found to have different metastatic potentials. The origin of this biological heterogeneity remains unclear but it is believed that the accumulation of mutations, epigenetic alterations and micro-environmental factors are the main responsible for the initial movement of some of the primary tumour’s cells into the surrounding tissues. The tumour cells that acquire the capacity of moving can then invade the surrounding tissues along the extracellular matrix and enter into the blood or lymphatic circulation. In the vessels, some of the circulating tumour cells transit to other organs where they arrest and proliferate to form a metastatic site [7] (Figure 1).

Figure 1: Formation of cancer metastasis. The tumour cells that acquire the migratory capacity leave the primary tumour and invade the surrounding tissues along the extracellular matrix. Some of these cells migrate towards the neighbouring blood vessels and enter the bloodstream in a process called intravasation, in which cells migrate through endothelial cell junctions and enter into the blood or lymphatic circulation. Part of the circulating tumour cells spread throughout the body and leave the circulation at potential secondary tumour sites through a process called extravasation. Extravasation consists in the interaction of cancer cells with vascular endothelial cells through cell adhesion- and chemokine-related processes. The cancer cells that extravase the vascular endothelial cells transmigrate through the endothelial barrier and invade the basement membrane surrounding the blood vessels. In these secondary sites, the majority of cancer cells undergo cell death; while the rest survive and proliferate within this new microenvironment forming a metastatic site. In breast cancer, brain, lung and bone are the most common organs for metastases formation. Figure adapted from [8], [9], [10] and [11].

The kinetics of metastasis progression is different according with the tissue where the cancer is originated. In the case of breast cancer, the time between the primary tumour diagnosis and the clinical detection of metastasis can range from years to decades. It has been proposed that the cells disseminated from the breast cancer primary tumour can enter in a state of metastatic dormancy which can be induced by their inability to colonise or prevented by the microenvironment of the other organs.

Interestingly, after a prolonged period of latency, the disseminated cells can undergo a metastatic speciation, a protracted evolution that was found to be organ-specific. It has been recently proved that some of the dormant disseminated cells acquire irreversible and very specific genetic and epigenetic modifications while residing in a particular organ micro-environment. Consistently, a number of independent studies demonstrated that the ability of breast cancer cells to form typical osteolytic metastases specifically require the production of osteoclast-activating factors, as it is the case of PTHRP, IL-11, IL-6 and TNFα [12].

1.3.3 Cell migration in metastatic cancer

The molecular basis of cancer cell migration

The metastasis process requires the ability of cancer cells to migrate from the primary tumour to other organs, being cell migration triggered in these cells by an increase in the cell asymmetry which is followed by an alteration of the cell cytoskeletal rearrangement and a maturation of the cell protrusions and focal adhesions (Figure 2)

[13]. A cancer cell acquires the capacity to migrate upon:

1. A change in the cell polarity and symmetry

The polarization of a cancer cell, which consists in an asymmetry of the cell molecular and functional phenotype, is mainly generated by the presence of microenvironmental factors, as chemo-attractants, which induce an asymmetric activation of cell receptors and signalling pathways. Several members of the Rho GTPase family have been found to be crucial in the maintenance of cell asymmetry and polarity. Among them, Cdc42, a small GTPase, has been demonstrated to play a central role in the cell polarisation by controlling the distance and position of the nucleus with respect to the centrosome

27 through MRCK and by regulating the microtubule dynamics through PAR3, PAR6 and

PKCα [14].

2. The actin cytoskeletal rearrangement and the formation of cell protrusions

The localised actin polymerisation of a polarised cell promotes the formation of local extensions of the plasma membrane named cell protrusions. The type and number of actin-binding proteins associated with the actin filaments dictate the type of cell protrusions adopted. Among the actin-binding proteins that have been found to be involved in the formation of cancer cell protrusions are: Arp2/3 complex which induces actin branching in a process mediated by the WASP and WAVE family proteins [15]; non-muscle myosin II and fascin that promote actin bundling by cross-linking parallel and anti-paralell actin filaments [16-19]; mDia2 which is important for the maintenance of polymerization-competent free filament barbed ends [20]; and the capping and Ena-

VASP proteins which promote and inhibit, respectively, the elongation of actin filaments by binding to the barbed ends of growing actin filaments [21, 22].

Cell protrusions are classified according to their morphology, molecular structure and function [23]. Lamellipodia are broad and flat protrusions with a high content of actin.

Their branched shape is initiated by actin polymerisation that pushes the plasma membrane forward [24]. In opposition, filopodia are long and thin protrusions which work as exploratory braches to help the cell investigating its surroundings. As lamellipodia, filopodia are originated from actin polymerisation but the long filaments are stabilised by the actin-bundling protein fascin [18]. Considered to be the functional analogues of lamellipodia, the plasma membrane blebs or pseudopodia are tiny protrusions formed by the rupture of the actin cortex which promotes the membrane blebbing and the protrusions’ retraction upon elongation [25]. Finally, invadopodia are protrusions constituted by actin and exocytotic vesicules containing matrix-degrading proteases.

Invadopodia confere cells the ability of degrading the basement membrane which underlies the stroma and the capillary barriers during the first step of metastatic extravasation [26].

3. A maturation of the cell focal adhesions

Upon the formation of the cell protrusions, the ability of cells to migrate is only possible due to the maturation of focal adhesions, which are complexes of proteins that mediate the interaction between the actin cytoskeleton of the migratory cell and the fibrils of the extracellular matrix. Focal adhesions have two main roles during cell motility: one is to activate signalling pathways that reinforce cell polarisation and the second is to act as the cell physical support during the movement. Although the composition of focal adhesions can vary from cell to cell, the majority are constituted by integrins α and β, talin, vinculin, paxiline, FAK and src [13, 27].

4. The establishment of the cell body contraction and retraction

Following the formation of the cell protrusions and focal adhesions, the cancer cell movement is then propelled by a coordinative process of the cell body contraction and retraction. The cell contraction is induced by the shrinkage of the non-muscle myosin II cytoskeleton, while cell retraction is promoted by the adhesion disassembly in a process mediated by dynamin [16].

Cell migration: step by step

1. A change in the cell polarity 2. The actin cytoskeleton rearrangement and the and symmetry formation of cell protrusions

Filopodia

Lamellipodia or Dia2 pseudopodia ENA/VASP F-actin Myosin-X IRSp53

Fascin Invadopodia F-actin ARP2/3 Capping nucleus protein

3. A maturation of the cell 4. The establishment of the cell body focal adhesions contraction and retraction

contraction retraction

Figure 2: Cell migration step by step. The molecular basis of the different steps of cell migration: a change in the cell polarity and symmetry; the actin cytoskeletal rearrangement and the formation of cell protrusions; a maturation of the cell focal adhesions; and the establishment of the cell body contraction and retraction. The figure was adapted from [14], [28], [29], [27] and [16].

MARK4 and cell migration

MARK4 belongs to the MAP/microtubule affinity-regulating kinase (MARK) family, a group of kinases implicated in the stability of the cell microtubule cytoskeleton. MARKs induce the dissociation of microtubule-associated proteins (MAPs) from microtubules by phosphorylating their microtubule-binding domain. Since MAPs are involved in the microtubules organization and dynamics, the dissociation of MAPs from microtubules induced by MARKs alters the stability of the cell microtubule cytoskeleton [30].

MARK4 was first reported by Kato et al whilst screening for genes regulated by β- catenin/Tcf complex in hepatocellular carcinoma. MARK4 levels were found to be elevated in response to the decreased activity of TCF/LEF1, a transcription factor implicated in Wnt signaling [31]. Several years later, MARK4 was found to localise to the cell centrosomes, midbody and nucleus and to interact with actin, myosin and α/β/γ- tubulins, indicating that this enzyme is, as its family members, implicated in the regulation of cell cytoskeleton dynamics [32-34].

Several evidences suggest that MARK4 might play an important role in the development of different human cancers. MARK4 expression was found to be upregulated in hepatocellular carcinoma and glioma [31, 35] and to interact with a number of proteins that play a determinant role in cancer cell migration [36, 37]. These binding partners are: 14-3-3 proteins and phosphatase phosphatase 2A (PP2A) which regulate the integrin signaling, implicated in the cell adhesion process during cancer cell migration, and Rho/Rac guanine nucleotide exchange factor 2 (ARHGEF2), a critical component of the locomotory machinery which coordinates multiple RhoA-dependent signaling pathways during cell migration [38-42]

1.4 ER-positive breast cancer

1.4.1 Clinical relevance of ERα status

Estrogen receptor status is a critical marker to guide breast cancer treatment. 70-

80% of breast cancers are ER-positive, being their progression highly dependent on estrogens [43]. Therefore, not surprisingly, the therapies that are most successful treating ER-positive breast cancers consist in the blockage of estrogen production or action [44]. Estrogen production can be blocked through surgical/chemical ovarian ablation or administration of aromotase innhibitors (AIs), drugs that inhibit estrogen production by targeting aromatase, the enzyme that converts androgens into estrogens

(Figure 3) [44, 45]. Ovarian ablation is commonly used to treat young women who produce most of their estrogens in ovaries; while the administration of AIs is more frequently used to treat post-menopausal women who produce most of their estrogens in the adrenal glands and fat cells [46, 47]. Alternatively, ERα-positive breast cancer patients can be treated with selective ER modulators (SERMs), drugs that inhibit the action of the estrogen receptors. Among these, tamoxifen (TAM), which blocks the estrogen binding to the ERα, is primarily responsible for the dramatic increase in survival of breast cancer patients over the past 30 years [48].

Figure 3: Scheme of aromatase reaction. Aromatase catalyses the conversion of testosterone into estradiol [49].

1.4.2 E2/ER-signalling in breast cancer

Estrogens

Apart from their essential and beneficial role in the development, maturation and function of women reproductive system, estrogens are known to stimulate the growth and spread of ER-positive breast cancer tumours [43]. In 1897, the demonstration that ovarieoctomy was able to induce the regression of metastatic breast cancer lesions indicated, for the first time, that anti-estrogen therapies could treat breast cancer patients

[50]. Since then, as aforementioned, estrogen receptor-positive breast cancers have been successfully treated through the blockage of estrogen production or action [44].

There are different classes of estrogens but the estrogen which is produced by aromatase and linked to the progression of ER-positive breast cancers is 17β-estradiol, often cited as E2 [51].

Estrogen receptor α

The action of E2 in cells is mainly initiated by its direct interaction with two intracellular receptors of the nuclear hormone family: estrogen receptor α (ERα) and estrogen receptor β (ERβ). ERα and ERβ have different genomic precursors (ESR1 and

ESR2, respectively) but share a high degree of homology, have a similar affinity to E2 and bind to identical DNA response elements [51]. These two receptors have been found to mediate E2 action on breast cancer development and progression through different signalling pathways, being the role of ERα in breast cancer more extensively studied than ERβ.

E2 activates ERα by inducing its homodimerization. The activated ERα is then translocated to the nucleus where it acts as a transcription factor. Depending on the co- factors and on the region where it binds, ERα can either activate or repress gene transcription. The most common co-factors that constitute a transcription-regulating complex with ERα are SRC-1, GRIP1, AIB1, CBP/p300, TRAP220, PGC-1α, p68 RNA helicase, and SRA [51] and, though ERα dimmers can bind to different regions in genome, they always interact with an identical region. The ERα form activated by E2 has a high affinity with specific response elements in genome named estrogen responsive elements (EREs). EREs contain the consensus sequence

AGGTCAnnnTGACCT, where ‘n’ represents any nucleotide [52]. These EREs are cis- acting enhancers located in either non-coding or coding regions (gene precursors). The interactions between ERα and its co-activators induce the formation of the transcription pre-initiation complex which then causes the disruption of chromatin at the ERE site

(Figure 4). Alternatively, ERα regulates gene transcription by an indirect manner through its binding to other transcription factors, as AP1 and SP1 [53].

Selective ER modulators Aromatase inhibitors (e.g. tamoxifen) E2

Cytoplasm E2 E2 Nucleus

E2 ERα ERα Aromatase

ERα ERα Gene transcription activation or repression

T Breast cancer cell Estrogen responsive elements AGGTCAnnnTGACCT Ovarian cell

Figure 4: Blockage of E2 production and action to treat ER-positive breast cancer. Estradiol (represented as E2) produced from testosterone (represented as T) activates estrogen receptor α (represented as ERα) by inducing its dimerisation. ERα dimer then enters in the nucleus where it acts as a transcription factor of non-coding or coding genes by binding to the Estrogen Responsive Elements (EREs) of their regulatory regions. As the activation of ERα by E2 promotes the progression of ER-positive breast cancer, the most commonly used and successful therapies to treat this type of breast cancer correspond to

34 the administration of aromatase inhibitors and selective ER modulators which respectively cause the blockage of E2 production and action.

1.4.3 Estrogen receptor α and Sphingosine kinase 1

Sphingolipids

Sphingolipids are a class of lipids in which fatty acids are linked to sphingoid bases by an amide bond [54]. For almost one hundred years, sphingolipids were thought to exclusively play a structural role in cell by conferring stability and resistance to the cell membrane [55]. However, over the last three decades, these molecules have also been found to be important mediators of several biological processes, including cell survival and proliferation [56]. Although dozens of sphingolipids have been identified to date, only three sphingolipids have been described to be determinant to govern cell survival and proliferation: ceramide and sphingosine, which induce cell apoptosis and cell cycle arrest, and sphingosine 1-phosphate, which mainly promotes cell survival and cell cycle progression. As a result, among other factors, cell fate is determined by the balance of the levels of ceramide, sphingosine and sphingosine 1-phosphate in the cell. This balance is commonly referred as sphingolipid rheostate and is modulated by the interconversion and sub-cellular localisation of these three sphingolipids (Figure 5) [56].

Sphingomielin Ceramide

Ceramide Sphingosine

Golgi Sphingosine 1- Ceramide Phosphate

Plasma membrane ER Sphingosine

Sphingosine 1- Phosphate Sphingosine 1- Phosphate nucleus

Figure 5: Sphingolipids interconversion and localisation in cell. Ceramide synthesised de novo in the endoplasmatic reticulum can be translocated to the Golgi apparatus where it is converted into sphingomyelin. The sphingomyelin generated in the Golgi apparatus can be in turn transported to the plasma membrane where it is converted into ceramide by sphingomyelinase. Still in the plasma membrane, ceramide can be converted into sphingosine 1-phosphate in a two-step process catalised by ceramidase and sphingosine kinase 1.The produced sphingosine 1-phosphate can be then translocated to the endoplasmic reticulum where it is reconverted into sphingosine. In this compartment, sphingosine can be either reconverted into ceramide by ceramide synthase or be transported to the nucleus where is converted into sphingosine 1-phosphate by sphingosine kinase 2 [56].

Sphingosine kinase isoforms

Sphingosine kinases are important modulators of the sphingolipid rheostate because they catalyse the conversion of sphingosine into sphingosine 1-phosphate

(S1P). There are two different human sphingosine kinase isoforms: Sphingosine Kinase

1 (SK1) and Sphingosine Kinase 2 (SK2). SK1 and SK2 have a similar sequence but distinct functions due to their different size, N-terminal sequence, adult tissue distribution and sub-cellular localisation. SK1 mainly promotes cell survival and proliferation through the production of sphingosine 1-phosphate in the plasma membrane, while SK2 mainly induces cell apoptosis by generating sphingosine 1-phosphate in the nucleus. The S1P produced by SK1 acts as an agonist or a secondary signalling messenger and the S1P

36 produced by SK2 controls gene expression by modulating histone acetylation (Table 2)

[57].

SK1 and SK2 are localised in different cellular compartments due to the distinct mechanisms through which they are activated. SK1 is activated by the extracellular signal-regulated kinase 1/2 (ERK1/2) through a phosphorylation on its Ser225. This phosphorylation, apart from inducing a 14 fold increase in its activity, promotes its translocation from the cytoplasm to the cell membrane as it induces a conformational change in SK1 that permits its association with particular membrane microdomains called lipid rafts [58]. SK2 is also activated by ERK1/2 but the phosphorylation occurs on its Ser351 and/or Thr578, as Ser225 site is not present [59]. Contrarily to SK1, after being phosphorylated by ERK1/2, SK2 is phosphorylated a second time by protein kinase D. This phosphorylation induces the translocation of SK2 from the cytoplasm to the nucleus through a mechanism that remains unclear to date (Table 2) [60].

SK1 SK2

Activation Erk1/2 Erk1/2 and PKD

Sub-cellular localisation of Plasma membrane nucleus the activated form

Function of the produced Agonist or secondary Genetic or epigenetic S1P messenger factor

Phenotype: Cell survival, proliferation Mostly Induction Mostly repression and migration

Table 2: The major differences between SK1 and SK2. The activation, sub-cellular localisation, function and phenotype of SK1 and SK2.

Sphingosine kinase 1 signalling

The S1P produced in the plasma membrane by SK1 has been found to activate numerous signalling pathways. It can act either intracellularly as a second messenger

[61] or outside the cell as an agonist of cell membrane receptors [62]. In the latter alternative, S1P is exported by ATP binding cassette (ABC) transporters to the cell outside where it binds to one of five G protein-coupled receptors called S1P1-5 receptors. Figure 6 summarises the signalling pathways which have been found so far to be activated by the most extensively studied S1P receptors: S1P1, S1P2 and S1P3

[63-66].

The effect of the S1P generated by SK1 in a certain cell or tissue is determined by the cell molecular environment, in particular by the levels of S1P receptors as their activation have been found to induce distinct cellular phenotypes [63]. S1P1 activation promotes cell survival, proliferation and migration in different types of cells [67-70]. Its activation was shown to induce tumour angiogenesis of endothelial cells by inducing their proliferation and neovascularization and to promote the migration of fibrosarcoma and ovarian cancer cells by inducing cytoskeletal rearrangements and membrane ruffling

[67]. Contrarily to S1P1, S1P2 is a controversial receptor as its activation was found to induce opposing effects depending on the cells where it is expressed. Its activation was found to inhibit cell migration of chinese hamster ovarian cells [71], to promote cell rounding and apoptosis of human embryonic kidney cells [72] and to induce cell survival and proliferation of hepatoma cells [73]. Similarly to S1P1, the activated S1P3 has been found to promote cell proliferation and migration of different cellular types. S1P3 activation was shown to induce the growth of rat hepatoma cells [73] and to promote motility of chinese hamster ovarian cells [69] and human umbilical vein endothelial cells

[67, 70]. Contrarily to the S1P1, S1P2 and S1P3, S1P4 and S1P5 are believed to be tissue-specific receptors as they are mainly expressed in immune and brain cells,

38 respectively. The activation of both receptors was shown to repress cell proliferation:

S1P4 inhibits the proliferation of T cells [74]; while S1P5 inhibits the proliferation of brain white matter cells [75, 76].

Figure 6: Schematic representation of S1P receptors and their respective pathways and functions. In plasma membrane, sphingosine is converted to sphingosine 1-phosphate by sphingosine kinase 1. After being exported through ABC transporters, S1P molecule can bind to five cognate G-coupled receptors named S1P receptors. These receptors induce the activation of different pathways that lead to different cell processes [63-66].

Sphingosine kinase 1 regulation

Although several roles of Sphingosine kinase 1’s have been identified since the discovery of this enzyme, little is known about the regulation of its expression. In terms of epigenetics, it was shown that SK1 gene, located in the chromosome 14, contains a

CpG island, which is a target for tissue-dependent DNA methylation. The overexpression of endogenous antisense transcripts originated from this CpG island, overlapping a

39 tissue-dependent differentially methylated region (T-DMR), is responsible for the demethylation of CG sites in the T-DMR. This RNA-directed demethylation was related with DNA methylation at three CC(A/T)GG sites in the T-DMR, revealing a possible novel mechanism of epigenetic regulation which links RNA-directed CG demethylation and non-CG methylation [77, 78].

At the transcriptional level, five transcription factors have been found to directly regulate SK1 expression: specificity protein 1 (SP1), activator protein 1 (AP1), activator protein 2 (AP2), hypoxia-inducible factor 1 (HIF1) and hypoxia-inducible factor 2 (HIF2).

The SP1, AP1 and AP2 factors were shown to respectively bind to the SP1, AP1 and

AP2 binding motifs of the first SK1 exon; whereas HIF1 and HIF2 were demonstrated to interact with the hypoxia-inducible factor-responsive-elements (HRE) present in the SK1 gene promoter [79].

Finally, in terms of post-trascriptional regulation, two AU-rich region-binding proteins

(AUBPs) have been found to be implicated in the stabilization of SK1 mRNA: AUF1 was shown to induce the decay of SK1 transcripts and in contrast HuR was demonstrated to promote the stabilization of the SK1 transcripts [80]. In addition to these AUBPs, several growth factors (e.g. Epidermal Growth Factor and Vascular Endothelial Growth Factor) and cytokines (e.g. interleukin-1 and Tumour Necrosis Factor alpha) have been reported to indirectly promote SK1 activation. These activators induce the upregulation of ERK1/2 which, as mentioned above, promotes an increase of 14 fold of SK1 activity [81, 82].

Sphingosine kinase 1 in cancer

SK1 has been shown to play an oncogenic role in different cancer types by promoting cellular processes that contribute to cancer progression, as cell survival, proliferation and migration. Reinforcing its importance in cancer, SK1 was also found to

40 serve as a “sensor” to chemotherapy [83], being its inhibition shown to have the potential to synergise the effects of chemotherapy [84-87]. Finally, demonstrating its clinical relevance, SK1 was shown to be highly expressed in a wide range of human malignant tumours and this higher expression proved to have a negative impact on the survival of cancer patients [88-92]. As a result of these evidences that highlight the importance of

SK1 in cancer progression, numerous SK1 inhibitors were designed, synthesised and tested for the treatment of different cancers. Among these, two SK1 inhibitors, phenoxiol and resveratrol, have recently reached the clinical phase due to their strong chemotherapeutic effect in prostate and colon cancer in vivo models, respectively [93].

Sphingosine kinase 1 and ERα in cancer

In breast cancer, SK1 has a 4-fold higher expression in approximately 80% of the cases in comparison with the normal tissues and promotes cell survival, proliferation and migration [94, 95]. The blockage of SK1 expression using small interfering RNAs

(siRNAs) or selective inhibitors was shown to abolish the response of ERα-positive breast cancer tumours to estrogens in vivo, proving that SK1 is necessary for an E2- dependent tumorigenesis [96]. E2 was shown to induce a rapid and transient activation of SK1 but the molecular mechanism underlying this effect has remained unclear until the time that the study reported in this thesis was initiated [92, 97].

1.5 MicroRNAs and breast cancer

1.5.1 microRNAs definition

MicroRNAs are small non-coding RNA molecules of ~17-25 nucleotides implicated in the regulation of diverse cellular and biological processes. MicroRNAs are produced from the stem zone of hairpin-shaped RNA molecules and modulate gene expression at the post-transcriptional level mainly by imperfectly binding to the mRNA of their target genes.

1.5.2 microRNAs historical overview

MicroRNAs were discovered in 1993 by Ambros’ and Ruvkun's laboratories whilst studying the heterochronic genes that control the temporal development pattern of C. elegans’ larval stages [98, 99]. They found that one of these genes, lin-4 gene, did not encode a protein but, instead, a RNA molecule with 22 nucleotides in length that was able to downregulate lin-14 expression by binding to the complementary sequences of the 3’UTR region of lin-14 mRNA [98, 99]. Although lin-4 RNA was proved to play a role which had not been previously described for other RNA molecule until that date, the importance of microRNAs was only recognised seven years later when the second microRNA, let-7, was identified by Brenda Reinhart [100]. Let-7 was found to play a role similar to lin-4 in the embryonic development of C. elegans but its conservation across different species, including humans, proved that these small non-coding RNAs were not a simple oddity of C. elegans [101]. Since then, thousands of microRNAs have been identified and their roles proved to be crucial for many biological processes and determinant for the progression of diverse human diseases, including cancer.

Nowadays, microRNAs are widely recognised as a separate class of small non-coding

RNA molecules due to their distinct biogenesis and mechanism of action [102].

1.5.3 The biogenesis of microRNAs

Although biogenesis can be specific for certain subsets of microRNAs [103], classically, microRNAs are generated and maturated through the following steps:

1. Transcription

MicroRNAs are transcribed from genes which can be categorised as intragenic or intergenic according to their genomic localisation. The intragenic microRNAs are located within the transcriptional units (TUs) of protein-coding and non-coding annotated genes and intergenic microRNAs reside in genomic regions that do not overlap with transcriptional units of annotated genes. As a result, the transcription of intragenic microRNAs is under the control of the regulatory regions of their host genes, while the transcription of intergenic microRNA is modulated by their own regulatory elements

[104].

The transcription of both intragenic and intergenic microRNAs is mainly preceded by the

RNA polymerase II (Pol II), having the generated transcripts typical features of Pol II transcripts such as the poly A tail at the 3’ end and the capped 5’ ends [105, 106]. The transcripts generated from microRNA genes are named primary-microRNAs (pri- microRNAs) and correspond to hairpin-shaped stems containing one or more loop structures along the stem formed by the imperfect complementarity between the nucleotides of the two arms (Figure 7).

2. Drosha processing

The pri-microRNAs are then converted into precursor-microRNAs by a Microprocessor complex in a process designated by “cropping”. This Microprocessor is constituted by

Drosha, a RNAse III, and the DiGeorge Syndrome Critical Region 8 (DGCR8), a RNA- binding protein also referred to as Pasha [107]. Drosha excises the 5' and 3' arms of the pri-miRNAs precisely 11 bp away from the ssRNA/dsRNA junction at the base of the hairpin stems, with the assistance of DGCR8 which act as a molecular ruler by recognising the cleavage sites [108] (Figure 7).

3. Nuclear export by exportin-5

The pre-microRNAs are then exported from the nucleus to the cytoplasm. The two- nucleotide 3’ overhang of the pre-microRNAs, typical of a RNAse III-mediated cleavage, is recognised by exportin-5, one of the soluble receptors present in the nucleus. Upon recognition, the exportin-5 in complex with the pre-microRNA docks at the nuclear pore complex (NPC), a large proteinacous channel embedded in the nuclear membrane [109].

This interaction allows the nuclear export of the pre-microRNAs through the NPC, being the release of the pre-microRNAs from exportin-5 only pursued upon the hydrolysis of its co-factor, Ran-GTP, which is converted into Ran-GDP [110] (Figure 7).

4. Dicer processing and guide strand selection

In the cytoplasm, Dicer, a RNAses III endonuclease, recognises the 3’ end of the pre- microRNAs and cleave off their hairpin loop [111]. The produced RNA duplex of ~22 nucleotides is then unwound and separated into two strands: a guide strand (miRNA), which correspond to the mature microRNA, and a passenger strand (miRNA*), which is degraded [112]. Interestingly, in some occasions, one of the strands is not degraded and two mature microRNAs are generated from the same RNA duplex and therefore co- expressed within the same cells [113, 114]. These microRNAs are denoted with the suffixes “5p” and “3p” to indicate that they were respectively originated from the 5′ arm and the 3′ arm of the same hairpin precursor [115] (Figure 7).

Pol II Pre-microRNA

DGCR8 Exportin-5 5’ Cap 5’ Cap AAAAA AAAAA Pri-microRNA Drosha Nucleus Cytoplasm RISC Dicer AGO2 NPC

MicroRNA:microRNA* Mature microRNA duplex

Figure 7: MicroRNA biogenesis. MicroRNAs have a precursor in genome which is transcribed to the primary-microRNA (pri-microRNA) form mainly by RNA polymerase II. The pri-microRNAs are cleaved and converted into precursor-microRNAs by Drosha and its cofactor, Pasha. The pre-microRNAs are then exported to the cytoplasm through NPC upon binding to exportin-5. In the cytoplasm, after being cleaved by Dicer, the microRNA:microRNA* duplex is then unwound. The mature microRNA as a single stranded RNA then is incorporates into RISC complex.

1.5.4 The mechanism of action of microRNAs

Upon biogenesis, the mature microRNAs as single stranded RNA molecules are incorporated into the RNA-induced silencing complex (RISC), a multiprotein complex constituted by Argonaute 2 (Ago2) and several RNA-binding proteins [116]. In the RISC complex, microRNAs recognise their targets by interacting with the 3’UTR, 5’UTR or coding region of their mRNAs in a sequence-specific manner [117-119]. MicroRNAs do not perfectly match with their target mRNAs in all occasions but there is a region of the microRNAs that is in most cases complementary and perfectly pairs with their target mRNAs. This region is commonly referred as “seed region” and in the case of the 3’UTR targeting corresponds to the positions 2 to 7 from the 5’ end of the microRNA [120].

After recognising their targets and still within the RISC complex, microRNAs negatively modulate the expression of their targets by translation repression or mRNA

45 degradation, being the mechanism of gene expression inhibition dictated by the degree of complementarity between the microRNA and their targets’ mRNAs (Figure 8).

When the complementarity of the microRNA:mRNA interaction is perfect or nearly perfect, the degradation of the mRNAs is induced by a Ago2-mediated endonucleolytic cleavage. Ago2 cleaves the phosphodiester bond at the mRNA sequence corresponding to the tenth and eleventh nucleotides of the microRNA from its 5’ end. Upon the cleavage, the GW18, a RISC-associated co-factor, translocates the RISC complex to the

P-bodies which are sites of the cell where the mRNA turnover occurs. In the P-bodies, the 3' fragment of the mRNA is destroyed by the exonuclease XRN1, while the 5' fragment of the mRNA is degraded in the exosome, a sub-cellular compartment containing exonucleases [121] (Figure 8).

On the other hand, when the microRNA:mRNA interaction is imperfect, protein expression is mostly inhibited through one of the following mechanisms:

. Repression of translation initiation

MicroRNAs can repress the translation of their targets at the initiation step with the assistance of Ago2 or anti-association factor eIF6 (aIF6), a RISC associated co-factor.

The anti-association factor eIF6 was shown to prevent the assembly of the 80S ribosome; while Ago2 was found to impede the recruitment of the cap-binding protein eukaryotic translation initiation factor 4E (eIF4E) to the m7G cap of the mRNAs [122,

123].

. Repression of translation elongation

MicroRNAs can repress the elongation of their targets’ translation by either inducing the dissociation of actively translating polyribosomes from the mRNA or promoting the proteolytic cleavage of the nascent polypeptides [124, 125].

. mRNA deanylation and degradation

MicroRNAs can induce the deanylation of their target mRNAs in a process assisted by

GW18, the RISC-associated co-factor above mentioned [126, 127]. GW18 mediates the

46 deanylation of the microRNA targets’ mRNAs by recruiting the deadenylation complex C-

C chemokine receptor 4 (CCR4)-NOT1 to the location of the RISC complex [127]. Upon deanylation, the RISC complexes together with the mRNAs are translocated to the P- bodies through a mechanism that remains unclear to date. In the P-bodies, the mRNAs are decapped by the decapping enzyme 1 and 2 complexes and then degraded through exonucleolytic digestion by XRN1 [128] (Figure 8).

Perfect miRNA:mRNA mRNA degradation XRN1 interaction Endonucleolyticcleavage GW18 5’ miRNA mRNA Exonuclease

RISC AGO2 3’

Repression of translation initiation

eIF4E Ribosomes The recruitment of eIF4E and/or 80s ribosomal unit is inhibited

Imperfect miRNA:mRNA Repression of translation elongation interaction Proteolysis of nascent polypeptides

Ribosomes drop-off OR

mRNA deadenylation and degradation

mRNA deanylation mRNA decapping GW18 NOT1/2 XRN1 P-bodies CCR4-NOT1

Figure 8: The different mechanisms through which microRNA negatively regulate gene expression at the post-transcriptional level. MicroRNAs complexed in the RISC machinery repress the expression of their target mRNAs through different mechanisms. When the complementarity of the microRNA:mRNA interaction is perfect or nearly perfect, microRNAs promote mRNA degradation by inducing a Ago2-mediated endonucleolytic cleavage. When the microRNA:mRNA interaction is imperfect, protein production is mostly repressed by: impeding the translation initiation through the inhibition of eIF4E or 80s ribosomal unit’s recruitment; repressing translation elongation through ribosome “drop off” or proteolytic cleavage of the nascent polypeptide; and inducing deadenylation of the target mRNA through the recruitment of CCR4-NOT complex with the assistance of the RISC-associated cofactor GW182.

Although microRNAs typically regulate gene expression by repressing translation or degrading mRNA, some subsets of microRNAs have been found to control gene expression of their targets through other mechanisms. Table 3 summarises the different microRNA non-canonical mechanisms of action that have been described so far.

MicroRNA non-canonical Example References mechanism of action

miR-320 silences the transcription of the cell cycle gene DNA- Repression of the directed RNA polymerase III subunit RPC4 by inducing the transcription of protein- [129] recruitment of a number of epigenetic mediators at the promoter coding genes region.

miR-373 activates the transcription of the E-Cadherin and cold-shock Activation of the domain containing protein C2 (CSDC2) by promoting the recruitment transcription of protein- [130] of Pol II at the transcription starting sites of these genes through its coding genes binding to the complementary regions of their promoters.

miR-369-3p stimulates protein translation of the tumour necrosis Activation of protein factor α (TNFα) with the assistance of fragile X-related protein 1 [131, 132] translation (FXR1) and Ago2 by directly interacting with the adenylate-uridylate- rich elements in TNFα mRNA.

Repression of the miR-671 downregulates the transcription of a circular antisense transcription of long non- transcript of the Cerebellar Degeneration-Related protein 1 (CDR1) [133] coding RNAs locus by directly cleaving it in an Ago2-slicer-dependent manner.

Regulation of other miR-709 suppresses miR-15a/16-1 biogenesis by blocking the [134] microRNAs biogenesis processing of pri-miR-15a/16-1 to pre-miR-15a/16-1.

Table 3: MicroRNA non-canonical mechanisms of action. Different microRNA non-canonical mechanisms of action and respective examples.

1.5.5 Identification of microRNA targets

The recognition of gene targets by microRNAs has been extensively studied but there are still no rules that permit us to predict whether a certain microRNA:mRNA interaction is functional or not. As a result, since the recognition of microRNAs as an independent class of non-coding RNAs, several bioinformatic tools have been developed to identify “in silico” microRNA targets. Table 4 indicates the most widely used tools for microRNA target prediction and the criteria in which are based their respective computational strategies [120].

Tool Sequence analysed Criteria

Stringent seed pairing, site number, site type, site context (for example, TargetScan site accessibility); option of ranking by likelihood of preferential 3’UTR conservation rather than site context

Stringent seed pairing for at least one of the sites for the miRNA, site Pictar 3’UTR number, overall predicted pairing stability, site conservation

Moderately stringent seed pairing, site number, pairing to most of the miranda 3’UTR miRNA, site conservation

Moderately stringent seed pairing, site number, overall pairing, site miRbase targets 3’UTR conservation

RNAhybrid ANY Moderately stringent seed pairing, overall predicted stability

Table 4: Bioinformatic tools for the prediction of microRNA targets. The most widely used tools for the prediction of mammalian microRNA targets; modified from [120].

As the bioinformatic outputs are merely predictive, an experimental validation is required to prove that the predicted targets are true targets of a certain microRNA. First, to investigate whether the interaction between the microRNA and the target mRNA is functional, the region of the target mRNA which is predicted to interact with the microRNA is cloned into a plasmid vector containing a reporter gene (e.g. luciferase).

The microRNA-mRNA interaction is considered to be functional if the microRNA overexpression induces a significant downregulation of the expression or activity of the report gene in the plasmid-transfected cells. As the reporter system is artificial, it is still necessary to verify whether the binding of the microRNA to the mRNA of the predicted target occurs physiologically. With this aim, the mRNA and protein levels of the target need to be analysed upon microRNA over-expression. As microRNAs are known to negatively modulate the expression of their targets, the tested gene is finally proved to be a microRNA target if its mRNA and/or protein expression is significantly downregulated upon the microRNA overexpression [135].

Although microRNA targets can be identified through the bioinformatic prediction and experimental validation of each microRNA-mRNA interaction in a trial and error process, nowadays the high throughput identification of microRNA targets is possible in a time-effective way due to the advances in molecular biology. Table 5 indicates the existing experimental approaches that allow the high throughput identification of microRNA targets.

Approaches for high throughput identification Description References of microRNA targets

The transcriptome of cells overexpressing the microRNA of interest is analysed using probe-target hybridization (e.g. microarray) or sequencing Transcriptome (e.g. RNA-seq) technology. Upon analysis, only the transcripts which are [136, 137] profiling simultaneously downregulated in the microRNA-overexpressing cells and predicted to be microRNA targets are experimentally validated.

crosslinking The RNA-protein interactions are crosslinked in cell overexpressing the immunoprecipitation microRNA of interest.The RNA in the cell lysates which is bound to the [138] (CLIP) of RISC RISC components is then isolated by immunoprecipitation and then components sequenced to identify microRNA target sites.

Cells are transfected with biotinylated miRNAs which permits the isolation biotin-tagged of the microRNA:mRNA duplexes from cell lysates using streptavidin [139] microRNAs beads. The isolated RNA is sequenced to identify microRNA target sites.

A strong detergent is used to disassociate proteins and RNA of cells overexpressing the microRNA of interest. A transcriptase is then used for Reverse transcription cDNA synthesis of the target mRNAs, being the endogenous microRNAs [140] of targets used as primers. The resultant cDNA is then cloned and sequenced to identify the bound mRNA.

Stable isotope Protein abundance of microRNA-transfected cells is measured by mass labelling with amino spectrometry of samples labelled with different isotopes. Upon analysis, [141] acids in cell only the proteins which are downregulated and predicted to be microRNA culture (SILAC) targets are experimentally validated.

Cyclohexamide is added to microRNA-transfected cells to trap elongating ribosomes. Upon cell lysis, centrifugation through a sucrose gradient is used separate mRNAs with no associated ribosomes from those with Translation profiling bound ribosomes and therefore undergoing translation. Poly(A) RNAs from [142] bound and unbound pools are isolated, amplified, coupled to Cy5 and Cy3 dyes, respectively, and competitively hybridised to DNA microarrays for mRNA identification.

Table 5: Approaches for high throughput identification of microRNA targets. Description of different approaches to identify microRNA targets in a high throughput mode.

1.5.6 The role of microRNAs in breast cancer

microRNA deregulation in cancer

The link between microRNAs and cancer was first reported in 2002 by Calin et al. who showed that miR-15 and miR-16 genes were located at a genomic region frequently silenced in chronic lymphocytic leukemia [143]. Since then, the expression of several microRNAs has been found to be altered in a wide range of human cancers, being an increased expression of certain microRNAs correlated with an acceleration of the development of different cancer diseases [144].

MicroRNA expression is deregulated in cancer at three different levels: pre- transcriptional, transcriptional and post-transcriptional [145]. Previously to transcription,

51 the expression of some microRNAs was found to be altered in cancer due to the amplification, deletion, mutation or epigenetic modifications (e.g. DNA methylation and histone acetylation) of their respective genomic locis. This is the case of the miR-124 gene which was shown to be silenced in different cancers due to the hypermethylation of a CpG island located in its promoter region [146]. At the transcriptional level, the expression of certain microRNAs was found to be deregulated in cancer diseases due to the upregulation or downregulation of transcription factors. For example, the expression of miR-17-92 cluster was shown to be upregulated in different types of lymphoma and solid tumours because of the elevated levels of myc, a transcription factor that directly activates the transcription of this microRNA cluster [147-152]. Upon transcription, the expression of some subsets of microRNAs was found to be mainly deregulated due to an alteration in the levels of proteins involved in their maturation process. This is the case of the processing of miR-16-1, miR-143 and miR-145 by Drosha which was shown to be directly repressed by p53, a tumour suppressor protein frequently downregulated in cancer [153].

Onco-microRNAs

MicroRNAs that are deregulated in cancer and play a determinant role in cancer progression are named onco-microRNAs [154]. Onco-microRNAs can act either as oncogenes and be upregulated in cancer (e.g. miR-155 and miR-21), or as tumour suppressors and be downregulated in cancer (e.g.let-7 and miR-34) [155]. As an overall downregulation of microRNAs was observed in different types of cancer, it is believed that the majority of microRNAs play a tumour suppressive role [156-160].

Numerous onco-microRNAs have been identified in breast cancer. Figure 9 represents a network of oncogenic and tumour suppressive genes targeted by microRNAs in breast cancer. Among the represented microRNAs, miR-21, miR-373 and

52 miR-520 have been found to be frequently overexpressed and to play an oncogenic role in breast cancer. miR-21 was shown to promote cell proliferation by inducing cell death resistance through PDCD4, BCL2 and PTEN targeting and to induce cell migration by promoting anoikis through tropomyosin1 targeting [161-163]. miR-373 and miR-520 were found to promote metastasis formation by inducing cancer cell motility through the targeting of cell adhesion and extracellular matrix molecules (e.g. CD44 and MMP9, respectively) [164, 165]. On the other hand, other microRNAs, as miR-34a, miR-17-5p, miR-206 and miR-221/222, have been found to be downregulated and to play a tumour suppressive role in breast cancer. miR-34a and miR-17-5p were demonstrated to repress cancer cell growth by directly and indirectly inhibiting the expression of proteins required for cell survival and cell cycle progression [166-168]; while miR-206 and miR-

221/222 were shown to sensitise hormonal-resistant cells by suppressing E2-dependent proliferation and motility through estrogen receptor α targeting [169-172].

METASTASIS PROLIFERATION

miR-200 ZEB2 E2F1 CDk6 E2F3 MMP16 Let-7 miR-17-5p miR-155

miR-34a CCND1 HRAS

RHOA MYC CELL DEATH

BCL2 RESISTANCE ESR1 miR-373/520c miR-206, miR-221/222 miR-31 miR-21 CD44 HER2 PTEN

CHEMORESISTANCE miR-125b

RDX ITGA5

Figure 9: MicroRNA targeting network in breast cancer. MicroRNAs implicated in the progression of breast cancer, their target genes and the oncogenic processes in which they are mainly involved. Diagram modified from [173].

53 microRNAs and metastasis

Since the discovered link between microRNAs and cancer, several studies have shown that microRNAs are important modulators of cancer metastasis [174]. The role of microRNAs in metastasis progression has been extensively studied in different types of cancers, including breast cancer [175]. Table 6 indicates some of the breast cancer metastasis-promoter and metastasis-suppressor microRNAs identified so far and the mechanism through which they are believed to control metastasis formation.

MicroRNAs have been found to control the development of breast cancer metastasis by modulating both cancer phenotype and microenvironment. miR-31 and miR-10b are among the microRNAs that govern the cancer phenotype required for metastasis development. miR-31 was found to suppress cell invasion by targeting components of cell cytoskeleton important for cell-cell and cell-ECM interactions, while miR-10b was shown to promote cell migration by targeting HOXD10, a transcription factor that represses the expression of several genes involved in cell migration and extracellular matrix remodelling, including RhoC and α3-integrin [176, 177]. On the other hand, some microRNAs modulate the metastatic breast cancer microenvironment by regulating the expression of cell membrane-bound and secreted proteins or by being transmitted between different cell types that constitute metastasis. For example, miR-126 was shown to target SDF-1α, a chemokine which induces the recruitment of mesenchymal stem cells and inflammatory monocytes that inhibit breast cancer metastasis [178]; while miR-105 expressed in metastatic breast cancer cells was found to be transferred through exosomes to nearby vascular endothelial cells where it promotes the dissemination of cancer cells to the blood circulation by targeting ZO-1, a protein of the tight junction

[179].

Metastasis-promoter microRNAs Metastasis-suppressor microRNAs

miR-373/520c - migration & invasion miR-31 – suppression of anoikis resistance & Validated targets: invasion  CD44, a cell adhesion molecule that mediates Validated targets: communication and adhesion between cells  RhoA mediates the formation of membrane and between cells and the extracellular matrix protrusions at the cell rear and retraction at the [164]; cell trailing edge during cell migration [176].;  mTOR and SIRT1, negative regulators of the  radixin, a component of the cell cytoskeleton that metalloproteinase MMP9 involved in the promotes the linkage between the actin filaments extracellular matrix degradation during invasion and the cell membrane [176].; [165].  integrin α 5, a subunit of the integrin α5β1, a receptor implicated in the adhesion and miR-9 – migration communication between cells [176]. Validated target:  E-cadherin, molecule that mediate the adhesion let-7 – migration and invasion repression between cells within an epithelial tissue [180]. Validated targets:  HMGA2, implicated in cancer epithelial-to- miR-10b – migration & invasion mesenchymal transition [184, 185]; Validated target:  BACH-1, a transcription factor that regulates the  HOXD10, a transcription factor that represses expression of several genes that modulate the expression of several genes involved in cell metastasis formation [186]. migration and extracellular matrix remodeling, including RhoC and α3-integrin [177]. miR-126 – tumour environment modulation Validated target: miR-29b - tumour environment modulation  SDF-1α, a chemokine which promotes the Validated targets: recruitment of mesenchymal stem cells and  ANGPTL4,protects cells from anoikis modulates inflammatory monocytes that inhibit cancer vascular permeability [181]; metastasis [178].  LOX, mediates collagen crosslinking required for fibrosis-enhanced metastasis [181]; miR-335 – suppression of migration & metastatic  MMP9, degrade extracellular matrix proteins colonisation [181]; Validated targets:  VEGF-A, promotes vascular permeability [181].  SOX4, a transcription factor that activates several genes that induce epithelial-mesenchymal miR-21 – migration transition [187]; Validated target:  Tenascin-C, an extracellular matrix glycoprotein  tropomyosin1, a cytoskeleton protein implicated that promotes the survival and outgrowth of in the cell contraction and relaxation during metastasis [187]. movement that induces cell anoikis [161]. miR-105 - tumour environment modulation miR-200 – tumour environment modulation Validated target: Validated target:  TJP1, a protein of the tight junction of blood  Sec23a, a protein mediates secretion of vessels cells (miR-105 is secreted within metastasis-suppressive proteins [182]; exosomes which enter in the blood vessel cells to  ZEB1, a crucial inducer of epithelial target TJP1) [179]. mesenchymal transition [183].

Table 6: MicroRNAs implicated in the formation of breast cancer metastasis. MicroRNAs that suppress and promote breast cancer metastasis and their validated targets.

55 microRNAs and ER2/ERα-signaling

As the expression of certain microRNAs is deregulated in breast cancer and E2 is an important regulator of gene expression, the link between E2 and microRNA expression was investigated in several studies [168-171, 188-193]. These studies revealed that E2 deregulates the expression of numerous microRNAs, including miR-21, miR-17-5p and let-7 family members, previously shown to play a determinant role in breast cancer progression [168, 188]. Notably, in one of these studies, the authors demonstrated that a considerable part of the E2-deregulated microRNAs contain ERα binding sites in their transcriptional units [168]. Therefore, not surprisingly, ERα was then proved to regulate the transcription of several microRNAs by directly interacting with their promoter region [168, 169, 192, 193]. Adding more complexity to the link between

ERα and microRNAs, a fraction of the microRNAs deregulated by ERα was also found to directly or indirectly regulate ERα expression [169, 171, 193, 194]. The reciprocal interaction between these microRNAs and ERα inevitably results in regulatory feedback loops: some microRNAs are implicated in a positive feedback loop with ERα as miR-375 which is upregulated by ERα and targets RASD1, a repressor of ERα transcription; while other microRNAs are involved in a negative feedback loop with ERα as miR-206 which is downregulated by ERα and directly targets ERα (Figure 10) [171, 193, 194].

miR-18a RASD1 mRNA ZBTB10 mRNA miR-145 miR-22

miR-375 miR-27a ZBTB10 RASD1 ERα mRNA ERα

ERα ERα

ZBTB10

RASD1 ERα mRNA miR-21 miR-206 miR-190 miR-221/222 miR-375 miR-27b miR-23a

Figure 10: MicroRNAs and ERα. ERα expression is regulated by several microRNAs: miR-18a, miR-145, miR-22, miR-206 and miR-221/222 directly target ERα [170, 171, 189-191], while miR-27a and miR-375 respectively target ZBTB10 and RASD1, two transcription factors that repress ERα transcription [194, 195]. On the other hand, the expression of certain microRNAs is regulated by ERα. ERα promotes the transcription of miR-21 and miR-190 and represses the transcription of miR-206, miR-221/222, miR-27b and miR-23a by directly binding to their promoter regions [168, 169, 192, 193].

Clinical applications of microRNAs in cancer

The proved importance of microRNAs in cancer attracted the attention of both clinicians and researchers who started contemplating the clinical application of microRNAs. Some microRNAs deregulated in cancer diseases have been proposed as diagnostic and prognostic cancer biomarkers [196], while other microRNAs playing a decisive role in carcinogenesis, malignant transformation or/and metastasis formation have emerged as potential candidates for the development of novel anti-cancer therapies [197]. As shown in Table 1.7, several microRNAs are currently being tested in clinics for the diagnosis and prognosis of different types of cancer. However, at the present time, only one study evaluating the therapeutic use of a microRNA in cancer has reached the clinical phase. MRX34, the investigated biologic, consists in a liposome formulation (Smarticles, Marina Biotech, Washington) containing synthetic microRNAs

57 which mimic the endogenous miR-34a, a very well-known tumour suppressor microRNA lost in a wide range of cancer diseases. MRX34 was developed by Mirna Therapeutics and its potential use is currently being studied for the treatment of non-invasive liver cancer [198].

Clinical trials that analysed the potential application of microRNAs as cancer biomarkers

Clinical Cancer type Trials.gov Study title Study aim Sponsor Identifier

ICORG- All Circulating miRNAs Analysing circulating microRNAs as Ireland NCT01722 ICORG 10-11, V2 biomarkers of response to guide and Cooperative 851 monitor response to chemotherapy Oncology Research Group

MIRNA Profiling of Breast Cancer in Patients Undergoing Studying the microRNA profiling of breast Neoadjuvant or NCT01231 cancer in patients with locally advanced & City of Hope Adjuvant Treatment 386 inflammatory breast cancer undergoing Medical Center for Locally neoadjuvant or adjuvant treatment Advanced & Inflammatory Breast Cancer

Evaluating the feasibility to detect in the Circulating miRNAs circulating blood of metastatic invasive as Biomarkers of breast cancer or locally advanced breast Breast cancer NCT01612 Institut Claudius Hormone Sensitivity cancer patients, before treatment, the 871 Regaud in Breast Cancer? presence of the fifteen tissular microRNAs Pilot Study described in preclinical studies as possibly involved in hormone resistance/sensitivity.

A Combined GWAS and miRNA for the Identification of Spanish Breast Identifying miRNA signatures in whole NCT01598 Bevacizumab Cancer blood as bevacizumab response predictors 285 Response Research in metastatic breast cancer patients. Predictors in Group, Spain Metastatic Breast Cancer Assessing whether changes in expression Circulating of selected circulating microRNAs in MicroRNA as West NCT02065 serum could comprise a sensitive and Biomarker of Pomeranian 908 specific biomarker of cardiotoxicity in Cardiotoxicity in Cancer Center cancer patients treated with anthracyclines Breast Cancer based chemotherapy.

Analysing the use of microRNA as Micro-RNA Wuerzburg Prostate NCT01220 biomarkers to predict the response of high- Expression Profiles University cancer 427 risk prostate cancer tumour to different in High Risk Hospital Prostate Cancer treatments

Biological Factors Identifying a predictive molecular Predicting Response European Lung NCT00864 signature, including microRNAs, for the to Chemotherapy in Cancer Working 266 response of patients with non-small cell Advanced Non Party lung cancer chemotherapy Small Cell Lung Cancer Lung cancer Plasma microRNA Profiling as First Fondazione Line Screening Test Assessing the efficacy of plasma miRNA NCT02247 IRCCS Istituto for Lung Cancer profiling as the first line screening test for 453 Nazionale dei Detection: a lung cancer detection. Tumori, Milano Prospective Study (BIOMILD) Association Between VEGF-C Lung cancer and miRNA and Evaluating miR-326 as a biomarker for the China Medical and NCT01240 Clinical Non-small early prediction of cancer of non-small cell University Esophagus 369 Cell Lung Cancer lung cancer and esophagus squamous cell Hospital carcinoma and Esophagus carcinoma Squamous Cell Carcinoma Biomarkers in Studying biomarkers, including certain Tissue Samples Endometrial NCT01119 microRNAs, in tissue samples from Gynecologic From Patients With cancer 573 patients with stage I or stage III Oncology Group Stage I or Stage III endometrial cancer. Endometrial Cancer Search for Predictors of Comparing profiles of miRNA expression Ovarian NCT01391 Therapeutic from serum of 2 patient populations: with Centre Francois cancer 351 Response in or without recurrence 6 months after Baclesse Ovarian Carcinoma completion of chemotherapy. (miRSa)

Analysing the use of a microRNA panel to identify thyroid malignancy in leftover cells Hadassah NCT01964 microRNA in Thyroid of ultrasound guided fine needle aspiration Medical 508 Cancer biopsy and the effect of these microRNAs Organization on target genes. Thyroid cancer Evaluating whether the expression of a Biomarkers to microRNA panel blood, fine-needle NCT01433 Distinguish Benign aspirate biopsies (FNAB) of thyroid Norman 809 From Malignant nodules and post-surgical fresh-frosen Eberhardt Thyroid Neoplasm thyroid cancer accurately distinguishes benign from malignant thyroid neoplasms. Evaluating the Testing the hypothesis that in primary Expression Levels of Dartmouth- NCT01849 glioma samples mir-10b expression Glioma MicroRNA-10b in Hitchcock 952 patterns will serve as a prognostic and Patients With Medical Center diagnostic marker. Gliomas Kaohsiung Biomarkers for the Medical Early Diagnosis, Identifying microRNAs in Circulating NCT01828 University Prediction, Tumour Cells that may be useful to guide 918 Chung-Ho Prognosis for change in treatment decisions. Memorial Colorectal Cancers Colorectal Hospital carcinoma Identifying relevant miRNAs that may serve as biomarkers to stratify CRC Quantifying Micro NCT01712 patients according to their clinical Meir Medical RNA Levels of 958 characteristics such as disease stage, Center Colon (CRC) specific treatment, prognosis and disease recurrence

Evaluating the use of microRNAs in the Central blood, cerebrospinal fluid and urine of Ann & Robert H nervous NCT01595 Longitudinal Study patients with central nervous system Lurie Children's system 126 of Biomarkers tumours as a markers for an early Hospital of cancer predictor of cancer and response to Chicago therapy

Table 7: Examples of clinical trials analysing the potential application of microRNAs as biomarkers for cancer diagnosis and prognosis. Title, clinicaltrials.gov identifier, aim and sponsors of clinical trials testing the potential application of microRNAs as cancer biomarkers [199].

miR-515-5p and breast cancer

Despite the existence of an extensive literature on the role of some microRNAs in cancer, there is still a long list of microRNAs of which functions are still unclear. This list includes miR-515-5p, a microRNA located in the C19MC cluster, the longest microRNA cluster identified in the human genome so far.

Prior to our study, there were only two known facts about miR-515-5p in cancer, both concluded from experiments using the breast cancer MCF7 cells. The first was reported by Cortez at al who demonstrated that the transcription of miR-515-5p was under the control of an epigenetic modification, the methylation of a CpG island localised in the promoter region of the C19MC cluster [200]. The second was revealed by Lee at al who found that miR-515-5p levels were downregulated upon E2 stimulation [201].

1.6 Aims

Since SK1 is required for an E2-dependent breast cancer tumorgenesis and the molecular mechanism underlying SK1 upregulation by E2 remained unclear until the beginning of this study, our initial aim was to elucidate this mechanism by investigating whether microRNAs targeting SK1 could mediate the effect of estradiol in SK1 expression.

In the first part of this study, we proved that in breast cancer the upregulation of SK1 expression by estradiol is mediated by the downregulation of miR-515-5p, a microRNA that directly targets SK1. Therefore, our second aim was to analyse the role of miR-515-

5p in breast cancer cell proliferation and migration by examining the phenotype of miR-

515-5p-overexpressing cells and by screening for direct targets of miR-515-5p that could mediate the observed phenotype.

Chapter 2 Material and Methods

Material and methods

2.1 Material

2.1.1 Mammalian cell culture

Cell Line Cell type Type of tumour Medium and Additives

MCF7 Breast Adenocarcinoma DMEM (Dulbecco’s Modified Eagle’s Medium - Sigma-Aldrich, Dorset, UK); 10% fetal calf serum (FCS); ZR75 Breast Adenocarcinoma 2mM glutamine, 100U/ml penicillin, 0.1mg/ml streptomycin

MDA-MB-231 Breast Ductal carcinoma

Table 8: Human breast cancer cell lines and respective growth media.

2.1.2 Plasmids

Plasmid name Insert Tag/ Position Backbone Origin

HA-MARK4 MARK4L HA N-Terminal pCDNA3 Trinczek et al JBC 2004

Table 9: Backbone and insert of the HA-MARK4 plasmid.

AGGGGGAGGTTGCAGTGAGCCAAGACCGTGCTACTGCACTCCAGCCTGGCA ACAGAGCAAGACTCCGTCTCAAATAATAATAATAATAATAATTTAAAAAAAAAA TAAAAAATAAGAAATAAAAGAATGAACCACAATGTCCCTTTGTTCTACAGGCCA CTCTTGTGCTATCCACAGTTGCCAAAAACCCCCACCTGCCTTGTCTATGGACG CCCATCGGGTGATACCACAAACACCTTTGATCCTACAGATACCAGGGTAAATT CAGCATCCAAGTCCTTGAATTCTTCCCGTGAGTATAAGCTGCCCTGGCTCAGC miR-515-5p promoter region CCTCTTCTTTGCCAGGAACCCAATCCTTCAAACTTCCCTCAAAGACTGGGTAC (human genome) TTTTAGCATGACTGCTCTAGAGATTCTGGGTGGGGCTCAAAGTTTAACCTTCA GGGTCAGCAACATTTTATGGAAAGAGCCACAGAGTAAATATTTGAAACATTCT CGTACACTGACACTCAGCTGGACTGGCAAAGCAGAATATCTGTGTGTCAGTG TGCGTTTTATTCAGCCGTTGTTTGGGTCAGGGTCTGTGGGCGGCCCCTGGCA GCTAATGCCCTCCTGTGAGGAACAATACCTCACCATCAAGCACAGTCCGTCA CATATTCTTGCCCTGGTCGCTCGCTCTTTCTTTCTTTTCTTTCTCTTCTTTCTTT CCTTTCTTTCTCATTTTCGTTTCTTTCTTTCTTTCCTTCTTTTCTTTTTCTTTCTG TCTCTCTTTATCTATTTCTGTTTCTTTCTCTCTCTCTCACCTTTCTTTCTTTCCTT

CCTTCCTTCTTTCTCTTCCTT

Table 10: Sequence of the miR-515-5p promoter region cloned into the pGl2-basic plasmid.

2.1.3 siRNA oligonucleotides

Supplier Target Gene siRNA ID Sequence

Applied SPHK1 1181 - AM51331 Not disclosed by the company Biosystems

D-005345-06 GCUGUACUCUCGAGCAAAU

D-005345-05 GGAUCAACAUCGGCUAUGA Dharmacon MARK4 D-005345-02 GGAAGUACCGGGUCCCUUU

D-005345-01 GAUCGAAGCUGGACACGUU

Table 11: Sequence of MARK4 siRNAs.

2.1.4 Primers

Primers used in plasmid construction

Primers Sequence (5´to 3´)

Forward: CGCACTAGTTAGTGTCTACTTGCAGGACC SK1 3'UTR Reverse: CGCAAGCTTAGGCTGGGAATGTCACTTTA

Forward:TGGAGGAGACTCCTCTGGTTCCGGGTGAGAAGGTGGAGGC SK1 3'UTRMUT Reverse: GCCTCCACCTTCTCACCCGGAACCAGAGGAGTCTCCTCCA

Table 12: Primers used for the SK1 3'UTR amplification.

Primers used for qRT-PCR

Target Forward (5’ to 3’) Reverse (5’ to 3’) Supplier

SPHK1 TATGAATGCCCCTACTTGGTATTG GCCTCGCTAACCATCAATTCC Primer Design

NRAS TGTGATTTGCCAACAAGGAC CAACACCCTGTCTGGTCTT Sigma-Aldrich

PI3KC2B TACCTCGTCCATCTCCAAGA GGAAGTCTCCATCAGCCAG Sigma-Aldrich

FZD4 CAACGTGACCAAGATGCC AGGAAGAACTGCAGCTGG Sigma-Aldrich

CDC42BPA GTGATTGGTCGAGGAGCT ACATGCTGTCTCAGCTCTTT Sigma-Aldrich

MARK4 GTCAACAGACTGTGAGAGCATCC GCTCTGTGTATGGCTTCAACTCC Sigma-Aldrich

GAPDH AGCCACATCGCTCAGACAC GCCCAATACGACCAAATCC Sigma-Aldrich

Table 13: List of primers used in real-time qPCR.

2.1.5 Buffers and reagents

Buffers used in western blots

Reagent Supplier Components

Sigma-Aldrich TG solution (10x) 288g glycine, 60g Tris dissolved in 2L ddH2O (all components)

Sigma-Aldrich 200ml 10xTG, 20ml SDS (20%) dissolved 1x SDS running buffer (all components) in 2L ddH2O

Sigma-Aldrich 200ml 10xTG, 400ml methanol dissolved 1x Transfer buffer (all components) in 2L ddH2O

Phosphate buffered Sigma-Aldrich saline-Tween 20 0.05% Tween 20 dissolved in 1xPBS (all components) (PBS-T wash buffer)

Dried skimmed milk Sigma-Aldrich 5% dried skimmed milk powder dissolved in powder blocking buffer (all components) PBS-T wash buffer

Table 14: Western Blotting buffers.

Buffers and reagents used in SK1 activity assays

For 1 ml buffer

Volume and initial Component Supplier Final Concentration Concentration

Santa Cruz Tris HCl buffer pH 7.4 200µl of 0.1M 20 mM Biotechnology

Glycerol Sigma-Aldrich 200µl neat 20%

β-mercaptoethanol Sigma-Aldrich 1µl of 14.3M 1 mM

EDTA Sigma-Aldrich 100µl of 10mM 1 mM

PMSF Sigma-Aldrich 10µl of 0.1M 1 mM

NaF Sigma-Aldrich 30µl of 0.5M 15 mM

Leupeptin Roche 10µl of 1mg/ml 10 µg/ml

Aprotinine Sigma-Aldrich 10µl of 1mg/ml 10 µg/ml

Soybean trypsin inhibitor Invitrogen 10ul of 1mg/ml 10 ug/ml

4-Deoxypyridoxine Sigma-Aldrich 10 µl of 50mM 0.5mM

β-glycerophosphate Millipore 20µl of 2M 40 mM

Sodium orthovanadate Sigma-Aldrich 10µl of 0.1M 1 mM

H2O distilled NA 483µl NA

Table 15: SPHK1 buffer components.

Name Supplier Components

SPHK1 buffer NA Up to 170 µl/reaction – table 2.8

Sphingosine Sigma-Aldrich 10 µl 1 mM sphingosine/assay

MgCl2 Sigma-Aldrich 2μl of20mM

ATP Sigma-Aldrich 0.2µl of 10mM/sample

(γ-32P)- ATP Perkin-Elmer 1ul ATP-32P (10uCi/sample)

Sigma-Aldrich per reaction: 200μl ddH O, 20μl HCl 1M, 240μl KCl N, SPHK1 washing solution 2 2 (all components) 506μl chloroform, 533μl methanol

Sigma-Aldrich per reaction: 50μl 1M HCl, 800μl Stop solution (all components) Chloroform/Methanol/HCl, 240μl CHCl3 and 240μl 2M KCl

Sigma-Aldrich Dissolving Solution per sample: 40 μl Chloroform/Methanol (2:1, v/v) (all components)

Sigma-Aldrich 200-300ml 1-butanol/ethanol/acetic acid/water Migratory Solution (all components) (80:20:10:20, v/v)

Table 16: Reagents used in SPHK1 activity assay.

Buffers used in immunoprecipitation

Buffer Supplier Components

Sigma-Aldrich 20mM Tris-HCL pH7.5, 150mM KCL, 0.5% NP40, Lysis buffer (all components) 2mM EDTA and 1mM NaF

Sigma-Aldrich 8 50mM Tris-HCL pH7.5, 300mM NaCl, Washing buffer (all components) 5mM MgCl2, NP40 0.05%

Sigma-Aldrich PBS-T 0.05% Tween-20 in 1xPBS (all components)

Sigma-Aldrich Dried milk solution 5% non-fat dried milk powder in 0.05% Tween-20 in PBS (all components)

Table 17: Immunoprecipitation buffers and reagents.

Buffers used in cell proliferation assays

Buffer Supplier Components

MTT 3-(4,5-dimethylthiazol-2- Sigma-Aldrich yl)-2,5-diphenyltetrazolium 5 mg/mL MTT (Sigma Aldrich) in PBS (all components) bromide (MTT) solution

Sigma-Aldrich Extraction buffer 100ml SDS 20% + 3.87ml Dimethylformamide (DMF) (all components)

Table 18: Cell proliferation assay buffers and reagents.

2.1.6 Antibodies

Antibodies used in western blots

Protein target Species Clone Supplier Dilution

ERK1/2 Mouse 9106 Cell signalling 1:1000 (phospho)

Santa Cruz GAPDH Mouse sc-69778 1:5000 Biotechnology

Anti-mouse Goat 7076 Cell Signalling 1:2000

Table 19: Western blot antibodies.

Antibodies used in immunoprecipitation

Protein Species Supplier

Ago2 Rat Sigma-Aldrich

IgG Rat Sigma-Aldrich

Table 20: Immunoprecipitation antibodies.

2.1.7 Probes used in immunofluorescence

Name Target Supplier Dilution

Alexa Fluor® 488-conjugated Actin Invitrogen 1:1000 Phalloidin

4’, 6-diamidino-2-phenylindole DNA Invitrogen 1:200, 1 μg/ml (DAPI)

Table 21: Immunofluorescence probes.

2.1.8 Clinicopathological details of patients

Patient samples used in Chapter 3

Patients Sex Age Type of Cancer Grade ER PR HER2

1 F 82 IDC 2 +++ +++ -

2 F 82 IMC 2 +++ +++ -

3 F 48 IDC 2 +++ +++ -

4 F 57 ILC 2 +++ + -

5 F 56 IDC 2 +++ + -

6 F 70 IDC 2 +++ ++ -

7 F 60 IDC 2 +++ - -

8 F 64 IDC 1 +++ +++ -

9 F 48 IDC 3 - - +

10 F 49 IDC N/A - - -

11 F 54 IDC 3 +++ - -

12 F 54 IDC 3 - - -

13 F 66 IDC 2 - - -

14 F 56 IDC 3 - - -

15 F 68 IDC 3 - - +

16 F 60 IDC 2 - - +

17 F 50 IDC 3 +++ + -

18 F 47 IDC 3 +++ ++ +

19 F 54 IDC 2 +++ ++ +

20 F 70 IDC 2 +++ +++ -

21 F 52 ILC 2 +++ ++ -

22 F 65 IDC 2 +++ +++ -

23 F 46 ILC 2 +++ +++ -

24 F 37 IDC 2 +++ +++ -

25 F 63 ILC 2 +++ ++ -

26 F 77 IDC 3 - ++ -

27 F 56 IDC 3 - - -

28 F 68 IDC 2 - - -

29 F 54 IDC 2 - - -

30 F 70 IDC 3 - + +

31 F 48 IDC 3 - ++ -

32 F 71 IDC 2 - - +

33 F 67 IDC 3 - - -

34 F 70 IDC 3 - - -

Table 22: Clinico-pathological information of patients whose tumours were analysed in chapter 3.

(IDC: invasive ductal carcinoma; IMC: invasive mixed carcinoma; ILC: invasive lobular carcinoma; ER: estrogen receptor; PR: progestrone receptor; HER2: human epidermal growth factor receptor 2).

Patient samples used in Chapter 4

Patients Sex Age Type of Cancer Grade ER PR HER2

1 F 50 IDC 3 - - 3+

2 F 94 IDC 2 - - -

3 F 33 IDC 2 - - -

4 F 57 IDC 3 - 1+ 3+

5 F 68 IDC 3 - - -

6 F 59 IDC 2 - 1+ -

7 F 52 IDC 3 - - -

8 F 45 IDC 3 - - 3+

9 F 57 IDC 3 - - 3+

10 F 79 IDC 2 - 1+ -

11 F 70 IDC 2 - 1+ -

12 F 44 IDC 3 - 1+ 3+

13 F 39 IDC 3 2+ 2+ -

14 F 44 IDC 3 1+ 1+ 3+

15 F 40 IDC 3 1+ - -

16 F 64 IDC 3 1+ - -

17 F 47 ILC 2 3+ 3+ -

18 M 62 IDC 2 3+ 2+ 2+

Table 23: Clinico-pathological information of patients whose tumours and lymph nodes were analysed in chapter 4. (IDC: invasive ductal carcinoma; IMC: invasive mixed carcinoma; ILC: invasive

71 lobular carcinoma; ER: estrogen receptor; PR: progestrone receptor; HER2: human epidermal growth factor receptor 2 ).

2.2 Methods

2.2.1 Mammalian cell culture

Cell lines origin, growth and passage

Human breast (MCF7, ZR-75-1 and MDA-MB-231) cell lines were obtained from

ATTC (Manassas, VA, USA) and maintained in Dulbecco’s modified Eagle’s medium

(DMEM; Gibco, Carlsbad, CA, USA) supplemented with 10% fetal calf serum (FCS;

Gibco, Carlsbad, CA, USA), 2mM glutamine (Sigma-Aldrich, Dorset, UK), 100U/ml penicillin (Sigma-Aldrich, Dorset, UK) and 0.1mg/ml streptomycin (Sigma-Aldrich,

Dorset, UK) at 37ºC with 5% CO2 (Table 8). Cells were routinely passaged when a confluence of ~90% was reached, depending on the growth curve of each cell line. To passage cells, medium was aspirated, cells were washed once with PBS and then trypsinised with trypsin-EDTA (Sigma-Aldrich, Dorset, UK) at 37°C for 3 to 10 minutes to allow them to detach. Dulbecco’s modified Eagle’s medium (DMEM; Gibco, Carlsbad,

CA, USA) supplemented with 10% fetal calf serum (FCS; Gibco, Carlsbad, CA, USA), was added to inactivate the trypsin (4:1 ratio) and cell clumps were disrupted through gentle pipetting. Suspension of cells was pipetted out of flask, and transferred to a sterile centrifuge tube. The cell suspension was centrifuged for 3 min at 1300 rpm. Upon centrifugation, the supernatant was aspirated and the cell pellet was resuspended in the appropriate volume of medium. The resulting suspension was split to the desired dilution into new flasks and fresh media was added. Cells were used between passages 4 and

20 in all experiments.

Estrogen and anti-estrogen treatments

Prior to treatment, cells were maintained for 3 days in DMEM without phenol red

(Gibco, Carlsbad, CA, USA) supplemented with 10% double charcoal-stripped FCS

(Gibco, Carlsbad, CA, USA), 2mM glutamine (Sigma-Aldrich, Dorset, UK), 100U/ml penicillin (Sigma-Aldrich, Dorset, UK) and 0.1mg/ml streptomycin (Sigma-Aldrich,

Dorset, UK) at 37ºC with 5% CO2. On the day of treatment, the media was changed to

DMEM without phenol red (Gibco, Carlsbad, CA, USA) supplemented with 10% FCS

(Gibco, Carlsbad, CA, USA), 100U/ml penicillin and 100mg/ml streptomycin. Cells were then treated with ethanol (used as negative control; Sigma-Aldrich, Dorset, UK), 10nM estradiol (17β-estradiol, E2; Sigma-Aldrich, Dorset, UK) or 100nM tamoxifen (4- hydroxytamoxifen, TAM; Sigma-Aldrich, Dorset, UK).

2.2.2 Luciferase assays

SK1 3’UTR luciferase reporter system

Amplification of the 3'UTR region of SK1 was achieved by Polymerase Chain

Reaction (PCR) using genomic DNA and the primers described in the Table 12. The

DNA fragment obtained was then cloned into the HindIII and SpeI sites of the multiple cloning sites (MCS) of pMIR-REPORT Firefly Luciferase vector (Applied Biosystems,

Life Technologies Ltd, Paisley, UK). After selecting the positive clones, sequencing confirmed the presence of the 3'UTR region of SK1 in the plasmid. Site-directed mutagenesis of the SK1 3’UTR was performed by using the Quick Change Site-Directed

Mutagenesis Kit (Stratagene, Agilent Technologies, Santa Clara, CA) according to the manufacturer’s instructions and using specific primer sets indicated in Table 12. The mutation was finally confirmed by sequencing.

For transfection, cells were plated in 24 well plates. Cells were then co-transfected in the following day with 100ng of pMIR-3’UTR SK1 or pMIR-3’UTRMUT SK1 (Firefly),

50ng of pRLTK (Renilla; Applied Biosystems, Life Technologies Ltd, Paisley, UK) and

50nM of precursor miRNA mimics (pre-miRs; Applied Biosystems, Life Technologies Ltd,

Paisley, UK) using 1.3µl of Lipofectamine™ 2000 (Invitrogen, Life Technologies Ltd,

Paisley, UK). After 24 hours, both Firefly and Renilla luciferase activities were quantified using the Dual-Glo® Luciferase Assay System (Promega, Madison, WI, USA). Firefly luciferase activity was normalised to the Renilla luciferase activity.

miR-515-5p’s promoter luciferase reporter system

Cloning of the miR-515-5p promoter region into the pGl2-basic plasmid (Promega,

Madison, WI, USA) was performed by Eurogentec (Southampton, UK; Table 10). Cells were plated in 24 well plates. After starving for 3 days, cells were treated with 10nM E2 and transfected with 200ng of miR-515-5p promoter pGl2-basic and 100ng of pRLTK

(Applied Biosystems, Life Technologies Ltd, Paisley, UK) using 1.5 µl of Fugene reagent

(Roche, Lewes, UK). After 24 hours, both Firefly and Renilla luciferase activities were quantified using the Dual-Glo® Luciferase Assay System. Firefly luciferase activity was normalised to the Renilla luciferase activity.

NRAS, MARK4 and PI3KC2B 3’UTR luciferase reporter systems

MCF7 and MDA-MB-231 cells were plated in 24 well plates and were allowed to adhere and grow overnight at 37°C with 5% CO2. In the following day, cells were co- transfected with pLightSwitch_3UTR GoClone vectors (Switchgear Genomics, Menlo

Park CA, USA) and 50 nM of microRNA precursors (pre-miRs mimics, Ambion, Applied

Biosystems, Life Technologies Ltd, Paisley, UK) using lipofectamine™ 2000 (Invitrogen,

Life Technologies Ltd, Paisley, UK). After 24 hours, cells were lysed using a passive

75 lysis buffer (Promega, Madison, WI, USA) and processed with the LightSwitch Assay

System (Switchgear Genomics, Menlo Park CA, USA). Lysates were transferred into an

Opti-plate 96-well plate, mixed with 50μl/well of Assay Solution and incubated for 30 min in the dark. Luminescence signals were read by a luminometer and averages of triplicates were calculated. Luciferase activity detection was performed using a GLOMAX

96 Microplate luminometer (Promega, Madison, WI, USA).

2.2.3 Transfections

miRNA and siRNA transfections

Cells were plated the day before transfection. Cells were transfected with 50nM of microRNA precursors (pre-miRs mimics, Ambion, Applied Biosystems, Life Technologies

Ltd, Paisley, UK), 200nM of anti-microRNAs (Applied Biosystems, Life Technologies Ltd,

Paisley, UK) or 20nM of siRNA (SK1 siRNA - Applied Biosystems, Life Technologies Ltd,

Paisley, UK; MARK4 siRNA - Dharmacon, Table 11) using the HiPerFect Transfection

Reagent (Qiagen, West Sussex, UK). For transfections in 24 well plates, microRNAs or siRNAs were diluted in 100µl of Opti-MEM® I Medium (Life Technologies Ltd, Paisley,

UK) and after adding 3µl of Hiperfect Transfection Reagent (Qiagen, West Sussex, UK), the solution was mixed by vortexing. The samples were incubated for 10 minutes at room temperature to allow the formation of transfection complexes. The complexes were then added drop-wise onto the cells and the plates were gently swirled before being incubated under normal growth conditions.

Plasmid transfections

Cells were plated and allowed to adhere and grow overnight at 37ºC with 5% CO2.

On the day of transfection, plasmids (control - pCDNA 3.0; HA-MARK4 plasmid) were diluted in 500µl of Opti-MEM® I Medium (Life Technologies Ltd, Paisley, UK) and after adding 1.5µl of Attractene Reagent (Qiagen, West Sussex, UK), the solution was mixed by vortexing. The samples were incubated for 10 minutes at room temperature and complexes were then added drop-wise onto the cells. Plates were gently swirled before being incubated under normal growth conditions. Details of both backbone and insert of the HA-MARK4 plasmid are indicated in Table 9.

2.2.4 Quantitative real-time Reverse Transcription-PCR

RNA extraction

RNA was extracted using Trizol (Invitrogen, Life Technologies Ltd, Paisley, UK) according to the protocol recommended by the manufacturer. Chloroform was used to separate the homogenised samples (Trizol + lysed cells) into three phases: phenol- chloroform phase, an interphase, and a colourless upper aqueous phase. The former two phases contain the protein and DNA content, respectively, and RNA is exclusively in the aqueous solution. Isopropanol was added to the aqueous phase to precipitate the

RNA which was then washed using ethanol. In the end, RNA was resuspended in 20-30

μl RNA-free water. After both extractions, RNA concentration was measured at 260nm and 280nm wavelengths using a NanoDrop ND-100 Spectrophotometer (NanoDrop

Technologies). RNA quality was analysed loading the RNA samples in an agarose gel.

As the presence of two bands around 2Kb and 5Kb correspond, respectively, to 18S and

28s ribosomal RNA subunits, the 28S:18S intensity ratio of approximately 2 indicates the presence of pure and non-degradated RNA.

mRNA reverse transcription and qRT-PCR

To analyse mRNA expression, both cDNA conversion and qPCR amplification were performed as recommended by Applied Biosystems using the High-Capacity cDNA

Reverse Transcription Kit and Fast SYBR® Green Master Mix (Applied Biosystems, Life

Technologies Ltd, Paisley, UK), respectively. The primers used for NRAS, PI3KC2B,

FZD4, CDC42BPA, MARK4 and GAPDH are detailed in Table 13. The thermal cycles were carried out using an ABI PRISM 7900 sequence detection system (Applied

Biosystems, Life Technologies Ltd, Paisley, UK) and were the following: 10 minutes at

95°C, 40 cycles of 15 seconds at 95°C and 60 seconds at 60°C followed by the dissociation step, 15 seconds at 95°C, 15 seconds at 60°C and 15 seconds at 95°C.

Each sample was quantified in triplicate and the data was analysed using the ΔΔCT method. The expression of the genes was normalised to the expression of the

Glyceraldehyde-3-phosphate dehydrogenase (GAPDH) housekeeping gene.

microRNAs reverse transcription and qRT-PCR

Both cDNA conversion and PCR amplification were performed as recommended in the manual “TaqMan MicroRNA Assays” (Applied Biosystems, Life Technologies Ltd,

Paisley, UK). For the first reaction, 10ng of pure RNA were diluted in 5µl of RNA-free water and 0.15µl of 100mM dNTPs, 1µl of MultiScribe™ Reverse Transcriptase 50 U/μL,

1.5µl of 10x Reverse Transcription Buffer, 0.19µl of RNase Inhibitor 20U/μL and 4.16µl of Nuclease-free water (TaqMan MicroRNA Reverse Transcription kit) were mixed to prepare the RT Master mix solution. The 5µl of pure RNA was then combined with 7µl of master mix and 3µl of 5x Taqman primers (Applied Biosystems, Life Technologies Ltd,

Paisley, UK). cDNA synthesis was finally performed using ABI Thermal Cycler (Applied

Biosystems, Life Technologies Ltd, Paisley, UK) and the following cycles: 16°C for 30 minutes, 42°C for 30 minutes and 95°C for 5 minutes to inactivate the reverse transcriptase enzyme. For each PCR reaction, 10ul of TaqMan 2x Universal PCR Master

Mix (No AmpErase UNG) were diluted in 7.67 ul of RNA-free water and mixed with 1µl of

20x Taqman primers and 1.33µl of the obtained cDNA. After adding the samples to a

Fast Optical 96-well reaction plate, the microRNAs were amplified using the following cycles: 10 minutes at 95°C and 40 cycles of 15 seconds at 95°C and 60 seconds at

60°C (ABI PRISM 7900 sequence detection system (Applied Biosystems, Life

Technologies Ltd, Paisley, UK). Each sample was quantified in quadruplicate and using the ΔΔCT method, the microRNAs expression was normalised to the U6 small nuclear

RNA expression.

2.2.5 Western blot

Growth media was aspirated and the cells were washed once with PBS. 150µl of 1x

SDS loading buffer was added to each well and cells were harvested by scraping. The samples were then sonicated to reduce the viscosity of the sample by fragmenting nuclear DNA and the total protein was quantified using the Dc Protein Assay kit (Bio-Rad

Laboratories, Hercules, CA, USA). 10µg of protein from each sample was mixed with 5x

SDS loading buffer and boiled for 5 minutes at 95°C. Proteins were separated using a

5% acrylamide stacking gel [5% (v/v) acrylamide, 0.125 M Tris, pH 6.8, 0.1% (w/v) SDS,

0.075% (w/v) APS and 0.083% (v/v) TEMED], and a 12.5% acrylamide resolving gel

[12.5% (v/v) acrylamide, 0.375 M Tris pH 8.8, 0.1% (w/v) SDS, 0.06% (w/v) APS, 0.07%

(v/v) TEMED]. Gels were run using 1x SDS running buffer (Table 14) at 120 V until the dye reached the bottom of the gel. Separated proteins were transferred onto

Polyvinylidene fluoride (PVDF) membrane (Millipore, Billerica, MA, USA) in 1x transfer buffer (Table 14) at 100V for 1 hour using a tank blotting system (Bio-Rad Laboratories,

Hercules, CA, USA) in wet transfer conditions. The membranes were then incubated in

Ponceau S red solution (Sigma-Aldrich, Dorset, UK) for 10 minutes and rinsed with water to check if the transfer was successful. In order to prevent non-specific binding of antibody, PVDF membranes were incubated with 5% non-fat dried milk powder in 0.05%

Tween-20 in 1xPBS (PBS-T, Table 14) with gentle shaking for one to two hours at room temperature. Membranes were then washed with PBS-T for 3 times for 5 minutes before being incubated with a primary antibody. Primary antibodies (Table 19) were diluted in

5% (w/v) BSA/PBS-T containing sodium azide and incubated overnight at 4 ˚C with gentle shaking. Membranes were then washed three times for 5 minutes in PBS-T.

Secondary antibodies conjugated to horseradish peroxidase (HRP) against mouse

(Table 19) were diluted in 5% (w/v) milk/ PBS-T and incubated for 1 to 2 hours at room temperature. Membranes were washed again 3-4 times for 5 minutes with agitation before being exposed to ECL reagents (Millipore, Watford, UK) and exposed to Kodak

GRI autoradiography film. Band intensities were measured and quantified by ImageJ.

2.2.6 Immunoprecipitation assay

Cell pellets were collected, resuspended in lysis buffer (Table 17) and passed 3 times through a 25G needle and once through a 27G needle. The lysate was then pre- cleared by being rotated with the beads (Protein G Sepharose Fast Flow; Sigma-Aldrich,

Dorset, UK) for 2 hours at 4ºC. In parallel, the beads were washed three times (washing buffer – Table 17) and then conjugated with 10µg of each antibody (Table 20) and heparin (final concentration of 1mg/ml (AppliChem GmbH, Darmstadt, Germany)) by rotation for 2 hours at 4ºC. Next, the pre-cleared lysate was divided into the appropriate

80 antibodies-beads samples and rotated for 4 hours at 4ºC. The beads were then washed once with lysis buffer and three times with washing buffer (Table 17). Next, the beads were reconstituted by adding proteinase K (New England Biolabs, Herts, UK) and DNA was degraded by adding DNAse (Promega, Madison, WI, USA). The RNA fraction was then isolated by phenol-chloroform extraction (same method above mentioned using

Trizol) using Acid-Phenol:Chloroform, pH 4.5 (Invitrogen, Life Technologies Ltd, Paisley,

UK) and the RNA was precipitated by adding 100% ethanol and leaving the samples at -

20ºC for at least 20 minutes. The RNA pellet was washed with 70% ethanol and finally resuspended in RNase free water.

2.2.7 SK1 activity assay

Cell pellets were resuspended in SPHK1 buffer, sonicated for 3-5 seconds and centrifuged at 20,000g for 30 minutes at 4°C. The protein concentration of the supernatant was measured by Bradford protein assay (Bio-Rad Laboratories, Hercules,

CA, USA) according to the manufacturer’s instructions. For the enzymatic assay, 100 µg of total protein of each sample were diluted in 180µl of SPHK1 buffer and mixed with 2µl of 20mM MgCl2, 0.2µl of 10mM ATP, 6.8µl of H2O. 10µl of 50 mM sphingosine were then added to each mixture, being this sphingosine solution prepared by drying 10 µl of 50 mM sphingosine diluted in ethanol (Avanti Polar Lipids, Alabaster, AL, USA), resuspending the dried sphingosine was in 500µl 0.25% triton and sonicating the obtained solution for 15 minutes in a water bath. 1µl of 10 µCi [γ-32P]-ATP (Perkin-Elmer,

Waltham, MA) was finally added to each enzymatic reaction immediately before its incubation. In each assay, two standards reactions were also prepared as positive control. For the sphingosine of these standards, 10µl of 50 µM sphingosine in ethanol were dried as described above; being in this case the dried sphingosine resuspended in

81 only 10 µl of 0.25% triton. The obtained sphingosine was then mixed with 2µl of 20mM

MgCl2, 0.2µl of 10mM ATP, 6.8µl of H2O, 180µl of SPHK1 buffer and 1 or 2µl of SPHK1.

As for the sample reactions, 1µl of 10 µCi [γ-32P]-ATP (Perkin-Elmer, Waltham, MA) was added immediately before their incubation. Both sample and standard reactions were incubated for 1 hour at 37 °C in a 5% CO2. To stop the reaction, solutions were added in the following order; 50µl 1M HCl, 800µl Chloroform/Methanol/HCl mix, 240µl CHCl3 and

240µl 2M KCL. All mixtures were then vortexed and centrifuged for 10 minutes at

2500rpm. The lower (organic) phase was transferred into a new 2ml micro-centrifuge tube with 800µl of SPHK washing solution. Samples were centrifuged as described previously and the lower phase was again transferred into a new 1.5ml micro-centrifuge tube and left to vaporise overnight. The samples were solubilised in 40µl

Chloroform/Methanol (2:1 v/v) and added to a thin layer chromatography on silica gel

G60 plate (Whatman, GE Healthcare, Waukesha, WI, USA). The plate was placed into a closed container with 200-300ml of 1-butanol/ethanol/acetic acid/water (80:20:10:20 v/v) and left to migrate for a minimum of 4 hours. Plates were left to dry for another 30 minutes under the fume hood before being visualised by autoradiography. The radioactive spots were quantified using standards and ImageJ software. All the buffers and reagents used in this assay are described in Table 15 and 16.

2.2.8 Cell proliferation assays

MTT assays

Cells were plated in 24 well plates and were allowed to adhere and grow overnight at 37ºC with 5% CO2. After 24 hours, the cells were transfected with microRNAs using the protocol described in section 2.2.3. and the amounts recommended for 24 well plates

(Qiagen, WestSussex, UK). In the following days after transfection, 150µl of MTT

82 solution (5 mg/mL in PBS) was added to the wells containing media, cells and the transfection complexes (600µl). After 3,5 hours of incubation at 37°C with 5% CO2, the media were carefully removed and 600 µl of Extraction buffer (100ml SDS 20% + 3.87ml

DMF) was added to dissolve the formazan produced by the cells. The plates were incubated 2 more hours at 37°C with 5% CO2 and after pipetting up and down to completely dissolve the formazan crystals, the solution from each well was transferred to

5 wells of a 96 well plate. The absorbance was measured at 570nm using a spectrometer. Table 18 indicates the components of the buffers used in this assay.

Trypan blue assays

Cells from each well were trypsinised and diluted in cell growth media. The detached cells were then stained using Trypan Blue (Sigma-Aldrich, Dorset, UK) in a 1:1 ratio, being the live cells (unstained) then counted using a haemocytometer.

Cell apoptosis assays

Cell apoptosis was determined by measuring the activity of caspases 3 and 7

(Caspase-Glo® 3/7 Assay, Promega, Madison, WI, USA) and the assay conducted using 24 well plates. At the indicated times, the culture media from each well was collected; cells were trypsinised (100µl) and then washed with the respective culture media (500µl). The resulting solution was then mixed and 50µl was transferred to a white

96-well plate. 100µl of Caspase-Glo reagent was then added to each well and the plate incubated in the dark at room temperature. After 1 hour, the luminescence was measured and normalised to the total protein, quantified using BCA Protein Assay Kit

(Pierce, Rockford, USA).

2.2.9 Immunofluorescence

Cells were fixed in 4% paraformaldehyde (PFA) at 37 ºC for 15 min, washed in PBS and then blocked in 3% BSA at room temperature for 1 hour. Actin cytoskeleton was stained with Alexa 488 Phalloidin (1:1000; Invitrogen, Life Technologies Ltd, Paisley,

UK) and nuclear DNA was visualised using DAPI (1:200; Invitrogen, Life Technologies

Ltd, Paisley, UK) at 4ºC overnight (Table 21). The following day, the stained cells were washed twice with PBS. Images were acquired using an ImageXpress High Content

Screening microscope (Molecular Devices, Sunnyvale, CA, USA).

2.2.10 Cell migration assays

Cell tracking assay

1x103 cells per well were plated in black, flat and clear bottomed 96 well plates (BD

Biosciences, Oxford, UK). Time-lapse imaging was performed for 18 hours (1 image for every 10 minutes) using a motorised-staged environment-controlled ImageXpress high content screening microscope powered by MetaXpress 2.0 (Molecular Devices,

Sunnyvale, CA, USA). Two sites were acquired per well, being each condition present in triplicate. To quantify the degree of migration a minimum of 30 cells per condition were tracked using the MetaXpress Track Points application. Tracks were analysed using a previously published Mathematica (Wolfram Research) notebook [202].

Boyden chamber assay

For transwell migration assays, 5 x 104 cells were seeded on the top of uncoated membranes with 8.0 µm pores (BD Biosciences, Oxford, UK). Cell were plated in serum

84 free-medium and allowed to migrate toward a complete growth medium for 8 hours. The cells on the bottom part of the filter were then stained using crystal violet (Sigma-Aldrich,

Dorset, UK). Stained cells were imaged using Olympus BX51 microscope and quantified using ImageJ [203].

2.2.11 RNA-Seq and ChIP-Seq analysis

Two micrograms of total RNA from each sample were used to produce cDNA libraries from polyA enriched RNA using the True-Seq RNA preparation kit (Illumina Inc.,

San Diego, CA, USA) according to the manufacturer instructions. Briefly, to isolate the mRNAs from the non-coding RNA, the extracted RNA from each condition was initially used as the input to an mRNA capture with oligo-dT coated beads. The obtained mRNA was then cleaved and a second strand of cDNA was synthesised for each fragment using random primers. The resulting double stranded fragments were then phosphorylated and A-tailed for adapter ligation. Different adapters were added to the end of the fragments for each condition and after PCR amplification, the cDNA library was ready for clustering and sequencing. The HiSeq 2000 instrument (Illumina Inc., San

Diego, CA, USA) generated paired end sequences (reads) 100 nucleotides in length and the Fastq files containing the sequenced reads, obtained at the end of the sequencing, were mapped to the University of California at Santa Cruz (UCSC) human genome

(hg19 assembly) with TopHat version 1.4.1 (http://tophat.cbcb.umd.edu), using default settings. The mapped bam files obtained at the end of the runs were then analysed using Cufflinks version 2.1.0 (http://cufflinks.cbcb.umd.edu) for RNA quantification and analysis. To evaluate whether ERα interacts with the miR-515-5p promoter, we re- analysed ChIP-Seq published data. We downloaded the samples GSM614606,

GSM614607, GSM614610 and GSM614611 from the GSE25021 dataset and after

85 adaptor removal and bioinformatics pre-processing of the samples, we used maqs

(liulab.dfci.harvard.edu/MACS/ ) to find significant peaks of ERα interaction. We used

UCSC genome browser (http://genome.ucsc.edu) to map the obtained bed files, containing the significant peaks, onto the Hg19 human assembly and zoomed on the locus containing miR-515-5p.

2.2.12 Tumour tissues

Analysis of miRNAs in formalin-fixed paraffin embedded (FFPE) tissues was approved by a UK national research ethics committee (London; 07/Q0401/20) and by

Imperial College Healthcare NHS Trust. To analyse microRNA expression according to the ER status, three sections of 5µm sections were obtained from FFPE tumour samples

(n=34) and macrodissected. Clinicopathological information of the patients is provided in

Table 22. To analyse microRNA expression in primary tumours and corresponding lymph node metastasis, we obtained breast cancer tissues at Imperial College

Healthcare NHS Trust, London, UK, between January 2008 and February 2010. All samples were formalin-fixed and paraffin-embedded and the clinicopathological details of patients are provided in Table 23.

2.2.13 Statistical analysis

The statistical analysis of the data presented in Chapters 3 and 4 was performed using GraphPad Prisma 5. To analyse the variance of unpaired and paired populations, all experiments were performed independently at least three times. Paired and unpaired t-test (95% confidence) was used to assess the variance between populations. To

86 measure the statistical dependence between two variables, Spearman’s Rank

Correlation test was used [204]. The Spearman coefficient (R) varies between -1 and 1; the sign (positive or negative) indicates the direction of the correlation between the two variables and its magnitude increases as X and Y become closer to being a perfect monotonic function of each other. To assess the effect of a gene or microRNA on the prognosis of patients, survival values were divided into two cohorts (high and low expression of the analysed factor) by a Kaplan-Meier plot that defined the survival function for each cohort [205]. The Hazard Risk (HR) test (with 95% of confidence interval) was assessed to compare the survival distribution of the two cohorts. HR test corresponds to a ratio of death probabilities, it indicates the number of times the factor increases the risk of death [206]. In all statistical analysis, P value lower than 0.05 were considered to be statistically significant.

Chapter 3 Downregulation of miR-515-5p by ERα leads to tumorigenesis via SK1

Downregulation of miR-515-5p by estrogen receptor

leads to tumorigenesis via SK1

3.1 Overview

About 75% of breast cancers are estrogen receptor-alpha (ERα) positive and cell proliferation in these tumours is highly estrogen-dependent [43]. ERα-positive breast cancers have been successfully treated with hormonal therapies that either block estrogen production or antagonise its action [44]. Among these, tamoxifen (TAM) blocks estrogen binding to the ERα and is primarily responsible for the dramatic increase in survival of breast cancer patients over the past 30 years [48].

Sphingosine kinase 1 (SK1) is the key regulator of the sphingolipid rheostat, which helps maintain the balance between cell growth / survival and cell cycle arrest / apoptosis. It converts the pro-apoptotic sphingosine into the mainly pro-survival sphingosine 1-phosphate, mediating breast cell survival, proliferation and migration [95].

In ERα-positive breast cancer, SK1 has been demonstrated to mediate estrogen- dependent tumorigenesis [96]. Blocking SK1 expression using siRNAs or selective inhibitors abolishes the response of breast cancer tumours in vivo to estrogens and restores the sensitivity of breast cancer cells resistant to anti-estrogens and TAM [207].

This indicates the requirement of SK1 in the estrogen response mechanism and the importance of this enzyme in breast cancer hormonal chemoresistance [207]. Although estrogens have been described to contribute to a higher activation of S1P receptors, by stimulating the sphingosine 1-phosphate export [208], the mechanism underlying the upregulation of SK1 activity by estrogens is still unclear.

MicroRNAs are small non-coding RNAs which negatively regulate target gene expression. By binding to the 3'UTR, 5'UTR or coding region of their gene targets, they induce mRNA destabilization/degradation and translational repression [117-119]. They are able to act as tumour suppressors or oncogenes, and we and others have shown that a subset of microRNAs is regulated by estrogen [168, 188, 209]. This indicates that microRNAs may be a new generation of anti-cancer molecules and miRNA-based therapies could provide an alternative to standard chemotherapeutics [210].

In this chapter, we aimed to elucidate the mechanism by which estrogen upregulates SK1 levels. Since dysregulation of microRNA expression has been observed in breast cancer cell lines and tumours upon estrogen or TAM treatment [188], we investigated whether microRNAs control the oncogenic role of SK1 in estrogen- dependent breast cancer. We found two microRNAs able to directly regulate SK1 expression, and we showed that loss of miR-515-5p in particular results in increased oncogenic SK1 activity. Furthermore, miR-515-5p overexpression in breast cancer cells inhibited cell proliferation and induced caspase-dependent apoptosis. Analysis of ChIP-

Seq data and RNA interference (RNAi) of ERα revealed that ligand-activated ERα directly suppresses miR-515-5p gene transcription, and accordingly TAM treatment upregulated mature levels and decreased SK1 activity. RNA-Seq of breast cancer cells after miR-515-5p overexpression revealed several downregulated transcripts involved in apoptosis and in the Wnt pathway. Lastly, we found that miR-515-5p expression is significantly lower in ERα-positive than ERα-negative breast cancer tumours. Together these evidences suggest miR-515-5p replacement strategies as a possible novel therapeutic approach in hormone resistant ERα-positive breast cancer.

3.2 Results

3.2.1 SK1 mRNA directly interacts with Ago2

As SK1 has a short 3'UTR similar in length to the 3'UTR of housekeeping genes, we started investigating whether SK1 mRNA was regulated by any microRNAs. We performed an immunoprecipitation (IP) in MCF7 cells using Ago2 and IgG (negative control) antibodies and in the subsequent RNA isolated from both IP fractions we analysed SK1 mRNA expression. We observed an enrichment of SK1 mRNA levels

(50%, P<0.05; Figure 11) in the Ago2 fraction in comparison with the IgG fraction. As

Ago2 is part of the RNA-induced silencing complex (RISC) through which mature microRNAs cause gene silencing, the presence of SK1 mRNA in the Ago2 fraction suggested the existence of endogenous microRNAs that are able to directly interact with

SK1 mRNA and regulate its expression.

Figure 11: SK1 mRNA expression is higher in Ago2 immunoprecipitates. Immunoprecipitation (IP) was performed in MCF7 cells using Ago2 and IgG antibodies. SK1 mRNA expression was then analysed in the RNA extracted from each IP fraction. Data are the mean of three independent experiments ± SEM. P values were calculated by paired t-test comparing Ago2 values with IgG (negative control) values (*:P<0.05).

3.2.2 miR-515-5p directly regulate SK1 expression

To identify human microRNAs that potentially target SK1, we used the Targetscan web-based software (Figure 12A). Since SK1 has an oncogenic role, we selected for further investigation the microRNAs that are known to act as tumour suppressors or known to be downregulated in at least one cancer type. Apart from miR-124, miR-342 and miR-668 which have been described to be downregulated or act as a tumour suppressors across different tissues, all the other microRNAs are more controversial because their tumorigenicity is dependent on the type of cancer tissue (Table 24).

Human Literature microRNAs

 is epigenetically silenced in hepatocellular carcinoma [211], cervical cancer [212, 213] and myeloid leukemia [213];  is downregulated in medulloblastoma [214] and inhibits the growth of medulloblastoma cells miR-124 [215, 216];  inhibits proliferation of glioblastoma multiforme cells [217];  suppresses multiple steps of breast cancer metastasis [218];  inhibits tumorigenesis of prostate cancer cells [219, 220].

 is silenced in germinal center-derived lymphoma [221];  (miR 28-5p) is downregulated in colorectal cancer and reduces colorectal cancer cell miR-28 proliferation, migration and invasion in vitro [222];  is upregulated in human cervical cancer [223] and renal cell carcinoma [224];  is upregulated in malignant gliomas [225].

 is downregulated in intestinal tumours [226]; miR-766  is upregulated in serum of prostate cancer patients [227], in renal cancer cells [228] and tissues of pituitary carcinomas [229] and lymphoblastic leukemia [230].

 is upregulated in human MCF7 breast cancer cell line by epigenetic therapy (treatment with inhibitors of DNA melhyltransferases which reactivate the transcription of silenced genes miR-515-5p and slow cell proliferation) [231];  is overexpressed in oral carcinoma [231, 232].  Play an oncogenic role in hepatocellular carcinoma [277],

 inhibits cell growth and invasion of prostate cancer cell lines [233];  is lost during colorectal, breast, lung, gastric, parathyroid, liver and bile ducts tumour miR-296 progression [234];  contributes significantly to angiogenesis in endothelial cells [235];  is upregulated in esophageal carcinoma and inhibits growth of esophageal cancer cells in vitro and in vivo [236].

miR-608  is lost in colorectal tumour cells [237];  Its high expression is associated with a poor prognosis of large B cell lymphoma [238].

 is commonly suppressed in human colorectal cancer by a genetic mechanism of silencing miR-342 (CpG island methylation) [239];  inhibits colorectal cancer proliferation and invasion [240];  its downregulation is associated with tamoxifen-resistant breast cancer tumours [241].

miR-668  is downregulated in head and neck squamous cell carcinoma cell lines [242] and in breast tumour tissues [243].

 targets ERα receptor in breast cancer cell lines and inhibits MCF7 cell growth [193, 244]; miR-206  attenuates cell growth and promotes cell apoptosis of 95D, a lung cancer cell line [245];  blocks human rhabdomyosarcoma growth in vivo [246];  its levels are elevated in the sera of rhabdomyosarcoma patients [247].

Table 24: Bibliographic references of the microRNAs predicted to target the SK1 3’UTR.

In order to verify whether these microRNAs directly target SK1, we cloned the SK1

3'UTR region into the pMIR-REPORT vector. Among the nine selected miRNAs, miR-

206, miR-515-5p and miR-766 significantly reduced luciferase activity (40%, P<0.01;

50%, P<0.001 and 20%, P<0.05 respectively), proving for the first time that these microRNAs directly interact with the SK1 3'UTR (Figure 12B). As miR-766 showed the least reduction in the luciferase activity, it was not considered further.

Figure 12: Identification of microRNAs that directly interact with SK1 3’UTR. (A) microRNAs predicted to target SK1 3’UTR using TargetScan. microRNAs targeting SK1 3’UTR were predicted using the web- based software, TargetScan. The microRNAs highlighted were selected for further analysis and experiments. (B) miR-206, miR-515-5p and miR-515-5p directly interact with SK1 3’UTR. Cells were co- transfected with 50nM of pre-miRs, 100ng of pMIR-SK1 3’UTR and 50ng of pRLTK (Renilla luciferase vector). After 24 hours, both Firefly and Renilla luciferase activities were measured. Data were normalised to the negative control result (miR NC) and presented as the mean of three independent experiments ± SEM. P values were calculated by paired t-test comparing individual miRNA values with miR NC values (*:P<0.05, **:P<0.01, ***:P<0.001).

To validate SK1 as a direct target of miR-206 and miR-515-5p, we next measured

SK1 activity levels in MCF7 cells upon overexpression of each microRNA (Figure 13A). miR-206 and miR-515-5p induced a reduction of 75% (P<0.001) and 45% (P<0.001) in

SK1 activity, respectively. For miR-515-5p, the decrease in SK1 activity was also

94 accompanied by a minor but significant reduction in SK1 mRNA levels (P<0.001; Figure

13B), without any change in activated (phospho-) ERK expression (Figure 13C)1.

Figure 13: miR-515-5p and miR-206 regulate SK1 expression. (A) Effect of miR-206 and miR-515-5p overexpression on SK1 activity. MCF7 cells were transfected with 50nM of precursor miRNA. After 48 hours, SK1 activity was measured. Data are presented as the mean ± SEM and P values were calculated by t-test comparing individual miRNA values with miR NC values (n=3, ***:P<0.001). (B) The effect of miR-206 and miR-515-5p overexpression on SK1 mRNA levels and (C) phospho-ERK 1/2 expression. MCF7 cells were transfected with 50nM of precursor miRNAs. After 48 hours, SK1 mRNA levels were measured by RT-qPCR and phospho-ERK 1/2 protein levels were analysed by immuno-blotting. Data are presented as the mean ± SEM and P values were calculated by paired t-test comparing individual miRNA values with miR NC values (n=3, ***:P<0.001).

We then mutated the miR-515-5p seed region in the SK1 3’UTR (at the point detected by the TargetScan analysis) located in the luciferase plasmid to evaluate whether miR-515-5p regulates SK1 by directly binding to this region. miR-515-5p

1 We tested all the SK1 antibodies which are commercially available, but we were unable to find any that was capable of recognizing the SK1 protein band by western blot. The antibodies highlighted a number of non-specific bands that did not correspond to the molecular weight of SK1 and we further confirmed this using siRNAs against SK1 at several time points. We did not see any effect on the intensity of these non-specific bands, demonstrating that the tested antibodies were unable to detect SK1 in our samples. Therefore, we used surrogate markers to quantify SK1 expression. It is known that a change of SK1 activity can be due to 3 factors: change in SK1 mRNA expression, alteration in SK1 protein levels and an increase in SK1 phosphorylation (post-translation modification). For this reason, apart from analysing SK1 activity levels, to assess SK1 expression we have always quantified the SK1 mRNA levels and the levels of the protein responsible to activate SK1 by phosphorylation, the activated form of ERK 1/2 (P-ERK 1/2). These data are presented in the Figures 13 and 24. 95 overexpression significantly increased the luciferase activity of the cells transfected with the plasmid containing the mutated sequence (SK1-3’UTRMUT pMIR) in comparison with the cells transfected with the plasmid containing the wild type sequence, proving that miR-515-5p directly targets SK1 by directly interacting with the TargetScan-predicted seed region of 3’UTR (Figure 14).

Figure 14: miRNA-515-5p directly interacts with SK1 3’UTR. Cells were co-transfected with 50nM of pre- miRs, 100ng of SK1-3’UTR pMIR or SK1-3’UTRMUT pMIR and 50ng of pRLTK (Renilla luciferase vector). After 24 hours, both Firefly and Renilla luciferase activities were measured. Data were normalised to the negative control result (miR NC) and presented as the mean of three independent experiments ± SEM. P values were calculated by paired t-test comparing individual miR values with miR NC values. (**:P<0.01, ***:P<0.001).

3.2.3 miR-515-5p inhibits breast cancer cell proliferation

Since SK1 mediates cell proliferation in cancer cells [248], we used MTT assays to determine the effect of miR-206 and miR-515-5p overexpression on breast cancer cell viability and proliferation (Figure 15A). miR-206 overexpression has been previously associated with decreased MCF7 cell viability [193], and therefore miR-206 was not considered further. miR-515-5p induced a 70% decrease in MCF7 cell viability (P<0.001;

Figure 15A), which was accompanied by a significant decrease in the expression of the anti-apoptotic Bcl-2 protein (P<0.01, Figures 15B and 15C).

Figure 15: miR-515-5p inhibits cell proliferation in MCF7. MCF7 cells were transfected with 50nM of precursor miRNAs. After 4 days, (A) MTT assays were performed and and (B, C) Bcl-2 protein expression was measured in MCF7. Data are presented as the mean of three independent experiments ± SEM and P values were calculated by paired t-test comparing individual miRNA values with miR NC values (**:P<0.01, ***:P<0.001).

To clarify the cause behind the reduced cell proliferation upon miR-515-5p, we conducted trypan blue (Figures 16A and 16B) and caspase 3/7 activity (Figures 16C and 16D) assays to further evaluate the effect of miR-515-5p on cell viability and apoptosis, respectively. In two breast cancer cell lines, proliferation was dramatically inhibited by miR-515-5p (70% of viable MCF7, P<0.001; 42% of viable ZR-75-1,

P<0.001); Figure 16A and 16B, respectively). Additionally, the activities of caspases 3 and 7 were significantly increased after 48 hours of treatment, indicating that apoptosis

97 had been activated by miR-515-5p (in both MCF7 and ZR-75-1 cells, P< 0.001; Figures

16C and 16D, respectively).

Figure 16: miR-515-5p reduces cell growth and induces cell apoptosis in MCF7 and ZR75-1. MCF7 cells were transfected with 50nM of precursor miRNAs. (A, B) The effect of miR-515-5p in MCF7 and ZR75- 1 cell growth. In the indicated days, viable cells were counted using Trypan Blue. presented as the mean of three independent experiments ± SEM and P values were calculated by paired t-test comparing individual miRNA values with miR NC values (**:P<0.01, ***:P<0.001). (C, D) The effect of miR-515-5p in MCF7 and ZR75-1 cell apoptosis. In the indicated days, caspase 3/7 activities were measured in MCF7 and ZR75-1 cells. Data are presented as the mean of three independent experiments ± SEM and P values were calculated by paired t-test comparing individual miRNA values with miR NC values (*:P<0.05, **:P<0.01).

3.2.4 Silencing SK1 partially rescues the effect of miR-515-5p inhibition on breast cancer cell proliferation

We next investigated the effect of miR-515-5p by loss-of-function on cell proliferation and apoptosis (Figure 17). We observed that inhibition of miR-515-5p in MCF7 cells caused an increase in cell proliferation (P<0.001; Figure 17A) and a decrease in cell

98 apoptosis activation (P<0.001; Figure 17B). In order to investigate whether an increase in SK1 expression was mediating this increase in cell proliferation and inhibition of cell apoptosis caused by the anti-miR-515-5p treatment, we silenced SK1 using a siRNA

(siSK1) together with miR-515-5p inhibition in MCF7 cells (Figures 17C and 17D). SK1 silencing induced a decrease in the viability of cells co-transfected with anti-miR-515-5p compared with siRNA negative control (siNC; Figures 17C and 17D). However, silencing SK1 was not enough to rescue the total effect of anti-miR-515-5p (neither the decrease in cell proliferation (P<0.01; Figure 17C), nor the increase in caspase-7 activity (P<0.01; Figure 17D)).

Figure 17: SK1 silencing partially rescues the enhanced cell proliferation after miR-515-5p inhibition. Cells were transfected with 200nM of anti-miRNAs and/or 20nM of siRNAs (Ambion). Viable cells were counted using (A, C) trypan blue and (B, D) caspase 3/7 activities were measured over time. presented as the mean of three independent experiments ± SEM and P values were calculated using paired t-test (**:P<0.01; ***:P<0.001).

3.2.5 miR-515-5p targets many oncogenes with an enrichment of genes belonging to the Wnt Pathway

Our data so far indicated that apart from SK1, it is likely that other oncogenic targets are involved in the tumour suppressor role of miR-515-5p. To confirm this hypothesis, we overexpressed miR-515-5p continuously in breast cancer cells for 6 days and performed

RNA-Seq analysis firstly to identify other targets of this miRNA and secondly to evaluate whether any of these regulated transcripts is involved in cell proliferation (Figure 18).

Notably, SylArray analysis (http://www.ebi.ac.uk/enright-srv/sylarray/) indicated that the downregulated transcripts are significantly enriched with miR-515-5p ‘seeds’ (Figure

18A), indicating that there were direct targets among the downregulated genes. In order to identify the high confidence direct targets among these, we considered transcripts that were commonly downregulated (<-1.2 fold) between two replicates and that were predicted as potential targets by TargetScan. This analysis identified 212 high confidence miR-515-5p directly regulated transcripts. Interestingly, GO terms and pathway enrichment analysis identified apoptosis and Wnt pathways as significantly regulated by this miRNA (Figure 18B).

Notably, the genes downregulated by miR-515-5p that also belong to the Wnt signaling pathway are all known to positively regulate cell proliferation. B-cell

CLL/lymphoma 9 (BCL9) protein improves β-catenin-mediated transcription [249]; DIX domain containing 1 (DIXDC1) targets p21 and cyclin D1 through the activation of the

PI3K/AKT pathway [250]; frequently rearranged in advanced T-cell lymphomas 2

(FRAT2) activates the β-catenin-TCF signaling pathway [251]; frizzled 4 (FZD4) is a mediator of the ERG oncogene–induced Wnt signaling [252]; and transcription factor 7- like 1 (TCF7L1) is a transcription factor activated by β-catenin that mediates Wnt signaling [253]. In addition, we identified important genes that have been shown to promote breast cancer, such as fibroblast growth factor receptor 2 (FGFR2) and

100 phosphatidylinositol-4-phosphate 3-kinase C2 domain-containing beta polypeptide

(PIK3C2B) [254, 255], and these seem to be directly regulated by miR-515-5p. We further validated miR-515-5p mediated regulation of a few of these genes (FGFR2,

TCF7L1 and PIK3C2B) by using RT-qPCR (Figures 18C, 18D and 18E, respectively).

Thus, we demonstrate here that although SK1 plays an important role in miR-515-5p cell proliferation regulation, it is not the only gene regulated by miR-515-5p that is responsible for its effects on cancer cell proliferation.

Figure 18: miR-515-5p regulates the expression of genes involved in wnt signaling and cell apoptosis. (A) SylArray enrichment analysis showing that the downregulated genes upon miR-515-5p expression are enriched with miR-515-5p ‘seeds’. (B) Downregulated transcripts predicted to interact with miR-515-5p and involved in wnt signaling and cell apoptosis (C, D, E) FGFR2, PI3K and TCF7L1 mRNA levels are downregulated by miR-515-5p (RNA-Seq validation). MCF7 cells were transfected twice with 50nM of miR-515-5p over 6 days. FGFR2 (C), PI3K (D) and TCF7L1 (E) mRNA levels were then measured by taqman RT-qPCR. presented as the mean of three independent experiments ± SEM and P values were calculated using paired t-test (*:P<0.05).

101

3.2.6 Estradiol downregulates miR-515-5p expression and upregulates SK1 activity

Previous studies have reported that SK1 mediates the estrogen-dependent tumorigenesis of MCF7 cells [256]. However, the molecular processes underlying the upregulation of SK1 expression by estradiol (E2) are still unclear. As miR-515-5p was previously shown be downregulated by E2 [201] and here shown to target SK1, we hypothesised that miR-515-5p downregulation could mediate the upregulation of SK1 expression by estradiol. In order to test this hypothesis, we measured miR-515-5p expression and SK1 activity levels in MCF7 cells treated with 10nM of E2 (Figures 19 and 20). Remarkably, we found that E2 significantly downregulated miR-515-5p levels in a time course experiment (up to 60% reduction at 48 hours, P<0.001; Figure 19A).

We also observed a reduction in miR-515-5p levels after E2 treatment in another

ERα-positive breast cancer cell line (ZR-75-1; Figure 19B). In contrast, miR-515-5p

Figure 19: ERα downregulates miR-515-5p expression in MCF7 and ZR75. After 72 hours of starvation, (A) MCF7 and (B) ZR75 were treated with 10nM estradiol (E2) and ethanol (negative control, where estradiol is dissolved). miR-515-5p expression measured by RT-qPCR in MCF7 and ZR75 treated with either ethanol or E2 after 3, 6, 12, 24 and 48 hours. For each time point, miR-515-5p represents the ratio between the values for E2 and ethanol. and P values were calculated by paired t-test comparing the value of each. Data are presented as the mean of three independent experiments ± SEM time point with the value of the 0 hours time point (*:P<0.05, **:P<0.01, ***:P<0.001).

induced a biphasic upregulation of SK1 activity: an early 4-fold upregulation occurred after 6 hours of E2 treatment (P<0.05), followed by a later 2-fold upregulation after 48

102 hours of E2 treatment (P<0.05) (Figures 20A). A trend towards an increase in SK1 mRNA expression was also observed but the difference was not significant (P>0.05;

Figure 20B).

Since the upregulation of SK1 after 6 hours of E2 treatment (Figure 20A) could not be explained by miR-515-5p mediated regulation, as no significant change in miR-515-

5p expression was observed until 12 hours from the E2 treatment (Figure 19A), we next verified whether SK1 activation by E2, at 48 hours, was caused by the downregulation of miR-515-5p. To do this, we measured SK1 activity levels in MCF7 cells transfected with miR-515-5p precursor (overexpression), together with E2 treatment (Figure 20C). When miR-515-5p is overexpressed, SK1 activity levels are not altered by E2, proving that the increase in SK1 activity after E2 treatment is caused by loss of miR-515-5p (Figure

20C).

Figure 20: SK1 activity upregulation by estradiol is rescued upon miR-515-5p overexpression. (A, B) The effect of estradiol on SK1 mRNA levels and activity. After 72 hours of starvation, MCF7 were treated

103 with 10nM estradiol (E2) and ethanol (negative control). SK1 activity was analysed (A) and SK1 mRNA levels (B) were measured by RT-qPCR in MCF7 treated with either ethanol or E2 after 3, 6, 12, 24 and 48 hours. For each time point, miR-515-5p levels represent the ratio between the values for E2 and ethanol. presented as the mean of three independent experiments ± SEM and P values were calculated by paired t- test comparing the values obtained for each time point compared to 0 hours (*:P<0.05). (C) The combined effect of E2 treatment and miR-515-5p overexpression on SK1 activity. SK1 activity assay was performed in MCF7 cells treated with E2 for 48 hours. Before E2 treatment and whilst under starvation, MCF7 cells were transiently transfected with 50nM of miRNAs. Data are presented as the mean of three independent experiments ± SEM and P values were calculated using paired t-test (*:P<0.05, **:P<0.01).

3.2.7 ERα and SK1 expression are positively correlated

As ERα is an important mediator of E2 action in breast cancer, we next investigated whether ERα might be implicated in the link between E2, miR-515-5p and SK1 by examining the correlation between ERα and SK1 expression. We analysed two publically available datasets, one from the GEO database (mRNA levels) and one from THPA (The

Human Protein Atlas) immunohistochemistry database (protein levels). A significant, but weak, positive correlation was observed between ERα and SK1 mRNA expression in

ERα positive breast cancers (R=0.181; P=0.0233; Figure 21A). In support of this finding, we also observed a significant and stronger positive correlation between ERα and SK1 protein expression in ERα positive cell lines (R=0.352; P=0.0389; Figure 21B). This observation suggest that the two genes might be correlated through miR-515-5p regulation because we previously found that miR-515-5p had a stronger effect inhibiting protein translation, compared to mRNA destabilization in our in vitro assays (it reduces

SK1 mRNA up to ~15% (Figure 13B) and SK1 activity expression to up to ~60% (Figure

13A)). Interestingly, mRNA and protein expression of ERα and SK1 were not significantly correlated in ERα negative cell lines (Figures 21C and 21D, respectively).

104

A C

B D

Figure 21: ERα and SK1 expression are positively correlated in ERα positive cell lines or tissues. (A, B) Correlation between ERα and SK1 mRNA levels. The mRNA analysis was conducted in RNA isolated from 216 breast tumours (ERα-negative, n=82; ERα-positive, n=128) of a public mRNA dataset (GSE22220) using GEO2R. (C, D) Correlation between ERα and SK1 protein levels. The protein analysis was conducted in 57 breast cancer cell lines (ERα-negative: n=31; ERα-positive: n=26) using Tissue Human Protein Atlas (THPA). P values were calculated using Spearman's rank correlation test, corresponding the indicated R to the Spearman coefficient.

3.2.8 SK1 activity downregulation by TAM requires miR-515-5p

As ERα and SK1 are positively correlated, we have then investigated whether ERα might mediate miR-515-5p downregulation shown here to be responsible for the upregulation of SK1 expression by E2 (Figure 20). To do this, we measured miR-515-5p levels and analysed SK1 activity in MCF7 cells treated with tamoxifen (TAM), an antagonist of estrogen receptor α (Figures 22A and 22B, respectively). Notably, miR-

515-5p expression was significantly upregulated by TAM (2.8 fold at 24 hours, P<0.01;

105

Figure 22A), indicating that ERα is most likely involved in the downregulation of miR-

515-5p by E2 treatment. In addition to the upregulation of miR-515-5p, we observed a

significant reduction of SK1 activity in MCF7 cells treated with TAM (P<0.001) (Figure

22B). In contrast, no significant change in SK1 mRNA levels and (P>0.05; Figure 22C)

and activated (phospho-) ERK expression (P>0.05; Figure 22D) was observed.

Interestingly, we did not observe a significant downregulation of SK1 activity in

MCF7 cells that were treated with TAM and anti-miR-515-5p (Figure 22B). The absence

of a response in SK1 activity to TAM in the presence of a miR-515-5p inhibitor suggests

the requirement of this microRNA for the downregulation in SK1 activity caused by the

anti-estrogen TAM (Figure 22B).

A B

Figure 22: Reduction in SK1 activity by TAM requires miR-515-5p. (A) miR-515-5p expression in MCF7 cells treated with TAM. After 72 hours of starvation, MCF7 were treated with 100nM tamoxifen (TAM) and ethanol (negative control). miR-515-5p levels were quantified by RT-qPCR after 3, 6, 12, 24 and 48 hours of treatment. For each time point, miR-515-5p levels represent the ratio between the values for TAM and ethanol treatment. presented as the mean of three independent experiments ± SEM. P values were calculated by paired t-test comparing the values obtained for each time point compared to 0 hours (*:P<0.05,

106

**:P<0.01). (B) SK1 activity levels in MCF7 cells treated with TAM. After 72 hours of starvation and 6 hours before being treated with TAM or ethanol, MCF7 cells were transiently transfected with 50nM of anti-miRNAs (anti-miRs). SK1 activity assay was then performed after 48 hours. presented as the mean of three independent experiments ± SEM. P values were calculated by paired t-test comparing the value obtained for each sample with the value relative to the ethanol and anti-miR NC sample (***:P<0.001). (C, D) SK1 mRNA levels and phospho-ERK 1/2 protein expression in MCF7 cells treated with 100nM TAM and ethanol. After 48 hours, SK1 mRNA levels were measured by RT-qPCR (C) and phospho-ERK 1/2 protein levels were analysed by immuno-blotting (D). presented as the mean of three independent experiments ± SEM and P values were calculated by paired t-test (ns-non-significant: P>0.05).

To further investigate the effect of TAM on miR-515-5p levels, we next performed an

IP of Ago2 in MCF7 cells treated with TAM and extracted total RNA from the IP (Figure

23). RT-qPCR showed that miR-515-5p levels in the RNA bound to Ago2 were significantly higher in the MCF7 cells treated with TAM, compared to the levels measured in the cells treated with ethanol (~4 fold, P<0.05; Figure 23). Therefore, not only were miR-515-5p levels upregulated upon TAM treatment, but miR-515-5p was also recruited to the RISC where Ago2 is known to be responsible for the silencing of sequence-specific transcripts.

Figure 23: miR-515-5p expression in Ago2 immunoprecipitates is upregulated by tamoxifen. After treating MCF7 cells with TAM and ethanol (negative control, NC) for 48 hours, immunoprecipitation (IP) was performed using Ago2 and IgG antibodies. miR-515-5p expression was then analysed by RT-qPCR in the total RNA isolated from each IP fraction. Data represent the mean of two independent experiments ± SEM and P value was calculated using paired t-test (*:P<0.05).

107

3.2.9 miR-515-5p levels are not altered in ERα-negative breast cancer cells after estradiol (E2) and TAM treatment

Next, we analysed the effect of E2 and TAM on miR-515-5p expression and SK1 activity in MDA-MB-231 breast cancer cells which, unlike MCF7 cells, do not express the

ERα. This demonstrated that miR-515-5p and SK1 activity levels were unaltered after both treatments (Figure 24), reinforcing our hypothesis that ERα mediates the downregulation of miR-515-5p by E2.

Figure 24: miR-515-5p and SK1 activity levels in MDA-MB-231 treated with estradiol or tamoxifen. After 72 hours of starvation, MDA-MB-231 were treated with either 10nM estradiol (E2; A,B), 100nM tamoxifen (TAM; C,D) or ethanol (negative control; A-D). (A, C) miR-515-5p expression in MDA-MB-231 cells treated with E2 or TAM. miR-515-5p levels were quantified by RT-qPCR after 24 and 48 hours of treatment. For each time point, miR-515-5p levels represent the ratio between the values for E2/TAM and ethanol treatment. Data presented are the mean of three independent experiments ± SEM. P values were calculated by paired t-test comparing the values obtained for each time point compared to 0 hours (ns-non- significant,P>0.05). (B, D) SK1 activity in MDA-MB-231 cells treated with E2 or TAM. SK1 activity was analysed after 24 and 48 hours of treatment. For each time point, SK1 activity levels represent the ratio between the values for E2/TAM and ethanol treatment. Data presented are the mean of three independent experiments ± SEM. P values were calculated by paired t-test comparing the values obtained for each time point compared to 0 hours (ns-non-significant,P>0.05).

108

3.2.10 ERα mediates miR-515-5p downregulation by estradiol (E2) by directly binding to its promoter region

To determine the mechanism by which ERα regulates miR-515-5p expression levels, we re-analysed published ChIP-seq data from MCF7 cells (sequencing data at

ArrayExpress accession number E-TABM-828 or Gene Expression Omnibus (GEO) accession number GSE25021) [257]. E2 treatment induced 116.77 fold enrichment of

ERα (Bonferroni-corrected P=0.00) close to the transcription start site (TSS) of miR-515-

5p (Figure 25). Accordingly, the level of ERα on the same site was substantially reduced after vehicle treatment (23.78 fold, Bonferroni-corrected P=0.24) (Figure 25), indicating that E2 induces the binding of ERα on the miR-515-5p promoter.

Figure 25: ERα directly binds to the miR-515-5p promoter region. Re-analysis of published ChIP-Seq data in which ERα enrichment was analysed in the chromosome 19 of the MCF7 genome [257].

Therefore, to confirm the importance of ERα in the downregulation of miR-515-5p, we silenced ERα and measured miR-515-5p levels. Consistent with our proposed mechanism, this led to a dose response increase of miR-515-5p levels (Figures 26A and 26B), providing further evidence that the binding of ERα to the miR-515-5p gene

109 promoter downregulates mature miR-515-5p expression. Importantly, E2 treatment induced a significant decrease in the luciferase activity of a reporter construct transfected into MCF-7 cells that contained the miR-515-5p promoter sequence, indicating that the binding of ERα to this region actually results in downregulation of upstream genes (Figure 26C). This finding reinforces our hypothesis that ERα reduces miR-515-5p levels by directly binding to its promoter.

Figure 26: miR-515-5p transcription is repressed by ERα. (A, B) miR-515-5p expression is upregulated by ERα silencing. After (A) 24 hours and (B) 48 hours of ERα siRNA transfection, miR-515-5p levels were quantified by TaqMan RT-qPCR. For each concentration, miR-515-5p levels represent the ratio between the values for ERα siRNA and NC siRNA. Data pre presented are the mean of three independent experiments ± SEM sented are the mean ± SEM and P values were calculated by paired t-test comparing the values obtained for each concentration compared to 0 nM (**:P<0.01, ***:P<0.001). (C) E2 treatment induces miR- 515-5p transcription downregulation. After starving for 3 days, cells were treated with 10nM E2 and transfected with 200ng of miR-515-5p promoter pGl2-basic and 100ng of pRLTK. After 24 hours, both Firefly and Renilla luciferase activities were measured. Data were normalised to the negative control result (miR NC) and presented as the mean of three independent experiments ± SEM. P values were calculated by paired t-test comparing the values obtained for E2 treatment compared to NC (ethanol) treatment (**:P<0.01).

3.2.11 Ex vivo analysis reveals downregulated miR-515-5p levels in ERα- positive breast tumours compared to ERα-negative breast tumours

110

To prove the clinical relevance of our findings, we correlated miR-515-5p expression with the presence of ERα in breast tumours (Figure 27A). In accordance with our in vitro findings (i.e. miR-515-5p downregulation is mediated by E2 via ERα), miR-515-5p levels were found to be significantly lower in ERα-positive breast tumours compared to ERα- negative lesions (P<0.01; Figure 27A). In addition, we used a publicly available dataset

[258] to assess the relationship between miR-515-5p and ERα status. Consistent with our in vitro and ex vivo data, we found that, in a large cohort of breast cancer patients, miR-515-5p expression was significantly lower in ERα-positive compared to ERα- negative tumours, whilst there was no association between miR-515-3p levels and ERα status (Figure 27B).

Figure 27: miR-515-5p levels are upregulated in ERα-negative breast tumours compared to ERα- positive tumours. (A) Ex vivo analysis reveals miR-515-5p levels are downregulated in ERα-positive breast tumours compared to ERα-negative breast tumours. RNA was isolated from 34 breast cancer tumour tissues (ER-negative: n=16, ERα-positive: n=18). miR-515-5p levels were quantified by TaqMan RT-qPCR and analysed according to ERα status. P value was calculated using unpaired t-test (**: P<0.01). (B) miR-515-5p levels are confirmed to be upregulated in ERα-negative breast tumours in an independent dataset. GEO2R analysis of miR-515-5p expression in ERα-negative (n=82) and ERα-positive (n=128) using the dataset GSE22220 [258]. miR-515-3p expression was also analysed as a control. P values were calculated using unpaired t-test (ns-non-significant:P>0.05, ***:P<0.001).

111

3.3 Discussion

In this chapter our aim was to investigate whether miRNAs control the oncogenic role of SK1 in estrogen-dependent breast cancer. Initially, by screening miRNAs that are predicted to regulate SK1, we identified that miR-515-5p directly interacted with SK1

3'UTR and regulated its expression (Figures 12, 13 and 14). We found that miR-515-5p expression was downregulated by estrogen in two different breast cancer cell lines

(MCF7 and ZR75-1) with a concomitant biphasic activation of SK1 (Figures 19 and 20).

Transient transfection with precursor miR-515-5p abrogated the later stimulation of SK1 activity (48 hours) by estrogen (Figure 20). This highlighted miR-515-5p as a potential, previously unidentified link, between estrogen and SK1-mediated oncogenesis.

Conversely, TAM, which antagonises estrogen action, upregulated miR-515-5p expression, and SK1 activity was subsequently reduced (Figure 22). Following silencing of miR-515-5p in ERα-positive MCF7 cells treated with TAM, we did not observe any change in SK1 activity levels (Figure 22), suggesting that reduced levels of miR-515-5p are required for the upregulation of SK1 after estrogen stimulation. Since TAM has been described to have an antagonistic effect on the activation of several estrogen receptors

[259], we hypothesised that ERα mediates the estrogen-induced downregulation of miR-

515-5p. In MDA-MB-231 cells, miR-515-5p expression is not altered upon estrogen or

TAM treatment (Figure 24). The absence of the effect of both treatments in these cells, which express ERβ, but not ERα, indicated that ERα was the mediator of the downregulation and upregulation of miR-515-5p induced by estrogens and TAM respectively [260]. In accordance with our experimental findings, our patient data also demonstrated that miR-515-5p expression is significantly higher in ERα-negative breast cancers, compared to ERα-positive breast cancers (Figure 27). This indicated again the importance of ERα expression in regulating miR-515-5p levels and also the potential clinical relevance of this study.

112

By re-analysing published ChIP-seq data [257], we found that one of the ERα enriched zones in the MCF7 genome (examined after E2 treatment) corresponds to the promoter of C19MC microRNA cluster where miR-515-5p is located, indicating that this particular region is responsible for the miR-515-5p transcription downregulation upom E2 treatment (Figure 25). Therefore, the upregulation of miR-515-5p expression when ERα is silenced (Figure 26), together with our findings in ERα-negative MDA-MB-231 cells

(Figure 24), and the clinical data (Figure 27), points to ERα as the mediator for miR-

515-5p transcriptional repression upon E2 treatment. Since we showed that ERα specifically interacts upstream of the TSS of a cluster of miRNAs, that are probably derived from the processing of a unique primary transcript, then the other miRNAs belonging to this cluster may also participate in the ERα mediated cell proliferation mechanism. However, this would also require further investigation.

Although we demonstrated a positive correlation between ERα and SK1 (Figure 21) which could be mediated by miR-515-5p, the negative effect of miR-515-5p on cell proliferation could not be explained only due to the negative regulation of SK1 by miR-

515-5p. In accordance with this hypothesis, we showed that silencing SK1 only partially rescues the effect of miR-515-5p inhibition on breast cancer cell proliferation (Figure

17). These findings strongly indicate that miR-515-5p also regulates other genes involved in cell proliferation regulation in addition to SK1. RNA-Seq assessment of the transcriptome after miR-515-5p overexpression revealed that genes which positively regulate Wnt signaling and also contain the seed region for miR-515-5p interaction in their 3’UTRs were significantly downregulated (Figure 18). In addition, other important oncogenes that contain a miR-515-5p seed region and were downregulated by this miRNA include FGFR2, which promotes breast tumorigenicity through maintenance of breast tumour-initiating cells [254], and IL6R, which induces STAT-3 activation and the production of the anti-apoptotic proteins Bcl-xL and BCL2 [261].

113

Apart from having a critical role in breast cancer tumour growth and formation, estrogens and the ER have been described to be responsible for breast cancer chemoresistance. Since ERα-positive tumours represent the majority of all breast cancers, understanding the mechanisms involved in ERα-mediated breast cancer chemoresistance is essential to finding novel treatments for resistant-tumours. Cell apoptosis is one of the pathways whose dysregulation has been associated with breast cancer chemoresistance [262]. We elucidated one of the mechanisms by which estrogen downregulates cell apoptosis and induces cell proliferation. By downregulating miR-515-

5p, estrogen upregulates SK1 activity and the expression of many other oncogenes, which then stimulate cell survival by inhibiting apoptosis (Figure 28).

Figure 28: miR-515-5p downregulation by ERα leads to an increase in cell proliferation via SK1. After its activation by E2, ERα negatively regulates miR-515-5p gene transcription. The loss of mature miR-515- 5p leads to a de-repression and increase of SK1 activity, which then promotes cell viability by inhibiting apoptosis.

We conducted the first functional studies to evaluate the role of miR-515-5p in cancer and highlighted its therapeutic potential for the treatment of breast cancer. Re- expression of miR-515-5p was found to dramatically inhibit breast cancer cell

114 proliferation and induce caspase-dependent apoptosis (Figures 15, 16 and 17).

Therefore, in ERα-positive breast cancers, miR-515-5p replacement could stop the stimulation of cell growth by estrogen and may be an adjunct to current anti-estrogen therapies.

115

Chapter 4 miR-515-5p inhibits breast cancer cell migration via MARK4

116

miR-515-5p inhibits breast cancer

cell migration via MARK4

4.1 Overview

In Chapter 3, we identified a new role for miR-515-5p in breast cancer. We demonstrated that miR-515-5p transcription was directly downregulated by the estrogen receptor and that miR-515-5p inhibited breast cell proliferation by inducing caspase- dependent cell apoptosis. During this work, we observed a change in MCF7 and MDA-

MB-231 cell cytoskeletal rearrangement upon miR-515-5p overexpression. As this change in cell morphology might influence cell motility, we started investigating the role of miR-515-5p in breast cancer cell migration. We performed an RNA-seq following the overexpression of miR-515-5p in MCF7 and MDA-MB-231 cells, which revealed the downregulation of five transcripts linked to cell migration: NRAS, FZD4, CDC42BPA,

PIK3C2B and MARK4 [34, 252, 263-265]. Luciferase reporter systems confirmed that miR-515-5p directly regulated NRAS, PIK3C2B and MARK4 expression. As the strongest downregulation of expression by miR-515-5p was seen with MARK4 in both

MCF7 and MDA-MB-231, we decided to focus our study on investigating the direct regulation of MARK4 expression by miR-515-5p. We showed that MARK4 silencing mimics the cell migration phenotype observed for miR-515-5p overexpression in MDA-

MB-231 breast cancer cells. Conversely, transfection of exogenous MARK4 into miR-

515-5p overexpressing cells rescued the observed decrease in cell migration, suggesting that MARK4 downregulation is a crucial mechanism through which miR-515-

5p reduces cell motility. By analysing miR-515-5p expression in tumour samples derived from breast cancer patients, we observed its expression to be inversely correlated with metastasis, providing clinical relevance to our findings. In addition, increased miR-515-

5p or decreased MARK4 expression were indicative of improved survival in metastatic

117 breast cancer. In short, our data demonstrate that miR-515-5p dramatically inhibits cell migration by directly downregulating MARK4 expression and suggests a role for miR-

515-5p as a possible therapeutic target and as a potential biomarker for metastatic breast cancer patients.

118

4.2 Results

4.2.1 miR-515-5p changes MCF7 and MDA-MB-231 cell morphology

Changes in cell morphology were observed within the breast cancer cell lines MCF7 and MDA-MD-231 upon overexpression of miR-515-5p, while studying the role of miR-

515-5p in breast cancer proliferation. By visualising the actin cytoskeleton we confirmed a more rounded phenotype characterised by a loss of lamellipodia protrusions suggesting a loss of cell polarity (Figure 29). Within MDA-MB-231, which commonly display a lack of contact inhibition, overexpression of miR-515-5p resulted in a change in cell to cell adhesions with a more pebble-stone appearance being seen (Figure 29, right hand panel).

Figure 29: miR-515-5p changes MCF7 and MDA-MB-231 cell morphology. Actin (green) and cell nucleus (blue) were stained using Phalloidin and DAPI, respectively. Images presented are representative of images of three independent experiments.

119

4.2.2 miR-515-5p inhibits breast cancer cell migration

Since regulation of the cell cytoskeleton and cell polarity is crucial for cell motility, we wished to verify whether this change in morphology affects cancer cell migration.

However, we only analysed the effect of miR-515-5p in MDA-MB-231 cell migration because MCF7 cells are known to display low motility. We performed random and

Boyden chamber directional cell migration assays in MDA-MB-231 cells upon miR-515-

5p overexpression and observed a decrease of 48% and 68% in random and directional cell migration assays, respectively, in miR-515-5p-transfected MDA-MB-231 in comparison to miR NC-transfected cells (P<0.001; Figure 30).

Figure 30: miR-515-5p inhibits MDA-MB-231 cell migration. (A) miR-515-5p inhibits MDA-MB-231 random migration. MDA-MB-231 were transfected with miR-515-5p for 48 h before time-lapse imaging was performed for 18 h (1 image/10 min). Plots show overlays of representative trajectories travelled by miR NC and miR-expressing MDA-MB-231 cells during time-lapse motility assays. The distance of migration was quantified by manual tracking using the MetaXpress 2.0 software. Data presented are the mean of three independent experiments ± SEM. P values were calculated by paired t-test comparing individual miRNA values with miR NC values (***:P<0.001). (B) miR-515-5p inhibits MDA-MB-231 directional migration. MDA- MB-231 were transfected with mirR NC or miR-515-5p for 48 h before transwell migration assays were performed for 9 h. The graph on the right indicates cell migration expressed as a percentage of the average number migratory cells per field (n=5 fields per transfection). Data presented are the mean of three independent experiments ± SEM. P values were calculated by paired t-test comparing individual miRNA values with miR NC values (***:P<0.001).

120

4.2.3 miR-515-5p directly regulates NRAS, MARK4 and PIK3C2B expression

In order to identify miR-515-5p’s direct targets responsible for its effect on breast cancer cell morphology and migration, we performed a RNA-seq analysis in MCF7 and

MDA-MB-231 overexpressing miR-515-5p. Interestingly, we found 5 downregulated transcripts which are predicted to interact with miR-515-5p and which have also been implicated in cell migration: NRAS, FZD4, CDC42BPA, PIK3C2B and MARK4 (Figure

31) [34, 252, 263-265].

Figure 31: RNA-seq analysis of MCF7 and MDA-MB-231 treated with miR-515-5p. RNA-seq of MCF7 and MDA-MB-231 transfected with miR-515 revealed the downregulation of five transcripts, NRAS, FZD4, CDC42BPA, PIK3C2B and MARK4, which have been implicated in the control of cell migration.

To validate the RNA-seq results, we analysed the levels of these five transcripts in

MCF7 and MDA-MB-231 upon the overexpression of miR-515-5p (Figures 32A and

32B) and miR-515-5p sponge vectors (Figures 32C and 32D), which reduce the levels of miR-515-5p by directly interacting with its mature form. We observed a significant downregulation of the expression of all five transcripts in the miR-515-5p-transfected

MCF7 (Figure 32A) but only a decrease in N-RAS, PI3KC2B and MARK4 mRNA levels

121 in miR-515-5p-transfected MDA-MB-231 (Figure 32B). Among the tested transcripts, the higher reduction was observed in MARK4 m RNA levels, which have reduced by more than 95% in both MCF7 and MDA-MB-231 cells upon miR-515-5p overexpression

(P<0.001; Figures 32A and 32B). In opposition, miR-515-5p sponge vectors induced an increase in the levels of the 5 transcripts in MCF7 (Figure 32C) but not within MDA-MB-

231 (Figure 32D). The absence of miR-515-5p sponges’ effect in MDA-MB-231 might be related to the lower expression of miR-515-5p expression in MDA-MB-231 compared to

MCF7 (Figure 32E).

Figure 32: NRAS, FZD4, CDC42BPA, PIK3C2B and MARK4 mRNA levels in MCF7 and MDA-MB-231 upon miRs or sponges overexpression. (A, B) The effect of miR-515-5p in NRAS, FZD4, CDC42BPA, PIK3C2B and MARK4 mRNA levels in MCF7 (A) and MDA-MB-231 (B). Cells were transfected with miR NC or miR-515-5p 48 h before NRAS, FZD4, CDC42BPA, PIK3C2B and MARK4 mRNA levels were quantified by SYBR Green qRT-PCR. Data is displayed as the mean of three independent experiments ± SEM. P values were calculated by paired t-test comparing individual miRNA values with miR NC values (*:P<0.05,

122

**:P<0.01, ***:P<0.001). (C, D) The effect of miR-515-5p sponges in NRAS, FZD4, CDC42BPA, PIK3C2B and MARK4 mRNA levels in MCF7 (C) and MDA-MB-231 (D). The miR-515-5p sponges were constructed by annealing, purifying and cloning oligonucleotides containing six tandem bulged miRNA binding motifs, into the HindIII and BamHI sites of the pEGFP-C1 plasmid (Contech, Saint-Germain-en-Laye, France). After 48 h of miR-515-5p sponges transfection, NRAS, FZD4, CDC42BPA, PIK3C2B and MARK4 mRNA levels were quantified by Syber Green qRT-PCR. Data were normalised to U6 snRNA values and are presented as the mean of three independent experiments ± SEM. P values were calculated by paired t-test comparing individual miR-515-5p sponges values with NC sponges values (**:P<0.01, ***:P<0.001). (E) miR-515-5p expression in MCF7 and MDA-MB-231. miR-515 levels were quantified by Taqman qRT-PCR. Data were normalised to U6 snRNA values and are presented as the mean of three independent experiments ± SEM.

We then wished to assess whether the transcripts for N-RAS, MARK4 and PI3KC2B

could be directly regulated by the binding of miR-515-5p to their 3’UTRs. We used a B reporter system where luciferase expression was under the control of the 3’UTRs of our

proposed targets. We showed that miR-515-5p negatively regulated the levels of

luciferase reporter expression in both MCF7 and MDA231 cells by directly interacting

with the 3’UTR regions of N-RAS, MARK4 and PI3KC2B (P<0.001; Figures 33A and

33B). This demonstrated that the mRNAs for these three proteins were true direct

targets of miR-515-5p.

Figure 33: miR-515-5p directly interacts with NRAS, MARK4 and PI3KC2B’s 3’UTR. Relative luciferase activity levels were measured 24 h after co-transfection of MCF-7 (A) and MDA-MB-231 (B) with 3’UTR- luciferase reporter constructs and either with miR-515-5p or miR NC. Data shown are mean of three independent experiments ± SEM. P values were calculated by paired t-test comparing each value with miR NC values (**:P<0.01, ***:P<0.001).

123

4.2.4 Silencing of MARK4 mimics the effect of miR-515-5p on breast cancer migration

As it was observed a strong downregulation of MARK4 expression by miR-515-5p in both MCF7 and MDA-MB-231, we next verified whether silencing MARK4 would mimic the effect of miR-515-5p on cell migration by performing random migration and Boyden chamber directional migration assays (Figures 34). As previously mentioned, since

MCF7 are known to display poor motility, we did not analyse the effect of either miR-515-

5p expression or MARK4 silencing on MCF7 cell migration [266]. In accordance with our previous evidence, we observed a reduction of 48% and 61% in random and directional

MDA-MB-231 cell migration upon MARK4 knockdown, respectively (P<0.001; Figures

34).

Figure 34: MARK4 knockdown inhibits MDA-MB-241 cell migration. (A, C) MARK4 silencing reduces MDA-MB-231 random migration. Cells were transfected either with siMARK4 or siNC for 48 h before time- lapse imaging was performed for 18 h (1 image/10 min). Plots show overlays of representative trajectories described by siRNA silenced cells during time-lapse motility assays (A). The degree of migration was quantified by manual tracking using the MetaXpress 2.0 software. Data are presented as the mean of three independent experiments ± SEM (C). P values were calculated by paired t-test comparing individual siRNA values with siRNA NC values (***:P<0.001). (B, D) Directional migration of MDA-MB-231 cells is inhibited upon MARK4 silencing. Cells were transfected either with siMARK4 or siNC for 48 h before transwell migration assays of transfected cells were performed for 9 h. Figures are representative fields of siRNA transfected cells (B). Graphs indicate cell migration expressed as a percentage of the average of migratory cells per field (n=5 fields per transfection, D). Data are mean of three experiments ± SEM. P values were calculated by paired t-test comparing individual siRNA values with siRNA NC values (D, ***:P<0.001).

124

4.2.5 miR-515-5p inhibits cell migration through MARK4 downregulation in breast cancer

To clarify the importance of MARK4 in the cell migration inhibition by miR-515-5p, we transfected MDA-MB-231 with miR-515-5p and, after 24 h, we overexpressed

MARK4 in the transfected cells. We then analysed whether MARK4 overexpression was able to rescue the effect of miR-515-5p on MDA-MB-231 random and directional migration (Figure 35). There was no significant difference between the distance travelled by MDA-MB-231 co-overexpressed with miR-515-5p and MARK4 and the distance travelled by the same cell lines co-transfected with miR NC and NC plasmid (P > 0.05;

Figures 35A and 35B), confirming that MARK4 overexpression rescues the miR-515-5p phenotype. However, overexpression of MARK4 only results in an 80% rescue of directional migration inhibition by miR-515-5p in MDA-MB-231 (P<0.01; Figures 35C and 35D).

Figure 35: miR-515-5p inhibits breast cancer cell migration by downregulating MARK4. (A, B) MARK4 overexpression rescues MDA-MB-231 random migration inhibition by miR-515-5p. Cells were transfected with miR-515-5p / miR NC. 24 h later cells were again transfected with a control (pNC) or MARK4 plasmid (pMARK4). After 24 h, time-lapse imaging was performed for 18 h (1 image/10 min). Plots show overlays of representative trajectories travelled by MDA-MB-231 cells over 18 h (E). Random migration was quantified by manual tracking using the MetaXpress 2.0 software. Data are presented as the mean of three independent experiments ± SEM. P values were calculated by paired t-test comparing individual values with miR NC/pNC values (**: P<0.01, ***:P<0.001). (C, D) MARK4 overexpression partially rescues MDA-MB-231 directional migration inhibition by miR-515-5p. Cells were transfected with miR-515-5p / miR NC. 24 h later cells were again transfected with a control (pNC) or MARK4 plasmid (pMARK4). The following day, transwell migration assays performed for 9 h. Figures are representative fields of migratory MDA-MB-231 (G). Graphs indicate cell migration expressed as a percentage of the average of migratory cells per field (n=5 fields per

125 transfection). Data are presented as the mean of three experiments ± SEM. P values were calculated by paired t-test comparing individual values with miR NC/pNC values (H; **:P<0.01, ***:P<0.001).

4.2.6 miR-515-5p levels are inversely correlated with metastasis in breast cancer

In order to evaluate the clinical relevance of our findings, we quantified miR-515-5p expression in metastasis in vivo and in breast cancer patients (Figures 36 and 37). The levels of miR-515-5p were quantified in MDA-MB-231 cells that had been inoculated into the mammary fat pads of nude mice and allowed to form metastatic deposits [267]. We observed that miR-515-5p levels were upregulated in the primary tumour cells at the inoculated sites, compared with metastatic cells at distant sites: lungs, bones, brain and adrenal glands (brain - P<0.05; adrenal glands - P<0.01; lung and bone - P<0.001;

Figure 36A). As expected from our in vitro findings, MARK4 mRNA levels were significantly lower in all the cells isolated from the metastatic sites compared to the primary tumour cells (lung and brain - P<0.05; adrenal glands - P<0.01; bone - P<0.001;

Figure 36B). Interestingly, NRAS expression did not correlate with metastasis (P>0.05;

Figure 36C) and PI3KC2B expression was only significantly upregulated in the metastatic cells isolated from brain and adrenal glands in comparison to the cells from the primary tumour (P<0.01; Figure 36D).

126

Figure 36: miR-515-5p and MARK4 expression of primary tumour and metastatic cells in vivo. MDA- MB-231 cells were inoculated into the mammary fat pads of nude mice and allowed to form metastatic deposits [267]. The expression of miR-515-5p (A), MARK4 (B), NRAS (C) and PI3KC2B (D) in both primary tumour and metastatic cells were analysed. miR-515-5p levels were quantified by TaqMan qRT-PCR and MARK4 mRNA levels were measured by Syber Green qRT-PCR. Data were normalised to u6 snRNA values and graph displays the mean of three independent experiments ± SEM. P values were calculated by paired t-test (*:P<0.05, **:P<0.01, ***:P<0.001).

Subsequently, we measured miR-515-5p levels in paired primary and metastatic breast cancer patient samples. We observed a higher mean expression of miR-515-5p in primary tumours compared with their corresponding lymph-node metastases (Figure

37), either by comparing the absolute values (P<0.001; Figure 37A) or the paired values for each patient (P<0.01; Figure 37B).

127

Figure 37: miR-515-5p expression of primary tumour and metastatic cells of breast cancer patients. RNA was isolated from 18 breast cancer tumour and metastatic tissues. miR-515-5p levels were quantified by TaqMan qRT-PCR and analysed either by comparing the absolute values (A) or the paired values for each patient (B). Data were normalised to u6 snRNA values and the mean of three independent experiments ± SEM. P values were calculated by unpaired and paired t-test, respectively (**:P<0.01, ***:P<0.001).

4.2.7 miR-515-5p and MARK4 as prognostic markers for metastatic cancer patients

To further analyse the clinical implications of our work, we evaluated the use of miR-

515-5p and MARK4 as prognostic markers for metastatic breast cancer (Figure 38).

Since we have previously proved that miR-515-5p expression is correlated with ERα expression and used now MDA-MB-231, an ER-negative breast cancer cell line, as our in vitro model, we decided to focus our study in analysing the relevance of miR-515-5p

128 and MARK4 in the prognosis of ER-negative breast cancer patients. We analysed the overall survival time of ER-negative metastatic breast cancer patients according to the levels of miR-515-5p and MARK4 mRNA.

According to miR-515-5p’s suppressive role in cancer cell migration, miR-515-5p expression is positively correlated with survival times in metastatic breast cancer patients. We demonstrate here that ER-negative metastatic breast cancer patients with lower levels of miR-515-5p had significantly reduced distant relapse-free survival times, compared to the patients with higher levels of miR-515-5p (Figure 38A, P=0.0002). In contrast, though not significant, survival time is reduced in ER-negative metastatic breast cancer with high levels of MARK4 mRNA compared to the patients with low levels of

MARK4 mRNA (Figures 38B, P=0.13).

A B

Figure 38: MARK4 and miR-515-5p as prognostic molecular markers in ER-negative metastatic breast cancer. (A) Overall survival of ER-negative metastatic breast cancer patients according to the levels of MARK4 mRNA. Analysis of MARK4 mRNA expression and overall survival of ERα-negative lymph node- positive breast cancer patients, n=226, using the Kaplan Meier-plotter [268]. The two patient cohorts, high and low MARK4 expression, are compared by a Kaplan-Meier survival plot. (B) Distant relapse-free survival of ER-negative metastatic breast cancer patients according to the levels of miR-515-5p. GEO2R analysis of miR-515-5p expression and distant relapse free survival of ERα-negative lymph node-positive breast cancer patients, n=18, using the dataset GSE22216 [258]. Survival values were divided in two cohorts by a Kaplan- Meier plot using GraphPad Prism: high miR-515-5p expression (> 8.20) and low miR-515-5p expression (< 8.20). (A, B) The t hazard ratio with 95% confidence intervals and logrank P value are calculated and presented.

129

4.3 Discussion

Having previously demonstrated the mechanism underlying the effect of miR-515-5p on ER-positive breast cancer cell proliferation (Chapter 3), we have here investigated the role of miR-515-5p on cell migration (Chapter 4). We initiated this study because we observed a cell cytoskeletal rearrangement in MCF7 and MDA-MB-231 cells that overexpressed miR-515-5p. Both MCF7 and MDA-MB-231 cells appeared more rounded which suggested a loss of polarity more characteristic of non-migratory cells (Figure 29).

Although the increased roundness in MCF7 cells upon overexpression of this microRNA could be explained by miR-515-5p-induced apoptosis, the cell morphology change in

MDA-MB-231 (in which apoptosis is not induced by miR-515-5p, [269]) indicated that the effects of miR-515-5p on apoptosis and morphology may have been caused by different targets of miR-515-5p and consequently through different signalling pathways.

Additionally, because MDA-MB-231 is an ER-negative breast cancer cell line, we propose that the estrogen receptor α is not involved in the miR-515-5p-induced change in cell morphology.

Altered cytoskeletal dynamics and cell polarity are known to influence cell motility; for this reason, we then investigated the effect of miR-515-5p on cell migration. As expected from the morphological changes (Figure 29), miR-515-5p inhibited random migration, basic MDA-MB-231 cell migration movements, and directional cancer cell migration, MDA-MB-231 cell motility towards an environmental stimulus, in this study from starved to complete media (Figure 30). In order to ascertain the mechanism through which miR-515-5p altered cell morphology and migration and to identify its gene targets, we performed an RNA-seq analysis using RNA obtained from miR-515-5p- transfected MCF7 and MDA-MB-231 cells. We identified 5 transcripts downregulated upon miR-515-5p overexpression: NRAS, FZD4, CDC42BPA, PIK3C2B and MARK4

130

(Figures 31 and 32). These transcripts have been described as being involved in cell migration and are predicted by TargetScan to be directly targeted by miR-515-5p [34,

252, 263-265]. Furthermore, we validated this using luciferase report assays that demonstrated that miR-515-5p directly interacts with the 3’UTR of NRAS, PIK3C2B and

MARK4 (Figure 33). Interestingly, MARK4 levels were reduced by more than 90% upon miR-515-5p overexpression, in contrast to the lesser reduction (10-50%) that microRNAs more commonly cause in their direct targets’ mRNA levels. Since the luciferase reporter system demonstrated that miR-515-5p only reduced 18% of the luciferase activity

(Figure 33), we believe that the direct interaction between miR-515-5p and the 3’UTR region of MARK4 mRNA levels might not be the only cause of the dramatic effect of miR-

515-5p on MARK4 expression.

As the most dramatic decrease in expression was seen in MARK4 mRNA levels (90-

95%) upon miR-515-5p overexpression (Figures 32), we decided to focus this study on investigating whether MARK4 downregulation was the main target responsible for the effect of miR-515-5p on cell migration. MARK4, initially termed MARKL1, has been demonstrated to localise to the centrosomes, midbody and nucleus [32, 33] and to interact with tubulin, myosin and actin [33, 34]). More recently, MARK4 was found to interact with 14-3-3 proteins, ARHGEF2 and Protein Phosphatase 2A (PP2A), which are known to control the fundamental processes of cell migration [36, 37]. Although together these evidences suggest a potential role for MARK4 in the regulation of cell motility, no studies have yet proved a direct link between MARK4 and cell migration. Here, we present the first clear evidence of the role of MARK4 in breast cancer cell motility by demonstrating that both random and directional cell migration are reduced in MDA-MB-

231 cells upon MARK4 silencing (Figure 34). Additionally, because MARK4 silencing mimics the random and directional migration inhibition by miR-515-5p in MDA-MB-231

(Figure 34), we propose that the effect of miR-515-5p on cell migration is mainly via

MARK4.

131

Testing our hypothesis that MARK4 downregulation is the main cause of miR-515-

5p’s effect on breast cancer cell migration, we verified whether overexpressing MARK4 in miR-515-5p–transfected cells could rescue the reduced cell motility (Figure 35). In accordance with our hypothesis, MARK4 overexpression rescued the effect of miR-515-

5p on MDA-MB-231 random cell migration (Figures 35A and 35B). However, the increased MARK4 expression did not totally rescue the effect of miR-515-5p overexpression on MDA-MB-231 directional migration (Figures 35C and 35D). Although there was apparently a total rescue in the case of the random cell migration, we should note that small differences in random cell migration assays may not be as visible as in directional migration assays due to the sample size. The directional migration assay is based on the analysis of thousands of cells (counting) and the second is based on the tracking of only 90 cells per condition. For this reason, we believe that, apart from

MARK4, other miR-515-5p direct targets may contribute towards the effect of this microRNA in cell migration. However, since the rescue of both types of migration is almost total (80-100%), we do prove here that MARK4 downregulation is the main mediator of breast cancer cell migration inhibition by miR-515-5p.

To assess the clinical relevance of our in vitro findings, we analysed miR-515-5p expression in primary tumour and metastatic tissues in vivo and in breast cancer patients

(Figures 36 and 37). We first quantified miR-515-5p levels in MDA-MB-231 cells that had been inoculated into the mammary fat pads of nude mice and allowed to form metastatic deposits. We observed a significant downregulation of miR-515-5p expression in metastatic MDA-MB-231 compared to MDA-MB-231 isolated from the primary tumour (Figure 36A). Interestingly, we confirmed the opposite correlation for

MARK4 mRNA levels (Figure 36B) but not for NRAS and PI3KC2B mRNA levels

(Figures 36C and 36D), suggesting again the importance of MARK4 as the main miR-

515-5p target involved in the role of this microRNA in breast cancer cell migration. We then quantified miR-515-5p levels in tissues of breast cancer patients. Consistent with

132 the mice experiments, miR-515-5p expression was significantly reduced in lymph nodes compared to primary tumours derived from patients (Figure 37), indicating the reduction of miR-515-5p levels as one of possible causes of the increased cell motility of the metastatic breast cancer cells. Additionally, we proved that miR-515-5p expression is indicative of the outcome of ER-negative metastatic breast cancer patients, the survival time of patients with lower levels of miR-515-5p is significantly reduced compared to the patients with higher levels of miR-515-5p (Figure 38).

In summary, we demonstrate here that miR-515-5p overexpression inhibits cell migration by downregulating MARK4 mRNA levels in breast cancer (Figure 39). As we observed a significant lower expression of miR-515-5p in metastatic breast cancer tissues versus the respective tumour tissues (Figure 37), a miR-515-5p-based therapy might treat metastasis or prevent the initiation of the metastasis process in breast cancer patients by inhibiting the movement of migratory cancer cells from the breast tumour to the surrounding tissues and then to other organs. Apart from the possibility of a future therapeutic application for miR-515-5p, our evidence that the lower levels of miR-515-5p expression are indicative of a poor outcome of ER-negative metastatic breast cancer highlights the potential use of miR-515-5p as a prognostic molecular biomarker for metastatic cancer patients with ER-negative breast cancer tumours.

133

Figure 39: miR-515-5p downregulation leads to an increase in cell motility via MARK4. Some cells in the primary tumour invade the surrounding tissues and then enter in blood and lymphatic circulation. These cells have a lower expression of miR-515-5p promote cell motility by downregulating MARK4 expression.

134

Chapter 5 Discussion and conclusions

135

Discussion and conclusions

5.1 miR-515-5p as a member of the C19MC miRNA cluster, the largest miRNA cluster discovered in the human genome so far

miR-515-5p is a context-dependent C19MC microRNA

Some microRNAs precursors reside alone in the genome, while others are located in clusters with other microRNAs working in an operon-like genomic structure [106, 270,

271]. The C19MC cluster is the largest miRNA cluster identified in the human genome so far, it harbors 46 pre-miRNA genes, including the genomic precursor of miR-515-5p

[272] (Figure 40). The function of all C19MC microRNAs are still not fully understood but the very short time of development and conservation of this cluster during primate evolution suggest the importance of the microRNAs residing in this genomic region

[273]. Accordingly, it has been shown that some of these microRNAs play an important role in cancer. However, they do not all have identical roles and expression profiles: some C19MC microRNAs act as tumour suppressors and are downregulated in cancer

(e.g. miR-520h in pancreatic cancer [274]), while other C19MC microRNAs have an oncogenic function and are upregulated in cancer (e.g. miR-517a in gastric cancer,

[275]). Interestingly, some of the microRNAs located in this cluster have even been found to have opposite roles when present in different cancer tissues. This is the case of miR-515-5p which was already shown to have an oncogenic role in hepatocellular carcinoma [276] and here has been demonstrated to have a tumour suppressive function in breast cancer. These microRNAs have been classified as context-dependent microRNAs and it is believed that their contradictory role within different molecular contexts might be explained by the existence of different transcriptomes, which results in

136 the presence of different targets and, consequently, in very specific and distinct combinatorial effects [210].

Figure 40: Scheme of the C19MC microRNA cluster. All microRNA genes located in the C19MC cluster are represented in red and the CpG island is represented in green. This scheme was taken from the UCSC Genome Browser on Human Feb. 2009 (GRCh37/hg19) Assembly.

The expression of miR-515-5p’s cluster neighbours is altered upon E2 stimulation

Although in this study we have only quantified the levels of miR-515-5p in MCF7 cells upon E2 stimulation, Yee-Ming Lee and his team have recently analysed the changes in the expression of other C19MC microRNAs in the same condition (MCF7 treated with E2) by performing a qPCR-based TLDA assay [201]. The authors demonstrated that the E2 treatment induces the upregulation of miR-524, miR-520b, miR-520c and miR-513-3p and the downregulation of miR-515-5p, miR-526b and miR-

518f and miR-518e [201]. The reported downregulation of miR-515-5p upon E2 stimulation is consistent with our evidences (Figure 19) but we did not expect an upregulation of some of the C19MC microRNAs (miR-524, miR-520b, miR-520c and miR-513-3p) by E2. As we showed here evidence indicates that E2 downregulates miR-

515-5p through the direct interaction of ERα with the C19MC promoter (Figures 25 and

26) and since it is believed that cluster miRNAs are initially transcribed into a long non- coding transcript [277], we presume that the transcription of all C19MC microRNAs might

137 be downregulated by E2. Therefore, we propose that the upregulation of miR-524, miR-

520b, miR-520c and miR-513-3p expression might be a result of a post-transcriptional mechanism which compensates the reduced levels of the primary form of these microRNAs. In fact, it has been recently found that hormones control the microRNAs maturation [278, 279] and that this process can be specific for certain subsets of microRNAs [280].

The C19MC cluster harbours microRNAs that are derived from either the 3’ arm or the 5’ arm of their respective duplex hairpin precursors [281], and interestingly, the

C19MC microRNAs that are upregulated by E2 are originated from their precursors’ 3’ arm, while the C19MC microRNAs that are downregulated by E2 are generated from their precursors’ 5’ arm (Figure 41). Therefore, the contradictory expression profile observed for these C19MC microRNAs upon E2 stimulation might be explained by a deregulation of the expression of RNA-binding proteins specifically involved in the maturation process of microRNAs originated from the 3’ or 5’ arm of their hairpin precursors. As miR-515-5p is among the microRNAs maturated from the 5’ arm of their precursors, the observed downregulation of miR-515-5p in cells treated with E2 might be simultaneously caused by the repressive effect of ERα on miR-515-5p’s transcription and a change in miR-515-5p’s maturation process.

Figure 41: Maturation of some microRNAs located in the C19MC cluster. Some microRNAs are matured from the 3’ arm of their precursors, while others are originated from their precursors’ 5’ arm [281]

138

5.2 The role of miR-515-5p in ER-positive breast cancer tumorigenesis

miR-515-5p downregulation is responsible for the positive correlation between E2 and SK1

SK1 has been demonstrated to mediate estrogen-dependent tumorigenesis [96] and, despite estrogens having been described to contribute to a higher activation of S1P receptors by stimulating the sphingosine 1-phosphate export [208], the mechanism underlying the upregulation of SK1 activity by E2 has been unclear. We have here revealed that miR-515-5p directly targets SK1 and proved that miR-515-5p downregulation was the main factor responsible for the positive correlation between E2 and SK1 activity (Figure 42) by demonstrating that SK1 activity upregulation by estradiol is rescued upon miR-515-5p overexpression (Figure 21).

miR-515-5p inhibits breast cancer cell proliferation by inducing caspase- dependent cell apoptosis

SK1 has been shown to rescue cells from apoptosis by unbalancing the sphingolipid rheostat towards cell survival and proliferation. SK1 converts the pro-apoptotic sphingosine into the pro-survival sphingosine 1-phosphate [58] which inhibits the activation of executioner caspases by binding to S1P receptors and activating the PLC,

ERK and Akt signaling pathways [63-65]. In this study, miR-515-5p was proved to directly downregulate SK1 expression, and in accordance with the pro-survival function of SK1, miR-515-5p was here shown to inhibit breast cancer cell proliferation by inducing caspase-dependent cell apoptosis (Figures 16 and 17). However, we have also

139 demonstrated that SK1 silencing was only able to partially rescue the caspase-induced cell apoptosis promoted by miR-515-5p silencing (Figure 17), indicating that other targets are involved in the miR-515-5p’s repressive effect on breast cancer cell proliferation and survival. Confirming this hypothesis, a RNA-seq analysis of cells overexpressing miR-515-5p revealed that miR-515-5p targets other genes, apart from

SK1, that are known to promote cell survival and proliferation (Figure 18), namely

FGFR2, which promotes breast tumorigenicity through maintenance of breast tumour- initiating cells [254], and IL6R, which induces STAT-3 activation and the production of the anti-apoptotic proteins bcl-xl and BCL2 [261].

Figure 42: Proposed mechanism underlying the effect of E2 on cell apoptosis via SK1. Estradiol (E2) activates estrogen receptor α (ERα) by inducing its dimerisation. Upon activation, ERα is translocated to the nucleus where it represses miR-515-5p transcription by directly interacting with its promoter region. The reduced expression of miR-515-5p leads to the upregulation of SK1 expression. The higher levels of SK1 increase the conversion of sphingosine (Sph) into sphingosine 1-phosphate (Sph-1P). S1P is then exported through ABC transporters and outside the cell S1P binds to five cognate G-coupled receptors named S1P receptors. These receptors induce the activation of pathways that lead to the repression of caspase- dependent cell apoptosis. At the same time, E2 stimulates S1P export though a different mechanism [208].

140

5.3 The role of miR-515-5p in cell motility in breast cancer

miR-515-5p inhibits breast cancer cell migration by targeting MARK4

We have here revealed that miR-515-5p inhibits cell migration in breast cancer and, though miR-515-5p was demonstrated to target a number of genes linked to cancer cell motility (Figure 31), the inhibitory effect of miR-515-5p on cell random and directional movements was proved to be mainly via MARK4 targeting (Figure 35). MARK4 belongs to the family of MARK proteins which are implicated in the regulation of the cell microtubule dynamics by phosphorylating microtubule-associated proteins (MAPs) [30].

MARK4 was found to co-localise with microtubules [32, 33] and to directly interact with a number of proteins involved in cell cystoskeletal rearrangement [36, 37]. Among the identified binding partners of MARK4 is ARHGEF2, a Rho-dependent GTPase which has been found to coordinate multiple RhoA-dependent signaling pathways during the cell migration, namely the actin-myosin-based contractility and the focal adhesions turnover

[40-42]. It is therefore conceivable that the inhibitory effect of miR-515-5p on cell motility might be at least partially mediated by the reduction of the interaction of MARK4 with

ARHGEF2 upon the downregulation of MARK4 levels by miR-515-5p (Figure 43A).

MARK4 mRNA contains miR-515-5p’s putative binding sites in both 3’UTR and 5’UTR regions

In this study we have shown that miR-515-5p downregulates MARK4 mRNA levels by more than 90% (Figure 32), a pronounced reduction of a degree that is not commonly seen to be directly caused by microRNAs [282]. MARK4 was here proved to be a miR-515-5p’s direct target but miR-515-5p was only able to reduce the luciferase

141 activity by 18% in cells transfected with a reporter system containing MARK4’s 3’UTR

(Figure 33). This indicates that the downregulation of MARK4 expression observed upon miR-515-5p overexpression might not be only induced by the binding of miR-515-5p to the 3’UTR region of MARK4 mRNA. Thus, we suggest that this pronounced reduction of

MARK4 transcript by miR-515-5p might be also caused by the downregulation of the expression of other miR-515-5p’s direct targets involved in the regulation of MARK4 transcription and mRNA stability or/and by the interaction between miR-515-5p and other regions, that not only the 3’UTR, of MARK4 mRNA.

Initially, it was believed that microRNAs exclusively downregulate gene expression by interacting with the 3’UTR of their targets’ mRNA. However, it has been recently proved that the large-scale change in the expression of some microRNA’s targets can be caused by the binding of these microRNAs to both 3’UTR and 5’UTR regions of their targets’ transcripts [283]. Interestingly, we found that the 5’UTR region of MARK4 mRNA contains putative binding site for miR-515-5p and that the interaction between miR-515-

5p and this site is predicted to be energetically favourable (Figure 43B), suggesting that the pronounced MARK4 downregulation by miR-515-5p might be induced by its direct interaction with both 3’UTR and 5’UTR’ regions of MARK4’s transcript.

microtubule

↓contractility miR-515-5p

5’UTR 3’UTR B MARK4 mRNA

↓ ARHGEF2 MARK4 ↓ MARK4 ↓MARK4 mRNA

↓cell migration

↓focal adhesions turnover

Figure 43: Proposed mechanism underlying the inhibitory effect of miR-515-5p on cell migration. (A) miR-515-5p downregulates MARK4 expression by directly interacting with both 3’UTR and 5’UTR regions of

142

MARK4 transcripts. The reduced expression of MARK4 leads to a reduction of its interaction with ARHGEF2, a member of the Rho GTPase family known to promote cell migration by inducing cell asymmetry and polarization. (B) The interaction between miR-515-5p and MARK4’s 5’UTR region. The minimum free energy of the interaction between miR-515-5p and MARK4’s 5’UTR region was calculated using RNAhybrid, a publically available tool that predicts the minimum free energy of the hybridization between microRNAs and the mRNA of possible targets (details in appendice) [284].

5.4 Conclusions and future directions

ER-positive breast cancers, which account for about 75% of all breast cancer tumours, have been successfully treated with hormonal therapies that block estrogens production or antagonise its action [44]. However, in 40% of the cases, these tumours become resistant and stop responding to the treatment [285]. In ER-positive breast cancer tumours, SK1 has been reported to mediate estrogen-dependent tumorigenesis

[96] and to play an important role in hormonal chemoresistance [207]. Thus, as a further understanding of mechanisms involved in breast cancer hormonal chemoresistance is essential to finding novel treatments for resistant-tumours, we have here investigated the molecular signalling pathway underlying the link between estrogens and SK1, which has been unclear to date. We proved that miR-515-5p directly targets SK1 and showed that the miR-515-5p downregulation is the main factor responsible for the SK1 activity upregulation by E2 (Chapter 3). Furthermore, we have here revealed a new role for miR-515-5p in cancer by proving that miR-515-5p inhibits breast cancer cell proliferation and migration. miR-515-5p was shown to induce caspase-dependent cell apoptosis partially by targeting SK1 (Chapter 3) and to inhibit breast cancer cell motility mainly by targeting MARK4 (Chapter 4).

In this work we present evidences that improve our understanding of miR-515-5p’s role in breast cancer. However, there are still few aspects of miR-515-5p’s mode of action that need to be addressed. In order to fully understand the tumour suppressive

143 role of miR-515-5p in ER-positive breast cancer, it will still be required to evaluate the importance of miR-515-5p in breast cancer chemo-resistance by analysing the effect of miR-515-5p on the response of chemo-resistant breast cancer cells to estrogens.

Additionally, to improve our comprehension of the molecular basis behind the inhibitory effect of miR-515-5p in breast cancer cell migration, it will still be necessary to validate our hypothesis that MARK4 downregulation by miR-515-5p is caused by the simultaneous binding of miR-515-5p to both 3’UTR and 5’UTR regions of MARK4 mRNA.

In conclusion, though more research is required before translating these results into pre-clinical and clinical studies, the findings presented here that miR-515-5p represses cell proliferation and motility in breast cancer suggest the possibility of using miR-515-5p in a future anti-cancer therapy to treat breast cancer patients by suppressing tumour cell growth or/and preventing metastasis progression. Nevertheless, caution is needed when developing a miR-515-5p-based therapy to treat cancer patients because miR-515-5p has been shown to have an oncogenic function in hepatocellular carcinoma [276], indicating that the role it plays in cancer is dependent on the type of tissue in which it is expressed and acts.

144

References

145

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

2. www.cancerresearchuk.org/

3. Weigelt B, Reis-Filho JS: Histological and molecular types of breast cancer: is there a unifying taxonomy? Nat Rev Clin Oncol 2009, 6(12):718-730.

4. Eroles P, Bosch A, Pérez-Fidalgo JA, Lluch A: Molecular biology in breast cancer: intrinsic subtypes and signaling pathways. Cancer Treat Rev 2012, 38(6):698-707.

5. Rakha EA, Reis-Filho JS, Baehner F, Dabbs DJ, Decker T, Eusebi V, Fox SB, Ichihara S, Jacquemier J, Lakhani SR et al: Breast cancer prognostic classification in the molecular era: the role of histological grade. Breast Cancer Res 2010, 12(4):207.

6. Singletary SE, Allred C, Ashley P, Bassett LW, Berry D, Bland KI, Borgen PI, Clark G, Edge SB, Hayes DF et al: Revision of the American Joint Committee on Cancer staging system for breast cancer. J Clin Oncol 2002, 20(17):3628- 3636.

7. Fidler IJ: The pathogenesis of cancer metastasis: the 'seed and soil' hypothesis revisited. Nat Rev Cancer 2003, 3(6):453-458.

8. Reymond N, d'Água BB, Ridley AJ: Crossing the endothelial barrier during metastasis. Nat Rev Cancer 2013, 13(12):858-870.

9. http://www.bbc.co.uk/schools/gcsebitesize/pe/appliedanatomy/2_an atomy_skeleton_rev4.shtml

10. http://tw.aisj-jhb.com/theblackplaguepaper/2013/05/19/why-is-lung- important/

11. http://uanews.org/story/good-vibrations-mediating-mood-through-brain- ultrasound

12. Nguyen DX, Bos PD, Massagué J: Metastasis: from dissemination to organ- specific colonization. Nat Rev Cancer 2009, 9(4):274-284.

13. Ridley AJ, Schwartz MA, Burridge K, Firtel RA, Ginsberg MH, Borisy G, Parsons JT, Horwitz AR: Cell migration: integrating signals from front to back. Science 2003, 302(5651):1704-1709.

14. Vicente-Manzanares M, Webb DJ, Horwitz AR: Cell migration at a glance. J Cell Sci 2005, 118(Pt 21):4917-4919.

15. Takenawa T, Suetsugu S: The WASP-WAVE protein network: connecting the membrane to the cytoskeleton. Nat Rev Mol Cell Biol 2007, 8(1):37-48.

16. Vicente-Manzanares M, Ma X, Adelstein RS, Horwitz AR: Non-muscle myosin II takes centre stage in cell adhesion and migration. Nat Rev Mol Cell Biol 2009, 10(11):778-790.

146

17. Betapudi V, Licate LS, Egelhoff TT: Distinct roles of nonmuscle myosin II isoforms in the regulation of MDA-MB-231 breast cancer cell spreading and migration. Cancer Res 2006, 66(9):4725-4733.

18. Machesky LM, Li A: Fascin: Invasive filopodia promoting metastasis. Commun Integr Biol 2010, 3(3):263-270.

19. Hashimoto Y, Parsons M, Adams JC: Dual actin-bundling and protein kinase C-binding activities of fascin regulate carcinoma cell migration downstream of Rac and contribute to metastasis. Mol Biol Cell 2007, 18(11):4591-4602.

20. Gupton SL, Eisenmann K, Alberts AS, Waterman-Storer CM: mDia2 regulates actin and focal adhesion dynamics and organization in the lamella for efficient epithelial cell migration. J Cell Sci 2007, 120(Pt 19):3475-3487.

21. Sinnar SA, Antoku S, Saffin JM, Cooper JA, Halpain S: Capping protein is essential for cell migration in vivo and for filopodial morphology and dynamics. Mol Biol Cell 2014, 25(14):2152-2160.

22. Krause M, Dent EW, Bear JE, Loureiro JJ, Gertler FB: Ena/VASP proteins: regulators of the actin cytoskeleton and cell migration. Annu Rev Cell Dev Biol 2003, 19:541-564.

23. Friedl P, Wolf K: Tumour-cell invasion and migration: diversity and escape mechanisms. Nat Rev Cancer 2003, 3(5):362-374.

24. Ponti A, Machacek M, Gupton SL, Waterman-Storer CM, Danuser G: Two distinct actin networks drive the protrusion of migrating cells. Science 2004, 305(5691):1782-1786.

25. Fackler OT, Grosse R: Cell motility through plasma membrane blebbing. J Cell Biol 2008, 181(6):879-884.

26. Poincloux R, Lizárraga F, Chavrier P: Matrix invasion by tumour cells: a focus on MT1-MMP trafficking to invadopodia. J Cell Sci 2009, 122(Pt 17):3015- 3024.

27. McCulloch CA, Downey GP, El-Gabalawy H: Signalling platforms that modulate the inflammatory response: new targets for drug development. Nat Rev Drug Discov 2006, 5(10):864-876.

28. Nürnberg A, Kitzing T, Grosse R: Nucleating actin for invasion. Nat Rev Cancer 2011, 11(3):177-187.

29. Mattila PK, Lappalainen P: Filopodia: molecular architecture and cellular functions. Nat Rev Mol Cell Biol 2008, 9(6):446-454.

30. Drewes G, Ebneth A, Preuss U, Mandelkow EM, Mandelkow E: MARK, a novel family of protein kinases that phosphorylate microtubule-associated proteins and trigger microtubule disruption. Cell 1997, 89(2):297-308.

31. Kato T, Satoh S, Okabe H, Kitahara O, Ono K, Kihara C, Tanaka T, Tsunoda T, Yamaoka Y, Nakamura Y et al: Isolation of a novel human gene, MARKL1, homologous to MARK3 and its involvement in hepatocellular carcinogenesis. Neoplasia 2001, 3(1):4-9.

147

32. Magnani I, Novielli C, Bellini M, Roversi G, Bello L, Larizza L: Multiple localization of endogenous MARK4L protein in human glioma. Cell Oncol 2009, 31(5):357-370.

33. Naz F, Anjum F, Islam A, Ahmad F, Hassan MI: Microtubule affinity-regulating kinase 4: structure, function, and regulation. Cell Biochem Biophys 2013, 67(2):485-499.

34. Trinczek B, Brajenovic M, Ebneth A, Drewes G: MARK4 is a novel microtubule- associated proteins/microtubule affinity-regulating kinase that binds to the cellular microtubule network and to centrosomes. J Biol Chem 2004, 279(7):5915-5923.

35. Beghini A, Magnani I, Roversi G, Piepoli T, Di Terlizzi S, Moroni RF, Pollo B, Fuhrman Conti AM, Cowell JK, Finocchiaro G et al: The neural progenitor- restricted isoform of the MARK4 gene in 19q13.2 is upregulated in human gliomas and overexpressed in a subset of glioblastoma cell lines. Oncogene 2003, 22(17):2581-2591.

36. Brajenovic M, Joberty G, Küster B, Bouwmeester T, Drewes G: Comprehensive proteomic analysis of human Par protein complexes reveals an interconnected protein network. J Biol Chem 2004, 279(13):12804-12811.

37. Sontag JM, Nunbhakdi-Craig V, White CL, Halpain S, Sontag E: The protein phosphatase PP2A/Bα binds to the microtubule-associated proteins Tau and MAP2 at a motif also recognized by the kinase Fyn: implications for tauopathies. J Biol Chem 2012, 287(18):14984-14993.

38. Wilker E, Yaffe MB: 14-3-3 Proteins--a focus on cancer and human disease. J Mol Cell Cardiol 2004, 37(3):633-642.

39. Larsen M, Tremblay ML, Yamada KM: Phosphatases in cell-matrix adhesion and migration. Nat Rev Mol Cell Biol 2003, 4(9):700-711.

40. Chang YC, Nalbant P, Birkenfeld J, Chang ZF, Bokoch GM: GEF-H1 couples nocodazole-induced microtubule disassembly to cell contractility via RhoA. Mol Biol Cell 2008, 19(5):2147-2153.

41. Yamahashi Y, Saito Y, Murata-Kamiya N, Hatakeyama M: Polarity-regulating kinase partitioning-defective 1b (PAR1b) phosphorylates guanine nucleotide exchange factor H1 (GEF-H1) to regulate RhoA-dependent actin cytoskeletal reorganization. J Biol Chem 2011, 286(52):44576-44584.

42. Nalbant P, Chang YC, Birkenfeld J, Chang ZF, Bokoch GM: Guanine nucleotide exchange factor-H1 regulates cell migration via localized activation of RhoA at the leading edge. Mol Biol Cell 2009, 20(18):4070-4082.

43. Simpson E, Misso M, Hewitt K, Hill R, Boon W, Jones M, Kovacic A, Zhou J, Clyne C: Estrogen - the good, the bad, and the unexpected. Endocrine Reviews 2005, 26(3):322-330.

44. Ali S, Buluwela L, Coombes R, Caskey C: Antiestrogens and Their Therapeutic Applications in Breast Cancer and Other Diseases. Annual Review of Medicine, Vol 62, 2011 2011, 62:217-232.

148

45. Simpson ER, Mahendroo MS, Means GD, Kilgore MW, Hinshelwood MM, Graham-Lorence S, Amarneh B, Ito Y, Fisher CR, Michael MD: Aromatase cytochrome P450, the enzyme responsible for estrogen biosynthesis. Endocr Rev 1994, 15(3):342-355.

46. Dellapasqua S, Colleoni M, Gelber RD, Goldhirsch A: Adjuvant endocrine therapy for premenopausal women with early breast cancer. J Clin Oncol 2005, 23(8):1736-1750.

47. Osborne CK: Aromatase inhibitors in relation to other forms of endocrine therapy for breast cancer. Endocr Relat Cancer 1999, 6(2):271-276.

48. Jordan V: Tamoxifen: A most unlikely pioneering medicine. Nature Reviews Drug Discovery 2003, 2(3):205-213.

49. http://www.anoleannals.org/2011/05/27/sex-requires-more-than-just- testosterone%E2%80%A6/

50. Boyd S: On Oöphorectomy in the Treatment of Cancer. Br Med J 1897, 2(1918):890-896.

51. Hall JM, Couse JF, Korach KS: The multifaceted mechanisms of estradiol and estrogen receptor signaling. J Biol Chem 2001, 276(40):36869-36872.

52. Klein-Hitpass L, Schorpp M, Wagner U, Ryffel GU: An estrogen-responsive element derived from the 5' flanking region of the Xenopus vitellogenin A2 gene functions in transfected human cells. Cell 1986, 46(7):1053-1061.

53. Saville B, Wormke M, Wang F, Nguyen T, Enmark E, Kuiper G, Gustafsson JA, Safe S: Ligand-, cell-, and estrogen receptor subtype (alpha/beta)- dependent activation at GC-rich (Sp1) promoter elements. J Biol Chem 2000, 275(8):5379-5387.

54. Pruett ST, Bushnev A, Hagedorn K, Adiga M, Haynes CA, Sullards MC, Liotta DC, Merrill AH: Biodiversity of sphingoid bases ("sphingosines") and related amino alcohols. J Lipid Res 2008, 49(8):1621-1639.

55. Brown DA, London E: Structure and function of sphingolipid- and cholesterol-rich membrane rafts. J Biol Chem 2000, 275(23):17221-17224.

56. Pitson SM: Regulation of sphingosine kinase and sphingolipid signaling. Trends Biochem Sci 2011, 36(2):97-107.

57. Hait NC, Allegood J, Maceyka M, Strub GM, Harikumar KB, Singh SK, Luo C, Marmorstein R, Kordula T, Milstien S et al: Regulation of histone acetylation in the nucleus by sphingosine-1-phosphate. Science 2009, 325(5945):1254- 1257.

58. Pitson S, Moretti P, Zebol J, Lynn H, Xia P, Vadas M, Wattenberg B: Activation of sphingosine kinase 1 by ERK1/2-mediated phosphorylation. Embo Journal 2003:5491-5500.

59. Hait NC, Bellamy A, Milstien S, Kordula T, Spiegel S: Sphingosine kinase type 2 activation by ERK-mediated phosphorylation. J Biol Chem 2007, 282(16):12058-12065.

149

60. Ding G, Sonoda H, Yu H, Kajimoto T, Goparaju SK, Jahangeer S, Okada T, Nakamura S: Protein kinase D-mediated phosphorylation and nuclear export of sphingosine kinase 2. J Biol Chem 2007, 282(37):27493-27502.

61. Spiegel S, Milstien S: Exogenous and intracellularly generated sphingosine 1-phosphate can regulate cellular processes by divergent pathways. Biochemical Society Transactions 2003, 31:1216-1219.

62. Fletcher JI, Haber M, Henderson MJ, Norris MD: ABC transporters in cancer: more than just drug efflux pumps. Nat Rev Cancer 2010, 10(2):147-156.

63. Brinkmann V: Sphingosine 1-phosphate receptors in health and disease: Mechanistic insights from gene deletion studies and reverse pharmacology. Pharmacology & Therapeutics 2007, 115(1):84-105.

64. Hla T: Sphingosine 1-phosphate receptors. Prostaglandins & Other Lipid Mediators 2001, 64(1-4):135-142.

65. Kluk M, Hla T: Signaling of sphingosine-1-phosphate via the S1P/EDG-family of G-protein-coupled receptors. Biochimica Et Biophysica Acta-Molecular and Cell Biology of Lipids 2002, 1582(1-3):72-80.

66. Takuwa Y: Subtype-specific differential regulation of Rho family G proteins and cell migration by the Edg family sphingosine-1-phosphate receptors. Biochimica Et Biophysica Acta-Molecular and Cell Biology of Lipids 2002, 1582(1-3):112-120.

67. Lee MJ, Thangada S, Claffey KP, Ancellin N, Liu CH, Kluk M, Volpi M, Sha'afi RI, Hla T: Vascular endothelial cell adherens junction assembly and morphogenesis induced by sphingosine-1-phosphate. Cell 1999, 99(3):301- 312.

68. Zondag GC, Postma FR, Etten IV, Verlaan I, Moolenaar WH: Sphingosine 1- phosphate signalling through the G-protein-coupled receptor Edg-1. Biochem J 1998, 330 ( Pt 2):605-609.

69. Kon J, Sato K, Watanabe T, Tomura H, Kuwabara A, Kimura T, Tamama K, Ishizuka T, Murata N, Kanda T et al: Comparison of intrinsic activities of the putative sphingosine 1-phosphate receptor subtypes to regulate several signaling pathways in their cDNA-transfected Chinese hamster ovary cells. J Biol Chem 1999, 274(34):23940-23947.

70. Paik JH, Chae Ss, Lee MJ, Thangada S, Hla T: Sphingosine 1-phosphate- induced endothelial cell migration requires the expression of EDG-1 and EDG-3 receptors and Rho-dependent activation of alpha vbeta3- and beta1- containing integrins. J Biol Chem 2001, 276(15):11830-11837.

71. Okamoto H, Takuwa N, Yokomizo T, Sugimoto N, Sakurada S, Shigematsu H, Takuwa Y: Inhibitory regulation of Rac activation, membrane ruffling, and cell migration by the G protein-coupled sphingosine-1-phosphate receptor EDG5 but not EDG1 or EDG3. Mol Cell Biol 2000, 20(24):9247-9261.

72. Van Brocklyn JR, Tu Z, Edsall LC, Schmidt RR, Spiegel S: Sphingosine 1- phosphate-induced cell rounding and neurite retraction are mediated by the G protein-coupled receptor H218. J Biol Chem 1999, 274(8):4626-4632.

150

73. An S, Zheng Y, Bleu T: Sphingosine 1-phosphate-induced cell proliferation, survival, and related signaling events mediated by G protein-coupled receptors Edg3 and Edg5. J Biol Chem 2000, 275(1):288-296.

74. Wang W, Graeler MH, Goetzl EJ: Type 4 sphingosine 1-phosphate G protein- coupled receptor (S1P4) transduces S1P effects on T cell proliferation and cytokine secretion without signaling migration. FASEB J 2005, 19(12):1731- 1733.

75. Im DS, Heise CE, Ancellin N, O'Dowd BF, Shei GJ, Heavens RP, Rigby MR, Hla T, Mandala S, McAllister G et al: Characterization of a novel sphingosine 1- phosphate receptor, Edg-8. J Biol Chem 2000, 275(19):14281-14286.

76. Malek RL, Toman RE, Edsall LC, Wong S, Chiu J, Letterle CA, Van Brocklyn JR, Milstien S, Spiegel S, Lee NH: Nrg-1 belongs to the endothelial differentiation gene family of G protein-coupled sphingosine-1-phosphate receptors. J Biol Chem 2001, 276(8):5692-5699.

77. Imamura T, Yamamoto S, Ohgane J, Hattori N, Tanaka S, Shiota K: Non-coding RNA directed DNA demethylation of Sphk1 CpG island. Biochem Biophys Res Commun 2004, 322(2):593-600.

78. Imamura T, Miyauchi-Senda N, Tanaka S, Shiota K: Identification of genetic and epigenetic similarities of SPHK1/Sphk1 in mammals. J Vet Med Sci 2004, 66(11):1387-1393.

79. Schwalm S, Döll F, Römer I, Bubnova S, Pfeilschifter J, Huwiler A: Sphingosine kinase-1 is a hypoxia-regulated gene that stimulates migration of human endothelial cells. Biochem Biophys Res Commun 2008, 368(4):1020-1025.

80. Sobue S, Murakami M, Banno Y, Ito H, Kimura A, Gao S, Furuhata A, Takagi A, Kojima T, Suzuki M et al: v-Src oncogene product increases sphingosine kinase 1 expression through mRNA stabilization: alteration of AU-rich element-binding proteins. Oncogene 2008, 27(46):6023-6033.

81. Taha TA, Osta W, Kozhaya L, Bielawski J, Johnson KR, Gillanders WE, Dbaibo GS, Hannun YA, Obeid LM: Down-regulation of sphingosine kinase-1 by DNA damage: dependence on proteases and p53. J Biol Chem 2004, 279(19):20546-20554.

82. Lebman DA, Spiegel S: Cross-talk at the crossroads of sphingosine-1- phosphate, growth factors, and cytokine signaling. J Lipid Res 2008, 49(7):1388-1394.

83. Pchejetski D, Golzio M, Bonhoure E, Calvet C, Doumerc N, Garcia V, Mazerolles C, Rischmann P, Teissie J, Malavaud B et al: Sphingosine kinase-1 as a chemotherapy sensor in prostate adenocarcinoma cell and mouse models. Cancer Research 2005:11667-11675.

84. Sauer L, Nunes J, Salunkhe V, Skalska L, Kohama T, Cuvillier O, Waxman J, Pchejetski D: Sphingosine kinase 1 inhibition sensitizes hormone-resistant prostate cancer to docetaxel. International Journal of Cancer 2009:2728-2736.

85. Pchejetski D, Doumerc N, Golzio M, Naymark M, Teissie J, Kohama T, Waxman J, Malavaud B, Cuvillier O: Chemosensitizing effects of sphingosine kinase-1

151

inhibition in prostate cancer cell and animal models. Molecular Cancer Therapeutics 2008:1836-1845.

86. Guillermet-Guibert J, Davenne L, Pchejetski D, Saint-Laurent N, Brizuela L, Guilbeau-Frugier C, Delisle M, Cuvillier O, Susini C, Bousquet C: Targeting the sphingolipid metabolism to defeat pancreatic cancer cell resistance to the chemotherapeutic gemcitabine drug. Molecular Cancer Therapeutics 2009:809-820.

87. Pchejetski D, Bohler T, Brizuela L, Sauer L, Doumerc N, Golzio M, Salunkhe V, Teissie J, Malavaud B, Waxman J et al: FTY720 (Fingolimod) Sensitizes Prostate Cancer Cells to Radiotherapy by Inhibition of Sphingosine Kinase- 1. Cancer Research 2010:8651-8661.

88. French K, Schrecengost R, Lee B, Zhuang Y, Smith S, Eberly J, Yun J, Smith C: Discovery and evaluation of inhibitors of human sphingosine kinase. Cancer Research 2003:5962-5969.

89. Malavaud B, Pchejetski D, Mazerolles C, de Paiva GR, Calvet C, Doumerc N, Pitson S, Rischmann P, Cuvillier O: Sphingosine kinase-1 activity and expression in human prostate cancer resection specimens. Eur J Cancer 2010.

90. Li J, Guan H, Gong L, Song L, Zhang N, Wu J, Yuan J, Zheng Y, Huang Z, Li M: Clinical Significance of Sphingosine Kinase-1 Expression in Human Astrocytomas Progression and Overall Patient Survival. Clinical Cancer Research 2008:6996-7003.

91. Li W, Yu C, Xia J, Zhang L, Weng G, Zheng H, Kong Q, Hu L, Zeng M, Zeng Y et al: Sphingosine Kinase 1 Is Associated with Gastric Cancer Progression and Poor Survival of Patients. Clinical Cancer Research 2009:1393-1399.

92. Ruckhaberle E, Rody A, Engels K, Gaetje R, von Minckwitz G, Schiffmann S, Grosch S, Geisslinger G, Holtrich U, Karn T et al: Microarray analysis of altered sphingolipid metabolism reveals prognostic significance of sphingosine kinase 1 in breast cancer. Breast Cancer Research and Treatment 2008:41-52.

93. Pitman MR, Pitson SM: Inhibitors of the sphingosine kinase pathway as potential therapeutics. Curr Cancer Drug Targets 2010, 10(4):354-367.

94. Ruckhäberle E, Rody A, Engels K, Gaetje R, von Minckwitz G, Schiffmann S, Grösch S, Geisslinger G, Holtrich U, Karn T et al: Microarray analysis of altered sphingolipid metabolism reveals prognostic significance of sphingosine kinase 1 in breast cancer. Breast Cancer Res Treat 2008, 112(1):41-52.

95. Sarkar S, Maceyka M, Hait N, Paugh S, Sankala H, Milstien S, Spiegel S: Sphingosine kinase 1 is required for migration, proliferation and survival of MCF-7 human breast cancer cells. Febs Letters 2005, 579(24):5313-5317.

96. Nava V, Hobson J, Murthy S, Milstien S, Spiegel S: Sphingosine kinase type 1 promotes estrogen-dependent tumorigenesis of breast cancer MCF-7 cells. Experimental Cell Research 2002, 281(1):115-127.

152

97. Sukocheva OA, Wang L, Albanese N, Pitson SM, Vadas MA, Xia P: Sphingosine kinase transmits estrogen signaling in human breast cancer cells. Mol Endocrinol 2003, 17(10):2002-2012.

98. Lee RC, Feinbaum RL, Ambros V: The C. elegans heterochronic gene lin-4 encodes small RNAs with antisense complementarity to lin-14. Cell 1993, 75(5):843-854.

99. Wightman B, Ha I, Ruvkun G: Posttranscriptional regulation of the heterochronic gene lin-14 by lin-4 mediates temporal pattern formation in C. elegans. Cell 1993, 75(5):855-862.

100. Reinhart BJ, Slack FJ, Basson M, Pasquinelli AE, Bettinger JC, Rougvie AE, Horvitz HR, Ruvkun G: The 21-nucleotide let-7 RNA regulates developmental timing in Caenorhabditis elegans. Nature 2000, 403(6772):901-906.

101. Pasquinelli AE, Reinhart BJ, Slack F, Martindale MQ, Kuroda MI, Maller B, Hayward DC, Ball EE, Degnan B, Müller P et al: Conservation of the sequence and temporal expression of let-7 heterochronic regulatory RNA. Nature 2000, 408(6808):86-89.

102. He L, Hannon G: Micrornas: Small RNAs with a big role in gene regulation. Nature Reviews Genetics 2004:522-531.

103. Winter J, Jung S, Keller S, Gregory RI, Diederichs S: Many roads to maturity: microRNA biogenesis pathways and their regulation. Nat Cell Biol 2009, 11(3):228-234.

104. Saini HK, Griffiths-Jones S, Enright AJ: Genomic analysis of human microRNA transcripts. Proc Natl Acad Sci U S A 2007, 104(45):17719-17724.

105. Cai X, Hagedorn CH, Cullen BR: Human microRNAs are processed from capped, polyadenylated transcripts that can also function as mRNAs. RNA 2004, 10(12):1957-1966.

106. Lee Y, Kim M, Han J, Yeom KH, Lee S, Baek SH, Kim VN: MicroRNA genes are transcribed by RNA polymerase II. EMBO J 2004, 23(20):4051-4060.

107. Denli AM, Tops BB, Plasterk RH, Ketting RF, Hannon GJ: Processing of primary microRNAs by the Microprocessor complex. Nature 2004, 432(7014):231-235.

108. Han J, Lee Y, Yeom KH, Nam JW, Heo I, Rhee JK, Sohn SY, Cho Y, Zhang BT, Kim VN: Molecular basis for the recognition of primary microRNAs by the Drosha-DGCR8 complex. Cell 2006, 125(5):887-901.

109. Okada C, Yamashita E, Lee SJ, Shibata S, Katahira J, Nakagawa A, Yoneda Y, Tsukihara T: A high-resolution structure of the pre-microRNA nuclear export machinery. Science 2009, 326(5957):1275-1279.

110. Yi R, Qin Y, Macara IG, Cullen BR: Exportin-5 mediates the nuclear export of pre-microRNAs and short hairpin RNAs. Genes Dev 2003, 17(24):3011-3016.

111. Kim VN: MicroRNA biogenesis: coordinated cropping and dicing. Nat Rev Mol Cell Biol 2005, 6(5):376-385.

153

112. Khvorova A, Reynolds A, Jayasena SD: Functional siRNAs and miRNAs exhibit strand bias. Cell 2003, 115(2):209-216.

113. Okamura K, Phillips MD, Tyler DM, Duan H, Chou YT, Lai EC: The regulatory activity of microRNA* species has substantial influence on microRNA and 3' UTR evolution. Nat Struct Mol Biol 2008, 15(4):354-363.

114. Huang CJ, Nguyen PN, Choo KB, Sugii S, Wee K, Cheong SK, Kamarul T: Frequent co-expression of miRNA-5p and -3p species and cross-targeting in induced pluripotent stem cells. Int J Med Sci 2014, 11(8):824-833.

115. Griffiths-Jones S, Grocock RJ, van Dongen S, Bateman A, Enright AJ: miRBase: microRNA sequences, targets and gene nomenclature. Nucleic Acids Res 2006, 34(Database issue):D140-144.

116. Chendrimada TP, Gregory RI, Kumaraswamy E, Norman J, Cooch N, Nishikura K, Shiekhattar R: TRBP recruits the Dicer complex to Ago2 for microRNA processing and gene silencing. Nature 2005, 436(7051):740-744.

117. Lai EC: Micro RNAs are complementary to 3' UTR sequence motifs that mediate negative post-transcriptional regulation. Nat Genet 2002, 30(4):363- 364.

118. Lytle JR, Yario TA, Steitz JA: Target mRNAs are repressed as efficiently by microRNA-binding sites in the 5' UTR as in the 3' UTR. Proc Natl Acad Sci U S A 2007, 104(23):9667-9672.

119. Rigoutsos I: New tricks for animal microRNAS: targeting of amino acid coding regions at conserved and nonconserved sites. Cancer Res 2009, 69(8):3245-3248.

120. Bartel DP: MicroRNAs: target recognition and regulatory functions. Cell 2009, 136(2):215-233.

121. Orban TI, Izaurralde E: Decay of mRNAs targeted by RISC requires XRN1, the Ski complex, and the exosome. RNA 2005, 11(4):459-469.

122. Chendrimada TP, Finn KJ, Ji X, Baillat D, Gregory RI, Liebhaber SA, Pasquinelli AE, Shiekhattar R: MicroRNA silencing through RISC recruitment of eIF6. Nature 2007, 447(7146):823-828.

123. Kiriakidou M, Tan GS, Lamprinaki S, De Planell-Saguer M, Nelson PT, Mourelatos Z: An mRNA m7G cap binding-like motif within human Ago2 represses translation. Cell 2007, 129(6):1141-1151.

124. Petersen CP, Bordeleau ME, Pelletier J, Sharp PA: Short RNAs repress translation after initiation in mammalian cells. Mol Cell 2006, 21(4):533-542.

125. Nottrott S, Simard MJ, Richter JD: Human let-7a miRNA blocks protein production on actively translating polyribosomes. Nat Struct Mol Biol 2006, 13(12):1108-1114.

126. Wu L, Fan J, Belasco JG: MicroRNAs direct rapid deadenylation of mRNA. Proc Natl Acad Sci U S A 2006, 103(11):4034-4039.

154

127. Braun JE, Huntzinger E, Fauser M, Izaurralde E: GW182 proteins directly recruit cytoplasmic deadenylase complexes to miRNA targets. Mol Cell 2011, 44(1):120-133.

128. Behm-Ansmant I, Rehwinkel J, Doerks T, Stark A, Bork P, Izaurralde E: mRNA degradation by miRNAs and GW182 requires both CCR4:NOT deadenylase and DCP1:DCP2 decapping complexes. Genes Dev 2006, 20(14):1885-1898.

129. Kim DH, Saetrom P, Snøve O, Rossi JJ: MicroRNA-directed transcriptional gene silencing in mammalian cells. Proc Natl Acad Sci U S A 2008, 105(42):16230-16235.

130. Place RF, Li LC, Pookot D, Noonan EJ, Dahiya R: MicroRNA-373 induces expression of genes with complementary promoter sequences. Proc Natl Acad Sci U S A 2008, 105(5):1608-1613.

131. Vasudevan S, Tong Y, Steitz JA: Switching from repression to activation: microRNAs can up-regulate translation. Science 2007, 318(5858):1931-1934.

132. Vasudevan S, Steitz JA: AU-rich-element-mediated upregulation of translation by FXR1 and Argonaute 2. Cell 2007, 128(6):1105-1118.

133. Hansen TB, Wiklund ED, Bramsen JB, Villadsen SB, Statham AL, Clark SJ, Kjems J: miRNA-dependent gene silencing involving Ago2-mediated cleavage of a circular antisense RNA. EMBO J 2011, 30(21):4414-4422.

134. Tang R, Li L, Zhu D, Hou D, Cao T, Gu H, Zhang J, Chen J, Zhang CY, Zen K: Mouse miRNA-709 directly regulates miRNA-15a/16-1 biogenesis at the posttranscriptional level in the nucleus: evidence for a microRNA hierarchy system. Cell Res 2012, 22(3):504-515.

135. Kuhn DE, Martin MM, Feldman DS, Terry AV, Nuovo GJ, Elton TS: Experimental validation of miRNA targets. Methods 2008, 44(1):47-54.

136. Lim LP, Lau NC, Garrett-Engele P, Grimson A, Schelter JM, Castle J, Bartel DP, Linsley PS, Johnson JM: Microarray analysis shows that some microRNAs downregulate large numbers of target mRNAs. Nature 2005, 433(7027):769- 773.

137. Xu G, Fewell C, Taylor C, Deng N, Hedges D, Wang X, Zhang K, Lacey M, Zhang H, Yin Q et al: Transcriptome and targetome analysis in MIR155 expressing cells using RNA-seq. RNA 2010, 16(8):1610-1622.

138. Hafner M, Landthaler M, Burger L, Khorshid M, Hausser J, Berninger P, Rothballer A, Ascano M, Jungkamp AC, Munschauer M et al: Transcriptome- wide identification of RNA-binding protein and microRNA target sites by PAR-CLIP. Cell 2010, 141(1):129-141.

139. Orom UA, Lund AH: Isolation of microRNA targets using biotinylated synthetic microRNAs. Methods 2007, 43(2):162-165.

140. Vatolin S, Navaratne K, Weil RJ: A novel method to detect functional microRNA targets. J Mol Biol 2006, 358(4):983-996.

155

141. Vinther J, Hedegaard MM, Gardner PP, Andersen JS, Arctander P: Identification of miRNA targets with stable isotope labeling by amino acids in cell culture. Nucleic Acids Res 2006, 34(16):e107.

142. Hendrickson DG, Hogan DJ, McCullough HL, Myers JW, Herschlag D, Ferrell JE, Brown PO: Concordant regulation of translation and mRNA abundance for hundreds of targets of a human microRNA. PLoS Biol 2009, 7(11):e1000238.

143. Calin GA, Dumitru CD, Shimizu M, Bichi R, Zupo S, Noch E, Aldler H, Rattan S, Keating M, Rai K et al: Frequent deletions and down-regulation of micro- RNA genes miR15 and miR16 at 13q14 in chronic lymphocytic leukemia. Proc Natl Acad Sci U S A 2002, 99(24):15524-15529.

144. Calin GA, Croce CM: MicroRNA signatures in human cancers. Nat Rev Cancer 2006, 6(11):857-866.

145. Deng S, Calin GA, Croce CM, Coukos G, Zhang L: Mechanisms of microRNA deregulation in human cancer. Cell Cycle 2008, 7(17):2643-2646.

146. Lujambio A, Ropero S, Ballestar E, Fraga MF, Cerrato C, Setién F, Casado S, Suarez-Gauthier A, Sanchez-Cespedes M, Git A et al: Genetic unmasking of an epigenetically silenced microRNA in human cancer cells. Cancer Res 2007, 67(4):1424-1429.

147. O'Donnell KA, Wentzel EA, Zeller KI, Dang CV, Mendell JT: c-Myc-regulated microRNAs modulate E2F1 expression. Nature 2005, 435(7043):839-843.

148. Wang M, Gu H, Qian H, Zhu W, Zhao C, Zhang X, Tao Y, Zhang L, Xu W: miR- 17-5p/20a are important markers for gastric cancer and murine double minute 2 participates in their functional regulation. Eur J Cancer 2013, 49(8):2010-2021.

149. Yu J, Ohuchida K, Mizumoto K, Fujita H, Nakata K, Tanaka M: MicroRNA miR- 17-5p is overexpressed in pancreatic cancer, associated with a poor prognosis, and involved in cancer cell proliferation and invasion. Cancer Biol Ther 2010, 10(8):748-757.

150. Hayashita Y, Osada H, Tatematsu Y, Yamada H, Yanagisawa K, Tomida S, Yatabe Y, Kawahara K, Sekido Y, Takahashi T: A polycistronic microRNA cluster, miR-17-92, is overexpressed in human lung cancers and enhances cell proliferation. Cancer Res 2005, 65(21):9628-9632.

151. Mu P, Han YC, Betel D, Yao E, Squatrito M, Ogrodowski P, de Stanchina E, D'Andrea A, Sander C, Ventura A: Genetic dissection of the miR-17~92 cluster of microRNAs in Myc-induced B-cell lymphomas. Genes Dev 2009, 23(24):2806-2811.

152. Thapa DR, Li X, Jamieson BD, Martínez-Maza O: Overexpression of microRNAs from the miR-17-92 paralog clusters in AIDS-related non- Hodgkin's lymphomas. PLoS One 2011, 6(6):e20781.

153. Suzuki HI, Yamagata K, Sugimoto K, Iwamoto T, Kato S, Miyazono K: Modulation of microRNA processing by p53. Nature 2009, 460(7254):529- 533.

156

154. Esquela-Kerscher A, Slack FJ: Oncomirs - microRNAs with a role in cancer. Nat Rev Cancer 2006, 6(4):259-269.

155. Zhang B, Pan X, Cobb GP, Anderson TA: microRNAs as oncogenes and tumor suppressors. Dev Biol 2007, 302(1):1-12.

156. Lu J, Getz G, Miska EA, Alvarez-Saavedra E, Lamb J, Peck D, Sweet-Cordero A, Ebert BL, Mak RH, Ferrando AA et al: MicroRNA expression profiles classify human cancers. Nature 2005, 435(7043):834-838.

157. Melo SA, Ropero S, Moutinho C, Aaltonen LA, Yamamoto H, Calin GA, Rossi S, Fernandez AF, Carneiro F, Oliveira C et al: A TARBP2 mutation in human cancer impairs microRNA processing and DICER1 function. Nat Genet 2009, 41(3):365-370.

158. Hill DA, Ivanovich J, Priest JR, Gurnett CA, Dehner LP, Desruisseau D, Jarzembowski JA, Wikenheiser-Brokamp KA, Suarez BK, Whelan AJ et al: DICER1 mutations in familial pleuropulmonary blastoma. Science 2009, 325(5943):965.

159. Melo SA, Moutinho C, Ropero S, Calin GA, Rossi S, Spizzo R, Fernandez AF, Davalos V, Villanueva A, Montoya G et al: A genetic defect in exportin-5 traps precursor microRNAs in the nucleus of cancer cells. Cancer Cell 2010, 18(4):303-315.

160. Rosenfeld N, Aharonov R, Meiri E, Rosenwald S, Spector Y, Zepeniuk M, Benjamin H, Shabes N, Tabak S, Levy A et al: MicroRNAs accurately identify cancer tissue origin. Nat Biotechnol 2008, 26(4):462-469.

161. Zhu S, Si ML, Wu H, Mo YY: MicroRNA-21 targets the tumor suppressor gene tropomyosin 1 (TPM1). J Biol Chem 2007, 282(19):14328-14336.

162. Wickramasinghe NS, Manavalan TT, Dougherty SM, Riggs KA, Li Y, Klinge CM: Estradiol downregulates miR-21 expression and increases miR-21 target gene expression in MCF-7 breast cancer cells. Nucleic Acids Res 2009, 37(8):2584-2595.

163. Frankel LB, Christoffersen NR, Jacobsen A, Lindow M, Krogh A, Lund AH: Programmed cell death 4 (PDCD4) is an important functional target of the microRNA miR-21 in breast cancer cells. J Biol Chem 2008, 283(2):1026- 1033.

164. Huang Q, Gumireddy K, Schrier M, le Sage C, Nagel R, Nair S, Egan DA, Li A, Huang G, Klein-Szanto AJ et al: The microRNAs miR-373 and miR-520c promote tumour invasion and metastasis. Nat Cell Biol 2008, 10(2):202-210.

165. Liu P, Wilson MJ: miR-520c and miR-373 upregulate MMP9 expression by targeting mTOR and SIRT1, and activate the Ras/Raf/MEK/Erk signaling pathway and NF-κB factor in human fibrosarcoma cells. J Cell Physiol 2012, 227(2):867-876.

166. Li L, Yuan L, Luo J, Gao J, Guo J, Xie X: MiR-34a inhibits proliferation and migration of breast cancer through down-regulation of Bcl-2 and SIRT1. Clin Exp Med 2013, 13(2):109-117.

157

167. Hossain A, Kuo MT, Saunders GF: Mir-17-5p regulates breast cancer cell proliferation by inhibiting translation of AIB1 mRNA. Mol Cell Biol 2006, 26(21):8191-8201.

168. Bhat-Nakshatri P, Wang G, Collins NR, Thomson MJ, Geistlinger TR, Carroll JS, Brown M, Hammond S, Srour EF, Liu Y et al: Estradiol-regulated microRNAs control estradiol response in breast cancer cells. Nucleic acids research 2009, 37(14):4850-4861.

169. Di Leva G, Gasparini P, Piovan C, Ngankeu A, Garofalo M, Taccioli C, Iorio MV, Li M, Volinia S, Alder H et al: MicroRNA cluster 221-222 and estrogen receptor alpha interactions in breast cancer. J Natl Cancer Inst 2010, 102(10):706-721.

170. Zhao JJ, Lin J, Yang H, Kong W, He L, Ma X, Coppola D, Cheng JQ: MicroRNA- 221/222 negatively regulates estrogen receptor alpha and is associated with tamoxifen resistance in breast cancer. J Biol Chem 2008, 283(45):31079- 31086.

171. Adams BD, Furneaux H, White BA: The micro-ribonucleic acid (miRNA) miR- 206 targets the human estrogen receptor-alpha (ERalpha) and represses ERalpha messenger RNA and protein expression in breast cancer cell lines. Mol Endocrinol 2007, 21(5):1132-1147.

172. Adams BD, Cowee DM, White BA: The role of miR-206 in the epidermal growth factor (EGF) induced repression of estrogen receptor-alpha (ERalpha) signaling and a luminal phenotype in MCF-7 breast cancer cells. Mol Endocrinol 2009, 23(8):1215-1230.

173. O'Day E, Lal A: MicroRNAs and their target gene networks in breast cancer. Breast Cancer Res 2010, 12(2):201.

174. Pencheva N, Tavazoie SF: Control of metastatic progression by microRNA regulatory networks. Nat Cell Biol 2013, 15(6):546-554.

175. Negrini M, Calin GA: Breast cancer metastasis: a microRNA story. Breast Cancer Res 2008, 10(2):203.

176. Valastyan S, Reinhardt F, Benaich N, Calogrias D, Szász AM, Wang ZC, Brock JE, Richardson AL, Weinberg RA: A pleiotropically acting microRNA, miR-31, inhibits breast cancer metastasis. Cell 2009, 137(6):1032-1046.

177. Ma L, Teruya-Feldstein J, Weinberg RA: Tumour invasion and metastasis initiated by microRNA-10b in breast cancer. Nature 2007, 449(7163):682-688.

178. Zhang Y, Yang P, Sun T, Li D, Xu X, Rui Y, Li C, Chong M, Ibrahim T, Mercatali L et al: miR-126 and miR-126* repress recruitment of mesenchymal stem cells and inflammatory monocytes to inhibit breast cancer metastasis. Nat Cell Biol 2013, 15(3):284-294.

179. Zhou W, Fong MY, Min Y, Somlo G, Liu L, Palomares MR, Yu Y, Chow A, O'Connor ST, Chin AR et al: Cancer-secreted miR-105 destroys vascular endothelial barriers to promote metastasis. Cancer Cell 2014, 25(4):501-515.

180. Ma L, Young J, Prabhala H, Pan E, Mestdagh P, Muth D, Teruya-Feldstein J, Reinhardt F, Onder TT, Valastyan S et al: miR-9, a MYC/MYCN-activated

158

microRNA, regulates E-cadherin and cancer metastasis. Nat Cell Biol 2010, 12(3):247-256.

181. Melo SA, Kalluri R: miR-29b moulds the tumour microenvironment to repress metastasis. Nat Cell Biol 2013, 15(2):139-140.

182. Korpal M, Ell BJ, Buffa FM, Ibrahim T, Blanco MA, Celià-Terrassa T, Mercatali L, Khan Z, Goodarzi H, Hua Y et al: Direct targeting of Sec23a by miR-200s influences cancer cell secretome and promotes metastatic colonization. Nat Med 2011, 17(9):1101-1108.

183. Burk U, Schubert J, Wellner U, Schmalhofer O, Vincan E, Spaderna S, Brabletz T: A reciprocal repression between ZEB1 and members of the miR-200 family promotes EMT and invasion in cancer cells. EMBO Rep 2008, 9(6):582-589.

184. Yu F, Yao H, Zhu P, Zhang X, Pan Q, Gong C, Huang Y, Hu X, Su F, Lieberman J et al: let-7 regulates self renewal and tumorigenicity of breast cancer cells. Cell 2007, 131(6):1109-1123.

185. Park SM, Shell S, Radjabi AR, Schickel R, Feig C, Boyerinas B, Dinulescu DM, Lengyel E, Peter ME: Let-7 prevents early cancer progression by suppressing expression of the embryonic gene HMGA2. Cell Cycle 2007, 6(21):2585-2590.

186. Yun J, Frankenberger CA, Kuo WL, Boelens MC, Eves EM, Cheng N, Liang H, Li WH, Ishwaran H, Minn AJ et al: Signalling pathway for RKIP and Let-7 regulates and predicts metastatic breast cancer. EMBO J 2011, 30(21):4500- 4514.

187. Tavazoie SF, Alarcón C, Oskarsson T, Padua D, Wang Q, Bos PD, Gerald WL, Massagué J: Endogenous human microRNAs that suppress breast cancer metastasis. Nature 2008, 451(7175):147-152.

188. Castellano L, Giamas G, Jacob J, Coombes RC, Lucchesi W, Thiruchelvam P, Barton G, Jiao LR, Wait R, Waxman J et al: The estrogen receptor-alpha- induced microRNA signature regulates itself and its transcriptional response. Proc Natl Acad Sci U S A 2009, 106(37):15732-15737.

189. Liu WH, Yeh SH, Lu CC, Yu SL, Chen HY, Lin CY, Chen DS, Chen PJ: MicroRNA-18a prevents estrogen receptor-alpha expression, promoting proliferation of hepatocellular carcinoma cells. Gastroenterology 2009, 136(2):683-693.

190. Spizzo R, Nicoloso MS, Lupini L, Lu Y, Fogarty J, Rossi S, Zagatti B, Fabbri M, Veronese A, Liu X et al: miR-145 participates with TP53 in a death-promoting regulatory loop and targets estrogen receptor-alpha in human breast cancer cells. Cell Death Differ 2010, 17(2):246-254.

191. Pandey DP, Picard D: miR-22 inhibits estrogen signaling by directly targeting the estrogen receptor alpha mRNA. Mol Cell Biol 2009, 29(13):3783- 3790.

192. Chu HW, Cheng CW, Chou WC, Hu LY, Wang HW, Hsiung CN, Hsu HM, Wu PE, Hou MF, Shen CY et al: A novel estrogen receptor-microRNA 190a-PAR-

159

1-pathway regulates breast cancer progression, a finding initially suggested by genome-wide analysis of loci associated with lymph-node metastasis. Hum Mol Genet 2014, 23(2):355-367.

193. Kondo N, Toyama T, Sugiura H, Fujii Y, Yamashita H: miR-206 expression is down-regulated in estrogen receptor alpha-positive human breast cancer. Cancer Research 2008:5004-5008.

194. de Souza Rocha Simonini P, Breiling A, Gupta N, Malekpour M, Youns M, Omranipour R, Malekpour F, Volinia S, Croce CM, Najmabadi H et al: Epigenetically deregulated microRNA-375 is involved in a positive feedback loop with estrogen receptor alpha in breast cancer cells. Cancer Res 2010, 70(22):9175-9184.

195. Li X, Mertens-Talcott SU, Zhang S, Kim K, Ball J, Safe S: MicroRNA-27a Indirectly Regulates Estrogen Receptor {alpha} Expression and Hormone Responsiveness in MCF-7 Breast Cancer Cells. Endocrinology 2010, 151(6):2462-2473.

196. Calin G, Croce C: MicroRNA signatures in human cancers. Nature Reviews Cancer 2006:857-866.

197. Bader AG, Brown D, Stoudemire J, Lammers P: Developing therapeutic microRNAs for cancer. Gene Ther 2011, 18(12):1121-1126.

198. Bouchie A: First microRNA mimic enters clinic. Nat Biotechnol 2013, 31(7):577.

199. Clinicaltrials.gov

200. Hother C, Cortez C, Miranda T, Yoo C, Workman C, Jones P, Gronbak K: Upregulation of the miR-515 cluster in the MCF7 breast cancer cell line by epigenetic therapy. Apmis 2008, 116(5):437-437.

201. Lee YM, Lee JY, Ho CC, Hong QS, Yu SL, Tzeng CR, Yang PC, Chen HW: miRNA-34b as a tumor suppressor in estrogen-dependent growth of breast cancer cells. Breast Cancer Res 2011, 13(6):R116.

202. Katso RM, Pardo OE, Palamidessi A, Franz CM, Marinov M, De Laurentiis A, Downward J, Scita G, Ridley AJ, Waterfield MD et al: Phosphoinositide 3- Kinase C2beta regulates cytoskeletal organization and cell migration via Rac-dependent mechanisms. Mol Biol Cell 2006, 17(9):3729-3744.

203. Schneider CA, Rasband WS, Eliceiri KW: NIH Image to ImageJ: 25 years of image analysis. Nat Methods 2012, 9(7):671-675.

204. Spearman C: The proof and measurement of association between two things. By C. Spearman, 1904. Am J Psychol 1987, 100(3-4):441-471.

205. kaplan EL, Meier P: Nonparametric Estimation from Incomplete Observations. In., vol. Vol. 53. Journal of the American Statistical Association; 1958: 457-481.

206. Case LD, Kimmick G, Paskett ED, Lohman K, Tucker R: Interpreting measures of treatment effect in cancer clinical trials. Oncologist 2002, 7(3):181-187.

160

207. Sukocheva O, Wang L, Verrier E, Vadas M, Xia P: Restoring Endocrine Response in Breast Cancer Cells by Inhibition of the Sphingosine Kinase-1 Signaling Pathway. Endocrinology 2009, 150(10):4484-4492.

208. Takabe K, Kim R, Allegood J, Mitra P, Ramachandran S, Nagahashi M, Harikumar K, Hait N, Milstien S, Spiegel S: Estradiol Induces Export of Sphingosine 1-Phosphate from Breast Cancer Cells via ABCC1 and ABCG2. Journal of Biological Chemistry 2010, 285(14):10477-10486.

209. Maillot G, Lacroix-Triki M, Pierredon S, Gratadou L, Schmidt S, Benes V, Roche H, Dalenc F, Auboeuf D, Millevoi S et al: Widespread estrogen-dependent repression of micrornas involved in breast tumor cell growth. Cancer research 2009, 69(21):8332-8340.

210. Kasinski A, Slack F: MicroRNAs en route to the clinic: progress in validating and targeting microRNAs for cancer therapy. Nature Reviews Cancer 2011, 11(12):849-864.

211. Furuta M, Kozaki KI, Tanaka S, Arii S, Imoto I, Inazawa J: miR-124 and miR-203 are epigenetically silenced tumor-suppressive microRNAs in hepatocellular carcinoma. Carcinogenesis 2010, 31(5):766-776.

212. Wilting S, van Boerdonk R, Henken F, Meijer C, Diosdado B, Meijer G, le Sage C, Agami R, Snijders P, Steenbergen R: Methylation-mediated silencing and tumour suppressive function of hsa-miR-124 in cervical cancer. Molecular Cancer 2010, 9.

213. Vazquez I, Maicas M, Marcotegui N, Conchillo A, Guruceaga E, Roman-Gomez J, Calasanz M, Agirre X, Prosper F, Odero M: Silencing of hsa-miR-124 by EVI1 in cell lines and patients with acute myeloid leukemia. Proceedings of the National Academy of Sciences of the United States of America 2010, 107(44):E167-E168.

214. Li K, Pang J, Ching A, Wong C, Kong X, Wang Y, Zhou L, Chen Z, Ng H: miR- 124 is frequently down-regulated in medulloblastoma and is a negative regulator of SLC16A1. Human Pathology 2009, 40(9):1234-1243.

215. Silber J, Felix T, Hariono S, Gupta N, James C, Bigner D, Hodgson G: MIR-124 INHIBITS CDK6 AND INDUCES G(1) CELL CYCLE ARREST IN MEDULLOBLASTOMA CELLS. Neuro-Oncology 2008, 10(5):872-873.

216. Hashizume R, Silber J, Hariono S, James D, Hodgson G, Gupta N: EXPRESSION OF miR-124 INHIBITS GROWTH OF CDK6-POSITIVE MEDULLOBLASTOMA CELLS. Neuro-Oncology 2011, 13:I22-I22.

217. Silber J, Lim DA, Petritsch C, Persson AI, Maunakea AK, Yu M, Vandenberg SR, Ginzinger DG, James CD, Costello JF et al: miR-124 and miR-137 inhibit proliferation of glioblastoma multiforme cells and induce differentiation of brain tumor stem cells. BMC Med 2008, 6:14.

218. Lv XB JY, Qing Y, Hu H, Cui X, Lin T, Song E, Yu F.: miR-124 suppresses multiple steps of breast cancer metastasis by targeting a cohort of pro- metastatic genes in vitro. 2011, 12:821-830.

161

219. Shi X, Xue L, White R: MIR-124 TARGETS ANDROGEN RECEPTOR AND INHIBITS TUMORIGENESIS OF PROSTATE CANCER CELLS. Journal of Urology 2012, 187(4):E92-E92.

220. Xia J, Wu Z, Yu C, He W, Zheng H, He Y, Jian W, Chen L, Zhang L, Li W: miR- 124 inhibits cell proliferation in gastric cancer through down-regulation of SPHK1. Journal of Pathology 2012, 227(4):470-480.

221. Schneider C, Maute R, Sumazin P, Brahmachary M, Setty M, Pasqualucci L, Rosolen A, Leslie C, Basso K, Dalla-Favera R: MiR-28 Silencing In Germinal Center-Derived Lymphomas. Blood 2010, 116(21):310-311.

222. Almeida M, Nicoloso M, Zeng L, Ivan C, Spizzo R, Gafa R, Xiao L, Zhang X, Vannini I, Fanini F et al: Strand-Specific miR-28-5p and miR-28-3p Have Distinct Effects in Colorectal Cancer Cells. Gastroenterology 2012, 142(4):886-U355.

223. Lui W, Pourmand N, Patterson B, Fire A: Patterns of known and novel small RNAs in human cervical cancer. Cancer Research 2007:6031-6043.

224. Gottardo F, Liu C, Ferracin M, Calin G, Fassan M, Bassi P, Sevignani C, Byrne D, Negrini M, Pagano F et al: Micro-RNA profiling in kidney and bladder cancers. Urologic Oncology-Seminars and Original Investigations 2007, 25(5):387-392.

225. Malzkorn B, Wolter M, Liesenberg F, Grzendowski M, Stuhler K, Meyer H, Reifenberger G: Identification and Functional Characterization of microRNAs Involved in the Malignant Progression of Gliomas. Brain Pathology 2010, 20(3):539-550.

226. Haller F, von Heydebreck A, Zhang JD, Gunawan B, Langer C, Ramadori G, Wiemann S, Sahin O: Localization- and mutation-dependent microRNA (miRNA) expression signatures in gastrointestinal stromal tumours (GISTs), with a cluster of co-expressed miRNAs located at 14q32.31. J Pathol 2010, 220(1):71-86.

227. Lodes MJ, Caraballo M, Suciu D, Munro S, Kumar A, Anderson B: Detection of cancer with serum miRNAs on an oligonucleotide microarray. PLoS One 2009, 4(7):e6229.

228. Liu W, Zabirnyk O, Wang H, Shiao YH, Nickerson ML, Khalil S, Anderson LM, Perantoni AO, Phang JM: miR-23b targets proline oxidase, a novel tumor suppressor protein in renal cancer. Oncogene 2010, 29(35):4914-4924.

229. Mao ZG, He DS, Zhou J, Yao B, Xiao WW, Chen CH, Zhu YH, Wang HJ: Differential expression of microRNAs in GH-secreting pituitary adenomas. Diagn Pathol 2010, 5:79.

230. Zhang H, Yang JH, Zheng YS, Zhang P, Chen X, Wu J, Xu L, Luo XQ, Ke ZY, Zhou H et al: Genome-wide analysis of small RNA and novel MicroRNA discovery in human acute lymphoblastic leukemia based on extensive sequencing approach. PLoS One 2009, 4(9):e6849.

231. Cortez C: Upregulation of the miR-515 cluster in the MCF7 breast cancer line by epigenetic therapy. 2008, 116(5):437-437(431).

162

232. Scapoli L, Palmieri A, Lo Muzio L, Pezzetti F, Rubini C, Girardi A, Farinella F, Mazzotta M, Carinci F: MicroRNA expression profiling of oral carcinoma identifies new markers of tumor progression. Int J Immunopathol Pharmacol 2010, 23(4):1229-1234.

233. Wei J, Wu X, Peng Y, Shi G, Olca B, Yang X, Daniels G, Osman I, Ouyang J, Hernando E et al: Regulation of HMGA1 Expression by MicroRNA-296 Affects Prostate Cancer Growth and Invasion. Clinical Cancer Research 2011, 17(6):1297-1305.

234. Vaira V, Faversani A, Dohi T, Montorsi M, Augello C, Gatti S, Coggi G, Altieri D, Bosari S: miR-296 regulation of a cell polarity-cell plasticity module controls tumor progression. Oncogene 2012, 31(1):27-38.

235. Wurdinger T, Tannous B, Saydam O, Skog J, Grau S, Soutschek J, Weissleder R, Breakefield X, Krichevsky A: miR-296 Regulates Growth Factor Receptor Overexpression in Angiogenic Endothelial Cells. Cancer Cell 2008:382-393.

236. Hong L, Han Y, Zhang H, Li M, Gong T, Sun L, Wu K, Zhao Q, Fan D: The Prognostic and Chemotherapeutic Value of miR-296 in Esophageal Squamous Cell Carcinoma. Annals of Surgery 2010, 251(6):1056-1063.

237. Cummins J, He Y, Leary R, Pagliarini R, Diaz L, Sjoblom T, Barad O, Bentwich Z, Szafranska A, Labourier E et al: The colorectal microRNAome. Proceedings of the National Academy of Sciences of the United States of America 2006:3687- 3692.

238. Lawrie C, Chi J, Taylor S, Tramonti D, Ballabio E, Palazzo S, Saunders N, Pezzella F, Boultwood J, Wainscoat J et al: Expression of microRNAs in diffuse large B cell lymphoma is associated with immunophenotype, survival and transformation from follicular lymphoma. Journal of Cellular and Molecular Medicine 2009, 13(7):1248-1260.

239. Grady W, Parkin R, Mitchell P, Lee J, Kim Y, Tsuchiya K, Washington M, Paraskeva C, Willson J, Kaz A et al: Epigenetic silencing of the intronic microRNA hsa-miR-342 and its host gene EVL in colorectal cancer. Oncogene 2008:3880-3888.

240. Wang H, Wu J, Meng X, Ying X, Zuo Y, Liu R, Pan Z, Kang T, Huang W: MicroRNA-342 inhibits colorectal cancer cell proliferation and invasion by directly targeting DNA methyltransferase 1. Carcinogenesis 2011, 32(7):1033- 1042.

241. Cittelly D, Das P, Spoelstra N, Edgerton S, Richer J, Thor A, Jones F: Downregulation of miR-342 is associated with tamoxifen resistant breast tumors. Molecular Cancer 2010, 9.

242. Shin K, Pucar A, Kim R, Bae S, Chen W, Kang M, Park N: Identification of senescence-inducing microRNAs in normal human keratinocytes. International Journal of Oncology 2011, 39(5):1205-1211.

243. Romero-Cordoba S, Rodriguez-Cuevas S, Rebollar-Vega R, Quintanar-Jurado V, Maffuz-Aziz A, Jimenez-Sanchez G, Bautista-Pina V, Arellano-Llamas R, Hidalgo-Miranda A: Identification and Pathway Analysis of microRNAs with No Previous Involvement in Breast Cancer. Plos One 2012, 7(3).

163

244. Adams B, Furneaux H, White B: The micro-ribonucleic acid (miRNA) miR-206 targets the human estrogen receptor-alpha (ER alpha) and represses ER alpha messenger RNA and protein expression in breast cancer cell lines. Molecular Endocrinology 2007, 21(5):1132-1147.

245. Wang X, Ling C, Bai Y, Zhao J: MicroRNA-206 Is Associated With Invasion and Metastasis of Lung Cancer. Anat Rec (Hoboken) 2011, 294(1):88-92.

246. Taulli R, Bersani F, Foglizzo V, Linari A, Vigna E, Ladanyi M, Tuschl T, Ponzetto C: The muscle-specific microRNA miR-206 blocks human rhabdomyosarcoma growth in xenotransplanted mice by promoting myogenic differentiation. Journal of Clinical Investigation 2009, 119(8):2366- 2378.

247. Miyachi M, Tsuchiya K, Yoshida H, Yagyu S, Kikuchi K, Misawa A, Iehara T, Hosoi H: Circulating muscle-specific microRNA, miR-206, as a potential diagnostic marker for rhabdomyosarcoma. Biochemical and Biophysical Research Communications 2010, 400(1):89-93.

248. Xia P, Gamble J, Wang L, Pitson S, Moretti P, Wattenberg B, D'Andrea R, Vadas M: An oncogenic role of sphingosine kinase. Current Biology 2000:1527- 1530.

249. de la Roche M, Worm J, Bienz M: The function of BCL9 in Wnt/beta-catenin signaling and colorectal cancer cells. BMC Cancer 2008, 8:199.

250. Wang L, Cao XX, Chen Q, Zhu TF, Zhu HG, Zheng L: DIXDC1 targets p21 and cyclin D1 via PI3K pathway activation to promote colon cancer cell proliferation. Cancer Sci 2009, 100(10):1801-1808.

251. Saitoh T, Katoh M: FRAT1 and FRAT2, clustered in human chromosome 10q24.1 region, are up-regulated in gastric cancer. Int J Oncol 2001, 19(2):311-315.

252. Gupta S, Iljin K, Sara H, Mpindi JP, Mirtti T, Vainio P, Rantala J, Alanen K, Nees M, Kallioniemi O: FZD4 as a mediator of ERG oncogene-induced WNT signaling and epithelial-to-mesenchymal transition in human prostate cancer cells. Cancer Res 2010, 70(17):6735-6745.

253. Balaz P, Plaschke J, Krüger S, Görgens H, Schackert HK: TCF-3, 4 protein expression correlates with beta-catenin expression in MSS and MSI-H colorectal cancer from HNPCC patients but not in sporadic colorectal cancers. Int J Colorectal Dis 2010, 25(8):931-939.

254. Kim S, Dubrovska A, Salamone RJ, Walker JR, Grandinetti KB, Bonamy GM, Orth AP, Elliott J, Porta DG, Garcia-Echeverria C et al: FGFR2 promotes breast tumorigenicity through maintenance of breast tumor-initiating cells. PLoS One 2013, 8(1):e51671.

255. Courtney KD, Corcoran RB, Engelman JA: The PI3K pathway as drug target in human cancer. J Clin Oncol 2010, 28(6):1075-1083.

256. Sukocheva O, Wang L, Albanese N, Pitson S, Vadas M, Xia P: Sphingosine kinase transmits estrogen signaling in human breast cancer cells. Molecular Endocrinology 2003, 17(10):2002-2012.

164

257. Schmidt D, Schwalie P, Ross-Innes C, Hurtado A, Brown G, Carroll J, Flicek P, Odom D: A CTCF-independent role for cohesin in tissue-specific transcription. Genome Research 2010, 20(5):578-588.

258. Buffa FM, Camps C, Winchester L, Snell CE, Gee HE, Sheldon H, Taylor M, Harris AL, Ragoussis J: microRNA-associated progression pathways and potential therapeutic targets identified by integrated mRNA and microRNA expression profiling in breast cancer. Cancer Res 2011, 71(17):5635-5645.

259. Glaeser M, Niederacher D, Djahansouzi S, Hanstein B, Dittrich R, Beckmann M, Fasching P, Ackermann S: Effects of the antiestrogens tamoxifen and raloxifene on the estrogen receptor transactivation machinery. Anticancer Research 2006, 26(1B):735-744.

260. Tong D, Schuster E, Seifert M, Czerwenka K, Leodolter S, Zeillinger R: Expression of estrogen receptor beta isoforms in human breast cancer tissues and cell lines. Breast Cancer Research and Treatment 2002, 71(3):249- 255.

261. Jiang XP, Yang DC, Elliott RL, Head JF: Down-regulation of expression of interleukin-6 and its receptor results in growth inhibition of MCF-7 breast cancer cells. Anticancer Res 2011, 31(9):2899-2906.

262. Simstein R, Burow M, Parker A, Weldon C, Beckman B: Apoptosis, chemoresistance, and breast cancer: insights from the MCF-7 cell model system. Experimental biology and medicine 2003, 228(9):995-1003.

263. Wilkinson S, Paterson HF, Marshall CJ: Cdc42-MRCK and Rho-ROCK signalling cooperate in myosin phosphorylation and cell invasion. Nat Cell Biol 2005, 7(3):255-261.

264. Cain RJ, Ridley AJ: Phosphoinositide 3-kinases in cell migration. Biol Cell 2009, 101(1):13-29.

265. Omholt K, Karsberg S, Platz A, Kanter L, Ringborg U, Hansson J: Screening of N-ras codon 61 mutations in paired primary and metastatic cutaneous melanomas: mutations occur early and persist throughout tumor progression. Clin Cancer Res 2002, 8(11):3468-3474.

266. Nagaraja GM, Othman M, Fox BP, Alsaber R, Pellegrino CM, Zeng Y, Khanna R, Tamburini P, Swaroop A, Kandpal RP: Gene expression signatures and biomarkers of noninvasive and invasive breast cancer cells: comprehensive profiles by representational difference analysis, microarrays and proteomics. Oncogene 2006, 25(16):2328-2338.

267. Helbig G, Christopherson KW, Bhat-Nakshatri P, Kumar S, Kishimoto H, Miller KD, Broxmeyer HE, Nakshatri H: NF-kappaB promotes breast cancer cell migration and metastasis by inducing the expression of the chemokine receptor CXCR4. J Biol Chem 2003, 278(24):21631-21638.

268. Gyorffy B, Lánczky A, Szállási Z: Implementing an online tool for genome- wide validation of survival-associated biomarkers in ovarian-cancer using microarray data from 1287 patients. Endocr Relat Cancer 2012, 19(2):197- 208.

165

269. Pinho FG, Frampton AE, Nunes J, Krell J, Alshaker H, Jacob J, Pellegrino L, Roca-Alonso L, de Giorgio A, Harding V et al: Downregulation of microRNA- 515-5p by the estrogen receptor modulates sphingosine kinase 1 and breast cancer cell proliferation. Cancer Res 2013, 73(19):5936-5948.

270. Altuvia Y, Landgraf P, Lithwick G, Elefant N, Pfeffer S, Aravin A, Brownstein MJ, Tuschl T, Margalit H: Clustering and conservation patterns of human microRNAs. Nucleic Acids Res 2005, 33(8):2697-2706.

271. Lee Y, Jeon K, Lee JT, Kim S, Kim VN: MicroRNA maturation: stepwise processing and subcellular localization. EMBO J 2002, 21(17):4663-4670.

272. Noguer-Dance M, Abu-Amero S, Al-Khtib M, Lefevre A, Coullin P, Moore G, Cavaille J: The primate-specific microRNA gene cluster (C19MC) is imprinted in the placenta. Human Molecular Genetics 2010, 19(18):3566-3582.

273. Flor I, Bullerdiek J: The dark side of a success story: microRNAs of the C19MC cluster in human tumours. J Pathol 2012, 227(3):270-274.

274. Wang F, Xue X, Wei J, An Y, Yao J, Cai H, Wu J, Dai C, Qian Z, Xu Z et al: hsa- miR-520h downregulates ABCG2 in pancreatic cancer cells to inhibit migration, invasion, and side populations. Br J Cancer 2010, 103(4):567-574.

275. Toffanin S, Hoshida Y, Lachenmayer A, Villanueva A, Cabellos L, Minguez B, Savic R, Ward SC, Thung S, Chiang DY et al: MicroRNA-based classification of hepatocellular carcinoma and oncogenic role of miR-517a. Gastroenterology 2011, 140(5):1618-1628.e1616.

276. Kim CJ, Park NH, Chung YH: Roads towards a new tailored therapy for hepatocellular carcinoma: diagnostic, therapeutic and prognostic implications of microRNAs. Liver Int 2012, 32(5):695-697.

277. Baskerville S, Bartel DP: Microarray profiling of microRNAs reveals frequent coexpression with neighboring miRNAs and host genes. RNA 2005, 11(3):241-247.

278. Yamagata K, Fujiyama S, Ito S, Ueda T, Murata T, Naitou M, Takeyama K, Minami Y, O'Malley BW, Kato S: Maturation of microRNA is hormonally regulated by a nuclear receptor. Mol Cell 2009, 36(2):340-347.

279. Paris O, Ferraro L, Grober OM, Ravo M, De Filippo MR, Giurato G, Nassa G, Tarallo R, Cantarella C, Rizzo F et al: Direct regulation of microRNA biogenesis and expression by estrogen receptor beta in hormone- responsive breast cancer. Oncogene 2012, 31(38):4196-4206.

280. Jacob J, Stebbing J, Frampton A, Pellegrino L, Castellano L: MicroRNA and RNA binding proteins. In. Advances in breast cancer; 2010: 14-17.

281. Bentwich I, Avniel A, Karov Y, Aharonov R, Gilad S, Barad O, Barzilai A, Einat P, Einav U, Meiri E et al: Identification of hundreds of conserved and nonconserved human microRNAs. Nat Genet 2005, 37(7):766-770.

282. Baek D, Villén J, Shin C, Camargo FD, Gygi SP, Bartel DP: The impact of microRNAs on protein output. Nature 2008, 455(7209):64-71.

166

283. Lee I, Ajay SS, Yook JI, Kim HS, Hong SH, Kim NH, Dhanasekaran SM, Chinnaiyan AM, Athey BD: New class of microRNA targets containing simultaneous 5'-UTR and 3'-UTR interaction sites. Genome Res 2009, 19(7):1175-1183.

284. Rehmsmeier M, Steffen P, Hochsmann M, Giegerich R: Fast and effective prediction of microRNA/target duplexes. RNA 2004, 10(10):1507-1517.

285. Ring A, Dowsett M: Mechanisms of tamoxifen resistance. Endocr Relat Cancer 2004, 11(4):643-658.

167

Appendices

168

RNA-seq data:

Supplementary Table. Transcripts downregulated in MCF7 and MDA-MB-231 cells upon miR-515-5p overexpression.

MCF7 MDA-MB-231 Transcripts Transcripts downregulated by downregulated by miR- Transcripts Transcripts downregulated miR-515-5p that are 515-5p that are predicted downregulated by miR- by miR-515-5p predicted to interact to interact with miR-515- 515-5p with miR-515-5p 5p (Target scan) (Target scan) FAM60A FGFR2 C3 F2RL3 LOC100616668 HMGN2 C19orf29-AS1 BMF ZNF514 ZNF84 F2RL3 USP6NL PROCA1 ZNF140 ECHDC2 TNFRSF9 CRNDE CA4 ASB2 ZNF681 ITGA2B ZWINT WNT9A FZD4 SFXN2 POLR2D KLRC2,KLRC3 WDR72 C11orf80 IL6R PIEZO2 ATM NPIP LRRC10B BMF FLCN MSS51 NYNRIN ZNF846 ZNF701 UPK1A DYM USP6NL LONRF2 UQCRC2 TRADD ASB3 NR2F2 VPS37B BEND7 TNFRSF9 SYT2 ADAMTS17 CENPA ZNF841 TMEM229B ANKRD35 CDH1 ZBTB25 IL1R1 TM6SF1 ATG13 C17orf57 ZFHX3 MFGE8 NUDT15 STON2 DUSP8 DYM ZNF18 ZNF681 LRP5 C1orf63 GLTP CDC42BPB KIAA1598 RASA3 ZNF266 TRAPPC12 RCSD1 BAIAP2 ZFYVE21 C10orf10 ATXN7L3B GALNS FLVCR2 LOC100652768 DYNC1LI2 PMCH RAVER2 FZD4 PLEKHH1 GNG7 CAMKK1 PIAS3 ZNF506 CHTOP C18orf1 POLG2 CD46 MAF LTBP1 WDR72 HCAR1 ECHDC2 WIPI1 CSF3 DNAL1 ITGB6 PPT1 C16orf72 ELF3 INCA1 ADAM8 ATM MR1 ATPIF1 ZNF714 NEBL ID3 LOC100188947 CBX1 RAP1GAP2 ITPKB PITPNM3 GINS2 RILPL1 ZNF441 NBPF6 SLK FLCN HES2 HES2 TXNDC12 ZNF701 ATG13 ALDH3A1 ZNF468 MPPE1 STX2

169

ZNF211 EIF2B1 LONRF2 PARP11 KRT23 SIRT2 RFXAP TMOD2 PIK3CD PLA2G4F ABCC3 ZNF254 TEX14 CDC42EP4 NR2F2 PTPN14 ZNF429 GPT2 THBS1 C19orf40 C10orf26 CCDC142 C14orf45 ZNF91 PCBP1-AS1 PDCD7 ATP2B4 ZNF678 FZD4 STARD13 LOC399744 ZNF43 DIXDC1 SPATS2 SYT2 ZMIZ1 USP6NL C10orf81 GBP4 IRF2BPL ZNF83 RPUSD4 TMEM229B DIXDC1 CTTN CC2D1B HAS3 PHKB CYP19A1,GLDN C15orf57 IL1R1 TET3 HP1BP3 SSH3 ZNF772 PRKG1 CYP26B1 CYP19A1 ZFHX3 IGFBP4 DMAP1 NIT1 CDC16 BCAS3 C15orf40 CENPBD1 RHOJ SUFU TM2D1 ASCL1 GBP2 IMPACT LOC728537 ZNF506 RPSAP52 TCF4 SLC26A10 PRADC1 DUSP8 SLC37A2 PCDH9 METTL9 C11orf71 ANKFY1 LMF1 PSKH1 ZNF583 ERMAP MCF2L FCGR1A LMBR1L PPP1R3C HDAC7 C11orf49 LOC100128398 EGLN3 PDE4DIP MFSD11 DMBT1 CSNK1G1 SYDE2 ANKRD11 ANKRD10 ZNF845 KCNJ8 ZNF814 IL1B DYM RAB3GAP1 MCF2L MKNK1 ZNF557 OLFML3 ZNF813 LRP5 MFSD11 SMUG1 ZNF83 NAGK BACE1 NYNRIN OPN3 RNASEL ZNF160 ZNF506 ZCCHC24 ZNF85 ERCC6 EXOC6B MED29 KIAA1598 RPUSD4 VPS35 BTG2 TAPBPL ARHGAP32 C11orf49 PIK3C2B ERVK13-1 ZNF84 IL6R TIMM50 RCSD1 ETV6 SLC7A8 EMP2 ATXN7L3B SVIL KIAA0090 ZNF443 DYNC1LI2 PGM2L1 KLC3 DNALI1 PLEKHH1 SSTR2 ZBTB48 SETBP1 ZNF525 CYB561D1 NEXN EFCAB2 COL13A1 ODF2L ATRNL1 C12orf65 JUP SLC35E2 BTG1,LOC256021 DQX1 AGRN C11orf95 FANCA SH2B1 TNFSF18 ZNF708 LOC284379 MRPS11 TPT1-AS1 TXNIP

170

TRAPPC12 LDLRAD3 CD58 ZNF28 CRLF1 SLC38A7 ZNF836 RASSF4 PIAS4 B3GNT9 C12orf26 HDHD2 PPP2R1B ZNF638 ZNF506 ZNF592 ZNF69 ZFYVE27 MAP3K8 AKAP6 METTL7A ZFP36L1 LOC678655 ARHGEF40 C10orf81 SMARCE1 SAV1 ATP11A ENDOV GLS2 ZSWIM4 PDE4DIP PPIE PRKCE LOC100129387 INSR A1BG,ZNF497 MUM1 PRRT2 LTBP1 SYTL2 TAF13 CD46 ADAL LRTOMT TMOD3 LZTS2 SH3PXD2A SPG7 CAB39L ZNF616 TXNDC12 ANKRD11 CARKD SLC25A29 ZBTB8A MBD3 PPP2R1B HCAR1 IER5 YJEFN3 ZNF32 GAS2L3 ATPBD4 DIO3 LAT DNAL1 APH1B GPR132 COG4 ELF3 EFNB2 KLK11 TNFRSF11A USP54 HP1BP3 CDH1 SLC35E2 PPFIA4 GPC6 FARP1 VSIG10 LYPD1 WSB1 ZNF570 TMEM101 MFAP2 TMC8 GNPTG SLC39A9 MR1 ANKRD11 ZNF404 PCDH9 TRIM54 PCDH9 TRAF4 ARHGAP30 ABCB9 F11R LOC100506123 ZNF263 SIDT2 ZNF420 MORN2 NR2F2 ID3 SMARCD1 SERTAD4 WBP4 ITPKB SLC38A7 DOT1L SVIL ZNF441 CYTH2 LETMD1 FBRSL1 TNFAIP2 IGSF3 GUCY1B2 USP6NL HES2 ZCCHC24 ANKRD13D TRIM45 ATG13 ZNF268 RINL MAF STX2 SP2 DUSP12 SLC1A5 SLC44A5 SGPL1 CLCN7 C1orf115 DHRS7 EEF2K LRP1B SMARCD1 PARP11 C15orf52 ATP5S NACC1 TMOD2 ZNF626 SLK PRX IL32 FREM2 ANO9 FGFBP3 FARP1 EXD2 PSTK EGLN3 MPP3 SLC16A7 ZNF18 KIAA0556 LHPP MLXIP SLC35E2 TMEM91 ZNF254 CAMKK1 BRF1 KDM5B MBD5 FAM63B ARHGAP33 FRAT2 PTPN14 KDM5A PDXDC2P ZNF134 LOC100288069 PRICKLE1

171

CASC5 C17orf103 C19orf40 NFIA MIRLET7I PLEKHM2 OSBPL5 EMR2 ZNF263 LRRC20 ITGA10 NPAS2 SLC3A1 HNRNPA1 ZNF91 PTPN21 C14orf45 SLC35E3 ZNF678 RAVER2 ERO1L HP1BP3 KRT86 FUT4 OR7E14P PPP1R9B NR4A2 MARVELD1 SERPINF1 WSB1 PLXDC2 GALNT1 RNF166 MALL ZNF43 CASP7 CHID1 C2orf56 DAAM1 USP22 NASP NEXN NMB TMEM8A TMEM80 BCL9 LOC728752 ZNF470 MAPKAPK5 SLC14A1 ZMIZ1 RSBN1 PTBP2 TCF7L1 IRF2BPL CACNA2D4 GOLGA8A NFATC4 CSF1 ZNF544 AMN1 GPR137C GTF2A1 ZNF527 GBAP1 ITSN2 MPHOSPH9 PSKH1 MAST1 RINL PIGN SYNRG LOC401010 CROCC ANXA9 FAM46C UPF3A SLC26A11 ZC3H12A ZNF430 NCF2 PTP4A2 FCGBP RC3H1 GCLM CEP164 PLEKHB1 STX12 FGGY HES2 DIXDC1 ZNF10 SYNC N4BP2L1 PHKB HTATIP2 MOB2 DDX51 TET3 MYSM1 RGS6 MARK4 PRKG1 ZNF776 C1orf86 CACNG4 CPNE7 SUV420H1 LOC728730 EFNA1 TCF25 MBOAT2 ZNF177,ZNF559,ZNF559- SPTB ZFYVE9 RASSF3 ZNF177 SIRT2 TMEM120B LRRC28 ADAMTS15 SLC35E3 PDE4DIP IGFBP4 LCOR ATG13 DOK4 BCAS3 MTF1 GPR137C MAN1C1 SUFU ERC1 SIPA1L1 PTGER3 TCTN2 KYNU ZNF84 ERMAP FAM211A MAP4K4 DLEU2,MIR3613 BRSK2 IMPACT HCFC2 MNS1 HEXA TCF4 SUMO2 SRSF4 MEGF8 SLC37A2 TANC2 BNIPL CHST3 ANKFY1 NFKBIA TCHH RAD23A SESN3 ZNF83 LOC100507217 IGHMBP2 ZNF426 CYTH1 RHBG SUFU GABPB2 PTPRE SLC12A4 PGPEP1 ERMAP RBL2 PRPF3 ENSA PPP1R3C DR1

172

RGL3 TRAF4 EGLN3 ERRFI1 BTBD11 SESN2 CSNK1G1 POLI PKP1 FLCN ZNF845 FLRT2 C18orf1 POLE DCAKD SOX9 RAB11FIP5 ABCC9 ZNF383 ZNF33B ASCL1 GPR132 SMAD6 KHNYN ERP27 SMAD9 DYM ZNF207 DQX1 ZNF720 IL18BP DBT IFITM10 PLA2G15 C14orf159 RIMKLA ANKS1B DIXDC1 CSAD ZSCAN29 BEND7 PGR KBTBD4 CMKLR1 CCAR1 NEDD4L ACOT4 POLR2D FBXO3 PITPNM3 ZNF557 PACS1 CALCOCO1 IRGQ MFSD11 SEMA7A CCDC85A IFITM10 BAIAP3 RAB2B TCP11L2 FAM109A DPP8 RUFY2 NAGK ESRRA UPF3A GXYLT1 LPAR5 B3GALT6 BACE1 TAOK1 CD82 PACS1 ZNF160 ATXN1L LOC100131320 ARHGEF17 SYMPK FAM160B1 LOC284408 DCUN1D3 ZNF276 NRAS ANKDD1A AMFR NEAT1 FBXL20 ARHGAP30 ARHGAP35 FAM70B NAV2 ODC1 JUND SLFN5 BMPR1A C17orf103 ULK3 RCAN3 PIK3C2B ZNF554 MBD3 ERCC6 MON2 ZNF490 SRCIN1 LACC1 ABTB2 MIR17HG WDR72 MIR4746 FOXN3 LOC284454,MIR24-2 PDDC1 RPUSD4 RAB11FIP2 SLC26A11 FZD4 RORA CHST15 ZNF784 ZNF652 TSPAN31 DSC2 MIR210HG KCTD15 ARHGAP32 MAGI3 FLJ45983 CNN2 MIR100HG,MIRLET7A2 CCDC3 SEMA6C GOLGA3 ZNF84 ZBTB40 MRPS31 CYTH2 ZNF25 PLEKHB2 HAS3 PLEKHM1 ETV6 ZNF134 MIR365A VASH1 SVIL XPR1 SAMD11 METTL7A ACTA2 OTUD1 ZNF444 ZNF490 BIRC3 ZFP90 ZCCHC24 C11orf95 LRRC37BP1 ZNF397 C2orf27A LRRC27 PGM2L1 ANO6 THRA BMF SSTR2 TMX1 PWWP2B SLC43A2 SF1 CD34 TTF2 FGFR2 CYB561D1 ALG10B PAX9,SLC25A21 HMGN2 PTGFRN OPN3

173

LOC100130987 ZNF84 ODF2L CBX1 CPAMD8 ZNF140 SLC35E2 FMNL3 TMEM99 CA4 C11orf95 ZNF140 KCNJ3 ZWINT RABGAP1L YTHDF2 EVL POLR2D ZNF227 NT5DC3 LTBP2 IL6R ZNF708 MEF2D TNKS1BP1 LRRC10B KCNS3 IRS2 MOK NYNRIN ZNF702P FBXO31 TARBP1 DYM TXNIP OTUB2 CTSF TRADD POLR3GL ABCC9 TCF7L1 BEND7 ZNF28 CENPA C16orf72 CENPA RASSF4 TCF12 IGSF9 CDH1 ZNF569 TRADD PLEKHA7 ATG13 MIR210HG VDR KIFC3 NUDT15 GJB2 ZNF37A PTPN6 ZNF18 SIN3B PUM1 SLC18A2 GLTP NRP1 MIB1 NHLRC2 ZNF266 CUBN DPYD FBRSL1 ZFYVE21 HDHD2 ARID1A ING4 FLVCR2 SIPA1L2 YEATS4 CENPBD1 RAVER2 ANKRD12 ALG9 TRPV3 CAMKK1 ZNF592 ATP7B PTGER3 C18orf1 FBXL19-AS1 RIMKLB LOC100131060 LTBP1 TARSL2 WDR59 PIGL WIPI1 AKAP6 NUDT15 BANP PPT1 ARHGEF40 SMAD3 TMEM86B ADAM8 ATP11A IRGQ BRIP1 ZNF714 PDE4DIP ZNF611 PPP1R13L CBX1 LOC339535 PPP1R12B HERC2P2 GINS2 LOC728730 CEP128 ZNF468 SLK SLAIN1 THEM4 TNFRSF11A TXNDC12 KIF3C QSER1 WDR72 ZNF468 GBP3 SGMS1 CBX8 EIF2B1 LOC440434 ITSN2 ZNF714 SIRT2 DOT1L KLF6 ETNK2 PLA2G4F INSR CDC42BPA NUPL1 CDC42EP4 SYT1 ILDR2 SRR GPT2 FAM63A C18orf54 TBC1D17 CCDC142 PYGO1 TMTC2 MPP3 PDCD7 TUBA1A KIF20B FLJ37453 STARD13 ZNF700 PHACTR4 CCDC63 SPATS2 LTBP1 ATG14 ENSA C10orf81 ADAL KDM5B SGSM2 RPUSD4 SH3PXD2A FAM76A RAB11FIP3 CC2D1B TXNDC12 MBP

174

XPOT C15orf57 ZNF624 PLA2G15 EVI5L SSH3 ZBTB8A SNX19 ADPRHL1 CYP19A1 CSNK1A1L GMFB KIAA0895L NIT1 IER5 ITPRIPL2 MYCL1 CENPBD1 ZNF766 CBX5 SLC16A4 ASCL1 COBLL1 ABLIM1 LOC219347 ZNF506 ATPBD4 C1orf144 SMARCD1 PRADC1 PPP1R3E PANX1 WSB1 METTL9 APH1B ZNF675 MAP2K7 PSKH1 AMIGO1 MAP3K2 PSMD7 FCGR1A EIF2C4 HN1L MCOLN1 C11orf49 GOLGA6L5,UBE2Q2P1 CDC42SE1 NANOS1 MFSD11 EFNB2 CHSY1 SCMH1 ANKRD11 ZNF320,ZNF468 STXBP4 ZNF20,ZNF625,ZNF625- ZNF814 HP1BP3 ZNF697 ZNF20 LTBP1 MCF2L ELMOD3 PLXNC1 FGFR2 ZNF813 GPC6 PLEKHA1 WDR81 ZNF83 WSB1 DBNDD1 C1orf226,NOS1AP OPN3 ZNF669 CENPBD1 C11orf95 ZCCHC24 TMC8 MED29 UPK2 MED29 ANKRD11 F2RL3 B3GNT9 BTG2 PCDH9 BMF MGAT5B PIK3C2B KCNN4 USP6NL CLASRP TIMM50 F11R TNFRSF9 MYO15B EMP2 RIC8B ZNF681 PHLDB1 ZNF443 ICAM1 FZD4 THBS3 DNALI1 ZNF420 WDR72 PHTF1 SETBP1 SMARCD1 ATM FBF1 EFCAB2 CIITA FLCN BTG2 C12orf65 LOC100499405 ZNF701 PRX DQX1 ZNF606 LONRF2 ADM SH2B1 TSTD1 NR2F2 SLC43A2 MRPS11 ZNF17,ZNF548 SYT2 HNMT LDLRAD3 SLC38A7 TMEM229B MAN2B1 SLC38A7 TRIB2 IL1R1 HOOK2 B3GNT9 ZNF790 ZFHX3 ERMAP ZNF638 FAM214A DUSP8 GABARAPL1 ZFYVE27 ATP10A LRP5 TMEM91 ZFP36L1 CYTH2 KIAA1598 GBA SMARCE1 ZNF642 RCSD1 TRIM45 GLS2 IGSF3 ATXN7L3B ITSN2 PRKCE NLRP1 DYNC1LI2 LOC100499489 MUM1 ZCCHC24 PLEKHH1 ZNF234 TAF13 KCTD12 ZNF506

175

EFCAB2 TMOD3 AGAP2 CD46 UCP2 CAB39L MIR4712 HCAR1 NXPH4 CARKD TCP11L1 DNAL1 PROX2 PPP2R1B ZNF268 ELF3 PALM ZNF32 C2orf27A MR1 ELF5 LAT SP2 ID3 LOC100288778 COG4 GBAP1 ITPKB CHMP3,RNF103,RNF103- CARD14 TNFRSF11A ZNF441 CHMP3 SPTB SLC35E2 AMN1 HES2 ABCC9 VSIG10 THRA ATG13 TMOD3 TMEM101 SERPINA1 STX2 LOC100506649 SLC39A9 TRIM22 PARP11 FAM213A PCDH9 NUPL1 TMOD2 DYRK1B ARHGAP30 LOC646214 ZNF254 MST1P2 ZNF263 H6PD PTPN14 ZNF670,ZNF670- NR2F2 C1R C19orf40 ZNF695,ZNF695 TSSK3 WBP4 ZNF613 ZNF91 CEMP1 SVIL WDTC1 ZNF678 ZNF814 FBRSL1 CD82 ZNF43 TSPAN9 USP6NL MUC1 ZMIZ1 ZNF226 TRIM45 ZNF585B IRF2BPL POLG2 MAF SGPL1 DIXDC1 PDCD7 SLC1A5 METTL21B PHKB FANK1 C1orf115 EEF2K TET3 C1orf115 SMARCD1 KIAA0895L PRKG1 KIAA0430 NACC1 C15orf52 IGFBP4 SMAD9 PRX RAB37 BCAS3 TATDN3 FGFBP3 ZNF626 SUFU RPL13AP20 EGLN3 FREM2 IMPACT TIMM50 KIAA0556 ARHGAP5 TCF4 ST3GAL2 TMEM91 LOC388692 SLC37A2 HMGN2 KDM5B BICC1 ANKFY1 AGBL5 FRAT2 EXD2 ERMAP WDR83 ZNF134 RBM19 PPP1R3C ZNF177,ZNF559,ZNF559- ASTL C17orf103 EGLN3 ZNF177 FBXL12 PLEKHM2 GLCE CSNK1G1 SYNE2 LRRC20 HIF1AN ZNF845 SH2B1 HNRNPA1 SLC16A7 DYM NBPF4 SLC35E3 MLXIP ZNF557 C14orf79 HP1BP3 ZBTB3 MFSD11 MMP19 PPP1R9B CAMKK1 BACE1 LOC440461 WSB1 FAM63B ZNF160 LRRC28 MALL NUDT4,NUDT4P1 ERCC6 LINC00085 C2orf56 HIST2H2BE RPUSD4

176

SUFU NEXN KDM5A ARHGAP32 OPN3 BCL9 KLF13 ZNF84 ZNF778 SLC14A1 CHFR ETV6 DCUN1D3 TCF7L1 DCLRE1C SVIL PPP6R3 NFATC4 LOC338588 PGM2L1 CCDC114 GPR137C PRICKLE1 SSTR2 INPP5K ITSN2 TGFB3 CYB561D1 TCF25 RINL ZFP28 ODF2L CACNG4 CROCC ZNF844 SLC35E2 PRKD1 SLC26A11 ACCS C11orf95 ZNF749 PTP4A2 C15orf38-AP3S2 ZNF708 ZNF181,ZNF302 CEP164 EPAS1 TXNIP CA4 HES2 NFIA ZNF28 MAP3K10 N4BP2L1 EMR2 RASSF4 LINC00294 DDX51 PTPRJ HDHD2 LOC100506668 MARK4 NPAS2 ZNF592 C11orf35 CACNG4 TTC5 AKAP6 FUT4,PIWIL4 EFNA1 TIMP2 ARHGEF40 HCN2 SPTB NS3BP ATP11A NBPF1 MARK4

ZNF682 TMEM120B COL4A1 PDE4DIP LOC284385 PDE4DIP ZNF546 INSR ZNF709 DOK4 ZRANB3 LTBP1 LOC729013 MAN1C1 ANKRD36 ADAL MMP25 PTGER3 PTPN21 SH3PXD2A CCDC18 ERMAP RAVER2 TXNDC12 GRB7 BRSK2 FUT4 ZBTB8A MIB2 HEXA MARVELD1 IER5 SERPINC1 MEGF8 TOM1L1 ATPBD4 PLEKHM1 CHST3 GALNT1 APH1B PIGM RAD23A NBEA EFNB2 ZNF616 IGHMBP2 ABCA10,ABCA5 HP1BP3 DPF1 SUFU USP32P2 GPC6 CRHR1,MGC57346 PGPEP1 CASP7 WSB1 LRRC20 ENSA GNPAT TMC8 MBTPS1 TRAF4 APC2 ANKRD11 DAP3 SESN2 USP22 PCDH9 FAM98C FLCN TMEM8A F11R TMEM120B POLE TMEM87A ZNF420 TAF13 ABCC9 ZNF470 SMARCD1 MIR1287,PYROXD2 GPR132 RSBN1 SLC38A7 LAMB3,MIR4260 SMAD9 CACNA2D4 CYTH2 CNIH2 ZNF720 ZNF137P IGSF3 THUMPD2 PLA2G15 ZNF544 ZCCHC24 FLJ45445 DIXDC1 NELL2 ZNF268

177

PGR PGR TTC32 SP2 VAV3 NEDD4L ZNF527 SGPL1 PITPNC1 PITPNM3 PSKH1 EEF2K LOC100506233 IRGQ ZNF643 C15orf52 AMZ2P1 IFITM10 PLEKHH2 ZNF626 LIN52 FAM109A SYNRG FREM2 BNIP2 ESRRA LIG3 EXD2 SPSB2 B3GALT6 E2F7 SLC16A7 ATXN7L2 PACS1 LTBP2 MLXIP MED16 ARHGEF17 KIAA1704 CAMKK1 NCAPD3 DCUN1D3 MEGF6 FAM63B SLC25A42 AMFR ZNF71 KDM5A FBXL16 ARHGAP35 SOS2 PRICKLE1 MAST2 JUND FAM46C NFIA NUDT13 ULK3 ZNF430 EMR2 COMMD9 MBD3 ZNF69 NPAS2 NR2F2 SRCIN1 RC3H1 PTPN21 SLC14A1 WDR72 TBC1D8 RAVER2 ZNF720 PDDC1 FLJ39051 FUT4 SBF2 FZD4 MARK4 MARVELD1 CCNL2 ZNF652 SLC35F5 GALNT1 LOC146880,MIR4315- KCTD15 STX12 CASP7 1,PLEKHM1P HSPA2 CNN2 ZNF10 USP22 ZNF790 GOLGA3 MEIS3P1 TMEM8A RPUSD4 CYTH2 IKBKE ZNF470 SPNS1 PLEKHM1 LOC339290 RSBN1 CPEB3 VASH1 HTATIP2 CACNA2D4 CMTM3 METTL7A MYSM1 ZNF544 IRGQ ZNF490 C16orf45 ZNF527 AS3MT,C10orf32,C10orf32- C11orf95 LOC100506233 PSKH1 AS3MT ZNF274 LRRC27 ZNF431,ZNF714 SYNRG NFATC4 BMF ZNF776 FAM46C SLC5A2 SLC43A2 GAS6 ZNF430 TRIM8 FGFR2 SUV420H1 RC3H1 FAM171A1 HMGN2 MBOAT2 STX12 MTHFSD ZNF84 RASSF3 ZNF10 GJB3 ZNF140 ITGB1 HTATIP2 MPZL3 CA4 ZFYVE26 MYSM1 ZNF134 ZWINT ADAMTS15 ZNF776 NAGPA POLR2D GABARAPL2 SUV420H1 PSKH1 IL6R LCOR MBOAT2 PIDD LRRC10B ERO1L RASSF3 HNRNPA1 NYNRIN PPAP2B ADAMTS15 PLA2G15 DYM MTF1 LCOR

178

PRKCE TRADD ERC1 MTF1 FOXRED1 BEND7 KYNU ERC1 NCKAP5L CENPA PAX8 KYNU PLAC2 CDH1 MAP4K4 MAP4K4 PTPRCAP ATG13 ZFP112 HCFC2 ARAP1 NUDT15 INF2 SUMO2 ACSF2 ZNF18 PAFAH2 TANC2 PLD3 GLTP FNBP1L NFKBIA KLC2 ZNF266 HCFC2 ZNF83 PLA2G4F ZFYVE21 SUMO2 CYTH1 MEF2B,MEF2BNB,MEF2BNB- FLVCR2 LRTOMT PTPRE MEF2B ACSM3 RAVER2 NGFR RBL2 RPAIN CAMKK1 MEX3C DR1 ZNF710 C18orf1 TANC2 ERRFI1 ANAPC5 LTBP1 PER3 POLI AKT1 WIPI1 ATG2B FLRT2 TRIM11 PPT1 ZNF223 SOX9 C16orf7 ADAM8 NFKBIA ZNF33B MEX3D ZNF714 ZNF83 KHNYN RNF214 CBX1 CYTH1 ZNF207 KIF3C GINS2 PTPRE DBT AKR1C3 SLK BAZ2A RIMKLA PRRG2 TXNDC12 EIF2C3 ZSCAN29 CHST3 ZNF468 MMP14 CMKLR1 OSGEP EIF2B1 RBL2 POLR2D RBP7 SIRT2 SNN PACS1 LOC100507577 PLA2G4F DR1 SEMA7A ZNF606 CDC42EP4 ERRFI1 RAB2B ELMOD3 GPT2 ZNF567 RUFY2 FEZ1 CCDC142 POLI GXYLT1 ALDH3B2 PDCD7 FLRT2 TAOK1 ZNF253 STARD13 ZNF45 ATXN1L ITGB4 SPATS2 MAML2 FAM160B1 TMEM101 C10orf81 SOX9 NRAS ZNF287 RPUSD4 C17orf63 FBXL20 SYT8 CC2D1B STYX NAV2 NEK8 C15orf57 ZNF33B BMPR1A ZNF565 SSH3 KHNYN PIK3C2B LOC388152 CYP19A1 CLSTN3 MON2 TXNDC12 NIT1 FLG ABTB2 VPS4A CENPBD1 ATRNL1 FOXN3 KCNC4 ASCL1 ING4 RAB11FIP2 GOLGA3 ZNF506 ZNF207 CHST15 PBXIP1 PRADC1 APPBP2 DSC2

179

NXPH3 METTL9 RBBP4 MAGI3 C1orf159 PSKH1 DBT CCDC3 KCTD13 FCGR1A RIMKLA ZBTB40 MUM1 C11orf49 ZSCAN29 PLEKHB2 SVIL MFSD11 KIAA1383 ZNF134 LRRC10B ANKRD11 RNF207 XPR1 ELK3 ZNF814 CMKLR1 OTUD1 TSGA10 MCF2L BTG1,LOC256021 ZFP90 VEGFB ZNF813 KRCC1 ZNF397 CDKN2C ZNF83 ZNF347 ANO6 PGPEP1 OPN3 BABAM1 TMX1 YPEL3 ZCCHC24 POLR2D CD34 SETBP1 MED29 VPS36 ALG10B TYRO3 BTG2 PACS1 OPN3 PTRHD1 PIK3C2B ORAI3 CBX1 DNAJC22 TIMM50 SEMA7A FMNL3 CWC25 EMP2 UBE3B ZNF140 GPX2 ZNF443 ETV3 YTHDF2 BCL9 DNALI1 ITFG1 NT5DC3 LRRC27 SETBP1 C14orf135 MEF2D HEXA EFCAB2 RAB2B IRS2 FAM101B C12orf65 RASAL2 FBXO31 C15orf33 DQX1 PRKACB OTUB2 LOC400027 SH2B1 RUFY2 ABCC9 PIK3C2B MRPS11 INSIG2 CENPA NPHP4 LDLRAD3 TBC1D3 TCF12 ZNF813 SLC38A7 GXYLT1 TRADD CYP2R1 B3GNT9 TAOK1 VDR MAN1C1 ZNF638 ZNF432,ZNF614 ZNF37A C19orf57 ZFYVE27 ATXN1L PUM1 KIF1C ZFP36L1 EML5 MIB1 MYO9A SMARCE1 CCDC57 DPYD ABCD4 GLS2 MEX3A ARID1A ZNF285 PRKCE ZNF211 YEATS4 REXO1 MUM1 FAM160B1 ALG9 VAX2 TAF13 NRAS ATP7B LOC100129361 TMOD3 CCDC144B RIMKLB LOC254128 CAB39L BCL2L11 WDR59 F11R,TSTD1 CARKD FBXL20 NUDT15 C1orf53 PPP2R1B NAV2 SMAD3 ZNF32 ZNF32 DLG5 IRGQ ATF7IP2 LAT PACS2 ZNF611 KRT16 COG4 MIR3916 PPP1R12B NRAS PCDH9 LARP6 TNFRSF11A BMPR1A CEP128

180

SLC38A7 SLC35E2 PIK3C2B THEM4 SRCIN1 VSIG10 MON2 QSER1 PPP2R5E TMEM101 ABR SGMS1 GLS2 SLC39A9 ABTB2 ITSN2 VMP1 NRAS MIR1287,PYROXD2 KLF6 LYPD3 ARHGAP30 FOXN3 CDC42BPA MEX3A ZNF263 RAB11FIP2 ILDR2 ZNF77 NR2F2 ARHGAP21 C18orf54 BRSK2 WBP4 ZCCHC14 TMTC2 NEDD4L SVIL CHST15 KIF20B ADAM8 FBRSL1 PLAU PHACTR4 ARG2 USP6NL MAPRE3 ATG14 TTC9C TRIM45 ZNF264 KDM5B CALHM2 MAF SNRPD1 FAM76A FLJ22184 SLC1A5 DSC2 MBP TLCD2 C1orf115 BAZ2B PLA2G15 IGHMBP2 SMARCD1 NRXN3 SNX19 TMCO4 NACC1 MAGI3 GMFB LIN37 PRX ZNF14 ITPRIPL2 CD58 FGFBP3 ZNF791 CBX5 GLTP EGLN3 SLC18A2 ABLIM1 PLEKHM2 KIAA0556 CCDC3 C1orf144 DUSP2 TMEM91 DTNA PANX1 CD163L1 KDM5B ZBTB40 ZNF675 OBSCN FRAT2 S100A5 MAP3K2 MUCL1 ZNF134 ZNF260 HN1L FAM108A1 C17orf103 ANKRD20A9P CDC42SE1 PQLC3 PLEKHM2 R3HDM2 CHSY1 C14orf93 LRRC20 PLEKHB2 STXBP4 LINC00482 HNRNPA1 ZNF134 ZNF697 LOC127841,PLEKHA6 SLC35E3 XPR1 PLXNC1 CCDC57 HP1BP3 OTUD1 PLEKHA1 TAF1B PPP1R9B HIVEP3 DBNDD1 FHOD3 WSB1 ZFP90 CENPBD1 POLR2D MALL TIRAP MED29 CMTM1 C2orf56 FAM213A ZNF443 NEXN ZNF605 SLC4A5 BCL9 ZNF397 LOC100506314 SLC14A1 LOC283624 KRT8 TCF7L1 SPRED1 CSNK2A2 NFATC4 RNF138 RARG GPR137C AGL HEXIM2 ITSN2 ANO6 CENPT RINL FCHSD2 PAQR6 CROCC DOCK1

181

FOS SLC26A11 TMX1 ARHGEF19 PTP4A2 CD34 GLTPD2 CEP164 ALG10B PARP6 HES2 OPN3 NUMA1 N4BP2L1 ARGLU1 RIN1 DDX51 CBX1 DCAF6 MARK4 SLC6A9 ARID3A CACNG4 FMNL3 PPHLN1 EFNA1 ZNF140 TUBB8 SPTB YTHDF2 SOX13 TMEM120B TECPR2 KIF7 PDE4DIP NT5DC3 ZNF587,ZNF814 DOK4 APOLD1 ZNF585B MAN1C1 MEF2D NIT1 PTGER3 ZNF738 PAK1 ERMAP IRS2 PACS1 BRSK2 FAM18B1 EIF3F HEXA FBXO31 LOC729513 MEGF8 DNHD1 NBR2 CHST3 VAMP3 CTF1 RAD23A ZNF224 RFNG IGHMBP2 CPD SEMA4A SUFU TNKS2 AKAP8L PGPEP1 OTUB2 CDC42BPG ENSA ZFP1 CCDC165 TRAF4 TTC8 MARK4 SESN2 LRIF1 GINS2 FLCN ARHGAP33 CCDC142 POLE ABCC9 METTL9 ABCC9 CENPA MAZ GPR132 FAM149B1 CYP19A1 SMAD9 ZDHHC17 KPTN ZNF720 TCF12 CNTD2 PLA2G15 TRADD CNKSR1 DIXDC1 KDM3A CBX1 PGR VDR FGFBP3 NEDD4L RAB31 TBC1D2B PITPNM3 ZNF37A STK11 IRGQ PUM1 HINFP IFITM10 LIX1L C17orf82 FAM109A ANKRD36B EEF2 ESRRA MIB1 PRSS23 B3GALT6 NAA40 FICD PACS1 DPYD PTPRS ARHGEF17 TCF7L2

182

CUEDC1 DCUN1D3 DENND5A KLHL17 AMFR ANKRD26 SRPR ARHGAP35 ARID1A BMF JUND YEATS4 ACOT4 ULK3 ALG9 POLR3GL MBD3 NANOS1 ARHGEF17 SRCIN1 AK4 POLE WDR72 CFH C16orf55 PDDC1 NEO1 C10orf107 FZD4 ATP7B FUK ZNF652 RIMKLB SMG6 KCTD15 WDR59 FLCN CNN2 ZC3H10 DDB2 GOLGA3 NUDT15 FLJ90757 CYTH2 PDIK1L NADSYN1 PLEKHM1 LOC100131691,MZF1 POLG VASH1 SMAD3 CES2 METTL7A IRGQ PIGS ZNF490 GOLPH3L ZFYVE27 C11orf95 SRSF11 ULK3 LRRC27 TLE3 ADSSL1 BMF TIA1 PINK1 SLC43A2 ZNF304 KIAA0226L FGFR2 AQR KCTD1 HMGN2 PPIP5K1 C19orf47 ZNF84 BBS2 LOC100506686 ZNF140 ZNF611 ST6GALNAC2 CA4 PPP1R12B CEP89 ZWINT CEP128 CAB39L POLR2D THEM4 CAPRIN2 IL6R QSER1 NUDT15 LRRC10B SGMS1 NEK5 NYNRIN ZNF724P ARHGAP35 DYM DUSP16 WIZ TRADD ITSN2 RFXAP BEND7 KLF6 LOC100507501 CENPA DENR MFSD11 CDH1 CDC42BPA STARD13 ATG13 ILDR2 GLTSCR2 NUDT15 C18orf54 FRAT2 ZNF18 TMTC2 C1orf151- GLTP IFT140 NBL1,MINOS1,NBL1 FCHSD2 ZNF266 SECISBP2L NENF ZFYVE21 KIF20B

183

GPR176 FLVCR2 WSB2 ALX3 RAVER2 PHACTR4 FLJ31306 CAMKK1 KIAA0182 B3GALT6 C18orf1 PKN2 ATAD3C LTBP1 ATG14 ZFP36L1 WIPI1 PRMT6 PPIAL4G PPT1 C16orf62 PPP2R5C ADAM8 KDM5B ZWINT ZNF714 FAM76A IQCC CBX1 MBP CSK GINS2 PLA2G15 C12orf65 SLK SNX19 BBC3 TXNDC12 GMFB GSTZ1 ZNF468 ITPRIPL2 HSD17B7P2 EIF2B1 CBX5 DCDC2B,TMEM234 SIRT2 TMED8 SPTBN2 PLA2G4F GPSM2 MED29 CDC42BPA ABLIM1 BLNK GPT2 C1orf144 PCSK7 CCDC142 LRP10 KCNMB4 PDCD7 PFKFB3 LINC00483,LUC7L3 STARD13 WDFY2 NDUFS2 SPATS2 PANX1 METTL17 C10orf81 ZNF675 POLL RPUSD4 MGAT5 CNN2 CC2D1B CCDC92 ZNF692 C15orf57 EMP1 AMDHD2 SSH3 MAP3K2 QPCTL CYP19A1 HN1L NUDT17 NIT1 CTTNBP2NL GALNTL1 CENPBD1 BICD1 KAT8 ASCL1 LPCAT2 ZNF213 ZNF506 CDC42SE1 LOC440905 PRADC1 KIAA0907 VASH1 METTL9 ZNF407 CENPA PSKH1 ZNF33A RAD51D,RAD51L3- FLJ35776 FCGR1A RFFL,RFFL MACROD1 C11orf49 CHSY1 TMEM106C MFSD11 ZNF286A LINC00152 ANKRD11 STXBP4 C17orf70 ZNF814 TMEM198B CROCC MCF2L ZNF697 RECQL5 ZNF813 PLXNC1 SLC39A13 ZNF83 PLEKHA1

184

FCGR1A OPN3 HTR7 N4BP2L1 ZCCHC24 DBNDD1 GRTP1 MED29 CCDC115 LOC642846 BTG2 SBF2 EXOC7 PIK3C2B IPO9,NAV1 LOC339803 TIMM50 PTGS2 ZNF785 EMP2 CENPBD1 LOC100130417 ZNF443 IFFO2 EGLN3 DNALI1 MED29 ACYP1 SETBP1 TMCC3 LDLRAD3 EFCAB2 ACAD10 C12orf65 CLK2 DQX1 CDH24 SH2B1 TCF3 MRPS11 SIPA1L2 LDLRAD3 ALMS1 SLC38A7 CAMKK1 B3GNT9 GAS6 ZNF638 MALL ZFYVE27 ZFYVE21 ZFP36L1 HPS1,MIR4685 SMARCE1 PMEL GLS2 HSPB2,HSPB2-C11orf52 PRKCE BAHCC1 MUM1 ACSF3 TAF13 SLC1A5 TMOD3 RPS7 CAB39L ETHE1 CARKD TBC1D15 PPP2R1B CARKD ZNF32 TMEM121 LAT THSD1P1 CC2D1B PSD PIF1 ZNF438

C14orf102

FAM109A

ACADVL

MIR3917,STMN1

SMARCE1

MC1R

185

ESPN

MCEE

FAM113B

FN3K

PRPSAP1

VIM

SLC39A1

CDK3,TEN1-CDK3

ZNF846

KXD1

ADCY6,MIR4701

PKD1P1

ZGLP1

CCDC92

ZSCAN2

MZF1

AKR1C2

TSC2

ZNF652

ZNF17

AMFR

TBC1D10B

MAPRE3

ZNF771

ZNF605

TUBA1A

PDDC1

LOC100130776

PSD4

GCFC2

PLEKHH3

MTERFD3

TAGLN

ALDH4A1,MIR4695

186

CEP95

C10orf137

TSEN54

CRADD

SF3A2

OAZ2

SPATS2

RAVER2

C2orf56

CERS5

C15orf57

IFT88

KIAA0913

SPSB3

NACC1

DPH5

ZNF668

SUV420H2

JMJD1C

ZNF358

IER3IP1

DNAJC4

KDM5B

DSTNP2,LRRC23,RPL13P5

SSBP4

HYI

PPP1R37

ZSCAN18

IRX5

CEP164

LRRC56

ZNF518A

TAF1C

ATE1

187

AP2A2

LOC100499467

LOC100506054

JPH3

HOXC13

PARP16

SLC48A1

C1orf85

PIGQ

MAD2L2 FOXM1 PPP4R1 PPP1R14B KCTD15 TBX2 OSCP1 TBC1D8 TRADD ZCCHC17 CIRBP JUND ESRRA SHC2 LMTK3 AHSA2 SRCAP CLK3 ATAD3B PQLC1 DYRK2 ADAT3,SCAMP4 ANKS3 MICU1 R3HDM4 ACBD6 JMJD7,JMJD7-

PLA2G4B,PLA2G4B ZNF266 SMCR7 SLC44A3 CTU2 SLC39A9

188

TCTN3 ZNF529 MLLT1 LRRC37A3 EIF2D ZIC2 PLEKHG4 VSIG10 ZFYVE1 FTH1 TOM1L1 CCDC159 DDX51 HIP1R COTL1 BCKDK LRRC45 MUC1 SSBP3 AP2A1 ERBB3 TTC7B DPYSL5 SMAD6 ARHGEF25 MUTYH SMYD3 PRC1 PABPC3 CDK10 SLC25A21 WWP2 IFT140 DNALI1 SYTL1 CWC15 DOK4 MAN2A2 SLMO1 MLLT11 FAM22A PRADC1 ZNF140 EPC1 UBR4

189

LOC100507567 RILPL2 GPN2 KBTBD7 PDK2 WDR24 C16orf62 MIEN1 CYTH2 ATG16L2 BTBD6 EIF2B1 MAN2C1,SIN3A FOLR1 FLVCR2 SOS1 PPCDC ARHGEF1 TACC2 JMJD5 UBB CTDNEP1 NEU3 C11orf2 GATA3 GAS5 WBP4 APC2 MYOF HDAC5 GPT2 HHLA3 LOC654342 WIPI1 PLIN5 LOC144481 STX4 PPP1R21 JUP ACP6 PPT1 C18orf21 ZNF271 DOCK6 ABCB9

190

RAB20 TP53I13 DEAF1 SLC16A1 CDK5RAP3 B3GNT1 ST7L PKMYT1 FTH1P3 RPL13A KIAA1522 PAWR C10orf82 MYO1E ZNF836 TMEM218 KIAA0556 LAMP1 COG4 OS9 RNF115 LAT MRPS11 SGCG THOP1 CD68 CHERP NKPD1 TMEM222 ZNF638 SCAF1 CDC42BPA EMP2 INO80B-WBP1,WBP1 ZNF837 CC2D1A MUDENG RPL23AP7 NAP1L1 PUS3 C17orf58 SSH3 NMNAT1 ZMYND11 TPT1

191

KCNS3 ABHD15 WASH7P ZC3H10 LIN7B SAMD1 IFT43 EFNA1 PCBP2 RAD23A PTP4A2 PBX4 MEGF8 TYMS SYMPK RILPL1 SESN2 PRPF40B ZNF821 SFT2D3 PARD6G CFD TMEM54 EML3 PPP1R9B TMEM39B UBTD1 LPPR2

192

Prediction of the interaction between miR-515-5p and MARK4 5’UTR:

1. miR-515-5p sequence from miRbase

>hsa-miR-515-5p MIMAT0002826 UUCUCCAAAAGAAAGCACUUUCUG

2. MARK4 mRNA sequence

 Step A: MARK4 cDNA from ensemble (human)

CTGACGTCCCTTCCTCCCTCCCCAGCCCCTCCACCGCCTCCCTCCGCCGCCGCTTGGGCCGGCTCCG CGCCCCCTCCGCGGCCCCCGCCCGCCCGCCTGCCCGCCGCCCCCATGGCGCCCGGGGTCCCCGCT GCACGGGGCCACTAGGACCCTCGGCGTCCCTTCCCCTCCCCCGCCCTGCCCCCTCTCCCGCCGCGC GGACCCGGGCGTTCTCGGCGCCCAGCTTTTGAGCTCGCGTCCCCAGGCCGGCGGGGGGGGAGGGGA AGAGAGGGGACCCTGGGACCCCCGCCCCCCCCACCCGGCCGCCCCTGCCCCCCGGGACCCGGAGAA G

 Step B: convertion of cDNA into mRNA in Sequence editor

CUUCUCCGGG UCCCGGGGGG CAGGGGCGGC CGGGUGGGGG GGGCGGGGGU CCCAGGGUCC CCUCUCUUCC CCUCCCCCCC CGCCGGCCUG GGGACGCGAG CUCAAAAGCU GGGCGCCGAG AACGCCCGGG UCCGCGCGGC GGGAGAGGGG GCAGGGCGGG GGAGGGGAAG GGACGCCGAG GGUCCUAGUG GCCCCGUGCA GCGGGGACCC CGGGCGCCAU GGGGGCGGCG GGCAGGCGGG CGGGCGGGGG CCGCGGAGGG GGCGCGGAGC CGGCCCAAGC GGCGGCGGAG GGAGGCGGUG GAGGGGCUGG GGAGGGAGGA AGGGACGUCA G

3. Prediction of the interaction between miR-515-5p and MARK4 mRNA in RNAhybrid

193

Manuscript where part of the work reported here is published:

Pinho FG, Frampton AE, Nunes J, Krell J, Alshaker H, Jacob J, Pellegrino L, Roca-Alonso L, de Giorgio A, Harding V, Waxman J, Stebbing J, Pchejetski D,Castellano L. Downregulation of microRNA-515-5p by the estrogen receptor modulates sphingosine kinase 1 and breast cancer cell proliferation. Cancer Research. 2013 Oct 1;73(19):5936-48.

194