INVESTIGATING THE MECHANISM OF OBESITY - ASSOCIATED ESTROGEN RECEPTOR NEGATIVE BREAST CANCER TO IDENTIFY NOVEL AGENTS FOR CHEMO PREVENTION By Vanessa Benham A DISSERTATION Submitted to Michigan State University i n partial fulfillment of the requirements for the degree of Pharmacology and Toxicology Environmental Toxicology Doctor of Philosophy 2020 ABSTRACT INVESTIGATING THE MECHANISM OF OBESITY - ASSOCIATED ESTROGEN RECEPTOR NEGATIVE BREAST CANCER TO IDENTIFY NOVEL AGENTS FOR CHEMOPREVENTION By Vanessa Benham The United States is undergoing an obesity epidemic as nearly 40% of adults are obese. This number is only expected to increase. Evidence over the last two decades demonstrates th e obesity public health crisis has devastating impacts on cancer incidence, specifically breast cancer. As a breast cancer risk factor, elevated body mass index (BMI) is associated primarily with postmenopausal, estrogen receptor (ER) positive (ER+) breast cancer. However, BMI does not describe adipose tissue distribution. Evidence cancer risk. Now we are learning that central obesity reflects accumulation of a type of adipose tissue that is particularly harmful because it is a strong predictor of pre - menopausal ER negative (ER ) breast cancers: triple negative breast cancer (TNBC) and HER2+ER . Currently, there are limited pharmacolog ical means that target ER breast cancer including PARP inhibitors and anti - PD - L 1 inhibitors . Furthermore, TNBC, a subtype lacks the ER, is an aggressive form of breast cancer that affects 10 - 20% of all patients . These patients would greatly benefit from preventative agents . S elective estrogen receptor modulators (SERMs) , like tamoxifen, effectively reduce the risk of ER+ breast cancers by 50%. Unfortunately, SERMS are ineffective against ER breast cancers. Because the mechanisms of how obesity promotes E R breast cancer are unknown, thus developing targeted prevention strategies is difficult. Therefore, the objective of this dissertation is to determine the mechanism s of visceral obesity driven ER breast cancer and use this mechanism to identify chemopre ventive compounds . L iterature suggest s fibroblast growth factor 2 (FGF2) and its main receptor fibroblast growth factor receptor 1 (FGFR1) could be a potential mechanism of obesity - promote d pre - menopausal ER breast cancer. I previously demonstrated that o ne ER mammary epithelial cell line undergo es malignant transformation when treated with factors from visceral adipose tissue ( VAT ) studies on FGF2 - FGFR1 mediated malignant transformation, another group recently identified FGFR1 activation as a primary pathway for obesity - associated progression of ER+ breast cancer after estrogen deprivation. This exciting discovery also implicates FGFR1 signaling as a primary node of breast cancer progression independent of estrogen signal ing. Together , this suggests FGFR1 activation in the context of obesity may contribute to both malignant transformation and progression. Therefore, I hypothesize FGF2/FGFR1 activation is a critical mechanism in VAT - associated breast epithelial cell transformation and is a potential target for chemoprevention. Herein, I have demonstrated that VAT and FGF2 transforms ER breast epithelial cells and this is prevented/attenuated by a selective FGFR1 inhibitor. This reveal ed FGF2/FGFR1 as a critical sign aling mechanism that I utilized t o develop a target - based, phenotypic transformation high throughput screen (HTS) to identify chemopreventive compounds that prevent/attenuate FGF2 - stimulated transformation. With this assay, fluvastatin was identified as th e lead candidate for chemoprevention . Fluvastatin inhibits HMG - CoA reductase, the rate - limiting enzyme in the mevalonate pathway. Interestingly, there is no known mechanistic connection between FGF2/FGFR1 signaling and the mevalonate pathway. My data revealed that factors from VAT upregulated protein expression of mevalonate pathway enzymes such as HMG - CoA synthase 1 (HMGCS1), farnesyltransferase (FNTA), squalene synthesis (FDFT1), and HRas . Furthermore, a selective FGFR1 inhibitor effectively pr evented this VAT - induced upregulation of these enzymes . This suggests that FGF2/FGFR1 influence s the mevalonate pathway. P roducts of the mevalonate pathway include cholesterol and isoprenoids like farnesyl pyrophosphate (FPP) and geranylgeranyl pyrophospha te (GGPP) . These isoprenoids could be important in transformation as GTP binding proteins (GTPase) like oncogenic Ras require FPP to undergo prenylation, a process that is necessary for oncogenic Ras activation . While previous literature has demonstrated H Ras activation stimulates transformation of ER breast epithelial cells, my data did not show that H Ras activation as a critical step in FGF2/VAT stimulated transformation. Further studies are needed to identify exactly how FGF2/FGFR1 and the mevalonate pa thway s are interconnect ed . Overall , these studies implicate the mevalonate pathway in FGF2/FGFR1 signaling and suggest that components of the pathway may serve as targets for prevention. Copyright by VANESSA BENHAM 2020 v This dissertation is dedicated to my family, Thank you for your love and support . vi ACKNOWLEDGMENTS There are many people that have contributed to the completion of this dissertation. First, I want to thank my advisor Dr. Jamie J. Bernard for her support , supervision , and scientific knowledge . I also want to thank the Bernard lab members including Blair Bullard and Wyatt Shoemaker for their constant help and support. Additionally, I want to thank Dr. Anne Dorrance for her guidance and advice as my Graduate Program Director. I am also extremely thankful to my committee, Dr. Stephanie Watts, Dr. Karen Liby, Dr. Kathy Gallo, and Dr. Anne Dorrance. The help and encouragement I received from my committee was instrumental. Thank you for believing in me as a scientist. Lastly, I want to thank my frien ds and family for their love and support. vii TABLE OF CONTENTS LIST OF TABLES ................................ ................................ ................................ ................................ .......... x LIST OF FIGURES ................................ ................................ ................................ ................................ ....... xi KEY TO ABBREVIATIONS ................................ ................................ ................................ ......................... xiii CHAPTER 1 ................................ ................................ ................................ ................................ .................. 1 The complex relationship between obesity and estrogen receptor negative breast cancer risk and the immediate need for novel prevention strategies ................................ ................................ ........................... 1 Overview ................................ ................................ ................................ ................................ ............... 2 The Obesity Cancer Association ................................ ................................ ................................ ........... 2 What is Obesity? ................................ ................................ ................................ ................................ 2 Obesity: a public health concern ................................ ................................ ................................ ........ 3 Ob esity - associated adipose tissue alterations ................................ ................................ ................... 4 Obesity altered adipokine profile ................................ ................................ ................................ .... 5 Obesity - associated inflammation ................................ ................................ ................................ ... 6 Combating Obesity ................................ ................................ ................................ ............................. 6 Obesity and Cancer ................................ ................................ ................................ ........................... 8 Breast cancer: defining triple negative breast cancer ................................ ................................ ........ 8 Obesity and estrogen receptor negative breast cancer ................................ ................................ ..... 9 Molecular mechanisms of obesity - associated cancer ................................ ................................ ........ 10 Modeling transformation ................................ ................................ ................................ ................... 10 Receptor tyrosine kinase dysregulation in triple negative breast cancer ................................ ......... 11 FGF2/FGFR1 in triple negative breast cancer ................................ ................................ ................. 13 FGF2 in obesity ................................ ................................ ................................ ................................ 15 FGF2/FGFR1 in Transformation ................................ ................................ ................................ ...... 16 Prevention of obesity driven cancer ................................ ................................ ................................ .... 17 Chemoprevention: success and failures ................................ ................................ .......................... 17 Chemoprevention drug discovery ................................ ................................ ................................ .... 17 Chemoprevention HTS ................................ ................................ ................................ ..................... 18 Fluvastatin as a chemopreventive agent ................................ ................................ ......................... 19 Statins and cancer ................................ ................................ ................................ ........................... 22 Statins and breast cancer prevention ................................ ................................ .............................. 23 Scope of the Project ................................ ................................ ................................ ............................ 24 CHAPTER 2 ................................ ................................ ................................ ................................ ................ 25 A role for FGF2 in visceral adiposity - associated mammary epithelial transformation ................................ 25 Abstract (120 words) ................................ ................................ ................................ ........................... 26 Introduction ................................ ................................ ................................ ................................ ......... 26 Results ................................ ................................ ................................ ................................ ................ 29 Discussion ................................ ................................ ................................ ................................ ........... 32 Methods ................................ ................................ ................................ ................................ .............. 36 Cell Culture ................................ ................................ ................................ ................................ ....... 36 Human fat tissue filtrate ................................ ................................ ................................ ................... 37 Anchorage - Independent Colony Formation Soft Agar Assay ................................ .......................... 37 FGFR1 Ab Treatment ................................ ................................ ................................ ....................... 38 Statistics ................................ ................................ ................................ ................................ ........... 38 CHAPTER 3 ................................ ................................ ................................ ................................ ................ 39 Identifying chemopreventive agents for obesity - associated cancers using an efficient, 3D high - throughput trans formation assay ................................ ................................ ................................ ................................ ... 39 viii Abstract ................................ ................................ ................................ ................................ ............... 40 Introduction ................................ ................................ ................................ ................................ ......... 40 Results ................................ ................................ ................................ ................................ ................ 42 Development and optimization of transformation HTS ................................ ................................ .... 42 FGFR1 is critical in FGF2 - stimulated transformation ................................ ................................ ....... 46 2D vs 3D growth ................................ ................................ ................................ ............................... 46 Screening of compound libraries ................................ ................................ ................................ ..... 48 Hit confirmation and soft agar validation ................................ ................................ .......................... 48 Picropodophyllin and fluvastatin do inhibit 2D prolifera tion ................................ ............................. 50 Efficacy of statins and IGF1R inhibitors ................................ ................................ ........................... 53 Discussion ................................ ................................ ................................ ................................ ........... 53 Materials/Methods ................................ ................................ ................................ ............................... 58 Cell culture ................................ ................................ ................................ ................................ ....... 58 Reagents ................................ ................................ ................................ ................................ .......... 59 Generating fat tissue filtrate ................................ ................................ ................................ ............. 59 Soft agar assay ................................ ................................ ................................ ................................ 59 Transformation HTS ................................ ................................ ................................ ......................... 60 Compound libraries ................................ ................................ ................................ .......................... 61 Cell viability and proliferation assessment by flow cytometry ................................ .......................... 61 Statistical analysis ................................ ................................ ................................ ............................ 61 CHAPTER 4 ................................ ................................ ................................ ................................ ................ 62 Investigating the mechanistic connection between visceral adipose tissue and the mevalonate pathway in ER breast epithelial cell transformation ................................ ................................ ................................ .... 62 Abstract ................................ ................................ ................................ ................................ ............... 63 Introduction ................................ ................................ ................................ ................................ ......... 63 Results ................................ ................................ ................................ ................................ ................ 66 Statins have varying efficacy in preventing MFTF - stimulated transformation ................................ . 66 MFTF upregulates protein expression of enzymes in the mevalonate pathway .............................. 68 FGF2/MFTF does not alter mRNA expression of mevalonate pathway related proteins ................ 70 Acute FGF2 or VAT stimulation failed to stimulate Ras activation in MCF - 10A cells ...................... 71 Discussion ................................ ................................ ................................ ................................ ........... 73 Materials/Methods ................................ ................................ ................................ ............................... 78 Cell culture ................................ ................................ ................................ ................................ ....... 78 Mouse fat tissue filtrate ................................ ................................ ................................ .................... 78 Soft agar assay ................................ ................................ ................................ ................................ 78 Western blot ................................ ................................ ................................ ................................ ..... 78 RT - PCR ................................ ................................ ................................ ................................ ............ 80 Ras Activation Assay ................................ ................................ ................................ ....................... 81 Immunofluorescence ................................ ................................ ................................ ........................ 81 CHAPTER 5 ................................ ................................ ................................ ................................ ................ 82 General Conclusions ................................ ................................ ................................ ................................ ... 82 Limitations ................................ ................................ ................................ ................................ ........... 87 Unanswered questions ................................ ................................ ................................ ....................... 89 How does adipose tissue secrete FGF2? ................................ ................................ ........................ 89 Which FGF2 isoforms are produced and secreted by adipose tissue? ................................ ........... 90 Potential contraindications targeting FGFR1 ................................ ................................ ................... 91 Future Directions ................................ ................................ ................................ ................................ . 93 Alternative hypotheses ................................ ................................ ................................ ........................ 94 IGF/IGF1R signaling ................................ ................................ ................................ ........................ 94 Estrogen metabolism ................................ ................................ ................................ ....................... 95 Conclusions ................................ ................................ ................................ ................................ ......... 97 APPENDICES ................................ ................................ ................................ ................................ ............. 98 Appendix A: Fluvastatin decreases cholesterol in MCF - 10A cells ................................ ..................... 99 ix Appendix B: FNTA and FDFT1 have co - occurring mRNA alterations with FGFR1 in breast cancer patients ................................ ................................ ................................ ................................ .............. 101 REFERENCES ................................ ................................ ................................ ................................ .......... 106 x LIST OF TABLES Table 1.1 . BMI and weight class .. Table 2.1. Clinical characteristics of donors ... 29 Table 3.1. Funnel Strategy - 2,532 compounds screened Table 4.1. Antibody parameters used for western blot protein detection . 79 Table 4.2. Primer sequences for RT - PCR - test comparing the mRNA of proteins in the Mevalonate pathway in FGFR1 Amplified vs Diploid Breast Cancer Samples Table B.2. mRNA alterations of FGFR1 and/or proteins in the mevalonate pathway and their tendencies to co - occur xi LIST OF FIGURES Figure 1.1. Obesity is linked to 13 different types of cancer Figure 1.2 . Signaling pathways activated by RTKs Figure 1.3. Transforming activity of HuFTF is associated with FGF2 levels Figure 1. 4 . FGF2/FGFR1 is a critical mechanism in VAT - associated, ER independent, epithelial cell transformation Figure 1. 5 . The mevalonate pathway . 20 Figure 1. 6 . Prenylation of HRas enables plasm a membrane localization and in turn activation . . 21 Figure 2.1. Inhibition of FGFR1 attenuates HuFTF - stimulated transformation of MCF - 10A cells . 28 Figure 2.2. FGF2 transforms MCF - 10A cells in a concentration - dependent manner Figure 2.3. FGF2 - tranformed MCF - 10A cells are morphologically and functionally distinct from parent MCF - 10A cells Figure 2.4. HuFTF - stimulated transformation of MCF - 10A cells is moderately associated with the HuFTF FGF2 concentration and BMI . . 32 Figure 3.1. Optimization and development of the transformation HTS . 44 Figure 3.2. PPP and fluvastatin significantly attenuate FGF2 - stimulated transformation of epithelial cells Figure 3.3. PPP, fluvastatin or PD166866 do not inhibit 2D proliferation of JB6 P + cells .51 Figure 3.4. IGF1R inhibitors prevent FGF2 - stimulated transformation whereas efficacy of HMG - CoA reductase inhibitors varies Figure 4.1. Lipophilic statins, fluvastatin and simvastatin concentration - dependently inhibit MFTF - stimulated transformation of MCF - 10A cells Figure 4.2. MFTF upregulates protein expression of enzymes in the mevalonate pathway and this upregulation is prevented by PD166866, a selective FGFR1 inhibitor .68 Figure 4.3. MFTF induces sustained upregulation of HMGCS1 protein and transient upregulation o f FDFT1 protein at 24 hours Figure 4.4. FGF2 and MFTF do not alter mRNA expression of enzymes in the mevalonate pathway in MCF - 10A cells Figure 4.5. Isoprenoids GPP and MVL fail ed to stimulate transformation of MCF - 10A cells and do not rescue the attenuating effects of PD166866 Figure 4.6. FGF2 upregulates H R as protein expression but does not induce membrane localization or activation Figure 4.7. The molecular structure of statins determines if they are Type I or Type II statins Figure 4.8. The FGF2/FGFR1 signaling pathway and the mevalonate pathway xii Figure 5.1. Overall findings: FGFR1 and HMGCR inhibition prevents VAT - stimulated ER breast epithelial cell transformation and VAT upregulation of HMGCS1 may be part of this mechanism Figure 5.2. 2D and 3D cell culture methods . 88 Figure A.1. Fluvastatin treatment for 24 hours significantly decreases the cholesterol content in MCF - 10A cells Figure B.1. Using four TCGA breast cancer databases, mRNA aberrations in FDPS, FNTA, and FDFT1 were associate d with FGFR1 mRNA amplification in breast cancer patients xiii KEY TO ABBREVIATIONS ADDRC Assay Drug and Development Repurposing Core ACAT1 Acetyl - CoA C - acetyltransferase ASC Adipocyte - derived stem cell Atorv Atorvastatin BaP Benzo(a)pyrene BMD Bone mineral density BMI Body mass index BRCA1 Breast cancer 1 BSA Bovine serum albumin CDC Center for Disease Control and Prevention cDNA Complementary deoxyribonucleic acid CHX Cycloheximide CRP C - reactive protein CSC Cancer stem cell CT Computerized tomography DCIS Ductal carcinoma in situ DGAC Dietary Guidelines Advisory Committee DMSO Dimethyl sulfoxide DTP Developmental Therapeutics Program ECM Extracellular matrix EGF Epidermal growth factor EGFR Epidermal growth factor receptor EndoR Endoplasmic Reticulum ER Estrogen receptor ER Estrogen receptor negative ER+ Estrogen receptor positive xiv ERE Estrogen response elements ERK Extracellular - signal - regulated kinase EtOH Ethanol FBS Fetal bovine serum FDPS Farnesyl pyrophosphate synthase FDFT1 Squalene synthesis FFA Free fatty acid FGFa Fibroblast growth factor acidic FGFb Fibroblast growth factor basic FGF1 Fibroblast growth factor 1 FGF2 Fibroblast growth factor 2 FGFR Fibroblast growth factor FGFR1 Fibroblast growth factor receptor 1 FGFR1 Ab Fibroblast growth factor receptor 1 antibody Fluv Fluvastatin FNTA Farnesyl transferase FPP Farnesyl pyrophosphate FTF Fat tissue filtrate FTO Fat mass and obesity - ass ociated protein GAP GTPase - accelerating proteins GAPDH Glyceraldehyde 3 - phosphate dehydrogenase GDP Guanosine diphosphate GEF Guanine nucleotide exchange factor G GPP Geranylgeranyl pyrophosphate GPP Geranyl pyrophosphate GGPS1 Geranylgeranyl pyrophosphate synthase 1 GH Growth hormone GTP Guanosine triphosphate xv GTPase GTP binding protein NCHS National Center for Health Statistics HDL High - density lipoproteins HER2 Human epidermal growth factor receptor 2 HFD High - fat diet HMG - CoA 3 - h ydroxy - 3 - methyl - glutaryl - coenzyme A HMGCR HMG - CoA reductase HMGCS1 HMG - Co A synthase 1 HMW H igh molecular weight HPG axis Hypothalamic - pituitary - gonadal axis HRas H arvey rat sarcoma viral oncogene homolog \ HS Horse serum HSPG Heparan - sulfate proteoglycans HTS High - throughput screen HuFTF Human fat tissue filtrate IARC International Agency for Research on Cancer ICAM - 1 Intracellular adhesion molecule IBET I - BET - 762 IDT Integrated developmental technologies iFGFR1 Inducibl e fibroblast growth factor receptor 1 IGF Insulin - like growth factor IGF1 Insulin - like growth factor 1 IGF1R Insulin - like growth factor receptor 1 IGF2R Insulin - like growth factor receptor 2 IGF3R Insulin - like growth factor receptor 3 IGFBP Ins ulin - like binding protein IgI - III Ig - like domains I - III IHC Immunohistochemistry xvi IL6 Interleukin 6 IPP Isopentenyl pyrophosphate IR Insulin receptor J AK/STAT Janus kinase/Signal transducer and activators of transcription JNK c - Jun N - terminal kinase KRas Kirsten rat sarcoma 2 viral oncogene homolog LDL Low - density lipoprotein LFD Low - fat diet LMW Low molecular weight MAPK Mitogen - activated protein kinase MAT Mammary adipose tissue MCP Monocyte chemoattractant MC4R Mel anocortin - 4 - receptor gene MEK Mitogen - activated protein kinase MEM Minimum essential media MeOH Methanol MetS Metabolic syndrome MFI Median fluorescence intensity MFTF Mouse fat tissue filtrate MM Multiple myeloma MMP9 Matrix metalloprotein ase 9 MmC Mitomycin C MSU Michigan State University MVL Mevalonate NCD Non - communicable disease NCI National Cancer Institute NFDM Non - fat dairy milk NFkB Nuclear kappa - light chain enhancer of activated B cells xvii NMU N - methyl - N - nitrosourea NR as N euroblastoma RAS viral (v - ras) oncogene homolog NVP NVP - AWD742 Ob - R Leptin Receptor P/s Penicillin/streptomycin PAI - 1 Plasminogen activator inhibitor - 1 PD PD166866 PI3K Phosphoinositide 3 - kinase PIP 2 Phosphatidylinositol 4,5 - bisphosphate PIP 3 Phosphatidylinositol (3,4,5) - trisphosphate POMC Proopiomelanocortin PPP Picropodophyllin PR Progesterone receptor Prav Pravastatin PR+ Progesterone receptor positive PR Progesterone receptor negativ e PTEN Phosphatase and tensin homolog RNA Ribonucleic acid Rosu Rosuvastatin RTK Receptor tyrosine kinase RT - PCR Reverse transcription polymerase chain reaction SAT Subcutaneous adipose tissue SERM Selective estrogen receptor modulator SHP2 Src homology region 2 - containing protein kinas phosphate 2 SHUB Spectrum Health Universal Biorepository Sim Simvastatin SNP Single nucleotide polymorphism SREBP Sterol regulatory element binding proteins xviii SREBF2 Sterol regulatory e lement binding protein 2 TBST 1X tris - buffered saline, 0.1% triton X TLR4 Toll - like receptor 4 TNBC Triple negative breast cancer UVB Ultra - violet radiation B VAT Visceral adipose tissue WHO World Health Organization WHR Waist - to - hip ratio 1 CHAPTER 1 The complex relationship between obesity and estrogen receptor negative breast cancer risk and the immediate need for novel prevention strategies 2 Overview In this chapter, I will discuss the epidemiology evaluating the complex relationship between obesity and cancer with a focus on pre - menopausal estrogen receptor ( ER ) negative (ER ) breast cancer . Next , I will address why the molecular mechanisms of obesity - associated cancers continue to el ude modern science . Furthermore, I will explain how understanding the driving molecular mechanisms of obesity associated ER breast cancer will reveal biomarkers for a t - risk individuals and identify targets for prevention. I will discuss preliminary data and current literature that suggests FGF2/FGFR1 signaling is critical in transformation of ER breast epithelial cells . Next , I will explain how the drug discovery process using a target based (FGF2/FGFR1) phenotypic screening can help elucidate the complex mechanisms of obesity associated ER breast cancer . Because the developed HTS identified fluvastatin as the lead candidate, I will discuss the current l iterature and what is known of the relationship between statins and cancer. The Obesity Cancer Association What is Obesity ? Obesity is an emerging di sease of global epidemic proportions as it coincide s with surge s in heart - disease, cancer, and diabetes creating a heavy burden for the health care system. The Center for Disease Control and P r evention (CDC) defines being over overweight or obese as having a weight that is higher than what is considered healthy for a g iven height 1 , defined by b ody mass index (BMI) . BMI is used as a screening tool that determines if a person is underweight, normal, overweight, or obese (Table 1 .1 ) 1,2 . Using Table 1.1 . BMI and weight class Class BMI (kg/m 2 ) Underweight <18.5 Normal 18.5 24.9 Overweight 25.0 29.9 Obese Class 1 30.0 - 34.9 Class 2 Class 3 35.0 - 39.9 >40.0 3 BMI to determine health is controversial because it does not consider body stature, muscle mass, or adiposity distribution, making BMI a crude measure of fat mass. While BMI is a reliable screening tool in are better indicators for determining health risks. Obesity: a public health concern The obesity epidemic emerge d within the first two decades of the 21 st century. First, t he early 20 th century saw people in developed countries gaining proportionally more weight than height (increasing the average BMI). While this did n o t capture the attention of the media or general public for a few decades, it was the start of a burgeoning obesity epid emic. Obesity progressively increased in industrialized counties so much so that the year 2000 was a historical landmark, as for the first time in human evolution, the number of adults with excess weight surpassed those who were underweight 3 . Obesity is a major public health concern since it contributes to many chronic diseases including metabolic syndrome, insulin resistance, type 2 diabetes mellitus 4 , cardiovascular disease 4,5 , and cancer 4 - 11 . Rising obesity rates coincide with increase s in obesity - related diseases that take an expensive toll on the health care system. From 2005 to 2010 the health care costs of obesity in adults rose from $212.4 billion to $315.8 billion in the Unite d States, a 48.7% increase 12 . According the National Center for Health Statistics ( NCHS ) brief reports, United States obesity reports rose ~1.5% from 2005 - 2010 13 . Therefore, a 1.5% increase in obesity rates translated to a $100 billion dollar increase in health care costs. This cost increase during a five year period is a culmination of increase in population, increase in cost per individual, and the increase in obesity prevalence 12 . Therefore, obesity has drastic negative health consequences on society and places a large economic burden on the health care system. Excess adiposity is considered a global non - communicable disease (NCD) target, an d the World Health Organization (WHO) Global Action Plan has a defined goal to halt rising obesity rates by 2025 14,15 . Ob esity rates are continuing to increase and obesity is now a pandemic 16 . In 2016, the NCD Risk Factor Collaboration performed the longest, most complete analysis of body weight trends from 1975 to 2014 16 . They accessed 1698 population - based data sources that included more than 19.2 million participants. In this study they showed rising obesity rates have slowed in high - income countries, while rates have 4 increased in other parts of the world, meaning the gl obal obesity rate has not slowed down 16 . Since 1980, the prevalence of global obesity has more than doubled and is at 12% in adults 17 . Furthermore, if present trends continue, by 2025, global obesity prevalence will reach 18% in men and surpass 21% in women 16 . Obesity in the United S tates is even higher as 42.4% of adults were considered obese (2017 - 2018) 13 . Interestingly , obesity rates vary among different ethnicities. Collec tively, African Americans have the highest obesity rates at 49.6%, followed by Hispanics at 44.8%, then Caucasians at 42.7% . Obesity rates are similar among men of different ethnicities but vary among women of different ethnicities. Hispanic men have the highest obesity rates at 45.7%, followed by Caucasian men at 44.7%, then African American men at 41.4%. African American women experience the highest obesity rates at 56.9%, followed by Hispanic women at 43.7%, then Caucasian women at 39.8% . However, age has less of an impact on obesity in women. The prevalence of obesity in women aged 20 - 39, 40 - 59, and 60+ is 39.7%, 43.3%, and 43.3% respectively 13 . Overall, obesity is serious health risk with pandemic proportions and needs swift effective strategies to prevent or reduce obesity related diseases. Obesity - associated adipose tissue alterations Ad ipose tissue , previously thought to be a passive energy store, is a complex and highly active endocrine organ . Adipose produces hormones and an array of cytokines called adipokines 18,19 . For example, adipose tissue produces growth factors , pro - and anti - inflammatory cytokines, and estrogen. Adipose tissue is composed of adipocytes, pre - adipocytes, immune cells, stromovascular cells, connective tissue matrix, and nerve tissue which forms an interactive network . This network is capable of releasing factors in a paracrine and/or endocrine manner that communicates/acts with local and distant sites 18 . C ohesively , this array of cell types and their a dipokine production and secretion are central to understanding the metabolic abnormalities associated with the development of obesity 20 . Adipose tissue depots exhibit unique adipokine expression, secretion profiles, and receptor expression densities 18,21,22 . Adipose tissue is distributed into two mai n depots, subcutaneous adipose tissue (SAT) and visceral adipose tissue ( VAT ) . SAT is found below the skin, whereas VAT surrounds the intra - abdominal organs (e.g. omental, mesenteric, retroperitoneal, gonadal, and pericardial) 22 . For example, VAT adipocytes are more sen sitive to catecholamine induced lipolysis and less sensitive to lipolysis inhibition 5 than SAT adipocytes 23 - 25 . T riacylglycerol lipolysis in VAT adipocytes produce and release more glycerol and free fatty acids into the body than SAT adipocytes 23,24 . Additionally, VAT and SAT differentially produce and secrete adipokines 21 and exhibit regional differences in receptor densities. Insulin upregulates adiponectin secretion from visceral adipocytes and has no effect on subcutaneous adipocytes 26 . Whereas leptin is primarily secreted from SAT 23 . This demonstrates that VAT and SAT are functionally different . Quantitatively, VAT accounts for 6% of total adipose tissue, but is biologically very active . Over the BMI range of 22 to 36, VAT increased by ~400 % and SAT increased by ~100% [(fat mass at 36 BMI Fat mass at 22 BMI / fat mass at 22 ) x 100] 23,27 . Individuals can gain weight in either or both depots but increases in VAT is associated with obesity - related diseases. Obesity altered adipokine profile A s the percentage of VAT increases, VAT undergoes more functional and structural changes than SAT , including more adipocyte hypertrophy . When adipocytes become hypertrophic, they become dysfunctional that can cause changes that have whole body effects as the adipokine profile can affect processes at distant si tes. Firstly, circulating levels of leptin is directly related to adipose tissue mass as adipose tissue in obese individuals as they produce and secrete higher leptin concentrations levels compared with normal - weight individual s 28 . Whereas, the adipokine adiponectin is downregulated in obese individuals. In addition, BMI is positively associated with increased aromatase expression and estrogen leve ls 29,30 . Furthermore, increases in VAT was positively associated with increased plasma levels of c - reactive protein (CRP), intracellular adhesion molecule 1 (ICAM - 1), monocyte chemoattractant protein (MCP), matrix metalloproteinase 9 (MMP9), and plasminogen activator inhibitor - 1 (PAI - 1) , whereas increases in SAT was only positively associated with increased plasma levels of CRP 21 . Additionally, a dipocytes contain aromat ase, a key enzyme in estrogen biosynthesis, but in normal, pre - menopausal women, ovarian granulosa cells are the main estrogen source with a small fraction of estrogen being produced from other tissues. I n obese premenopausal women, estrogens are predominantly produced in adipose tissue 31 . Additionally, e xcess VAT also exhibits low - grade chronic inflammation and produces increased inflammatory cytokines like interleukin 6 ( IL6 ) . These obesity related changes are theorized to be central to the underlying m olecular mechanisms of obesity - related diseases. 6 Obesity - associated inflammation Obesity results in low - grade, chronic inflammati on in adipose tissue 32,33 . When a dipocytes enlarge and increase in number, this can create a hypoxic environment and not all adipocytes have access to nutrients and/or oxygen. Cell dea th then causes the number of immune cells to increase within the tissue to phagocytize the debris , which results in low - grade chronic inflammation. In excess, VA T experiences changes in immune cell infiltration, macrophages being the major innate recruited immune cells and are the major source of inflammatory cytokines 34 . Functionally, macrophages phagocytize debris from adipocytes that underwent hypoxia - induced adipocyte cell death. A dipocyte cell death releases free fatty acids (FFAs) that act as toll - like receptor 4 (TLR4) agonists and this further induce s an inflammato ry response and enhances macrophage recruitment 35 . Furthermore, the macrophage phenotype ratio switches. M2 - like phenotype is an anti - inflammatory form and the M1 - like phenotype is a more pro - inflammatory form and in obese adipose tissue , the macrophage phenotype shift s to be predominantly M1. Additionally, the low - low grade inflammation is promoted by Leptin. Lepti n is pro - inflammatory whereas its counterpart adiponectin is anti - inflammatory, and increased leptin production leads to increased macrophage production by monocyte differentiation 33 . Macroph ages are phagocytic leukocytes that are impacts the adaptive immunological response 33 . T h is is a small snapshot of the long - recognized relationship between obesity, cancer and inflammation and definitively identifying inflammation as the causal mechanism has yet to be determined. Overall, these obesity - associated VAT changes are suspected to p romote the etiology of obesity - related diseases. Combating Obesity Reducing obesity rates would prevent the occurrence of obesity related diseases , however, this is an ineffective strategy as c ombating obesity is more complicated then limiting caloric intake . Obesity can be cause d by an interplay between behavior, environment, and genetics. An unhealthy d iet was recognized as a risk factor for colon and breast cancer by the Dietary Guidelines Advisory Committee (DGAC) in 2015 36,37 . D iet and cancer risk is still a relatively young field and current studies have significant inter - study heterogeneity 37,38 . It is difficult to holistically investigate diet ary patterns. Addit ionally , f or some people, a 7 healthy lifestyle is an unattainable privilege as l ow socio - economic status is disproportionately associated with obesity 39 . This includes physical activities such as going for a run outside or a walk in the park as these communit ies are not safe environments. Additionally, h ealthy eating is notoriously expensive compared to certain people have limited access to affordable nutritious food and only have corner stores , stocked with predominantly high - calorie food, as their main food source 40 . Furthermore, genetics can pre - dispos e people to obesity . For example, t he fat mass and obesity - associated gene (FTO) has variants associated with increased obesity rates and is found in up to 43% of the population 41,42 . Despite this complicated interplay, p roper diet and exercise reverse the risk for obesity - associated disea ses regardless of weight loss . However, p oor compliance makes this an ineffective strategy to combat obesity , leaving this disease with a dire need for effective prevention strategies. Figure 1.1. Obesity is linked to 13 different types of cancer. Obesity increases the risk of meningioma (cancer in the tissue covering the brain and spinal cord), thyroid cancer, multiple myeloma (cancer of the blood), liver cancer, gallbladder cancer, colon cancer, esophageal cancer (adenocarcinoma), breast c ancer (post - menopausal), stomach cancer, kidney cancer, pancreatic cancer, endometrial cancer, and ovarian cancer. Image was adapted from Centers for Disease Control & Prevention. Image made in BioRender. 8 Obesity and Cancer From 1959 - 1972, the American Cancer Society performed a long - term prospective study, the first large - scale study of mortality according to weight 43 . Here, they evaluated 750,000 men and women and found obese individuals had distinctly elevated mortality for colon and rectum cancer and breast, uterus, and endometrium cancer 43 . The magnitude of the obesity/cancer relationship was revealed in 2003 when the New England Journal of Medicine conducted a study evaluating almost a million U.S. adults demonstrating significant trends of increasing risk with eight cancer types 44 . Today, the epidemiological studies have linked excess weight to the increased risk of 13 different cancer types including breast cancer (Figure 1.2) 2,45,46 . Interestingly, obesity is differently associated to each cancer ty pe. These 13 cancer types account for 40% of all U.S. cancers diagnosed 47 burden. Breast cancer: defining triple negative breast cancer Breast cancer is the most frequently diagnosed cancer in women, affecting 1 in 8 women , after skin cancer , with around 1.3 million cases worldwide. B reast cancer is categorized based on differential expression of ER, progesterone receptor (PR), and human epidermal growth factor receptor 2 (HER2). Breast cancers are classified as either ER - positive (ER + ) or PR - positive (PR + ) , HER2 overexpressing, or triple negative. Triple negati ve breast cancer (TNBC) is characterized as ER , PR - negative (PR ) with no HER2 overexpression . Approximately 10% to 20% of all breast cancers are triple negative 48 . TNBC prognosis is very poor compared to non - TNBC types because it is a more aggressive breast cancer subtype with no effective target ed therapy. Furthermore, the oncogenic drivers , the mutations responsible for cancer initiation and maintenance of TNBC are not understood 48 . Identifying the molecular mechanisms of TNBC tumorigenesis will inform as to the aberrant signaling mechanisms that arise in TNBC and can be used to develop prevention strategies and targeted therapies. This illustrates several critical deficits in the prevention and treatment of TNBC because of its aggressive nature and poor prognosis. TNBC is hard to treat and does not have any clinically approved prevention strategies. While, s elective ER modulator s (SERM s ), like tamoxifen effectively prevent ER+ breast cancer regardless of menopausal status 46 , there are no clinically approved chemopreve ntive compounds for ER breast cancer. ER breast 9 cancer is often slated under TNB C 48,49 . TNBC is a heterogenous breast canc er subtype, therefore a one - strategy fits all is highly unlikel y, further emphasizing an unmet need of ER - independent chemopreventive strategies 50 . Preventing ER tumorigenesis ha s the potential to significantly reduce breast cancer burden 51 . Obesity and estrogen receptor negative b reast c ancer The relationship between obesity and breast cancer is complicated and controversial. Initial studies demonstrated obesity increased the risk of post - menopausal breast cancer and had an inverse relationship for premenopausal breast cancer, meaning obesity has a protective effect ag ainst premenopausal breast cancer . Unsurprisingly, o besity is officially recognized as a risk factor for post - menopausal breast cancer, but not premenopausal breast cancer. However, more current studies undermine th e previously generalized protective effec t of pre - menopausal obesity . Pre - and post - menopausal women exhibit significant biological differences in breast cancer clinical expression. S tratifying breast cancer according to hormone receptor status es reve aled that o besity increases the risk of postme nopausal ER+ breast cancer and premenopausal ER breast cancer . In postmenopausal women, obesity increases the risk of ER+ breast cancer and an inverse/no relationship with ER breast cancer 52 . Conversely, in premenopausal women, obesity increases the risk of ER breast cancer and has no association with ER+ breast cancer 52 . In 2013 , a meta analysis demonstrated premenopausal obese women (BMI > 30 kg/m 2 ) have a 42% higher risk of developing TNBC compared with premenopausal non obese women. This study also revealed no significant association between obesity and TNBC for postmenopausal women 53 . Since 2013, at least seven studies have further supported BMI ( 2 ) is associated with increased incidence of ER breast cancer in premenopausal patients 54 - 60 . A nother pooled analysis from 2011 2 ) was more frequent in case patients with ER /PR tumors compared with other hormone receptor expression combinations 61 . Overall, menopausal status is a significant factor in breast cancers clinical expression , and e pidemiology stratification by breast cancer subtype support s obesity as a risk fa ctor for premenopausal ER breast cancer 53 - 61 . Studies that focus on VAT mass instead of BMI de monstrate stronger association s with pre - menopausal ER breast cancer and not ER+ breast cancer . C omputed tomography (CT ) and waist - to - hip ratio (WHR), a measure of VAT, found generally higher abdominal VAT distributions was associated with 10 ER breast canc er , but not with ER+ breast cancer 62,63 . This suggests that VAT and not SAT promote ER breast cancer in pre - menopausal women . Despite ample evidence supporting visceral obesity as a risk factor for ER breast cancer in pre - menopausal women, the underlying molecular mechanisms have not been clearly elucidated. Therefore, the objective of this thesis is to determine the mechanism o f visceral obesity driven ER breast cancer in order to find new targets for prevention. Molecular m echanisms of o besity - associated c ancer Linking the epidemiology of obesity and cancer risk with a molecular, biological basis has proved difficult. The etiology of obesity associated cancer is hypothesized to involve growth factors , inflammation , and/or hormones . However, o besity is a complex m ultifactorial pathophysiology and these candidate which makes elucidating the exact biological mechanisms difficult . While , obesity - mediated alterations in adipose endocrine functions are hypothesized to be mechanistically involved in obesity associated tumorigenesis , the exact molecular mechanisms remain unclear. Modeling t ransformation In vitro assays are necessary for ex perimental analysis of cancer development on cellular and molecular levels because they offer insights that are often unattainable in whole animal studies 64 . Cancer development is a multistep process that begins with initiation 50 . One model of carcinogenesis involves initiation, promotion, and progression. Initiation is caused by an initiating agent such as a carcinogen that induce s an irreversible mutation in g ene ( s ) that control proliferation, survival, or apoptosis th at leads to abnormal cell growth . The next step is promotion by which the initiated cell proliferates . The last step is progression that results from continued proliferation and accumulation of fu rther mutations that have invasive and metastatic potential 50 . To model cancer development, researchers turn to in vitro assays that detect cell transformation, a process that describes the conversion of normal (non - tumorigenic) cells to transformed (tumori genic) cells in cell culture. In vitro transformation assays are easier to quantitate and control than tumor induction in an in vivo model 64 . 11 Transformation , the steps a normal cell undergoes to become malignant, is modeled in vitro by the soft agar assay that measures anchorage - independent proliferation . For normal cells, pr oliferation is inhibited by cell - cell contact or deprivation of a solid substrate for cells to anchor to. When cells transform and become tumorigenic, the y lose cell - cell contact inhibition and no longer require anchorage - dependence to proliferate , a hallm ark of transformed cells . Therefore, suspending cells in agar selects for cells that can only undergo anchorage - independent growth. Normal cells will remain as single cells suspended in agar as they are not anchored to a solid substrate and transformed cel ls will proliferate forming colonies. These colonies can be counted and used as a measure for transformation. Soft agar is considered the gold standard of transformation because the ability of cells to form colonies in agar directly correlates to the tumor igenicity of these cells in a mouse xenograft model. This signifies that in vitro cell transformation is a valid indicator of cancer development in vivo 65 . Overall, current literature lack s a fundamental understanding of the exact molecular mechanisms responsible for breast cancer development. It is difficult to define the sequential processes that cause a normal cell to lose anchorage - dependence because this can be achieved by different sy stems. Anchorage - dependent growth is regulated by dynamic cellular interactions between cells and extracellular matrix (ECM). When normal cells are not attached to a supportive matrix, cells undergo a specific type of apoptosis called anoikis 66 . Ano ikis is important for physiological function that removes displaced cells, maintains tissue homeostasis, and prevents disease. Cells achieve matrix anchorage through mechanisms including ECM components and cell adhesion molecules. Conversely, changes in th ese mechanisms can be hijacked inducing changes that lead a cell to lose anchorage dependence. It is unknown if or how ECM - cell interactions contribute to or promote transformation. R eceptor tyrosine kinase dysregulation in triple negative breast cancer D ysregulation of several receptor activated signaling pathways can lead to breast cancer development . One such family of receptors is Receptor Tyrosine Kinases (RTKs) and they bind g rowth factors , making RTK s a potential target for obesity associated ER breast cancer prevention 67 . RTK s are cell surface receptors that mediate a range of biological functions, including cell proliferation, migration, survival, and differentiation . Cancer initiation is the first step in carcinogenesis , which is a multistep process. Cancer 12 initiation involve s an irreversible mutation that causes alterations in cell proliferation, survival, and differentiation that allows the cells to evade apoptosis, expand within normal tissue without constrains on cell - cell contact. Not surprisingly, m utations that affect R TK signaling often lead to cell transformation, which is observed in a wide variety of malignancies 68,69 . Therefore, as RTKs regulate the se relevant biological processes, many RTK s are known oncogenes and suspected to be involved in tumorigenesis and progression. RTK signaling pathways are frequently dysregulated in TNBC patients 67 . RTKs activates multiple signal transduction pathways: the phosphoinositide 3 - kinase (PI3K)/A kt , Ras/mitogen - activated protein kinase (MAPK) pathway, and janus kinase/ s ignal transducer and activator of transcription (JAK/STAT) pathway and other pa thways (Figure 1. 2 ) . T argeting the Ras/MAPK pathway with selumetinib, a MEK inhibitor , in vitro inhibit s the invasiveness and motility of TNBC cell lines. Selumetinib also prevented lung metastasis in TNBC - bearing mouse xenograft model 67 . Regarding the PI3K/A kt pathway, RTKs activate PI3K which phosphorylates phosphatidylinositol - 4,5 - diphosphate (PIP 2 ) to phosphatidylinositol - 4,5 - triphosphate (PIP 3 ), which enable s Akt phosphorylat ion that results in protein synthesis and cell growth 67 . The PI3K/A kt pathways is frequently dysregulated in TNBC. Approximately 24% of TNBC tumors have activating PI3K mutations , and 25 - 30% of TNBC cases have loss of function mutations in Phosphatase and tensin homolog ( PTEN ) , a tumor suppressor that dephosphorylates PIP 3 to PIP 2 67 . PTEN also negatively regulates the Ras/ MAPK pathway, making PTEN one example of the high level of crosstalk occurring between Ras/ MA PK and PI3K/A kt pathways that has been detected in basal - like breast cancer models . This biochemical multiplicity limits the efficacy and applicability of RTK inhibitors for TNBC 67 . RTKs consist of several subfamilies of receptors including but not limited to epidermal growth factor receptors (EGFRs), fibroblast growth factor receptors (FGFRs), and insulin and insulin - like growth factor receptors (IR and IGFR respectively) 70 . RTK activity is tightly regulated under normal physiologic cond itions. However, RTKs can undergo potential mutations/alterations that induces the constitutive RTK activation seen in cancer: amplification, gain - of - function mutations, chromosomal rearrangements, and/or autocrine /constitutive activation 68 . For example, overexpression (a result of gene amplification) of RTKs, such as FGFRs, are found in breast cancer 68,70 . 13 FGF2/FGFR1 in triple negative breast cancer FGFRs are a subclass of RT K s that when bound to their endogenous ligands, serve multiple f unction s and dysregulation of FGF/FGFR signaling can lead to cancer formation 70 . The FGF family consists of 22 FGF ligands. The first two FGF ligands were isolated according to pH and received the names FGF acidic (FGFa) and FGF basic (FGFb). When more FGF ligands wer e discovered FGFa and FGFb were renamed FGF1 and FGF2 respectively. The FGFR family has four FGFR receptors (FGFR1 - 4). RTKS, including FGFR1 - 4, typically have an extracellular ligand binding domain with three Ig - like domains (IgI - III), a transmembrane doma in, and an intracellular tyrosine kinase domain. However, due to alternative splicing, FGFR1 - 3 has multiple isoforms with varying FGF ligand specificity 71 . Receptor activation requires two compuls ory events, ligand binding then receptor dimerization. Without cognate ligands, RTKs remain inactive by autoinhibitory mechanisms 70 . Ligands, like those from the FGF family, bind to FGFRs, leading to dimerization and subsequent activation of the intracellular kinase domain by phosphorylation and activation of the aforementioned downstream signaling pathways 7 0 . FGF2/FGFR1 signaling is associated with TNBC progression and may be involved in TNBC tumorigenesis. First, FGF signaling is essential for mammar y gland formation ; specifically FGF2 regulates ductal elongation in the mammary gland, through processes like cell proliferation and epithelial expansion 72,73 . In a breast cancer mouse model, tumors of obese mice had increased expression of FGF2 compared with tumors from lean mice 74,75 . In addition, 62% of basal - like b reast cancers expressed FGF2 compared with only 5% of non - basal - like breast cancers and b asal - like breast cancer cell lines also express FGF2 71 . While these cases are not exclusively triple negati ve, a pproximately 80% of TNBC are basal - like, therefore, these two terms describe similar cancer forms 48 . Another study demonstrated FGFR1 expression was upregulated in TNBC cells , and TNBC cell lines were highly sensitive to FGFR inhibition in anchorage - independent conditions 48,76 . As FGF2 and FGFR1 is induced in breast cancer, FGF2/FGFR1 signaling could promote tumorigenesis. 14 Figure 1.2 . Signaling pathways activated by RTKs . FGF2 is a n endogenous ligand for FGFR1, an RTK. and the Ras/MAPK pathway (blue). Image made in BioRende r. Abbreviations: FRS2 - fibroblast growth factor receptor subs trate 2; GRB2 - growth factor receptor - bound protein 2; GAB1 - GRB2 associated binding protein; SOS - son of sevenless; DAG - diacyl glycerol; CA 2+ - calcium; IP 3 - inositol triphosphate. 15 FGF2 in obesity VAT produces FGF2 and in obesity, FGF2 production and se cretion is significantly increased. E pidemiology studies have demonstrated FGF2 plasma and serum levels increase with high BMI and increased fat mass respectively 77,78 . In vivo , high - fat diet (HFD) - fed mice had elevated FGF2 levels in VAT and serum compared to low - fat diet (LFD) - fed mice. Additionally, when HFD - fed mice underwent surgical lipectomy of the parametrial fat pad (the largest VAT depository in mice), serum FGF2 levels decreased to the FGF2 serum levels seen in LFD - fed mic e 79 . This suggests that excess VAT is a main source for elevated circulating FGF2 levels. Physiologically, FGF2 exhibits a concentration dependent biphasic effect on adipogenesis. Low concentrations (2 ng/mL or lower) of FGF2 enhanced adipogenic expression whereas high concentrations ( 10 ng/mL or higher) suppressed adipogenesis 80 . A s FGF2 has a physiological role in VAT and mammary gland development, FGF2 could link obesity with cancer. Figure 1.3. Transforming activity of HuFTF is associated with FGF2 levels. Fat tissue filtrates were made from human visceral adipose tissue (HuFTF) from four donors undergoing hysterectomy. (a) Percentage of clones growing in soft agar (% colony formation) significantly increases in JB6 P+ and NMuMG cells cultured with HuFTF from donors 1, 2 and 4 compared no treatment (control;Cont). (b) HuFTF with higher concentrations of FGF2 are more potent at stimulating cell transformation 79 . 16 FGF2/FGFR1 in Transformation FGF2/FGFR1 signaling is potentially a critical signaling mechanism between obesity and TNBC. Animal models have shown that VAT promotes tumorigenesis. One study demonstrated that surgical removal of intra - abdominal VAT (lipectomy) in a rat model of HFD - promoted intestinal cancer significantly reduced obesity - associated tumorigenesis 81 . Additionally, another study used this concept with HFD - fed mice, the parametrial fat pads were either lipectomized or the mice underwent a sham - operation, and both groups we re irradiated with ultra - violet radiation (UVB). Compared with the sham operated mice, the lipectomized mice had significantly reduced squamous cell carcinomas by 75 - 80%, suggesting VAT in HFD - fed mice promoted skin tumorigenesis 82 . Our previous data in vestigate d how HFD - VAT promoted skin tumorigenesis. First, we showed FGF2 was elevated in the VAT and serum of HFD - fed mice compared to LFD fed mice. Next, they demonstrated factors from VAT (made into a fat tissue filtrate (FTF) ) stimulated transformation of mouse mammary epithelial (NMuMG) cells and skin epi dermal (JB6 P + ) cells as determined by the soft agar assay 79 . The ability of FTF to transform NMuMG and JB6 P + cells was associated with the concentration of FGF2 in the FTF. Furthermore, knocking out FGFR1 in JB6 P + cells prevented FTF from stimulating transformation 79 . Additionally, exogenous FGF2, subcutaneously injected, stimulated JB6 P + cells to form tumors in an immunocompromised mouse xenograft model 79 . These studi es identified FGF2/FGFR1 as a critical signaling pathway in VAT stimulated transformation of skin epidermal cells. Because FGF2/FGFR1 is characterized in ER breast cancer, we hypothesize that FGF2/FGFR1 is a critical mechanism in VAT - associated ER breast epithelial cell transformation and is a potential target for chemoprevention strategies (Figure 1.4) . Figure 1. 4 . FGF2/FGFR1 is a critical mechanism in VAT - associated, ER independent, epithelial cell transformation. Excess visceral adipose tissue produce s and secretes FGF2 into the circulation. FGF2 activates FGFR1 on breast epithelial cells to stimulat e transformation . Image made in BioRender. 17 Prevention of o besity d riven c ancer Chemoprevention: success and failures Cancer chemoprevention uses agents that prevent cancer development and can significantly reduce cancer burden. For chemopreventive compounds that target ER breast cancer to be successful , potential compounds must uphold high standards in order to successf ully 1) reduce cancer cases that 2) fosters consumer compliance in 3) high - risk individuals 4) that has a tolerable safety profile . First, to reduce cancer cases, effective chemopreventive agents prevent cancer formation by either eliminating premalignant cells or by protecting normal cells from undergoing malignant transformation 83,84 . Second, chemopreventive compounds must be well tolerated with minor to no side effects. Many chemotherapeutics have signific ant adverse side effects. Chemoprevention has different standards because it entails women taking a drug for a prolonged period of time to reduce the possibility cancer might develop. Therefore, the presence of side effects can severely impact consumer com pliance and safety . Tamoxifen is backed by strong data showing it does prevent ER+ breast cancer, but has significant adverse side effects including stroke, blood clots, and endometrial cancer and a long list of common side effects 46 . As a result, tamoxifen compliance is less than optimal. Third, target populations should consist of high - risk individuals in order for cancer - preventive drug s to have the highest benefit 85 . A targeted populati on shifts the risk/benefit ratio that further encourages compliance . Lastly, chemopreventive compounds need minimal to no side effects because they are being administered in a healthy population of people. Side effects can negatively effect consumer compli ance, lowering the effectiveness of the prevention strategy. Overall, p re - menopausal obese women with elevated circulating levels of FGF2 represent a high - risk population that would benefit from chemoprevention. In this scenario, FGF2 not only functions as a biomarker, but as a potential new target for obesity associated TNBC. Chemoprevention drug discovery The fundamental goal of chemoprevention drug discovery is to identify compounds that eliminate or prevent the normal cells from transforming while minimizing or prevent ing negative side effects on normal cells. This is most readily accomplished with high - t hroughput scree n s (HTS) that enables testing of compounds in a high - throughput efficient manner. Challenges faced with drug discovery culminate to the 18 paucity of novel targets and attrition of drugs in the pipeline due to insufficient clinical efficacy . Th is has spurred debates on the most effective drug discovery approaches which center on target - based or phenotypic based approaches. Target - based is self - descriptive and focuses on the target. I n the context of FGF2/FGFR1, it would be identifying compounds that specifically target FGF2 or FGFR1. Phenotypic screens focus on a function or process and in this context would be targeting the transformation process . Both approaches have advantages and disadvantages. Target - based is mechanism informed , however, tar get inhibition can be negated by RTK compensatory crosstalk and redundancy that often prevents clinical efficacy. Phenotypic approaches are at risk of identifying non - selective compounds and promiscuous compounds are more likely to have off - target effects and in turn side effects . This is problematic for chemoprevention as these compounds need to be efficacious with minimal to no side effects 86 . Chemoprevention HTS Utilizing high - throughput screening methods to identify chemopreventive compounds is an underdeveloped area and currently faces many challenges. C urrent chemoprevention assays predominantly use cancer cell lines in a 2D format, methodology that is not representative of chemoprevention. Cancer cells are morphologically and functionally dissimilar to non - tumorigenic cells and inducing cancer cell deat h is different than preventing a normal cell from transforming. Additionally, HTS s in a 2D format ( cells that attach to the cell cultured treated plate s that proliferate in a monolayer ) does not inform on the process of transformation. Both transformed and non - transformed can proliferate in 2D culture conditions, giving no way to distinguish between the two cell varieties , and cell death does not mean the process of transformation is bei ng inhibited or prevented. In contrast, 3D culture conditions (anchorage - independent growth) like those employed in the soft agar assay and ultra - low attachment conditions do distinguish between transformed and non - transformed cells. Furthermore, cell prol iferation in 2D and 3D utilize different factors and signaling pathways. This is demonstrated by a study that demonstrated E rb (HER4) overexpression (by lentivirus transduction ) transformed MCF - 10A cells but did not induce proliferation of TNBC cell line s whereas E rb (HER2) did not transform MCF - 10A cells but did stimulate proliferation of TNBC cell lines 87 . MC F - 10A cells are non - tumorigenic human breast epithelial cells. Consequently, mechanisms of cancer cell proliferation are not reflective of mechanisms of transformation. 19 The limitations and challenges that come from using cancer cell lines in a 2D format ha ve yet to be overcome with a novel HTS that models the process of transformation starting with non - transformed cells. Currently, no HTS assays model s the transformation process. The soft agar assay is not suitable for HTS as scaling down the assay is not f easible, nor would that method be easily quantified. One study demonstrated that growth of transformed or non - transformed cells in ultra - low attachment conditions, correlated to growth in soft agar , in a high - throughput format (384 - well) . Therefore, this p latform using 3D growth in ultra - low attachment conditions could be utilized to develop the first HTS to model the transformation process and enable screening of compound libraries to identify chemopreventive compounds against obesity - associated transforma tion of breast epithelial cells . This methodology enables the associated ER breast cancer . Fluvastatin as a chemopreventive agent The HTS methodology identified fluvastatin as the lead candidate for chemoprevention. Fluvastatin is a part of a large class of statins that inhibit 3 - hydroxy - 3 - methyl - glutaryl - CoA (HMG - CoA) reductase (HMGCR), the rate limiting step in the mevalonate pathw ay 88 . The mevalonate pathway begins with acetyl - CoA, the product of glycolysis. Two molecules of acetyl - CoA are combined by acetyl - CoA C - acetyltransferase (ACAT1) to make acetoacetyl - CoA. Here, acetoacetyl - CoA is condensed to form HMGCR by HMGCS1 and HMG - CoA is converted into mevalonate by HMGCR. Mevalonate is metabolized to isopentenyl pyrophosphate (IPP), an isoprenoid precursor to FPP and then GGPP by FDPS and GGPS1, respectively. FPP can also be converted to squalene by sq ualene synthesis (FDFT1) and squalene is the precursor of cholesterol (Figure 1. 5 ) 89 . Additionally, FPP and GGPP are essential for prenylation, a series of posttranslational modifications that enable protein function through plasma membrane localization and anchoring molecules (Figure 1. 6 ) 90,91 . Prenylation is a three - step process that leads to the addition of a prenyl moiety to a small GTP binding ( GTPase ) protein. There are two types of prenylation: farnesylation and geranylgeranylation. Farnesylated proteins include the Ras superfamily (K/N/HRas) and geranygeranylated proteins include RhoA/B/C, Rap1A, Rac - 1, and Cdc42. The Ras GTPase superfamily is a well - recognize d class of prenylated proteins. Protein prenylation is require d for protein activity. When 20 H/NRas has been farnesylated and palmitoylated (attachment of fatty acids), Ras has a 100 - fold higher affinity for the plasma membrane than farnesylated Ras. Plasma membrane localization places Ras within proximity of cell surface effector proteins where they can activate HRas and subsequent signaling (Figure 1. 6 ). While this is a small snapshot of the mevalonate pathway, it plays a key role in cellular metabolism and homeostasis. Figure 1. 5 . The mevalonate pathway. The mevalonate pathway begins with Acetyl CoA. Acetyl CoA can be derived from Acetate or can be absorbed through the cell or derived from citrate produced in the TCA cycle. Acet yl CoA and is converted to Acetoacetyl CoA by ACAT1. HMGCS1 converts Acetoacetyl CoA to HMG - CoA. Here, HMGCR produces mevalonate. Mevalonate produces intermediates FPP and then GGPP, isoprenoids that can undergo farnesylation and geranylgeranylation respectively. The FGF2 ligand binding to FGFR1 stimulated multiple signaling pathways including the Ras/MAPK pathway which regulates gene transcription. Abbreviations: ACAT1 - Acetyl - CoA Acetyltransferase 1 ; PGTB1 - geranylgeranyl transferase type 1 89 . 21 Statins are efficacious cholesterol lowering agents that are essential medical therapy in preventing heart attack and stroke. Statins are classified based on lipophilicity which reflects tissue selectivity. Lipophilic statins like fluvastatin, simvastatin, and at orvastatin penetrate hepatic and extrahepatic cells whereas hydrophilic statins like pravastatin and rosuvastatin have less absorption with non - hepatic cells. Regarding cholesterol - lowering and cardiovascular disease, different statin types have slightly d ifferent efficacies and there is currently no clinical standard for selecting statin type. Studies have shown similar effects and safety in patients with hydrophilic and lipophilic statins in coronary artery disease and acute coronary syndrome 92 . Statins also have a relatively good safety profile, making them ideal candidates for disease prevention. Figure 1. 6 . Prenylation of HRas enables plasma membrane localization and in turn activation. Prenylation is a three - step process that attaches the farnesyl moiety, proteolysis of the aaX residue, and carboxymethylation. Prenylated proteins like H Ras and NRas then undergo palmitoylation then membrane localization . This puts HRas in prox imity to cell surface receptors and adaptor proteins where they can be activated and participate is cell signaling pathways. Ras bound to GDP is inactive. GEF exchanges GDP for GTP and GAPs exchanges GTP for GDP. Active Ras with appropriate adaptor proteins. Abbreviations : GEFs - guanine nucleotide exchange factor ; GAPs - GTPase activating proteins; GTP - guanosine triphosphate; GD P - guanosine diphosphate. 22 The transformation HTS is not the first to associate the mevalonate pathway with tumorigenesis. The mevalonate pathway is often dysregulated in tumor cells in order to produce more cholesterol and accommodate higher energy demand s 93 . M evalonate pathway dysregulation can also promote tumorigenesis through protein prenylation. As Rho and Ras proteins are known oncogenes, the tumor suppressive effects of statins are likely from reduced prenylation of Ras and Rho, thus reduced oncogene activity 88 . However, the exact mechanisms of how mevalonate pathway dysregulation promotes tumorigenesis is unknown. Statins and c ancer Investigating fluvastatin as a potential chemopreventive agent in ER breast cancer poses many challenges as the epidemiological literature surround ing statins and cancer is highly heterogenous and inconsistent. Initial retrospective analyses investigating statins and cancer prevention demonstrated statins did not increase cancer risk nor decrease cancer risk. However, these initial studies were not designed to evaluate cancer risk but to evaluate cardiovascular outcomes 94 . When subsequent in vitro experiments indicated statins slow the growth of cancer cell growth 94 , it revealed a need for further investigations to evaluate s tatins for chemoprevention. Statins are investigated for chemopreventive properties as well as adjuvant chemotherapeutics. Today, meta - analyses show that adjuvant statin use is associated with modest reductions in cancer - related mortality 95 . There is still inter - study heterogeneity, but cumulative analyses show a reduction. For inst ance, a meta - analysis of 95 studies including, 18+ cancer types , found consistent evidence that statin use was associated with reduced cancer - specific mortality 96 . While the majority of these studies investigated prostate, breast, and colorectal cancer, this association was up held when stratified by cancer type 96 . Another more recent meta - analysis of 60 observational studies found similar findings regarding statin use and reduced cancer - specific mortality 95 . As a result, studies are assessing statins as adjuvant chemotherapeutics. Overall, this is still a burgeoning area of research and needs a better understanding of what tumor molecular characteristics equate to statin sensitivity for different cancer types. In contrast, e pidemiological studies investigating statin use with cancer prevention are overall inconsistent. This is a result of highly heterogeneous literature. Meta - analyses that evaluated the effect of 23 statin use and prostate cancer risk identified five studies whe re statin increased this risk, 10 studies that decreased the risk, and 18 studies that demonstrated there was no effect 97 . Literat ure on statin use and colorectal cancer predominantly describe s no effect of statin use on cancer risk 98 whereas stat in use was strongly associated with a lower risk of liver cancer 99 . Breast cancer literature is more challenging to investigate as statin chem o prevention is affected by statin type, tumor subtype and age. More research is needed to determine which cancers could benefit from statin chemoprevention including stratificati on by statin type (lipophilicity). Furthermore, a better understanding of how statins prevent tumorigenesis of different cancer types and would further inform on populations that could benefit from statin use. Statins and breast cancer prevention Statin c hemoprevention efficacy on breast cancer risk is influenced by lipophilicity, cancer site, and hormone receptor expression . This presents many challenges as this requires study designs to have enough statistical power to evaluate these endpoints. S tudies demonstrate that stratify ing by breast cancer subtype and statin type suggest lipophilic statins have a protective against ER breast cancer and not ER+ breast cancer. For instance, Langballe et al . demonstrated that statin use reduced the risk of c ontralateral breast cancer among patients with ER breast tumors and not with patients with ER+ tumors 100 . Interestingly, Kumar et al . had a striking observation in that lipophilic statin use (for more than one year) appeared to significantly reduce ER breast cancers by 37% relative to the proportion of ER+ breast cancers 101 . Another study revealed younger patients experienced more favorable outcomes with statin use reducing risk of breast cancer reoccurrence 102 . Moreover , two meta - analysis found lipophilic statins were associated with decreased breast cancer risk 103,104 , supporting the results of two previously published randomized control trials 105,106 . The studies analyzed by Manthravadi et al pertaining to breast cancer subtype , a limitation that might have influenced their conflicting results 104 . These data demonstrate that statin type, breast cancer subtype, and menopausal status all have the potential to influence or alter associations between statins and obesity - associated breast cancer risk. It is difficult to address all of these variables and maintain enough statistical power to determine associations. As a result, the effect of statin use on breast cancer risk is still controversial among epidemiology studies. There are several studies that do not demonstr ate a chemoprotective effect. For example, two retrospective studies, 24 107 108 , did not find a protecti ve effect with statins and ER breast cancer but evaluated lipophilic and hydrophilic statins collectively. Furthermore, there have not been epidemiological analyses on statin use and obesity - related cancers, therefore, there is no literature that could de termine i f premenopausal obese women would be a population to benefit from statin chemoprevention. Ultimately, the epidemiological complexities / should not be a deterrent against statins as chemoprevention. Because of s tatin tolerability and cardiovascular benefits, statins should not be overlooked when it comes to their potential to reduce the risk of ER breast cancer 101 . Scope of the Project The introduction has described the current knowledge gaps surrounding obesity - associated ER breast cancer and fluvastatin as a potential chemopreventive agent. In summary, i t is unknown how VAT stimulates transformation of ER breast epithelial cells nor how HMGCR is involved in this process. The preceding rationale was used to formulate the central hypothesis tha t FGF2/FGFR1 is a critical mechanism in ER - independent VAT - associated transformation and is a potential target for chemoprevention strategies . This dissertation went beyond this initial hypothesis and investigated the mechanisms of fluvastatin, the lead ca ndidate identified for chemoprevention, and determine the role of the mevalonate pathway in VAT - stimulated transformation of ER breast epithelial cells. 25 CHAPTER 2 A role for FGF2 in visceral adiposity - associated mammary epithelial transformation Published in Adipocyte 2016, 7:2, 113 - 120, DOI: 10,1080/21623945.2018.1445889 Vanessa Benham 1 , Debrup Chakraborty 1 , Blair Bullard & Jamie J Bernard 1 1 Department of Pharmacology and Toxicology, Michigan State University, East Lansing, MI 48824 26 Abstract (120 words) Obesity is a leading risk factor for post - menopausal breast cancer, and this is concerning as 40% of cancer diagnoses in 2014 were associated with overweight/obesity. Despite this epidemiological link, the underlying mechanism responsi ble is unknown. We recently published that visceral adipose tissue (VAT) releases FGF2 and stimulates the transformation of skin epithelial cells. Furthermore, obesity is differentially associated with many epithelial cancers, and this mechanistic link co uld be translational. As FGF2 and FGFR1 are implicated in breast cancer progression, we hypothesize that VAT - derived FGF2 plays a translational role in promoting adiposity - associated mammary epithelial cell transformation. In this brief report, data sugges t that FGF2/FGFR1 signaling is a potential mechanistic link in VAT - stimulated transformation of breast epithelial cells. Introduction Obesity is a well - established risk factor for post - menopausal breast cancer. 109 A greater waist to hip ratio [indicative of a higher content of visceral adipose tissue (VAT)] increases the risk of post - menopausal breast cancer 110 - 113 . In pre - menopausal breast cance r, when adjusted for weight or body mass index (BMI), women with the smallest waist to hip ratios have a 37% lower risk 114 . Thus, visceral obesity, an inc rease in adipose tissue surrounding the intra - abdominal organs, directly relates to the magnitude of obesity - related breast cancer risk 110 - 113 . However, the underlying mechanisms responsible for the VAT - breast cancer link are not fully elucidated 115 . Anatomical and physiological differences between VAT and subcutaneous adipose tissue (SAT) determine the extent of how these depots contribute to obesity, metabo lic syndrome (MetS), and cancer 23,116 . In obese individuals, adipocytes become hypertrophic, which makes them dysfunctional and insulin resistant. The pathophysiology of obesity - ind uced insulin resistance has been attributed to ectopic fat deposition, increased inflammation, oxidative stress, adipose tissue hypoxia and mitochondrial dysfunction, and impaired adipocyte expansion and angiogenesis 23,116 . Excess VAT contains a greater number of large adipocytes in contrast to SAT, which contains smaller, insulin sensitive adipocytes 23 . Moreover, expanding adipose tissue can induce hypoxia from insufficient vasculature and oxygen supply 23 . This hypoxia can induce immune cell infiltration, causing low - grade chronic inflammation 117 . Adipocyte hypertrophy and 27 immune cell infiltration alters the release of adipokines (cytokines derived from adipose tissue) that can exacerbate the immune response and induce systemic release of adi pokines that can act on neighboring and distant targets 11 5,117 . These obesity - related changes are associated with insulin resistance, which in turn le ads to hyperglycemia, hypertension, dyslipidemia, and other metabolic abnormalities 116,118 . Similarly, MetS is characterized by a cluster of three or more metabolic abnormali ties including visceral obesity, insulin resistance, dyslipidemia, hypertension, and hyperglycemia. However, these similar physiological aspects, they are not mutually exclusive; for example, not everyone who is obese has inflammation and metabolic syndrom e 119 . Regardless, many studies have concluded the rise in visceral obesity has led to an increase in MetS 116 . Epidemiological studies show both obesity and MetS are breast cancer risk factors 111,120 . Bridging the link between obesity, MetS and breast cancer risk, Kabat et al . showed obesity is associated with increased breast cancer risk, and metabolically unhealthy obese individuals had the highest risk 119 . However, epidemiological as sociations and obesity - related changes fall short of explaining the biological mechanisms by which adiposity contributes to cancer promotion and malignant transformation (a change a cell undergoes to become malignant). Animal models have given insight into the mechanistic link between visceral adiposity and cancer. In a rat model of intestinal cancer, removing VAT significantly attenuated obesity - associated intestinal tumorigenesis 81 . In addition, we previously demonstrated removing VAT in HFD - fed mice significantly reduces UVB - induced squamous cell carcinomas by 75 - 80% when compared to sham - operated cont rol animals 82 . These data suggest VAT - derived factors are critical for carcinogenesis 81,82 . We also utilized an ex vivo model to evaluate the ability of VAT - derived growth factors to stimulate transformation of non - tumorigenic JB6 P + mouse skin epithelial cells. Cellular transformation as indicated by anchorage - independent growth in soft agar is a well - established, stringent method for detecting the tumorigenic potential of transfo rmed cells 121 - 123 . JB6 P + cells can not proliferate in an anchorage - independent manner but have the ability to transform upon treatment of tumor promoters 121 - 123 . Using this model, we identified fibroblast growth factor 2 (FGF2) as the critical VAT - derived factor in stimulating JB6 P + growth in soft agar assay 79 . JB6 P + cells t to transform in the presence of VAT, suggesting the FGF2/FGFR1 signaling axis is critical in VAT - stimulated transformation of epithelial cells at distant targets. How gener alizable this mechanism is to other tissues 28 and human cells is unknown. Therefore, we hypothesize that VAT - stimulation of skin carcinogenesis through the FGF2/FGFR1 signaling is translational to VAT - associated breast cancer. The objective of this study is to determine the effects of human VAT on the transformation of MCF - 10A human mammary epithelial cells. MCF - 10A cells are non - tumorigenic and do not exhibit anchorage - independent growth in soft agar. We hypothesized that VAT w ill stimulate the transformation of MCF - 10A cells and this activity will be dependent on FGFR1. Establishing a human model of VAT - stimulated transformation will strengthen support for the direct role of VAT in adiposity - promoted carcinogenesis. There are f undamental differences in the transformation susceptibility of human and mouse cells specifically in the greater number of events required to transform human cells than those required for non - tumorigenic mouse cells 124,125 . Con sequently, establishing a human model of VAT - stimulated transformation of mammary epithelial cells shows adiposity - promoted carcinogenesis is relevant to both mouse and human models and is translational to obesity - associated breast cancer. Figure 2.1. I nhibition of FGFR1 attenuates HuFTF - stimulated transformation of MCF - 10A cells. HuFTF and HuFTF+Ab significantly stimulated colony formation above the untreated control. Cells were - stimulated growth in soft ag ar was significantly attenuated by the FGFR1 Ab (p=0.02). MCF - 10A ells were treated with a FGFR1 neutralizing antibody (FGFR1 Ab) at as described in Me thods, MCF - 10A cells were cultured as described in Methods. Data is presented as mean ± six biological replicates. Each biological replicate had three technical replicates. Confidence intervals (CI) were calculated for HuFTF treated MCF - 10A cells (95% CI 3 .024 - 7.658) and for HuFTF+Ab (95% CI 0.093 - 2.929). Both treatments did not contain the untreated control (0% colony formation). Statistical significance between HuFTF and HuFTF+FGFR1 Ab was determined by unpaired t - test (* p<0.05). 29 Results To test the effects of VAT on mammary epithelial transformation, human fat tissue filtrate (HuFTF) was generated from VAT of six different human donors. VAT was obtained from omental tissue of cancer - free female obese human subjects purchased from Spectrum Health Univ ersal Biorepository (SHUB, Grand Rapids, MI). These subjects were undergoing surgery for gastrointestinal conditions. Table 2. 1 describes the human donor characteristics including age, BMI, gender, and ethnicity. We were not able to obtain information on menopausal status, metabolic status or serum metabolites. To determine if HuFTF stimulates transformation of human mammary epithelial cells, MCF - 10A cells were incubated with 200 counted and a percent of colony formation was obtained by relating the number of colonies with the number of cells plated (750 cells/well). Wh ile MCF - 10A cells are non - tumorigenic epithelial cells, they have a low level of spontaneous transformation in contrast to tumorigenic epithelial cells, which have almost 100% transformation. Using this low, baseline, spontaneous transformation, tumor prom oters can be added to stimulate transformation, inducing colony formation to increase above the control. HuFTF significantly stimulated colony formation above the control with statistical significance (p<0.05) (Figure 2. 1). Table 2.1. Clinical characteri stics of donors Donor ID FGF2 (pg/mL) Age BMI Gender Ethnicity 01 1.0034 52 30.8 F African American 02 1.1472 76 28.1 F White 03 0.61684 70 26.0 F White 04 0.67632 82 27.6 F White 05 0.50780 62 22.0 F White 06 0.35249 50 24.0 F White Abbreviation: F, Female 30 To determine the role of FGFR1 signaling in HuFTF - stimulated transformation, MCF - 10A cells were incubated with a tyrosine kinase fibroblast growth factor receptor 1 antibody antagonist (FGFR1 Ab). Inhibiting FGFR1 receptor activity attenuated HuFTF - stimulated transformation of MCF - 10A cells (Figure 2. 1). - stimulated colony formation, indicating FGFR1 signaling is required for optimal HuFTF - stimulated transformation. Figure 2. 2. FGF2 transforms MCF - 10A cells in a concentration - dependent manner. FGF1 an d FGF2 significantly stimulated transformation of MCF - 10A cells at 10 and 20 ng/mL. FGF18 significantly stimulate transformation at 10 ng/mL but not 20 ng/mL and FGF21 was not statistically significant. MCF - 10A cells were cultured as described in Methods, control cells were untreated. Data is presented as mean ± SE of triplicate values. Statistical significance was determined by one - way ANOVA with multiple comparisons (*p < 0.05, **p < 0.01, ***p < 0.001). The percent of colony formation was calculated as d escribed in Methods. 31 FGFR1 is a receptor for many FGF ligands and to determine the optimal ligand for stimulating MCF - 10A transformation, MCF - 10A cells were incubated with FGF1, FGF2, FGF18, and FGF21. FGF1 binds to all four FGFR receptors, FGF18 has the highest affinity for F GFR3 IIIc, with some affinity for FGFR4, FGF21 has the highest affinity for FGFR4, followed by FGFR2 IIb and FGFR IIIc, and FGF2 has the highest affinity for FGFR1 IIIc, FGFR3 IIIc, and FGF4 with some affinity for FGFR2 IIIc 126,127 . While FGF1, FGF2, and FGF18, significantly increased colony formation in MCF - 10A cells, FGF2 was the only ligand to induce a concentration response at 10 and 20 ng/mL (Figure 2.2). A colony of FGF2 - stimulated MCF - 10A cells was isolated from soft agar and grown in traditional cell culture plates. The FGF2 - transformed MCF - 10A cells have a fibroblastic - spindle morphology compared to the parent MCF - 10A cells that are more epithelial - like with a polygonal shape (Figure 2.3). After several passages this spindle morphology remained and the FGF2 - transformed MCF - 10A cells demonstrated an irreversible ability to grow in soft agar (data not shown). Figure 2. 3. FGF2 - tr anformed MCF - 10A cells are morphologically and functionally distinct from parent MCF - 10A cells. (A) MCF - 10A cells have epithelial - like morphology and transformed MCF - 10A cells have spindle - like morphology. Transformed MCF - 10A cells were obtained by treatin g MCF - 10A cells with FGF2 in soft agar. After 14 days, a colony was isolated and cultured in traditional cell culture conditions for several passages, making transformed MCF - 10A cells. Untreated transformed MCF - 10A cells formed over 44% more colonies in so ft agar compared to untreated parent MCF - 10A cells (data not shown). 32 The concentration of FGF2 in each HuFTF donor was determined by ELISA and related to the percent of colony formation and BMI (Figure 2.4). The transforming activity of each HuFTF in th e soft agar assay was moderately associated with FGF2 concentration in the filtrates ( R 2 = 0.45 ) (Figure 2.4A) and with BMI ( R 2 = 0.64 ) (Figure 2.4B). Additionally, there was a moderate association between BMI and HuFTF FGF2 concentrations ( R 2 = 0.64 ) (Figure 2.4C). Discussion Visceral adiposity is significantly associated with breast cancer risk, and despite this strong association, t he mechanism is unknown 128,129 . We previously showed visceral obesity promoted skin tumor formation, with our mechanistic finding that VAT - derived FGF2 stimulates skin epithelial cell transformation through FGFR1 79 . As obesity is associated with many different cancers, we hypothesized our mechanistic finding may be translational to other obesity associated cancers, like breast cancer. Visceral obesity, as measured by large waist circumferences and waist - to - hip ratios, is strongly correlated with pre - and post - menopausal breast cancer risk 110 - 113 . Herein, we describe a translational role for VAT - derived FGF2 in stimulating mammary epithelial cell transformation through FGFR1. These findings h ighlight FGF2/FGFR1 signaling as a potential link between VAT and breast cancer risk. Figure 2.4. HuFTF - stimulated transformation of MCF - 10A cells is moderately associated with the HuFTF FGF2 concentration and BMI. (A) HuFTF with higher FGF2 concentration s more potently stimulated MCF - 10A transformation compared with HuFTF with lower FGF2 concentrations (R2 = 0.45). (B) HuFTF from donors with a higher BMI more potently stimulated MCF - 10A transformation compared to HuFTF from donors with a lower BMI (R2 = 0 .64). (C) Higher HuFTF FGF2 concentrations is moderately associated with a higher BMI ((R2 = 0.64). The % colony formation, HuFTF FGF2 concentration, and BMI of six HuFTF were used. MCF - 10A cells were cultured as described in Methods. Data were analyzed wi th Linear regression (performed in PRISM). 33 Previous research implicates FGF2/FGFR1 signaling in breast cancer 130 . Constitutive activation of FGFR1 in normal mouse mammary epithelium induced proliferation, invasive lesions, and antiapoptotic effects 131 . In breast cancer cells, FGF2 is a strong mitogen and potent antiapoptotic and induces invasiveness while subverting various chemotherapeutic agents, leading to drug resistance 132 - 134 . In addition, FGFR1 activation increases proliferation and invasion of breast cancer cell lines 135,136 . Clinical studies have shown that FGF2 levels in serum, nipple aspirate fluid, and tumor samples are higher in patients with cancerous breast tumors as compared with benign breast diseases/tumors 137 - 140 . In breast cancer patients, FGFR1 amplification is seen in up to 10 - 15% of all breast cancers and is associated with early relapse and poor survival 120,141 . Likewise, tumors overexpressing FGFR1 exhibi ted increased proliferation and decreased distant metastasis - free survival 141 . The role of FGF2/FGFR1 in breast cancer onset is less clear. One study demonstrated that genetic variants in FGFR1, FGFR3, or FGFR4 had no impact on breast cancer risk, 142 whereas an intronic single - nucleotide polymorphism (SNP) in the FGFR2 gene was associated with an increased risk of breast cancer, particularly estrogen receptor (ER) posi tive disease 142 . A separate study demonstrated no significant associations with SNPs in FGF2 and breast cancer risk 143 . However, the functional relevance of these FGF2 polymorphism for function are unknown. Our data demonstrate that FGFR1 activation is critical for optimal VAT - stimulated MCF - 10A cell transformation. Th ese data add additional relevance to the previous findings that FGFR1 activation by inducible dimerization of the receptor induced growth in soft agar of MCF - 10A cells 135 . Moreover, we showed VAT FGF2 concentrations were associated wit h VAT transforming activity (Figure 2 .4 A). Collectively, these data suggest that FGF2 from VAT stimulates mammary epithelial cell neoplastic transformation through FGFR1 activation. The downstream effects of FGFR1 activation of transformation are unknown, but it would be interesting to observe a potential subtype or breast cancer signature in tumors that arise from visceral adiposity - promotion. Breast cancer is characterized into different subtypes based on expression of ER, progesterone receptor (PR), and human epidermal growth factor receptor (HER). Studies show FGFR1 - overexpressing tumors are frequently ER positive, the primary subtype associated with obesity 120,144 . Therefore, we would hypothesize that FGF2 activation of FGFR1 would promote ER positive tumors in viscerally obese individuals. 34 Evaluating individual characteristics including age, BMI, gender, and ethnicity were not associated with transforming capacity of the HuFTFs. As obesity is more stro ngly associated with post - menopausal breast cancer, we requested samples from individuals of post - menopausal age. Menopause occurs on average in women at 51 years old, however, we were not able to confirm the menopausal status of the individual donors as t wo donors were 52 and 50 years old at the time of surgery. It would be interesting to investigate if menopause affects the quality of VAT, in turn affecting FGF2 levels. In our study, there was no relationship between age and the transforming capacity of t he HuFTF ( R 2 = 0.09 ). For BMI, there is a moderate association with BMI and the transforming capacity of the HuFTF, giving an R 2 value of 0.64 (Figure 2 .4 B). This suggests BMI might be an indicator of VAT FGF2 levels. We were not able to investigate ethnici ty as a variable. Additionally, we were not afforded any information on serum metabolites or the metabolic status of each donor. There are 22 structurally similar FGF ligands that mediate effects through activation of receptor tyrosine kinases (RTK), fibr oblast growth factor receptors (FGFR) 1 - 4. FGFs can have affinities for more than one receptor and each receptor can bind multiple FGFs. Similarities between receptors has resulted in receptor redundancy as they can converge on key downstream signaling ca scades 145,146 . All four FGFRs activate 145 . Activation of these pathways play important roles in migration, survival, differentiation, and proliferation 146 . However, studies suggest the strength and specificity of each signaling cascade can vary depending on the type of FGFR and FGF 145 . Additionally, pharmacological ly inhibiting FGFR1 partially attenuated VAT - stimulated mammary epithelial transformation (p=0.02) (Figure 2. 1). This partial attenuation suggests either ligands, receptors, and/or signaling cascades are influencing colony formation or the FGFR1 Ab does no t have complete inhibition of FGFR1. Furthermore, we found FGF1 and FGF18 induces colony formation, but not in the concentration - dependent manner as seen with FGF2 (Figure 2.2 ). One study using MCF - 10A cells revealed phenotypic distinctions in 3D growth s timulated by different RTKs, including EGFR and MET 147 . Other FGFRs or RTKs activated by other ligands could be inducing signaling pathways that attribute to VAT - stimulated transformation 146 . Therefore, w hile other RTKs could contribute to transformation, these data suggest biased agonism associated with FGF2/FGFR1 is optimal for VAT - stimulated MCF - 10A transformation. 35 HuFTF from donor 02 had the highest level of FGF2 at 1.14 pg/mL. Independently FGF2 requ ired at least 10 ng/mL to stimulate transformation of MCF - 10A cells. This could be due to a potential difference in the FGF2 isoform in the HuFTF compared to the recombinant protein. FGF2 exists in five different isoforms that are divided into two groups l ow molecular weight (LMW) and high molecular weight (HMW) proteins. Studies have suggested there are distinct biological activities of LMW and HMW proteins. For example, one study showed that overexpressing LMW FGF2 enhanced bone mineral density (BMD) wher eas overexpressing HMW FGF2 lowered BMD 148 , and another study showed LMW FGF2 suppressed hepatic fibrosis and HMW enhanced hepatic fibrosis 149 . In contrast, other studies ha ve showed FGF2 isoforms exhibit different potencies. Kole et al . showed that all FGF2 isoforms exhibited mitogenic activity in dermal fibroblasts, however, HMW isoforms were less efficient 150 . Additionally, a study by Mydlo et al . showed that FGF2 derived from omental VAT demonstrated greater mitogenic and angiogenic activity than FGF2 derived from either benign and cancerous renal ti ssue 151 . The recombinant protein used is an LMW FGF2 (18 kDa), and the ELISA used to detect FGF2 is nonspecific regardi ng FGF2 isoforms. Therefore, the recombinant protein might not be representative of the most active type of FGF2 in the HuFTF. Additionally, HMW FGF2 could be more potent than LMW FGF2, accounting for the difference in dose of FGF2 in HuFTF and the recombi nant FGF2 used. Currently, only LMW FGF2 is commercially available, and isolating FGF2 from HuFTF would provide a more accurate representation of the transformative capabilities of VAT - derived FGF2. FGF2 is classically considered to function in both an aut ocrine and paracrine manner, however, our research suggests FGF2 functions in an endocrine manner acting on distant targets. Our previous study showed an induction of serum FGF2 in HFD - fed mice compared to LFD - fed mice 79 . This serum induction was depleted following lipectomy of VAT su ggesting the circulating levels of FGF2 are of adipose tissue origin 79 . Circulating levels of FGF2 in these animals were associated with UVB - induced squamous cell carcinomas, suggesting that FGF2 secreted from VAT influences tumor promotion at distant sites 79 . In tandem, one study found FGF2 concentrations in ser um increased with higher BMIs 77 , and another found plasma FGF2 levels of obese Chinese men were correlated with adipose tissue mass 78 . Our recent study demonstrates that human serum samples with elevated FGF2 had greater efficacy in stimulating JB6 P + cell growth in soft agar. Future studies are needed to assess circulating FGF2 concentrations in relation to 36 visceral obesity and breast cancer risk to ascertain potential associations and a role for FGF2 as an endocrine growth factor and as a biomarker for at risk individuals. Obesity and breast cancer are independently complex diseases with multiple factors potentially influencing their etiology. FGF can be secreted from many different tissues along with other FGFR1 ligands. For example, FGF2 is secreted fr om skin 152 , heart 153 , liver 149 , lungs 154 , and SAT 155 and could contribute to circulating FGF2 levels. The contribution of FGF2 from VAT and other sourc es to mammary tumorigenesis will be determined in vivo in future investigations. Although VAT is more strongly correlated with breast cancer risk than its subcutaneous counterpart, there is an intimate and bidirectional interaction between mammary epithelium and adjacent subcutaneous mammary adipose tissue (M AT). The total absence of MAT in transgenic mice prevents non - tumorigenic mammary gland development and MAT supports and amplifies breast cancer progression 156 . Dialog between MAT and mammary epithelium might persist and influence breast cancer onset as a potential source of FGF2 156,157 . This exposes a limitation in our study as assessing VAT and mammary epithelial cells in our in vitro m odel does not evaluate whole body effects in vivo . In summary, we demonstrate FGF2 from human VAT stimulates transformation of non - tumorigenic mammary epithelial cells. Our data suggests differences the transformative ability of human VAT is associated with FGF2 levels and that inhibiting FGFR1 attenuated this transformation. These findings highlight FGF2/FGFR1 signaling as a potential link between visceral adiposity and elevated breast cancer risk. Future studies will use in vivo mouse models to determine the tumorigenicity of transformed MCF - 10A cells, th e ability of HFD to promote mammary tumorigenesis, and the effect of lipectomy on mammary tumorigenesis. FGF2/FGFR1 signaling could be a therapeutic target for breast cancer prevention strategies and/or a biomarker for identifying at risk individuals. Met hods Cell Culture MCF10A cells (human mammary epithelial cells) were obtained from ATCC (Manassas, VA, USA). penicillin/streptomycin, 100 ng/mL cholera toxin, 20 ng/mL epidermal growth fa ctor (EGF), 10 µg/mL insulin, 37 0.5 mg/mL hydrocortisone, 7.5% sodium bicarbonate, 15mM HEPES, and 2 mM L - Glutamine (growth media). MCF - with 20% horse serum and antibiotics ( resuspensions media). The FGFR1 Ab was purchased from R&D Systems (Minneapolis, MN, USA #MAB765). Human fat tissue filtrate VAT was homogenized in equal volume of serum free media on ice for 30 seconds using Tissue Ruptor (Qiagen, Hilden, Germany) on med ium speed. Homogenates were filtered through a hanging transwell insert 15 - mm wide 0.4 um filter (Millicell, cat# MCHT06H48) positioned in 6 - serum free media. Plates were incubated on a rocker for 1 hour to allow fat deriv ed factors to diffuse into the media. After incubation, filtrates were centrifuged at 4500 rpm for 5 minutes and the supernatant was of HuFTF was used for respective experiments. Anchorage - Independent Colony Formation Soft Agar Assay MCF10A cells were seeded at 750 cells per well in a 24 - HS, and 0.33% agar with or without HuFTF and/or inhibitors which was overl aid onto 350 µL of before 200 µL of growth media was gently added to each well and then stored at 37°C. Every 3 - 4 days, the growth media was removed from ea ch well and replenished with 200 µL of growth media. After two weeks, the colonies were fixed in 70% ethanol (EtOH) and stained with 150 µL of 0.01% crystal violet. Colonies were visually counted and used to calculate the percent of colony formation from t he number of cells plated ([Colonies counted x 100] / 750 cells). The % colony formation was then normalized to the control to determine the increase in % of colony formation. (% Colony formation of treatment the % colony formation of untreated control). 38 FGFR1 Ab Treatment FGFR1 Ab was added directly into the top layer of the soft agar assay. MCF - 10A cells were pre - incubated with the monoclonal FGFR1 Ab (2 µg/mL) in 37°C for 1.5 hours before being added to the top soft agar layer. Statistics Six biol ogical and three technical replicates were used to ensure adequate power to detect a significant change in growth in soft agar. Data are presented as mean ± s.e. Unpaired t - test and one - way ANOVA for multiple comparisons were used appropriately. For all st atistical tests, 0.05, 0.01, and 0.001 level of confidence, were accepted for statistical significance. FGF2 Quantification: FGF2 concentrations in HuFTF was measured by ELISA kit according to the FGF2 concentration was 0.625 pg/mL. 39 CHAPTER 3 Identifying chemopreventive agents for obesity - associated cancers using an efficient, 3D high - throughput transformation assay Published in Scientific Reports 2019, 9(1):10278 1 Vanessa Benham 1 , Blair Bullard1, Thomas S Dexheimer 1 , Matthew P Bernard 1 , Richard R Neubig 1,2 , Karen T Liby 1 & Jamie J Bernard 1,2 1 Department of Pharmacology and Toxicology, Michigan State University, East Lansing, MI 48824 2 Division of Dermatology, Department of Medicine, College of Human Medicine, Michigan State University, East Lansing 40 Abstract Obesity is associated with ~40% of cancer diagnoses but there are currently no effective preventive strategies, illustrating a need for chemoprevention. We previously demonstrated that fibroblast growth factor 2 (FGF2) from adipose tissue stimulates malignant transformation, as measured by growth in soft agar, the gold - standard in vitro transformation assay. Because the soft agar assay is unsuitable for high throughput screens (HTS), we developed a novel method using 3D growth in ultra - low attachment conditions as an alternative to growth in agar to discover compounds that inhibit transformation. Treating non - tumorigenic, skin epi thelial JB6 P + cells with FGF2 stimulates growth in ultra - low attachment conditions analogous to growth in the soft agar. This transformation HTS identified picropodophyllin, an insulin growth factor 1 receptor (IGF1R) inhibitor, and fluvastatin, an HMG - Co A reductase inhibitor, as potential chemopreventive agents. These compounds were validated for efficacy using two non - tumorigenic cell lines in soft agar. Another IGF1R inhibitor and other statins were also tested and several were able to inhibit growth in soft agar. This novel 3D HTS platform is fast, robust and has the potential to identify agents for obesity - associated cancer prevention. Introduction Obese individuals are a specific high - risk population that would benefit from targeted chemoprevention s trategies. Obesity is associated with 13 different types of cancers and is associated with over 40% of cancer diagnoses 47 . Because of the swelling obesity epidemic, over 38% of American adults were considered overweight/obese in 2014 158,15 9 . Rising obesity rates illustrate an immediate need for effective primary prevention strategies for obesity - associated cancers. Current prevention strategies like lifestyle changes are ineffective due to non - compliance. Although there are theories to explain the obesity - cancer association, the underlying mechanisms are poorly elucidated. While investigating this mechanism, our studies demonstrated that circulating fibroblast growth factor 2 (FGF2) from visceral adipose tissue (VAT) stimulates fibroblast growt h factor receptor 1 (FGFR1) on epithelial cells to drive malignant transformation 79,160,161 . We demonstrated that VAT depleted of FGF2 failed to transform epithelial cells and epithelial cells lacking FGFR1, the primary receptor for F GF2, also failed to exhibit VAT - induced transformation 79 . Although FGF2 is considered a paracrine growth factor, its circulating levels correlate with 41 adipose tissue mass in humans and are at levels sufficient to induce transformation in vitro 78,79 . This suggests that VAT - secreted FGF2 has a systemic role. In animals, we found that a high - fat diet increased serum FGF2 levels and that removing VAT via lipectomy depleted se rum FGF2 levels 79 . Therefore, the FGF2/FGF R1 signaling axis is a potential chemopreventive target for obesity - associated epithelial cancers. Compounds that lower cancer risk by delaying or preventing cancer, whether synthetic or natural, have the potential to significantly reduce cancer burden 51 . As the World Health Organization estimates that 30 50% of all cancer cases are preventable, lowering cancer incidence will in turn lower cancer mortality 84 . Effective chemopreventive agents either eliminate premalignant cells or protect normal/initiated cells from undergoing malignant transformation (i.e. the changes a non - transformed or normal cell undergoes to become carcinogenic) 83,84 . There are, however, drawbacks to implementing chemoprevention 162 . As these agents would benefit only a subset of the treated population, they must be efficacious and have minimal to no side effects so that they are tolerated for a long duration 163 . Cumulatively, these improbable standards are a challenge for the development and implementation of chemoprevention. Applying an approach of precision medicine to chemoprevention can help overcome these shortfalls 83 . Target - driven strategies to risk - stratify individuals would reduce strict restraints on side effects because the clinical benefits would presumably outweigh the risks 83 . Therefore, implementing targeted chemoprevention strategies in conditions such as obesity has the potential to greatly reduce cancer burden. Methodological limitation s make identifying compounds that prevent obesity - associated cancers challenging. Transformation is commonly modeled by the soft agar assay, a well - established technique that measures transformation by assessing anchorage - independent growth 122,164 - 167 . Non - transformed cel ls must be anchored to an extracellular matrix (ECM) or a treated - culture plate to proliferate. In contrast, tumorigenic cells, which have undergone transformation, lose their anchorage - dependence 168 . 3D models like the soft agar assay do not provide an ECM - like environment, so cells suspended in soft agar only proliferate and form colonies if they are transformed 168 . This gives 3D models a distinct advantage for modeling transformation over 2D culture methods and can disting uish transformed and non - transformed cells 168,169 . We previously used the soft agar assay to show that factors derived from VAT as well as FGF2 itself can stimulate epithelial cells to transform 79,160,161 . This experiment can identify chemopreventive compounds that prevent or inhibit FGF2 - stimulated transformation, but the soft agar assay is unsuitable for 42 high - throughput sc reening because it is laborious and inefficient, using a 6 - 24 well format with a 2 - week incubation period. Identifying compounds for precision chemoprevention requires a targeted high - throughput screening platform that models malignant transformation of no n - transformed cells. However, this is an underdeveloped area of cancer prevention as current models use cancer cells lines and 2D culture. Currently, there are not any high throughput assays that model the process of transformation 168,169 . Therefore, the objective of these studies was to develop a high - throughput 3D model of transformation to screen for chemopreventive agents and then validate hits in soft agar with two cell lines, JB6 P + and MCF - 10A. The se are, respectively, non - tumorigenic mouse epidermal cells and non - tumorigenic human breast epithelial cells. In a recent publication, Rotem et al. describe an HTS assay (384 - well) where growth in ultra - low attachment conditions in a round bottom plate (3 D growth) is strongly correlated to growth in soft agar 166 . Investigators measured growth of non - transformed and transforme d cell lines in soft agar and in low attachment conditions and demonstrated a strong correlation with an R 2 value of 0.82 166 . In this manuscript, we used growth in ultra - low attachment conditions to develop a novel phenotypic transformation HTS assay using FGF2/FGFR1 signaling as the target - based mechanism to identify chemopreventive agents for obesity - associated epithelial canc ers. Screened compounds that prevent growth in ultra - low attachment conditions may further the mechanistic understanding of malignant transformation and have the potential to be developed as precision chemopreventive therapies. While this screen could iden tify novel inhibitors of FGF2/FGFR1 signaling, it also has the potential to find compounds that interrupt the transformation process itself. Results Development and optimization of transformation HTS The transformation HTS was developed using JB6 P + mouse skin epithelial cells. JB6 P + cells cannot proliferate in an anchorage - independent manner, but they have the ability to transform upon treatment with tumor promoters . 16 7 JB6 P + cells are used below passage 15 to prevent spontaneous transformation. We previously demonstrated that visceral adipose tissue (VAT) - derived FGF2 stimulates JB6 P + cell growth in soft agar (transformation) FGF2 was used in the HTS assay to stim ulate transformation . 79 Using mouse fat 43 tissue filtrate (MFTF), a filtrate that contains factors from the VAT of high - fat diet (HFD) - fed mice, might be more physiologically relevant and constitute a screen with better face validity to identify chemopreventive compounds in obesity - assoc iated transformation. However, using biological matrices in HTS often result in significant variability. Therefore, the assay was developed and optimized with FGF2 as the stimulus to increase rigor and reproducibility. We used CellTiter - Glo (Promega) to qu antitate the ATP content in the cells, a surrogate marker for proliferation, as the amount of ATP is generally proportional to the number of cells (Figure 3. 1A). The transformation HTS assay parameters, including number of cells plated per well, incubatio n time, and FGF2 concentration were optimized. To determine the optimal cell number, JB6 P + cells were plated in 384 - well round - bottom, ultra - low - attachment plates at 50, 100, and 200 cells/well with FGF2 at 30 ng/mL. 200 cells/well gave an optimal z - facto r of 0.56 (Figure 3.1B). Using 200 cells/well, a concentration response study was performed. FGF2 at 30 ng/mL gave the optimal z - factor of 0.64 (Figure 3.1C). The highest concentration of FGF2 (100 ng/mL) increased the variability, decreasing the Z - factor (Figure 3.1C). Figure 1D illustrates FGF2 - stimulated growth over a 96 - hour period, whereas untreated or vehicle - treated JB6 P + cells congregate at the bottom of the well and fail to proliferate (Figure 3.1D). Overall, the transformation HTS was developed in a 384 - well, round bottom, ultra - low attachment plates and optimized to have 200 JB6 P + cells/well, 30 ng/mL of FGF2, 96 - hour incubation at 37°C, resulting in a Z - factor of 0.503 over eight independent assays. DMSO, the solvent for the compound libraries , was used below 0.05% because higher concentrations negatively impacted the Z - factor. Therefore, the compounds were screened with a final DMSO concentration of 0.025%. 44 Figure 3. 1. Optimization and development of the transformation HTS. A) Methodology schema of the transformation HTS. JB6 P + cells were plated in 384 - well round bottom ultra - low attachment conditions and stimulated with FGF2. JB6 P + cell growth in ultra - low attachment conditions was measured with CellTiter - Glo that gives a lum inescent signal stimulated by binding to ATP. ATP levels are proportional to the number of cells, and thus used as a measure of proliferation. B) JB6 P + cells were plated in 384 - well round bottom, ultra - low attachment plates and treated with 30 ng/mL of FG F with either 50, 100, or 200 cells/well. 200 cells/well was the optimal cell density giving a Z - factor of 0.56. Fifty (50) and 100 cells/well gave Z - factors of - 0.023 and 0.47 respectively. Each treatment group had 16 technical replicates. C) A concentrat ion response of FGF2 was performed with JB6 P + cells at 200/cells per well. The EC 50 was 15 ng/mL, however, 30 ng/mL gave the optimal Z - factor 0.644. Each treatment had 7 technical replicates. The concentration response stimulation was statistically analyz ed using a nonlinear regression, dose - response with PRISM. D) FGF2 - stimulated transformation of JB6 P + cells can be visually observed with JB6 P + cells at 200 cells/well with or without FGF2 at 30 ng/mL over the 96 - hour incubation. JB6 P + cells congregate at the bottom of the round - bottom wells and untreated, do not grow, but with FGF2, do proliferate. E) PD166866, - stimulated JB6 P + cells growth in ultra - low attachment conditions. Untreated (Unt.) and F GF2 controls had 16 technical replicates and PD166866 had seven technical replicates. F) PD166866 attenuates FGF2 (0.5 ng/mL) - stimulated JB6 P + cell growth in soft agar. The soft agar assay was performed as described in Material/Methods. G) PD166866 attenu ates FGF2 (20 ng/mL) - stimulated MCF - 10A cell growth in soft agar. 45 Figure 3.1 The soft agar assay was performed as described in Material/Methods. Data are presented as mean ± S.D., statistical significance was determined using a one - way ANOVA, mu ltiple comparisons (**p<0.01, ***p<.0001). H) IBET concentration - dependently inhibits FGF2 stimulated growth in ultra - low attachment conditions. JB6 P + cells were plated at 200 cells/well with FGF2 at 30 ng/mL. The IC50 of IBET inhibition is 0.12 µM. Each treatment had seven technical replicates. The concentration response inhibition was statistically analyzed using a nonlinear regression, dose - response with PRISM. I) JB6 P + cells were plated in 384 - well round - bottom, low attachment plates (100 and 200 cell s/well) and incubated at 37ºC for 48 hrs. FGF2 at 30 ng/mL significantly stimulates JB6 P + cell growth in low attachment conditions compared to untreated controls (Unt.). Each treatment had seven technical replicates and data was analyzed by one - way ANOVA. Data are presented as mean ± S.D. J) JB6 P + cells were plated in 384 - well flat - bottom, cell culture treated plates (100 and 200 cells/well) and incubated at 37ºC for 48 hrs. FGF2 at 30 ng/mL did not stimulate growth in 2D culture conditions compared to u ntreated controls (Unt.). Each treatment had seven technical replicates and data was analyzed by one - way ANOVA. Data are presented as mean ± S.D. (**p<0.01, ***p<.0001) NS, not significant. 46 FGFR1 is critical in FGF2 - stimulated transformation The transformation HTS was developed to be a FGFR1 target - based, phenotypic screen. Therefore, this assay can identify hits that act directly on FGF2 and/or FGFR1, as well as ones that target the transformation process, including any part of the FGF2/FGFR1 signaling axis that promotes transformation. To demonstrate that FGFR1 is critical in FGF2 - stimulated growth in ultra - low attachment conditions, we used PD166866, an FGFR1 - selective inhibitor (SelleckChem), in the screen as a positive contr ol. PD166866 at 0.5 µM completely inhibited FGF2 - stimulated transformation in the HTS (Figure 3. 1E), demonstrating that compounds which inhibit FGF2/FGFR1 signaling will inhibit JB6 P + cell transformation. This result was validated in soft agar with both J B6 P + and MCF - 10A cells. PD166866 significantly inhibited JB6 P + colony formation at 0.2 and 0.5 µM and MCF - 10A colony formation at 2.5, 5.0 and 10 µM (Figure 3. 1F, G). These data demonstrate that FGFR1 is critical in FGF2 stimulated epithelial cell transf ormation. To demonstrate efficacy and feasibility of this assay, we also tested I - BET - 762 (IBET), a bromodomain inhibitor, which we have previously shown to prevent FGF2 - stimulated transformation both in vitro and in vivo 170 . IBET concentration - dependently prevented FGF2 - stimulated growt h in low attachment conditions (Figure 3. 1H). 2D vs 3D growth Cell growth in 2D culture (proliferation) is not mechanistically analogous to cell growth in 3D culture (transformation) 168,169 . To demonstrate this, JB6 P + cells were plated in conventional 384 - well flat - bottom, cell culture plates at 100 or 200 cells/well, with or without FGF2 (30 ng/mL) for 48 hours (Figure 3. 1D). JB6 P + cells were also plated in 384 - well round - bottom, ultra - low attachment plates at 100 or 200 cells/well, with or without FGF2 (30 ng/mL) for 48 hours (Figure 3. 1I). 96 hrs in 2D growth was not optimal because all cells are proliferating and reach confluence before the 96 hours, a limitation that does not apply to ultra - low attachment cell growth because the cells are not restrained to the surface area of the well. FGF2 significantly stimulated proliferation in 3D conditions and not in 2D conditions with either 100 or 2 00 cells per well (Figure 3. 1I). FGF2 is a known mitogen that stimulates growth of cancer cells in 2D culture 20 . However, in our assay with non - tumorigenic cells, FGF2 is not a mitogen as proliferation of JB6 P + was not increased by FGF2 (Figure 3. 1J). Thi s corresponds to findings from Rotem et al., which demonstrate that 47 the oncogenic capacity of cells (growth in 3D) is independent of the proliferation rate 16 . Therefore, FGF2 stimulates anchorage - independent growth, a characteristic of malignant cells, whi ch suggests that FGF2 stimulates JB6 P + cell transformation but does not enhance traditional 2D proliferation. Table 3. 1. Funnel Strategy - 2,532 compounds screened Hit Parameter Identified Hits Hit Percentage Transformation HTS - >50% inhibition 178 7.0 Cytotoxicity - <25% decrease in HEK293 cell viability) 105 4.2 Promiscuity - active in <20% assays listed in Pubchem 58 2.3 Commercially available compounds - DTP/CC 33 1.3 Fresh powder confirmation - >40% inhibition 7 0.27 Abbreviations: HEK293 - human embryonic kidney; DTP - Developmental Therapeutics Program; CC - Cayman Chemical 48 Screening of compound libraries Over a thousand compounds from the Prestwick Chemical Library®, the National Cancer Institute (NCI) Natural products library, and the Michigan State University (MSU) Chemistry library of MSU - made analogs of natural products were screened. These libraries w ere used at 0.5 µM, a relatively low concentration to identify chemopreventive agents with a higher potency and to reduce potential for toxicity. The screen gave an average Z - factor of 0.503. Compounds that attenuated transformation 50% or more (which is 3 - 4 standard deviations from the mean) were considered primary hits. Compounds were then assessed for general cytotoxicity and eliminated if HEK293 cell viability was decreased by more than 25% at 10 µM (MSU screening core, unpublished). Next, compounds wer e assessed for promiscuity and eliminated if they demonstrated activity in more than 20% of the bioassays listed in PubChem, determined by number of hits/total assays screened (data acquired in August 2017). Promiscuous compounds are problematic for chemop revention due to the increased potential for side - effects. These parameters narrowed hits down to 58 compounds for validation and concentration response (Table 3. 1). Hit confirmation and soft agar validation Fresh powder was obtained from the Developmenta l Therapeutics Program (DTP) at the NCI or commercially (Cayman Chemical) for 33 compounds to confirm hits and eliminate false positives. Not all 58 prioritized hits were commercially available for order. Confirmation tests with new powders ensure that inh ibition of FGF2 - stimulated transformation corresponds to the intact compound and not to an impurity or degraded compound in DMSO library stocks which can occur if compounds have undergone several freeze - thaw cycles over time 171 . Of the prioritized hits from the transformation HTS, 2 hits were confirmed: picropodophyllin (PPP) and fluvastatin. PPP is an insulin - like growth factor - 1 receptor (IGF1R) inhibitor that is currently undergoing clinical trials as an adjuvant chemotherapeutic. PPP concentration - dependently attenuated FGF2 - stimulated growth in low attachment conditions (Figure 3. 2A). Fluvastatin, a statin drug used for lowering blood cholesterol and triglycerides, concentration - dependently attenuated FGF2 - stimulated growth in low attachment conditions (Figure 3. 2A). Both PPP and fluvastatin have been used clinically with no observed toxicity 172 - 175 . 49 Figure 3. 2. PPP and fluvastatin significantly attenuate FGF2 - stimulated transformation of epithelial cells. A) PPP and fluvastatin concentration - dependently inhibit JB6 P + cell growth in low attachment conditions. Dotted lines indicated FGF2 (top) and untreated controls (bottom). Each treatment had three technical replicates. JB6 P + cells were cultured with the op timized parameters. B) PPP and fluvastatin attenuate FGF2 (0.5 ng/mL) - stimulated JB6 P + cell growth in soft agar at 0.5 µM and 0.2 and 0.5 µM respectively. The soft agar was performed as described in Materials/Methods with three technical replicates. C) PP P and fluvastatin attenuates FGF2 (20 ng/mL) - stimulated MCF - 10A cell growth in soft agar at 1.0 and 2.5 µM and 0.5, 1.0, and 2.5 µM respectively. The soft agar was performed as described in Materials/Methods with three technical replicates. D) PPP and fluv astatin (Fluv) significantly attenuate MFTF (200 µg/mL) - stimulated JB6 P + cell growth (1000 cells/well) in soft agar at 2.5 µM and PD166866 (PD) was significant at 0.5 and 2.5 µM. The soft agar was performed as described in Materials/Methods with three tec hnical replicates. E) PPP and fluvastatin (Fluv) significantly attenuate MFTF (200 µg/mL) - stimulated MCF - 10A cell growth (1000 cells/well) in soft agar at 2.5 µM and PD166866 (PD) was significant at 10 µM. The soft agar was performed as described in Materi als/Methods with three technical replicates. The soft agar assays were analyzed by one - way ANOVA with multiple comparisons (*p<0.05; **p<0.01; ***p<0.001). 50 PPP and fluvastatin were validated in the soft agar assay using both JB6 P + and MCF - 10A cells to demonstrate their efficacy in preventing FGF2 - stimulated transformation. MCF - 10A cells are non - tumorigenic human mammary epithelial cells. We previously showed that MFTF and FGF2 can stimulate MCF - 10A cells to transform 160,170 . Here, PPP and fluvastatin concentration - dependently attenuated FGF2 - stimulated colony formation of JB6 P + (Figure 3. 2B) and MCF - 10A cells in soft agar (Figure 3. 2C), analogous to the inhibition observed in the HTS. JB6 P + cells were stimulated with 0.5 ng/mL of FGF2 and PPP significantly attenuated colony formation at 0.5 µM compared to the FGF2 - treated control; fluvast atin also significantly attenuated colony formation at 0.2 and 0.5 µM (Figure 3. 2B). MCF - 10A cells were stimulated with 20 ng/mL of FGF2 and PPP significantly attenuated colony formation at 1.0 and 2.5 µM; fluvastatin significantly attenuated colony format ion at 0.5, 1.0, and 2.5 µM (Figure 3. 2C). Overall, these data demonstrate that compounds discovered through the transformation HTS also inhibit colony formation in the gold - standard soft agar transformation assay. The objective of these studies is to ide ntify chemopreventive compounds for obesity - related cancers. Therefore, as a more relevant stimulating with FGF2, we tested the efficacy of PPP and fluvastatin for preventing MFTF - stimulated transformation. For JB6 P + cells were stimulated with 200 µg/mL o f MFTF, both PPP and fluvastatin significantly attenuated colony formation at 2.5 µM (Figure 3. 2D). The FGFR1 inhibitor PD166866 also significantly attenuated colony formation at 0.5 and 2.5 µM. For MCF - 10A cells stimulated with 200 µg/mL of MFTF, PPP and fluvastatin significantly attenuated colony formation at 2.5 µM and PD166866 showed significant inhibition at 10 µM (Figure 3. 2E). Picropodophyllin and fluvastatin do inhibit 2D proliferation Many anti - cancer agents inhibit cell proliferation and induce c ell death, however, for chemoprevention, compounds need to exhibit little to no toxicity. To determine if PPP and fluvastatin affected cell proliferation and viability. JB6 P + ture dye binds to cellular proteins containing primary amines. As the cells divide, the dye evenly distributes to both daughter cells. Reduction of the me dian fluorescence intensity (MFI) by approximately half indicates 51 Figure 3. 3. PPP, fluvastatin or PD166866 do not inhibit 2D proliferation of JB6 P + cells. Histograms of JB6 P + (A) PPP, (B) fluvastatin, PD166866 (C) or the vehicle, DMSO. Inhibition of proliferation is indicated by higher fluorescent signal that is represented by visually distinct rightward s hifted histograms as shown following treatment with mitomycin C (MmC) and cycloheximide (CHX). The same histograms for DMSO, MmC (1.0 µM), and CHX (10 µg/mL) are shown for comparison with PPP, fluvastatin, and PD166866. ( D) + ce lls fluvastatin, or PD166866 was graphically depicted. PPP, fluvastatin, and PD166866 did not significantly cle (DMSO) controls, whereas MMC and CHX (positive ( E) PPP, fluvastatin, or PD166866 did not significantly decrease the % of live cells (singlet, SYTOX Red negative cells). MmC and CHX significantl y induced cell death at 0.1 and 1.0 µM and 10 and 50 µg/mL respectively. Each treatment group had two replicates and analyzed by one - way ANOVA with multiple comparisons (**p<0.01; ***p<0.001). 52 a successful division. Therefore, cells treated with compounds that inhibit proliferation would have a higher i n cells that have undergone proliferation. PPP, fluvastatin, and PD166866 had minimal to no effect on JB6 P + between these treatments at any concentration and vehicle (DMSO) treatment. (Figure 3. 3A - C). Mitomycin C (MmC), a chemotherapeutic agent that inhibits DNA synthesis, and cycloheximide (CHX), a protein synthesis inhibitor, are shown as positive controls. The same histograms for vehicle (DMSO), MmC at 1.0 µM, and CHX at 10 µg/mL are shown for comparison with PPP (Figure 3. 3A), fluvastatin (Figure 3. 3B), an d PD166866 (Figure 3. 3C) demonstrating the substantial shift expected when cell proliferation is strongly 3. 3D. Figure 3. 4. IGF1R inhibitors prevent FGF2 - stimulated transformation whereas efficacy of HMG - CoA reductase inhibitors varies. A) Fluvastatin (Fluv), simvastatin (Sim), rosuvastatin (Rosu), PPP, and NVP - ADW742 (NVP) significantly attenuated FGF2 (5 ng/mL) - stimulated transformation of JB6 P + cells (2,000 cells/ well) at 5.0 µM. Colonies were counted via automated counting using the Cytation 3 imaging reader from Biotek using Gen5 3.04 software. Seven pictures were taken every 100 microns and superimposed together by the zprojection function. B) Fluvastatin (Fluv ), rosuvastatin (Rosu), atorvastatin (Ator), and NVP - ADW742 (NVP) significantly attenuated FGF2 (30 ng/mL) - stimulated transformation of MCF - 10A cells (1,000 cells/well) at 0.5 and 2.5 µM. Simvastatin (Sim) was significant at 2.5 µM. Each treatment group ha d three replicates and analyzed by one - way ANOVA with multiple comparisons (*p<0.05; **p<0.01). 53 48 - hour incubation before flow cytometric acquisition of sam ples to determine whether any of the compounds induced cell death. PPP, fluvastatin, and PD166866 did not significantly induce cell death, as the percent of SYTOX Red negative cells were similar to DMSO vehicle (Figure 3. 3H). Both MmC and CHX significantly induce cell death compared to DMSO - treated controls (0 µM), as indicated by reduction in the percent of SYTOX Red negative JB6 P + cells. Overall, PPP, fluvastatin, and PD166866 did not influence JB6 P + cell division or cell viability. Efficacy of statins and IGF1R inh i bitors To determine if the mechanisms identified by PPP and fluvastatin were compound - specific or applicable to similar classes of drugs, NVP - ADW742 (NVP), an IGF1R inhibitor, and 4 statins (simvastatin, pravastatin, rosuvastatin, ato rvastatin) were evaluated for efficacy for preventing FGF2 - stimulated transformation. PPP and NVP both inhibited FGF2 - stimulated significantly attenuated colony formation. For JB6 P + cells stimulated with 5 ng/mL of FGF2, both PPP and NVP significantly att enuated colony formation at 5.0 µM (Figure 3. 4A). For MCF - 10A cells stimulated with 30 µg/mL of FGF2, NVP significantly attenuated colony formation at 0.5 and 2.5 µM (Figure 3. 4B). Fluvastatin, simvastatin, and rosuvastatin significantly inhibited FGF2 - sti mulated JB6 P + colony formation at 5.0 µM (Figure 3. 4A). Fluvastatin, rosuvastatin, and atorvastatin significantly prevented FGF2 - stimulated MCF - 10A colony formation at 0.5 and 2.5 µM (Figure 3. 4B). Simvastatin significantly prevented FGF2 - stimulated MCF - 1 0A transformation at 2.5 µM (Figure 3. 4A). Discussion Primary prevention for high - risk patients is arguably the most effective tool for mitigating cancer burden. More effective chemotherapeutics and precise molecular targeting have contributed to a substantial decrease in worldwide mortality from 2005 to 2015 176 . In contrast, cancer incidence has steadily increased from 2005 to 2015, rising 33% and affecting more than 14 million people in 2015 176,177 . The cancers investigated in this manuscript, breast cancer and skin cancer, are two of the ten cancers that are still increasing in incidence 178,179 . Interestingly, obesity - associated cancers are preferentially on the rise, and 54 account for 40% of cancers diagnoses 47 . These epidemiological studies illustrate that cancer is still a major h ealth burden and suggests a need to prevent obesity - associated cancer incidence. We previously demonstrated a role for FGF2/FGFR1 in visceral obesity - associated epithelial cell transformation 79 . Therefore, using this mechanism, we aimed to develop an HTS method to screen for compounds that have the potential to prevent or attenuate FGF2/FGFR1 - stimulated transformation. Herein we describe the first HTS assay that models the process of transformation in an in vitro experimental setting. This transformation HTS assay is a tool to identify potential prevention strategies targeting obesity - associated epithelial cancers and if successful, could greatly reduce cancer burden. The transformation HTS is a novel FGF2/FGFR1 target - based, phenotypic screen that can evaluate the ability of compounds t o inhibit transformation. Target - based drug discovery may lack clinical efficacy negate target inhibition 86 . Additionally, using phenotypic approaches with no mechanism of action can identify non - selective agents that influe nce a plethora of processes by acting on other cell types, receptors, or pathways. A screen that utilizes both target - based and phenotypic qualities creates a mechanism - informed, phenotypic screen that overcomes the individual limitations of each screen ty pe and thus can identify compounds that are more likely to be efficacious in vivo 86 . The transformation HTS can identify hits that act directly on FGF2 an d/or FGFR1 or can target critical factors in the FGF2/FGFR1 signaling axis required to stimulate transformation. This HTS can identify potential chemopreventive agents and serve as a tool to further elucidate the mechanisms of transformation. PD166866, a s elective FGFR1 inhibitor, inhibited FGF2 - stimulated transformation, in both ultra - low attachment conditions and the soft agar assay, which confirms FGFR1 as the critical receptor in FGF2 - stimulated transformation (Figure 3. 1E - G). Furthermore, the transform ation HTS identified PPP and fluvastatin as effective chemopreventive agents (Figure 3. 2). However, based on our knowledge based on published literature, these two compounds do not directly interact with FGF2 or FGFR1. Therefore, compounds like PPP and flu vastatin can be used to gain mechanistic understanding on how FGF2/FGFR1 stimulates transformation. Screening more compound libraries should reveal more efficacious compounds that would further aid in elucidating the mechanisms of obesity - associated transf ormation. 55 The transformation HTS method is the first to stimulate a non - tumorigenic cell to transform in vitro while overcoming limitations of the soft agar assay and 2D cell proliferation/apoptosis assays for chemoprevention drug discovery. The soft agar assay in its traditional 6 - 24 well plate format is laborious, inefficient, and costly, and is not usable for high - throughput screening 79,180,181 . However, studies have scaled up the soft agar assay to a 96 - or 384 - well format to screen for chemotherapeutic compounds 182 - 184 . For example, Horman et al . developed an HTS - compatible 3D colony formation assay in a 384 - well plate by incubating 150 HCT116 human colorectal carcinoma cells with compounds for 5 days, quantifying colonies with a laser - scanning fluorescence cytometer 183 . Methods such as this successfully identify chemotherapeutics but are not suitable for identifying chemopreventive agents for two reasons. First, these soft agar assays use cancer cell lines which are functionally different from non - transformed cells regarding activated/inhibit ed signaling pathways, changing the druggable pathways in each cell type. Because the methodology for true prevention assays is underdeveloped, inhibiting colony formation of transformed cells is used to investigate both chemotherapeutic 185 and chemopreventive 186 compounds, highlighting a need for more effective models for chemoprevention that target the transformation process. Second, using non - transformed cell lines in a high - throughput soft agar assay results in only a small fraction of these cells forming colonies over the course of 10 - 14 days making visualizing colony formation and inhibition a challenge. Furthermore, achieving a Z - factor of 0.5 or above would be highly improbable. While traditional 2D prevention assays are advantageous because the y are easily scaled up to 384 - 1536 - well formats, mechanisms of 2D growth are different from mechanisms of 3D growth 169 . For exam ple, Figure 3. 1I - J demonstrates that FGF2 stimulates 3D growth of JB6 P + cells, but not 2D growth. Moreover, these traditional 2D prevention assays use cancer cells. Using cancer cells to gain mechanistic insight may not be fully representative of the mech anism(s) to prevent carcinogenesis 185,186 . One limitation to the transformation HTS, as wi th many anti - cancer screens, is that cytotoxic compounds will show up as hits. For example, Roridin A and CHX, cytotoxic compounds that inhibit protein synthesis, inhibited 3D growth by more than 60% in the initial screen and were selected as primary hits. Since our libraries consisted of known and/or FDA approved compounds, those compounds could be eliminated based on published/known toxicities in non - 450 proliferation experiment as a control that inhibits 2D proliferation (Figure 3. 3). For unknown or novel 56 compounds, following cytotoxicity studies should be performed to ensure that compounds are preventing the process of transformation and are not simply inducing cell death or inhibiting vital cell functions l ike protein synthesis. Our transformation HTS resulted in the identification of picropodophyllin (PPP) as an inhibitor of FGF2 - stimulated growth in ultra - low attachment conditions. PPP is a cyclolignan alkaloid from the mayapple plant family. PPP has been suggested to have anti - neoplastic activity by inhibiting IGF1R 187 , a receptor tyrosine kinase that is a key regulator of energy metabolism ant tumor growth. PPP inhibits the IGF1R by inducing the activation loop - specific inhibition of tyrosine phosphorylation. Although it has been suggested that PPP is specific for IGF1R at nanomolar concentrations, it is unknown if PPP at micromolar concentrations will inhibit FGF1R, a mechanism that will be explored in future studies. A role of IGF1R itself in transformation is supported by recent investigations into metformin as a chemopreventive agent. Metformin inhibits insulin like growth factor 1 (IGF1)/IGF1R signaling 188 . Furthermore, an additional IGF1R inhibitor, NVP - ADW 742 attenuated FGF2 - stimulated transformation of JB6 P + cells suggesting a potential role for IGF1R in FGF2/FGFR1 - driven transformation. Interestingly, there are elevated circulating and tissue levels of both insulin growth factor and insulin in obesity, s uggesting that a combination of elevated growth factors may increase cancer risk 174,188 . PPP has demonstrated both safety and efficacy in clinical studies and mous e models of tumorigenesis. In a phase I/II trial of four patients with squamous cell lung carcinoma, PPP treatments induced necrosis in the tumor and disease progression was halted for seven months 172 . None of the patients in this study showed dose - limiting toxicity 172 . These studies showed that PPP is a potential chemotherapeutic and has good tolerabi lity 172 . In vivo mouse models demonstrated that PPP decreased tumorigenesis with no associated toxicity. In a mouse model of Benzo(a)pyrene (BaP) - induced lung tumorigenesis, PPP decreased tumor volume, increased apoptosis (caspase - 3) and decreased proliferation (Ki - 67) in the tumor 173 . Additionally, these A/J mice were treated with PPP once a day, five t imes a week for 20 weeks and there were no changes in body weight and no overt side effects 173 . In another study using a xenograft model of multiple myeloma (MM), PPP was subcutaneously administered to mice with established MM tumors. PPP significantly decreased tumor burden and inhibited tumor - associated a ngiogenesis and osteolysis. PPP also significantly prolonged the life of the mice from 100 days to 150 days 174 . It is important 57 to note that current published studies examine effects of PPP on established tumors, whereas this manuscript investigates PPP as a chemopreventive agent, we test its ability to prevent the process of transformation, revealing a new clinical target for prev ention that has not been previously explored. Collectively, these studies show that PPP has oral clinical efficacy in humans and overall is well tolerated, suggesting that PPP has the potential to have utility for cancer prevention. Our transformation HTS also identified fluvastatin as an inhibitor of FGF2 - stimulated growth in ultra - low attachment conditions in the primary screen. Fluvastatin is one of several 3 - hydroxy - 3 - methylglutaryl coenzyme A (HMG - CoA) reductase inhibitors, cholesterol lowering agents that treat dyslipidemia and prevent cardiovascular disease 189 . Statins work by competitive inhibition of HMG - CoA reductase, the rate - limiting step in cholesterol biosynthesis, causing reductions in cholesterol and low - density lipoproteins (LDL) and an increase in hig h - density lipoproteins (HDL), that carry cholesterol from other parts of the body to the liver for removal 189,190 . Fluvastatin is a good candidate for chemoprevention because it has a favorable safety profile and has been shown to have anti - cancer activity 190 . Fluvastatin inhibits breast ca ncer cell proliferation and with a greater potency in estrogen receptor (ER) negative breast cancer cells 191,192 . Interestingly, fluvastatin inhibited FGF2 - stimulated transformation of MCF - 10A cells, which are ER negative. Recently, FGFR1 activation was identified as the primary mechanism by which obesity drives estrogen receptor positive mammary tumor progression following endocrine deprivation 193 . These studies suggest that fluvastatin may be efficacious for inhibiting obesity - promoted mammary tumor progression and a potential compound for secondary prevention in obese patients. Epidemiological st udies that evaluate statins and cancer risk have been inconclusive. A 2006 meta - analysis by Browning et al . reviewed the association between statins and cancer risk investigating 42 studies and concluded statins use is not associated with short - term cancer risk. However, these studies had relatively short follow - ups that were too brief to capture a true association between statin use and cancer incidence or mortality 194 . In more recent analyses, Yang et al . (2017) 195 analyzed four articles and came to the tentative conclusion that fluvastatin may reduce breast cancer risk but further high - quality research is needed to confirm this. Likewise, Liu et al . (2017) inves tigated seven studies and suggested that lipophilic statins (like fluvastatin, simvastatin, and atorvastatin) were more protective than hydrophilic stains (like pravastatin and rosuvastatin) but due to high heterogeneity between the studies made it difficu lt 58 to see an advantageous benefit for this population 196 . Ultimately, long - term data is lacking to support the role of statins in primary chemoprevention 197,198 . We investigated four additional statins including simvastatin, rosuvastatin, pravasta tin, and atorvastatin to determine if the effects on colony formation were specific to fluvastatin. Not all the statins attenuated colony formation, nor did efficacy correlate with statin lipophilicity. Additionally, similar effects were observed in both c ell lines, except for atorvastatin which had no effect on colony formation of JB6 P + cells but attenuated colony formation of MCF - 10A cells (Figure 3. 4). Follow - up studies focused on mechanisms underlying the efficacy of statins are warranted. Overall, we optimized a novel HTS of FGF2 - stimulated transformation utilizing growth in ultra - low attachment conditions. This is the first screen to stimulate non - tumorigenic cells to transform in vitro. Additionally, this assay has the potential to be optimized with other tumor promoters such as hepatocyte growth factor, epidermal growth factor, and phorbol esters, as well as with complete carcinogens such as BaP. The transformation HTS identified PPP and fluvastatin as potential chemopreventive agents. After these co mpounds were confirmed to concentration - dependently inhibit FGF2 stimulated transformation, they were validated in soft agar with two non - tumorigenic cell lines, JB6 P + and MCF - 10A cells. Future studies will test these compounds in vivo and evaluate their mechanism of action. Overall, the transformation HTS is a fast, robust and uniquely adept 3D screen to identify potential chemopreventive compounds. Materials/Methods Cell culture JB6 P + cells (mouse skin epidermal cells) were obtained from ATCC (Manassas, VA, USA). Cells were cultured in minimum essential medium (MEM) supplemented with 5% fetal bovine serum (FBS) and 1% penicillin/streptomycin (p/s) (JB6 P + growth media). JB6 P + cells we re trypsinized with 0.05% trypsin and quenched with MEM with 5% FBS and 1% p/s. JB6 P + cells are used below passage 15 prevent spontaneous transformation. MCF - 10A cells (human mammary epithelial cells) were obtained from ATCC (Manassas, VA, USA). Cells wer ng/mL cholera toxin, 20 ng/mL epidermal growth factor (EGF), 10 µg/mL insulin, 0.5 mg/mL hydrocortisone, 59 7.5% sodium bicarbonate, 15mM HEPES, and 2 mM L - Glutamine (MCF - 1 0A growth media). MCF - 10A antibiotics (resuspension media). MCF - 10A cells are used below passage 20 prevent spontaneous transformation. Reagents Mouse FGF2 is a recombinant protein purchased from Prospec (CYT - 386). Fresh powder for were obtained from the DTP (Figure 2). Fluvastatin (10010337), simvastatin (MK - 73 3), rosuvastatin (ZD 4522), pravastatin (10010342), atorvastatin (10493), cycloheximide (14126), and mitomycin C (11435) were ordered from Caymen Chemical. PD166866 (S8493) and NVP - ADW742 (S1088) was purchased from - 0842 - Cell Stain, for 633 or 635 nm (S34859) were purchased from ThermoFisher. I - BET - 762 was purchased from JSTAR Research Inc. Generating fat tissue filtrate Mouse fat tissue filtrate (MFTF) was made as pre viously described 79 . Briefly, the parametrial fat pad was removed from 13 - week - old mice that had been on a HFD for 4 weeks. This fat was placed in a transwell insert above serum free media to collect factors from the fat (MFTF). This fat was used in the soft agar assay at 200 µg/mL. A ll animal experiments were performed in adherence to principles stated in the Guide for the Care and Use of Laboratory Animals (NIH publication, 1996 edition). The IACUC and Animal Care Program of Michigan State University (East Lansing, MI, USA) approved all animal experiments under AUF 02/15 - 027 - 01. Soft agar assay JB6 P + cells were plated at 500 cells/well (or otherwise indicated) in a 24 - well plate in 200 µL of MEM media with 10% FBS in 0.33% agar overlaid onto 350 µL of MEM media with 10% FBS in 0.5% agar. FGF2 60 (Prospec, CYT - 386) was incubated with the cells at 0.5 ng/m L and compared to untreated controls. Soft agar plates were left at room temperature for 30 minutes then incubated at 37°C. MCF - 10A cells were seeded at 750 cells/well (or otherwise indicated) in a 24 - well plate in 200 µL of agar. FGF2 (prospec, CYT - 386) was incubated with the cells at 20 ng/mL and compared to untreated controls. Soft agar plates were left at room temperature for 30 minutes before 200 µL of MCF - 10 A growth media was gently added to each well and then stored at 37°C. Every 3 - 4 days, the growth media was removed from each well and replenished with 200 µL of MCF - 10A growth media. After two weeks, JB6 P + and MC2HitF - 10A soft agar plates were fixed in 70 % ethanol (EtOH) and stained with 150 µL of 0.01% crystal violet. Colonies were visually counted and used to calculate the percent of colony formation from the number of cells plated ([Colonies counted x 100] / number of cells plated). Transformation HTS JB6 P + cells were plated in 384 - well round - bottom low attachment plates at 200 cells/well in 40 µL of JB6 P + growth media. FGF2 (30 ng/mL) was used to stimulate JB6 P + cell growth in ultra - low attachment conditions. Cells were manually added to each plate with a multichannel, repeater pipette. Compounds were plated using a dual arm Biomek F X liquid ha n dl i ng robot with 384 - well pintool liquid handling system. Plates were incubated at 37°C for 96 hours. Then 40 µL of CellTiter Glo (Promega) was added to each well. Plates were shaken at 300 rpm for 5 mins on a plate shaker and then a Biotek synergy Neo HTS Multi - Mode Microplate Reader detected the luminescence signal of eac h well. Untreated cells were used as the negative control. For all compound screening, FGF2 and untreated controls were treated with the vehicle, DMSO. A Z - factor was used to evaluate the quality of a HTS. A Z - factor is a screening window coefficient that qualitatively assesses the ability of a screen to identify active compounds or hits while screening large compounds libraries. The Z - factor considers the means and standard deviation of the positive and negative controls (Z factor = 1 (3( p + n )) / | p n |). A high quality HTS assay has a Z - factor that ranges between 0.5 and 1.0. A Z - factor was calculated during method development to optimize the parameters of the screen to achieve a Z - factor of 0.5 or above during the screening assay. 61 Compound li braries The Prestwick Chemical Library®, the National Cancer Institute (NCI) Natural products library, and a Michigan State University (MSU) library of MSU - made compounds were used for screening in the MSU Assay Development and Drug Repurposing Core (ADDRC). The P restwick Chemical Library® is a unique collection of 1280 diverse, small molecules consisting of mostly FDA approved drugs, with known bioavailability and safety. The NCI Natural products library consists of 419 compounds selected from the Developmental Th erapeutics Program (DTP) open repository and has a variety of scaffold structures with multiple functional groups. The MSU compounds were synthesized by Dr. Jetze Tepe (MSU) and designed to mimic the diverse structural features found in natural products. Cell viability and proliferation assessment by flow cytometry JB6 P + + growth media and were plated at 50,000 cells/well in 6 - well plates. After 24 hours, cells were treated wi th PPP, fluvastatin, PD166866, or DMSO (vehicle control). After an additional 48 hours, cells were prepped for flow cytometric analysis. Cells were trypsinized, washed, resuspended in PBS, and filtered. Prepared JB6 P + ead Cell Stain at a concentration of 5 nM, which was added approximately 15 minutes before analyzing samples on a BD FACS Aria IIu located in the MSU South Campus Flow Cytometry Core. Flow cytometry data was analyzed using FCS Express (DeNovo Software). Vi ability was assessed as a percentage of singlet, SYTOX Red negative cells. Proliferation was assessed + cells (singlet, SYTOX Red negative cells). Statistical analysis Data are presented as mean ± SD. For soft agar experiments, three technical replicates were used and analyzed by one - way ANOVA, multiple comparisons. For HTS method development, Z - factor was used to define optimal parameters giving a quality HTS able to define prim ary hits. For all statistical tests, the 0.05, 0.01, and 0.001 level of confidence was accepted for statistical significance. 62 CHAPTER 4 Investigating the mechanistic connection between visceral adipose tissue and the mevalonate pathway in ER breast ep ithelial cell transformation 63 Abstract Cancer prevention has the potential to drastically reduce the number of breast cancer cases. However, ca ncer prevention strategies are difficult to develop and implement because compliance can be difficult to maintain in a general population because of the prevention paradox. The prevention paradoz states that w hile a population see s the benefits of general i zed prevention, it often times goes unnoticed on an individual basis. Therefore, identifying a high - risk population further encourages compliance resulting in more successful implementation prevention strategies. For example, pre - menopausal women who are o bese are h igh - risk for developing ER breast cancer and would benefit from cancer prevention strategies. However, there are no pharmacological means to prevent for ER breast cancer. My previous studies identified FGF/FGFR1 signaling as a mechanistic target for the inhibition of ER breast epithelial cell transformation. Our previous data demonstrated that FGF2 in VAT can stimulate malignant transformation of breast epithelial cells by activating FGFR1. Furthermore , I revealed that fluvastatin , an HMGCR inhibitor, could prevent FGF2 - assciated transformation , identifying fluvastatin as a potential chemopreventive agent . HMGCR is the rate limiting enzyme in the mevalonate pathway and interestingly, literature has not d emonstrated a mechanistic connection between the mevalonate pathway and FGF2/FGFR1 signaling. Therefore , the objective of these studies is to investigate how an HMGCR reductase inhibitor mechanistically inhibit s FGF2/FGFR1 stimulated transformation. Herein , I demonstrated that factors from VAT upregulates enzymes in the mevalonate pathway , suggesting that VAT does affect the mevalonate pathway . However, m ore research is needed to fully understand the relationship between FGF2/FGFR1 signaling and the mevalon ate pathway. Introduction Women who are high risk for o besity - associated pre - menopausal ER breast cancer are in need chemopreventive strategies. Obesity is now a global pandemic and obesity rates are expected to rise. The prevalence of obesity in the Un ited States was 42.4% in 2018, a ~12% increase from 2000 13 . Epidemiologically, high BMI/increased abdominal obesity is strongly asso ciated with premenopausal triple negative breast cancer ( TNBC) 54,199,200 . TNBC is partly defined by absence of the estrogen receptor (ER) and current breast cancer chemoprevention compounds are only effective against breast ca ncers 64 expressing the ER. T amoxifen, a selective ER modulator (SERM), reduces the risk of ER+ breast cancer by almost 50%, 46 but does not reduce the risk ER breast cancer. Therefore, premenopausal obese women are a potential high - risk target population that would benefit from chemoprevention strategies. Elucidating the mechanisms of how obesity promotes ER breast cancer would reveal targets for chemopreventive compounds. Our published and supporting data reveal ed a key role for visceral adipose tissue (VAT) - derived fibroblast growth factor 2 (FGF2) via fibroblast growth factor receptor 1 (FGFR1) signaling in estrogen receptor (ER) negative (ER ) tumor initiation and progression. In addition, there is a high rate of FGFR1 amplifications 201 - 203 and sustained activation of the Ras/MAPK pathway in TNBC and basal - like breast cancer (as opposed to other subtypes) 204,205 . This suggests breast carcinomas and their normal epithelial precursors may use the Ras/MAPK pathway for survival. The Ras/MAPK pathway is a ctivated by receptor tyrosine kinase receptors like FGFR1. Therefore, FGFR1/Ras signaling could be a viable target for obesity - associated ER breast cancer prevention. However, d irectly targeting FGFR1/Ras signaling is difficult. While preclinical FGFR1 in hibitors (and one clinical FGFR inhibitor, Balversa) exist, they likely lack the high safety threshold chemoprevention requires 206,207 and oncogenic Ras is considered with FGFR1/Ras . While activating Ras mutations in breast cancer are rare, Ras expression and activity is often en hanced in breast cancer 208 . To discover potentially safe chemopreventive and therapeu tics for obesity - associated pre - menopausal ER breast cancer, I developed a novel phenotypic high throughput screen (HTS) targeting FGFR1/Ras - driven malignant transformation to identify compounds that inhibit FGF2 - stimulated 3D growth 209 . Using our HTS, fluvastatin emerged as the lead candidate, showing concentration - response efficacy in our soft agar assay (inhibiting and FGF2 - stimulated transformation in MCF - 10A cells) 209 . Chemoprevention strategies must effectively reduce cancer b urden and have minimal to no side effects and fluvastatin is a clinically approved drug that is relatively safe 210 . Fluvastatin inhibits 3 - hydroxy - 3 - methylglutaryl coenzyme A (HMG - CoA) reductase (HMGCR), the rate limiting enzyme in the mevalonat e pathway. Before the transformation HTS screen, there has not been a described relationship between FGFR1 signaling and the mevalonate pathway . 65 Our HTS was not the first to implicate HMGCR in malignant transformation and breast cancer. Clendening et. al. demonstrated that d eregulated HMGCR expression (increased HMGCR activity) in MCF - 10A (non - tumorigenic ER human breast epithelial) cells stimulated transformation as determined by the soft agar assay 211 . Furthermore, another stud y demons trated that a microRNA ( MiRNA ) which represses HMGCR and HMGCS1, decreased through an in vitro model of breast cancer progression, starting with normal - like MCF10A to preneoplastic MCF10.AT1, and invasive MCF10.Ca1d cells 30 . This means that HMGCR and HMGCS1 activity increased in t his in vitro model of breast cancer progression 30 . Epidemiology also corroborates mevalonate involvement in early transformation events 93% of ductal carcinomas in situ ( DCIS ) express HMGCR 212 . Furthermore, mevalonate pathway activity is linked to visceral obesity as v isceral obesity, and not subcutaneous obesity, is positively associated with elevated serum levels of squalene , which is cholesterol pr ecursor 213 . Additionally, t he mevalonate pathway synthesizes sterol isoprenoids (e.g., cholesterol, farnesyl pyrophosphate (FPP), geranylgeranyl pyrophosphate (GGPP) ) and non - sterol isoprenoids (e.g., dolichol, heme - A, isopentenyl tRNA and ub iquinone 90 ). Isoprenoids like FPP and GGPP are used to prenylate Rho and Ras family proteins which induces Rho/Ras plasma membrane localization enabling GTPase activation. Therefore, inhibition of HMGCR has the potential to alter posttranslational modifications of signaling molecules 90,91 . D ysregulation of HMGCR and the mevalonate pathway may make a causal contribution to obesity - associated ER cancer, suggesting statins may be us eful for prevention 211 . However, studies have n o t identified how mevalonate pathway dysregulation can promote transformation. U nraveling the connection between VAT - derived growth factor stimulation and malignant transformation will provide new targets for chemoprevention. Identif ying fluvastatin further informs on the mechanisms involved in ER breast epithelial cell transformation as it suggests a role for HMGCR d ys regulation and the mevalonate pathway . Therefore, the objective of these studies is to determine the role of the mev alonate pathway in VAT - stimulated ER breast carcinogenesis. We hypothesize that FGF2 and factors from VAT activate the mevalonate pathway through FGFR1 activity and subsequent Ras protein prenylation/activation drives ER transformation. Herein, we demons trate that VAT stimulation of MCF - 10A cells induced protein expression of mevalonate pathway enzymes HMG - CoA synthase 1 (HMGCS1), farnesyl transferase (FNTA), squalene 66 synthase (FDFT1) and Harvey rat sarcoma viral oncogene homolog (HRas). In addition, this protein upregulation was prevented by PD166866, a selective FGFR1 inhibitor. While HRas induction was measured by both western blot and immunofluorescence by VAT and FGF2, we failed to confirm its activation status. Results Statins have varying efficacy in preventing MFTF - stimulated transformation Previous data demonstrated that fluvastatin attenuated FGF2 - stimulated transformation of MCF - 10A cells as measured by colony formation in the soft agar assay 209 . Mouse fat tissue filtrate (MFTF), made from factors derived from VAT, was used in soft a gar with MCF - 10A cells to stimulate transformation. Fluvastatin also prevented MFTF - stimulated MCF - 10A cell transformation. To evaluate if other statins can prevent MFTF - stimulated transformation simvastatin, atorvastatin, pravastatin, and rosuvastatin were also tested for efficacy (Figure 4.1). Fluvastatin and simvastatin significantly attenuated MFTF - stimulated transformation. Pravastatin and rosuvastatin, hydrophilic statins , did not attenuate MFTF - stimulated transformation. Atorvastatin is the most lipophilic statin, followed by simvastatin, fluvastatin, pravastatin, and then rosuvastatin 214 . While fluvastatin and simvastatin are both lipophilic statins, atorvastatin is the most lipophilic statin, thus demonstrating lipophilicity does not determine efficacy, but efficacious statins have a tendency t o be lipophilic. 67 Figure 4.1. Lipophilic statins, fluvastatin and simvastatin concentration - dependently inhibit MFTF - stimulated transformation of MCF - 10A cells. Fluvastatin significantly attenuated MFTF (200 µg/mL) - stimulated transformation of MCF - 10A cells (1,000 cells/well) at 2.0 µM. Simvastatin significantly attenuated MFTF - stimulated transformation of MCF - 10A cells at 1.0 and 2.0 µM. Pravastatin, rosuvastatin , and atorvastatin did not attenuate MFTF - stimulated transformation of MCF - 10A cells. Rosuvastatin and atorvastatin enhanced MFTF - stimulated transformation. Colonies were counted via automated counting using the Cytation 3 imaging reader from Biotek using Gen5 3.04 software. Seven pictures were taken every 100 microns and superimposed together by the zprojection function. Treatments (3 biological replicates) were analyzed by one - way ANOVA with multiple comparisons (*p<0.05, **p<0.01, ***p<0.001 vs 0 µM). Th is experiment was repeated one time. 68 MFTF upregulates protein expression of enzymes in the mevalonate pathway To determine if the mevalonate pathway is upregulated by MFTF, MCF - 10A cells were treated with MFTF for 24 hours with and without fluvastatin o r PD166866, a selective FGFR1 inhibitor. Protein lysates were probed for proteins involved in the mevalonate pathway as indicated in Figure 4.2. At 24 hours, MFTF upregulated HMGCS1, FNTA, and HRas protein expression. PD166866 prevented this MFTF - induced p rotein upregulation compared to untreated control. Fluvastatin upregulated HMGCS1, which is in line with published literature. Inhibiting HMGCR induces a feedback loop that upregulates the HMGCR in an attempt to gain HMGCR activity 215,216 . Fluvastatin also prevented MFTF - upregulation of HRas and FNTA. Figure 4.2. MFTF upregulates protein expression of en zymes in the mevalonate pathway and this upregulation is prevented by PD166866, a selective FGFR1 inhibitor. A) Diagram of the mevalonate pathway First, two Acetyl - CoA molecules are condensed into acetoacetyl - CoA by acetoacetyl - CoA thiolase. From here HMG - CoA synthase 1 (HMGCS1) produces the metabolite HMG - CoA. Next, HMG - CoA reductase, the rate limiting enzyme, converts HMG - CoA to mevalonate. Through multiple steps, this pathway produces geranyl pyrophosphate (GPP) and then farnesyl pyrophosphate synthase ( FDPS) adds an isopentenyl - 5 - pyrophosphate (IPP) to create FPP. FPP can either be converted to GGPP by geranylgeranyl pyrophosphate synthase 1 (GGPS1) or FPP can be converted to squalene by squalene synthase (FDFT1) which ultimately leads to cholesterol bi osynthesis. B) MCF - 10A cells were treated with MFTF (200 µg/mL) and/or fluvastatin (1 µM), PD166866 (10 µM) for 24 hours (fluvastatin and PD166866 were pre - treated 30 minutes before MFTF). The same lysates were run twice at 22 µg of protein per sample. Mem branes were re - probed for proteins. The protein probes appear above their respective actin bands. This experiment was repeated once. 69 To determine if proteins upregulation is transient or sustained, MCF - 10A cells were treated with 200 µg/mL of MFTF over 8 , 24, 48, and 72 hours (Figure 4.3). The relative intensity for each protein was normalized to a ctin (Figure 4.3 B - F). HMGCS1 is upregulated at 24 hours and this protein upregulation is maintained for an additional 48 hours (Figure 4.3 B). FDFT1, squalene synthesis is upregulated at 24 hours, but returns to baseline levels after 72 hours (Figure 4.3 C). The relative protein intensity of FDPS (Figure 4.3 D) and FNTA (Figure 4.3 F) moderately increased (Figure 4.3 A). GGPS1 protein expression remained unchang ed (Figure 4.3 A, E). Figure 4.3. MFTF induces sustained upregulation of HMGCS1 protein and transient upregulation of FDFT1 protein at 24 hours. B) MCF - 10A cells were treated with MFTF (200 µg/mL) for 8, 24, 48 and 72 hours. The same lysates were run twice at 20 µg o f protein p er sample. Membranes were re - probed for proteins. The protein probes appear above their respective actin bands. This experiment was performed once. 70 FGF2/MFTF does not alter mRNA expression of mevalonate pathway related proteins RT - PCR was used to investigate if MFTF - stimulated protein expression was a product of MFTF - activated gene transcription (Figure 4.4). FGF2 after 1, 4, and 24 hours did not affect mRNA expression of HMGCR, FDPS, GGPS1, FNTA, HRas, and SREBF2. Fluvastatin up regulated mRNA of HMGCR, FDPS, and SREBF2 after 24 hours, which is also a result of the afore mentioned positive feedback loop in line with Figure 4.4. FGF2 and MFTF do not alter mRNA expression of enzymes in the mevalonate pathway in MCF - 10A cells. MCF - 10A cells were untreated (Unt.) or FGF2 (30 ng/mL) treated, fluvastatin (Fluv) (2.5 µM), PD166866 (PD) (10 µM), or MFTF (200 µg/mL) for 1, 4, or 24 hours. mRNA expression levels of HMGCR, FDPS, GGPS1, FNTA, HRas, and SREBF2 were normalized to GAPDH. Fluva statin upregulated HRas at 1 hour and HMGCR, FDPS, and SREBF2 at 24 hours. Fluvastatin downregulated GGPS at 4 hours. PD166866 downregulated FDPS and GGPS at 1 hour and 24 hours. PD166866 upregulated SREBF2 at hours. Each treatment group has 3 biological r eplicates, 2 technical replicates. Treatment groups with 2 biological replicates did not have enough RNA recovered during sample prep. Data is presented as mean ± SD (*p<0.05, **p<0.01, ***p<0.001 vs Unt. control) 71 published literature. Interestingly, Fluv astatin upregulated HRas mRNA at 1 hour. PD166866 downregulated FDPS and GGPS mRNA at 1 hour and 24 hours and upregulated SREBF2 mRNA at 4 hours. This upregulation indicates fluvastatin is inhibiting HMGCR, its known mechanism of action. Acute FGF2 or VAT stimulation failed to stimulate Ras activation in MCF - 10A cells The mevalonate pathway begins with acetyl coenzyme A ( Acetyl - CoA ) which and can produced products like cholesterol and isoprenoids . Isoprenoids like farnesyl pyrophosphate (FPP) and geranylgeranyl pyrophosphate (GGPP) are used for proteins that undergo prenylation, the attachment of a moiety onto proteins like Rho or Ras, stimulating membrane localization and in turn activation (Figure 4.2). To determ ine if isoprenoid production is critical in FGF2/FGFR1 - stimulated transformation, geranyl pyrophosphate (GPP) and mevalonate (MVL) were exogenously added to MCF - 10A cells in the soft agar assay. As a positive control, FGF2 was used to stimulate transformat ion . PD166866 attenuated FGF2 - stimulated transformation, as previously demonstrated 209 . GPP and MVL did not stimulat e transformation of MCF - 10A cells at 10 o r 20 ng/mL and these isoprenoids did not rescue the inhibitory effects of PD166866 (Figure 4. 5 ). Figure 4.5. Isoprenoids GPP and MVL failed to stimulate transformation of MCF - 10A cells and do not rescue the atten uating effects of PD166866. A) MCF - 10A cells were treated with FGF2 (30 ng/mL), PD166866 (10 µM) and either 10 or 20 ng/mL of GPP or MVL. GPP and MVL are constituted in methanol (MeOH) and is the vehicle for this experiment. Colonies were counted via autom ated counting using the Cytation 3 imaging reader from Biotek using Gen5 3.04 software. Seven pictures were taken every 100 microns and superimposed together by the zprojection function. Treatments (3 biological replicates) were analyzed by one - way ANOVA w ith multiple comparisons (*p<0.05). 72 Isoprenoid production leads to protein prenylation which localizes Ras to the membrane where its GTPase activity activates signaling cascades 88 . Therefore MCF - 10A cells were treated with FGF2 to determine if FGF2 induced Ras membrane localization and/or Ras activation. MCF - 10A cells were treated with FGF2 for 24 hours and Ras expression patterns were evaluated using immunofluorescence (Figure 4.6 A). FGF2 increased HRas protein expression but did not induce HRas localization. A Ras activation assay (Ras GTPase Chemi ELISA Kit) demonstrated that FGF2 (30 ng/mL) failed to activate Ras in MCF - 10A cells (Figure 4.6 B). Figure 4.6. FGF2 upregulates H R as protein expression but does not induce membrane localization or activation. A) MCF - 10A cells were treated with FGF2 for 24 hours and immunofluorescence was used to detect HRas protein. B) MCF - 10A cells were treated with FGF2 (30 ng/mL) for 2, 5, 10, and 20 mins and did not stimulate Ras activation as determined by luminescence, as shown as relative light units (RLU) . PC is a supplied positive control of HeLa cells treated with EGF. 73 Discussion Fluvastatin was identified as a chemopreventive compound in obesity associated ER breast epithelial cell transformation with our recent HTS. The transformation HTS models FGF2/FGFR1 stimulated transformation, a critical signaling pathway in VAT - stimulated ER breast epithelial cell transformation. Herein, fluv astatin was confirmed and validated as the lead chemopreventive hit. Before the transformation HTS screen, there has not been a described relationship between FGFR1 signaling and the mevalonate pathway, which begs the questions, how does an HMG - CoA reducta se inhibitor prevent FGF2/FGFR1 stimulated transformation? Therefore, the objective of these studies is to determine the role of the mevalonate pathway in VAT - stimulated ER breast carcinogenesis. First, I investigated whether e effects are achieved by other statin types. These data revealed that lipophilic statins, simvastatin and fluvastatin, concentration dependently prevent s MFTF - stimulated ER breast epithelial cell transformation whereas hydrophilic statins, pravastatin an d rosuvastatin did not. However, atorvastatin, the most lipophilic statin was not efficacious in preventing transformation unlike the other lipophilic statins . This suggests lipophilic statins are more likely to have cancer prevention efficacy, but lipophilicity alone does not guarantee efficacy. These data also do not suggest efficacy is specific to Type I or Type II statins. Statin type is determined by the molecular structure, and fluvastatin is a Type II statin an d simvastatin is a Type I statin 217 . All statins are HMGCR competitiv e inhibitors and bind to the active site. Structural differen ces between statins are thought to account for potency differences. Rosuvastatin is the most potent HMGCR inhibitor, followed by simvastatin, atorvastatin, and then fluvastatin (this study did no t evaluate pravastatin). All statins have an HMG - like moiety that binds to the active site of HMGCR. The methylethyl group and decalin ring structure on Type I I and Type I statins respectively, involve equivalent and numerous hydrogen binding interactions. The fluorophenyl groups of Type II statins exhibit additional binding interactions (Figure 4.7) 217 . Rosuvastatin and atorvastatin form additional hydrogen bonds. However, structural differences and their respective binding interactions do not predict or associate with chemopreventive efficacy as both statin types did and d id not prevent transformation. While these data do not indicate that chemopreventive efficacy is specific to statin lipophilicity or molecular structure, fluvastatin and simvastatin are potential are candidates for ER breast tumorigenesis prevention. 74 S inc e statins are relatively safe, they are promising candidate s for chemoprevention. Chemopreventive agents, like tamoxifen, can effectively reduce ER positive (ER+) breast cancer but not ER breast cancer. TNBC, defined as being ER , progesterone receptor (P R) negative (PR ), with no human epidermal growth factor receptor 2 (HER2) overexpression, constitutes 10% to 20% of all breast cancers. TNBC is a heterogenous, aggressive form of breast cancer with high tendency of metastasis, poor prognosis, and has no a pproved targeted therapy 48,49 , and no targeted chemopreventive strategies 50 . The ER breast cancer subtype would an ideal candidate for statin chemoprevention. Figure 4.7. The molecular structure of statins determines if they are Type I or Type II statins. Type I statins have an HMG - like moiety group, a butyryl group, and a decalin group. Simvastatin and pravastatin are Type I Statins. Type II statins have their HMG - like moiety, a fluorophenyl group, and a methylethyl group. Fluvas tatin, rosuvastatin, and atorvastatin are Type II statins. 75 S tatins have a challenging history when it comes to cancer risk because epidemiology does not collectively conclude that statins are chemopreventive 210 . Initial retrospective analyses agreed statins did not increase cancer risk, a crucial finding if statins were to be used universally as a cholesterol lowering agent, but these retrospective studies did not demonstrate a chemoprotective effect either 94 . However, these initial retrospective epidemiological studies were designed to evaluate cardiovascular outcomes rather than cancer prevention, cancer reoccurrence, and/or cancer mortality 94 . Once studies were designed to look at cancer outcomes and stratified by breast cancer subtype, statin lipophilicity, and menopausal status, they demonstrated lipophilic statins have a protective agai nst premenopausal ER breast cancer and not ER+ breast cancer 100,101,103 . Studies that do not stratify by lipophilicity, breast cancer subtype, or menopausal status give inconsistent results 107,108 . However, because of statin s safety thresholds and health benefit s, they should not be overlooked for their ability to reduce the cancer burden of aggressive ER breast cancers 101 . In vitro and in vivo studies positively support statins as chemopreventive agents. Using a N - methyl - N - nitrosourea (NMU) - induced mammary carcinogenesis in female Sprague Dawley rats, fluvastatin suppressed mammary tumor frequency by 63% and tumor incidence by 33% in comparison with the controls 218 . In vitro studies also reveal der egulated HMGCR activity promotes/and is associated with transformation. HMGCR is composed of two contiguous domains, the N - terminal is integrated into the endoplasmic reticulum (EndoR) membrane and the c - terminal projects into the cytosol which is responsi ble - terminal domain, HMGCR has stable catalytic activity 219 . MCF - 10A cells that express HMGCR in this deregulated state form colonies in soft agar, indicating that constitutive HMGCR activity induces MCF - 10A transformation 211 . Additionally, one study demonstrat ed that miR - 140 - 3p - 1, a microRNA that represses HMGCR and HMGCS1, decreased through an in vitro modeled of breast cancer progression, starting with normal - like MCF10A to preneoplastic MCF10.AT1, and invasive MCF10.Ca1d cells. As the cell lines became more cancer like, there is more HMGCR and HMGCS1 activity 30 . Therefore, these studies suggest HMGCR activity promotes MCF - 10A cell transformation and supports our data showing fluvastatin prevents MCF - 10A cell transformation. 76 Furthermore, our data suggests that the mevalonate pathway i s altered in obesity - associated ER breast epithelial cell transformation. Factors from VAT upregulated protein expression of mevalonate pathway - involved enzymes including HMGCS1, FDFT1, and FNTA and PD166866, an FGFR1 inhibitor prevented this upregulation (Figure 4.1). As MFTF/FGF2 did not regulate the gene expression of these enzymes as determined by RT - PCR, MFTF must regulate protein translation and/or post - translational modifications of these enzymes. This suggests that FGF2/FGFR1 signaling stimulated b y MFTF alters the mevalonate pathway and in turn cholesterol biosynthesis. Along with cholesterol biosynthesis, the mevalonate pathway also produces isoprenoids, like FPP and GGPP are used for protein prenylation. The Rho and Ras superfamilies are examples of prenylated proteins and mevalonate pathway dysregulation can promote oncogenesis through this process . Prenylation is necessary to maintain malignant activity of oncogenic Ras. All Ras isoforms (KRas, NRas, and HRas) are farnesylated, but NRas and HRas are then palmitoylated , the attachment of one and two fatty acids to NRas and HRas respectively. Farnesylated and palmitoylated Ras has a 100 - fold higher membrane affinity than Ras that is only farnesylated. Membrane localization places Ras in proximity w ith cell surface receptors and effector proteins which is required for Ras activity. As statins deplete isoprenoid production by reducing cellular mevalonate levels, this isoprenoid depletion could inhibit farnesylation preventing oncogenic Ras membrane lo calization and activation. The mevalonate pathway is the only metabolic pathway that produces FPP and GGPP. Ras is a known oncogene and previous literature demonstrated that constitutive HRas activation induces MCF - 10A cell transformation. Moon et al . tran sfected MCF - 10A cells with a mutant HRas (HRas MCF - 10A cells) or mutant NRas (NRas MCF - 10A cells) using retroviral vectors 220 , and both mutant cell lines produced colonies in soft agar. However, constitutively active HRas induced an invasive phenotype but constitutively active N - Ras did not (as determined by transwell invasion assay) 220 . Furthermore, simvastatin treatment markedly decreased prenylated HRas in the membrane and incr eased unprenylated HRas in the cytosol of HRas MCF - 10A cells 221 . Simvastatin was also able to concentration - dependently attenuate the invasiveness of HRas MCF - 10A cells and reintroducing exogenous FPP rescued the invasive phenotype of HRas MCF - 10A cells 221 . The ability of exogenous FPP to restore simvastatin inhibited - 10A cells. Therefore, 77 if FGF2/FGFR1 requires HRas farnesylation in MFTF - st imulated transformation, FGF2/FGFR1 activity would need HRas membrane localization and exogenous isoprenoids would rescue inhibitory effects of PD166866. Because HRas i s an oncogenic driver in MCF - 10A cells, FGFR1 signaling activates Ras , and HRas is a prenylated protein, we hypothesized that HRas protein prenylation is critical in FGF2/FGFR1 - stimulated ER breast epithelial cell transformation (Figure 4.8) . However, we did not see HRas activation or prenylation in MCF - 10A cells following FGF2 trea tment (Figure 4.5). In addition, exogenous isoprenoids did not rescue the effects of PD166866, a selective FGFR1 inhibitor. This suggests that fluvastatin inhibition of HRas prenylation is not how fluvastatin prevents FGF2/FGFR1 - stimulated ER breast epith elial cell transformation. Figure 4.8. The FGF2/FGFR1 signaling pathway and the mevalonate pathway. This figure is a combination of Figure 1.2 and Figure 1.4. Farnesylated HRas undergoes membrane localization and can activate FGFR1 downstream signaling pat hways. Previous literature has not described a relationship between FGF2/FGFR1 pathways. 78 There are potential limitations to our studies. While we did not see HRas activation/prenylation, these assays can be technically challenging. For exogenous isopreno ids to rescue a phenotype, isoprenoids must cross the plasma membrane. While previous literature has demonstrated exogenous isoprenoids like mevalonate and FPP cross the plasma membrane to rescue phenotypes 222 , this has not been demonstrated in soft agar. The Ras Activation assay is technic ally involved, and technical issues could have resulted in the negative data. Ultimately, it is still unclear how statins inhibit FGF2/FGFR1 stimulated ER breast epithelial cell transformation and more studies are needed to clarify the underlying molecula r mechanisms. In conclusion, our data demonstrated that factors from VAT upregulated the protein expression of HMGCS1, FDFT1, FNTA, and HRas through FGFR1 signaling. More studies are needed to investigate the role of the mevalonate pathway in ER breast ep ithelial cell transformation to determine if the induction of these enzymes are drivers of transformation or bystanders . In addition, studies to knock out HMGCS1 are needed to determine if this protein is necessary, sufficient, or dispensable for transform ation by VAT. Ultimately, FGF2/FGFR1 and the mevalonate pathway are potential pathways by which cells transform breast epithelial cells independent of ER signaling that can be targeted for prevention. Materials/Methods Cell culture MCF - 10A cells were cultured as previously described . Mouse fat tissue filtrate Mouse filtrate was prepared as previously described . Soft agar assay The soft agar assay was performed as previously described . Western blot MCF - 10A cells were plated in 6 - well plates and incubated at 37°C. After 24 hours and/or time required for cells to reach the appropriate density, cells were treated according to each experimental parameter. 79 Each well was washed with PBS and cells were tryp sinized with 0.05% Trypsin - EDTA. Trypsin was quenched with resuspension media and cells were centrifuged at 600 g for 5 mins. Supernatant was decanted and cells were resuspended in chilled PBS. Cells were centrifuged at 600 g for 5 mins. All the PBS was re moved, and cells were resuspended in lysis buffer (RIPA buffer and cocktail or protease and phosphatase inhibitors). Samples were frozen in the - 20°C and then thawed at 4°C. Samples were sonicated 3 times (10 seconds each, 30 second break) and then centrif uged at 12,000 rpm for 10 mins at Fisher, 23225) to determine protein content. Equal amounts of protein and volume were used, with 8X loading dye. Sample s were heated at 100°C for 5 mins. Samples were running at 60 V for ~20 mins and the 100 V for about 60 minutes. Protein from the gel was transferred to nitrocellulose membranes and washed with 1X tris - buffered saline, 0.1% Triton X (TBST). Membranes were blocked in 4% bovine serum albumin (BSA) in TBST or 5% non - fat dairy milk (NFDM) in TBST. Membranes were washed in 1X TBST (5 mins on rocker at room temperature) and incubated with primary membranes on the rocker either for 3 hours at room temperature or o vernight at 4°C. Next, membranes were washed and incubated with secondary antibody (IRDye® 800CW Donkey Anti - Rabbit IgG, Invitrogen, 926 - 32213) at 1:10000 in 1% BSA in TBST for 1 hour on the rocker. Membranes were washed and fluorescence was read on the Li - Cor Odyssey CLX. Table 4.1. Antibody parameters used for western blot protein detection ANTIBODY Company Catalog # Species DF Blocking buffer 1 ° Dilution Buffer Actin Sigma A5060 Rabbit 1/500 5% NFDM/TBST 4% BSA/TBST FDFT1 Proteintech 13128 - 1 - AP Rabbit 1/1000 5% NFDM/TBST 3% NFDM/TBST FDPS Proteintech 16129 - 1 - AP Rabbit 1/1000 5% NFDM/TBST 5% NFDM/TBST FNTA Proteintech 12274 - 1 - AP Rabbit 1/1000 5% NFDM/TBST 3% NFDM/TBST GGPS1 Proteintech 14944 - 1 - AP Rabbit 1/1000 5% NFDM/TBST 3% NFDM/TBST HMGCS1 Cell Signal 176431 - 1 - APWB Rabbit 1/1000 5% NFDM/TBST 3% NFDM/TBST HRas Proteintech 18295 - 1 - AP Rabbit 1/1000 5% NFDM/TBST 3% NFDM/TBST Abbreviations: FDFT1 - squalene synthase ; FDPS - farnesyl pyrophosphate ; FNTA - farnesyl transferase ; GGPS1 - geranylgeranyl pyrophosphate synthase 1 ; HMGCS1 - HMG - CoA synthase 1 ; HRas - Harvey rat sarcoma viral oncogene 80 RT - PCR Gene expression for the following genes were evaluated using RT - PCR: farnesyl transferase (FNTA), farnesyl pyrophosphate (FDPS), glyceraldehyde 3 - phosphate dehydrogenase (GAPDH), geranylgeranyl pyrophosphate 1 (GGPS1), HMGCR, Harvey rat sarcoma viral oncog ene (HRas), and sterol regulatory element binding transcription factor 2 (SREBF2). Table 4.1 lists the genes with their corresponding forward and reverse primer sequences. MCF - 10A cells were plated in 12 - well plates and incubated at 37°C. After 24 hours an d/or time required for cells to reach the appropriate density, cells were treated according to each experimental parameter. Each well was washed twice with PBS. 200 µL of lysis buffer from RNA ea sy Mini Kit (Qiagen, 74104). After 2 - 3 minutes, lysis buffer/c ell contents were transferred to a microcentrifuge tube and stored at - 20°C RNA was isolated according to the Qiagen Quick - start protocol, Part 1, RNeasy Mini Kit. Ribonucleic acid (RNA) was quantitated on a nanodrop. cDNA was synthesized according to Appl iedBiosystems Protocol for High Capacity cDNA Reverse Transcription Kid (4368814). cDNA was used during reverse transcription polymerase chain reaction (RT - PCR) in a SYBR® Green reaction. Primers were designed and ordered through Integrated developmental t echnologies (IDT) (Table 4.2). Each sample was run in a 384 - well format with 10 µL of sample and RT - PCR master mix. Target gene values were normalized to the housekeeper gene GAPDH and graphed as 2 - Target gene GAPDH Rep 1 , CT GAPDH Rep 2 GAPDH Rep 3 )]. Table 4.2. Primer sequences for RT - PCR Gene Forward Sequence Reverse Sequence FNTA CACACGAGAGTGGTCCTTCC GCACCCAAGGAGCATCAGTT FDPS TGTGACCGGCAAAATTGGC GCCCGTTGCAGACACTGAA GAPDH CTGGGCTACACTGAGCACC AAGTGGTCGTTGAGGGCAATG GGPS1 ACAGCATCTATGGAATCCCATCT CAAAAGCTGGCGGGTAAAAAG HMGCR TGATTGACCTTTCCAGAGCAAG CTAAAATTGCCATTCCACGAGC HRas GCTACGGCATCCCCTACATC TTGTGCTGCGTCAGGAGAG SREBF2 AACGGTCATTCACCCAGGTC GGCTGAAGAATAGGAGTTGCC 81 Ras Activation Assay MCF - 10A cells were plated in 6 - well plates and incubated at 37°C. After 24 hours and/or time required for cells to reach the appropriate density, cells were treated according to each experimental parameter. The protocol for Ras GTPase Chemi ELISA Kit (Acti ve Motif, 52097) was followed for Ras Extraction and the ELISA. Immunofluorescence MCF - 10A cells were plated in 24 well - pates on top of glass coverslip and treated accordingly to experimental parameters. Cells were fixed with formaldehyde at a 1:10 dilut ion. After 15 minutes, media/formaldehyde was aspirated and washed with PBS then incubated in blocking buffer (1% bovine serum albumin in PBST) for 1 hour. Wells were incubated with a 1:500 dilution of HRas (same antibody used for western blot) and incubat ed on a rocker overnight at 4°C. Wells were incubated in secondary antibody for 1 hour. Wells were washed with PBS and coverslips were placed on DAPI on glass slides and allowed to adhere. Images were acquired by a confocal microscope. 82 CHAPTER 5 General Conclusions 83 The data presented in this dissertation reveals that factors from VAT can stimulate ER mammary epithelial cell transformation. My s tudies suggested that FGF2 may be one of the most critical factors in VAT for inducing this response and FGF2 exerts its effects through FGFR1 . Therefore, these data addressed the first part of my objective which is to determine the mechanism of visceral obesity driven ER breast cancer and use this mechanism to identify chemopreventive compounds . To address the second part of the objective, I developed a HTS that utilized FGF2 - stimulated epithelial cell growth in ultra - low attachment conditions and used this tool to discover compounds that have the potential to inhibit VAT - s timulated transformati on. From this HTS , fluvastatin was identified as a potential chemopreventive agent. These data support the overall hypothesis that FGF2/FGFR1 is a critical mechanism in ER - independent VAT - associated transformation and is a potential target for chemoprevention strategies and these data revealed a potential role for the mevalonate pathway in transformation . To determine the mechanism of visceral obesity driven ER breast cancer , I identif ied factors in VAT that stimulated malignant transformation . First, I demonstrate d th at factors from VAT in HuFTF stimulates transformation of MCF - 10A cells, non - tumorigenic ER breast epithelial cells 160 . Previous data I generated also demonstrated HuFTF stimulates NMuMG cells, non - tumorigenic mouse mammary epithelial cells 79 . A s determined by ELISA, the FGF2 protein levels in HuFTF is positively associated with HuFTF - stimulated MCF - 10A and NMuMG cell colony formation . Furthermore, exogenous FGF2 concentration - dependently stimulated MCF - 10A transformation. To verify that the transformative effects are specific to FGF2 , the transformative ability of other FGF ligands were investigated . First, we compared the ef fects of FGF2, FGF1, FGF18, and FGF22 in soft agar and determined that FGF2 was the only ligand that concentration - dependently stimulated MCF - 10A colony formation. As FGF2 has the highest affinity for FGFR1, two inhibitors were used, an FGFR1 a ntibody and PD166866, a selective FGFR1 inhibitor, to determine if FGF2 stimulates transformation through FGFR1. The FGFR1 Ab attenuated Hu FTF - stimulated MCF - 10A colony formation in soft agar and PD166866 concentration - dependently attenuates FGF2 and MFTF - stimulated c olony formation in soft agar. PD166866 also attenuate s FGF2 - stimulated growth in ultra - low attachment conditions. These studies demonstrate that FGF2 may be the most important factor in VAT that activates FGFR1, the critical receptor in VAT - stimulated ER breast epithelial cell transformation. 84 The next section of this dissertation address es the second part of the objective which was to identify chemopreventive agents. There are currently no targeted treatments for ER breast cancer prevention or therapy. Th is is in large part due to lack of targets as ER breast cancer is defined by a lack of a receptor and the etiology of obesity - associated premenopausal ER is unknown. Moreover, there was not a high - throughput method that model ed the transformation process which means the methodology did not exist to screen for compounds for chemoprevention. Because the soft agar assay is not suitable for a high - throughput format, we used an alternative model of anchorage - independent growth, ultra - low attachment conditions. I developed and optimized the first high - throughput screen to model the process of transformation using growth in low attachment conditions in a 384 - well format. First, I demonstrated FGF2 concentration dependently stimulated growth of skin epidermal cell s in ultra - low attachment conditions over 96 hours. MCF - 10A cells were not suitable for this method and were used to validate hits in soft agar. Growth in low attachment conditions was quantified by CellTiter Glo, a luminescent signal proportional to the A TP content, a measure of proliferation. To ensure the HTS could identify chemopreventive agents, PD166866 was used to attenuate FGF2 - stimulate growth in low attachment conditions. This was further supported with I - BET - 726 (IBET), a bromodomain inhibitor th at we previously demonstrated prevent s HuFTF - stimulated transformation of NMuMG and MCF - 10A cells 170 . IBET attenuated FGF2 - stimulated growth in low - attachment conditions. Furthermore, FGF2 did not stimulate growth of skin epithelial cells in 2D culture conditions. After screening three li braries and confirmation and validation studies, fluvastatin emerged as the lead candidate for chemoprevention . F luvastatin is an ideal candidate for prevention because of its relatively known safety profile. Overall, i dentifying fluvastatin as the lead candidate for chemoprevention, fulfill ed the second part of the objective . The last section of this dissertation investigates the role of the mevalonate pathway in VAT - stimulated transformation. Fluvastatin inhibits HMGCR, the rate - limiting enzyme in the mevalonate pathway but there has not been a described relationship between FGFR1 signaling and the mevalonate pathway. First, I established fluvastatin is working by its known mechanism of action. Inhibiting HMGCR triggers feedback loops that upregulat es HMGCR protein expression and mRNA. After MCF - 10A were treated with fluvastatin for 24 hours, HMGCS1 protein was upregulated and HMGCR mRNA was significantly elevated. Furthermore, fluvastatin decreased the cholesterol content of MCF - 10A cells after 24 h ours (Appendix A). 85 Therefore, fluvastatin is inhibiting HMGCR in MCF - 10A cells. Next, it was determined that MFTF upregulated enzymes in the mevalonate pathway including HMGCS1, FNTA, FDFT1, and HRas and PD166866, a selective FGFR1 inhibitor prevented this upregulation. Fluvastatin treatment is known to up upregulate both HMGCS1 and HMGCR protein expression. HMGCS1 produces HMG - CoA, the substrate for HMGCR. Interestingly, an FGFR1 inhibitor was able to block VAT - stimulated upregulation of HMGCS1, which woul d reduce the amount of HMG - CoA produced. This s uggest s that HMGCS1, and in turn HMG - CoA production, may be part of the mechanism by which VAT stimulates ER mammary epithelial cell transformation. Furthermore, aberrations in FGFR1, FNTA, and FDFT1 have sig nificant co - occurrence in breast cancer patients , meaning these aberrations have functional synergy (Appendix B). These data suggest FGF2 and FGFR1 signaling influences the mevalonate pathway by upregulating protein expression of enzymes. Experiments to fu rther understand this mechanistic connection gave negative data. We did not see HRas activation or membrane localization and exogenous isoprenoids did not rescue the effects of PD166866 on FGF2 - stimulated MCF - 10A transformation in soft agar. These data sug gest mevalonate pathway HRas prenylation is not critical in VAT - associated ER breast epithelial cell transformation . Overall, it is not clear how FGFR1 - stimulated upregulation of HMGCS1 contributes to VAT - stimulated ER breast epithelial cell transformati on (Figure 5.1). Lastly, t h is dissertation has made significant contributions to the development of three - dimensional (3D) platforms by creating the first high - throughput assay to model the transformation of a non - tumorigenic cell. This is a breakthrough i n the technological development of cancer prevention HTSs. Th e s tudies described within this dissertation are important steps in understanding the obesity - cancer link as i n vitro experiments are fundamental in deciphering mechanisms behind cell behavior in vivo 169 . However, i n vitro methodology has predominantly consisted of two - dimensional (2D) cell culture. While 2D models have made significant contributions to understanding cell behavior, 2D systems deviate considerably from cell behavior in vivo . Physiologically, cells are influe nced by interactions with their microenvironment . Unlike 2D systems, 3D models more closely resemble the in vivo microenvironment by more effectively inducing in vivo - like cell fates 63 . The soft agar assay is well - accepted as a model for transformation that use s a 3D experimental setting. Suspending single cells in agarose only allows cells to grow in an anch orage dependent manner, thus distinguishing between transformed and non - transformed cells. Previous cancer 86 prevention HTSs use cancer cells which reveals a significant limitation with c ancer prevention models as cancer cells behave differently than normal cells. Overall, the transformation HTS further extends the range of processes 3D platforms can model. This has the potential to be adapted and modified to model transformation by carcinogens and other factors. Figure 5.1. Overall findings: FGFR1 and HMGCR inhibition prevents VAT - stimulated ER breast epithelial cell transformation and VAT upregulation of HMGCS1 may be part of this mechanism. I t is unclear exactly how FGFR1 signaling and mevalonate pathway interconnects to stimulate ER breast epit helial cell transformation. 87 Limitations The experimental settings used in this dissertation were primarily in vitro assays. Using MFTF, made from the VAT of mice or humans mimics the factors from VAT in the in vitro environment. The soft agar assay is t he gold standard assay of transformation as the ability for cells to form colonies in soft agar directly correlates to the cells ability to form tumors in vivo . Regardless, MFTF - treated cells in soft agar does not fully mimic tumorigenesis in vivo . The pre mise of using VAT is based on SAT not being as biologically active in promoting transformation. Subcutaneous mammary adipose tissue (MAT) is an important component of the mammary gland as complete ablation of MAT in transgenic mice prevents mammary gland d evelopment. Furthermore, MAT can augment breast cancer progression 156,157 . This bidirectional dialog between MAT and breast epithelium reveals another limitation of these studies as they do not model the role of MAT in ER breast epithelial transformation. Overall, the findings of these studies need to be validated in an in vivo model of FGF2/FGFR1 stimulated ER breast epithelial cells. The previous chapter investigating the role of the mevalonate pathway in FGF2/FGFR1 stimulated transformation only used one cell line, MCF - 10A cells. One - cell line studies can be problematic because the effects could be cell line specific and not representative of primary breast epithelial cells. In addition, the premise of this dissertation focuses on ER transformation based on the epidemiology data that premenopausal obesity increases the risk for ER breast cancer. Complementary studies using an ER+ breast epithelial cell would reveal if this mechanism is specific to ER transformation. For a more comprehensive view, other non - tumorigenic breast epithelial cells should be analyzed for MFTF - induced upr egulation of mevalonate pathway related enzymes. Negative data regarding HRas protein prenylation limits data interpretation and conclusions. Interpreting negative data can be difficult because this can be a result of technical issues instead of being tru ly negative data. First, the Ras activation assay is a technically challenging assay, and technical errors for the MCF - 10A cells when compared to protocol examples. Further troubleshooting should be explored to verify the Ras activation assay results. In addition, exogenous isoprenoids did not rescue the effects of PD166866 on FGF2 - stimulated MCF - 10A cell colony formation in soft agar. While previou s publications have shown that isoprenoids like mevalonate, FPP, GPP, and GGPP can cross the plasma membrane and 88 rescue stain - - 10As and the soft agar assay. The inability of MVL or GPP to rescue the eff the cell. Additionally, MVL or GPP at 20 µM could be too low of a concentration to rescue the phenotype. Higher isoprenoid concentrations would likely see a vehicle effect due to the low stock co ncentrations and they are constituted in methanol , risking a vehicle effect . Additional experiments are needed in order to validate that isoprenoids can pass through the plasma membrane at these concentrations and induce prenylation. Figure 5.2. 2D and 3D cell culture methods . 2D monolayer conditions have unlimited access to the nutrients, oxygen, signaling molecules, etc. Spheroid structures form cell layers that limits exposure to internal cells as the outer layer are highly exposed to the medium where as the core cells receive less growth factors, nutrients, and oxygen and are often quiescent or hypoxic. Image made in BioRender. 89 Investigating the mechanisms of VAT - stimulated transformation would preferably be performed in 3D culture conditions, however this proves to be challenging. As the transformation HTS was developed for the purpose of this dissertation, it is still a new technique and has no t been adapted to investigate the underlying mechanisms of transformation. The cy totoxicity studies, western blots, RT - PCR, immunofluorescence , and Ras activation assay were all performed in 2D culture conditions. This deviates from the 3D experimental setting of the soft agar assay and transformation HTS. 2D culture conditions were us ed because the 3D equivalents have not been developed. This proves technically challenging because the premise of transformation assays is that non - transformed cells do no proliferate. Therefore, control or untreated cells would not produce enough cells to quantitate for assays like western blots or RT - PCR. Immunofluorescence assays would need to be optimized for 3D spheroids. Spheroids have diffusional limits so staining would consist of the outer cell layers, a potential limitation in 3D immunofluorescenc e (Figure 5.2). Ideally, mechanistic studies should be performed in 3D platforms. Future studies should look at adapting and optimizing methodologies like western blots and RT - PCR with 3D formats. This would give more representative mechanistic data regard ing transformation. Furthermore, these 3D formats would enable more mechanistic studies comparing differences between 2D and 3D cell culture methods. Unanswered questions How does adipose tissue secrete FGF2? Current literature has not identified the mec hanism by which FGF2 is produced and secreted into the circulation. Studies have demonstrated that high BMI is associated with elevated circulating FGF2 levels, however , how FGF2 is secreted from adipose tissue , enters the circulation, and is transported to distance sites is a complex process 78 . FGF ligands are either paracrine, intracrine or endocrine signaling molecules 223 . Paracrine and e ndocrine FGFs have different FGFR binding mechanisms. Paracrine FGFs, like FGF2, interact with heparan - sulfate proteoglycans (HSPG) to stabilize the FGF/FGFR complex, whereas endocrine FGFs require klotho proteins and have very low affinity to HSPG 223 . Functionally, HSPGs regulate FGF availability, and their transport within tissues because HSPGs are covalently linked to cell surface proteins like FGFRs . Here, they modulate diffusion by sequestering FGFs that keeps them as local signals 224 . In contrast, endocrine FGFs require klotho proteins binding to stabilize the FGF/FGFR 90 complex and have low binding affinity to HSPG. This low affinity allows endocrine FGFs to enter the extracellular space and permeate freely through HSPGs and enter the bloo d. Because FGF2 is a paracrine growth factor, this goes against studies that attribute s adipose tissue as the source of circulating FGF2 levels because HSPGs would prevent FGF2 systemic diffusion. Therefore, how FGF2 is secreted in an endocrine manner rema ins unclear. Interestingly, FGF2 lacks a signal peptide and is secreted unconventionally. Typically, canonical protein transport to the extracellular space requires a signal - peptide that utilizes an endoplasmic reticulum/Golgi apparatus secretory mechanis m 225 . FGF2 does not have a signal peptide and must undergo an alternative secretion mechanism. Unconventional secretion of FGF2 begins with phosphatidylinositol 4,5 - bisphosp hate (PIP 2 ) - dependent FGF2 oligomerization that forms a pore in the plasma membrane. Here, FGF2 exits the cytoplasm through this pore and once in the extracellular matrix, FGF2 is trapped and firmly bound by HSPG, keeping FGF2 proximal 225 . The intricacies of this unconventional secretory mechanis m have yet to be fully understood. Regardless, this does not explain endocrine FGF2 transport. Another potential protein transport mechanism utilizes exosomes, vesicles that entrap intracellular molecules and travel to distant sites 226 . Exosomes are involved in cell - to - cell communicati on by transferring proteins, mRNAs, and microRNAs 225 . In addition, adipocyte secreted exosomes are altered in obesity 226 . Specifically, adipose tissue in obese mice release more than twice the amount of exosomes 227 . Another stu dy demonstrated adipose - derived stem cells (ASCs) produced exosomes that contained FGF2 228 . Studies have not investigated circulating exosome FGF2 co ntent in obesity compared to normal weight. FGF2 - containing exosomes could be a mechanism of FGF2 secretion into the circulation, enabling FGF2 to act in an endocrine manner despite being a paracrine growth factor. Further studies are needed to understand how VAT produced FGF2 is released into the circulation. Which FGF2 isoforms are produced and secreted by adipose tissue? Human FGF2 can undergo alternative splicing that produces five FGF2 isoforms and our data does n o t inform as to which FGF2 isoforms ar e produced and secreted by VAT. FGF2 isoforms are referred to as either high molecular weight (HMW) (34 kDa) or low molecular weight (LMW) (24, 22.5, 22, and 18 kDa). Studies have described HMW and LMW FGF2 isoforms as having contrasting potencies and/or f unctions 91 on physiological processes. For example, HMW FGF2 overexpression lowered bone mineral density (BMD) while LMW FGF2 overexpression enhanced bone mineral density in mice 148 . Another mouse model demonstrated that HMW FGF2s enhanced hepat ic fibrosis whereas LMW FGF2 suppressed hepatic fibrosis 149 . In addition, LMW isoforms protected mice from myocardial dysfunction following ischemia - reperfusions, whereas HMW isoforms had a detrimental effect 229 . This indicates that HMW and LMW FGF2 isoforms can have opposing functions with LMW isoforms promoting favorable health outcomes compared to HMW FGF2. In contrast, FGF2 isoforms can also demonstrate a range of potencies. LMW isoforms were more mitogenic th an HMW isoforms in dermal fibroblasts 150 . Interestingly, FGF2 derived from omental VAT was more mitogenic than FGF2 from benign or ca ncerous renal tissue in endothelial cells 151 . In the soft agar assays, MCF - 10A cells were treated with recombinant FGF2 that is exclusively 18 kDa, an LMW isoform and we did not investigate the type of FGF2 isoform in the FTF. Further investigations need to determine the efficacies/potencies of FGF2 isoforms in MCF - 10A transformation and which FGF2 isoforms are in the circu lation/produced in the VAT of normal weight and obese individuals. In tandem, studies have not looked at FGF2 isoforms in the breast cancer, so there could be a specific FGF2 isoform expressed in ER breast cancer that promotes progression. Identifying spe cific FGF2 isoforms would reveal more specific targeting in ER breast cancer progression and treatment. Potential contraindications targeting FGFR1 Ablating FGFR1 in a healthy population could have significant unwanted side effects because FGF signaling controls a variety of processes in embryonic development and in adult organisms. FGFR1 mutations cause genetic disorders that effect bone development and the hypothalamic - pituitary - gonadal o n i c dysplasia, and - associated genetic disorders reveal FGFR1 is crucial in physiological processes . In addition, FGFR1 is essential for mammary gland development. The developing mammary bud expresses FGFR1 and FGFR1 is the primary FGF R during duct al morphogenesis in mammary epithelium. Prenatal FGFR1 deletion delays mammary gland development. Inducible FGFR1 activation increases lateral budding in mammary epithelium, induces alveolar hyperplasia, and invasive lesions 131,136 . While the target population are adults and are no longer undergoing mammary gland development, FGFR1 is crucial for mammary gland 92 homeostasis which is demonstrated by the structural changes generated by inducible FGFR1 (iFGFR1) activation. iFGFR1 activation induces proliferation, disrupts cell polarity, and promotes cell survival 136 . Furthermore, FGFR1 plays a role in osteogenesis and inhibiting FGFR1 could negatively affect this process . For example, MiR - 214, a microRNA, inhibits FGFR1 signaling and attenuates osteogenesis 230 . Therefore, targeting FGF2/FGFR1 signaling could have negative effects on osteogenesis in a healthy population. Overall, attenuating FGFR1 activity in a healthy population should be approached with caution to ensure critical physiological functions like osteogenesis and mammary gland homeostasis are not impeded and do not cause unwanted and/or ad verse side effects. Currently, there are no clinically approved compounds that selectively inhibit FGFR1 for chemoprevention. E r dafitinib, a pan - FGFR1 inhibitor, is the only approved FGFR inhibitor. E r dafitinib is clinically approved for advanced bladder c ancer with FGFR2 or FGFR3 genetic alterations. For some patients, Erdafitinib use caused ocular toxicity, an adverse side effect common to MAPK pathway inhibitors 231 . Unfortunately, this means erdafitinib lacks the high safety threshold required for chemoprevention 206,20 7 . Theoretically, selectively targeting FGFR1 would lower potential side effects because processes regulated by FGFR2 - 4 would not be affected. However, there are currently no clinically approved selective FGFR1 inhibitors. Regardless of a lack of FGFR1 inhibitors, targeting the transformation process itself instead of FGF2/FGFR1 directly is a more robust way to identify efficacious chemopreventive comp ounds. HTSs have two different types of approaches: target - based and phenotypic. Target - based approaches identifying compounds that directly inhibit FGF2 or FGFR1. However, this does not guarantee they will be FGFR1 selective nor will this guarantee chemop reventive efficacy. Moreover, RTKs have high redundancy and compensatory crosstalk, and compounds identified using a target - based approaches have a high attrition rate due to lack of efficacy when moved to in vitro and/or in vivo models due to redundancy a nd compensatory crosstalk. Therefore, we used a target - based, phenotypic assay to identify chemopreventive agents that inhibit the transformation process. This uses FGF2/FGFR1 as the informed mechanism but is a phenotypic assay to increase hit efficacy. Th is allows hits to target downstream signaling pathways involved in transformation and may not have the same contradictions that a direct FGFR1 inhibitor might 93 have. Regardless, toxicity studies would have to determine any potential side effects and/or adve rse outcomes. Future Directions Further studies are needed to examine the chemopreventive efficacy of fluvastatin in vivo against FGFR1 - promoted tumorigenesis. This would be effectively demonstrated using HC - 11 mouse mammary epithelial cells that have a transgenic iFGFR1 that is induced by a synthetic molecule in a Balb/c xenograft model 232 . This model has a slow tumor onset, making it well suited f or prevention studies. It would be interesting to determine whether fluvastatin decreases the percentage of mice that develop adenocarcinomas and delay tumor onset. Between tumors that arose from control/vehicle or fluvastatin treatment, the expression of Ras isoforms could be compared to see if FGFR - stimulated mammary tumorigenesis increased Ras protein expression/activity compared to benign mammary tis sue. While this in vivo experimental system models FGFR1 - stimulated mammary tumorigenesis, there are drawbacks to this model as HC - 11 cells are ER+ 233 . The ideal in vivo model would be ER , revealing a limitation of the HC - 11 iFGFR1 model. An alternative model would use MCF - 10A cells in a xenograft model in immunocompromised mice with exogen ous FGF2 injections. This would be modeling FGF2 - stimulated ER tumorigenesis. This in vivo model is immune deficient mice and precludes any role the immune system might have. However, this MCF - 10A xenograft model needs to be developed. While there are lim itations to both models, they would both function to further this research with in vivo data. More experiments are needed to assess mevalonate pathway activity. Protein upregulation does not guarantee pathway activation. Therefore, quantification of mevalo nate pathway endpoints is needed to determine in FGF2/MFTF activated the mevalonate pathway. Quantifying cholesterol content in MCF - 10A cells treated with FGF2, MFTF, fluvastatin, and/or PD166866 would determine the role of FGF2/MFTF on cholesterol biosynt hesis through the mevalonate pathway. Additionally, metabolomics analyses of the aforementioned treatments would also inform on mevalonate pathway activation. Our data demonstrated both fluvastatin and simvastatin prevented ER breast epithelial cell and f uture studies should follow up with both fluvastatin and simvastatin to determine if they have equal 94 chemopreventive efficacies and/or if one statin is more favorable over the other. Thus, the role of simvastatin should be investigated with in vitro and in vivo with fluvastatin. Alternative hypotheses While the data did not show HRas activation, further studies should investigate other prenylated proteins for a potential role in ER breast tumorigenesis. There are two types of prenylation: farnesylation an d geranylgeranylation. Farnesylated proteins include the Ras superfamily (K/N/HRas) and geranygeranylated proteins include RhoA/B/C, Rap1A, Rac - 1, and Cdc42. The data described focused on HRas however, the NRas and KRas isoforms could be responsible. If th ere was NRas or KRas activity, the Ras activation assay detects all Ras isoforms so this assay would have shown activation if FGF2 was activatin g NRas and/or KRas. In addition, other prenylated GTPases are deregulated in cancer and/or promote tumorigenesis . For example, Cdc42 activation induce transformation. While no activating Cdc42 mutations have been detected in human cancer, C dc42 is reportedly overexpressed in breast cancer 234 . Intriguingly, one study implicated Cdc42 activity in prom oting NIH 3T3 fibroblast transformation 234 . Furthermore, the oncoprotein Dbl is highly transforming and it activates Rho GTPases like Cdc42 234 . Additionally, transformation assays demonstrated RhoA NIH 3T3 mutants did not have the same robust transformation seen with Ras mutants 235 . Ultimately, there are many GTPase proteins that could be further investigated for a ro le in transformation. IGF/IGF1R signaling Mechanistically, there are several obesity - related alterations in the insulin and the insulin - like growth factor (IGF) system that may promote tumorigenesis. Along with fluvastatin, insulin - like growth factor receptor 1 (IGF1R) inhibitor picropodophyllin (PPP) was identified as a chemopreventive compound in the transformation HTS. Insulin resistance is one of the hallmarks of obesity which is often followed by hyperinsulinemia, a condition that may contr ibute to breast cancer development 236 . First, a hyper insulinemic state induces changes in circulating ligands of the IGF family. The IGF family comprises two ligands, IGF1 and IGF2, and three receptors (IGF1R, IGR2R, and IGF3R) and six IGF - binding proteins ( IGFBP1 - 6) 236,237 . Insulin increases hepatic growt h hormone (GH) receptor expression, leading to GH 95 mediated increases in IGF1 production as well as repression of IGFPB1,2 31,236 . IGFPBs normally binds to IGF1, inhibiting IGF1, and free IGF1 is the bioactive form. Obese individuals consistently have elevated free circulating IGF1 levels 236 . Additionally, the insulin receptor (IR) and IGF1R. have high homology meaning both insulin and IGF1 can interact with either IR or IGF1R 31,237 . Furthermore, epidemiology associated receptor expression with TNBC. One study demonstrated the IR was highly expressed in TNBC cases 238 . African American women, have higher IGF1R expression in normal breast tissue, are three times more li kely to develop TNBC then non - African American women 237,239 - 243 . High levels of IGF1 has been associated with increased risk of premenopausal breast cancer in women. Overall, epidemiological studies support a rol e for insulin and/or IGF1/IGF1R in obesity associated TNBC. As obesity has a complex multifactorial pathophysiology it is unlikely that obesity - promotes tumorigenesis through one pathway. More studies should focus on multi - factors stimulating transformatio n and pathway synergy in tumorigenesis. Estrogen metabolism Estrogens are known to influence isoprenoid metabolism through the mevalonate pathway, but little is known about the molecular mechanisms. Estrogen should not be overlooked when investigating t he mechanisms underlying pre - menopausal obesity and ER breast cancer. First, BMI is positively associated with increased aromatase expression and estrogen levels 29,30 . In normal weight preme nopausal women, ovarian granulosa cells are the main source of estrogens while a small fraction is produced by other tissues like adipose. However, in obese premenopausal women, estrogens are predominantly produce by adipose tissue as adipocytes express ar omatase, the rate limiting enzyme in estrogen biosynthesis 31 . Previously, it was thought estrogen was only involved in the development and progression of ER+ breast cancer, however estrogen can play a role in ER breast cancer as well 46 . Estrogen can act as a carcinogen through ER binding or estrogen metab olism . 46 First, estrogen can bind to ER and activate multiple pathways including the MAPK and PI3K/A kt pathways . E strogen can also bind to nuclear ER, dimerize with a complementary estorgen/ER, and act as a transcription factor that alter s gene expression . T he second pathway is through estrogen metabolism by the catechol pathway. In short, e strogen is metabolized (Phase I metabolism) by P450 enzymes to catechol estrogens, metabolites that forms unstable depurinating DNA adducts, and may generate cancer initia ting mutations and potentially lead to 96 transformation 46,244 assume estrogen does not influence TNBC . Furthermo re, epidemiological studies suggest estrogen metabolism is involved in cancer initiation 244 . Key et al . demonstrated high circulating levels of estradiol had a strong association with pre - menopausal breast cancer 245 . Interestingly, breast cancer 1 (BRCA1) tumors are predominantly ER and oophorectomy, which reduces estrogen levels, decreased the risk of BRCA1 breast cancer 246 . In tandem, ovariectomy inhibited the development of ER breast cancer 247 and exog e nous steroid hormones contributes to ER br east cancer progression 248 . Additionally, Gupta et al . demonstra ted circulating estrogens are required for the formation of ER tumors using a breast cancer xenograft model 249 . Lastly, a romatase inhibitors resulted in lower ER cancer incidence in the contralateral breast compared to women who received tamoxifen 46 . Collectively, these studies indicate estrogen promotes ER - independent tumorigenesis. ER - independen t e strogen carcinogenicity is also demonstrated in vitro and in vivo . For example, ER MCF - 10A cells are transformed by estrogen as measured by growth in soft agar, even in the prescence of tamoxifen 244 . In addition, estrogen - transformed MCF - 10A cells produced tumors when injected into immuno - deficient mice, indicating estrogen promotes breast tumorigenesis independent of ER 244 . tumorigenesis. High circulating estrogen levels in obese individuals cou ld be a mechanism of obesity - associated tumorigenesis. However, mechanistic studies investigating estrogen and cholesterol metabolism, use ER - expressing cell lines/models. For example, one study analyzed the promoter region of SREBF2 for estrogen response elements (ERE), a sequence of DNA that binds to estrogen/ER and regulates gene transcription. They found an ERE - like sequence in the SREBF2 promoter region and demonstrated estradial promoted SREBF2 expression in HepG2 cells (human heptatoblastoma cells), and an ER antagonist inhibited this 250 . While this study demonstrated that estrogen can regulate transcription of SREBF2, this is in an ER - dependent manner. Overal l, more studies are needed to further understand how estrogen regulates the mevalonate pathway in an ER - independent manner. 97 Conclusions Obesity is a complex disease and understanding the molecular mechanisms underlying how obesity - promotes pre - menopausal ER breast cancer will reveal targets for prevention and potential biomarkers for at - risk individuals. There is an immediate need to e lucidat e these molecular mechanisms as there are no current effective chemopreventive strategies for ER breast cancer. Id entifying FGF2 in VAT and the role of FGFR1 in transformation suggests the FGF2/FGFR1 may be important in the etiology of obesity associated ER breast cancer. In tandem, identifying fluvastatin as a chemopreventive agent suggests HMGCR activity may also b e important. Together, these data suggest this FGFR1/HMGCR axis may be potential target for prevention and FGF2 could function as a biomarker for at risk individuals. Furthermore, developing the first HTS to model transformation marks a new advancement for 3D cell culture models. The transformation HTS can evaluate the ability of compounds to inhibit transformation , making this is a breakthrough in the technological development of cancer prevention HTSs. 98 APPENDICES 99 Appendix A : Fluvastatin decreases cholesterol in MCF - 10A cells Statins are clinically used as cholesterol lowering agents. To determine if fluvastatin was inhibiting HMGCR in MCF - 10A cells, lipids were isolated from MCF - 10A cells treated with fluvastatin (1 µM) for 24 hours . Fluvastatin - treated MCF - 10A cells had significantly reduced cholesterol content then vehicle controls (Figure A.1 ) Figure A.1. Fluvastatin treatment for 24 hours significantly decreases the cholesterol content in MCF - 10A cells. MCF - 10A cells wer e treated with DMSO or fluvastatin (1 µM) for 24 hours. Cholesterol was quantitated as described in Methods. Significance was determined by - test (*p < 0.05). 100 Materials/Methods Amplex Red cholesterol quantitation assay: MCF - 10A cells were plated in 6 - well plates and incubated at 37°C and after 24 hours/or time required for cells to reach the appropriate density, cells were treated according to each experimental parameter. Cells were wa shed 3X with PBS. Cells were scraped in 200 µL of chilled PBS and transferred to sterile glass vials and centrifuged at 3,000 x g. PBS was decanted and 370 µL of chilled ( - 20°C) methanol was added to each vial. Vials were sonicated for 10 second (twice) an d then 270 µL of chloroform. Vials were vortexed for 1 - 2 minutes and then centrifuged for 5 minutes at 3,000 x g. The supernatant was transferred to new glass vial. For the glass vials with the pellet, 370 µL of chilled ( - 20°C) methanol was added to each v ial. Vials were sonicated for 10 seconds (twice) and then 270 µL of chloroform. Vials were vortexed for 1 - 2 minutes and then centrifuged for 5 minutes at 3,000 x g. The supernatant was transferred the same vial as previous supernatant. Samples were run on a speedvac to evaporate the methanol:chloroform mixture. Lipids were resuspended in 200 µL of 1X reaction buffer from the Cholesterol Quantitation Kit (Invitrogen, A12216). The cholesterol quantitation assay was performed according to the protocol in the C holesterol Quantitation Kit. 101 Appendix B : FNTA and FDFT1 have co - occurring mRNA alterations with FGFR1 in breast cancer patients The mRNA of FGFR1 and mevalonate - related proteins was analyzed in breast cancer patients using the public cBioPortal for Can cer Genomics of four databases: TCGA, Firehose Legacy; TCGA, Nature 2012; TCGA, PanCancer Atlas; Meta B ric, Nature 2012 & Nat Commun 2016 (Figure B .1). Figure A.1 is collectively summarized in Table B .1 listing the p values and how many data sets out of fou r reached significance . The mRNA of squalene synthase (FDFT1) was significantly reduced in FGFR1 amplified samples compared to FGFR1 diploid patients in three out of four data sets. FNTA mRNA was significantly elevated in FGFR1 amplified vs FGFR1 diploid p atients in three datasets. Additionally, FDPS and GGPS1 mRNA was significantly reduced in FGFR1 amplified compared to FGFR1 diploid patients in three datasets. 102 Figure B.1. Using four TCGA breast cancer databases, mRNA aberrations in FDPS, FNTA, and FDFT1 were associated with FGFR1 mRNA amplification in breast cancer patients. Breast cancer patients listed in the A) TCGA, firehose database, B) TCGA, METABRIC, Nature 2012 & Nat Commun 103 . 2016, C) TCGA, Nature 2012, and D) TCGA PanCancer Atlas were analyzed for mRNA gene expression in amplified and diploid FGFR - test (*p<0.05, **p<0.01, ***p<0.001). 104 Table B - test comparing the mRNA of proteins in the Mevalonate pathway in FGFR1 Amplified vs Diploid Breast Cancer Samples Gene Firehouse p value MetaBric p value Nature p value PanCancer p value Significance GGPS1 0.0532 <0.0001 0.0062 0.1120 2/4 FDPS 0.7077 0.0435 0.0269 0.0216 0/4 FNTA <0.0001 0.8883 0.0045 <0.0001 3/4 HMGCR 0.0915 - 0.1490 0.6676 0/3 HMGCS1 0.1417 <0.0001 0.5860 0.3595 1 /4 FDFT1 <0.0001 <0.0001 0.0009 <0.0001 4/4 HRas 0.6484 <0.0001 0.0003 0.5810 2/4 SREBF2 0.2301 <0.0001 0.2679 0.0911 1/4 Abbreviations: Firehose - TCGA firehose database; Metabric - TCGA METABRIC Nature 2012 & Nat Commun 2016; Nature - TCGA Nature 2012; PanCancer - TCGA PanCancer Atlas. 105 Furthermore, three of these datasets were used to analyze for patterns of co - occurrence or mutual exclusivity regarding the mRNA in breast cancer patients (the MetaBric dataset did not have co - occurrence/mutual exclusivity data available). Patterns of co - o ccurrence and mutual exclusivity can inform on functionally relevant mechanisms altered for oncogenesis. 251 Mutually exclusive alterations tend to suggest functional redundancy, meaning both alterations are unnecessary as the second will not be advantageous or could mean that both alterations indu ce synthetic lethality. 252 On the other hand, co - occurrent alterations suggest functional synergy. As calculated by cBioPortal, there is sig nificant co - occurrence with alterations between FGFR1 / FDFT1, FGFR1 / FNTA, and between FDFT1 / FNTA (Table B . 2 ). Interestingly, FDPS and GGPS1 tended to have co - occurring alterations. This suggest that FGFR1, FDFT1 and FNTA have functional synergy in bre ast cancer. FDPS and GGPS1 also had a tendency for co - occurrence, suggesting functional synergy Abbreviations: Firehose - TCGA firehose database ; Metabric - TCGA METABRIC Nature 2012 & Nat Commun 2016 ; Nature - TCGA Nature 201 ; PanCancer - TCGA PanCancer Atlas. 106 REFERENCES 107 REFERENCES 1 Defining Adult Overweight & Obesity , < https://www.cdc.gov/obesity/adult/defining.html > (2017). 2 Lauby - Secretan, B. et al. Body Fatness and Cancer Viewpoint of the IARC Working Group. New England Journal of Medicine 375 , 794 - 798, doi:10.1056/NEJMsr1606602 (2016). 3 Gardner, G. & Halweil, B. Underfed and Overfed: The Global Epidemic of Malnutrition. Worldwatch Paper , X - 68 (2000). 4 Singh, G. M. et al. The age - specific quantitative effects of metabolic risk factors on cardiovascular diseases and diabetes: a pooled analysis. PLoS One 8 , e65174, doi:10.1371/journal.pone.0065174 (2013). 5 Ni Mhurchu, C. et al. Body mass index and cardiovascular disease in the Asia - Pacific Region: an overview of 33 cohorts involving 310 000 participants. Int J Epidemiol 33 , 751 - 758, doi:10.1093/ije/dyh163 (2004). 6 Goossens, G. H. The Metabolic Phenotype in Obesity: Fat Mass, Body Fat Distribution, and Adipose Tissue Function. Obes Facts 10 , 207 - 215, doi:10.1159/000471488 (2017). 7 Berrington de Gonzalez, A. et al. Body - Mass Index and Mortality among 1.46 Million Whi te Adults. New England Journal of Medicine 363 , 2211 - 2219, doi:10.1056/NEJMoa1000367 (2010). 8 Zheng, W. et al. Association between Body - Mass Index and Risk of Death in More Than 1 Million Asians. New England Journal of Medicine 364 , 719 - 729, doi:10.1056/N EJMoa1010679 (2011). 9 The Emerging Risk Factors, C. Separate and combined associations of body - mass index and abdominal adiposity with cardiovascular disease: collaborative analysis of 58 prospective studies. The Lancet 377 , 1085 - 1095, doi:10.1016/S0140 - 6 736(11)60105 - 0 (2011). 10 Metabolic mediators of the effects of body - mass index, overweight, and obesity on coronary heart disease and stroke: a pooled analysis of 97 prospective cohorts with 1·8 million participants. The Lancet 383 , 970 - 983, doi:10.1 016/S0140 - 6736(13)61836 - X (2014). 11 Prospective Studies, C. Body - mass index and cause - specific mortality in 900 000 adults: collaborative analyses of 57 prospective studies. The Lancet 373 , 1083 - 1096, doi:10.1016/S0140 - 6736(09)60318 - 4 (2009). 12 Bi ener, A., Cawley, J. & Meyerhoefer, C. The High and Rising Costs of Obesity to the US Health Care System. Journal of General Internal Medicine 32 , 6 - 8, doi:10.1007/s11606 - 016 - 3968 - 8 (2017). 13 Hales, C. M., Carroll, M. D., Fryar, C. D. & Ogden, C. L. Preva lence of obesity and severe obesity among adults: United States, 2017 - 2018., (U.S Deparrment of Health and Human Services, Centers for Disease Control and Prevention, 2020). 14 Organization, W. H. Global action plan for the prevention and control of NCD 20 13 - 2020 , 2013). 15 Kontis, V. et al. Contribution of six risk factors to achieving the 25×25 non - communicable disease mortality reduction target: a modelling study. The Lancet 384 , 427 - 437, doi:10.1016/s0140 - 6736(14)60616 - 4 (2014). 108 16 Collaboration, N. R. F. Trends in adult body - mass index in 200 countries from 1975 to 2014: a pooled analysis of 1698 population - based measurement studies with 19·2 million participants. The Lancet 387 , 1377 - 1396, doi:10.1016/s0140 - 6736(16)30054 - x (2016). 17 Gregg, E. W. & Sha w, J. E. Global Health Effects of Overweight and Obesity. N Engl J Med 377 , 80 - 81, doi:10.1056/NEJMe1706095 (2017). 18 Kershaw, E. E. & Flier, J. S. Adipose tissue as an endocrine organ. J Clin Endocrinol Metab 89 , 2548 - 2556, doi:10.1210/jc.2004 - 0395 (2004 ). 19 Zhang, Y. et al. Positional cloning of the mouse obese gene and its human homologue. Nature 372 , 425 - 432, doi:10.1038/372425a0 (1994). 20 Frayn, K. N., Karpe, F., Fielding, B. A., Macdonald, I. A. & Coppack, S. W. Integrative physiology of human adip ose tissue. International Journal of Obesity 27 , 875 - 888, doi:10.1038/sj.ijo.0802326 (2003). 21 Wajchenberg, B. L., Giannella - Neto, D., da Silva, M. E. R. & Santos, R. F. Depot - Specific Hormonal Characteristics of Subcutaneous and Visceral Adipose Tissue a nd their Relation to the Metabolic Syndrome. Horm Metab Res 34 , 616 - 621, doi:10.1055/s - 2002 - 38256 (2002). 22 Ibrahim, M. M. Subcutaneous and visceral adipose tissue: structural and functional differences. Obesity Reviews 11 , 11 - 18, doi:10.1111/j.1467 - 789X. 2009.00623.x (2010). 23 Ibrahim, M. M. Subcutaneous and visceral adipose tissue: structural and functional differences. Obes Rev 11 , 11 - 18, doi:10.1111/j.1467 - 789X.2009.00623.x (2010). 24 Hellmér, J., Marcus, C., Sonnenfeld, T. & Arner, P. Mechanisms for d ifferences in lipolysis between human subcutaneous and omental fat cells. The Journal of Clinical Endocrinology & Metabolism 75 , 15 - 20, doi:10.1210/jcem.75.1.1320047 (1992). 25 Acta Medica Scandinavica 187 , 477 - 483, doi:10.1111/j.0954 - 6820.1970.tb02973.x (1970). 26 Motoshima, H. et al. Differential Regulation of Adiponectin Secretion from Cultured Human Omental and Subcutaneous Adipocytes: Effects of Insulin and Rosiglitazone. The Journal of Clinical Endocrinology & Metabolism 87 , 5662 - 5667, doi:10.1210/jc.2002 - 020635 (2002). 27 Ferrann ini, E., Sironi, A. M., Iozzo, P. & Gastaldelli, A. Intra - abdominal adiposity, abdominal obesity, and cardiometabolic risk. European Heart Journal Supplements 10 , B4 - B10, doi:10.1093/eurheartj/sum042 (2008). 28 Schmidt, S., Monk, J. M., Robinson, L. E. & M ourtzakis, M. The integrative role of leptin, oestrogen and the insulin family in obesity - associated breast cancer: potential effects of exercise. Obesity reviews : an official journal of the International Association for the Study of Obesity 16 , 473 - 487, doi:10.1111/obr.12281 (2015). 29 Gérard, C. & Brown, K. A. Obesity and breast cancer Role of estrogens and the molecular underpinnings of aromatase regulation in breast adipose tissue. Molecular and Cellular Endocrinology 466 , 15 - 30, doi:10.1016/j.mce.20 17.09.014 (2018). 30 Bhardwaj, A. et al. The isomiR - 140 - 3p - regulated mevalonic acid pathway as a potential target for prevention of triple negative breast cancer. Breast Cancer Research 20 , 150, doi:10.1186/s13058 - 018 - 1074 - z (2018). 31 Laudisio, D., Muscog iuri, G., Barrea, L., Savastano, S. & Colao, A. Obesity and breast cancer in premenopausal women: Current evidence and future perspectives. European Journal of Obstetrics & Gynecology and Reproductive Biology 230 , 217 - 221, doi:10.1016/j.ejogrb.2018.03.050 (2018). 109 32 Gregor, M. F. & Hotamisligil, G. S. Inflammatory Mechanisms in Obesity. Annual Review of Immunology 29 , 415 - 445, doi:10.1146/annurev - immunol - 031210 - 101322 (2011). 33 Kolb, R., Sutterwala, F. S. & Zhang, W. Obesity and cancer: inflammation bridges the two. Current Opinion in Pharmacology 29 , 77 - 89, doi:10.1016/j.coph.2016.07.005 (2016). 34 Balistreri, C. R., Caruso, C. & Candore, G. The role of adipose tissue and adipok ines in obesity - related inflammatory diseases. Mediators Inflamm 2010 , 802078, doi:10.1155/2010/802078 (2010). 35 Sun, K., Kusminski, C. M. & Scherer, P. E. Adipose tissue remodeling and obesity. J Clin Invest 121 , 2094 - 2101, doi:10.1172/JCI45887 (2011). 3 6 The 2015 Dietary Guidlines Advisory Committee. The Scientific Report of the 2015 Dietary Guidelines Advisory Committee , < https://health.gov/dietaryguidelines/2015 - scientific - report/PDFs/Scientific - Report - of - the - 2015 - Dietary - Guidelines - Advisory - Committee.pdf > (2015). 37 Liese, A. D. et al. The Dietary Patterns Methods Project: synthesis of findings across cohorts and relevance to dietary guidance. J Nutr 145 , 393 - 402, doi:10.3945/jn.114.205336 (2015). 38 Castello, A. et al. Adherence to the Western, Prudent and Mediterranean dietary patterns and breast cancer risk: MCC - Spain study. Maturitas 103 , 8 - 15, doi: 10.1016/j.maturitas.2017.06.020 (2017). 39 Volaco, A. Socioeconomic Status: The Missing Link Between Obesity and Diabetes Mellitus? Current diabetes reviews 14 , 321 - 326, doi:10.2174/1573399813666170621123227 (2018). 40 Sherman, S. et al. Corner Stores: The Perspective of Urban Youth. Journal of the Academy of Nutrition and Dietetics 115 , 242 - 248, doi:10.1016/j.jand.2014.10.017 (2015). 41 Zhao, X., Yang, Y., Sun, B. - F., Zhao, Y. - L. & Yang, Y. - G. FTO and Obesity: Mechanisms of Association. Current Diabetes Re ports 14 , 486, doi:10.1007/s11892 - 014 - 0486 - 0 (2014). 42 Sicat, J. in Obesity Medicine Association (Obesity Medicine Association, https://obesitymedicine.org/obesity - and - genetics/ , 2018). 43 Lew, E. A. & Garfinkel, L. Variations in mortality by weight among 750,000 men and women. Journal of Chronic Diseases 32 , 563 - 576, doi:10.1016/0021 - 9681(79)90119 - X (1979). 44 Calle, E. E., Rodriguez, C., Walker - Thurmond, K. & Thun, M. J. Overweight, Obesity, and Mortality from Cancer in a Prospectively Studied Cohort of U.S. Adults. New England Journal of Medicine 348 , 1625 - 1638, doi:10.1056/NEJMoa021423 (2003). 45 Cancer, I. A . f. R. o. IARC handbooks of cancer prevention: weight control and physical activity. Lyon, France: IARC Press 6 (2002). 46 Yager, J. D. & Davidson, N. E. Estrogen Carcinogenesis in Breast Cancer. New England Journal of Medicine 354 , 270 - 282, doi:10.1056/N EJMra050776 (2006). 47 Steele, C. et al. Vital Signs: Trends in Incidence of Cancers Associated with Overweight and Obesity United States, 2005 2014. MMWR. Morbidity and mortality weekly report 66 , 1052 - 1058, doi:10.15585/mmwr.mm6639e1 (2017). 48 Sharpe, R. et al. FGFR signaling promotes the growth of triple - negative and basal - like breast cancer cell lines both in vitro and in vivo. Clin Cancer Res 17 , 5275 - 5286, doi:10.1158/1078 - 0432.CCR - 10 - 2727 (2011). 110 49 Wahba, H. A. & El - Hadaad, H. A. Current approach es in treatment of triple - negative breast cancer. Cancer Biol Med 12 , 106 - 116, doi:10.7497/j.issn.2095 - 3941.2015.0030 (2015). 50 Siddiqui, I., Sanna, V., Ahmad, N., Sechi, M. & Mukhtar, H. Resveratrol nanoformulation for cancer prevention and therapy. Anna ls of the New York Academy of Sciences 1348 , doi:10.1111/nyas.12811 (2015). 51 Mehta, R. G., Murillo, G., Naithani, R. & Peng, X. Cancer chemoprevention by natural products: how far have we come? Pharm Res 27 , 950 - 961, doi:10.1007/s11095 - 010 - 0085 - y (2010). 52 Picon - Ruiz, M., Morata - Tarifa, C., Valle - Goffin, J. J., Friedman, E. R. & Slingerland, J. M. Obesity and adverse breast cancer risk and outcome: Mechanistic insights and strategies for intervention. CA: A Cancer Journal for Clinicians 67 , 378 - 397, doi: 10.3322/caac.21405 (2017). 53 Pierobon, M. & Frankenfeld, C. L. Obesity as a risk factor for triple - negative breast cancers: a systematic review and meta - analysis. Breast Cancer Research and Treatment 137 , 307 - 314, doi:10.1007/s10549 - 012 - 2339 - 3 (2013). 54 Sahin, S. et al. The association between body mass index and immunohistochemical subtypes in breast cancer. The Breast 32 , 227 - 236, doi:10.1016/j.breast.2016.09.019 (2017). 55 Kerlikowske, K. et al. Risk Factors That Increase Risk of Estrogen Receptor Posi tive and Negative Breast Cancer. JNCI: Journal of the National Cancer Institute 109 , doi:10.1093/jnci/djw276 (2016). 56 Li, H. et al. BMI, reproductive factors, and breast cancer molecular subtypes: A case - control study and meta - analysis. Journal of Epide miology 27 , 143 - 151, doi:10.1016/j.je.2016.05.002 (2017). 57 Jerônimo, A. F. d. A. & Weller, M. Differential Association of the Lifestyle - Related Risk Factors Smoking and Obesity with Triple Negative Breast Cancer in a Brazilian Population. Asian Pac J Can cer Prev 18 , 1585 - 1593, doi:10.22034/APJCP.2017.18.6.1585 (2017). 58 Chen, L. et al. Body mass index and risk of luminal, HER2 - overexpressing, and triple negative breast cancer. Breast Cancer Research and Treatment 157 , 545 - 554, doi:10.1007/s10549 - 016 - 3825 - 9 (2016). 59 Jeong, S. H. et al. Body mass index and risk of breast cancer molecular subtypes in Korean women: a case - control study. Breast Cancer Res Treat 179 , 459 - 470, doi:10.1007/s10549 - 019 - 05451 - 1 (2020). 60 Gershuni, V. et al. Breast cancer subtype distribution is different in normal weight, overweight, and obese women. Breast Cancer Res Treat 163 , 375 - 381, doi:10.1007/s10549 - 017 - 4192 - x (2017). 61 Yang, X. R. et al. Associations of breast cancer risk factors with tumor subtypes : a pooled analysis from the Breast Cancer Association Consortium studies. J Natl Cancer Inst 103 , 250 - 263, doi:10.1093/jnci/djq526 (2011). 62 Kim, M. S., Choi, Y. - J. & Lee, Y. H. Visceral fat measured by computed tomography and the risk of breast cancer. Translational Cancer Research 8 , 1939 - 1949, doi:10.21037/tcr.2019.09.16 (2019). 63 Millikan, R. C. et al. Epidemiology of basal - like breast can cer. Breast Cancer Res Treat 109 , 123 - 139, doi:10.1007/s10549 - 007 - 9632 - 6 (2008). 64 Coleman, W. B. Genomic Catastrophe and Neoplastic Transformation. Am J Pathol 185 , 1846 - 1849, doi:10.1016/j.ajpath.2015.05.001 (2015). 65 Cooper, G. in The Cell: A Molecula r Approach. 2nd edition (Sinauer Associates, 2000). 111 66 Chiarugi, P. & Giannoni, E. Anoikis: A necessary death program for anchorage - dependent cells. Biochemical Pharmacology 76 , 1352 - 1364, doi:10.1016/j.bcp.2008.07.023 (2008). 67 Wu, N. et al. Precisio n medicine based on tumorigenic signaling pathways for triple - negative breast cancer. Oncol Lett 16 , 4984 - 4996, doi:10.3892/ol.2018.9290 (2018). 68 Du, Z. & Lovly, C. M. Mechanisms of receptor tyrosine kinase activation in cancer. Mol Cancer 17 , 58, doi:10 .1186/s12943 - 018 - 0782 - 4 (2018). 69 Regad, T. Targeting RTK Signaling Pathways in Cancer. Cancers (Basel) 7 , 1758 - 1784, doi:10.3390/cancers7030860 (2015). 70 Butti, R. et al. Receptor tyrosine kinases (RTKs) in breast cancer: signaling, therapeutic implicat ions and challenges. Mol Cancer 17 , 34, doi:10.1186/s12943 - 018 - 0797 - x (2018). 71 Touat, M., Ileana, E., Postel - Vinay, S., André, F. & Soria, J. - C. Targeting FGFR Signaling in Cancer. Clinical Cancer Research 21 , 2684, doi:10.1158/1078 - 0432.CCR - 14 - 2329 (201 5). 72 Jeong, W. et al. The functional effects and mechanisms by which fibroblast growth factor 2 (FGF2) controls bovine mammary epithelial cells: Implications for the development and functionality of the bovine mammary gland. J Anim Sci 95 , 5365 - 5377, doi :10.2527/jas2017.1877 (2017). 73 Pandey, G. et al. ROR1 Potentiates FGFR Signaling in Basal - Like Breast Cancer. Cancers (Basel) 11 , doi:10.3390/cancers11050718 (2019). 74 Incio, J. et al. Obesity promotes resistance to anti - VEGF therapy in breast cancer by up - regulating IL - 6 and potentially FGF - 2. Science Translational Medicine 10 , eaag0945, doi:10.1126/scitranslmed.aag0945 (2018). 75 Bradley, C. A. Breast cancer: Obesity - tipping the scales of resistance. Nat Rev Cancer 18 , 265, doi:10.1038/nrc.2018.31 (2 018). 76 Li, M., Guo, Y., Feng, Y. - M. & Zhang, N. Identification of Triple - Negative Breast Cancer Genes and a Novel High - Risk Breast Cancer Prediction Model Development Based on PPI Data and Support Vector Machines. Frontiers in Genetics 10 , doi:10.3389/fg ene.2019.00180 (2019). 77 Kuhn, M. C. et al. Adipocyte - secreted factors increase osteoblast proliferation and the OPG/RANKL ratio to influence osteoclast formation. Mol Cell Endocrinol 349 , 180 - 188, doi:10.1016/j.mce.2011.10.018 (2012). 78 Hao, R. H. et al . Associations of Plasma FGF2 Levels and Polymorphisms in the FGF2 Gene with Obesity Phenotypes in Han Chinese Population. Sci Rep 6 , 19868, doi:10.1038/srep19868 (2016). 79 Chakraborty, D. et al. Fibroblast growth factor receptor is a mechanistic link bet ween visceral adiposity and cancer. Oncogene 36 , 6668 - 6679, doi:10.1038/onc.2017.278 (2017). 80 Kim, S., Ahn, C., Bong, N., Choe, S. & Lee, D. K. Biphasic effects of FGF2 on adipogenesis. PloS one 10 , e0120073 - e0120073, doi:10.1371/journal.pone.0120073 (2015). 81 Huffman, D. M. et al. Abdominal obesity, independent from caloric intake, accounts for the development of intestinal tumors in Apc(1638N/+) female mice. Cancer Prev Res (Phila) 6 , 177 - 187, doi:10.1158/1940 - 6207.C APR - 12 - 0414 (2013). 82 Lu, Y. P. et al. Surgical removal of the parametrial fat pads stimulates apoptosis and inhibits UVB - induced carcinogenesis in mice fed a high - fat diet. Proc Natl Acad Sci U S A 109 , 9065 - 9070, doi:10.1073/pnas.1205810109 (2012). 112 83 K ensler, T. W. et al. Transforming Cancer Prevention through Precision Medicine and Immune - oncology. Cancer Prev Res (Phila) 9 , 2 - 10, doi:10.1158/1940 - 6207.CAPR - 15 - 0406 (2016). 84 Hail, N., Jr., Cortes, M., Drake, E. N. & Spallholz, J. E. Cancer chemopreven tion: a radical perspective. Free Radic Biol Med 45 , 97 - 110, doi:10.1016/j.freeradbiomed.2008.04.004 (2008). 85 Uray, I. P. & Brown, P. H. Chemoprevention of hormone receptor - negative breast cancer: new approaches needed. Recent Results Cancer Res 188 , 147 - 162, doi:10.1007/978 - 3 - 642 - 10858 - 7_13 (2011). 86 Moffat, J. G., Rudolph, J. & Bailey, D. Phenotypic screening in cancer drug discovery - past, present and future. Nat Rev Drug Discov 13 , 588 - 602, doi:10.1038/nrd4366 (2014). 87 Faria, M. et al. cell line MCF - regulation of hypoxic signaling. Oncotarget 9 , 12201 - 12211, doi:10.18632/oncotarget.24134 (2018). 88 Mullen, P. J., Yu, R., Longo, J., Archer, M. C. & Penn, L. Z. The interplay between cell signalling and the mevalonate pathway in cancer. Nat Rev Cancer 16 , 718 - 731, doi:10.1038/nrc.2016.76 (2016). 89 Buhaescu, I. & Izzedine, H. Mevalonate pathw ay: A review of clinical and therapeutical implications. Clinical Biochemistry 40 , 575 - 584, doi:10.1016/j.clinbiochem.2007.03.016 (2007). 90 Buhaescu, I. & Izzedine, H. Mevalonate pathway: a review of clinical and therapeutical implications. Clin. Biochem. 40 , 575 - 584, doi:10.1016/j.clinbiochem.2007.03.016 (2007). 91 Goldstein, J. L. & Brown, M. S. Regulation of the mevalonate pathway. Nature 343 , 425 - 430, doi:10.1038/343425a0 (1990). 92 Bytyci, I. et al. Hydrophilic vs lipophilic statins in coronary artery disease: A meta - analysis of randomized controlled trials. J Clin Lipidol 11 , 624 - 637, doi:10.1016/j.jacl.2017.03.003 (2017). 93 Gobel, A., Rauner, M., Hofbauer, L. C. & Rachner, T. D. Cholesterol and beyond - The role of the mevalonate pathway in cancer b iology. Biochim Biophys Acta Rev Cancer 1873 , 188351, doi:10.1016/j.bbcan.2020.188351 (2020). 94 Osmak, M. Statins and cancer: current and future prospects. Cancer Lett 324 , 1 - 12, doi:10.1016/j.canlet.2012.04.011 (2012). 95 Yang, J. et al. Impact of statin use on cancer - specific mortality and recurrence: A meta - analysis of 60 observational studies. Medicine (Baltimore) 99 , e19596, doi:10.1097/MD.0000000000019596 (2020). 96 Mei, Z. et al. Effects of statins on cancer mortality and progression: A systematic r eview and meta - analysis of 95 cohorts including 1,111,407 individuals. Int J Cancer 140 , 1068 - 1081, doi:10.1002/ijc.30526 (2017). 97 Tan, P. et al. Effect of statins type on incident prostate cancer risk: a meta - analysis and systematic review. Asian J Andr ol 19 , 666 - 671, doi:10.4103/1008 - 682X.190327 (2017). 98 Renman, D., Lundberg, E., Gunnarsson, U. & Strigård, K. Statin consumption as a risk factor for developing colorectal cancer: a retrospective case study. World Journal of Surgical Oncology 15 , 222, do i:10.1186/s12957 - 017 - 1287 - 0 (2017). 113 99 Yi, C., Song, Z., Wan, M., Chen, Y. & Cheng, X. Statins intake and risk of liver cancer: A dose - response meta analysis of prospective cohort studies. Medicine 96 , e7435 - e7435, doi:10.1097/MD.0000000000007435 (2017). 100 Langballe, R. et al. Statin use and risk of contralateral breast cancer: a nationwide cohort study. British Journal of Cancer 119 , 1297 - 1305, doi:10.1038/s41416 - 018 - 0252 - 1 (2018). 101 Kumar, A. S. et al. Estrogen Receptor Negative Breast Cancer Is Less Likely to Arise among Lipophilic Statin Users. Cancer Epidemiology Biomarkers &amp; Prevention 17 , 1028, doi:10.1158/1055 - 9965.EPI - 07 - 0726 (2008). 102 Sakellakis, M., Akinosoglou, K., Kostaki, A., Spyropoulou, D . & Koutras, A. Statins and risk of breast cancer recurrence. Breast Cancer (Dove Med Press) 8 , 199 - 205, doi:10.2147/BCTT.S116694 (2016). 103 Islam, M. M. et al. Exploring association between statin use and breast cancer risk: an updated meta - analysis. Arc hives of Gynecology and Obstetrics 296 , 1043 - 1053, doi:10.1007/s00404 - 017 - 4533 - 3 (2017). 104 Manthravadi, S., Shrestha, A. & Madhusudhana, S. Impact of statin use on cancer recurrence and mortality in breast cancer: A systematic review and meta - analysis. I nternational Journal of Cancer 139 , 1281 - 1288, doi:10.1002/ijc.30185 (2016). 105 - risk individuals: a randomised placebocontrolled trial. The Lancet 360 , 7 - 22, doi:10.101 6/S0140 - 6736(02)09327 - 3 (2002). 106 Officers, T. A. & Group, C. f. t. A. C. R. Major Outcomes in Moderately Hypercholesterolemic, Hypertensive Patients Randomized to Pravastatin vs Usual CareThe Antihypertensive and Lipid - Lowering Treatment to Prevent Hear t Attack Trial (ALLHAT - LLT). JAMA 288 , 2998 - 3007, doi:10.1001/jama.288.23.2998 (2002). 107 Boudreau, D. M. et al. Statin Use and Breast Cancer Risk in a Large Population - Based Setting. Cancer Epidemiology Biomarkers &amp; Prevention 16 , 416, doi:10.115 8/1055 - 9965.EPI - 06 - 0737 (2007). 108 Cauley, J. A. et al. Statin Use and Breast Cancer: Prospective Results From the Women's Health Initiative. JNCI: Journal of the National Cancer Institute 98 , 700 - 707, doi:10.1093/jnci/djj188 (2006). 109 Ahn, J., Schatzki n, A., Lacey, J. V., Jr & et al. Adiposity, adult weight change, and postmenopausal breast cancer risk. Archives of Internal Medicine 167 , 2091 - 2102, doi:10.1001/archinte.167.19.2091 (2007). 110 Connolly, B. S. et al. A meta - analysis of published literatur e on waist - to - hip ratio and risk of breast cancer. Nutr Cancer 44 , 127 - 138, doi:10.1207/S15327914NC4402_02 (2002). 111 Gaudet, M. M. et al. Waist circumference, body mass index, and postmenopausal breast cancer incidence in the Cancer Prevention Study - II N utrition Cohort. Cancer Causes & Control 25 (2013). 112 Hajian - Tilaki, K. O., Gholizadehpasha, A. R., Bozorgzadeh, S. & Hajian - Tilaki, E. Body mass index and waist circumference are predictor biomarkers of breast cancer risk in Iranian women. Med Oncol 28 , 1296 - 1301, doi:10.1007/s12032 - 010 - 9629 - 6 (2011). 113 Huang, Z. et al. Waist circumference, waist:hip ratio, and risk of breast cancer in the nurses' health study. American Journal of Epidemiology 150 , 8 (1999). 114 114 Harvie, M., Hooper, L. & Howell, A. H. Central obesity and breast cancer risk: a systematic review. Obesity reviews : an official journal of the International Association for the Study of Obesity 4 , 157 - 173 (2003). 115 Schaffler, A., Scholmerich, J. & Buec hler, C. Mechanisms of disease: adipokines and breast cancer - endocrine and paracrine mechanisms that connect adiposity and breast cancer. Nat Clin Pract Endocrinol Metab 3 , 345 - 354, doi:10.1038/ncpendmet0456 (2007). 116 Kwon, H., Kim, D. & Kim, J. S. Bod y Fat Distribution and the Risk of Incident Metabolic Syndrome: A Longitudinal Cohort Study. Sci Rep 7 , 10955, doi:10.1038/s41598 - 017 - 09723 - y (2017). 117 Esser, N., Legrand - Poels, S., Piette, J., Scheen, A. J. & Paquot, N. Inflammation as a link between ob esity, metabolic syndrome and type 2 diabetes. Diabetes Res Clin Pract 105 , 141 - 150, doi:10.1016/j.diabres.2014.04.006 (2014). 118 Jung, U. J. & Choi, M. S. Obesity and its metabolic complications: the role of adipokines and the relationship between obesit y, inflammation, insulin resistance, dyslipidemia and nonalcoholic fatty liver disease. Int J Mol Sci 15 , 6184 - 6223, doi:10.3390/ijms15046184 (2014). 119 Kabat, G. C. et al. Metabolic obesity phenotypes and risk of breast cancer in postmenopausal women. Cancer Epidemiol Biomarkers Prev , doi:10.1158/1055 - 9965.EPI - 17 - 0495 (2017). 120 Elbauomy Elsheikh, S. et al. FGFR1 amplification in breast carcinomas: a chromogenic in situ hy bridisation analysis. Breast Cancer Res 9 , R23, doi:10.1186/bcr1665 (2007). 121 Wittwer, J. A. et al. Enhancing Mitochondrial Respiration Suppresses Tumor Promoter TPA - Induced PKM2 Expression and Cell Transformation in Skin Epidermal JB6 Cells. Cancer Prev ention Research 4 , 1476 - 1484, doi:10.1158/1940 - 6207.capr - 11 - 0028 (2011). 122 Chang, A. et al. Prevention of skin carcinogenesis by the beta - blocker carvedilol. Cancer Prev Res (Phila) 8 , 27 - 36, doi:10.1158/1940 - 6207.CAPR - 14 - 0193 (2015). 123 Lu, F. J. et al . Promoting neoplastic transformation of humic acid in mouse epidermal JB6 Cl41 cells. Chem Biol Interact 162 , 249 - 258, doi:10.1016/j.cbi.2006.07.007 (2006). 124 Pang, Y. et al. Development of human cell models for assessing the carcinogenic potential of c hemicals. Toxicol Appl Pharmacol 232 , 478 - 486, doi:10.1016/j.taap.2008.08.009 (2008). 125 Balmain, A. & Harris, C. Carcinogenesis in mouse and human cells - parallels and paradoxes. Carcinogenesis 12 , 371 - 377 (2000). 126 Zhang, X. et al. Receptor specificit y of the fibroblast growth factor family. The complete mammalian FGF family. J Biol Chem 281 , 15694 - 15700, doi:10.1074/jbc.M601252200 (2006). 127 Ornitz, D. M. et al. Receptor Specificity of the Fibroblast Growth Factor Family. Journal of Biological Chemis try 271 , 15292 - 15297, doi:10.1074/jbc.271.25.15292 (1996). 128 Schapira, D. et al. Visceral Obesity and Breast Cancer Risk. Cancer 74 , 632 (1994). 129 Rankinen, T., Kim, S. Y., Pérusse, L., Després, J. P. & Bouchard, C. The prediction of abdominal visceral fat level from body composition and anthropometry: ROC analysis. International Journal of Obesity 23 , 801 - 809, doi:10.1038/sj.ijo.0800929 (1999). 130 Brady, N., Chuntova, P., Bade, L. K. & Schwertfeger, K. L. The FGF/FGFR axis as a therapeutic target in b reast cancer. Expert Rev Endocrinol Metab 8 , 391 - 402, doi:10.1586/17446651.2013.811910 (2013). 115 131 Welm, B. E. et al. Inducible dimerization of FGFR1: development of a mouse model to analyze progressive transformation of the mammary gland. J Cell Biol 157 , 703 - 714, doi:10.1083/jcb.200107119 (2002). 132 Vandermoere, F., El Yazidi - Belkoura, I., Adriaenssens, E., Lemoine, J. & Hondermarck, H. The antiapoptotic effect of fibroblast growth factor - 2 is mediated through nuclear factor - kappaB activation induced via interaction between Akt and IkappaB kinase - beta in breast cancer cells. Oncogene 24 , 5482 - 5491, doi:10.1038/sj.onc.1208713 (2005). 133 Hu, Y. et al. miRNA - 205 targets VEGFA and FGF2 and regulates resistance to chemotherapeutics in breast cancer. Cell Deat h Dis 7 , e2291, doi:10.1038/cddis.2016.194 (2016). 134 Lui, J., Crepin, M., Liu, J., Barritault, D. & Ledoux, D. FGF - 2 and TPA induce matrix metalloproteinase - 9 secretion in MCF - 7 cells through PKC activation of the Ras - ERK pathway. Biochemical and Biophys ical Research Communications 293 , 8 (2002). 135 Xian, W. et al. Fibroblast growth factor receptor 1 - transformed mammary epithelial cells are dependent on RSK activity for growth and survival. Cancer Res 69 , 2244 - 2251, doi:10.1158/0008 - 5472.CAN - 08 - 3398 (200 9). 136 Xian , W., Schwertfeger , K. L., Vargo - Gogola , T. & Rosen , J. M. Pleiotropic effects of FGFR1 on cell proliferation, survival, and migration in a 3D mammary epithelial cell model. Journal of Cell Biology 171 , 663 - 673, doi:10.1083/jcb.200505098 (2005). 137 Rykala, J. et al. The - 553 T/A polymorphism in the promoter region of the FGF2 gene is associated with increased breast cancer risk in Polish women. Arch Med Sci 11 , 619 - 627, doi:10.5114/aoms.2013.35996 (2015). 138 Ryk ala, J. et al. Angiogenesis markers quantification in breast cancer and their correlation with clinicopathological prognostic variables. Pathol Oncol Res 17 , 809 - 817, doi:10.1007/s12253 - 011 - 9387 - 6 (2011). 139 Sauter, E. R. et al. Biomarkers associated with breast cancer are associated with obesity. Cancer Detect Prev 32 , 149 - 155, doi:10.1016/j.cdp.2008.06.002 (2008). 140 Hsiung, R. et al. High basic fibroblast growth factor levels in nipple aspirate fluid are correlated with breast cancer. The Cancer Journa l 8 , 7 (2002). 141 Turner, N. et al. FGFR1 amplification drives endocrine therapy resistance and is a therapeutic target in breast cancer. Cancer Res 70 , 2085 - 2094, doi:10.1158/0008 - 5472.CAN - 09 - 3746 (2010). 142 Agarwal, D. et al. FGF receptor genes and breast cancer susceptibility: results from the Breast Cancer Association Consortium. Br J Cancer 110 , 1088 - 1100, doi:10.1038/bjc.2013.769 (2014). 143 Slattery, M. L. et al. Associations with growth factor genes (FGF1, FGF2, PDGFB, F GFR2, NRG2, EGF, ERBB2) with breast cancer risk and survival: the Breast Cancer Health Disparities Study. Breast Cancer Res Treat 140 , 587 - 601, doi:10.1007/s10549 - 013 - 2644 - 5 (2013). 144 Maehle, B., Tretly, S., Skjaerven, R. & Thorsen, T. Premorbid body wei ght and its relations to primary tumour diameter in breast cancer patients; its dependence on estrogen and progesterone receptor status. Breast Cancer Research and Treatment 68 , 10 (2001). 145 Ornitz, D. M. & Itoh, N. The Fibroblast Growth Factor signaling pathway. Wiley Interdiscip Rev Dev Biol 4 , 215 - 266, doi:10.1002/wdev.176 (2015). 146 Harbinski, F. et al. Rescue screens with secreted proteins reveal compensatory potential of receptor tyrosine kinases in driving cancer growth. Cancer Discov 2 , 948 - 959, doi:10.1158/2159 - 8290.CD - 12 - 0237 (2012). 116 147 Shaw, K. R., Wrobel, C. N. & Brugge, J. S. Use of three - dimensional basement membrane cultures to model oncogene - induced changes in mammary epithelial morphogenesis. J Mammary Gland Biol Neoplasia 9 , 297 - 310, do i:10.1007/s10911 - 004 - 1402 - z (2004). 148 Homer - Bouthiette, C., Doetschman, T., Xiao, L. & Hurley, M. M. Knockout of nuclear high molecular weight FGF2 isoforms in mice modulates bone and phosphate homeostasis. J Biol Chem 289 , 36303 - 36314, doi:10.1074/jbc.M 114.619569 (2014). 149 Pan, R. L. et al. Low - molecular - weight fibroblast growth factor 2 attenuates hepatic fibrosis by epigenetic down - regulation of Delta - like1. Hepatology 61 , 1708 - 1720, doi:10.1002/hep.27649 (2015). 150 Kole, D. et al. High molecular we ight FGF2 isoforms demonstrate canonical receptor - mediated activity and support human embryonic stem cell self - renewal. Stem Cell Res 21 , 106 - 116, doi:10.1016/j.scr.2017.04.006 (2017). 151 Mydlo, J. H., Kral, J. G. & Macchia, R. J. PRELIMINARY RESULTS COMP ARING THE RECOVERY OF BASIC FIBROBLAST GROWTH FACTOR (FGF - 2) IN ADIPOSE TISSUE AND BENIGN AND MALIGNANT RENAL TISSUE. The Journal of Urology 159 , 2159 - 2163, doi:10.1016/S0022 - 5347(01)63298 - 1. 152 Grazul - Bilska, A. et al. Effects of basic fibroblast growth factor (FGF - 2) on proliferation of human skin fibroblasts in type II diabetes mellitus. Experimental and clinical endocrinology & diabetes 110 , 176 - 181, doi:10.1055/s - 2002 - 32149 (2002). 153 Itoh, N. & Ohta, H. Pathophysiological roles of FGF signaling in t he heart. Front Physiol 4 , 247, doi:10.3389/fphys.2013.00247 (2013). 154 Matsui, R., Brody, J. & Yu, Q. FGF - 2 Induces Surfactant Protein Gene Expression in Foetal Rat Lung Epithelial Cells Through a MAPK - Independent Pathway. Elsevier 11 , 221 - 228 (1999). 155 Gabrielsson, B. et al. Depot - Specific Expression of Fibroblast Growth F actors in Human Adipose Tissue. Obesity Research 10 , 608 - 617 (2002). 156 Wang, Y. Y. et al. Adipose tissue and breast epithelial cells: a dangerous dynamic duo in breast cancer. Cancer Lett 324 , 142 - 151, doi:10.1016/j.canlet.2012.05.019 (2012). 157 Zhao, M . et al. Mesenchymal stem cells in mammary adipose tissue stimulate progression of breast cancer resembling the basal - type. Cancer Biol Ther 13 , 782 - 792, doi:10.4161/cbt.20561 (2012). 158 Ligibel, J. A. et al. American Society of Clinical Oncology position statement on obesity and cancer. J Clin Oncol 32 , 3568 - 3574, doi:10.1200/JCO.2014.58.4680 (2014). 159 Flegal, K. M., Kruszon - Moran, D., Carroll, M. D., Fryar, C. D. & Ogden, C. L. Trends in Obesity Among Adults in the United States, 2005 to 2014. JAMA 315 , 2284 - 2291, doi:10.1001/jama.2016.6458 (2016). 160 Benham, V., Chakraborty, D., Bullard, B. & Bernard, J. J. A role for FGF2 in visceral adiposity - associated mammary epithelial transformation. Adipocyte , 00 - 00, doi:10.1080/21623945.2018.1445889 (2018). 16 1 Chakraborty, D., Benham, V. & Bernard, J. J. Elucidating the role of adipose tissue secreted factors in malignant transformation. Adipocyte , 00 - 00, doi:10.1080/21623945.2017.1388971 (2017). 162 Sporn, M. B. & Liby, K. T. Cancer chemoprevention: scientifi c promise, clinical uncertainty. Nat Clin Pract Oncol 2 , 518 - 525, doi:10.1038/ncponc0319 (2005). 117 163 Al Rabadi, L. & Bergan, R. A Way Forward for Cancer Chemoprevention: Think Local. Cancer Prev Res (Phila) 10 , 14 - 35, doi:10.1158/1940 - 6207.CAPR - 16 - 0194 (20 17). 164 Shin, S. I., Freedman, V. H., Risser, R. & Pollack, R. Tumorigenicity of virus - transformed cells in nude mice is correlated specifically with anchorage independent growth in vitro. Proc Natl Acad Sci U S A 72 , 4435 - 4439 (1975). 165 Jain, M. V. PKB /Akt kinase localization and role in stemness maintenance in cancer . (2016). 166 Rotem, A. et al. Alternative to the soft - agar assay that permits high - throughput drug and genetic screens for cellular transformation. Proc Natl Acad Sci U S A 112 , 5708 - 5713 , doi:10.1073/pnas.1505979112 (2015). 167 Colburn, N. H. et al. Correlation of Anchorage - independent Growth with Tumorigenicity of Chemically Transformed Mouse Epidermal Cells. Cancer Research 38 , 624 (1978). 168 Fang, Y. & Eglen, R. M. Three - Dimensional Cell Cultures in Drug Discovery and Development. SLAS Discov 22 , 456 - 472, doi:10.1177/1087057117696795 (2017). 169 Duval, K. et al. Modeling Physiological Events in 2D vs. 3D Cell Culture. Physiology (Bethesda) 32 , 266 - 277, doi:10.1152/physiol.00036.2016 (2017). 170 Chakraborty, D. et al. A BET bromodomain inhibitor suppresses adiposity - associated malignant transformation. Cancer Prev Res (Phila) , doi:10.1158/1940 - 6207.CAPR - 17 - 0262 (2017). 171 Roy, A. Early Probe an d Drug Discovery in Academia: A Minireview. High - Throughput 7 , 4, doi:10.3390/ht7010004 (2018). 172 Ekman, S. et al. Clinical Phase I study with an Insulin - like Growth Factor - 1 receptor inhibitor: experiences in patients with squamous non - small cell lung c arcinoma. Acta Oncol 50 , 441 - 447, doi:10.3109/0284186X.2010.499370 (2011). 173 Zhang, Q., Pan, J., Lubet, R. A., Wang, Y. & You, M. Targeting the insulin - like growth factor - 1 receptor by picropodophyllin for lung cancer chemoprevention. Mol Carcinog 54 Sup pl 1 , E129 - 137, doi:10.1002/mc.22206 (2015). 174 Menu, E. et al. Targeting the IGF - 1R using picropodophyllin in the therapeutical 5T2MM mouse model of multiple myeloma: beneficial effects on tumor growth, angiogenesis, bone disease and survival. Int J Cancer 121 , 1857 - 1861, doi:10.1002/ijc.22845 (2007). 175 McFarland, A. J. et al. Molecular mechanisms underlying the effects of statins in the central nervous system. Int J Mol Sci 15 , 20607 - 20637, doi:10.3390/ijms151120607 (2014). 176 Torre, L. A., Siegel, R. L., Ward, E. M. & Jemal, A. Global Cancer Incidence and Mortali ty Rates and Trends An Update. Cancer Epidemiology Biomarkers & Prevention 25 , 16 - 27, doi:10.1158/1055 - 9965.epi - 15 - 0578 (2016). 177 Ferlay, J. et al. Cancer incidence and mortality worldwide: Sources, methods and major patterns in GLOBOCAN 2012. Intern ational Journal of Cancer 136 , E359 - E386, doi:10.1002/ijc.29210 (2015). 178 Siegel, R. L., Miller, K. D. & Jemal, A. Cancer statistics, 2018. CA: A Cancer Journal for Clinicians 68 , 7 - 30, doi:10.3322/caac.21442 (2018). 179 Siegel, R. L., Miller, K. D. & Je mal, A. Cancer statistics, 2017. CA: A Cancer Journal for Clinicians 67 , 7 - 30, doi:10.3322/caac.21387 (2017). 118 180 Borowicz, S. et al. The soft agar colony formation assay. J Vis Exp , e51998, doi:10.3791/51998 (2014). 181 Horibata, S., Vo, T. V., Subramania n, V., Thompson, P. R. & Coonrod, S. A. Utilization of the Soft Agar Colony Formation Assay to Identify Inhibitors of Tumorigenicity in Breast Cancer Cells. J Vis Exp , e52727, doi:10.3791/52727 (2015). 182 Anderson, S. N., Towne, D. L., Burns, D. J. & Warr ior, U. A high - throughput soft agar assay for identification of anticancer compound. J Biomol Screen 12 , 938 - 945, doi:10.1177/1087057107306130 (2007). 183 Horman, S. R., To, J. & Orth, A. P. An HTS - compatible 3D colony formation assay to identify tumor - spe cific chemotherapeutics. J Biomol Screen 18 , 1298 - 1308, doi:10.1177/1087057113499405 (2013). 184 Thierbach, R. & Steinberg, P. Automated soft agar assay for the high - throughput screening of anticancer compounds. Analytical Biochemistry 387 , 318 - 320, doi:10 .1016/j.ab.2009.01.029 (2009). 185 Hirsch, H. A., Iliopoulos, D. & Struhl, K. Metformin inhibits the inflammatory response associated with cellular transformation and cancer stem cell growth. Proceedings of the National Academy of Sciences 110 , 972 - 977, do i:10.1073/pnas.1221055110 (2013). 186 Wang, W. et al. Prevention of prostate cancer by natural product MDM2 inhibitor GS25: In vitro and in vivo activities and molecular mechanisms. Carcinogenesis , doi:10.1093/carcin/bgy063 (2018). 187 Pollak, M. Insulin a nd insulin - like growth factor signalling in neoplasia. Nature Reviews Cancer 8 , 915, doi:10.1038/nrc2536 (2008). 188 Kasznicki, J., Sliwinska, A. & Drzewoski, J. Metformin in cancer prevention and therapy. Ann Transl Med 2 , 57, doi:10.3978/j.issn.2305 - 5839 .2014.06.01 (2014). 189 McFarland, A. et al. Molecular Mechanisms Underlying the Effects of Statins in the Central Nervous System. International Journal of Molecular Sciences 15 , 20607 (2014). 190 Awad, K. et al. Efficacy and Safety of Alternate - Day Versus Daily Dosing of Statins: a Systematic Review and Meta - Analysis. Cardiovasc Drugs Ther 31 , 419 - 431, doi:10.1007/s10557 - 017 - 6743 - 0 (2017). 191 Goard, C. A. et al. Identifying molecular features that distinguish fluvastatin - sensitive breast tumor cells. Brea st Cancer Research and Treatment 143 , 301 - 312, doi:10.1007/s10549 - 013 - 2800 - y (2014). 192 Campbell, M. J. et al. Breast Cancer Growth Prevention by Statins. Cancer Research 66 , 8707 - 8714, doi:10.1158/0008 - 5472.can - 05 - 4061 (2006). 193 Wellberg, E. A. et al. FGFR1 underlies obesity - associated progression of estrogen receptor - positive breast cancer after estrogen deprivation. JCI Insight 3 , doi:10.1172/jci.insight.120594 (2018). 194 Browning, D. R. & Martin, R. M. Statins and risk of cancer: a systematic review and metaanalysis. Int J Cancer 120 , 833 - 843, doi:10.1002/ijc.22366 (2007). 195 Yang, K. et al. Statin use and non - melanoma skin cancer risk: a meta - analysis of randomized controlled trials and observational studies. Oncotarget 8 , 75411 - 75417, doi:10.18632/oncotarget.20034 (2017). 119 196 Liu, B., Yi, Z., Guan, X., Zeng, Y. X. & Ma, F. The relationship between statins and breast cancer prognosis varies by statin type and exposure time: a meta - analysis. Breast Cancer Res Treat 164 , 1 - 11, doi:10.1007/s10549 - 017 - 4246 - 0 (2017). 197 Osmak, M. Statins and cancer: Current and future prospects. Cancer Letters 32 4 , 1 - 12, doi:10.1016/j.canlet.2012.04.011 (2012). 198 Alfaqih, M. A., Allott, E. H., Hamilton, R. J., Freeman, M. R. & Freedland, S. J. The current evidence on statin use and prostate cancer prevention: are we there yet? Nature Reviews Urology 14 , 107, doi :10.1038/nrurol.2016.199 (2016). 199 Gershuni, V. et al. Breast cancer subtype distribution is different in normal weight, overweight, and obese women. Breast Cancer Research and Treatment 163 , 375 - 381, doi:10.1007/s10549 - 017 - 4192 - x (2017). 200 Harris, H. R., Willett, W. C., Terry, K. L. & Michels, K. B. Body Fat Distribution and Risk of JNCI: Journal of the National Cancer Institute 103 , 273 - 278, doi:10.1093/jnci/djq500 (2010). 201 Theillet, C. et al. FGFRI and PLAT genes and DNA amplification at 8p12 in breast and ovarian cancers. Genes Chromosomes Cancer 7 , 219 - 226 (1993). 202 Perez - Garcia, J., Munoz - Couselo, E., Soberino, J., Racca, F. & Cortes, J. Targeting FGFR pathway in breast cancer. Breast 37 , 126 - 133, doi:10.1016/j.breast.2017.10.014 (2018). 203 Helsten, T. et al. The FGFR Landscape in Cancer: Analysis of 4,853 Tumors by Next - Generation Sequencing. Clin. Cancer Res. 22 , 259 - 267, doi:10.1158/1078 - 0432.CCR - 14 - 3212 (2016). 204 Hoeflich, K. P. et al. In vivo antitumor activity of MEK and phosphatidylinositol 3 - kinase inhibitors in basal - like breast cancer models. Clin. Cancer Res. 15 , 4649 - 4664, doi:10.1158/1078 - 0432.CCR - 09 - 0317 (2009). 205 Balko, J. M. et al. Activation of MAPK pathways due to DUSP4 loss promotes cancer stem cell - like phenotypes in basal - like breast cancer. Cancer Res. 73 , 6346 - 6358, doi:10.1158/0008 - 5472.CAN - 13 - 1385 (2013). 206 Valade, E. et al. Assessment of the effect of erdafitinib on cardiac safety: analysis of ECGs and ex posure - QTc in patients with advanced or refractory solid tumors. Cancer Chemother. Pharmacol. 84 , 621 - 633, doi:10.1007/s00280 - 019 - 03896 - 1 (2019). 207 Chae, Y. K. et al. Inhibition of the fibroblast growth factor receptor (FGFR) pathway: the current landsca pe and barriers to clinical application. Oncotarget 8 , 16052 - 16074, doi:10.18632/oncotarget.14109 (2017). 208 Galiè, M. RAS as Supporting Actor in Breast Cancer. Frontiers in Oncology 9 , doi:10.3389/fonc.2019.01199 (2019). 209 Benham, V. et al. Identifying chemopreventive agents for obesity - associated cancers using an efficient, 3D high - throughput transformation assay. Sci Rep 9 , 10278, doi:10.1038/s41598 - 019 - 46531 - y (2019). 210 Katz, M. S. Therapy Insight: potential of statins for cancer chemoprevention an d therapy. Nature Clinical Practice Oncology 2 , 82 - 89, doi:10.1038/ncponc0097 (2005). 211 Clendening, J. W. et al. Dysregulation of the mevalonate pathway promotes transformation. Proc Natl Acad Sci U S A 107 , 15051 - 15056, doi:10.1073/pnas.0910258107 (2010). 120 212 Butt, S. et al. The target for statins, HMG - CoA reductase, is expressed in ductal carcinoma - in situ and may predict patient response to radiotherapy. Ann. Surg. Oncol. 21 , 2911 - 2919, doi:10.1245/s10434 - 014 - 3708 - 4 (20 14). 213 Peltola, P. et al. Visceral Obesity is Associated with High Levels of Serum Squalene. Obesity 14 , 1155 - 1163, doi:10.1038/oby.2006.132 (2006). 214 Rageh, A. H., Atia, N. N. & Abdel - Rahman, H. M. Lipophilicity estimation of statins as a decisive phy sicochemical parameter for their hepato - selectivity using reversed - phase thin layer chromatography. Journal of Pharmaceutical and Biomedical Analysis 142 , 7 - 14, doi:10.1016/j.jpba.2017.04.037 (2017). 215 Jiang, S. - Y. et al. Discovery of a potent HMG - CoA re ductase degrader that eliminates statin - induced reductase accumulation and lowers cholesterol. Nature Communications 9 , 5138, doi:10.1038/s41467 - 018 - 07590 - 3 (2018). 216 Brown, M. S. & Goldstein, J. L. Multivalent feedback regulation of HMG CoA reductase, a control mechanism coordinating isoprenoid synthesis and cell growth. Journal of Lipid Research 21 , 505 - 517 (1980). 217 Istvan, E. Statin inhibition of HMG - CoA reductase: a 3 - dimensional view. Atherosclerosis Supplements 4 , 3 - 8, doi:10.1016/S1567 - 5688(03)0 0003 - 5 (2003). 218 Kubatka, P. et al. Preventive effects of fluvastatin in rat mammary carcinogenesis. European Journal of Cancer Prevention 22 , 352 - 357, doi:10.1097/CEJ.0b013e32835b385d (2013). 219 DeBose - Boyd, R. A. Feedback regulation of cholesterol syn thesis: sterol - accelerated ubiquitination and degradation of HMG CoA reductase. Cell Res 18 , 609 - 621, doi:10.1038/cr.2008.61 (2008). 220 Moon, A. et al. H - ras, but not N - ras, induces an invasive phenotype in human breast epithelial cells: A role for MMP - 2 in the h - ras - induced invasive phenotype. International Journal of Cancer 85 , 176 - 181, doi:10.1002/(SICI)1097 - 0215(20000115)85:2 (2000). 221 Kang, S., Kim, E. S. & Moon, A. Simvastatin and lovastatin inhibit breast cell invasion induced by H - Ras. Oncol Rep 21 , 1317 - 1322, doi:10.3892/or_00000357 (2009). 222 Gbelcova, H. et al. Isoprenoids responsible for protein prenylation modulate the biological effects of statins on pancreatic cancer cells. Lipids Health Dis 16 , 250, doi:10.1186/s12944 - 017 - 0641 - 0 (2017). 223 Korsensky, L. & Ron, D. Regulation of FGF signaling: Recent insight s from studying positive and negative modulators. Seminars in Cell & Developmental Biology 53 , 101 - 114, doi:10.1016/j.semcdb.2016.01.023 (2016). 224 Itoh, N., Nakayama, Y. & Konishi, M. Roles of FGFs As Paracrine or Endocrine Signals in Liver Development, Health, and Disease. Front Cell Dev Biol 4 , 30 - 30, doi:10.3389/fcell.2016.00030 (2016). 225 La Venuta, G., Zeitler, M., Steringer, J. P., Muller, H. M. & Nickel, W. The Startling Properties of Fibroblast Growth Factor 2: How to Exit Mammalian Cells without a Signal Peptide at Hand. J Biol Chem 290 , 27015 - 27020, doi:10.1074/jbc.R115.689257 (2015). 226 Kita, S., Maeda, N. & Shimomura, I. Interorgan communication by exosomes, adipose tissue, and adiponectin in metabolic syndrome. J Clin Invest 129 , 4041 - 4049, doi:10.1172/JCI129193 (2019). 227 Flaherty, S. E. et al. A lipase - independent pathway of lipid release and immune modulation by adipocytes. Science 363 , 989, doi:10.1126/science.aaw2586 (2019). 121 228 Kim, Y. - S. et al. Adipose stem cell - derived nanovesicles i nhibit emphysema primarily via an FGF2 - dependent pathway. Experimental & Molecular Medicine 49 , e284 - e284, doi:10.1038/emm.2016.127 (2017). 229 Liao, S. et al. The influence of FGF2 high molecular weight (HMW) isoforms in the development of cardiac ischemi a reperfusion injury. Journal of Molecular and Cellular Cardiology 48 , 1245 - 1254, doi:10.1016/j.yjmcc.2010.01.014 (2010). 230 Yang, L. et al. MiR - 214 Attenuates Osteogenic Differentiation of Mesenchymal Stem Cells via Targeting FGFR1. Cellular Physiology and Biochemistry 38 , 809 - 820, doi:10.1159/000443036 (2016). 231 Erdafitinib Efficacious in Bladder Cancer. Cancer Discovery 8 , OF6, doi:10.115 8/2159 - 8290.CD - NB2018 - 085 (2018). 232 Bohrer, L. R. et al. Activation of the FGFR - STAT3 pathway in breast cancer cells induces a hyaluronan - rich microenvironment that licenses tumor formation. Cancer Res 74 , 374 - 386, doi:10.1158/0008 - 5472.CAN - 13 - 2469 (2014 ). 233 Helguero, L. A., Faulds, M. H., Gustafsson, J. - Å. & Haldosén, L. - epithelial cell line HC11. Oncogene 24 , 6605 - 6616, doi: 10.1038/sj.onc.1208807 (2005). 234 Stengel, K. & Zheng, Y. Cdc42 in oncogenic transformation, invasion, and tumorigenesis. Cellular Signalling 23 , 1415 - 1423, doi:10.1016/j.cellsig.2011.04.001 (2011). 235 Philippe, F. & Anne, B. Rho Gtpases: Molecular Biolo gy In Health And Disease . (World Scientific Publishing Company, 2017). 236 Roberts, D., Dive, C. & Renehan, A. Biological Mechanisms Linking Obesity and Cancer Risk: New Perspectives. Annual review of medicine 61 , 301 - 316, doi:10.1146/annurev.med.080708.0 82713 (2009). 237 Farabaugh, S. M., Boone, D. N. & Lee, A. V. Role of IGF1R in Breast Cancer Subtypes, Stemness, and Lineage Differentiation. Front Endocrinol (Lausanne) 6 , 59 - 59, doi:10.3389/fendo.2015.00059 (2015). 238 Papa, V. et al. Elevated insulin re ceptor content in human breast cancer. J Clin Invest 86 , 1503 - 1510, doi:10.1172/JCI114868 (1990). 239 Carey, L. A. et al. Race, Breast Cancer Subtypes, and Survival in the Carolina Breast Cancer Study. JAMA 295 , 2492 - 2502, doi:10.1001/jama.295.21.2492 (200 6). 240 Stead, L. A. et al. Triple - negative breast cancers are increased in black women regardless of age or body mass index. Breast cancer research : BCR 11 , R18 - R18, doi:10.1186/bcr2242 (2009). 241 Boyle, P. Triple - negative breast cancer: epidemiological considerations and recommendations. Annals of Oncology 23 , vi7 - vi12, doi:10.1093/annonc/mds187 (2012). 242 Kalla Singh, S., Tan, Q. W., Brito, C., De León, M. & De León, D. Insulin - like growth factors I and II receptors in the breast cancer survival dispa rity among African - American women. Growth Horm IGF Res 20 , 245 - 254, doi:10.1016/j.ghir.2010.03.001 (2010). 243 Sarfstein, R., Maor, S., Reizner, N., Abramovitch, S. & Werner, H. Transcriptional regulation of the insulin - like growth factor - I receptor gene i n breast cancer. Molecular and Cellular Endocrinology 252 , 241 - 246, doi:10.1016/j.mce.2006.03.018 (2006). 122 244 Wen, C., Wu, L., Fu, L., Wang, B. & Zhou, H. Unifying mechanism in the initiation of breast cancer by metabolism of estrogen (Review). Mol Med Rep 16 , 1001 - 1006, doi:10.3892/mmr.2017.6738 (2017). 245 Sex hormones and risk of breast cancer in premenopausal women: a collaborative reanalysis of individual participant data from seven prospective studies. The Lancet Oncology 14 , 1009 - 1019, doi:10.1016/S1 470 - 2045(13)70301 - 2 (2013). 246 Rebbeck, T. R. et al. Prophylactic Oophorectomy in Carriers of BRCA1 or BRCA2 Mutations. New England Journal of Medicine 346 , 1616 - 1622, doi:10.1056/NEJMoa012158 (2002). 247 Systemic treatment of early breast cancer by hormo nal, cytotoxic, or immune therapy. 133 randomised trials involving 31,000 recurrences and 24,000 deaths among 75,000 women. Early Breast Cancer Trialists' Collaborative Group. Lancet 339 , 71 - 85. 248 Sun, H. et al. Triple - negative breast cancer and its asso ciation with obesity. Mol Clin Oncol 7 , 935 - 942, doi:10.3892/mco.2017.1429 (2017). 249 Gupta, P. B. et al. Systemic Stromal Effects of Estrogen Promote the Growth of Estrogen Receptor Negative Cancers. Cancer Research 67 , 2062, doi:10.1158/0008 - 5472.CAN - 06 - 3895 (2007). 250 Meng, Y. & Zong, L. Estrogen stimulates SREBP2 expression in hepatic cell lines via an estrogen response element in the SREBP2 promoter. Cellular & Molecular Biology Letters 24 , 65, doi:10.1186/s11658 - 019 - 0194 - 5 (2019). 251 Sanchez - Vega, F. et al. Oncogeni c Signaling Pathways in The Cancer Genome Atlas. Cell 173 , 321 - 337.e310, doi:10.1016/j.cell.2018.03.035 (2018). 252 Remy, E. et al. A Modeling Approach to Explain Mutually Exclusive and Co - Occurring Genetic Alterations in Bladder Tumorigenesis. Cancer Res 75 , 4042 - 4052, doi:10.1158/0008 - 5472.CAN - 15 - 0602 (2015).