Raju Khatri Email:[email protected] EDUCATION Doctor of Philosophy in Molecular and Mechanistic Toxicology, 2013 University of Maryland, Baltimore, Maryland, USA

Raju Khatri Email:Ghiraju@Gmail.Com EDUCATION Doctor of Philosophy in Molecular and Mechanistic Toxicology, 2013 University of Maryland, Baltimore, Maryland, USA

Role of Nrf2 Signaling in Therapeutic Resistance and Cancer Metastasis

Item Type dissertation

Authors Khatri, Raju

Publication Date 2013

Abstract The stress response transcription factor Nrf2 is a master regulator of cytoprotective genes. In non-stress situation, INrf2 (Keap1), constantly sequesters Nrf2 and facilitates its ubiquitination-mediated proteasomal degradation. Upon exposure to stre...

Keywords invasion; Nrf2-Keap1; Metastasis; Aromatase Inhibitors; Chemoprevention; Drug Resistance; Neoplasm Metastasis

Download date 10/10/2021 12:59:14

Link to Item http://hdl.handle.net/10713/3642 Raju Khatri Email:[email protected] EDUCATION Doctor of Philosophy in Molecular and Mechanistic Toxicology, 2013 University of Maryland, Baltimore, Maryland, USA

Master of Science in Medicinal Chemistry 2008 University of Maryland Baltimore County, Baltimore, Maryland, USA

Bachelor of Laws in Criminal Law 2004 Nepal Law Campus, TU, Kathmandu, Nepal

Master of Public Administration in Development Administration 2003 Tribhuvan University (TU), Kathmandu, Nepal

Master of Science in Organic Chemistry 1997 Tribhuvan University (TU), Kathmandu, Nepal

Bachelor of Science in Chemistry, Zoology and Botany 1995 Tri-Chandra Multiple Campus, TU, Kathmandu, Nepal

TRAINING Gas Chromatography for Identification of Narcotic Drugs 2004 United Nations, Office of Drugs and Crime Regional Precursor Control Project for SAARC Countries, Kathmandu, Nepal

Post Graduate Diploma in Forensic Chemistry and Toxicology 1999 National Institute of Criminology& Forensic Science, New Delhi, India

PUBLICATION Khatri R, Shah P, Rassool F, Tomkinson A, Brodie A, and Jaiswal AK. Inhibition of Nrf2 increased the sensitivity of aromatase inhibitors-resistant cells to chemotherapeutic drugs and decreased TIC cells subpopulation. Breast Cancer Research (Under revision).

Niture SK, Khatri R, Jaiswal AK. Regulation of Nrf2-An update, Review Article, Free Radical Biology & Medicine, 2013, Feb 19 [Epub ahead of print].

Khatri R. Importance of DNA Fingerprinting in Criminal Investigation, Special issue of Journal of Nepal police, 69-71, 2000

Khatri R, Jaiswal AK. Low level of Nrf2 promotes invasion and migration of cancer cells (in preparation) RESEARCH EXPERIENCE Graduate Research Assistant 2008-2013 Dept. of Pharmacology and Experimental Therapeutics School of Medicine, University of Maryland, Baltimore, Maryland

Investigated role of Nrf2 and its target genes in the aromatase inhibitor-resistant breast cancer cells to understand the mechanism of Nrf2-mediated resistance and to improve the therapeutic efficacy of aromatase inhibitors by developing inhibitors of Nrf2.

Explored the role of Nrf2 signaling in tumor metastasis by using stable cell lines with the altered levels of Nrf2 expression.

Graduate Teaching Assistant 2005-2008 Department of Chemistry and Biochemistry University of Maryland Baltimore County, Baltimore, Maryland

Involved in synthesis of unnatural nucleosides and novel carbocyclic nucleosides and purified organic compounds by extraction, fractional distillation, crystallization, sublimation and chromatographic techniques.

EMPLOYMENT EXPERIENCE Forensic Chemist 1998-2005 Narcotics /Chemistry & Toxicology Units Central Police Forensic Science Laboratory, Kathmandu, Nepal

Routinely carried out extraction, purification, detection of poisons, drugs and other chemicals from biological samples, and examined the adulteration of petroleum products.

MEMBER OF PROFESSIONAL ASSOCIATION Indian Academy of Forensic Sciences, India 1999 Indian Society of Analytical Scientists, India 1999 The Forensic Science Society, United Kingdom 1999 Nepal Forensics Society, Nepal 2004 American Association for Cancer Research, USA 2010 Society of Toxicology, USA 2012 ▪ National Capital Area Chapter 2012 ▪ Molecular Biology Specialty Section 2012 AWARD AND HONOR Graduate Student Travel Award, by Society of Toxicology Annual Meeting, San Antonio, TX (March10-14, 2013). Awarded $1,000.

Molecular Biology Specialty Section Graduate Student Research Award, Honorable Mention, by MBSS of Society of Toxicology Annual Meeting San Antonio, TX (March10-14, 2013).

Bern Schwetz Travel Award, first place, for „Outstanding Research, by National Capital Area Chapter of Society of Toxicology Annual Meeting, San Antonio, TX (March 10-14, 2013). (I had to decline it because I was also awarded another travel grant for the same conference).

Bern Schwetz Travel Award, Honorable Mention, for „Outstanding Research, by National Capital Area Chapter of Society of Toxicology Annual Meeting, San Francisco, CA (March 11-15, 2012).

Student Poster Competition Award, second place by National Capital Area Chapter of Society of Toxicology Spring Symposium, NIH, Bethesda, MD (March14, 2012).

CONFERENCE ABSTRACT AND PRESENTATION Khatri R, Shah P, Brodie A, and Jaiswal AK. Inhibition of Nrf2 signaling decreased drug resistance in aromatase inhibitor-resistant breast cancer cells. Annual Graduate Research Conference, University of Maryland, Baltimore, MD (April 11, 2013).

Khatri R and Jaiswal A. Lower expression of Nrf2 promotes proliferation, migration and invasion of prostate cancer cells. Annual meeting of Society of Toxicology, San Antonio, TX (March 10-14, 2013).

Khatri R, Brodie A, and Jaiswal A. Increase in Nrf2 contributes to reduced apoptosis and resistance to aromatase inhibitors. Annual meeting of Society of Toxicology, San Francisco, CA (March 11-15, 2012).

Khatri R, Niture SK and Jaiswal A. Activation of PKC-delta inhibits apoptosis leading to Mitomycin C resistance. Proceedings of the Annual Meeting of American Association for Cancer Research, Washington D.C. (April 18, 2010). ABSTRACT

Title of Dissertation: Role of Nrf2 Signaling in Therapeutic Resistance and Cancer Metastasis

Raju Khatri, Doctor of Philosophy, 2013

Dissertation directed by: Dr. Anil Jaiswal, Professor, Department of Pharmacology, School of Medicine

The stress response transcription factor Nrf2 is a master regulator of cytoprotective genes. In non-stress situation, INrf2 (Keap1), constantly sequesters

Nrf2 and facilitates its ubiquitination-mediated proteasomal degradation. Upon exposure to stress, Nrf2 stabilizes, translocates to nucleus and binds to antioxidant response elements (ARE) in the promoter of its target genes and activates transcription of many cellular defense genes, resulting in cell survival and cytoprotection. Consequently, upregulation of Nrf2 results in development of therapeutic resistance.

Aromatase inhibitors (AI) have been used for treatment of ERα positive breast cancer in post-menopausal women. However, persistent treatment with AI can lead to resistance. Little is known about the underlying mechanisms for the development of

AI-resistance. In this study, we examined the involvement of Nrf2 signaling in AI- resistant cells and found downregulation of INrf2 and upregulation of Nrf2 and its target genes such as anti-apoptotic proteins, antioxidants and drug efflux transporters.

Importantly, Nrf2- knock down AI-resistant cells showed fewer tumor-initiating cells

(TIC), formed fewer mammospheres and become more sensitive to doxorubicin and etoposide. A stable cell line containing a luciferase reporter under the control of an ARE-promoter was generated to identify Nrf2 inhibitors by high- throughput screening to develop an adjuvant therapy that contains an Nrf2 inhibitor to enhance the therapeutic efficacy of aromatase inhibitors.

We also investigated the role of Nrf2 in prostate cancer metastasis by generating stable cell lines expressing different levels of Nrf2. We investigated the role of Nrf2 signaling in colony formation, cell proliferation, migration and invasion.

Our findings suggested that loss of Nrf2 leads to increased anchorage independent cell growth, proliferation, migration and invasion.

In conclusion, persistent AI-treatment downregulated INrf2 leading to higher expression of Nrf2 and cytoprotective proteins that may, in part, result in increased

AI-resistance. As knocking down Nrf2 decreased TIC, the inhibition of Nrf2 signaling appears to be beneficial to reduce AI-resistance. However, lowering the levels of Nrf2 might program cancer cells towards invasion and migration. Thus, understanding the levels of Nrf2 in tumors would be important to determine whether inhibition or activation of Nrf2 is beneficial to a cancer therapy.

Role of Nrf2 Signaling in Therapeutic Resistance and Cancer Metastasis

Raju Khatri Program in Toxicology (Molecular and Mechanistic)

Dissertation submitted to the faculty of the Graduate School of the University of Maryland, Baltimore in partial fulfillment of the requirements for the degree of Doctor of Philosophy 2013 ©Copyright by Raju Khatri

ACKNOWLEDGEMENTS

It gives me a great pleasure to express my sincere gratitude to Prof. Jaiswal who had molded me into a cancer biologist from a chemist. He had trained me how to design a well-executed experiment and how to interpret and present data. His far sighted vision on a project is something I admire. I thank him for everything he has done for me.

I am indebted to my committee members Dr. Angela Brodie, Dr. Katherine Squibb, Dr. Amy Fulton, Dr. Srivastava, Dr. Arif Hussain and Dr. Vincent Njar for their constant encouragement and suggestions throughout the course of my study and I respectfully acknowledge their support. I also thank the University of Maryland, Baltimore for providing me the opportunity to come here and get trained in one of the best institutions of the country.

This study would not have been possible without the generosity of Dr. Brodie, Dr. Rassool, Dr. Srivastava, Dr. Njar, Dr. Qiu, Dr. Fulton, Dr. Hussain and their lab members, who provided us cell lines, antibodies and chemicals and I thank every one of them individually. I would also like to acknowledge Dr. Katherine Squibb for her guidance and suggestions.

I will not forget my colleagues in Dr. Jaiswal‟s laboratory. It was the cordial atmosphere I enjoyed there with the past and present members of the lab that transformed these inspirations into reality and I thank each one of them. I am extremely grateful to Dr. Emmanual Unni, Dr. Phillip Shelton and Andrew Kwegyir- Afful for their help.

Lastly but not the least, I thank my wife Yashu, son Rayan, my parents and in- laws for their support, understanding and patience during these years. The value they had given to education and to acquiring knowledge was truly the driving force to achieve my goal.

Table of Contents

CHAPTER Page ACKNOWLEDGEMENTS ...... i

TABLE OF CONTENTS ...... ii

LIST OF TABLES ...... vii

LIST OF FIGURES ...... viii

LIST OF ABBREVIATIONS ...... x

I. ROLE OF NRF2 SIGNALING IN CANCER THERAPEUTIC RESISTANCE ..1

Introduction ...... 1 Cancer Therapy ...... 1 Therapeutic Resistance ...... 2 Major Mechanisms of Therapeutic Resistance ...... 3 Two Schools of Thought on Drug Resistance ...... 4 Nuclear Factor Erythroid 2-Related Factor 2 (Nrf2) ...... 4 Regulation of Nrf2 Activity ...... 7 Activators and Activation of Nrf2 ...... 9 Nrf2 Activation and Drug Resistance ...... 10 Nrf2 Downstream Antioxidants and Phase II Detoxifying Enzymes ....11 Nrf2-regulated Drug-efflux Pumps and Anti-apoptotic Proteins...... 12 Nrf2, Cancer Stem Cells and Drug Resistance ...... 13

HYPOTHESES AND SPECIFIC AIMS ...... 15 Specific Aim 1 ...... 15 Specific Aim 2 ...... 15 Specific Aim 3 ...... 15

II.ROLE OF NRF2 SIGNALING IN AROMATASE INHIBITOR RESISTANCE 16

Abstract: ...... 16

Introduction: ...... 18 Aromatase Inhibitor (AI) and Resistance ...... 19

Materials and Methods: ...... 23

Chemicals and Reagents ...... 23 Cells and Cell Culture Conditions ...... 24 Generation of Stable LTLT-Ca Cells Expressing Nrf2shRNA ...... 25 Western Blotting ...... 25 Nrf2 Degradation Assay ...... 26 Immunoprecipitation and Ubiquitination Assay ...... 26 Cell Survival and Cell Death Assays ...... 27 ALDEFLUOR Staining ...... 27 Isolation of Tumor Initiating Cells (TIC) Using ALDEFLUOR Staining ...... 28 Mammosphere Assay ...... 28 Gene Expression Analysis ...... 29 Glutathione Quantification...... 29 Densitometry and Statistical Analyses...... 29

Results: ...... 30 Nrf2 was upregulated and INrf2 was downregulated in AI-resistant ... cells ...... 30 Anti-apoptotic proteins, drug transporters and glutathione levels were higher in AI-resistant cells ...... 32 Knocking down Nrf2 in LTLT-Ca cells decreased Nrf2 target gene expression and increased sensitivity to doxorubicin and etoposide...... 33 Nrf2 was less ubiquitinated and degraded slowly in LTLT-Ca cells.....34 Tumor initiating cells (TIC) expressed lower INrf2 and higher DNA repair proteins, HER2, Nrf2 and Nrf2 target gene GCLC ...... 36 shRNA-mediated inhibition of Nrf2 in LTLT-Ca cells significantly decreased TIC cells and mammosphere formation ...... 37

Discussion: ...... 39

III. THE MECHANISM OF INRF2 DOWNREGULATION IN AROMATASE INHIBITOR-RESISTANT CELLS ...... 44

Abstract: ...... 44

Introduction: ...... 46 INrf2 (Keap1)...... 46

iii

Regulation of INrf2 ...... 47 Epigenetic Regulation of INrf2 ...... 48 DNA Methylation ...... 48 Histone Modification ...... 49 Micro-RNA ...... 49 Autophagy ...... 49 Reactive Oxygen Species ...... 51

Materials and Methods: ...... 53 Chemicals and Reagents ...... 53 Cells and Cell Culture Conditions ...... 53 Western Blotting ...... 54 mRNA Stability Assay ...... 54 Gene Expression Analysis ...... 54 ROS Detection ...... 54 Densitometry and Statistical Analyses...... 55

Results: ...... 55 Inhibition of proteasome did not restore INrf2 ...... 55 Autophagy activity was high in LTLT-Ca cells ...... 57 Autophagy inhibitors did not increase INrf2 in LTLT-Ca cells ...... 58 Induction of autophagy decreased INrf2 protein ...... 60 Epigenetic modulators did not restore INrf2 in AI-resistant and .. sensitive cells ...... 61 HDAC inhibitors reduced INrf2 mRNA in AI-resistant and sensitive cells ...... 62 Withdrawal of letrozole increased mRNA of INrf2 ...... 64 mRNA of INrf2 was less stable in LTLT-Ca cells ...... 65

Discussion: ...... 66

IV. IDENTIFICATION OF COMPOUNDS INHIBITING NRF2 SIGNALING.....70

Abstract: ...... 70

Introduction: ...... 72

Materials and Methods: ...... 74

Chemicals and Reagents ...... 74 Cells and cell culture conditions ...... 74 Cloning of human NQO1-ARE in luciferase reporter vector ...... 75 Generation of stable MCF-7cells expressing INrf2shRNA and luciferase ...... 76 Western blotting ...... 76 Luciferase Activity...... 76 Densitometry and Statistical Analyses...... 77

Results: ...... 77 Natural products did not demonstrate the inhibitory effect on Nrf2 signaling in AI-resistant cells...... 77 Generation of cell-based luciferase system for identification of Nrf2 inhibitor ...... 79

Discussion: ...... 81

V. LOW LEVELS OF NRF2 PROMOTE MIGRATION AND INVASION OF PROSTATE CANCER CELLS ...... 85

Abstract: ...... 85

Introduction: ...... 87 Prostate Cancer and Metastasis ...... 87 Epithelial-Mesenchymal Transition (EMT) ...... 87 Matrix Metalloproteinase ...... 88 Cell adhesion molecules ...... 88 NF-κB and Cyclooxygenase-2 ...... 90 Nrf2 signaling in Metastasis ...... 91

Materials and Methods: ...... 93 Reagents and Chemicals ...... 93 Cell lines and generation of stable cells ...... 93 Cell culture conditions ...... 94 Western blotting ...... 94 Total RNA extraction, cDNA synthesis, and qRT-PCR ...... 95 Soft agar colony formation assay ...... 95 Boyden (BD) Migration and invasion Assays ...... 95 v

xCELLigence System ...... 96 Gelatin Zymography Assay ...... 97 Luciferase Activity...... 97 siRNA Transfection ...... 97 Densitometry and Statistical Analyses...... 98

Results: ...... 98 Highly metastatic cells had lower levels of Nrf2 and promoted colony formation ...... 98 Lowering Nrf2 level in less metastatic cells promoted anchorage- independent cell growth ...... 100 LNCaP cells with low level of Nrf2 proliferated faster, migrated and invaded more as compared to control cells ...... 101 Expressing higher levels of Nrf2 in highly metastatic C4-B2 cells lowered their ability to form colonies ...... 102 C4-B2 cells with higher levels of Nrf2 demonstrated higher proliferation, lower migration and invasion as compared to control cells ...... 103 Non-cancerous prostate epithelial cells (RWPE1) with low levels of Nrf2 did not form colonies in soft agar ...... 104 RWPE1-Nrf2KD cells showed a high degree of proliferation, migration and invasion...... 105 Low level of Nrf2 increased enzymatic activity of matrix metalloproteinase 9 ...... 106 Nrf2 mRNA in metastasized breast tumors ...... 108

Discussion: ...... 110

CONCLUSION ...... 115

Summary: ...... 115

Significance:...... 116

Future Direction: ...... 117

REFERENCES ...... 125

LIST OF TABLES

Table Page Table 6-4. Antibody dilution for western blotting ...... 122

vii

LIST OF FIGURES Page Figure 1-1. Schematic representation of different domains of Nrf2 ...... 5

Figure 1-2. Schematic diagram of INrf2-dependent regulation of Nrf2 ...... 6

Figure 1-3. Role and regulation of Nrf2 in chemoprevention and drug resistance .... 14

Figure 2-1. Chemical structure of aromatase inhibitors...... 19

Figure 2-2. Flow chart showing generation of aromatase inhibitor-resistant cells .... 21

Figure 2-3. Characterization of AI-sensitive and resistant breast cancer cells ...... 30

Figure 2-4. Levels of anti-apoptotic proteins, drug-efflux pumps and glutathione in AI-resistant and sensitive cells ...... 32

Figure 2-5. LTLTCa-Nrf2KD cells were more sensitive to doxorubicin and etoposide than control cells ...... 33

Figure 2-6. Nrf2 was less ubiquitinated and degraded slowly in LTLT-Ca cells ...... 35

Figure 2-7. Tumor initiating cell (TIC) cells expressed lower INrf2 and higher Nrf2, GCLC, DNA repair proteins and HER2 ...... 36

Figure 2-8. LTLT-Ca-Nrf2KDcells expressed lower levels of ALDH as compared to control LTLTCa-V cells ...... 38

Figure 2-9. Nrf2-Knocked down LTLT-Ca cells formed fewer mammospheres ...... 39

Figure 3-1. Schematic representation of various domains of INrf2 ...... 47

Figure 3-2. Cartoons showing different steps of autophagy ...... 50

Figure 3-3. Inhibition of proteasome did not restore INrf2 in LTLT-Ca cells ...... 56

Figure 3-4. Basal level of autophagy was high in LTLT-Ca cells ...... 57

Figure 3-5. Inhibition of Autophagy did not increase INrf2 ...... 59

Figure 3-6. Induction of autophagy decreased INrf2 protein ...... 60

Figure 3-7. Epigenetic modulators did not upregulate INrf2 ...... 62

Figure 3-8. HDAC inhibitors lowered INrf2 mRNA in AI-sensitive and resistant cells ...... 63

Figure 3-9. Withdrawal of letrozole increased transcription of INrf2 ...... 64

Figure 3-10. INrf2 mRNA was less stable in LTLT-Ca cells ...... 65

viii

Figure 4-1. Natural products did not demonstrate inhibitory effect on Nrf2 signaling in AI-resistant cells...... 78

Figure 4-2. Generation of a stable cell line expressing high levels of Nrf2 and luciferase ...... 80

Figure 4-3. Validation of luciferase expression in stable cells (MCF-7+INrf2KD+ Luc2) line expressing high levels of Nrf2 and luciferase ...... 81

Figure 5-1. Flow chart showing generation of bone metastasized C4-B2 cells...... 98

Figure 5-2. C4-B2 cells formed more colonies in soft agar ...... 99

Figure 5-3. LNCaP cells with less Nrf2 formed more colonies in soft agar ...... 100

Figure 5-4. Nrf2-knocked down LNCaP cells showed more colonies in soft agar ...101

Figure 5-5. LNCaP+Nrf2KD cells proliferated faster, migrated and invaded more as compared to control cells ...... 102

Figure 5-6. C4-B2 cells with higher levels of Nrf2 showed fewer colonies in soft agar ...... 103

Figure 5-7. C4-B2-INrf2-KD cells proliferated faster but showed less degree of migration and invasion ...... 104

Figure 5-8. Prostate epithelial cells (RWPE1) with low levels of Nrf2 did not form colonies in soft agar ...... 105

Figure 5-9. RWPE1+Nrf2KD cells proliferated faster, migrated and invaded more as compared to control cells ...... 106

Figure 5-10. Low levels of Nrf2 resulted in higher expression of Cox-2 and higher enzymatic activity of MMP9 ...... 107

Figure 5-11. Nrf2 mRNA in metastasized human breast tumors (Met-samples) as compared with that in organoid (Org-sample) ...... 109

Figure 5-12. Flow chart showing how low level of Nrf2 facilitates the invasion of tumor cells ...... 112

Figure 6-1. Schematic representation of development of drug resistance and progression of metastasis...... 116

Figure 6-2. Downregulation of Bcl-2 in AI-resistant cells ...... 119

Figure 6-3. Luciferase expression in prostate cancer cells with altered levels of Nrf2 ...... 121

LIST OF ABBREVIATIONS

AC1 MCF-7 cells transfected with human aromatase

AI Aromatase Inhibitor

AnaR Anastrozole-resistant AC1 cell

ARE Antioxidant response element

ATF3 Activating transcription factor 3

ATRA All trans-retinoic acid

Bach1 BTB and CNC homology 1

BCRP Breast cancer resistance protein

BODIPY Boron-dipyrromethane

CBT Center for Biomolecular Therapeutics

CHX Cycloheximide

CSC Cancer stem cell

DMSO Dimethyl sulfoxide

ExemR Exemestane resistant AC1 cell

FACS Fluoroscence-activated cell sorting

GCLC γ-glutamyl cysteine ligase catalytic unit

HER2 Human epidermal growth factor receptor 2

HO-1 Heme Oxygenase-1

INrf2 Inhibitor of Nrf2 (Keap1)

LC3B Microtubule-associated protein 1 light chain 3 beta (LC3)

LDH Lactate dehydrogenase

LTLT-Ca Long term letrozole treated letrozole-resistant MCF-7Ca cell

MCF-7Ca MCF-7 cells stably expressing human aromatase

MRP1 Multidrug resistance-associated protein 1

NF-κB Nuclear factor-κappa B

NIH National Institute of Health

NQO1 NAD (P) H quinone oxidoreductase1

Nrf2 Nuclear factor-erythroid 2 related factor 2

PI3K Phosphoinositide-3-kinase

PPAR γ Peroxisome proliferator-activated receptor γ

RAR α Retinoic acid receptor α

RIPA Radioimmunoprecipitation Assay

RON Recepteur d‟ origine nantais

ROS Reactive oxygen species

SAM S-adenosyl methionine

SCF Skp1-cul1-F-box shRNA short hairpin RNA t-BHQ tert-Butylhydroquinone

TIC Tumor initiating cell

UTR Untranslated region

WT Wild type

XRE Xenobiotic response element

β-TrCP β-transducin repeat-containing protein

CHAPTER ONE

Background and Significance

ROLE OF NRF2 SIGNALING IN CANCER THERAPEUTIC

RESISTANCE

Introduction

Cancer Therapy

Advancement in early diagnostic techniques and improvement in treatment regimens have reduced cancer related death by 20 % in the last two decades (Siegel et al., 2013). Based on stage and location of cancer initiation, the choice of therapy is surgery, radiation, chemo (chemical), hormonal and targeted therapy. Surgery is the primary approach for cancer treatment, since it removes visible and localized tumors.

Radiation therapy destroys cancer cells by generating free radicals within the cells that damage DNA. Radiation therapy can be used alone or in combination with other therapies (Calais, 1997). Chemotherapy is a systemic therapy and is one of the main therapeutic approaches for all types of cancer treatment. Modes of action of chemotherapy are based on cytotoxicity of anticancer agents that kill proliferating cancer cells as well as normal cells (Pavet et al., 2011). Targeted treatment, the latest therapeutic option introduced in clinical use, utilizes drugs that prevent cancer cells from proliferation and metastasis by interfering with specifically targeted molecules involved in tumor growth and progression (Raguz and Yague, 2008). For example, hormone therapy uses substances that block the synthesis or function of hormones like estrogen and androgen that stimulate the growth of hormone-dependent tumor cells.

Therapeutic Resistance

Therapeutic resistance can be categorized into intrinsic and acquired resistance. When tumor cells show insignificant or no response to the treatment upon first exposure is termed intrinsic resistance. Cancer cells originating from organs or tissues expressing high levels of drug detoxifying enzymes and cytoprotective molecules encounter intrinsic resistance. For example, melonoma and hepatoma cells are intrinsically resistant to chemotherapy (Liu, 2009). Whereas in acquired resistance, tumor cells that initially respond to an anti-cancer agent are no longer sensitive to the same agent (Goldie, 2001). Acquired resistance results from an adaptive process of neutralizing the mechanism of action of therapy that develops over time of treatment can be induced by the therapy through genetic or epigenetic alterations. Emergence of resistance in breast cancer treated with letrozole is an example of acquired drug resistance (Sabnis et al., 2008).

Resistance is a natural cellular self-defense process to protect even normal cells from any types of stress or xenobiotic. Previous studies revealed that therapeutic resistance is not limited to radiation and traditional chemotherapy but also includes hormonal and targeted therapy (Chumsri et al., 2011; Raguz and Yague, 2008).

Tumor cells can also develop resistance to a particular drug or multiple drugs that are structurally and mechanistically unrelated (Gillet and Gottesman, 2010). Therapeutic resistance is a continuous and major impediment to successful cancer treatment.

Understanding the molecular mechanisms of therapeutic resistance and developing new therapeutic agents to target those mechanisms allow us to design and develop new therapies.

Major Mechanisms of Therapeutic Resistance

A growing body of evidence demonstrates that several mechanisms such as gene amplification and upregulation of target molecules, reduction in drug uptake, alteration of the targeted pathways are responsible for drug resistance (Li et al., 2009).

Resistance to the drugs that target DNA damage arises from activation of DNA repair enzymes (Tobin et al., 2012). Reduction in the programmed cell death pathway is another mechanism of resistance due to upregulation of anti-apoptotic proteins (Gillet and Gottesman, 2010). Recently, autophagy and Tumor-Initiating Cells (TIC) have also been implicated in drug resistance (Chen et al., 2011; Chen and Debnath, 2010;

Gilani et al., 2012). Similarly, antioxidants mediate resistance to anti-cancer agents that generate ROS (reactive oxygen species) as a mode of action (Shim et al., 2009).

Drug inactivation or detoxification by activation of drug metabolizing systems (GST, drug transporters) can also lead to resistance (Raguz and Yague, 2008). Several studies have demonstrated that exposure of cancer cells to one drug often results in multidrug resistance (MDR), by which cancer cells are resistant to many structurally and functionally unrelated drugs (Gottesman, 2002). Drug efflux, mostly associated with MDR, is mediated by several members of the ATP-binding cassette (ABC) superfamily of membrane bound transporters which are subdivided into seven families designated ABC-A through ABC-G based on sequence similarities. Forty nine ABC transporters have been identified in humans, sixteen of them are implicated in cancer drug resistance (Gillet and Gottesman, 2010; Gottesman, 2002). The increased expression of drug-efflux pumps MDR1, MRP, BCRP is mainly associated with the acquisition of multidrug resistance in chemotherapeutic agent-induced drug resistance (Pawlowski et al., 2013; Singh et al., 2010).

Two Schools of Thought on Drug Resistance (Classical and Cancer Stem Cell)

Two models of drug resistance have been proposed to explain the emergence of drug resistance; clonal expansion and cancer stem cells (Abdullah and Chow,

2013). According to classical theory of drug resistance (clonal expansion or stochastic) model, cancer cells acquire resistance following the treatment due to genetic changes and selection of the drug resistant cells. On the other hand, a population of undifferentiated tumorigenic cells that can divide asymmetrically and produce identical daughter cells and more differentiated cells are termed cancer stem cells. In this second model, tumor cells are a heterogeneous mass of proliferating cells with different genetic variations originated from cancer stem cells. This small population of cancer stem cells within tumors is inherently resistant to therapy. Over the time of treatment, only the cancer stem cells survive, proliferate and repopulate the tumor with stem cells (Nicolini et al., 2011).

Nuclear Factor Erythroid 2-Related Factor 2 (Nrf2)

The stress response molecule, nuclear factor erythroid 2-related factor 2 (Nrf2) is a master regulator of more than 200 cytoprotective genes (Niture et al., 2013). Nrf2, belongs to the cap-n-collar (CNC) subfamily of basic leucine zipper (b-Zip) transcription factor and was first identified and characterized in 1994 (Moi et al.,

1994). As shown in Figure 1-1, human Nrf2 with 605 amino acids has six highly conserved Nrf2-ECH homology (Neh) domains. The N-terminal hydrophobic region of Nrf2 contains the INrf2 (Keap1) binding Neh2 domain with seven lysine residues that are targets for Keap1–mediated ubiquitination (Itoh et al., 1999a; Zhang et al.,

2004). Neh4 and Neh5 domains are necessary for transcriptional activation. The nuclear import signal is located in the Neh1 domain of Nrf2. The Neh1 domain also

4 contains CNC-bZip domains that are essential for heterodimerization with a small musculo-aponeurotic fibrosarcoma (Maf) protein and DNA binding (Itoh et al., 1997).

The Neh3 domain at the C-terminal of Nrf2 also showed transactivation activity and

Neh6 connects Neh5 and Neh1 (Li and Kong, 2009).

Figure 1-1. Schematic representation of different domains of Nrf2.

Nrf2 is expressed at higher levels in organs like liver and kidney that are involved in detoxification and also in skin, lung and GI tract that are exposed to external environment (Copple et al., 2008). Under basal conditions, the inhibitor of

Nrf2 (INrf2), a substrate adaptor protein for ubiquitinizing Cul3-Rbx1 complex, constitutively sequesters Nrf2 and facilitates the ubiquitination-mediated proteasomal degradation of Nrf2 (Dhakshinamoorthy and Jaiswal, 2001; Itoh et al., 1999b). A controlled level of Nrf2 is available in the nucleus to maintain cellular homeostasis.

Upon exposure to stress (oxidative, electrophilic, UV irradiation), the oxidation of thiol groups in Nrf2 and INrf2 triggers the stabilization of Nrf2 (He and Ma, 2009).

Consequently, Nrf2 translocates into the nucleus, heterodimerizes with small Maf proteins and binds to a core nucleotide sequence known as an electrophilic/ antioxidant response element (Ep/ARE) in the promoter region of its target genes and activates the transcription of cellular defense genes (Favreau and Pickett, 1991;

Friling et al., 1990; Venugopal and Jaiswal, 1996).

There is plentiful information to prove that a wide variety of cytoprotective genes are under the control of a master regulator, Nrf2 including aldo-keto-reductase, antioxidants and related molecules [superoxide dismutase, catalase, metal binding protein metallothionein and ferritin, glutathione, thioredoxins, γ-glutamylcysteine ligase/synthetase (γ-GCL, which has a γ-glutamylcysteine ligase catalytic subunit,

GCLC and γ-glutamyl-cysteine ligase modifier subunit, GCLM), glutathione peroxidase, glutathione reductase, NAD(P)H quinone oxidoreductases (NQO1 and

NQO2), oxidative response enzyme heme oxygenase-1 (HO-1)], activity and subunits of proteasomes, anti-apoptotic proteins (Bcl-2 and Bcl-xL), detoxifying enzymes and drug transporters.

Figure 1-2. Schematic diagram of INrf2 (Keap1) dependent regulation of Nrf2. (A) Under basal conditions, INrf2, an adaptor molecule for a Cul3/Rbx1-ubiquitin ligase complex, facilitates ubiquitination of Nrf2 and Nrf2 gets degraded. A small amount of Nrf2 translocates to the nucleus to maintain normal cellular functions. (B) In the presence of stress, cysteine residues in INrf2 get modified to cause conformational changes in INrf2, subsequently; newly synthesized Nrf2 accumulates in the nucleus and transactivates its target genes.

Indeed, in vivo evidence has demonstrated the importance of Nrf2 in protecting cells from the toxic and carcinogenic effects of many environmental insults. The Nrf2-knockout mouse was susceptible to acetaminophen, ovalbumin, cigarette smoke and pentachlorophenol-induced acute damage and had increased tumor formation when exposed to carcinogens such as benzo[a]pyrene, diesel exhaust

6 and N-nitroso-butyl (4-hydroxybutyl) amine (Aoki et al., 2001; Chan et al., 2001;

Enomoto et al., 2001; Hu et al., 2006b; Iida et al., 2004; Iizuka et al., 2005; Ramos-

Gomez et al., 2001). A large body of evidence has accumulated to support the fact that Nrf2 coordinately activates cytoprotective proteins including anti-inflammatory proteins, detoxifying enzymes, drug transporters, antioxidants, reduces electrophiles and ROS, decreases genomic instability and mutations, protects against many human diseases and prevents cancer metastasis (Cho et al., 2004; Rachakonda et al., 2010;

Satoh et al., 2010; Shih et al., 2005; Shu et al., 2010). Therefore, Nrf2 plays a significant role in cytoprotection and cell survival, chemoprevention and tumor suppression. However, upregulation of Nrf2 activates many of the molecules and enzymes that protect tumor cells from therapy and promotes cell proliferation and survival, supporting a role of Nrf2 in drug resistance and oncogenesis. Because of the dual role of Nrf2 in cancer prevention and cancer progression, Nrf2 is known as both a tumor suppressor as well as a tumor promoter (Shelton and Jaiswal, 2012).

Regulation of Nrf2 Activity

Multiple levels of Nrf2 regulation have been reported in the literature. As Nrf2 binds to antioxidant response element (ARE) sequences, aryl hydrocarbon receptor

(AhR) binds to xenobiotic response elements (XRE) sequences to activate their target genes. Another study showed that the transcription of Nrf2 was increased by treatment with Nrf2 activator (3H-1, 2-dithiole-3-thione) and xenobiotic (2, 3, 7, 8- tetrachlorodibenzo-p-dioxin), suggesting that the Nrf2 promoter has unique sequences of XREs and ARE (Kwak et al., 2002; Miao et al., 2005). Thus, the AhR and Nrf2 are known to regulate Nrf2 transcriptionally. In addition, oncogenes K-Ras, B-Raf and

Myc all increase Nrf2 transcription and increase basal Nrf2 expression (DeNicola et

7 al., 2011). More recently, identification of micro-RNAs that target Nrf2 mRNA has added an extra layer of regulation in Nrf2 signaling. miR-28, miR-34 and miR-144 all have been reported to regulate Nrf2 by targeting its mRNA for degradation (Bryan et al., 2012).

Furthermore, posttranslational modifications such as acetylation, ubiquitination and phosphorylation of Nrf2 were also shown to have a direct effect on

Nrf2 signaling. Acetylation of Nrf2 increased its nuclear localization and transcriptional activity (Kawai et al., 2011). Two opposite effects were identified when Nrf2 is phosphorylated. Phosphorylation of Nrf2 at serine 40 by PKC-δ results in stabilization of Nrf2 whereas phosphorylation of Nrf2 by Fyn and GSK3 β leads to degradation of Nrf2 (Jain and Jaiswal, 2007; Niture et al., 2009; Rada et al., 2011).

Skp1-cul1-F-box (SCF)/β-transducin repeat-containing protein (β-TrCP) recognize the GSK3 β-mediated phosphorylated Nrf2 that facilitates the ubiquitination- dependent degradation of Nrf2 (Rada et al., 2012). INrf2 (Keap1) is known as a negative regulator of Nrf2. INrf2-mediated proteasomal degradation of Nrf2 is a well- accepted mechanism for the posttranslational regulation of Nrf2 (Dinkova-Kostova et al., 2005a; Nguyen et al., 2004). It is noteworthy that SCF/ β-TrCP-mediated degradation of Nrf2 is INrf2-independent and redox-insensitive degradation pathway.

Therefore, there are two possible independent mechanisms that degrade Nrf2 under basal conditions.

There are also reports showing suppression of Nrf2 signaling by some negative regulators like Bach1, ERα, RARα ATF3, NF-κB, p53 and peroxisome proliferator-activated receptor γ (PPARγ). These endogenous molecules are shown to

8 antagonize the function of Nrf2 through various mechanisms (Dhakshinamoorthy et al., 2005; Ma and He, 2012).

Activators and Activation of Nrf2

Molecules that can stabilize Nrf2 and activate Nrf2 downstream gene transcription are known as Nrf2 activators. Many natural and some synthetic compounds are identified as Nrf2 activators. Curcumin, t-butyl hydroquinone (t-

BHQ), sulforaphane, cinnamaldehyde, resveratrol, and heavy metals including arsenic, cadmium and chromium are examples of Nrf2 activators (Lau et al., 2008;

Ma and He, 2012; Wondrak et al., 2010). However, all Nrf2 inducers do not activate the Nrf2 pathway through the same mechanism (Wang et al., 2008). Different models have been proposed to account for Nrf2 nuclear accumulation and Nrf2-dependent genes induction. Oxidation of INrf2, a redox sensor molecule, is the well-studied mechanism for Nrf2 activation. INrf2 is rich in cysteine (27 in human and 25 in mouse) residues (Lau et al., 2008). Cysteine residues located adjacent to basic amino acids are reactive with low pKa values that are excellent targets of oxidants and electrophiles (Ma, 2013). Oxidation of cysteine residues of INrf2 in response to stress or Nrf2 activators causes a conformational change of INrf2 that renders Nrf2 inaccessible by the ubiquitinizing complex and, thereby Nrf2 gets stabilized and translocates to the nucleus. Accordingly, cysteine modification of Nrf2 has also been reported to be involved in translocation of Nrf2 to the nucleus and activation of its downstream genes (He and Ma, 2009).

In addition to activation of Nrf2 by its activators through the INrf2-dependent pathway, some non-canonical pathways of Nrf2 activation have been reported.

Recently, a few molecules that compete with Nrf2 to interact with INrf2 leading to

Nrf2 activation have been identified. Dipeptidyl peptidase 3 (DPP3) activates Nrf2 by competing Nrf2 for interaction with INrf2 (Hast et al., 2013). Autophagy cargo receptor p62 protein sequesters INrf2 into aggregates resulting in stabilization and activation of Nrf2 (Inami et al., 2011; Komatsu et al., 2010). Similarly, sestrins, tumor suppressor p21 and BRCA2 (PALB2) are also known to stabilize Nrf2 by preventing

INrf2 from interacting with Nrf2 (Bae et al., 2013; Chen et al., 2009; DeNicola et al.,

2011). It is notable that activation of Nrf2 by these endogenous molecules is independent of cellular oxidative stress. Post-translational modification of INrf2 such as succination and nitration also results in activation of Nrf2 (Kinch et al., 2011).

Moreover, loss of function mutation of INrf2 in many cancer cells results in nuclear accumulation of Nrf2 (Padmanabhan et al., 2006; Singh et al., 2006).

Similarly, gain of function mutations in Nrf2 also leads to disruption of the interaction between Nrf2 and INrf2, and nuclear accumulation of Nrf2 (Shibata et al., 2008).

Mutations either in INrf2 and Nrf2 that disrupt their interaction have been reported in many cancers, including prostate and breast cancer, and results in increased nuclear accumulation of Nrf2 and persistent activation of cytoprotective proteins (Nioi and

Nguyen, 2007; Shibata et al., 2008). This increased activation of Nrf2 is thought to provide an advantage to cancer cells with increased cell growth, proliferation and cytoprotective gene expression. Downregulation of INrf2 is one of the mechanisms of

Nrf2 activation that was studied in detail in Chapter three.

Nrf2 Activation and Drug Resistance

Activation of Nrf2 protects normal cells from any environmental insults whereas upregulation of Nrf2 in cancer cells is associated with cell survival and

10 cytoprotection against stress and therapy. Activation of Nrf2 and its involvement in cancer cell survival has been confirmed by independent research groups (Hayes and

McMahon, 2009; Shibata et al., 2008; Singh et al., 2006). The silencing of Nrf2 using

Nrf2-siRNA or by over expression of INrf2 in lung carcinoma, breast adenocarcinoma and neuroblastoma rendered cancer cells more susceptible to chemotherapeutic drugs like cisplatin, etoposide and doxorubicin. Previous studies have provided clear evidence that constitutive activation of Nrf2 causes drug resistance due to overexpression of efflux pumps and phase II detoxification enzymes (Dinkova-

Kostova et al., 2005a; Lau et al., 2008). However, multiple molecular mechanisms can be involved in drug resistance. The persistent activation of Nrf2 due to mutations or deregulation of factors controlling Nrf2 also increases nuclear accumulation of

Nrf2 (Okawa et al., 2006; Vollrath et al., 2006), which lead to higher levels of cytoprotective proteins including detoxifying/biotransformation enzymes, drug transporters, antioxidants and anti-apoptotic proteins. This results in decreased apoptosis, increased cell survival and drug resistance. Therefore, Nrf2 has been implicated in cancer cell survival and drug resistance, making it an attractive target to improve the efficacy of chemotherapy.

Nrf2 Downstream Antioxidants and Phase II Detoxifying Enzymes

Gamma-glutamylcysteine synthetase (γ-GCS) catalyzes the rate limiting step in the biosynthesis of glutathione. The γ-GCS consists of GCLC and GCLM subunits, which are regulated by Nrf2 (Mulcahy et al., 1997). Glutathione, a cellular antioxidant and free radical scavenger, maintains intercellular redox homeostasis and detoxifies xenobiotics through enzymatic conjugation by Nrf2 target gene, (GST) glutathione-S-transferase (Yates et al., 2009). Glutathione-related gene (glutathione

11 reductase), glutathione-dependent enzyme (glutathione peroxidase), downstream genes of Nrf2 are key molecules in protection against oxidative stress (Hu et al.,

2006a). Glutathione peroxidase metabolizes H2O2, generating oxidized GSH (GSSG), and glutathione reductase regenerates reduced GSH. Nrf2 also activates the expression of thioredoxin and thioredoxin reductase, which also maintain cellular redox status during oxidative stress (Hu et al., 2006a).

A prototypical Nrf2 target gene NQO1 [NAD(P)H dehydrogenase, quinone 1 or NAD(P)H quinone oxidoreductase 1] catalyzes the two-electron reduction of quinone to hydroquinones, which are readily detoxified and exported out of cells after conjugation with endogenous polar molecules. Therefore, NQO1 protects cells from highly reactive and potentially damaging quinonesre (Klaassen and Reisman, 2010;

Siegel et al., 2012). Recently, NQO1 has been reported to have other novel biological functions including scavenging of superoxide anion radicals and stabilization of tumor suppressor p53 (Zhu and Li, 2012). Another Nrf2 target gene HO-1 (Heme oxygenase-1) catalyzes the first and rate-limiting step in heme catabolism (Choi and

Alam, 1996). The cytoprotective effect of HO-1 is exerted through degradation of pro-oxidant free heme into carbon monoxide, free iron and biliverdin, which can be further reduced to antioxidant bilirubin (Gozzelino et al., 2010).

Nrf2-regulated Drug-efflux Pumps and Anti-apoptotic Proteins

Antioxidant-induced Nrf2 was demonstrated to activate expression of anti- apoptotic proteins Bcl-2 and Bcl-xL that collectively help to develop resistance to apoptosis (Niture et al., 2013). Moreover, Nrf2 inducers were shown to increase the expression of MRPs (multidrug resistance-associated protein 1, 2, 3, 4 and 6), which

12 export a variety of substrates conjugated with glutathione, sulfate and glucoronide

(Aleksunes et al., 2008; Maher et al., 2005; Vollrath et al., 2006). Another Nrf2 target gene, BCRP (breast cancer resistance protein) is a drug-transporter, which exports cisplatin and cyclophosphamide (Pawlowski et al., 2013; Singh et al., 2010).

Nrf2, Cancer Stem Cells and Drug Resistance

Cancer stem cells, also named as side population (SP) or cancer/tumor- initiating cells (TIC) have recently been implicated in therapy resistance. Recent studies have implicated mammary tumor-initiating cells (TIC) in chemotherapy and radiation therapy resistance (Chumsri and Burger, 2008; Sabnis et al., 2010). TIC cells have a lowered ability to undergo apoptosis and a higher ability for DNA repair, making them more resistant to cancer therapy (Ma et al., 2008; Nicolini et al., 2011).

High aldehyde dehydrogenase (ALDH) activity and BCRP (breast cancer resistance protein) are found in chemoresistant cancer stem cells (Awad et al., 2010). ALDH is considered to be a marker of stem cells (Ginestier et al., 2007). One of the characteristics of TIC cells is their ability to form non-adherent mammospheres.

These are floating spherical colonies grown in serum free medium (Saadin et al.,

2013).

Nrf2 has been implicated with chemoprevention and drug resistance in cancer therapy. However, recent studies have revealed a role of Nrf2 in maintenance of cancer stem (CS) cells (Achuthan et al., 2011). In addition to other self-protection mechanisms, CS cells express high levels of Nrf2 and ABC drug-transporters (Singh et al., 2010). Expression of many membrane transporters in cancer stem cells establishes a correlation between cancer stem cells and drug resistance (Alison et al.,

2012; Raguz and Yague, 2008). Since CS cells express high levels of Nrf2 that might contribute to the development of therapeutic resistance, indicating a positive correlation between the expression of Nrf2 and the TIC population. Thus, Nrf2 inhibitors should be effective in eliminating both CS cells and other differentiated cancer cells by making those cells more sensitive to therapeutic agents.

Figure 1-3. Role and regulation of Nrf2 in chemoprevention and drug resistance. Endogenous inhibitors of Nrf2 (INrf2 and Bach1) repress Nrf2 signaling. Endogenous metabolic pathway or metabolites of anti-cancer drugs or oxidant or Nrf2 activators modify cysteine residues of INrf2 that stabilizes Nrf2 and activates Nrf2-mediated cytoprotective genes. Mutations either in INrf2 or Nrf2 that disrupt their interaction results in activation of Nrf2 function. Expression of endogenous activators of Nrf2 also stabilizes Nrf2. Activation of Nrf2 protects normal cells from harmful chemicals. Persistent activation of the Nrf2 pathway provides advantage to cancer cells for cytoprotection against anti-cancer agents leading to drug resistance.

In conclusion, Nrf2 expression under proper regulation by positive and negative factors is essential for normal cell growth and survival, and also protection against chemical and radiation stress. However, deregulation of Nrf2, either due to gain of function mutations in Nrf2 or loss of function mutations in regulatory factors, results in oncogenesis. Importantly, persistent activation of Nrf2 is detrimental to chemotherapy (Fig.1-3).

HYPOTHESES AND SPECIFIC AIMS

Increased Nrf2 signaling has been reported in various cancers. As a known regulator of cytoprotective genes, Nrf2 controls the expression of anti-apoptotic proteins, antioxidants and drug transporters. Upregulation of these molecules is known to increase resistance to apoptosis and anti-cancer agents. Therefore, we hypothesized that prolonged treatment of breast cancer cells with aromatase inhibitors

(AI) would result in AI-resistance and upregulation of Nrf2 and its target genes leading to resistance to chemotherapy. Cancer stem cells also express high levels of

Nrf2 and its target genes rendering them drug resistant. These observations had directed us to hypothesize that Nrf2 plays a role in maintaining tumor initiating cells

(TIC) and by creating a drug resistant environment, it supports tumor growth. INrf2, a repressor of Nrf2 normally keeps the level of Nrf2 under control. INrf2 is reported to be degraded by autophagy. So, it is reasonable to speculate that there could be an activation of Nrf2 in AI-resistant cells due to autophagy. On the contrary, detection of lower levels of Nrf2 in prostate cancer and anti-metastasis properties of Nrf2 has recently been reported. Therefore, we also hypothesized that lowering Nrf2 levels can initiate cancer metastasis.

These hypotheses were examined by addressing the following specific aims.

Specific Aim 1 To investigate the role of Nrf2 in the development of aromatase inhibitor-resistance.

Specific Aim 2 a) To determine the mechanism of INrf2 downregulation in AI-resistant cells. b) To identify potential Nrf2 inhibitors to combat AI-resistance.

Specific Aim 3 To examine the role of Nrf2 in cancer metastasis.

CHAPTER TWO

ROLE OF NRF2 SIGNALING IN AROMATASE INHIBITOR RESISTANCE 1

Abstract:

Aromatase inhibitors (AIs) are widely used for the treatment of ER positive breast cancers in post-menopausal women. However, persistent treatment with AI can lead to resistance. Nrf2-INrf2 (Keap-1) complex function as a sensor of drugs or radiation induced oxidative /electrophilic stress. In the absence of stress, INrf2 mediates the degradation of Nrf2 by ubiquitination through Cul3/Rbx1 complex.

Upon exposure to stress, Nrf2 stabilizes and translocates to the nucleus and coordinately induce more than 200 cytoprotective gene expressions. In this study we investigated the role of Nrf2 in AI-resistance in breast cancer cells.

Protein levels of INrf2, Nrf2 and its downstream genes were compared by immunoblots in AI-sensitive and resistant cells. Simultaneously mRNA levels of

INrf2 and Nrf2 were also determined in these cells. Cell survival assays were then carried out to determine the sensitivity of these cells to doxorubicin. Etoposide- induced apoptosis was measured by DNA fragmentation assay. Nrf2 was knocked down in AI-resistant cells by shRNA to evaluate the role of Nrf2-mediated cytoprotective genes in drug induced cell death, number of tumor initiating cells

(TIC) and mammosphere formation.

1 Some portion of this chapter has been communicated for publication in Breast Cancer Research.

The results of our study indicate that AI-resistant cells express lower levels of

INrf2 and higher levels of Nrf2, as compared to AI-sensitive cells. The increase in

Nrf2 was due to lower ubiquitination and a slower rate of Nrf2 degradation. Further studies revealed that high amounts of Nrf2 resulted in increased expression of antioxidant enzymes, drug-transporters, DNA repair and anti-apoptotic proteins, contributing resistance to doxorubicin and etoposide-induced apoptosis and cell death.

Knocking down Nrf2 in letrozole (AI) resistant LTLT-Ca cells reduced the resistance to doxorubicin and etoposide and sensitized the cells to these drugs. Interestingly,

LTLTCa-Nrf2KD cells also showed reduced levels of ALDH (aldehyde dehydrogenase), a marker of tumor initiating cells and significantly decreased mammosphere formation as compared to the control LTLTCa-V cells expressing higher levels of Nrf2.

In conclusion, the results together suggested that persistent AI-treatment down regulated INrf2 resulting in activation of Nrf2 signaling that led to increased number of tumor initiating cells. The identification of Nrf2 as a cause of drug resistance makes it a promising target for the development of Nrf2 inhibitors to overcome AI- resistance.

Introduction:

Breast cancer is the most common cancer among women (Siegel et al., 2013).

Estrogens contribute to growth and proliferation of ER positive breast cancer cells.

Knowing the fact that estrogen stimulates growth of breast tumor cells through estrogen receptor, anti-estrogen, tamoxifen was developed as a treatment for breast cancer. Although tamoxifen therapy has reduced a significant number of deaths from breast cancer, resistance to this treatment eventually develops (Raguz and Yague,

2008). Subsequently, inhibitors of estrogen synthesis were developed to target the aromatase, an enzyme that catalyzes the aromatization of androgen to yield estrogen

(Thompson and Siiteri, 1974). Since tamoxifen is also a weak or partial estrogen, the inhibition of estrogen was more effective than treatment with tamoxifen. Aromatase, a product of cytochrome P450 CYP19A1, is mainly expressed in the ovaries in premenopausal women under regulation by the gonadotropins, follicle-stimulating hormone (FSH) and luteinizing hormone (LH), which are under control of the pituitary-hypothalamus axis (Chen et al., 1988). Since estrogen has an inhibitory feedback effect on FSH and LH expression in premenopausal women, lowering the levels of estrogen by aromatase inhibitors activates FSH and LH expression. In contrast, in postmenopausal women, regulation of aromatase in the adipose tissue is not under gonadotropin control, but under regulation of glucocorticoids, whereas, in cancer cells, cAMP controls the expression of aromatase (Wong and Chen, 2012).

The expression of aromatase is reported to be higher in breast cancer cells and adipose stromal cells compared to normal cells (Chen et al., 2006). For this reason, breast cancer cells have higher levels of estrogens than non-cancerous cells. Therefore, inhibition of aromatase is very specific and effective in breast cancer treatment in post-menopausal women.

Aromatase Inhibitor (AI) and Resistance

Aromatase inhibitors (AIs) are effective first line treatment for ERα positive breast cancer that constitutes three-fourths of all types of breast cancers (Chumsri et al., 2011; Miller, 2010). Currently, letrozole (Femara), anastrozole (Arimidex) and exemestane (Aromasin) are the third-generation of aromatase inhibitors approved by the FDA for the treatment of ERα positive breast cancer in postmenopausal women in both the adjuvant and metastatic setting. Letrozole and anastrozole are non-steroidal and reversible inhibitors of aromatase whereas exemestane is steroidal but irreversible inhibitor of aromatase (Chen et al., 2006). Letrozole is more potent than other AIs in reducing plasma levels of estrogen in patients (Geisler et al., 2002).

Figure 2-1.Chemical structure of aromatase inhibitors. Letrozole and anastrozole are non-steroidal reversible inhibitors of aromatase. Eexemestane is a steroidal irreversible inhibitor of aromatase

Despite the fact that AI treatment has demonstrated superior therapeutic efficacy over tamoxifen in the clinic, not all ER positive breast cancers respond to

AI treatment (Chen et al., 2006). The treatment benefit of aromatase inhibitor is often limited by resistance. The emergence of resistance to AI occurs in a significant number of patients in the adjuvant setting and in the metastatic setting. Subsequent studies have demonstrated that gene amplification of HER2 [human epidermal growth factor receptor 2 (ErbB-2/neu)] caused the development of AI resistance (Sabnis et

19 al., 2005). Persistent treatment with AI can activate an alternative pathway of ERα activation in a ligand-independent manner by activating ER through EGFR signaling

(Wong and Chen, 2012).

Acquired drug resistance by cancer cells is a major problem associated with therapy. The factors that contribute to the development of drug resistance include

DNA damage repair and overexpression of efflux-drug transporters and deregulation of cell death by evasion of apoptosis (Raguz and Yague, 2008). Overexpression of apoptosis regulating proteins such as Bcl-xL, Bcl-2 and Mcl-1 was reported to be associated with drug resistance (Tian et al., 2012). In addition, the differential expression of membrane proteins such as solute carriers, channels and drug- transporters have been demonstrated to play important roles in drug resistance (Niture et al., 2013).

A growing body of evidence implicates Nrf2 in drug resistance. Nrf2 and

INrf2 (Keap1) are cellular sensors of xenobiotic, drugs and radiation induced

ROS/electrophilic stress (Kaspar et al., 2009; Niture et al., 2013). Nrf2 is a nuclear factor that controls the expression and coordinated induction of a battery of genes encoding detoxifying enzymes, antioxidants, drug transporters and anti-apoptotic proteins (Hayashi et al., 2003; Maher et al., 2007; Slitt et al., 2006). Thus, we hypothesized that prolonged treatment with aromatase inhibitors upregulates Nrf2 and its target genes that contributes to the AI-resistance. AI-sensitive cell lines (MCF-7Ca and AC1) and resistant cell lines (LTLT-Ca, AnaR and ExemR) have been used in this study. The main purpose of this study is to gain insights into molecular

20 mechanisms mediated by Nrf2 in AI resistance and develop an effective therapy to improve breast cancer treatment.

Figure 2-2. Flow chart showing generation of aromatase inhibitor-resistant cells. Human aromatase was introduced in MCF-7 breast cancer cells and deginated as MCF-7Ca or AC1, which were AI-sensitive cells. LTLT-Ca cells were long-term letrozole treated cells generated from mouse MCF-7Ca xenograft tumor treated with letrozole for 56 weeks. Likewise, AnaR and ExemR were generated and resistance to anastrozole and exemestane.

In this study, we have shown that aromatase inhibitor-resistant breast cancer

(long term letrozole treated letrozole-resistant LTLT-Ca, anastrozole-resistant AnaR and exemestane-resistant ExemR) cells express lower levels of INrf2 and higher levels of Nrf2, as compared to AI-sensitive MCF-7Ca and AC1 cells, respectively.

The studies also revealed that upregulation of Nrf2 was due to downregulation of

INrf2 and lower ubiquitination and a slower rate of degradation of Nrf2 in AI- resistant cells. Increased expression of cytoprotective molecules (HO-1, GCLC,

GSH), drug-transporters MRP1 and MRP4 (multidrug resistance-associated protein),

BCRP (breast cancer resistance protein), and Mcl-1 (anti-apoptotic proteins) mediated by higher levels of Nrf2 contributed to reduced apoptosis that led to chemoresistance.

Interestingly, Nrf2-knocked down LTLT-Ca cells with lowered levels of Nrf2 showed reduced levels of ALDH (aldehyde dehydrogenase), a marker of tumor-initiating cells and significantly decreased mammosphere formation as compared to control

LTLTCa-V cells expressing higher levels of Nrf2. These results collectively suggest that persistent AI treatment down regulated INrf2 leading to upregulation of Nrf2 and downstream cytoprotective proteins that resulted in increased drug resistance.

Materials and Methods:

Chemicals and Reagents

The chemicals and antibodies were purchased from the various companies as indicated. G418 sulfate (Cat.No.61-234-RG) from Cellgro, Manassas, VA. Penicillin/ streptomycin (Cat. No. A5955), doxorubicin (Cat. No.D1515), epoxomicin (Cat. No.

E1383) and anti-β-Actin antibody (Cat. No. A5441) from Sigma, St. Louis, MO.

RNeasy mini kit (Cat. No. 74104) from Qiagen, Valencia, CA. High capacity cDNA reverse transcription kit (part No. 4368814), primers and probes (Nrf2;

Hs00975960_ml, Keap1; Hs00202227_ml, GCLC; Hs00155249_ml, NQO1;

Hs00168547_ml, GUSB; Hs99999908_ml) used for quantitative real time-PCR were from Applied Biosystems, Foster city, CA. Puromycin dihydrochloride (sc-108071), control shRNA lentiviral particles-A (sc-108080), Nrf2 shRNA (sc-37030-V), Anti-

Nrf2 (sc-13032), anti-Keap1 (sc-15246), anti-HO-1 (sc-10789), anti-NQO1 (sc-

32793), anti-Bcl-2 (sc-492), anti-Bcl-xL (sc-8392), anti-Mcl-1 (sc-819), anti-Lamin B

(sc-6217), anti-Mdr-1 (sc-8318), anti-MRP1 (sc-13960), anti-HER-2 (sc-284), anti-

Ub (sc-8017), anti-Ku70 (sc-17789) antibodies were from Santa Cruz Biotechnology,

Paso Robles, CA. Glutathione assay kit (item No. 703002) was from Cayman

Chemical, Ann Arbor, MI. Ultra-low-attachment of 24 well plate (Cat. No. 3473) for mammosphere from Corning, Acton, MA. DCFDA Cellular ROS detection assay kit

(Cat. No. ab113851) and γ-glutamylcysteine synthatase (GCLC, ab40929) antibody obtained from Abcam, Cambridge, MA. Anti-ER-α (Cat. No. 1545) purchased from

Immunotech. Anti-LDH (Cat. No. 3558) from Cell Signaling, Danvers, MA, Anti-

MRP4 (Cat. No.ALX-801-038) from Enzo life science, anti-BCRP (Cat. No. OP191-

200UL), Ku80 (Cat. No.NA54) and MG-132 (Cat. No. 474790) from Millipore,

Billerica, MA. Aldefluor assay kit from Stem cell technologies, Vancouver, Canada.

Aromatase Inhibitors (Letrozole and Anastrazole) were provided by Dr. Angela

Brodie‟s laboratory, University of Maryland, Baltimore.

Cells and Cell Culture Conditions

Aromatase-inhibitor (AI) sensitive cells (MCF-7Ca and AC1) and AI-resistant cells (LTLTCa and AnaR) were obtained from Dr. Angela Brodie‟s laboratory.

Generation of aromatase inhibitor-resistant cells has been described previously

(Jelovac et al., 2005). Briefly, human breast cancer MCF-7 cells were stably transfected with the human aromatase gene to generate MCF-7Ca cells (Zhou et al.,

1990). Letrozole-resistant LTLT-Ca cells were isolated from MCF-7Ca mouse xenograft tumors treated with letrozole for 56 weeks. Similarly, AC1 cells were generated from the MCF-7 cells by stable transfection with human aromatase gene.

AnaR cells were anastrazole-resistant cells isolated form AC1 mouse xenograft tumors treated with anastrozole for 14 weeks (Jelovac et al., 2005). Examination of the levels of estrogen receptor alpha (ERα) in different passages of MCF-7Ca

(passage numbers 18, 23, 43, 73) and LTLT-Ca (passage numbers 22, 37, 58, 67) cells did not demonstrate any significant changes in ERα with cell passages. However, as already known, LTLT-Ca cells express significantly less ERα than MCF-7Ca cells

(Macedo et al., 2009). MCF7Ca cells were grown in DMEM containing 700 µg/ml

G418 sulfate and 5% FBS. DMEM with 700 µg/ml G418 sulfate and 10% FBS were used to culture AC1 cells. LTLT-Ca cells were maintained in phenol red free

Modified IMEM containing 700 µg/ml G418 sulfate, 1 µM letrozole and 5% charcoal stripped FBS. AnaR and ExemR cells were also grown in modified IMEM with 700

µg/ml G418 sulfate, 20 µM anastrozole for AnaR and 5 µM exemestane for ExemR and 10% charcoal stripped FBS. Nrf2-knocked down LTLT-Ca (LTLTCa-Nrf2KD) 24 cells were cultured in Modified IMEM medium supplemented with 700 µg/ml of

G418 and 5% charcoal stripped FBS. All cell lines were grown in monolayer in medium containing 1% Penicillin/streptomycin in an incubator at 37°C with 95% air and 5% CO2.

Generation of Stable LTLT-Ca Cells Expressing Nrf2shRNA

LTLT-Ca cells were transduced with lentiviral particles containing Nrf2 shRNA or control shRNA and stable cells were selected in presence of 10 µg/ml puromycin and designated as LTLTCa-Nrf2 knock down (LTLTCa-Nrf2KD) and

LTLTCa-Vector Control (LTLTCa-V) respectively.

Western Blotting

Cells were treated with DMSO (vehicle control) or t-BHQ as indicated in figures. The cells were harvested at the time of 95% confluence and washed with cold phosphate-buffered saline and lysed in RIPA buffer (50 mM Tris, pH 8.0, 150 mM

NaCl, 0.2 mM EDTA, 1% Nonidet P-40, 0.5% sodium deoxycholate) supplemented with 1x protease inhibitor cocktail (Roche). Subcellular fractionation of the cells was performed according to manufactures protocol (Active Motif, CA). Proteins were quantified using Bio-Rad protein assay. Total cell lysates (30-50 µg) were separated on SDS-PAGE and transferred to nitrocellulose membranes. After blocking in 5% non-fat milk solution in TBST (Tris Buffered Saline Tween -20), membranes were incubated with the primary antibodies over night at 4°C and washed 4 times with

TBST. Subsequently, the membranes were incubated with secondary antibodies conjugated with HRP (horseradish peroxidase) at room temperature for 1h and

25 washed four times with TBST, then, using chemiluminescence system, the protein bands were visualized.

Nrf2 Degradation Assay

Cells were treated with 25 µg/ml cycloheximide (CHX) for the indicated time points, washed twice with ice cold 1X PBS and lysed in RIPA buffer with protease inhibitors. 30 µg of total cell lysate was loaded per well on 10% SDS PAGE gel, transferred and immunoblotted with Nrf2 and β-actin antibodies. Nrf2 band intensity was quantified and normalized to β-actin. The relative levels of Nrf2 from sample with zero (0) minute was considered as base level. The graphs represent the natural logarithm of the relative levels of the Nrf2 protein as a function of the CHX inhibition of protein synthesis. The half-life of protein was determined in the linear range of the degradation curve using formula t1/2=ln2/k, k is the slope of the degradation curve.

Immunoprecipitation and Ubiquitination Assay

For ubiquitination assay, cells were treated with 2 µM of MG-132 for 16h and lysed in RIPA buffer. One mg of whole cell lysate was immunoprecipitated with 1 µg of rabbit IgG or Nrf2 antibody by incubating the reaction mixture overnight in RIPA buffer supplemented with 0.1% SDS at 4ºC. After adding 20 µl of washed protein

A/G plus beads (Santa Cruz Biotechnology), the mixture was incubated for 2h at 4ºC and centrifuged at 4000 RPM for 1 minute. The beads were washed twice with RIPA buffer. 30µl of SDS-sample dye was added to each tube and boiled for 5 minutes, centrifuged and the immunoprecipitated Nrf2 released in the supernatant was

26 separated by 8% SDS PAGE and immunoblotted with anti-ubiquitin antibody. The same blot was stripped and re-probed for Nrf2.

Cell Survival and Cell Death Assays

MCF-7Ca and LTLT-Ca cells were seeded at a density of 2 x 104 cells per well in 24-well plates and treated with different concentration of doxorubicin and etoposide for 72h. The cells were incubated with freshly prepared MTT dye (100

µl/well of 5 mg/ml MTT dye in PBS) for 1h. MTT dye is reduced by mitochondrial dehydrogenases of viable cells to form insoluble formazan crystals. The amount of formazan produced is proportional to viable cells. DMSO was added to the wells to dissolve the formazan crystals and the color developed was quantified spectrophotometrically at 570 nm. Cell viability was calculated from the absorbance and normalized to the value of the corresponding control cells. Each data point represents a mean ± S.D. from three independent experiments. For DNA fragmentation assay, 2 x 104 cells of MCF-7Ca, LTLTCa-V and LTLTCa-Nrf2KD were seeded per well in 12-well plates and treated with 20 µM of etoposide in triplicate wells for 72h. Following manufacturer‟s protocol, cytoplasmic histone- associated DNA fragments (mono and oligonucleosomes) was determined using Cell

Death Detection kit (Roche). The color obtained from enzyme linked immunosorbent assay (ELISA) was recorded at 405 nm wavelength using microplate reader (Bio-Tek

Ex-800). The reading was normalized against vehicle treated cells.

ALDEFLUOR Staining

ALDEFLUOR assay (Stem Cell Technologies, Vancouver, BC) was performed according to the manufacturer‟s instructions. MCF-7Ca and LTLT-Ca 27 cells, and LTLTCa-V and LTLTCa-Nrf2KD cells expressing ALDH (aldehyde dehydrogenase) were stained with aldefluor reagent (ALDH substrate) and identified by comparing the same sample with and without the ALDH inhibitor DEAB

(diethylaminobenzaldehyde). ALDH catalyze the conversion of non-fluorescence substrate to the fluorescence product. Cells were acquired using FACS Canto and analyzed using FloJo software (BD Biosciences, Franklin Lakes, NJ). Dead cells were excluded based on light scatter characteristics and using viability dye (propidium iodide) gating parameters.

Isolation of Tumor Initiating Cells (TIC) Using ALDEFLUOR Staining

Aldefluor assay/Aldehyde dehydrogenase assay (Stem Cell Technologies,

Vancouver, Canada) was performed according to the manufacturer‟s instructions.

Briefly, LTLT-Ca cells were stained with adlefluor reagent along with or without the inhibitor of ALDH, DEAB and sorted using FACS ARIA (BD Biosciences, Franklin

Lakes, NJ). All cells showing differential ALDH staining pattern were sorted and designated as ALDH-high and ALDH-low cells based on highest and lowest expression of ALDH enzyme respectively.

Mammosphere Assay

Mammosphere assay was performed using reagents from Stem Cell

Technologies, as per manufacturer‟s instructions. Briefly, LTLTCa-V and LTLTCa-

Nrf2KD cells were suspended in complete mammocult media and 1x104 cells per well were plated in ultra-low-attachment 24 well plates. Mammospheres were counted

28 after 3 weeks. Stabilized spheres with a colony count of at least 50 cells were considered as mammospheres (Gilani et al., 2012).

Gene Expression Analysis

Total RNA was isolated from the MCF-7Ca, LTLTCa-V and LTLTCa-

Nrf2KD cells using RNeasy mini kit, following manufacture‟s protocol. High cDNA reverse transcription kit was used to synthesize cDNA using 1µg of total RNA as template and the cDNA was amplified and quantified to determine the target gene expression using TaqMan specific primers and probes in 7500 real-time PCR system

(Applied Biosystems).

Glutathione Quantification

Total glutathione content was determined spectrophotometrically using

Cayman‟s glutathione assay kit following the manufacture‟s protocol. Briefly, 1x105 cells were seeded on 6-well plates and lysed in 1x buffer supplied in Glutathione detection kit and oxidized and reduced form of glutathione was quantified following the kit protocol using TECAN plate reader (405nm). Glutathione content is expressed as µM/µg protein.

Densitometry and Statistical Analyses

Image J software (NIH) was used to quantify the intensity of proteins bands.

The protein bands were normalized against loading controls. Data from cell survival, cell death assay, quantitative real-time PCR and intensity of protein bands were

29 analyzed using a two-tailed student‟s test. Data were presented as the mean ± standard deviation. Two data sets with p-value ≤0.05 were considered statistically significant.

Results:

Nrf2 was upregulated and INrf2 was downregulated in AI-resistant cell lines

Figure 2-3. Characterization of AI-sensitive and resistant breast cancer cells. (A-C) AI-resistant and sensitive cells were lysed and analyzed for Nrf2 and INrf2 by Western blot. (D) cells were treated with vehicle or 50 µM t-BHQ for12h and lysed and analyzed for relative levels of Nrf2, its target genes HO-1 and NQO1, and INrf2 (Keap1) proteins in MCF-7Ca and LTLT-Ca cells by Western blot. E) Higher level of Nrf2 protein was detected in nuclear fraction of LTLT-Ca cells as compared to MCF-7Ca cells.

Initial characterization of AI-resistant breast cancer cells by western blot analysis revealed that Nrf2 was upregulated and INrf2 was downregulated in three AI- resistant cells relative to their AI-sensitive counterparts (Fig. 2-3A-C). This data

30 suggested a possible mechanism by which upregulation of Nrf2 might be due to a decrease in INrf2 levels. As the expression pattern of Nrf2 and INrf2 in all three AI- resistant cell lines was similar and importantly, letrozole is the most potent aromatase inhibitor, most of the studies were performed utilizing letrozole resistant LTLT-Ca and MCF-7Ca as control cells. In some cancers, because of mutations of either Nrf2 or INrf2, antioxidant-mediated induction of Nrf2 may not be observed. Therefore, we determined whether Nrf2 signaling was intact or not, by treating the cells with t-BHQ

(tertiary-butyl hydroquinone), an activator of Nrf2. Basal and inducible levels of Nrf2 and its target genes GCLC, HO-1and NQO1 were evaluated. As expected, Nrf2 was stabilized by t-BHQ; consequently, target genes of Nrf2, GCLC and HO-1 were also induced upon treatment with t-BHQ (Fig.2-3D). Surprisingly, protein levels of NQO1, a prototype target gene of Nrf2 was lower in LTLT-Ca cells as compared to MCF-7Ca cells. We expected higher level of NQO1 in parallel with Nrf2 level. In this study, we did not investigate the underlying mechanism by which NQO1 was downregulated.

Based on observation from our lab, NQO1 is a relatively long-lived protein and proteasome inhibitor MG132 does not stabilize NQO1 (Fig.3.3C), we assumed that

NQO1 may be degraded by autophagy. However, INrf2 levels in both cell lines were decreased with t-BHQ treatment. This observation indicates that antioxidant t-BHQ may activate the mechanism by which INrf2 is degraded (Fig.2-3D). Nuclear accumulation of Nrf2 substantiated the upregulation of its target genes GCLC and

HO-1 in LTLT-Ca cells (Fig.2-3E).

Anti-apoptotic proteins, drug transporters and glutathione levels were higher in

AI-resistant cells

Western blot analysis showed that LTLT-Ca cells expressed significantly higher levels of drug transporters (MRP-1, MRP-4 and BCRP) and anti-apoptotic proteins (Bcl-xL and Mcl-1) as compared with MCF-7Ca cells (Fig.2-4A-B).

Surprisingly, Bcl-2 was not detected in LTLT-Ca cells. Expression of pro-apoptotic proteins Bax and Bad was not significantly different in the two cell lines (Fig.2-4C).

Figure 2-4. Levels of anti-apoptotic proteins, drug transporters and glutathione in AI- resistant cells. (A-C) cells were lysed and analyzed by western blot for (A) drug efflux pumps (MRP1, MRP4 and BCRP), (B) anti-apoptotic proteins (Bcl-2, Mcl-1 and Bcl-xL), and (C) pro-apoptotic proteins (Bax and Bad). (D) Total and reduced form of glutathione (GSH) in MCF-7Ca and LTLT-Ca cells was determined spectrophotometrically at 570 nm.

Since Nrf2 downstream gene GCLC, which is the rate limiting enzyme of glutathione synthesis, was highly upregulated in LTLT-Ca cells, cellular glutathione content was quantified. Cellular level of glutathione (total and reduced) was found to be significantly higher in LTLT-Ca cells, as compared to sensitive MCF-7Ca cells

(Fig.2-4D). These data together demonstrated that persistent exposure of cells with

AIs (AI-resistant cells were continuously cultured in medium containing AI) decreased INrf2 and increased nuclear Nrf2 and expression of Nrf2-regulated genes

(anti-apoptotic proteins Bcl-xL, drug transporters MRP1, MRP4 and BCRP, and

32 antioxidant related molecules GCLC, HO-1), Mcl-1 and glutathione. The results suggested an activation of Nrf2 and its target genes in the AI-resistant cells.

Knocking down Nrf2 in LTLT-Ca cells decreased Nrf2 target gene expression and increased sensitivity to doxorubicin and etoposide

To investigate the role of Nrf2 in drug resistance, we knocked down Nrf2 in

LTLT-Ca cells. MCF-7Ca, LTLTCa-V (vector control) and LTLTCa-Nrf2KD cells were immunoblotted for Nrf2 and INrf2; detoxifying proteins GCLC and NQO1; drug- transporters MRP1, MRP4 and BCRP; and anti-apoptotic proteins Mcl-1 and

Figure 2-5. LTLTCa-Nrf2KD cells were more sensitive to doxorubicin and etoposide than control cells (LTLTCa-V). (A -D) Knocking down of Nrf2 by Nrf2shRNA decreased the mRNA and protein levels of Nrf2 and some Nrf2-regualted genes (detoxifying enzyme, drug transporters and anti-apoptotic proteins) in LTLT-Ca cells. (E) The percentage of survival cells in LTLTCa-Nrf2KD cells was less than LTLTCa- V cells treated with doxorubicin for 72h. p-values were calculated by data set from LTLT-Nrf2KD and control cells (LTLTCa-V). (F) LTLTCa-Nrf2KD cells demonstrated increased DNA fragmentation upon etoposide treatment as compared to LTLTCa-V and MCF-7Ca cells. 33

Bcl-xL and actin (Fig.2-5A-D). Even though, there was no difference in Nrf2 mRNA levels in MCF-7Ca and LTLTCa-V cells, the Nrf2 protein level was increased in

LTLTCa-V cells because of the low levels of INrf2 protein. It is worth mentioning that INrf2 facilitates the degradation of Nrf2. Knocking down Nrf2 in LTLT-Ca cells reduced the mRNA and protein levels of Nrf2, GCLC and NQO1, but it did not change mRNA and protein levels of INrf2, as compared to the control LTLTCa-V cells (Fig.2-5A & B. Accordingly, expression levels of drug-efflux pumps MRP4 and

BCRP, and anti-apoptotic proteins Bcl-xL and Mcl-1 were also decreased in LTLT-

Ca-Nrf2KD cells. However, we did not observe any significant decrease in the levels of MRP1 in LTLTCa-Nrf2KD cells. Since letrozole does not have a direct cytotoxic effect, the sensitivity of LTLTCa-Nrf2KD cells to doxorubicin and etoposide were determined (Fig.2-5E and F). As expected, LTLTCa-V cells were more resistant to doxorubicin compared to MCF-7Ca cells. Interestingly, shRNA-mediated inhibition of Nrf2 sensitized LTLTCa-Nrf2KD cells to doxorubicin (Fig.2-5E). We found that the concentrations of doxorubicin (0.25µM to 0.75µM) significantly decreased cell survival in LTLTCa-Nrf2KD cells as compared to control LTLTCa-V cells.

Accordingly, shRNA-mediated inhibition of Nrf2 in LTLT-Ca cells also increased etoposide-induced DNA fragmentation (Fig.2-5F). These results together indicate that lowering the expression of Nrf2 signaling sensitizes AI-resistant breast cancer cells to doxorubicin and etoposide.

Nrf2 was less ubiquitinated and degraded slowly in LTLT-Ca cells

Nrf2 is constantly degraded by the INrf2-mediated ubiquitin-dependent proteasome pathway. Lower levels of INrf2 may result in less degradation of Nrf2 in

LTLT-Ca cells. Therefore, a degradation assay was performed by blocking new

34 synthesis of Nrf2 with cycloheximide to determine the half-life of Nrf2. The results presented in (Fig.2-6A) indicated that Nrf2 had a longer half-life (52.1 minutes) in

LTLT-Ca cells than in MCF-7Ca cells (28.9 minutes). The slower rate of Nrf2

Figure 2-6. Nrf2 was less ubiquitinated and degraded slowly in LTLT-Ca cells. (A) Cells were treated with 25 µg/ml cycloheximide (CHX) for the indicated time points and 30 µg of total cells lysate was immunoblotted with Nrf2 and β-actin antibodies. The graphs represent the natural logarithm (Ln) of the relative levels of the Nrf2 protein versus the CHX chase time and the half-life of Nrf2 was determined using the linear part of the degradation curve. (B) 1 mg of total cell lysate from the cells treated with MG-132 (proteasome inhibitor) was immunoprecipitated with 1µg of rabbit IgG or Nrf2 antibody. The immunoprecipitated Nrf2 was immunoblotted for ubiquitin and Nrf2.

degradation in LTLT-Ca cells directed us to examine the ubiquitination status of Nrf2.

AI-sensitive and resistant cells were treated with proteasome inhibitor MG-132 to accumulate the ubiquitinated Nrf2 by blocking its degradation. LTLT-Ca cells showed decreased Nrf2 ubiquitination as compared to MCF-7Ca cells (Fig.2-6B).

Nrf2 degradation and ubiquitination assay revealed that the rate of degradation of

Nrf2 was significantly slower in LTLT-Ca cells compared to MCF-7Ca cells

(Fig.6A). These results collectively suggested that higher level of Nrf2 in LTLT-Ca cells was due to decreased ubiquitination and longer half-life of Nrf2.

Tumor initiating cells (TIC) expressed lower INrf2 and higher DNA repair proteins, HER2, Nrf2 and Nrf2 target gene GCLC

Figure 2-7. Tumor initiating cell (TIC) cells expressed lower INrf2 and higher Nrf2, GCLC, DNA repair proteins and HER2. LTLT-Ca cells were subjected to aldefluor staining and the cells expressing ALDH were gated into high ALDH-and low ALDH cells. Cells were sorted using FACS ARIA as high ALDH (TIC) and low ALDH (non-TIC) fractions. The cells were lysed and total cell lysate was immunoblotted for (A) for Nrf2, INrf2 and GCLC and (B) for human HER2, and DNA repair enzymes (Ku70 and Ku80). * Control cells that were sorted by propidium iodide exclusion.

Enhanced DNA repair is one of the reported mechanisms of drug resistance.

Furthermore, tumor initiating cells (TIC) have a lowered ability to undergo apoptosis and a higher ability for DNA repair, making them more resistant to DNA targeting cancer therapy (Ma et al., 2008; Nicolini et al., 2011). TIC cells from AI-resistant

LTLT-Ca were acquired and isolated by FACS sorting based on the activities of aldehyde dehydrogenase (ALDH), which is considered as marker of TIC cells. We then examined the levels of Nrf2, INrf2, GCLC, HER2 and DNA repair proteins

Ku70 and Ku80 in TIC cells (high-ALDH) and compared that with low ALDH (non–

TIC) cells (Fig.2-7A & B). MCF-7Ca and LTLT-Ca cells were also analyzed by western blotting as control along with non-TIC and TIC cells. We noticed that the levels of Nrf2 and INrf2 were not much different as we have observed in rest of the other experiment. During flow cytometry analysis, these cells were at room temperature for 4-5 hours that might be one of the reasons for fluctuation of Nrf2 and

INrf2 levels. LTLTCa-high ALDH (TIC) cells expressed lower levels of INrf2 and higher levels of Nrf2 and Nrf2 downstream GCLC gene expression than LTLTCa-low

ALDH (non-TIC) cells. Intriguingly, those TIC cells that contain high levels of

ALDH expressed higher levels of NHEJ (non-homologous end-joining) DNA repair proteins Ku80 and Ku70, as compared to MCF-7Ca and LTLT-Ca cells. This observation may be very significant since TIC cells are believed to contribute to drug resistance through multiple mechanisms including increase in DNA repair system.

shRNA-mediated inhibition of Nrf2 in LTLT-Ca cells significantly decreased

TIC cells and mammosphere formation

To investigate whether there was any link between levels of Nrf2 and TIC population, we determine the percentage of TIC cells in LTLT-Ca cells. MCF-7Ca and LTLT-Ca cells, LTLTCa-V and two clones of LTLTCa-Nrf2KD cells with different levels of Nrf2 were stained for stem cell marker, ALDH in the presence and absence of ALDH inhibitor, DEAB (diethylaminobenzaldehyde). Flow cytometry analysis of MCF-7Ca and LTLT-Ca (Fig.2-8A & B) demonstrated that LTLT-Ca cells have a greater number of tumor-initiating cells with high ALDH (5.19%) as compared to MCF-7Ca cells with high-ALDH.(0.3%). The higher percentage of TIC cells in

LTLT-Ca cells was corresponding to high level of Nrf2 in LTLT-Ca cells (Fig.2-8B).

Figure 2-8. LTLT-Ca-Nrf2KDcells expressed lower levels of ALDH as compared to control LTLTCa-V cells. Two clones of LTLT-Ca cells expressing Nrf2shRNA were selected. (A and C) Cells were subjected to Aldefluor staining and the percentage of cells expressing high ALDH was determined by comparing the same sample with and without the ALDH inhibitor DEAB (diethylaminobenzaldehyde). LTLTCa-Nrf2KD cells contained lower percentage of tumor initiating cells. Data was acquired using FACS CANTO. (B and D) Cells were lysed and immunoblotted for Nrf2 to show different levels of Nrf2.

Knocking down Nrf2 in LTLT-Ca (Fig.2-8D, clone #2 and 1) cells dramatically decreased the number of tumor initiating cells (1.58 % and 0.05 %), as compared to

LTLT-Ca-V cells (2.27 %) (Fig.2-8C).The lowest expression of Nrf2 in LTLTCa-

Nrf2KD clone #1 correlated with minimal percentage of cells with high ALDH

(0.05%). The results demonstrated a direct relationship between the levels of Nrf2 and magnitude of ALDH expression, suggesting that shRNA inhibition of Nrf2 led to

Nrf2 dependent decrease in number of TIC cells with high ALDH.

One of the properties of TIC cells is their ability to form mammospheres (Gilani et al., 2012). To investigate whether Nrf2 has any role in mammosphere formation,

38 we then examined the ability of LTLTCa+Nrf2KD cells to form mammospheres

(Fig.2-9). The results (compare Fig.2-8D and 2-9) revealed that cells expressing lower levels of Nrf2 have less ability to form mammospheres. In other words, shRNA inhibition of Nrf2 in LTLT-Ca cells led to Nrf2 concentration-dependent decrease in mammosphere formation. In conclusion, there is a positive correlation between Nrf2,

ALDH high-TIC cells and mammosphere formation with implications in AI- resistance that warrants further studies.

Figure 2-9. Nrf2-Knocked down LTLT-Ca cells formed fewer mammospheres. LTLTCa-V and LTLT-Nrf2KD (1X104) cells were seeded in complete Mammocult media per well in ultra-low-attachment 24 well plates. Mammospheres were counted after 3 weeks and photographed. p-values were calculated using number of mammosphere of LTLT-Ca cells and that of LTLTCa-Nrf2KD.

Discussion:

Nrf2 is a transcription factor that in response to oxidative and electrophilic stress coordinately induces expression of a battery of genes encoding detoxifying enzymes, antioxidants such as glutathione, membrane transporters, and anti-apoptotic proteins (Kaspar et al., 2009; Niture and Jaiswal, 2013). Therefore, a role of Nrf2 was posited in breast cancer drug resistance but remained unknown. The study was the first report demonstrating a role of Nrf2 in AI-resistance breast cancer cells. LTLT-Ca

39 cells showed lower INrf2 and higher Nrf2, as compared with MCF-7Ca cells. The same pattern was observed when Nrf2 and INrf2 were analyzed in anastrozole- resistant AnaR and exemstane-resistant ExemR cells. Likewise, anti-apoptotic proteins Mcl-1 and Bcl-xL were upregulated in three AI-resistant cells, surprisingly,

Bcl-2 was downregulated in AnaR and ExemR cells, and it was not detected in LTLT-

Ca cells. The decrease in INrf2 led to decreased ubiquitination and degradation of

Nrf2 leading to the stabilization and nuclear translocation of Nrf2. High levels of Nrf2 in the nucleus led to coordinated induction of Nrf2 downstream cytoprotective proteins including detoxifying enzymes, membrane transporters and anti-apoptotic proteins. The increased cytoprotective proteins have been known to contribute to drug resistance (Wu et al., 2012). Higher level of glutathione content has correlated with reduced doxorubicin sensitivity of LTLT-Ca cells, strongly suggesting a role of Nrf2 in AI resistance. The role of Nrf2 in AI resistance was further confirmed by observations that inhibition of Nrf2 in LTLT-Ca cells sensitized the cells to doxorubicin and etoposide. However, it is important to know whether the inhibition of

Nrf2 in AI-resistant cells can sensitize them to aromatase inhibitors.

The studies also revealed that lower levels of INrf2 in AI-resistant LTLT-Ca cells is not due to proteasomal degradation of INrf2 as inhibitors of proteasomes failed to increase the level of INrf2. Further studies demonstrated that down regulation of INrf2 in letrozole-resistant LTLT-Ca cells was due to significant decrease in INrf2 transcription. This raised an intriguing question regarding the mechanism of AI-mediated down regulation of INrf2 gene expression that was investigated and described in chapter three.

Increase in expression of drug transporters is one of the mechanisms of drug resistance in cancer cells. We detected significantly higher levels of drug transporters like BCRP, MRP1 and MRP4 in AI-resistant LTLT-Ca cells, as compared to AI- sensitive MCF-7Ca cells (Fig. 1D). Inhibition of Nrf2 by shRNA significantly decreased the levels of Nrf2 and Nrf2 downstream genes such as GCLC and NQO1

(Fig 2A). Accordingly, protein levels of anti-apoptotic proteins, Bcl-xL and Mcl-1, drug transporters BCRP (breast cancer resistance protein) and MRP4 (multidrug resistance-associated protein 4) were also decreased (Fig 2 B). Several studies have shown that Nrf2 regulates many drug transporters like MRP1 and MRP4 through

Nrf2-ARE pathway (Ji et al., 2013; Maher et al., 2007; Song et al., 2009). siRNA- mediated downregulation of Nrf2 has recently been demonstrated to reduce the expression of MRP1 in small cell lung cancer cell line H69AR (Ji et al., 2013).

However, we did not observe any significant reduction in the levels of MRP1 in Nrf2- knocked down LTLT-Ca cells. One of the possible reasons might be that MRP1 is not only regulated by Nrf2 but also other transcription factors including NF-кB and c-Jun

(Ronaldson et al., 2010). Therefore, the role of Nrf2 in MRP1 regulation might be less significant than in regulation of other MRPs such as MRP4 (Fig.2B).

Previous studies from our laboratory and other groups have demonstrated that

Nrf2 upregulates biotransformation enzymes (McMahon et al., 2001), drug transporters (Maher et al., 2007) and anti-apoptotic proteins (Niture and Jaiswal,

2012) and contributed to reduced efficacy of drugs and aversion to apoptosis leading to drug resistance. LTLTCa-V cells with more Nrf2 as compared to MCF-7Ca cells express more drug transporters like MRP4 and BCRP and anti-apoptotic proteins like

Bcl-xL and Mcl-1 (Fig. 2A & B). A recent publication from our laboratory had shown

41 that knocking down of Bcl-xL by siRNA and challenging with etoposide increased apoptosis (Niture and Jaiswal, 2013). Similarly, it has also been shown in breast cancer cells that down regulation of MRP1 (ABCC1) and BCRP (ABCG2) by siRNA enhanced chemosensitivity and decreased drug resistance (Pawlowski et al., 2013).

Aromatase inhibitor-resistant cells showed a higher population of TIC cells that are believed to contribute to resistance to chemotherapy and radiation. TIC cells showed lower levels of INrf2 and higher levels of Nrf2. Although, a link between

Nrf2 signaling and DNA repair mechanism has not been established yet and the fact that LTLT-Ca cells express high levels of Nrf2 prompt us to speculated that Nrf2 might have a role in DNA repair. TIC cells have slightly higher expression of Ku70 and Ku80 proteins, which participate in the NHEJ (non-homologous end-joining) pathway. The NHEJ pathway is known to be active in repairing DNA double strand breaks, one of the most lethal forms of DNA damage. Moreover, NHEJ DNA repair is also known to be active in G0/G1 phase of the cell cycle in which many “stem” cells reside (Rassool and Tomkinson, 2010). These observations suggested that high levels of Nrf2 in TIC cells contributed to AI-resistance. While a previous publication found Ku proteins to be reduced in LTLT cells (Tobin et al., 2012), their high expression levels in TIC cells are interesting observation. NHEJ pathway may participate in drug resistance in “stem”-like TICs by increasing repair of DNA damage, leading to cell survival. This conclusion was further strengthened from the following observations. First, shRNA-mediated inhibition of Nrf2 led to Nrf2 concentration-dependent reduction in TIC subpopulation. It is noteworthy that a role of Nrf2 in cell survival by inhibiting apoptosis is well established in other systems

(Niture and Jaiswal, 2012). Second, in related experiments, inhibition of Nrf2 led to

42 significant decrease in mammosphere formation. The decrease in mammosphere paralleled the inhibition of Nrf2. Thus, it is reasonable to conclude that higher levels of Nrf2 in TIC cells may contribute to AI-resistance.

Collectively, the results led to the hypothesis that prolonged treatment with AI down regulates the transcription of INrf2 that leads to increased Nrf2 and Nrf2 downstream cytoprotective proteins including detoxifying enzymes, antioxidants and drug- efflux pumps leading to detoxification of drugs and drug resistance (Fig. 2-10).

In addition, activation of Nrf2 increased expression of anti-apoptotic proteins that led to reduced apoptosis and enhanced drug resistance. Furthermore, higher Nrf2 increased number of TIC cells and mammosphere formation that also contributed to

AI-resistance.

In conclusion, the current studies have presented strong evidence for a role of

Nrf2 in AI- resistance in breast cancer and added an important mechanism of AI- resistance. We believe that this information can be useful for the development of new treatment strategy against, not only AI resistance but also, other therapeutic resistance. The mechanism involving Nrf2 to combat drug resistance in these types of cancer is especially interesting since it is estrogen independent. More importantly, our studies have demonstrated a direct correlation between protein levels of Nrf2 and TIC cells, suggesting that Nrf2 inhibitors could be explored for use as adjuvants with aromatase inhibitors to treat AI-resistant breast cancer.

CHAPTER THREE

THE MECHANISM OF INRF2 DOWNREGULATION IN AROMATASE INHIBITOR-RESISTANT CELLS

Abstract:

INrf2 (Keap1) is an adapter molecule for a cullin3-dependent ubiquitin ligase (E3) complex (Cul3/Rbx1) that constantly ubiquitinates Nrf2, and is degraded by ubiquitination-dependent proteasomal pathway. A feedback auto-regulatory loop between Nrf2 and INrf2 that determines their cellular abundance has been reported.

Importantly, Nrf2 and INrf2 are sensor molecules of various environmental insults, specifically oxidative or electrophilic stress. In the presence of stress, oxidation of reactive cysteine residues in these proteins modulates their normal functions that have recognized them as sensor molecules. Antioxidants or activators of Nrf2 are known to oxidize cysteine residues of INrf2. Modification of cysteine residues of INrf2 triggers the conformational change in INrf2 homodimer that leads to inactivation of INrf2 and stabilization of Nrf2. Deregulation of INrf2 results in activation of Nrf2 signaling which contributes to drug resistance in cancer treatment. Therefore, understanding the regulation of INrf2 will be useful for developing new therapeutic approaches. Our initial characterization of aromatase inhibitor-resistant cells revealed that letrozole- resistant cells express higher levels of Nrf2 and lower levels of INrf2. In this study, we tested our hypothesis that prolonged treatment with AI decreases the expression of

INrf2. Some molecular mechanisms have been reported to elucidate the regulation of

INrf2. However, which mechanism that is involved in downregulation of INrf2 in AI- resistant breast cancer has not yet been investigated. We examined proteasomal and autophagy-mediated protein degradation pathway in AI-resistant cells to determine

44 the mechanism of INrf2 degradation. Inhibition of proteasome system and autophagy

(major protein degradation pathways) did not increase the level of INrf2 in AI- resistant LTLT-Ca cells. We also examined epigenetic mechanisms that might regulate the expression of INrf2. DNA methyltransferase inhibitor (DNMTi) and inhibitors of histone deacetylase (HDACi) failed to upregulate INrf2. Higher level of autophagy activity was detected in LTLT-Ca cells as compared to MCF-7Ca cells.

Conversely, inhibition of autophagy further decreased the levels of INrf2. Moreover, less stability of INrf2 mRNA was observed in LTLT-Ca cells. In conclusion, shorter half-life of INrf2 mRNA in LTLT-Ca cells suggested that INrf2 was, at least in part, regulated post transcriptionally.

Introduction:

Acting as an adapter protein for an ubiquitin E3 ligase complex (Cul3/Rbx1), an endogenous inhibitor of Nrf2 (INrf2), Keap1 controls ubiquitination of Nrf2 that determines the protein level of Nrf2. Previous studies from our laboratory have shown that the promoter of INrf2 has antioxidant response elements (ARE), to which Nrf2 binds to activate transcription of INrf2 (Lee et al., 2007). Therefore, Nrf2 regulates

INrf2 at transcription level. Reduced levels of INrf2 protein were detected in AI- resistant LTLT-Ca cells (Fig.2-1A). This chapter examined the mechanism of regulation of INrf2 in LTLT-Ca cells and compared to those reported in literature.

INrf2 (Keap1)

Kelch-like erythroid cell-derived protein with CNC homology (ECH)- associated protein 1 (Keap1/ KLHL19) is an endogenous inhibitor of Nrf2 (hence also named INrf2) (Dhakshinamoorthy and Jaiswal, 2001; Dhanoa et al., 2013; Itoh et al.,

1999a). INrf2 is known as a substrate adapter molecule for a Cullin3-dependent E3 ligase complex (Cul3/Rbx1) involved in ubiquitination of Nrf2 (Cullinan et al., 2004;

Itoh et al., 2004; Zhang et al., 2004). The N-terminal BTB domain of INrf2 binds to

Rbx1 bound Cul3. The intervening region (IVR) is enriched with reactive cysteine residues. The C-terminal contains double glycine region (DGR/Kelch) domain which binds to Nrf2. Human INrf2 is a 624 amino acid-protein with 27 cysteine residues

(Copple et al., 2008). Homodimerization of INrf2 is essential for functional sequestration, ubiquitination and degradation of Nrf2. Under basal condition, Keap1 constitutively facilitates the ubiquitination-mediated degradation of Nrf2 in the cytosol. It is believed that, in the presence of oxidative or electrophilic stress, some of the reactive cysteine residues of INrf2 are readily oxidized and form either adducts

46 with electrophiles or disulfide-bonds within the INrf2 homodimer. These stress mediated oxidation of INrf2 results in a conformational change in the INrf2 homodimer that causes the INrf2-Cul3/Rbx1 complex to impair ubiquitination of Nrf2

(Zhang and Hannink, 2003).

Figure 3-1. Schematic representation of various domains of INrf2.

Regulation of INrf2

Very few studies have investigated the regulation of INrf2. The first study of the regulation of INrf2 demonstrated a feedback auto-regulatory loop between Nrf2 and INrf2 that controls their cellular abundance (Lee et al., 2007). Nrf2 regulates transcription of INrf2 through interaction of Nrf2 and AREs in the promoter region of

INrf2. On the other hand, INrf2 regulates Nrf2 through ubiquitination-mediated degradation. The auto-regulatory loop between Nrf2 and INrf2/Cul3-Rbx1 was reported to be deregulated in cells with mutations in one of the components of the auto-regulatory loop (Lee et al., 2007). In addition, INrf2 is controlled at the transcription level by promoter hyper methylation. Other cellular events such as ubiquitination, oxidation of INrf2 and autophagy are also known to determine the protein levels of INrf2 (Hong et al., 2005; Taguchi et al., 2012; Zhang et al., 2005).

Epigenetic Regulation of INrf2 DNA methylation and histone modification are the main epigenetic mechanisms involved in the alterations of gene expression in many cancer cells.

These mechanisms determine the degree of interaction between transcriptional activators or co-repressors and promoters that control the expression of that particular gene. A growing body of evidence has suggested an important role of miRNA in regulation of many genes (Eades et al., 2011; van Jaarsveld et al., 2012).

DNA Methylation Aberrant DNA methylation is considered to be a critical epigenetic mechanism in cancer (Sebova and Fridrichova, 2010). Promoter hyper methylation is the dominant event that has been implicated in the transcriptional silencing of tumor suppressor genes in cancer pathogenesis (Cameron et al., 1999). DNA methyltransferases (DNMTs) catalyze the transfer of a methyl group from the universal methyl donor S-adenosyl methionine (SAM) to cytosine in the CpG dinucleotide sequence (Patra and Bettuzzi,

2009). There are primarily three DNA methyltransferases (DNMT1, DNMT3A and

DNMT3B) which are reported to be overexpressed in cancer cells (Daniel et al., 2011;

Dhe-Paganon et al., 2011). A growing body of evidence has already proved that inhibition of DNMTs activates the expression of tumor suppressor genes that were silenced by promoter hyper methylation in various cancers (Karahoca and Momparler, 2013). The expression of Keap1 was found to be suppressed by promoter hyper methylation of

Keap1 in prostate cancer cells (Zhang et al., 2010).

Histone Modification Post-translational modification of histone determines chromatin remodeling mainly through acetylation at lysine (Marks and Xu, 2009). This modification changes the secondary structure of the histone protein tails, increases the distance between DNA strands and histones, allows transcription factors to bind to the promoter region of the gene and subsequently activates gene transcription. Histone acetylation is regulated by the opposing actions of two enzymes, histone acetyltransferases (HATs) and histone deacetylases (HDACs). Abnormal recruitment and overexpression of HDACs have been associated with a wide range of cancers, causing the silencing of tumor suppressor genes (Hoshino and Matsubara, 2010).

Thus, inhibition of HDAC has been proposed as a potential therapeutic strategy to reactivate the transcription of the tumor suppressor genes that have been silenced by histone deacetylation.

Micro-RNA miRNAs are non-coding single stranded RNA of 21-23 nucleotides in length that regulate genes post transcriptionally. miRNA targets mRNA at the 3‟UTR region and exerts its function through degradation of mRNA or inhibition of protein translation (Yang et al., 2011). In addition to transcriptional regulation, INrf2 is also subjected to post transcriptional control. Recently, miR-141 and miR-200a have been shown to negatively regulate INrf2 (Eades et al., 2011).

Autophagy Autophagy (self-eating), known as programmed cell death (PCD) type II, degrades and recycles aged organelles, long-lived, damaged or aggregated proteins to

49 maintain the amino acid pool during starvation. An anti-apoptotic protein Bcl-2 was reported to inhibit autophagy by binding to the Beclin1 that interferes with formation of the phagophore (isolation membrane). The phagophore is a double membrane structure, which elongates and encloses cytoplasmic material in vesicles

(autophagosomes). After maturation, the autophagosome subsequently fuses with the lysosome to form the autolysosome, which degrades its contents (Yang and Klionsky,

2010).

Figure 3-2. Cartoons showing different steps of autophagy. Vescicle nucleation starts after interaction between beclin1 and PI3K-III and other components necessary for the formation of isolation membrane. This step can be blocked by wortmannin or 3MA. Bcl-2 inhibits autophagy by interacting with beclin1, autophagy protein. Upon induction of autophagy, LC3 is immediately processed by Atg4 into a cytosolic form, LC3-I. LC3-I is converted to LC3-II which is embeded in autophagosome membrane. Mature autophagosome fuses with lysosomes and autolysosomes degrades its contents. If autophagy is inhibited by Bafilomycin A1 or chloroquine, the process of autolysosome will be blocked that results in no degradation of its contents.

Autophagy is activated in response to various insults such as amino acid and glucose deprivation, oxidative stress, chemotherapy drugs and hypoxia. Autophagy also protects cells from oxidative stress as oxidized proteins are degraded by

50 autophagy. Thus, autophagy is considered as a pro-survival and drug resistance mechanism (Kongara and Karantza, 2012).

Autophagy is a multi-step process that can be inhibited at different stages. For instance, autophagy inhibitors 3-methyladenine (3-MA) and wortmannin block the association of molecules required for formation of autophagosome by inhibiting class

III PI3K; bafilomycin A1 (specific inhibitor of vacuolar-type H (+)-ATPase) inhibits the acidification of lysosomes and endosomes whereas lysosome lumen alkalizer, chloroquine impairs lysosome. While wortmannin and 3-MA block the initial step, bafilomycin A1 and chloroquine inhibit the late stage of autophagy (Fig.3-2).

Autophagy related protein (Microtubule-associated protein 1 light chain 3 beta-I) LC3B-I is lipidated and embedded into autophago-somal membranes during induction of autophagy. Detection of the LC3B-II (lipidated form of LC3B-I) signifies the activation of autophagy or inhibition of late autophagy. LC3B and P62 molecules, the best characterized autophagy markers, are degraded by autophagy (Pankiv et al.,

2007). Interestingly, p62 interacts with INrf2 through KIR (Keap1 interacting region) domain. It was reported that INrf2 is at least partly degraded by autophagy and the basal level of INrf2 was found to be increased in autophagy-deficient cells (Copple et al., 2010).

Reactive Oxygen Species Reactive oxygen species (ROS), short-lived by-products of oxygen metabolism, is a group of molecules containing one or more oxygen atoms with highly reactive free radicals (molecules with unpaired electron) such as hydroxyl

51 radical (HO·) and superoxide radical (OO·). Leaky electron phosphorylation is the main source of ROS in mitochondria. Upon exposure to ozone, UV radiation, heavy metals like chromium or cadmium, ROS can be generated in cells and acts as a signal transduction molecule. However, high levels of ROS can cause oxidative state resulting in oxidation of proteins (Valko et al., 2006).

In this study, utilizing qRT-PCR, western blotting, quantification of ROS, treating the AI-resistant cells LTLT-Ca with antioxidants, inhibitors of proteasome, autophagy, DNMT and HDAC, molecular mechanism accounting for the downregulation of Keap1 was investigated. We found that mRNA of INrf2 was lower in LTLT-Ca cells as compared to MCF-7Ca cells. High levels of ROS were detected in LTLT-Ca cells. We also found that INrf2 was not upregulated upon treatment of proteasome and autophagy inhibitors, suggesting that INrf2 may not be degraded by proteasome and autophagy in AI-resistant LTLT-Ca cells. In addition, we also assessed the epigenetic regulation of INrf2 by inhibiting histone deacetylases (HDAC) and DNA methyl transferases (DNMT). Inhibitors of histone deacetylases (HDACi) and DNA methyl transferases (DNMTi) did not increase the expression of INrf2 in

LTLT-Ca cells, indicating that promoter hyper methylation and histone modification may not be the mechanism of INrf2 downregulation. The studies showed that INrf2 was downregulated at least partially at post transcription level in LTLT-Ca cells.

Materials and Methods:

Chemicals and Reagents

The chemicals and antibodies were purchased from the various companies as indicated. G418 sulfate (Cat.No.61-234-RG) from Cellgro, Manassas, VA. HBSS

(Ref. No.14025-092) from gibco, Grand Island, NY. 3-methyladenine (Cat. No.BML-

AP502) from Enzo, Farmingdale, NY. Penicillin/streptomycin (Cat. No. A5955), epoxomicin, p62 antibody (Cat. No. P0067) and β-Actin antibody (Cat. No. A5441), curcumin (Cat. No. C1386) , bafilomucin A1 (Cat. No. B1793), actinomycin D (Cat.

No. A-1410), decitabine (Cat. No.A3656) and trichostatin A (Cat. No. T8552) from

Sigma, St. Louis, MO. RNeasy mini kit (Cat. No. 74104) from Qiagen, Valencia,

CA. High capacity cDNA reverse transcription kit (part No. 4368814), primers and probes used for real time-PCR from Applied Biosystems, Foster city, CA. Anti-Nrf2

(sc-13032), anti-Keap1 (sc-15246), anti-HO-1 (sc-10789), anti-NQO1 (sc-32793), anti-Bcl-2 (sc-492) antibodies from Santa Cruz Biotechnology, Paso Robles, CA.

DCFDA Cellular ROS detection assay kit (Cat. No. ab113851) and γ- glutamylcysteine synthatase (GCLC, ab40929) antibody from Abcam, Cambridge,

MA, MG-132 (Cat. No. 474791), wortmannin (Cat. No. 68165) from Millipore,

Billerica, MA, Chemiluminescence system, (Product No.32209) from Thermo

Scientific (Chicago, IL). Chloroquine, MS-275 and suberoylanilide hydroxamic acid

(SAHA) were kindly provided by Dr. Njar‟s laboratory, University of Maryland,

Baltimore.

Cells and Cell Culture Conditions

MCF-7Ca cells and AI-resistant LTLT-Ca cells were grown and maintained exactly the same way as described in materials and methods of chapter two.

Western Blotting

Cells were treated with different chemicals or vehicle (DMSO or ethanol) control as indicated in figures. Western blot was performed as described in materials and methods of chapter two.

mRNA Stability Assay

Cells were treated with 2 µg/ml actinomycin D for indicated time points and washed with 1x PBS and total RNA was isolated. The relative levels of mRNA of

INrf2 from sample with zero (0) minute was considered as base level. The graphs represent the natural logarithm of the relative levels of the mRNA as a function of the actinomycin D chase time. The half-life of mRNA was determined in the linear range of the degradation curve using formula t1/2=ln2/k, where k is the slope of the degradation curve.

Gene Expression Analysis

mRNA was quantified as described in materials and methods (gene expression analysis) of chapter two.

ROS Detection

DCFDA cellular ROS detection assay kit (Abcam) was used to measure the cellular levels of ROS. Cells were trypsinized and washed with PBS. The cells were suspended in 2‟,7‟-dichlorofluorescein diacetate (DCFDA) and incubated at 37°C for

30 min in the dark and washed with 1x buffer. 1x106 DCFDA-stained cells were suspended in 1 ml of 1x supplemented buffer and 1x105 cells in 100 µl of the cell

54 suspension buffer were added to each well of 96 well black plates. Fifty µM of t-butyl hydroperoxide were added and incubated at 37°C for 3h to generate ROS as positive control. ROS-mediated fluorescence intensity was recorded with excitation at 485 nm and emission at 535 nm wavelength using TECAN Infinite M1000 PRO-plate reader.

Densitometry and Statistical Analyses

Densitometry and statistical analysis was performed as described in materials and methods of chapter two.

Results:

Inhibition of proteasome did not restore INrf2

Our initial characterization of AI-resistant cells revealed that letrozole- resistant cells express higher levels of Nrf2 and lower levels of INrf2 (Fig.3-3A). We began investigating the mechanism of INrf2 regulation in letrozole-resistant cells by examining the expression of INrf2 stabilizing proteins (HSP90, CK2 and PKC epsilon), which are recently shown to stabilize INrf2 by our laboratory (Niture and

Jaiswal, 2010). Casein kinase 2 (CK2) phosphorylated INrf2 at threonine 55 that resulted in increased Hsp90-INrf2 interaction and led to stabilization of INrf2 (Niture and Jaiswal, 2010). Protein levels of INrf2 stabilizing molecules were moderately decreased in LTLT-Ca cells that may result in less stabilization of INrf2 in LTLT-Ca cells (Fig.3-3B). Nrf2 is degraded through proteasomal degradation. This is one of the main pathways for protein turnover. Therefore, we also examined whether inhibition of proteasome prevents INrf2 degradation. We treated cells with two inhibitors of proteasome (MG132 and epoxomicin). Stabilization of Nrf2 and its target gene HO-1

55 confirmed the inhibition of the proteasome. However, inhibition of proteasome did not increase the protein levels of INrf2 and Nrf2 target gene NQO1 (Fig.3-3C & D).

Figure 3-3. Inhibition of proteasome did not restore INrf2 in LTLT-Ca cells (A-F) Cells were lysed and analyzed by Western blot for Nrf2 and INrf2 (A), INrf2 stabilizing proteins (B). (CD) cells were treated with DMSO or inhibitors of proteasome, lysed and analyzed for Nrf2, HO1, NQO1 and INrf2(C) and for Nrf2 and INrf2 (D). (E) Cells were treated with calpeptin (inhibitor of calpain), lysed and analyzed for INrf2 and calpain. (F) Cells treated with cycloheximide and harvested at indicated time points and analyzed for INrf2 for degradation assay. (G) INrf2 bands from (F) were quantified, normalized to actin, then ploted against CHX chase time.

Similarly, inhibition of calpain (calcium-dependent non-lysosomal cysteine protease) by calpeptin did not increase INrf2 protein in LTLT-Ca cells (Fig. 3-3E).

Next, a degradation assay was performed to evaluate the rate of INrf2 turnover using cycloheximide, inhibitor of protein translation and we found that the rate of INrf2 degradation in LTLT-Ca was not significantly different from that of MCF7-Ca cells

(Fig.3-3F & G). These data suggested that the rate of INrf2 degradation was comparable in both AI-sensitive and resistant cells, and the decrease in INrf2 level was independent of the proteasome-mediated protein degradation pathway.

Autophagy activity was high in LTLT-Ca cells

Recently, it has been reported that INrf2 is degraded by autophagy, which can be induced by ROS. One of the reasons for degradation of INrf2 could be due to oxidation of INrf2 in LTLT-Ca cells. We then quantified the ROS level by utilizing

DCFDA Cellular ROS detection assay kit and measuring the ROS-mediated fluorescence intensity and we found that the ROS level was higher in LTLT-Ca cells as compared to MCF-7Ca cells (Fig. 3-4A).

Figure 3-4. Basal level of autophagy was high in LTLT-Ca cells. A) Cellular reactive oxygen species (ROS)-mediated fluorescence in MCF-7Ca and LTLT-Ca cells. Cells were suspended in 2‟7‟-dichlorofluorescein diacetate (DCFDA) and incubated at 37°C for 30 minutes and 1x105 cells per well were incubated in 96 well plate at 37°C for 3h and cellular ROS-mediated fluorescence intensity was recorded using microplate reader. (B & D) Cells were lysed and analyzed by Western blot for Nrf2, INrf2 and Bcl-2, autophagy inhibitor, p62 and LC3 (LC3B), autophagy marker, autophagy protein beclin1 and autophagy substrate RARα. (C) Total RNA was extracted, cDNA was synthesized using the RNA as template and cDNA was utilized to quantify mRNA of Nrf2, INrf2 and Bcl2.

To explore the autophagy-mediated degradation of INrf2, we determined the expression levels of some important molecules associated with autophagy. For example, anti-apoptotic protein Bcl-2, an endogenous inhibitor of autophagy, was undetected at transcription as well as protein level in LTLT-Ca cells (Fig.3-4B- C).

Beclin-1, a very essential molecule for initiation of autophagy, is a Bcl-2 interacting protein. The expression levels of Beclin-1 were similar in both cell lines. The

57 decrease in protein levels of autophagy markers LC3, p62 and autophagy substrate protein (retinoic acid receptor α) RARα, which are degraded by autophagy, suggested that LTLT-Ca cells have high basal level of autophagy activity (Fig. 3-4D).

Autophagy inhibitors did not increase INrf2 in LTLT-Ca cells

To find the optimal conditions for inhibition of autophagy, we treated MCF-7Ca and

LTLT-Ca cells with 50 nM of bafilomycin A1 for different periods of time. We found that autophagy was inhibited at 12h but there was not much effect on INrf2 levels in either cell lines (Fig. 3-5A). However, LC3 levels increased with the duration of treatment. Since LTLT-Ca cells contain less INrf2, we concentrated our studies mainly in LTLT-Ca cells. We lowered the concentration of bafilomycin A1 to 1nM-

10nM but extended the duration of treatment. LC3-II levels increased in response to bafilomycin A1 but no increase in INrf2 levels was observed, in fact, INrf2 levels decreased when LTLT-Ca cells were treated with bafilomycin A1 for 72h (Fig. 3-5B).

Again, treating the LTLT-Ca cells with 2nM bafilomycin A1 for up to 24h did not change the protein levels of INrf2 (Fig.3-5C). mRNA of INrf2 was decreased at both

6h and 24h treatments consistent with the observation that INrf2 protein levels were not increased by bafilomycin A1 (Fig.3-5D). As treatment with bafilomycin A1 failed to increase INrf2 protein levels, we screened the potency of other autophagy inhibitors (cholroquine, 3-methyladenine and wortmannin) to evaluate their effects on the levels of INrf2. The treatment with those autophagy inhibitors also failed to increase INrf2 levels in LTLT-Ca cells (Fig.3-5E-G). However, INrf2 levels in MCF-

7Ca was slightly increased at 6h time point and not changed at 9h time point with chloroquine treatment. LC3-II levels did increase demonstrating that autophagy was inhibited. Neither combination treatment with autophagy inhibitors (bafilomycin A1

58 and wortmannin) nor proteasome inhibitor MG132 with chloroquine resulted in increased INrf2 levels (Fig.3-5H-I).

Figure 3-5. Inhibition of Autophagy did not increase INrf2. (A) Cells were treated with Bafilomycin A1, lysed and analyzed for INrf2 and LC3 (LC3B). (B) LTLT-Ca cells were treated with Bafilomycin A1, lysed and analyzed for INrf2, p62 and LC3. (C) LTLT-Ca cells were treated with Bafilomycin A1, lysed and analyzed for INrf2, p62 and LC3 by western blot, (D) Total RNA was extracted, cDNA was synthesized and quantified for mRNA of Nrf2 and INrf2 by qRT-PCR. (E) Cells were treated with chloroquine (CQ), lysed and analyzed for INrf2 and LC3. (F & G) LTLT-Ca cells were treated with 3-methyladenine (3-MA) (F) and with wortmannin (G), lysed and analyzed for INrf2 and LC3. Cells were treated with Bafilomycin A1 and wortmannin (H), chloroquine and MG132 (I), lysed and analyzed for INrf2 and LC3. (J) Cells were treated with autophagy inhibitors, total RNA was extracted, and cDNA was synthesized and quantified for mRNA of Nrf2 and INrf2.BW, bafilomycin A1 and wortmannin.

Evaluation of INrf2 mRNA in LTLT-Ca cells treated with inhibitors of autophagy showed that bafilomycin A1 and chloroquine decreased mRNA of INrf2 (Fig.3-5J).

That might be the reason for further downregulation of INrf2 upon treatment with chloroquine (Fig.3-5E) and longer exposure (72h) to bafilomycin A1 (Fig.3-5B)

Induction of autophagy decreased INrf2 protein

In chapter two, we noticed that INrf2 level was decreased on treatment with t-

BHQ. Therefore, we examined whether t-BHQ has any effect on autophagy in MCF-

7Ca cells, which express higher levels of INrf2. Treatment with t-BHQ decreased the levels of LC3-II, suggesting induction of autophagy. During induction of autophagy,

LC3-I is lipidated to produce LC3-II, which is degraded by autophagy, and LC3-I levels did not decrease as its expression is induced. Another autophagy marker p62 and also INrf2 levels were decreased upon t-BHQ treatment (Fig.3-6A).

Figure 3-6. Induction of autophagy decreased INrf2 protein. (A) MCF-7Ca cells were treated with antioxidant t-BHQ and analyzed for Nrf2, INrf2, and autophagy markers p62 and LC3 (LC3B). (B) Cells were treated with curcumin, lysed and analyzed for Nrf2, INrf2 and autophagy markers p62 and LC3.

We then examined the effect of another antioxidant curcumin on induction of autophagy. Consistent with our previous observation, curcumin also decreased the levels of INrf2 and autophagy marker LC3-II in both MCF-7Ca and LTLT-Ca cells

(Fig. 3-6B). As expected, Nrf2 protein was stabilized with the treatment of

60 antioxidants. The levels of p62, which is a target gene of Nrf2, slightly increased with

20h curcumin treatment.

Epigenetic modulators did not restore INrf2 in AI-resistant and sensitive cells

To understand how protein levels of INrf2 decreased in LTLT-Ca cells, we extended our studies by evaluating the mRNA level of INrf2. We showed that the mRNA levels of INrf2 were significantly decreased in LTLT-Ca cells as compared to

MCF-7Ca cells (Fig.3-7A). One of the possibilities of lower levels of INrf2 mRNA in

LTLT-Ca cells could be an epigenetic regulation. To examine the epigenetic regulation of INrf2, we first used decitabine (5-aza-2′-deoxycytidine), inhibitor of

DNA methyltransferases (DNMTi) to determine whether DNA methyltransferases was involved in the downregulation of INrf2 transcription. Cells were treated with different concentrations of DNMTi for different time points and protein levels of

INrf2 were analyzed by western blot. Treatment with DNMTi failed to increase INrf2, indicating a possible promoter hyper methylation may not be involved in downregulation of INrf2 (Fig.3-7B). Another epigenetic modifying agent, HDACi

(inhibitor of histone deacetylase) was utilized to explore the epigenetic regulation of

INrf2. First, cells were treated with the HDACi TSA (Trichostatin A) at different concentrations and different times. Western blot of histone acetylation confirmed the inhibition of HDAC by TSA but INrf2 protein level was unchanged (Fig.3-7C).

Accordingly, treatment with two other HDAC inhibitors, MS-275 and SAHA also did not increase the INrf2 levels (Fig. 3-7D). The HDACi (SAHA and MS-275), in fact, caused INrf2 to degrade in both AI-sensitive and resistant cells (Fig.3-7D & E).

Therefore, the fact that expression of INrf2 was not increased by way of inhibitors of

HDAC and DNMT, suggesting that epigenetic regulation may not be the cause of lower levels of INrf2 in LTLT-Ca cells.

Figure 3-7. Epigenetic modulators did not upregulate INrf2. (A) Total RNA was extracted; cDNA was synthesized and quantified for mRNA of INrf2. (B) Cells were treated with inhibitor of DNMT (decitabine), lysed and analyzed by Western blot for Nrf2 and INrf2. (C) Cells were treated with inhibitor of HDAC (TSA), lysed and analyzed for INrf2 and histone acetylation. (D & E) cells were treated with inhibitor of HDAC (5 µM of MS-275 and SAHA), lysed and analyzed for INrf2 and histone acetylation.

HDAC inhibitors reduced INrf2 mRNA in AI-resistant and sensitive cells

Using HDAC inhibitors, we did not observe an increase level of INrf2 in

MCF-7Ca and LTLT-Ca cells. We then asked if these HDAC inhibitors have any effect on transcription of INrf2. We examined the mRNA levels of INrf2 in MCF-7Ca and LTLT-Ca cells treated with HDAC inhibitors (TSA, MS-275 and SAHA). All three HDACi did decrease the mRNA of INrf2 in both cell lines (Fig.3-8A). In fact,

SAHA dramatically reduced mRNA levels of INrf2 regardless of cell types.

Figure 3-8. HDAC inhibitors lowered INrf2 mRNA in AI-sensitive and resistant cells. (A) Cells were treated with inhibitors of HDAC (Trichostatin A, MS-275 and SAHA, suberoylanilide hydroxamic acid) total RNA was extracted, and cDNA was synthesized and quantified for mRNA of INrf2. (B & C)) cells were treated with inhibitor of HDAC, SAHA for 24h, lysed and analyzed by Western blot (B) for Nrf2, INrf2, NQO1 and acetylation of histone 3 and total RNA was extracted, cDNA was synthesized and quantified for mRNA of Nrf2 and INrf2 by qRT-PCR (C).

Therefore, we treated both MCF-7Ca and LTLT-Ca cells with SAHA and determined mRNA and protein levels of INrf2. SAHA, indeed, decreased the protein and mRNA of INrf2 in MCF-7Ca and LTLT-Ca (Fig. 3-8B & C). Conversely, treatment with

SAHA did not decrease mRNA and protein level of Nrf2 in LTLT-Ca cells. Even though, Nrf2 mRNA level was lower in LTLT-Ca cells, protein level of Nrf2 was higher due to lower level of INrf2. NQO1mRNA was reduced by SAHA treatment in both cell lines but protein level of NQO1was not altered. Nrf2 mRNA was decreased

63 with SAHA treatment in MCF-7Ca cells but in LTLT-Ca cells the levels of Nrf2 mRNA was not decreased.

Withdrawal of letrozole increased mRNA of INrf2

Figure 3-9. Withdrawal of letrozole increased transcription of INrf2. (A) LTLT-Ca cells were grown in absence of letrozole for 12 weeks, then MCF-7Ca, LTLT-Ca+ and LTLT-Ca(-) were lysed and analyzed by western blot (A) for (HER-2, Nrf2, INrf2 and GCLC) and by qRT-PCR for Nrf2 and INrf2 (B).(C) MCF-7Ca cells were treated with vehicle(10% ethanol) or 1µM letrozole, then lysed and analyzed by western blot (C) for (Nrf2 and INrf2) and by qRT-PCR for INrf2. *p-value, data was compared between the levels of INrf2 in MCF-7Ca vs INrv2 in LTLT-Ca+ cells. **p- value was calculated from levels of INrf2 in LTLT-Ca+ cells vs INrf2 in LTLT-Ca- cells.

LTLT-Ca cells, which are continuously grown in presence of 1µM letrozole, have low levels of INrf2 compared to parental MCF-7Ca cells. We were curious whether withdrawal of letrozole from the medium can affect the expression of INrf2 in LTLT-

Ca cells. Interestingly, removal of letrozole did increase INrf2, decreased HER2, Nrf2 and its target gene GCLC (Fig.3-9A). Quantification of INrf2 mRNA also demonstrated the increase in mRNA level in LTLT-Ca cells growing in absence of letrozole. The results were consistent with the protein level of INrf2 (Fig.3-9A & B).

We further examined whether growing MCF-7Ca cells in the presence of letrozole can decrease the expression of INrf2. Although, mRNA of INrf2 decreased at 96h of letrozole treatment, the protein levels of INrf2 did not change (Fig.3-9C & D).

mRNA of INrf2 was less stable in LTLT-Ca cells

Figure 3-10. INrf2 mRNA was less stable in LTLT-Ca cells. (A) Cells were treated with transcription inhibitor actinomycin D (2µg/ml) for the indicated time points and total RNA was extracted, 1µg of total RNA was used to synthesize and cDNA was quantified for mRNA of INrf2 by qRT-PCR. The graphs represent the natural logarithm of the relative levels of the Nrf2 protein versus the actinomycin D chase time and the half-life of mRNA of INrf2 was determined in the linear part of the degradation curve using formula t1/2=ln2/k, where k is the slope of the degradation curve.

To understand why INrf2 transcripts were reduced in LTLT-Ca cells as compared to MCF-7Ca cells, we examined the rate of degradation of its mRNA in both cell lines treated with actinomycin D for various time points. The half-life of mRNA of INrf2 in LTLT-Ca cells (Fig.3-10A & B) was lower (2.1h) compared to

MCF-7Ca cells (3.0h).

Discussion:

INrf2, a cytosolic repressor of Nrf2, controls the nuclear accumulation of Nrf2 that activates the Nrf2 signaling. It has been shown that the homodimer of INrf2 binds to single molecules of Nrf2 (McMahon et al., 2006). Thus, alteration in expression of

INrf2 will have a direct impact on protein levels of Nrf2 and its target genes. In line with that notion, downregulation of INrf2 in AI-resistant LTLT-Ca cells increased

Nrf2 signaling (chapter two). The underlying mechanisms of downregulation of INrf2 in LTLT-Ca cells were unknown.

We quantified the ROS levels and found that letrozole-resistant LTLT-Ca cells generated significantly higher levels of ROS (Reactive oxygen species), as compared to AI-sensitive MCF-7Ca cells. ROS is known to oxidize proteins and oxidized proteins are subjected to degradation (Margaritopoulos et al., 2013). Therefore, we speculated that one of the reasons for degradation of INrf2 could be due to high ROS level in LTLT-Ca cells. Accumulating evidence demonstrated that ROS induces autophagy (Khan et al., 2011; Kongara and Karantza, 2012). This increase in ROS was noteworthy considering the fact that the increase in ROS can oxidize cysteine residues of INrf2; thereby ROS-mediated oxidized INrf2 can be degraded by autophagy.

Inhibition of the proteasome did not increase the INrf2 protein level. That ruled out the possibility of INrf2 degradation by proteasome. In fact, we did noticed that the levels of INrf2 and NQO1 (Nrf2 target gene) were decreased upon treatment with inhibitors of proteasome. INrf2 and NQO1 are relatively long-lived proteins and proteasome inhibitors do not stabilize those (Fig.3.3C & D). The degradation of INrf2

66 by autophagy has already been reported (Taguchi et al., 2012). The pathway by which

NQO1 is degraded remains unknown. We presumed that NQO1 may be degraded by autophagy. It is possible that when proteasome-dependent pathway of protein degradation is blocked, another pathway of protein degradation, autophagy might be activated. That explanation also supports the possibility of INrf2 degradation by autophagy.

Autophagy is an another mechanism of protein degradation and Bcl-2 is known to inhibit autophagy by interacting with Beclin-1, an essential molecule for the initiation of autophagy (Kongara and Karantza, 2012). In LTLT-Ca cells, Bcl-2 was not detected (Fig 2-4B, 3-4B & C), which further supported our hypothesis that autophagy flux is higher in AI-resistant cells than in AI-sensitive cells (Fig.3-4). We then compared the autophagy flux in MCF-7Ca and LTLT-Ca cells. Interestingly, we detected high autophagy flux in LTLT cells. That observation suggested that higher activity of autophagy may contribute, in part, to the downregulation of INrf2 in

LTLT-Ca cells. However, inhibition of autophagy by various chemicals (wortmannin, bafilomycin A1, 3-methyladenine and chloroquine) did not increase the protein levels of INrf2. We then tried to explore why INrf2 level was not increased upon inhibition of autophagy. Evaluation of mRNA by quantitative RT-PCR indicated that mRNA levels of INrf2 were decreased in LTLT-Ca cells treated with two autophagy inhibitors (bafilomycin A1 and chloroquine) suggesting that these inhibitors may promote the degradation of INrf2 mRNA by unknown mechanism.

Furthermore, we examined the regulation of INrf2 at the level of transcription.

Quantitative RT-PCR data showed that mRNA level of INrf2 was lowered in LTLT-

Ca cells as compared to MCF-7Ca cells. Inhibitors of DNMT and HDAC were utilized to investigate the epigenetic regulation of INrf2. DNMTi did not increase the levels of INrf2 in either MCF-7Ca or LTLT-Ca cells. We then employed inhibitors of

HDAC, hydroxamic acid analog TSA (trichostatin A), suberoylanilide hydroxamic acid (SAHA) and benzamides (MS-275) to study the epigenetic regulation of INrf2 and expected to upregulate the expression of INrf2 in LTLT cells. Treatment with

DNMT inhibitor, decitabine, and HDAC inhibitors (TSA, MS-275 and SAHA) failed to upregulate INrf2. While we were trying to understand why HDACi down regulated the INrf2 protein, we started to examine other possible pathways that would decrease the expression of INrf2. It has been reported that INrf2 is degraded by autophagy, which is also induced by inhibitor of HDAC, SAHA (Shao et al., 2004). We speculated that the reason for the degradation of INrf2 upon treatment with HDACi might be due to induction of autophagy. However, SAHA dramatically reduced the mRNA of INrf2.

Treatment with letrozole in MCF-7Ca cells did not decrease INrf2 protein but decreased its mRNA. This may be due to the discordance between mRNA and protein expression or relatively short time of letrozole treatment and longer half-life of INrf2 protein (Taguchi et al., 2012). It is worth mentioning that LTLT-Ca cells were generated by treating MCF-7Ca cells with letrozole for 56 weeks. However, the decrease in mRNA with letrozole treatment in MCF-7Ca cells also supports the possibility of letrozole-mediated generation of miRNA that may target INrf2 mRNA.

Letrozole-induced generation of miRNA could be another possible mechanism of

INrf2 downregulation in LTLT-Ca cells. Our studies also demonstrated that letrozole may regulate INrf2 in LTLT-Ca cells at the level of post transcription.

Lower half-life of INrf2 mRNA in LTLT-Ca cells indicated that the rate of mRNA degradation of INrf2 was faster in LTLT-Ca cells than in MCF-7Ca cells.

Recently, some micro-RNA have been shown to regulate INrf2 in breast cancer cells

(Eades et al., 2011). We presumed that letrozole may have an effect on generation of a miRNA that targets mRNA of INrf2 for degradation.

In conclusion, this study revealed that proteasome and autophagy pathways were not involved in the degradation of INrf2. Lower stability of INrf2 mRNA was one of the possible mechanisms for downregulation of INrf2 in AI-resistant LTLT-Ca cells. These results would open up a new direction for further study on regulation of

INrf2. Thus, understanding the multiple mechanisms of INrf2 regulation might provide a new therapeutic intervention to sensitize chemoresistance cancers to anti- cancer therapy.

CHAPTER FOUR

IDENTIFICATION OF COMPOUNDS INHIBITING NRF2

SIGNALING

Abstract:

Drug resistance is a major problem in chemotherapy. Resistance can be developed mainly due to upregulation of drug detoxifying enzymes, increased DNA damage repair system, and efflux of drug by overexpression of various transporters and most importantly resistance to apoptosis. Most chemotherapeutic agents are designed to target apoptosis. Since chemotherapy approaches have been extended in an attempt to treat cancers not responding to targeted therapy, finding a way to overcome drug resistance is becoming increasingly important. Research of the last two decades has elucidated that nuclear factor Nrf2 is a master regulator of cytoprotective genes that have a direct involvement in cellular protection against many kinds of environmental insults. Constitutive activation of Nrf2 pathways has been reported in many types of cancer originated from different organs (Homma et al., 2009; Zhang et al., 2010). It is also established that upregulation of Nrf2 signaling is associated with therapeutic resistance. In previous studies, genetic inhibition of

Nrf2 was shown to sensitize cancer cells to anti-cancer therapies (Homma et al., 2009;

Zhang et al., 2010). Thus, a search for Nrf2 inhibitors is highly warranted. Use of natural products is an alternative and promising strategy to deal with therapeutic resistance. In this study, we have used some natural compounds, which have already been identified as anticancer agents, to examine if they could inhibit the function of

Nrf2 in AI-resistant cells. A stable cell line in which luciferase was expressed will be useful to identify inhibitors of Nrf2 utilizing a high-throughput screening technique.

Introduction:

Extensive research on Nrf2 in the last two decades has allowed us to understand how Nrf2 protects normal cells from environmental insults. This stimulated the search to identify and develop potent Nrf2 activators as chemopreventive agents. On the other hand, many cancers express very high levels of Nrf2 due to various genetic or epigenetic alterations (Homma et al., 2009; Inami et al.,

2011). The studies have also revealed that upregulation of Nrf2 signaling is mostly associated with therapeutic resistance. In previous studies, genetically inhibited Nrf2 has been shown to sensitize cancer cells to anti-cancer therapies, suggesting that pharmacological manipulation of Nrf2 activity may be a useful approach to overcome the phenomenon of therapeutic resistance (Homma et al., 2009; Zhang et al., 2010).

Thus, a search for Nrf2 inhibitors is highly warranted.

Several classes of natural compounds have been demonstrated to have anticancer activity. Silibinin, a biologically active polyphenolic compound, has been reported to possess anticancer activity that induces apoptosis in various tumor cells

(Singh and Agarwal, 2004). A polyphenolic compound, honokiol, used in the traditional Japanese medicine has anti-tumor activity (Singh and Katiyar, 2013).

Likewise, curcumin is a well-known anti-cancer agent through multiple signaling pathways (Goel and Aggarwal, 2010). AKBA (acetyl-11-keto-beta-boswellic acid) is a pentacyclic triterpene natural product and treatment with AKBA demonstrated anti- cancer activity against human gastric carcinomas (Zhang et al., 2013). Furthermore,

ATRA (all-trans retinoic acid), one of the natural retinol derivatives, is known to modulate a wide range of biological events such as cell differentiation, proliferation, apoptosis, morphogenesis in vertebrates, vision, reproduction and immune responses

72 and exerts its biological functions through retinoic acid receptors (RARα, β and γ).

A number of natural compounds and endogenous molecules have been shown to inhibit Nrf2. For example, angiotensin II, a vasoconstrictor peptide, repressed Nrf2 signaling by enhancing the expression levels of ATF3 (activating transcription factor

3), one of the negative regulators of Nrf2 (Kang et al., 2011). Since estrogen-related receptor β (ERRβ) was shown to physically interact with Nrf2 in a complex that hinder binding of Nrf2 to ARE elements of Nrf2 target genes, ERRβ was reported as a potent inhibitor of Nrf2 (Zhou et al., 2007). Recently, a major alkaloid component of coffee and fenugreek, trigonelline has been shown to block the translocation of Nrf2 into the nucleus and suppress the transcriptional activity of Nrf2 (Arlt et al., 2012).

There are reports of some potential compounds that inhibit Nrf2. For example, procyanidins in cinnamomi cortex extract was shown to suppress Nrf2 and enhanced sensitivity of A549 cells to the cytotoxic action of anticancer drugs (Ohnuma et al.,

2011). Brusatol, 4-(2-Cyclohexylethoxy) aniline (IM3829) and Luteolin have also been identified as inhibitors of Nrf2 (Lee et al., 2012; Tang et al., 2011). Brusatol inhibits Nrf2 by decreasing its protein levels through enhanced ubiquitination and degradation whereas luteolin reduced Nrf2 at the levels of transcription and protein.

In summary, to identify novel compounds that specifically inhibit Nrf2 signaling, we examined silibinin, honokiol, AKBA (acetyl-keto-β-boswellic acid), curcumin and ATRA for their ability to inhibit Nrf2 signaling. A stable cells expressing luciferase under control of Nrf2 was generated to screen several natural and synthetic compounds for potential inhibitors of Nrf2.

Materials and Methods:

Chemicals and Reagents

The chemicals and antibodies were purchased as indicated. MEGM™

(Mammary Epithelial Cell Growth Medium, Cat. No. CC-3150 from Lonza,

Walkersville, MD), Hygromycin B (Cat. No.10687-010) from Invitrogen. Penicillin/ streptomycin (Cat. No. A5955), anti-β-Actin antibody (Cat. No. A5441) were from

Sigma (St. Louis, MO). RNeasy mini kit (Cat. No. 74104) was obtained from Qiagen

(Valencia, CA). High capacity cDNA reverse transcription kit (part No. 4368814), primers and probes used for real time-PCR were from Applied Biosystems (Foster city,CA). Puromycin dihydrochloride (sc-108071), control shRNA lentiviral particles-A (sc-108080), Keap1 (h) shRNA (sc-43878-V), anti-Nrf2 (sc-13032), anti-

Keap1 (sc-15246), anti-NQO1 (sc-32793) antibodies were from Santa Cruz

Biotechnology (Paso Robles, CA). L-glutamine (Gibco, Cat No. 25030-081) was from

Life technologies and human insulin (Cat # I9278) was purchased from Sigma. Pierce

Chemiluminescence western blotting substrate (Product No.3220) was from Thermo

Scientific (Rockford, IL). Complete protease inhibitor (Ref. No. 11 697 498 001) was obtained from Roche Applied Science. Dual luciferase system and luciferase reporter vector pGL4.27 (Part No. E845A) were obtained from Pomega Corporation

(Madison, WI).

Cells and cell culture conditions

Human breast cancer MCF-7 cells were obtained from ATCC and maintained in MEM containing 10% FBS, 1 % glutamine, 0.1% human insulin, 1%

Penicillin/streptomycin. Human mammary epithelial MCF10A cells were obtained

74 from Dr. Feyruz V. Rassool‟s laboratory and grown in MEGM™ supplemented with

5% FBS. MCF-7+INrf2 shRNA cells were grown in same medium as MCF-7 plus 10

µg/ml puromycin and MCF-7+INrf2 shRNA+Luc2 cells were cultured in the same medium with 50 µg/ml hygromycin B. All cell lines were cultured in monolayer in the medium with 1% Penicillin/streptomycin in an incubator at 37°C with 95% air and

5% CO2.

Cloning of human NQO1-ARE in luciferase reporter vector

Oligonucleotides containing three repeats of human NQO1 gene antioxidant response elements (ARE), TGACTCAGCA with intervening sequences and restriction enzyme sequences at the extreme ends were custom synthesized (IDT

Technologists) along with its complementary sequence, annealed at 720C, double digested with the restriction enzymes, purified and cloned between Xho1 and HindIII sites of multiple cloning sites (MCS) of pGL4.27 luciferase reporter plasmid

(Promega, NJ). Similarly, oligonucleotides containing three repeats of mutated ARE

(GTCAGACACA) were synthesized and cloned to Xho1 and Hind III sites of pGL4.27 vector. The plasmids were sequenced to confirm the correct orientation and sequence. The oligonucleotides sequences of the (+) strand of wild type ARE sequence was 5‟-ctcg ag ACAG TGA CTC AGC AGA ATA CAG TGA CTC AGC

AGA ATA CAG TGA CTC AGC AGA AT aagctt-3‟ and that of the (+) strand of mutant-ARE was 5‟-ctcg agA CAG GTC AGA CAC AGA ATA CAG GTC AGA

CAC AGA ATA CAG GTC AGA CAC AGA ATaagctt-3‟.

Generation of stable MCF-7cells expressing INrf2shRNA and luciferase

MCF-7 and MCF10A cells were transduced with lentiviral particles containing

INrf2 shRNA or control shRNA and cells stably expressing INrf2 shRNA or control shRNA were selected in presence of 10 µg/ml puromycin and designated as MCF-7-

INrf2 knock down (MCF-7+INrf2KD) and (MCF10A+INrf2KD) cells, respectively.

MCF-7-INrf2KD+Luc2 stable cells were generated by transfecting MCF-7-INrf2KD cells with luciferase constructs under control of human NQO1-ARE and luciferases expressing stable cells were selected using 50 µg/ml of hygromycin B.

Western blotting

Stock solution of silibinin, honokiol, AKBA, curcumin and ATRA was prepared in DMSO. Cells were treated with curcumin or ATRA or t-BHQ or DMSO

(vehicle control) as indicated in figures. Western blotting was performed as described in materials and methods of chapter two.

Luciferase Activity

For transiently transfection, cells plated in 12 well plates were transfected with

100 ng of luciferase plasmids of either ARE-hNQO1 or mutant ARE-hNQO1 plus 10 ng of TK-renilla plasmid. 24h after transfection, cells were harvested. Stable cells were treated with 50 µM of t-BHQ for 24 hours. Cells were lysed in 1x passive lysis buffer and luciferase activity was recorded using TECAN Infinite M1000 PRO plate reader. In the case of transient transfection, luciferase activity of firefly was normalized to activity of renilla luciferase, whereas in stable cells luciferase activity was plotted per µg protein. Promoter activity values were expressed as arbitrary units

76 after normalization with protein concentration. Experiments were done in triplicates and the standard deviation is indicated.

Densitometry and Statistical Analyses

Image J software (NIH) was used to quantify the intensity of proteins bands.

The protein bands were normalized against loading controls. Data from quantitative real-time PCR, luciferase activity and intensity of protein bands were analyzed using as a two-tailed student‟s test. Data were presented as the mean ± standard deviation.

Two data sets with p-value ≤ 0.05 were considered statistically significant.

Results:

Natural products did not demonstrate the inhibitory effect on Nrf2 signaling in AI-resistant cells.

We tested whether some natural compounds with anti-cancer activities have any inhibitory effect on AI-resistant cells. AI-sensitive MCF-7Ca and AI-resistant

LTLT-Ca cells were treated with silibinin, honokiol or AKBA and analyzed them for

Nrf2 and molecules involved in Nrf2 signaling. Interestingly, instead of the expected

Nrf2 degradation, we observed that treatment with any of these natural products degraded INrf2, the endogenous inhibitor of Nrf2 in concentration-dependent manner in both cell lines (Fig.4-1A & B). Moreover, Nrf2 proteins were further stabilized upon treatment with AKBA in MCF-7Ca cells (Fig.4-1A). Accordingly, MCF-7Ca and LTLT-Ca cells were treated with curcumin for 2h and 18h and analyzed for Nrf2,

INrf2 and Nrf2 target genes GCLC and HO-1. Contrary to our expectation, Nrf2 was stabilized with treatment of curcumin within 2h as well as at 18h time points. It is

77 remarkable that protein levels of INrf2 decreased upon treatment of curcumin in a time dependent manner (Figure 4-1C). Accordingly, the expression of Nrf2 target genes GCLC and HO-1 were induced.

Figure 4-1. Natural products did not demonstrate inhibitory effect on Nrf2 signaling in AI-resistant cells. (A and B) Cells were treated with different concentration of Silibinin, Honokiol and AKBA for 24h and lysed and analyzed by western blot for protein levels of Nrf2, its target gene NQO1 and INrf2 in MCF-7Ca cells (A) and in LTLT-Ca cells (B). (C and D) Cells were treated with curcumin for 18h and with ATRA for 4d and lysed and analyzed by Western blot for Nrf2, its target genes GCLC and HO-1 and INrf2 (C) and RARα proteins (D).

Previously published data demonstrated that ATRA repressed basal and inducible expression of Nrf2 target genes. That interesting inhibitory effect of ATRA on Nrf2 signaling prompted us to test its effect in AI-resistant breast cancer cells. ATRA treated cells were analyzed for expression of Nrf2, INrf2, GCLC and RARα.

Surprisingly, we did not observe any inhibited expression of Nrf2 target gene GCLC

(Figure 4-1D). Wang et al reported that ATRA antagonized the transcription activity of

Nrf2 by enhancing the interaction between RARα and Nrf2 that prevents Nrf2 from binding to the ARE (Wang et al., 2007b). Western blot analysis revealed that the protein levels of RARα were significantly lower in LTLT cells as compared to MCF-

7Ca cells (Figure 4-1D).

Generation of cell-based luciferase system for identification of Nrf2 inhibitor

Cells expressing high levels of Nrf2 were generated by knocking down of

INrf2. Breast cancer cells, MCF-7 and immortalized human mammary epithelial cells,

MCF10A were transduced with lentiviral particles containing control shRNA or INrf2 shRNA. Stable cells expressing INrf2 shRNA were selected and designated as MCF-7 cells with INrf2 knocked down (MCF-7+INrf2KD) and MCF10A cells with INrf2 knocked down (MCF10A+ INrf2KD). Similarly, MCF-7+VP and MCF10A+VP were selected as control cells expressing control shRNA. The short hairpin-mediated inhibition of INrf2 was confirmed by western and qRT-PCR (Figure 4-2A & B).

Since the transcription of INrf2 was knocked down to 75% and Nrf2 target genes

NQO1 was upregulated, only MCF-7+INrf2KD cells were used for further experiments. Human wild type NQO1-ARE or mutant NQO1-ARE were separately cloned in luciferase reporter vector (Figure 3-2C) and confirmed by DNA sequencing

(Figure 4-2D). MCF-7+INrf2KD or MCF-7+VP cells were transiently transfected

79 with those luciferase construct with wild type or mutant NQO1-ARE to validate the expression of luciferase with respect to Nrf2 protein levels (Figure 4-2D & E). As expected, MCF-7+INrf2KD cells with high levels of Nrf2 showed more luciferase activity as compared to the control cells. Cells expressing mutant-ARE did not increase luciferase activity regardless of protein levels of Nrf2.

Figure 4-2. Generation of a stable cell line expressing high levels of Nrf2 and luciferase. (A & B) Stable cell lines (MCF-7 and MCF10A) expressing INrf2 shRNA were analyzed by western blot (A) and quantitative real-timePCR (B) for determination of protein and mRNA fo Nrf2, INrf2 and NQO1. (C) Human wild type NQO1-ARE or mutant NQO1-ARE was cloned to luciferase reporter vector. (D) Expression of Nrf2, INrf2 and NQO1 proteins in puromycin dihydrochloride-selected stable cell lines. E) Luciferase activity in MCF-7+INrf2KD and control MCF-7+V cells transfected with wild type or mutant-ARE plasmid.

We further generated stable cell lines which express higher level of Nrf2 and luciferase. To this effect, MCF-7-INrf2KD+Luc2 stable cells were generated by transfecting MCF-7-INrf2KD cells with the wild type or mutant-ARE-luciferase construct and luciferase expressing stable cells were selected. Utilizing stable cell line

(MCF-7-INrf2KD+Luc2) expressing high level of Nrf2 and luciferase under control of Nrf2, we showed that the cells transfected with wild type ARE have higher luciferase activity in basal and in inducible conditions, whereas MCF-7-INrf2KD cells transfected with mutant-ARE-luciferase construct did not increase luciferase activity even in inducible condition by treating with t-BHQ (Figure 4-3A & B). These results confirmed that MCF-7+INrf2KD cells expressing ARE luciferase can be utilized for screening of natural or synthetic compounds for the identification of Nrf2 inhibitors.

Figure.4-3. Validation of luciferase expression in stable cells (MCF- 7+INrf2KD+Luc2) expressing high levels of Nrf2 and luciferase. (A)Basal and inducible levels of Nrf2 and its target gene NQO1, and (B) luciferase activity in stable cells expressing wild type or mutant-ARE.

Discussion:

Activation of Nrf2 signaling in cancer cells has limited the use of anti-cancer agents regardless of their potency against cancer. An increasing body of evidence has

81 suggested that inhibition of Nrf2 is essential for restoring the therapeutic effect of the drugs. Screening of natural products with anti-cancer properties is the first preferable approach for the identification of Nrf2 inhibitors in terms of their availability and cost for their production. Natural compounds, silibinin, Honokiol and AKBA have shown their anti-cancer effect. We were curious if they could repress Nrf2 signaling in AI- resistant LTLT-Ca cells. We treated MCF-7Ca and LTLT-Ca cells with those natural compounds. Unexpectedly, none of them were able to degrade or inhibit Nrf2 activity in LTLT-Ca cells. Stabilization of Nrf2 in AI-sensitive cells with lower levels of INrf2 was observed. Degradation of INrf2 by those compounds explains the stabilization of

Nrf2 in MCF-7Ca cells. This observation suggests that these natural products

(antioxidants) may activate a pathway by which INrf2 is degraded. It is possible that autophagy might be activated by these natural compounds.

Curcumin exerts its anti-cancer effect through multiple pathways. Induction of apoptosis by curcumin is one of these pathways that sensitize cancer cells to therapy.

Therefore, curcumin is also known as a chemosensitizer. We presumed that the chemosensitizing property of curcumin could be due to Nrf2 inhibition. However, as other antioxidants do, curcumin also stabilized Nrf2 in both AI-sensitive and resistant cells, indicating that inhibition of Nrf2 signaling may not be involved in curcumin- mediated sensitization of cancer cells. All-trans retinoic acid (ATRA) has been shown to repress basal and inducible expression of Nrf2 target genes (Wang et al., 2007b).

That interesting inhibitory effect of ATRA on Nrf2 signaling prompted us to test its effect in AI-resistant breast cancer cells. Treatment of ATRA did not show any effect on Nrf2 activity in LTLT-Ca cells. Antagonizing effect of ATRA on the expression of

Nrf2 target genes was the interaction between RARα and Nrf2 that prevented Nrf2

82 from binding to the ARE. As ATRA mediates the inhibition of Nrf2 activity through

RARa, we examined the expression of RARα. Downregulation of retinoic acid receptor α (RAR α) in LTLT cells may be the reason for not observing any inhibitory effect of ATRA on Nrf2 signaling.

Although development of natural product-based activators of Nrf2 as effective agents of chemoprotection is demanding, development of inhibitors of Nrf2 is also expected to help reduce the therapeutic resistance and improve therapeutic efficacy of current treatments. To search for the potential inhibitors of Nrf2, we generated a stable cell line expressing high levels of Nrf2 by genetically knocking down INrf2. We introduced ARE in the promoter of luciferase reporter gene and generated the MCF-

7+INrf2KD cells stably expressing luciferase. Those cells were given to the Center for Biomolecular Therapeutics (CBT) and are being used to screen thousands of compounds for the identification of Nrf2 inhibitors. The MCF-7+INrf2KD cells expressing ARE-luciferase will be a useful cell-based luciferase system for screening of several natural or synthetic compounds for the identification of Nrf2 inhibitors.

Some of the clinically approved drugs for cancer treatment may have an inhibitory effect on Nrf2 signaling cascade. Therefore, in addition to screening new compounds with unknown biological activity for identification of Nrf2 inhibitors, the luciferase expressing cells can also be used to screen known compounds that have already been used in the clinic or in clinical trials.

In summation, a stable cell with expression of luciferase reporter under control of human NQO1-ARE will be a valuable cell-based system to find Nrf2 inhibitors by

83 screening various compounds with known or unknown biological activity, so that we could use those Nrf2 inhibitors along with aromatase inhibitors to enhance their therapeutic effect. Upon identification of Nrf2 inhibitors, future studies will focus on mechanisms involved in inhibition of Nrf2 signaling.

CHAPTER FIVE

Background and Significance

LOW LEVELS OF NRF2 PROMOTE MIGRATION AND INVASION OF PROSTATE CANCER CELLS

Abstract:

The nuclear factor Nrf2 is known to play a critical role in cellular protection against oxidative stress and cellular transformation. However, unabated nuclear accumulation of Nrf2 is also known to reduce apoptosis, promote cancer cell survival and drug resistance. Mutations in INrf2 and Nrf2 leading to nuclear accumulation of

Nrf2 in many cancers including prostate and breast cancer are known. These also raise interesting questions regarding the role of Nrf2 in cancer metastasis and/or metastasis progression that remains elusive. In this study we have investigated the role of Nrf2 in initiation of metastasis in prostate cancer. We used human prostate epithelial cells

(RWPE1), LNCaP and its highly metastatic variant C4-B2 prostate cancer cell lines to test our hypothesis. The analysis revealed that highly metastatic C4-B2 cells expressed higher INrf2 and lower Nrf2 levels as compared to less metastatic LNCaP cells. We used control, INrf2 and Nrf2 shRNA to generate C4-B2, LNCaP and

RWPE1 cells with altered levels of INrf2 and Nrf2 to determine the role of Nrf2 in metastasis and metastasis progression using soft agar colony formation, proliferation, migration and invasion assays. INrf2-knocked down C4-B2 (C4-B2-+INrf2KD) cells expressing higher levels of Nrf2 showed fewer colonies in soft agar and proliferated faster, migrated and invaded less, as compared to control C4-B2-V cells. Similarly,

Nrf2-knocked down LNCaP (LNCaP-Nrf2KD) cells expressing reduced levels of

Nrf2 in less metastatic LNCaP cells demonstrated significantly higher number of

85 colonies in soft agar, proliferated faster and showed greater migration , as compared to control LNCaP-V cells. RWPE1 cells with low levels of Nrf2 (RWPE1+Nrf2KD) proliferated faster and migrated more as compared to cells with high levels of Nrf2.

However, they did not show tumorigenic properties in soft agar assays. RWPE1-

Nrf2KD cells demonstrated high migration and invasion. These results together suggest that lower Nrf2 levels are associated with higher proliferation, migration and invasion of the cells. The studies revealed that cells with lower levels of Nrf2 and its target gene NQO1 expressed high level of Cox-2 and MMP9. Cells expressing high levels of Cox-2 and MMP9 may facilitate the initial steps of metastasis invasion and migration of cells.

Introduction:

Prostate Cancer and Metastasis

Prostate cancer is one of the leading causes of death in men and every year alarming number of cases are reported (Siegel et al., 2013). Metastasis is a multifactor and multi-step complex process by which tumor cells spread from the organ of origin to distant parts of the body. In advanced stage of prostate cancer, tumor cells starts spreading to various parts of body such as lung, liver, skin, distant lymph nodes, brain and bone (Heidenreich et al., 2008). Regardless of improvements in diagnosis and recent advances in treatment options, most of the deaths related to cancer result from metastasis. It is important to understand the multiple pathways involved in metastasis that allow us to target them for therapeutic intervention. Despite intensive research efforts very little is known about underlying molecular mechanisms of prostate cancer metastasis. Some important steps in cancer metastasis are degradation of the extracellular matrix, epithelial to mesenchymal transitions, dissociation of tumor cells from the organ of origin, invasion into surrounding tissues and blood circulation, cell migration, anchorage independent growth, intravasation and colonization to form tumors in different sites (Talmadge and Fidler, 2010).

Epithelial-Mesenchymal Transition (EMT)

Substantial evidence supports the fact that cytokines, pro-angiogenic factors, growth factors, transcription factors (SLUG, SNAIL, TWIST) are highly expressed in metastatic cells and have been suggested to have a role in inducing EMT (Kannappan et al., 2012; Yang et al., 2004). Invasion of tumor cells requires degradation of extra cellular matrix (ECM). Proteases such as MMPs facilitate tumor invasion by

87 degrading components of ECM. Tumor cells also need to pass through basement membrane, a thin sheet of specialized form of extracellular matrix underlying the epithelial cells. The main component of the basement membrane is collagen type IV.

Matrix Metalloproteinase

Matrix metalloproteinases (MMPs), zinc-dependent endopeptidases, are known to degrade extracellular matrix (ECM) proteins. MMPs are upregulated while cells undergo EMT that result in disruption of cell to cell and cell to ECM interactions. Alteration in expression and activities of MMPs are important events in early stage of metastasis. Among other MMPs, type IV collagenases, MMP2

(gelatinase A) and MMP9 (gelatinase B) have been extensively implicated in cancer invasion (Egeblad and Werb, 2002; Martin and Matrisian, 2007). MMP-2 and 9 have been reported to have their role in tumor-promoting and suppressing activities in a substrate and context-dependent manner as they hydrolyze ECM and non-ECM proteins. MMP-9 has been shown to be upregulated and involved in migration and invasion of tumor cells.

Cell adhesion molecules

Adhesion between cells is necessary for the maintenance of epithelial cell layers. Cell adhesion molecule E-cadherin is expressed predominantly in epithelial cells. A tumor and invasion suppressor protein E-cadherin maintains intercellular adhesion (Berx and Van Roy, 2001; Pećina-Slaus, 2003). Downregulation or loss of

E-cadherin was reported in some cancers due to various factors including its transcription repression that was linked to invasion of tumor cells (Talmadge and

Fidler, 2010). Consequently, cells loose epithelial properties such as cell adhesion and 88 polarity. The transcriptional repressors Twist and Slug are known to regulate expression of several genes including E-cadherin and mediate epithelial to mesenchymal transition, which has been implicated in tumor metastasis (Batlle et al.,

2000; Yang et al., 2004). Importantly, loss of E-cadherin function is necessary, but not enough for EMT. N-cadherin is a member of the cadherin gene family and its expression correlates with invasion and motility. Thus, E-cadherin is a marker for epithelial cells whereas N-cadherin is a mesenchymal marker. Vimentin is expressed in mesenchymal cells. It is the major cytoskeletal component of mesenchymal cells and is used as a marker of mesenchymally-derived cells or cells undergoing EMT. β- catenin is a subunit of the cadherin protein complex that is necessary for the maintenance of epithelial cell layers. It has a dual role as a transcriptional co-activator in Wnt signaling pathway and as an adhesion molecule in the cadherin-mediated cell adhesion complex (Nelson and Nusse, 2004). Previous studies reported that β-catenin is frequently absent or downregulated in metastases (Han et al., 2006; Tang et al.,

2002). However, β-catenin was also reported as an essential survival factor for metastatic melanoma cells (Sinnberg et al., 2011).

Integrins, a family of cell adhesion receptors, are transmembrane glycol- proteins composed of α and β subunits and they regulate cell signaling, cellular shape, motility, and a variety of cellular functions required for tumor metastasis (Bradley et al., 2010). During prostate cancer metastasis, integrins β1, β3 and β6 were reported to be upregulated (Goel et al., 2008). Integrins also interact with other components of extracellular matrix and activate signaling pathways by integrin-activated signaling molecules including focal adhesion kinase (Fornaro et al., 2001). Focal adhesion kinase (FAK) is a cytoplasmic protein tyrosine kinase and plays a critical role in

89 regulating integrin-mediated cell functions such as cell adhesion, invasion and migration. Increased expression of FAK is directly linked to tumor metastasis (Slack et al., 2001). Upon loss of cell adhesion, the majority of cells undergo detachment- induced cell death due to insufficient cell-matrix interactions (Sakamoto and

Kyprianou, 2010). However, some tumor cells are able to survive, then invade locally and enter the blood stream.

NF-κB and Cyclooxygenase-2

NF-κB (nuclear factor kappa-light-chain-enhancer of activated B cells) is a transcription factor that controls expression of many genes involved in various cellular events. NF-κB is normally retained in the cytoplasm by its inhibitory partner

IκB and activated in cellular responses to various stimuli like lipopolysacchrides, cytokines, UV-irradiation and oxidative stress (Jeong et al., 2004). During oxidative stress, IKK phosphorylates IκB that causes the degradation of IκB. Consequently, free

NF-κB translocates to the nucleus and activates the transcription of downstream genes

(Gopalakrishnan and Tony Kong, 2008). NF-κB is known to bind to the cis-acting elements in the promoter of the pro-inflammatory enzyme cyclooxygenase-2 (Cox-2) and MMP-9 (Fang et al., 2013; Kundu et al., 2006; Wang et al., 2013). Recent findings demonstrated that Cox-2 also upregulates MMP-9 expression (Bu et al., 2011).

Cox-2, an inducible enzyme for synthesis of prostaglandins, is upregulated during inflammation. Prostaglandin PGE2 is one of the major prostaglandins produced during inflammation and acts through G-protein coupled membrane receptors, known as E-prostanoid (EP) receptors. EP1 receptors function as a metastasis suppressor whereas EP4 receptors promote metastasis in breast, lung, and colon cancer (Doherty et al., 2009; Kim et al., 2010; Ma et al., 2010).

Nrf2 signaling in Metastasis

Since the discovery of Nrf2 20 years ago, it has been the focus of intense research in its role in chemoprevention and drug resistance (He et al., 2008; Jaiswal,

2004; Moi et al., 1994). Nrf2 is induced under various stress conditions and is a master regulator of cytoprotective genes (Ma, 2013). Under basal conditions, endogenous inhibitor of Nrf2 (INrf2/Keap1) constitutively sequesters Nrf2 in the cytoplasm and prevents Nrf2 from inducing expression of its target genes (Lee et al.,

2007). Upon exposure to carcinogens, xenobiotics, and oxidative stress, key cysteine residues of INrf2 become modified and Nrf2 stabilizes, allowing the transcription factor to translocate into the nucleus (Ma and He, 2012). Nuclear Nrf2 heterodimerizes with small Maf proteins, binds to a core nucleotide sequence in the promoter region of its target genes termed the antioxidant response element (ARE), and transactivates the expression of cellular defense genes (Venugopal and Jaiswal,

1998). Activation of Nrf2 signaling in cancer cells provides them with increased cytoprotection and therefore, Nrf2 has been implicated in cell survival and drug resistance. However, various types of cancer like human head and neck squamous cell carcinoma, lung adenocarcinoma, ovarian serous adenocarcinoma and hepatocellular carcinoma express very low levels of Nrf2 (Frohlich et al., 2008). This suggests that a low level of Nrf2 has also a role in cancer progression. Brave et al. observed that the expression of Nrf2 and its downstream genes were gradually suppressed during the progression of prostate cancer in TRAMP mice (Transgenic adenocarcinoma of the mouse prostate), a model for prostate cancer (Barve et al., 2009; Greenberg et al.,

1995). More specifically, the decreased level of Nrf2 was also found in human metastatic prostate cancer (Frohlich et al., 2008). The authors further suggested that

Nrf2 and its downstream genes were suppressed in human prostate cancer, and

91 continued to decrease during progression to metastasis. Furthermore, using mouse lung carcinoma and Nrf2-deficient mice, Satoh et al. have performed bone marrow transplantation experiment to demonstrate the anti-metastatic activity of Nrf2 in lung metastasis(Satoh et al., 2010). It has also been reported that Nrf2-deficient cancer cell lines increased cell migration and plasticity (Rachakonda et al.), indicating that lower levels of Nrf2 may promote metastasis. These observations pointed out that Nrf2 has anti-metastatic activity. Though, low level of Nrf2 was previously reported in prostate primary tumors and metastasized tumors (Frohlich et al., 2008), no correlation to its functional role has been postulated. In this study, we examined the functional role of

Nrf2 on colony formation, proliferation, migration and invasion in prostate cancer cells. We also evaluated the expression of molecules involved in metastasis to establish a molecular mechanism mediated by Nrf2.

Materials and Methods:

Reagents and Chemicals The chemicals and antibodies were purchased from the various companies.

G418 sulfate (Cat.No.61-234-RG) was from Cellgro (Manassas,VA). Penicillin/ streptomycin (Cat. No. A5955) and β-Actin antibody (part No. 5316) were from

Sigma (St. Louis, MO). RNeasy kit (Cat. No. 74104) was from Qiagen (Valencia,

CA). High capacity cDNA reverse transcription kit (part No. 4368814), primers and probes (Nrf2; Hs00975960_ml, Keap1; Hs00202227_ml, NQO1; Hs00168547_ml,

GUSB; Hs99999908_ml) used for real time-PCR were obtained from Applied

Biosystems (Foster city,CA). matrigel (BD Falcon, Cat # 354234, 1:30 dilution).

Puromycin dihydrochloride (sc-108071), control shRNA lentiviral particles-A (sc-

108080), Nrf2 shRNA (sc-37030-V), Anti-Nrf2 (sc-13032), anti-Keap1 (sc-15246), anti-NQO1 (sc-32793) antibodies from Santa Cruz Biotechnology (Paso Robles, CA).

PageBlue protein staining solution (Cat.No. R0571) from Fermentas (Vilnius,

Lithuania). Amicon Ultra-0.5,Ultracel-10 membrane (10 kDa)centrifugal filters (Cat.

No. UFC501024) from Amicon (Ireland). Hematoxylin solution ,Gill No.2 (GHS216) from Sigma-Aldrich (St. Louis, MO).

Cell lines and generation of stable cells

RWPE1, a non-cancerous prostate epithelial cell line (immortalized by HPV),

LNCaP, a less metastatic prostate cancer cell line, and C4-B2, a highly metastatic

LNCaP-derived cell line were used as models. To generate stable LNCaP cells overexpressing INrf2, the cells were transfected with pcDNA of mINrf2 and selected in presence of 500 µg/ml G418. Using lentiviral particles expressing Nrf2shRNA or

93 control shRNA, LNCaP or RWPE1 cells were infected to generate LNCaP or RWPE1 cells stably expressing Nrf2shRNA (LNCaP-Nrf2KD or RWPE1-Nrf2KD) or control shRNA (LNCaP-V or RWPE1-V). Likewise, C4-B2 cells were infected with lentiviral particles expressing INrf2shRNA or control shRNA to generate C4-B2 cells stably expressing INrf2shRNA (C4-B2-INrf2KD) or control shRNA (C4-B2-V). All stable cell lines generated using virus particles were selected in presence of 10 µg/ml puromycin.dihydrochloride

Cell culture conditions

LNCaP cells were cultured in RPMI with 10% FBS and 1%

Penicillin/streptomycin. C4-B2 cells were grown in T-medium which contains 400 ml of DMEM (Invitrogen, Cat #11995) and 100 ml of Ham‟s F12 K (Cat #21127-022,

Invitrogen) and supplemented with 5 % FBS, 5 mg/ml Insulin, 13.65 pg/ml T3

(Griiodo-Thyronine), 5 mg/ml apo-transferrin, 0.244 mg/ml d-biotin and 25 mg/ml

Adenine hemi-sulfate (Sigma). Keratinocyte Serum Free Media Kit (Cat. No.17005-

042, Invitrogen) supplemented with 0.05 mg/ml BPE (Bovine pituitary extract) and 5 ng/ ml EGF1-53 (Epidermal growth factor 1-53) was used to grow RWPE1 cells.

RWPE1, LNCaP and C4-B2 cells were grown in monolayer. All cells used in this study were incubated at 37°C with 95% air and 5% CO2.

Western blotting

Western blot was performed as described in material and methods of chapter two.

Total RNA extraction, cDNA synthesis, and qRT-PCR

mRNA was quantified as described in material and methods (gene expression analysis) of chapter two.

Soft agar colony formation assay

Anchorage-independent growth is a stringent in vitro assay to evaluate the ability of tumor cells to form colonies after detaching from primary tumor. The assay was performed following a common protocol. Briefly, cells were trypsinized and cell suspension with various numbers of cells (0.1-2 x 104) in 0.35 % soft agar was added on top of 0.7 % agar in 24 well plates. The plate was incubated overnight at 37°C with

5% CO2 and 95% air. The following day, 200µl of normal media was added to each well. After 3 weeks, 200µl of INT (0.1% iodonitrotetrazolium chloride in PBS) solution was added to each well and incubated further for 16h to visualize cell colonies. Colonies were photographed using Bio-rad chemidoc and numbers of colonies were counted under microscope.

Boyden (BD) Migration and Invasion Assays

Migration and invasion assays were performed in a 24-well Boyden chamber following the protocol recommended by manufacturer. The Boyden chamber has an upper chamber (insert) and bottom chamber separated by polycarbonated membrane of 8µm pore. For invasion assay, the insert membrane was coated with matrigel at a concentration of 30µg/insert and uncoated inserts were used for migration assay.

Briefly, cells were trypsinized and 2.5x104 cells in 250 µl of serum-free medium were plated onto upper chamber. The bottom chamber contained 300 µl of complete

95 medium with 10% FBS as a chemoattractant. The cells were then incubated for 36h.

The top chambers were removed and fixed in 4% paraformaldehyde for 15 minutes at room temperature. Cells were washed with PBS twice and stained with hematoxylin for 20 minutes at room temperature. Non-migratory and non-invasive cells in upper surface of the insert were removed with a cotton swab and washed with water 3 times.

Migrated or invaded cells were photographed and counted manually from randomly selected fields under microscope and quantified.

xCELLigence System

xCELLigence system is a Real Time Cell Analyzer (RTCA) used for monitoring biological process of living cells. The principle of this system is based on measurement of electric impedance of cells present in a well of e-plate, which is interconnected with other wells by gold microelectrodes (Ke et al., 2011). Electric impedance of cells is expressed as cell index, which is directly proportional to viable cell numbers. Cell proliferation, migration and invasion assays were performed following the protocol of xCELLigence System. Briefly, cells were incubated in medium with 0.5% FBS for 16 hours. The cells were then trypsinized and resuspended in 5 ml of the serum free medium. 100 µl cells suspension was added to each well (1x104 cells for proliferation and 4x104 cells for migration and invasion) of

CIM-Plate 16. The cells were incubated at room temperature for 30 minutes and transferred to RTCA DP station (xCELLigence Instrument), which is placed inside a tissue culture incubator and cell indices were recorded at intervals of 15 minutes for

24 to 96 hours. Data were analyzed by using RTCA SP software.

Gelatin Zymography Assay

Activity of matrix metalloproteinase-9 (MMP-9) was assayed by zymography gel on the basis of ability of gelatinases (MMP-9) to hydrolyze the denatured collagen

IV (gelatin). Cells (1x106) were seeded on 60mm plates and the medium was replaced with 3 ml of serum-free medium and incubated for 48h. The secreted proteins in the medium were concentrated using Amicon/Millipore Ultracel 10K (Cat. No. UFC

501024) cutoff column and 25ug of the protein of the concentrated medium was loaded onto 8% gels co-polymerized with 0.1% gelatin, without adding 2- mercaptoethanol and heating. The gel was washed with water twice for 5 minutes and incubated in 1x renaturing buffer at room temperature for 1h to remove SDS. After washing with water twice for 5 minutes, the gel was incubated with 1x developing buffer at 37° C for 48h, followed by staining with PageBlue protein staining solution at room temperature for 2h. Gels were washed with water and the gel was scanned and bands were quantified.

Luciferase Activity

LNCaP-Nrf2KD and C4-B2-INrf2KD cells were transfected with pGL4.51

(Promega) luciferase reporter vector and stable cell lines that constitutively express luciferase were selected with 500 µg/ml G418. The constitutive expression of luciferase activity was confirmed using TECAN Infinite M1000 PRO plate reader.

siRNA Transfection

Opti-mem medium (1ml), 20µl of RNAimax (Invitorgen) and 75 and 150 nM siRNA were mixed in an eppendorf tube and vortex for 10 seconds and incubated the tube for 15 minutes at room temperature. The resulting mixture was uniformly spread 97 over on 60mm plates and incubated for 15 minutes at room temperature. Cell suspension (2x106 cells) in 3 ml medium without antibiotic was added to the pre- coated plate with siRNA mixture and harvested the cells 24h after transfection for knock down of Nrf2.

Densitometry and Statistical Analyses

Data from colonies in soft agar, migration and invasion assay and real time

PCR were analyzed and tested for significance using a two-tailed student‟s t-test.

Data were presented as the mean ± standard deviation. Two data sets with p-value ≤

0.05 were considered as statistically significant.

Results:

Highly metastatic cells had lower levels of Nrf2 and promoted colony formation

Figure 5-1. Flow chart showing generation of bone metastasized C4-B2 cells.

To study the role of Nrf2 in metastasis, we wanted to use cell lines with low level of Nrf2. Therefore, we initially screened several prostate cancer cell lines

(DU145, C81, LNCaP, C4-B2, C4, VCaP-1, PC3 and NCIH660) for determination of the levels of Nrf2 (Fig.5-2A). Of these LNCaP and VCaP are androgen dependent and

DU145, C81, C4-B2, C4, and PC3 were androgen independent cell lines (Saraon et al., 2012). DU145 cells express high level of Nrf2. Cell lines C81, C4 and PC-3 were relatively slow growing cells. NQO1 was not detected in VCaP cells and NCIH660 cell line was suspension cells. Therefore, on the basis of metastatic ability (Fig. 5-1) and Nrf2 levels (Fig5-2B), we selected LNCaP and C4-B2 for further studies. LNCaP cells have more Nrf2 but less metastatic. C4-B2 cells have less Nrf2 but are highly metastatic, even though, C4-B2 cells are derived from LNCaP. Initially we examined the ability of these cell lines to form colonies in soft agar. C4-B2 cells formed more colonies in soft agar (Fig.5-2C & D).

Figure 5-2. C4-B2 cells formed more colonies in soft agar. (A) Low levels of Nrf2 were detected in highly metastatic cells. Cells were treated with DMSO or t-BHQ and total cell lysates were used to determine the levels of Nrf2 and INrf2. (B) Comparison of colony formation in soft agar. Cells were seeded in agar and grown for 21 days and stained with iodonitrotetrazolium chloride for 16h and colonies were counted under microscope. (C) Levels of Nrf2 in LNCaP cells overexpressing INrf2. (D) Anchorage independent cell growth of the LNCaP+INrf2-V5 cells with reduced levels of Nrf2 and comparative graph of colonies of LNCaP and LNCaP+INrf2-V5 cells.

Figure 5-3. LNCaP cells expressing lower levels of Nrf2 formed more colonies in soft agar. (A) Levels of Nrf2, INrf2 and NQO1 in LNCaP cells overexpressing INrf2 tagged with V5. (B) Anchorage independent cell growth of the LNCaP+INrf2-V5 cells with reduced levels of Nrf2. (C) Comparative graph of colonies of LNCaP and LNCaP+INrf2-V5 cells.

To lower the Nrf2 levels in LNCaP cells and to be in parallel with the levels of

Nrf2 in C4-B2 cells, LNCaP cells stably expressing INrf2 (LNCaP+INrf2-V5) was generated (Fig.5-3A). Overexpression of INrf2 in LNCaP cells decreased Nrf2 and

NQO1 proteins (Fig.5-3A). We then compared the ability of LNCaP and

LNCaP+INrf2-V5 cells to form colonies in soft agar. LNCaP+INrf2-V5 cells expressing lower levels of Nrf2 formed more colonies in soft agar (Fig.5-3B & C).

Collectively, these results suggested that low levels of Nrf2 led to increase in anchorage independent cell growth.

Lowering Nrf2 level in less metastatic cells promoted anchorage-independent cell growth To evaluate the role of Nrf2 signaling in initiation and progression of metastasis, we knocked down Nrf2 in LNCaP cells using lentiviral particles expressing short hairpin RNA of Nrf2 and generated a stable cell line with lower Nrf2 expression (LNCaP+Nrf2KD). They were then analyzed for the expression of Nrf2,

INrf2 and NQO1 at the level of transcription and protein (Fig.5-4A & B). Both

100 mRNA and protein levels of Nrf2 and NQO1 were decreased when Nrf2 was knocked down. But INrf2 mRNA and protein levels were increased. We analyzed these cell lines for their ability to form colonies in soft agar and found that LNCaP+Nrf2KD cells showed more colonies in soft agar as compared to control cells (Fig.5-4C & D).

Figure 5-4. Nrf2-knocked down LNCaP cells showed more colonies in soft agar. (A & B) cells were lysed and analyzed for Nrf2, INrf2 and NQO1 by western blot (A) and total RNA was extracted, cDNA was synthesized and quantified for mRNA of Nrf2, INrf2 and NQO1 by qRT-PCR (B). (C) Anchorage independent cell growth of the LNCaP+Nrf2KD cells. D) Comparative graph of colonies in control and LNCaP+Nrf2-KD cells.

LNCaP cells with low level of Nrf2 proliferated faster, migrated and invaded more as compared to control cells

We utilized xCELLigence system to test the functional role of lower level of

Nrf2 by examining proliferation and migration properties of LNCaP+Nrf2KD cells.

The assays demonstrated that LNCaP-Nrf2KD cells with reduced levels of Nrf2 in less metastatic LNCaP cells proliferated faster and increased migration as compared to LNCaP+VP cells (Fig.5-5A & B). Invasion and migration properties of these genetically altered LNCaP cells were analyzed in Boyden chambers coated with or

101 without matrigel. The results revealed that there was a greater degree of invasion and migration in LNCaP cells expressing lower levels of Nrf2 (Fig.5-5C-F).

Figure 5-5. LNCaP+Nrf2KD cells proliferated faster, migrated and invaded more as compared to control cells. xCELLigence System; (A) Proliferation Assay, LNCaP+VP and LNCaP+Nrf2-KD (1x104) cells were seeded per well and cell indices were measured up to 96h. (B) Migration assay, 4x104 cells were seeded per well and cell indices were recorded up to 96h.(C) Protein levels of Nrf2, INrf2 and NQO1. (D) Migration (E) Invasion assays were performed in a 24-well Boyden chamber. 250 µl of cell suspension (2.5x104 cells) in serum-free medium were plated on upper chamber (insert) and the bottom chamber contained 300µl of complete medium as a chemoattractant and incubated for 36h. The assay was stopped by fixing cells with 4% paraformaldehyde for15minutes at room temperature. Cells in upper chamber of the insert were removed with a cotton swab. (F)Migrated or invaded cells were quantified by counting them manually from randomly selected fields under microscope and photographed. For invasion assay, the insert membrane was coated with matrigel.

Expressing higher levels of Nrf2 in highly metastatic C4-B2 cells lowered their ability to form colonies

Accordingly, we demonstrated that knocking down INrf2 in C4-B2 cells increased the protein levels of Nrf2 and NQO1 (Fig.5-6A). Simultaneously, mRNA levels of Nrf2, INrf2 and NQO1 were quantified (Fig.5-6B). Though, there is no increase in level of Nrf2 transcripts, the levels of NQO1 mRNA and protein were increased in C4-B2+INrf2KD cells. As expected, INrf2 mRNA decreased in C4-

102

B2+INrf2KD cells (Fig.5-6B). Colony formation assay showed fewer colonies of C4-

B2+INrf2KD cells in soft agar (Fig.5-6C & D).

Figure 5-6. C4-B2-INrf2-KD cells with higher levels of Nrf2 showed fewer colonies in soft agar. (A & B) Protein and transcription level of Nrf2, INrf2 and NQO. (C) Anchorage independent cell growth of the C4-B2+INrf2KD cells. (D) Comparative graph of colonies in C4-B2+VP and C4-B2+INrf2KD cells in bottom right.

C4-B2 cells with higher levels of Nrf2 demonstrated higher proliferation, lower migration and invasion as compared to control cells

We further examined the proliferation and migration properties of C4-B2-

INrf2KD cells using xCELLigence system. C4-B2-INrf2KD cells with high levels of

Nrf2 proliferated faster but did not migrate as efficiently as control cells (Fig.5-7A &

B). However, Boyden migration and invasion assay demonstrated that C4-

B2+INrf2KD cells showed lesser migration and invasion as compared to C4-B2+VP cells (Fig.5-7C-F). Together, these results suggested that high levels of Nrf2 may inhibit cell migration and invasion.

103

Figure 5-7. C4-B2-INrf2-KD cells proliferated faster but showed less degree of migration and invasion. xCELLigence System; (A) Proliferation Assay, C4-B2+VP and C4-B2+INrf2-KD(1x104) cells were seeded per well and (B) Migration assay, 4x104 cells were seeded per well and cell indices were recorded up to 48h for proliferation, 72h for migration, respectively. (C) Protein levels of Nrf2, INrf2 and NQO1. (D) Migration (E) Invasion assays were performed in a 24-well Boyden chamber. 250µl of cell suspension (2.5x104 cells) in serum-free medium were plated on upper chamber (insert) and the bottom chamber contained 300µl of complete medium as a chemoattractant and incubated for 24h. The assay was stopped by fixing cells with 4% paraformaldehyde for 15 minutes at room temperature. Cells in upper chamber of the insert were removed with a cotton swab. Migrated or invaded cells were quantified by counting them manually from randomly selected fields under microscope and photographed. For invasion assay, the insert membrane was coated with matrigel.

Non-cancerous prostate epithelial cells (RWPE1) with low levels of Nrf2 did not form colonies in soft agar

We were curious how non-cancerous prostate epithelial cells (RWPE1) behave if we knocked down Nrf2. Knocking down Nrf2 by shRNA decreased the levels of

Nrf2 and NQO1 (Fig.5-8A). In parallel, the levels of mRNA of Nrf2 and NQO1 were also decreased (Fig.5-8B). However, knocking down Nrf2 in RWPE1 did not increase the ability to form more colonies in soft agar (Fig.5-8C & D).

104

Figure 5-8. Prostate epithelial cells (RWPE1) with low levels of Nrf2 did not form colonies in soft agar. (A & B) cells were lysed and analyzed for Nrf2, INrf2 and NQO1 by western blot (A) and total RNA was extracted, cDNA was synthesized and quantified for mRNA of Nrf2, INrf2 and NQO1 by qRT-PCR (B). (C) Anchorage independent cell growth and (D) comparative graph of colonies in control and RWPE1+Nrf2KD cells.

RWPE1-Nrf2KD cells showed a high degree of proliferation, migration and invasion.

xCELLigence system showed that RWPE1 cells with low levels of Nrf2 can proliferate faster and migrate more as compared to cells with high levels of Nrf2

(Fig.5-9A & B). Similarly, we evaluated migration and invasion properties of

RWPE1-Nrf2KD cells using BD migration and invasion method. RWPE1-Nrf2KD cells demonstrated higher migration and invasion (Fig.5-9C-E), which were consistent with that of xCELLigence assay. These results further supported our hypothesis that lower level of Nrf2 is associated with migration and invasive properties of cells regardless of cell types.

105

Figure 5-9. RWPE1+Nrf2KD cells proliferated faster, migrated and invaded more as compared to control cells. xCELLigence System;(A) for proliferation assay, 1x104 cells and (B) for migration, 4x104 cells were seeded per well and cell indices were recorded 60h for proliferation,50h for migration, respectively.(C) Migration (D) Invasion assays were performed in a 24 well Boyden chamber. 250 µl cell suspensions (2.5x104 cells) in serum-free medium was plated on upper chamber (insert) and the bottom chamber contained 300 µl complete medium as a chemoattractant and incubated for 24h. The assay was stopped by fixing cells with 4% paraformaldehyde for 15 minutes at room temperature. Cells in upper chamber were removed with a cotton swab. Migrated or invaded cells were photographed and counted manually from randomly selected fields under microscope and quantified (E). For invasion assay, the membrane in the insert was coated with matrigel.

Low level of Nrf2 increased enzymatic activity of matrix metalloproteinase 9

To substantiate the functional assays, we examined the expression of proteins involved in initiation of metastasis. In this study we knocked down Nrf2 in LNCaP cells which express more Nrf2 and INrf2 was knocked down in C4-B2 cells which express more INrf2. The logic of this experiment was to examine independently how the other proteins involved in metastasis were altered with this manipulation.

Enzymatic activity of MMP9 (gelatinase) was examined in the secreted medium by zymography assay. Cells were grown in serum-free conditioned medium which was concentrated by centrifugation and electrophoresed under non-reducing conditions to

106 maintain the enzymatic activity. Enzymatic activity of MMP9 was higher in LNCaP-

Nrf2KD cells compared to LNCaP+VP cells. Interestingly, in C4-B2-INrf2KD cells, low activity of MMP-9 was observed as compared to the control C4-B2 cells

(Fig.10A). When we examined the metastatic related molecules such as Cox-2, EP1,

EP4 and Smad3, we were not able to detect any dramatic changes in the expression of

EP1 in the Nrf2 and INrf2 knocked down cells compared to their parental cells. There were 50% reduction in EP4 expression and two fold increase in expression of Cox-2 in LNCaP+Nrf2KD cells, however, cox-2 expression was decreased in C4-B2 cells, which have a high level of Nrf2 (Fig 10B). These results indicated that the Nrf2 level was inversely proportional to the expression of Cox-2 and MMP-9 activity.

Figure 5-10. Low levels of Nrf2 resulted in higher expression of Cox-2 and higher enzymatic activity of MMP-9. (A-C) Cells were lysed and analyzed for Nrf2-related proteins INrf2, NQO1 and CEBPα, molecules involved in inflammation, NF-κB, Cox- 2, EP-1 and EP-4, transcription factor (repressor), smad3 and slug, metastasis-related molecules integrin α5, vimentin, E and N-cadherin, β-catenin, Pyk2 and FAK by western blot.

We also determined levels of cell adhesion molecules such as N-Cadherin, E-

Cadherin, β-catenin and vimentin (Fig.10C). The expression of E-Cadherin was not changed in both parental and knocked down cells and N-cadherin and vimentin were

107 not detected in these cells. The levels of Smad3 were increased, E-cadherin was slightly increased and β-catenin was decreased in C4-B2+INrf2KD cells with high levels of Nrf2 as compared to parental cell line. To examine the effect of different levels of Nrf2 on the expression of Cox-2 and NF-кB, we also transiently overexpressed Nrf2, and knocked down Nrf2 using siRNA in LNCaP cells and analyzed them for Nrf2, Cox-2 and NF-кB. However, we did not observe any effect of different levels of Nrf2 on the expression of Cox-2 and NF-кB, implying that Nrf2 may not have a direct and short-term effect on Cox-2 expression (data not shown).

Nrf2 mRNA in metastasized breast tumors

To examine the significance of our findings from cell models, we wanted to analyze prostate tumor cells metastasized to distant organs for the expression of Nrf2 and we were not able to get those samples, but we were able to obtain cDNA prepared from mRNA that were collected from human breast tumors that had metastasized to distant organs such as gall bladder, liver, peritoneum, lung and bone. Using qRT-

PCR, we analyzed the cDNA for the quantification of Nrf2 mRNA. On the average, there was tenfold decrease in Nrf2 mRNA levels in all the metastasized tumor samples analyzed as compared to normal breast tissue (organoids 163, 266 and org

267, Fig.5-11A-C). These organoids and metastasized tumor samples were not from same patient. In future we need to compare primary tumor with metastasized tumor for the expression of Nrf2. Consistent with our results from cell models, excitingly our studies demonstrated that mRNA levels of Nrf2 in metastatic tumor samples were significantly lower as compared to normal breast tissues.

108

Figure 5-11. Low levels of Nrf2 mRNA in metastasized breast tumors (Met-samples) as compared with that in organoid (Org-sample). 200 ng of cDNA obtained from Dr. Sarah Sarswati‟s laboratory (John Hopkins University) was subjected to qRT-PCR for quantification of Nrf2 mRNA by taqman gene expression analysis using primers and probes from Applied Biosystems. Org-163, 266 and 267 are mRNA from normal breast tissues as control. Each Mets-sample was from different metastasized breast tumors collected from gall bladder, liver, peritoneum, lung and bone. Initial number of Met-sample indicates individual patient. There were no match organoid-primary- met samples.

109

Discussion:

Until recently, the impact of repressed expression of Nrf2 on cancer progression has not gained much attention. Low levels of Nrf2 have been reported in some metastatic prostate cancer cells, warranting more research to understand the mechanism and impact of reduced levels of Nrf2 in prostate cancer metastasis. A few reports on the role of Nrf2 suggested an inverse correlation between the level of Nrf2 protein and metastasis (Satoh et al., 2010). However, it is not clear whether metastasis causes the downregulation of Nrf2 or downregulation of Nrf2 initiates metastasis.

Importantly, how the reduced level of Nrf2 contributes to the initiation or progression of prostate cancer metastasis has not been studied previously.

Several prostate cancer cell lines were initially screened to identify levels of

Nrf2 and INrf2. We had selected LNCaP and its highly metastatic variant C4-B2 for our studies to test the hypothesis that reduced levels of Nrf2 initiate tumor metastasis.

Interestingly, we detected lower level of Nrf2 in highly metastatic C4-B2 cells as compared to less metastatic LNCaP cells. Highly metastatic C4-B2 cells with low

Nrf2 formed more colonies in soft agar. The colony formation assay basically evaluates the anchorage independent cell growth, which is considered as a surrogate phenotype of metastasis (Mori et al., 2009). Using this as a model, we examined whether less-metastatic cells (LNCaP) could be transformed into metastatic phenotype by decreasing the Nrf2 levels and whether highly metastatic cells C4-B2 could be changed into less metastatic phenotype by increasing Nrf2. A stable cell line

(LNCaP+INrf2-V5) with lower level of Nrf2 due to overexpression of INrf2 promoted anchorage-independent cell growth. Similarly, LNCaP cells expressing lower levels of Nrf2 (LNCaP+Nrf2KD) also demonstrated more colonies in soft agar

110 as compared to the control cells, indicating that the reduced level of Nrf2 may contribute to anchorage independent cell growth. Moreover, LNCaP+Nrf2KD cells proliferated faster and migrated and invaded more as compared to the control cells. In contrast to the established role of Nrf2, how lower levels of Nrf2 can result in high proliferation is not clear at this time (Homma et al., 2009; Ohta et al., 2008). The results also demonstrate that increasing the levels of Nrf2 in highly metastatic C4-B2 cells by knocking down INrf2 (C4-B2+INrf2KD cells) decreased the formation of colonies in soft agar; though proliferated faster but did not migrate and invade as compared to the vector transduced control cells (C4-B2+VP). High levels of Nrf2 have also been shown to increase cell proliferation (Ohta et al., 2008; Shelton and

Jaiswal, 2012). At this point we do not know why C4-B2+INrf2KD cells, which have more Nrf2, did not form more colonies as compared to C4-B2 cells with low level of

Nrf2. These results suggest that lower levels of Nrf2 may be associated with more anchorage independent cell growth, higher proliferation, enhanced migration and invasion.

We then compared these observations with normal non-cancerous prostate epithelial cells. RWPE1 cells with low levels of Nrf2 proliferated faster and migrated more, however, did not form more colonies as compared to cells with high levels of

Nrf2, suggesting that even in normal cells, low levels of Nrf2 is related to faster proliferation and more migration of cells. These results further supported our hypothesis that lower levels of Nrf2 are associated with enhanced migration and invasive properties of prostate cancer cells regardless of cell type. Irrespective of Nrf2 levels, RWPE1+Nrf2KD prostate epithelial cells behaved differently in colony formation as compared to cancer cells, suggesting that there may be other pathways

111 involved in anchorage independent cell growth in addition to Nrf2 signaling that need to be investigated in the future.

The more we understand the molecular mechanisms involved in each step of metastasis, the better we would be able to develop an effective therapeutic approach for prevention and reduction of prostate cancer metastasis. Previously, a protective role of Nrf2 signaling against inflammation-associated tumorigenesis was reported

(Osburn and Kensler, 2008). Moreover, some Nrf2 activators inhibited NF-κB and showed potent anti-inflammatory property (Dinkova-Kostova et al., 2005b; Ma et al.,

2003). Loss of Nrf2 results in production of ROS that increases oxidative stress and inflammation pathway, which damage proteins and nucleic acids inside cells as well as in extracellular matrix compartment (Dinkova-Kostova et al., 2005b). Importantly, inhibition of Nrf2 may also lead to activation of NF-κB signaling that increases inflammation pathway. Chronic inflammation may facilitate the creation of a metastatic microenvironment. However, the levels of NF-κB in our cell models were not varied with different levels of Nrf2 (data not shown).

Figure 5-12. Flow chart showing how low level of Nrf2 facilitates the invasion of tumor cells. (A) Model exhibiting invasion of tumor cells. (B) Low level of Nrf2 resulted in lower level of NQO1 and high level of Cox-2. ROS can be accumulated in Nrf2-deficient cells. ROS and Cox-2 are known to activate MMP9, which is involved in tumor invasion by degrading extracellular matrix.

112

Findings of the present study substantiated that low expression of Nrf2 resulted in reduction of NQO1 in LNCaP-Nrf2KD cells. Likewise, high expression of

Nrf2 activated transcription and protein levels of NQO1 in C4-B2-INrfKD cells.

MMP9 was upregulated in Nrf2KO mice after spinal injury as opposed to wild type mice (Mao et al., 2010). Consistent with the previous studies, our mechanistic studies demonstrated that down regulation of Nrf2 and NQO1, increased the activity of

MMP9. The protein level of Cox-2 was increased in LNCaP cells with lower expression of Nrf2. Since Nrf2 signaling is associated in anti-inflammatory mechanisms, lower levels of Nrf2 may facilitate the inflammatory microenvironment that could result in the overexpression of Cox-2 in some cancers. Furthermore, a novel inhibitory role of Nrf2 in the expression of the oncogene Recepteur d‟ origine nantais (RON) was reported. RON oncogene has a key role in cell invasion and metastasis and is upregulated in various cancers of epithelial origin (Leonis et al.,

2007; Thangasamy et al., 2011; Wang et al., 2007a).

In conclusion, the invasion of tumor cells through connective tissue is a crucial prerequisite for initiation of metastasis. Based on our observation, reduced level of Nrf2 resulted in lower level of NQO1 and higher level of Cox-2 and increased enzymatic activity of MMP9 in LNCaP+Nrf2KD cells as compared to LNCaP cells with higher level of Nrf2 (Fig.5-10 A). At this point, it is not clear whether lower level of Nrf2 has any role in enhanced activity of MMP9 or decreased level of NQO1 has any role in activation of MMP9 activity. Since Nrf2 was reported to detoxify

ROS, LNCaP+Nrf2KD cells may generate ROS. Cox-2 and ROS have been implicated in activation of MMP9 (Bu et al., 2011; Zhang et al., 2005). Thus, we believed that reduced levels of Nrf2-mediated increase in MMP9 activity may facilitate the degradation of extracellular matrix (ECM) and reduce inter cellular 113 adhesion that promote tumor invasion (Fig. 5-12B). Hence, determination of Nrf2 levels in primary tumors may allow us to predict metastasis. Taken together, these results revealed that higher activity of MMP9 may facilitate the invasion by degrading

ECM, indicating that Nrf2 could be a crucial target for prevention of metastasis by activating Nrf2 by its activators. Our studies have just developed the hypothesis that lower level of Nrf2 can result in initiation of metastasis. Therefore, further studies are necessary to understand the effect of lower level of Nrf2 on cancer metastasis.

114

CHAPTER SIX

CONCLUSION

Summary:

Our studies found that chronic treatment with letrozole decreased the stability of INrf2 mRNA that resulted in higher levels of Nrf2 and activation of Nrf2 signaling leading to upregulation of anti-apoptotic proteins, antioxidant, drug detoxifying enzyme and efflux pumps, increase in tumor initiating cell (TIC) population and reduction in drug-induced cell death. Nrf2-mediated activation of cytoprotective system collectively can increase chemoresistance that might induce aromatase inhibitor resistance too. Therefore, we believe that Nrf2 inhibitors may enhance the therapeutic efficacy of anti-cancer agents including aromatase inhibitors.

Our studies also demonstrated that there were a greater degree of invasion and migration in prostate cancer cells with low levels of Nrf2. Similarly, the levels of Nrf2 mRNA were dramatically lower in metastasized breast tumor samples from patients.

These intriguing observations suggest that a low level of Nrf2 could be used as an indicator of potential metastasis of primary tumors. Extension of the present study into in vivo models would further highlight our understanding in the role of Nrf2 in cancer metastasis.

As is seen in our studies, the level of Nrf2 protein directly correlates with a subpopulation of tumor initiating cells, therefore a reduction of drug resistance using

Nrf2 inhibitor as adjuvant may be an effective therapeutic approach for early cancer treatment, but this might promote metastasis (Fig.6-1). Resistant cells with high levels

115 of Nrf2 may have anti-metastasis properties. However, any modification due to genetic or epigenetic mechanisms may lower the expression level of Nrf2 in the resistant cells that might program cancer cells towards invasion and migration.

Fig.6-1. Schematic representation of development of drug resistance and progression of metastasis. Activation on Nrf2 in cancer cells protects them from anti-cancer drug- induced cell death that causes resistance to the cancer therapy. Cells with high levels of Nrf2 may also have resistance to metastasis. However, any genetic or epigenetic modifications may decrease Nrf2 levels that may facilitate the initiation of metastasis.

Significance:

We investigated the effect of different levels of Nrf2 protein on the expression of some molecules involved in initiation and progression of metastasis. Examination of a number of genes involved in metastasis in Nrf2-knocked down cells will provide some insights into the molecular mechanisms of prostate cancer metastasis mediated by Nrf2 signaling. Many compounds inducing Nrf2 activators have been identified as chemopreventive agents. Considering the key role of Nrf2 in cellular detoxification and cell survival, Nrf2 signaling has been implicated in chemo-prevention and drug resistance. On one hand, inhibition of Nrf2 is an important approach to enhance the efficacy of chemotherapeutic drugs by blocking the Nrf2-mediated drug resistance pathways. On the other hand, reduced levels of Nrf2 may facilitate cancer metastasis, 116 implying that activators of Nrf2 can be used as anti-metastatic agents. The protein levels of Nrf2 will have a reciprocal effect on chemoresistance and metastasis and cancer cells can take advantage of both activation and downregulation of Nrf2 signaling. Therefore, understanding the fact that whether induction or inhibition of

Nrf2 is effective in cancer therapy based on the stage of cancer progression will provide more insights on the use of inhibitors or inducers of Nrf2. Since cancer cells are smart to take advantage of both upregulation and downregulation of Nrf2 signaling, we need to understand the levels of Nrf2 in tumor cells to decide at what stage of cancer the inhibition of Nrf2 pathway is beneficial to combat drug resistance and the activation of Nrf2 signaling to prevent cancer cells from metastasis.

Future Direction:

Nrf2 was upregulated in aromatase inhibitor-resistant cell lines due to downregulation of INrf2. Our results supported the hypothesis that autophagy activity was high in LTLT-Ca cells. Additionally, autophagy is known to degrade INrf2.

However, inhibition of autophagy did not increase protein level of INrf2. Therefore, further research is necessary to explore the autophagy-mediated degradation of INrf2 degradation in AI-resistant LTLT-Ca cells. The results demonstrated that lower stability of INrf2 mRNA suggests a post transcription regulation. One of the possibilities of the post transcriptional regulation might be miRNA-mediated degradation of INrf2 mRNA. We presumed that letrozole might have a role in generation of miRNA that targets INrf2 mRNA. Therefore, future study should focus on letrozole-mediated miRNA regulation of INrf2 in LTLT-Ca cells.

117

Polyadenylation has a critical role in the maturation and stability of mRNA.

The stability of mRNA is also determined by the expression of poly (A) binding protein (PABP) and RNA binding proteins (RBP) (Wu and Brewer, 2012). Letrozole may have an effect on the binding of RBP to cis-elements and the expression of

PABP and RBP. Therefore, future study should direct towards the effect of letrozole on the regulation of poly (A) binding proteins (PABP) and RNA binding proteins.

We observed that INrf2 levels were decreased when MCF-7Ca and LTLT-Ca cells were treated with natural products as shown in figure (4-1A-C). Elucidation of a mechanism by which INrf2 was degraded with the treatment of these natural products might further our understanding how INrf2 could be downregulated in AI-resistant cells. LTLT-Ca have high levels of ROS, even though, high levels of glutathione

(GSH) was detected in these cells. Thus, an examination of the expression levels of antioxidant-related molecules such as superoxide dismutase (SOD), catalase, thioredoxin and thioredoxin reductase would shed light in understanding why the high levels of ROS in LTLT-Ca cells persists. Therefore, future studies need to be directed to examine the levels of these molecules.

Activation of Nrf2 protects normal cells from environmental insult. However, it also protects cancer cells from anti-cancer drugs. Nrf2 is activated by its inducers through different mechanisms. Our study showed that not only activators of Nrf2, but also inducers of autophagy can elicit Nrf2 signaling through degradation of INrf2. It is possible that many anti-cancer drugs or their metabolites could induce Nrf2. The efficacy of any new chemotherapeutic agent can be further improved by ruling out the possible activation of Nrf2 mediated defense genes during preclinical studies. It is

118 important that during the development of a new drug, induction of Nrf2 by the drug or its metabolites needs to be evaluated. Future studies on the regulation of INrf2 can provide additional insights in our understanding on molecular mechanism of Nrf2 signaling and in the reduction of drug resistance.

Understanding the mechanism of INrf2 regulation will allow us to identify compounds which can induce/stabilize INrf2 which in turn, will open up a new arena to study the downregulation of Nrf2 resulting in sensitizing cancer cells to therapy.

For example, we noticed that withdrawal of letrozole had decreased the levels of Nrf2 in LTLT-Ca cells. Therefore, quantification of Nrf2 protein in tumors will help to decide whether continuation of letrozole treatment is beneficial or not.

Since Bcl-2 mRNA and protein were undetectable in LTLT-Ca cells, and significantly lower levels in other AI-resistant cells suggest that either aromatase inhibitors or their metabolites may have a role in the regulation of Bcl-2 (Fig.2-4B &

C and 6-2A & B). It would be interesting to elucidate the mechanism of Bcl-2 downregulation and to identify molecules that can inhibit the expression of Bcl-2.

Compounds that inhibit Bcl-2 might be used in the treatment of other cancers that are resistant to apoptosis due to overexpression of Bcl-2.

Figure 6-2. Downregulation of Bcl-2 in AI-resistant cells. Cells were lysed and analyzed by western blot for anti-apoptotic protein Bcl-2 (A-B). The Bcl-2 expression is lowered in all AI-resistant cells.

119

We tested the sensitivity of Nrf2-knocked down LTLT-Ca cells to anti-cancer drugs such as doxorubicin and etoposide, which have cytotoxic effect as a mode of action. Future study should examine the sensitivity of the Nrf2-knocked down LTLT-

Ca cells to aromatase inhibitor to determine whether inhibition of Nrf2 signaling can sensitize the LTLT-Ca cells to letrozole.

The present study demonstrated that metastatic related molecules like Cox-2 expression and MMP9 activity were increased while Nrf2 was knocked down, which in effect, downregulated NQO1. In addition, little change in levels of E-cadherin was also observed. Therefore, further studies are warranted to elucidate the molecular mechanism by which prostate cancer cells expressing low levels of Nrf2 can have a greater degree of migration and invasion.

To utilize the bioluminescence imaging system, we have generated stable prostate cancer cell lines expressing lower or higher levels of Nrf2 with constitutively active luciferase expression to investigate the functional role of altered levels of Nrf2 in metastasis progression in an in vivo model (Fig. 6-3A-F). That study will provide us new directions for the use of Nrf2 inducers or inhibitors for better clinical management of cancer. In parallel with breast tumor study, determination of levels of

Nrf2 in metastasized prostate cancer samples would help to further understand how prostate tumors metastasize to different organs.

In this dissertation study, we have used aromatase inhibitor-resistant breast cancer cells (LTLT-Ca) to study the role of Nrf2 in drug resistance and used prostate cancer cells LNCaP and its highly metastatic variant C4-B2 cells to investigate

120 whether lower levels of Nrf2 can initiate metastasis by promoting cell invasion and migration. Based on the results from this study, we concluded that inhibition of Nrf2 appeared to be beneficial to reduce AI-resistance. However, the possibility of greater degree of cell invasion and migration due to inhibition of Nrf2 in AI-resistance cells was not investigated. Future research should address this phenomenon in AI-resistant cells.

Figure 6-3. Luciferase expression in prostate cancer cell lines with altered levels of Nrf2. Luciferase activity were expressed after normalization with 1µg protein. Experiments were done in triplicates, and the standard deviation is indicated. (A & D) Protein levels. (B & E) Transcription level of Nrf2, INrf2 and NQO1. (C & F) Luciferase activity.

121

Antibodies used for western blotting

Protein Primary Secondary Catalogue Species Detected Antibody Antibody- Number (kDa) HRP Acetyl-Histone 3 1:2000 1:5000 CS 9675 Rabbit 17 (Lys 18) Androgen 1:1000 1:5000 sc-816 Rabbit 110 Receptor (AR) BAD 1:1000 1:5000 CS 9268 Rabbit 21 Bax 1:2000 1:5000 sc-493 Rabbit 20 Bcl-2 1:500 1:1000 sc-492 Rabbit 26 Bcl-xL 1:5000 1:10000 sc-8392 Rabbit 33 BCRP 1:100 1:5000 Millipore Mouse 70 OP191 Beclin-1 1:1000 1:5000 CS 3738 Rabbit 60 C/EBPα 1:1000 1:5000 sc-61 Rabbit 42 Calpain 1:1000 1:10000 Calbio- Mouse 28/80/94 208730 Casein Kinase 2α 1:1000 1:5000 CS 2656 Rabbit 42 Caspase-3 1:1000 1:5000 CS 9662 Rabbit 17,19,35 Cox-2 1:1000 1:5000 Invitrogen Mouse 72 35-8200 Cul3 1:1000 1:5000 sc-8556 Rabbit 88 DNA Ligase III 1:1000 1:5000 BD-611877 Mouse 103 E-cadherin 1:5000 1:10000 sc-7870 Rabbit 120,135 EP-1 1:1000 1:10000 Cayman Rabbit 42 101740 EP-4 1:1000 1:10000 αDiagnositic Rabbit 47 EP41-S Estrogen 1:1000 1:5000 Immunotech Mouse 66 Receptor α (ER α) 1545 Focal Adhesion 1:1000 1:5000 CS 3285 Rabbit 125 Kinase GAPDH 1:10000 1:20000 sc-32233 Mouse 38 GCLC 1:1000 1:5000 ab40929 Rabbit 73 Heat Shock 1:10000 1:10000 Sigma Mouse 90 Protein 90 H-1775 HER-2(Erbb-2) 1:1000 1:10000 sc-284 Rabbit 185 HO-1 1:1000 1:2000 sc-10789 Rabbit 34 Hox B13 1:500 1:1000 sc-28333 Mouse 34

122

INrf2 1:500 1:10000 sc-15246 Goat 64-68 Integrin α5 1:1000 1:5000 sc-6614 Goat 150 Ku-70 1:5000 1:10000 sc-17789 Mouse 70

Ku-80 1:5000 1:10000 Millipore Mouse 80 NA54 Lamin B (Nuc.) 1:1000 1:5000 sc-6217 Goat 70 LC3B 1:2000 1:5000 CS 3868 Rabbit 14,16 LDH (cytosol) 1:5000 1:10000 CS 3558 Rabbit 34 MCL-1 1:5000 1:10000 sc-819 Rabbit 42 MDR 1:1000 1:5000 sc-8318 Rabbit 170 MRP-1 1:1000 1:1000 sc-13960 Rabbit 191 MRP-4 1:500 1:1000 Enzo-ALX- Rabbit 159 801-038 N-cadherin 1:500 1:5000 CS 4061 Rabbit 140 NF-κB 1:5000 1:5000 CS 4764 Rabbit 65 NQO1 1:10000 1:10000 sc-32793 Mouse 31 NQO-2 1:5000 1:10000 Proteintech Rabbit 26 15767-1-AP Nrf2 1:1000 1:5000 sc-13032 Rabbit 100 Nrf2 pS40 1:10000 1:25000 ab76026 Rabbit 100 p-AKT 1:1000 1:5000 CS 4058 Rabbit 60 PARP 1:1000 1:5000 CS 9542 Rabbit 116, 89 PGDH 1:1000 1:5000 Cayman Rabbit 25 160615 PKCδ 1:1000 1:5000 CS 2058 Rabbit 78 Pyk2 1:1000 1:5000 sc-1514 Goat 116 PKCδ-V5 1:10000 HRP Invitrogen Rabbit 84 P/N 46-0708 PKCε 1:1000 1:5000 sc-214 Rabbit 82 Rbx1 1:1000 1:5000 Invitrogen Rabbit 15 AHO0402 Retinoic acid 1:1000 1:5000 sc-551 Rabbit 52 Receptor α Slug/Snail 1:1000 1:5000 CS 9585 Rabbit 30 SMAD 1:1000 1:5000 CS 9523 Rabbit 52 SQSTM1(p62) 1:5000 1:10000 Sigma- Rabbit 62 P0067 Ubiquitin 1:5000 1:10000 sc-8017 Mouse 8.5

123

Vimentin 1:500 1:5000 AMF-17b-s Mouse 57 β-Actin 1:1000 1:10000 Sigma- Mouse 43 A5441 β-catenin 1:5000 1:10000 CS 9582 Rabbit 92 β-Tubulin 1:5000 1:10000 sc-9104 Rabbit 55

Table 6-4. Antibody dilution for western blotting

PS: Abcam (ab), Santa Cruz Biotechnology (sc), Cell Signaling (CS), Horseradish Peroxidase (HRP). HRP-conjugated second antibody (Millipore).

124

REFERENCES

Abdullah, L.N., and E.K.-H. Chow. 2013. Mechanisms of chemoresistance in cancer stem cells. Clin Transl Med. 2:3-3.

Achuthan, S., T.R. Santhoshkumar, J. Prabhakar, S.A. Nair, and M.R. Pillai. 2011. Drug-induced senescence generates chemoresistant stemlike cells with low reactive oxygen species. The Journal of biological chemistry. 286:37813- 37829.

Aleksunes, L.M., A.L. Slitt, J.M. Maher, L.M. Augustine, M.J. Goedken, J.Y. Chan, N.J. Cherrington, C.D. Klaassen, and J.E. Manautou. 2008. Induction of Mrp3 and Mrp4 transporters during acetaminophen hepatotoxicity is dependent on Nrf2. Toxicol Appl Pharmacol. 226:74-83.

Alison, M.R., W.R. Lin, S.M. Lim, and L.J. Nicholson. 2012. Cancer stem cells: in the line of fire. Cancer treatment reviews. 38:589-598.

Aoki, Y., H. Sato, N. Nishimura, S. Takahashi, K. Itoh, and M. Yamamoto. 2001. Accelerated DNA adduct formation in the lung of the Nrf2 knockout mouse exposed to diesel exhaust. Toxicology and Applied Pharmacology. 173:154- 160.

Arlt, A., S. Sebens, S. Krebs, C. Geismann, M. Grossmann, M.L. Kruse, S. Schreiber, and H. Schafer. 2012. Inhibition of the Nrf2 transcription factor by the alkaloid trigonelline renders pancreatic cancer cells more susceptible to apoptosis through decreased proteasomal gene expression and proteasome activity. Oncogene.

Awad, O., J.T. Yustein, P. Shah, N. Gul, V. Katuri, A. O'Neill, Y. Kong, M.L. Brown, J.A. Toretsky, and D.M. Loeb. 2010. High ALDH activity identifies chemotherapy-resistant Ewing's sarcoma stem cells that retain sensitivity to EWS-FLI1 inhibition. PloS one. 5:e13943.

Bae, S.H., S.H. Sung, S.Y. Oh, J.M. Lim, S.K. Lee, Y.N. Park, H.E. Lee, D. Kang, and S.G. Rhee. 2013. Sestrins Activate Nrf2 by Promoting p62-Dependent Autophagic Degradation of Keap1 and Prevent Oxidative Liver Damage. Cell Metab. 17:73-84.

Barve, A., T.O. Khor, S. Nair, K. Reuhl, N. Suh, B. Reddy, H. Newmark, and A.-N. Kong. 2009. Gamma-tocopherol-enriched mixed tocopherol diet inhibits prostate carcinogenesis in TRAMP mice. Int J Cancer. 124:1693-1699.

Batlle, E., E. Sancho, C. Franci, D. Dominguez, M. Monfar, J. Baulida, and A. Garcia De Herreros. 2000. The transcription factor snail is a repressor of E-cadherin gene expression in epithelial tumour cells. Nat Cell Biol. 2:84-89.

125

Berx, G., and F. Van Roy. 2001. The E-cadherin/catenin complex: an important gatekeeper in breast cancer tumorigenesis and malignant progression. Breast Cancer Res. 3:289-293.

Bradley, D.A., S. Daignault, C.J. Ryan, R.S. Dipaola, K.A. Cooney, D.C. Smith, E. Small, P. Mathew, M.E. Gross, M.N. Stein, A. Chen, K.J. Pienta, J. Escara- Wilke, G. Doyle, M. Al-Hawary, E.T. Keller, and M. Hussain. 2010. Cilengitide (EMD 121974, NSC 707544) in asymptomatic metastatic castration resistant prostate cancer patients: a randomized phase II trial by the prostate cancer clinical trials consortium. Invest New Drugs. 29:1432-1440.

Bryan, H.K., A. Olayanju, C.E. Goldring, and B.K. Park. 2012. The Nrf2 cell defence pathway: Keap1-dependent and -independent mechanisms of regulation. Biochem Pharmacol.

Bu, X., C. Zhao, and X. Dai. 2011. Involvement of COX-2/PGE(2) Pathway in the Upregulation of MMP-9 Expression in Pancreatic Cancer. Gastroenterology research and practice. 2011:214269.

Calais, G. 1997. [Role of radiotherapy in soft tissue sarcoma]. Cancer Radiother. 1:457-461.

Cameron, E.E., K.E. Bachman, S. Myohanen, J.G. Herman, and S.B. Baylin. 1999. Synergy of demethylation and histone deacetylase inhibition in the re- expression of genes silenced in cancer. Nature genetics. 21:103-107.

Chan, K., X.D. Han, and Y.W. Kan. 2001. An important function of Nrf2 in combating oxidative stress: detoxification of acetaminophen. Proc Natl Acad Sci U S A. 98:4611-4616.

Chen, L.-H., C.-C. Loong, T.-L. Su, Y.-J. Lee, P.-M. Chu, M.-L. Tsai, P.-H. Tsai, P.- H. Tu, C.-W. Chi, H.-C. Lee, and S.-H. Chiou. 2011. Autophagy inhibition enhances apoptosis triggered by BO-1051, an N-mustard derivative, and involves the ATM signaling pathway. Biochemical pharmacology. 81:594- 605.

Chen, N., and J. Debnath. 2010. Autophagy and tumorigenesis. FEBS letters. 584:1427-1435.

Chen, S., S. Masri, X. Wang, S. Phung, Y.C. Yuan, and X. Wu. 2006. What do we know about the mechanisms of aromatase inhibitor resistance? J Steroid Biochem Mol Biol. 102:232-240.

Chen, S.A., M.J. Besman, R.S. Sparkes, S. Zollman, I. Klisak, T. Mohandas, P.F. Hall, and J.E. Shively. 1988. Human aromatase: cDNA cloning, Southern blot analysis, and assignment of the gene to chromosome 15. DNA. 7:27-38.

Chen, W., Z. Sun, X.J. Wang, T. Jiang, Z. Huang, D. Fang, and D.D. Zhang. 2009. Direct interaction between Nrf2 and p21(Cip1/WAF1) upregulates the Nrf2- mediated antioxidant response. Molecular cell. 34:663-673.

126

Cho, H.Y., S.P. Reddy, M. Yamamoto, and S.R. Kleeberger. 2004. The transcription factor NRF2 protects against pulmonary fibrosis. FASEB J. 18:1258-1260.

Choi, A.M., and J. Alam. 1996. Heme oxygenase-1: function, regulation, and implication of a novel stress-inducible protein in oxidant-induced lung injury. Am J Respir Cell Mol Biol. 15:9-19.

Chumsri, S., and A.M. Burger. 2008. Cancer stem cell targeted agents: therapeutic approaches and consequences. Curr Opin Mol Ther. 10:323-333.

Chumsri, S., T. Howes, T. Bao, G. Sabnis, and A. Brodie. 2011. Aromatase, aromatase inhibitors, and breast cancer. J Steroid Biochem Mol Biol. 125:13- 22.

Copple, I.M., C.E. Goldring, N.R. Kitteringham, and B.K. Park. 2008. The Nrf2- Keap1 defence pathway: role in protection against drug-induced toxicity. Toxicology. 246:24-33.

Copple, I.M., A. Lister, A.D. Obeng, N.R. Kitteringham, R.E. Jenkins, R. Layfield, B.J. Foster, C.E. Goldring, and B.K. Park. 2010. Physical and functional interaction of sequestosome 1 with Keap1 regulates the Keap1-Nrf2 cell defense pathway. The Journal of biological chemistry. 285:16782-16788.

Cullinan, S.B., J.D. Gordan, J. Jin, J.W. Harper, and J.A. Diehl. 2004. The Keap1- BTB protein is an adaptor that bridges Nrf2 to a Cul3-based E3 ligase: oxidative stress sensing by a Cul3-Keap1 ligase. Molecular and cellular biology. 24:8477-8486.

Daniel, F.I., K. Cherubini, L.S. Yurgel, M.A. de Figueiredo, and F.G. Salum. 2011. The role of epigenetic transcription repression and DNA methyltransferases in cancer. Cancer. 117:677-687.

DeNicola, G.M., F.A. Karreth, T.J. Humpton, A. Gopinathan, C. Wei, K. Frese, D. Mangal, K.H. Yu, C.J. Yeo, E.S. Calhoun, F. Scrimieri, J.M. Winter, R.H. Hruban, C. Iacobuzio-Donahue, S.E. Kern, I.A. Blair, and D.A. Tuveson. 2011. Oncogene-induced Nrf2 transcription promotes ROS detoxification and tumorigenesis. Nature. 475:106-109.

Dhakshinamoorthy, S., A.K. Jain, D.A. Bloom, and A.K. Jaiswal. 2005. Bach1 competes with Nrf2 leading to negative regulation of the antioxidant response element (ARE)-mediated NAD(P)H:quinone oxidoreductase 1 gene expression and induction in response to antioxidants. The Journal of biological chemistry. 280:16891-16900.

Dhakshinamoorthy, S., and A.K. Jaiswal. 2001. Functional characterization and role of INrf2 in antioxidant response element-mediated expression and antioxidant induction of NAD(P)H:quinone oxidoreductase1 gene. Oncogene. 20:3906- 3917.

127

Dhanoa, B.S., T. Cogliati, A.G. Satish, E.A. Bruford, and J.S. Friedman. 2013. Update on the Kelch-like (KLHL) gene family. Hum Genomics. 7:13-13.

Dhe-Paganon, S., F. Syeda, and L. Park. 2011. DNA methyl transferase 1: regulatory mechanisms and implications in health and disease. International journal of biochemistry and molecular biology. 2:58-66.

Dinkova-Kostova, A.T., W.D. Holtzclaw, and T.W. Kensler. 2005a. The role of Keap1 in cellular protective responses. Chemical research in toxicology. 18:1779-1791.

Dinkova-Kostova, A.T., K.T. Liby, K.K. Stephenson, W.D. Holtzclaw, X. Gao, N. Suh, C. Williams, R. Risingsong, T. Honda, G.W. Gribble, M.B. Sporn, and P. Talalay. 2005b. Extremely potent triterpenoid inducers of the phase 2 response: correlations of protection against oxidant and inflammatory stress. Proc Natl Acad Sci U S A. 102:4584-4589.

Doherty, G.A., S.M. Byrne, E.S. Molloy, V. Malhotra, S.C. Austin, E.W. Kay, F.E. Murray, and D.J. Fitzgerald. 2009. Proneoplastic effects of PGE2 mediated by EP4 receptor in colorectal cancer. BMC Cancer. 9:207-207.

Eades, G., M. Yang, Y. Yao, Y. Zhang, and Q. Zhou. 2011. miR-200a regulates Nrf2 activation by targeting Keap1 mRNA in breast cancer cells. The Journal of biological chemistry. 286:40725-40733.

Egeblad, M., and Z. Werb. 2002. New functions for the matrix metalloproteinases in cancer progression. Nature reviews. Cancer. 2:161-174.

Enomoto, A., K. Itoh, E. Nagayoshi, J. Haruta, T. Kimura, T. O'Connor, T. Harada, and M. Yamamoto. 2001. High sensitivity of Nrf2 knockout mice to acetaminophen hepatotoxicity associated with decreased expression of ARE- regulated drug metabolizing enzymes and antioxidant genes. Toxicological sciences : an official journal of the Society of Toxicology. 59:169-177.

Fang, Y., H. Chen, Y. Hu, Z. Djukic, W. Tevebaugh, N.J. Shaheen, R.C. Orlando, J. Hu, and X.L. Chen. 2013. Gastroesophageal reflux activates the NFkB pathway and impairs esophageal barrier function in mice. Am J Physiol Gastrointest Liver Physiol.

Favreau, L.V., and C.B. Pickett. 1991. Transcriptional regulation of the rat NAD(P)H:quinone reductase gene. Identification of regulatory elements controlling basal level expression and inducible expression by planar aromatic compounds and phenolic antioxidants. The Journal of biological chemistry. 266:4556-4561.

Fornaro, M., T. Manes, and L.R. Languino. 2001. Integrins and prostate cancer metastases. Cancer metastasis reviews. 20:321-331.

Friling, R.S., A. Bensimon, Y. Tichauer, and V. Daniel. 1990. Xenobiotic-inducible expression of murine glutathione S-transferase Ya subunit gene is controlled

128

by an electrophile-responsive element. Proc Natl Acad Sci U S A. 87:6258- 6262.

Frohlich, D.A., M.T. McCabe, R.S. Arnold, and M.L. Day. 2008. The role of Nrf2 in increased reactive oxygen species and DNA damage in prostate tumorigenesis. Oncogene. 27:4353-4362.

Geisler, J., B. Haynes, G. Anker, M. Dowsett, and P.E. Lonning. 2002. Influence of letrozole and anastrozole on total body aromatization and plasma estrogen levels in postmenopausal breast cancer patients evaluated in a randomized, cross-over study. Journal of clinical oncology : official journal of the American Society of Clinical Oncology. 20:751-757.

Gilani, R.A., A.A. Kazi, P. Shah, A.J. Schech, S. Chumsri, G. Sabnis, A.K. Jaiswal, and A.H. Brodie. 2012. The importance of HER2 signaling in the tumor- initiating cell population in aromatase inhibitor-resistant breast cancer. Breast Cancer Res Treat. 135:681-692.

Gillet, J.P., and M.M. Gottesman. 2010. Mechanisms of multidrug resistance in cancer. Methods Mol Biol. 596:47-76.

Ginestier, C., M.H. Hur, E. Charafe-Jauffret, F. Monville, J. Dutcher, M. Brown, J. Jacquemier, P. Viens, C.G. Kleer, S. Liu, A. Schott, D. Hayes, D. Birnbaum, M.S. Wicha, and G. Dontu. 2007. ALDH1 is a marker of normal and malignant human mammary stem cells and a predictor of poor clinical outcome. Cell Stem Cell. 1:555-567.

Goel, A., and B.B. Aggarwal. 2010. Curcumin, the golden spice from Indian saffron, is a chemosensitizer and radiosensitizer for tumors and chemoprotector and radioprotector for normal organs. Nutr Cancer. 62:919-930.

Goel, H.L., J. Li, S. Kogan, and L.R. Languino. 2008. Integrins in prostate cancer progression. Endocrine-related cancer. 15:657-664.

Goldie, J.H. 2001. Drug resistance in cancer: a perspective. Cancer metastasis reviews. 20:63-68.

Gopalakrishnan, A., and A.N. Tony Kong. 2008. Anticarcinogenesis by dietary phytochemicals: cytoprotection by Nrf2 in normal cells and cytotoxicity by modulation of transcription factors NF-kappa B and AP-1 in abnormal cancer cells. Food and chemical toxicology : an international journal published for the British Industrial Biological Research Association. 46:1257-1270.

Gottesman, M.M. 2002. Mechanisms of cancer drug resistance. Annu Rev Med. 53:615-627.

Gozzelino, R., V. Jeney, and M.P. Soares. 2010. Mechanisms of cell protection by heme oxygenase-1. Annual review of pharmacology and toxicology. 50:323- 354.

129

Greenberg, N.M., F. DeMayo, M.J. Finegold, D. Medina, W.D. Tilley, J.O. Aspinall, G.R. Cunha, A.A. Donjacour, R.J. Matusik, and J.M. Rosen. 1995. Prostate cancer in a transgenic mouse. Proc Natl Acad Sci U S A. 92:3439-3443.

Han, S.A., H. Chun, C.M. Park, S.J. Kang, S.H. Kim, D. Sohn, S.H. Yun, and W.Y. Lee. 2006. Prognostic significance of beta-catenin in colorectal cancer with liver metastasis. Clin Oncol (R Coll Radiol). 18:761-767.

Hast, B.E., D. Goldfarb, K.M. Mulvaney, M.A. Hast, P.F. Siesser, F. Yan, D.N. Hayes, and M.B. Major. 2013. Proteomic analysis of ubiquitin ligase KEAP1 reveals associated proteins that inhibit NRF2 ubiquitination. Cancer Res. 73:2199-2210.

Hayashi, A., H. Suzuki, K. Itoh, M. Yamamoto, and Y. Sugiyama. 2003. Transcription factor Nrf2 is required for the constitutive and inducible expression of multidrug resistance-associated protein 1 in mouse embryo fibroblasts. Biochemical and biophysical research communications. 310:824- 829.

Hayes, J.D., and M. McMahon. 2009. NRF2 and KEAP1 mutations: permanent activation of an adaptive response in cancer. Trends Biochem Sci. 34:176-188.

He, X., M.G. Chen, and Q. Ma. 2008. Activation of Nrf2 in defense against cadmium- induced oxidative stress. Chemical research in toxicology. 21:1375-1383.

He, X., and Q. Ma. 2009. NRF2 cysteine residues are critical for oxidant/electrophile- sensing, Kelch-like ECH-associated protein-1-dependent ubiquitination- proteasomal degradation, and transcription activation. Molecular pharmacology. 76:1265-1278.

Heidenreich, A., G. Aus, M. Bolla, S. Joniau, V.B. Matveev, H.P. Schmid, F. Zattoni, and U. European Association of. 2008. EAU guidelines on prostate cancer. European urology. 53:68-80.

Homma, S., Y. Ishii, Y. Morishima, T. Yamadori, Y. Matsuno, N. Haraguchi, N. Kikuchi, H. Satoh, T. Sakamoto, N. Hizawa, K. Itoh, and M. Yamamoto. 2009. Nrf2 enhances cell proliferation and resistance to anticancer drugs in human lung cancer. Clin Cancer Res. 15:3423-3432.

Hong, F., K.R. Sekhar, M.L. Freeman, and D.C. Liebler. 2005. Specific patterns of electrophile adduction trigger Keap1 ubiquitination and Nrf2 activation. The Journal of biological chemistry. 280:31768-31775.

Hoshino, I., and H. Matsubara. 2010. Recent advances in histone deacetylase targeted cancer therapy. Surgery today. 40:809-815.

Hu, R., C. Xu, G. Shen, M.R. Jain, T.O. Khor, A. Gopalkrishnan, W. Lin, B. Reddy, J.Y. Chan, and A.N. Kong. 2006a. Gene expression profiles induced by cancer chemopreventive isothiocyanate sulforaphane in the liver of C57BL/6J mice and C57BL/6J/Nrf2 (-/-) mice. Cancer Lett. 243:170-192.

130

Hu, X., J.R. Roberts, P.L. Apopa, Y.W. Kan, and Q. Ma. 2006b. Accelerated ovarian failure induced by 4-vinyl cyclohexene diepoxide in Nrf2 null mice. Molecular and cellular biology. 26:940-954.

Iida, K., K. Itoh, Y. Kumagai, R. Oyasu, K. Hattori, K. Kawai, T. Shimazui, H. Akaza, and M. Yamamoto. 2004. Nrf2 is essential for the chemopreventive efficacy of oltipraz against urinary bladder carcinogenesis. Cancer Research. 64:6424-6431.

Iizuka, T., Y. Ishii, K. Itoh, T. Kiwamoto, T. Kimura, Y. Matsuno, Y. Morishima, A.E. Hegab, S. Homma, A. Nomura, T. Sakamoto, M. Shimura, A. Yoshida, M. Yamamoto, and K. Sekizawa. 2005. Nrf2-deficient mice are highly susceptible to cigarette smoke-induced emphysema. Genes Cells. 10:1113- 1125.

Inami, Y., S. Waguri, A. Sakamoto, T. Kouno, K. Nakada, O. Hino, S. Watanabe, J. Ando, M. Iwadate, M. Yamamoto, M.S. Lee, K. Tanaka, and M. Komatsu. 2011. Persistent activation of Nrf2 through p62 in hepatocellular carcinoma cells. The Journal of cell biology. 193:275-284.

Itoh, K., T. Chiba, S. Takahashi, T. Ishii, K. Igarashi, Y. Katoh, T. Oyake, N. Hayashi, K. Satoh, I. Hatayama, M. Yamamoto, and Y. Nabeshima. 1997. An Nrf2/small Maf heterodimer mediates the induction of phase II detoxifying enzyme genes through antioxidant response elements. Biochemical and biophysical research communications. 236:313-322.

Itoh, K., K.I. Tong, and M. Yamamoto. 2004. Molecular mechanism activating Nrf2- Keap1 pathway in regulation of adaptive response to electrophiles. Free Radic Biol Med. 36:1208-1213.

Itoh, K., N. Wakabayashi, Y. Katoh, T. Ishii, K. Igarashi, J.D. Engel, and M. Yamamoto. 1999a. Keap1 represses nuclear activation of antioxidant responsive elements by Nrf2 through binding to the amino-terminal Neh2 domain. Genes Dev. 13:76-86.

Itoh, K., N. Wakabayashi, Y. Katoh, T. Ishii, K. Igarashi, J.D. Engel, and M. Yamamoto. 1999b. Keap1 represses nuclear activation of antioxidant responsive elements by Nrf2 through binding to the amino-terminal Neh2 domain. Genes Dev. 13:76-86.

Jain, A.K., and A.K. Jaiswal. 2007. GSK-3beta acts upstream of Fyn kinase in regulation of nuclear export and degradation of NF-E2 related factor 2. The Journal of biological chemistry. 282:16502-16510.

Jaiswal, A.K. 2004. Nrf2 signaling in coordinated activation of antioxidant gene expression. Free Radic Biol Med. 36:1199-1207.

Jelovac, D., G. Sabnis, B.J. Long, L. Macedo, O.G. Goloubeva, and A.M. Brodie. 2005. Activation of mitogen-activated protein kinase in xenografts and cells

131

during prolonged treatment with aromatase inhibitor letrozole. Cancer Res. 65:5380-5389.

Jeong, W.S., I.W. Kim, R. Hu, and A.N. Kong. 2004. Modulatory properties of various natural chemopreventive agents on the activation of NF-kappaB signaling pathway. Pharm Res. 21:661-670.

Ji, L., H. Li, P. Gao, G. Shang, D.D. Zhang, N. Zhang, and T. Jiang. 2013. Nrf2 pathway regulates multidrug-resistance-associated protein 1 in small cell lung cancer. PloS one. 8:e63404.

Kang, S.J., A. You, and M.K. Kwak. 2011. Suppression of Nrf2 signaling by angiotensin II in murine renal epithelial cells. Arch Pharm Res. 34:829-836.

Kannappan, R., S.C. Gupta, J.H. Kim, and B.B. Aggarwal. 2012. Tocotrienols fight cancer by targeting multiple cell signaling pathways. Genes Nutr. 7:43-52.

Karahoca, M., and R.L. Momparler. 2013. Pharmacokinetic and pharmacodynamic analysis of 5-aza-2'-deoxycytidine (decitabine) in the design of its dose- schedule for cancer therapy. Clin Epigenetics. 5:3.

Kaspar, J.W., S.K. Niture, and A.K. Jaiswal. 2009. Nrf2:INrf2 (Keap1) signaling in oxidative stress. Free Radic Biol Med. 47:1304-1309.

Kawai, Y., L. Garduno, M. Theodore, J. Yang, and I.J. Arinze. 2011. Acetylation- deacetylation of the transcription factor Nrf2 (nuclear factor erythroid 2- related factor 2) regulates its transcriptional activity and nucleocytoplasmic localization. The Journal of biological chemistry. 286:7629-7640.

Ke, N., X. Wang, X. Xu, and Y.A. Abassi. 2011. The xCELLigence system for real- time and label-free monitoring of cell viability. Methods Mol Biol. 740:33-43.

Khan, M.I., A. Mohammad, G. Patil, S.A.H. Naqvi, L.K.S. Chauhan, and I. Ahmad. 2011. Induction of ROS, mitochondrial damage and autophagy in lung epithelial cancer cells by iron oxide nanoparticles. Biomaterials. 33:1477- 1488.

Kim, J.I., V. Lakshmikanthan, N. Frilot, and Y. Daaka. 2010. Prostaglandin E2 promotes lung cancer cell migration via EP4-betaArrestin1-c-Src signalsome. Mol Cancer Res. 8:569-577.

Kinch, L., N.V. Grishin, and J. Brugarolas. 2011. Succination of Keap1 and activation of Nrf2-dependent antioxidant pathways in FH-deficient papillary renal cell carcinoma type 2. Cancer cell. 20:418-420.

Klaassen, C.D., and S.A. Reisman. 2010. Nrf2 the rescue: effects of the antioxidative/electrophilic response on the liver. Toxicol Appl Pharmacol. 244:57-65.

132

Komatsu, M., H. Kurokawa, S. Waguri, K. Taguchi, A. Kobayashi, Y. Ichimura, Y.S. Sou, I. Ueno, A. Sakamoto, K.I. Tong, M. Kim, Y. Nishito, S. Iemura, T. Natsume, T. Ueno, E. Kominami, H. Motohashi, K. Tanaka, and M. Yamamoto. 2010. The selective autophagy substrate p62 activates the stress responsive transcription factor Nrf2 through inactivation of Keap1. Nat Cell Biol. 12:213-223.

Kongara, S., and V. Karantza. 2012. The interplay between autophagy and ROS in tumorigenesis. Frontiers in oncology. 2:1-13.

Kundu, J.K., Y.K. Shin, and Y.J. Surh. 2006. Resveratrol modulates phorbol ester- induced pro-inflammatory signal transduction pathways in mouse skin in vivo: NF-kappaB and AP-1 as prime targets. Biochem Pharmacol. 72:1506-1515.

Kwak, M.K., K. Itoh, M. Yamamoto, and T.W. Kensler. 2002. Enhanced expression of the transcription factor Nrf2 by cancer chemopreventive agents: role of antioxidant response element-like sequences in the nrf2 promoter. Molecular and cellular biology. 22:2883-2892.

Lau, A., N.F. Villeneuve, Z. Sun, P.K. Wong, and D.D. Zhang. 2008. Dual roles of Nrf2 in cancer. Pharmacol Res. 58:262-270.

Lee, O.H., A.K. Jain, V. Papusha, and A.K. Jaiswal. 2007. An auto-regulatory loop between stress sensors INrf2 and Nrf2 controls their cellular abundance. The Journal of biological chemistry. 282:36412-36420.

Lee, S., M.J. Lim, M.H. Kim, C.H. Yu, Y.S. Yun, J. Ahn, and J.Y. Song. 2012. An effective strategy for increasing the radiosensitivity of Human lung Cancer cells by blocking Nrf2-dependent antioxidant responses. Free Radic Biol Med. 53:807-816.

Leonis, M.A., M.N. Thobe, and S.E. Waltz. 2007. Ron-receptor tyrosine kinase in tumorigenesis and metastasis. Future oncology. 3:441-448.

Li, M., J.F. Chiu, A. Kelsen, S.C. Lu, and N.K. Fukagawa. 2009. Identification and characterization of an Nrf2-mediated ARE upstream of the rat glutamate cysteine ligase catalytic subunit gene (GCLC). J Cell Biochem. 107:944-954.

Li, W., and A.N. Kong. 2009. Molecular mechanisms of Nrf2-mediated antioxidant response. Mol Carcinog. 48:91-104.

Liu, F.S. 2009. Mechanisms of chemotherapeutic drug resistance in cancer therapy--a quick review. Taiwanese journal of obstetrics & gynecology. 48:239-244.

Ma, Q. 2013. Role of nrf2 in oxidative stress and toxicity. Annual review of pharmacology and toxicology. 53:401-426.

Ma, Q., and X. He. 2012. Molecular basis of electrophilic and oxidative defense: promises and perils of Nrf2. Pharmacological reviews. 64:1055-1081.

133

Ma, Q., K. Kinneer, J. Ye, and B.J. Chen. 2003. Inhibition of nuclear factor kappaB by phenolic antioxidants: interplay between antioxidant signaling and inflammatory cytokine expression. Molecular pharmacology. 64:211-219.

Ma, S., T.K. Lee, B.J. Zheng, K.W. Chan, and X.Y. Guan. 2008. CD133+ HCC cancer stem cells confer chemoresistance by preferential expression of the Akt/PKB survival pathway. Oncogene. 27:1749-1758.

Ma, X., N. Kundu, O.B. Ioffe, O. Goloubeva, R. Konger, C. Baquet, P. Gimotty, J. Reader, and A.M. Fulton. 2010. Prostaglandin E receptor EP1 suppresses breast cancer metastasis and is linked to survival differences and cancer disparities. Mol Cancer Res. 8:1310-1318.

Macedo, L.F., G. Sabnis, and A. Brodie. 2009. Aromatase inhibitors and breast cancer. Ann N Y Acad Sci. 1155:162-173.

Maher, J.M., X. Cheng, A.L. Slitt, M.Z. Dieter, and C.D. Klaassen. 2005. Induction of the multidrug resistance-associated protein family of transporters by chemical activators of receptor-mediated pathways in mouse liver. Drug Metab Dispos. 33:956-962.

Maher, J.M., M.Z. Dieter, L.M. Aleksunes, A.L. Slitt, G. Guo, Y. Tanaka, G.L. Scheffer, J.Y. Chan, J.E. Manautou, Y. Chen, T.P. Dalton, M. Yamamoto, and C.D. Klaassen. 2007. Oxidative and electrophilic stress induces multidrug resistance-associated protein transporters via the nuclear factor-E2-related factor-2 transcriptional pathway. Hepatology. 46:1597-1610.

Mao, L., H. Wang, L. Qiao, and X. Wang. 2010. Disruption of Nrf2 enhances the upregulation of nuclear factor-kappaB activity, tumor necrosis factor-alpha, and matrix metalloproteinase-9 after spinal cord injury in mice. Mediators of inflammation. 2010:238321.

Margaritopoulos, G.A., E. Tsitoura, N. Tzanakis, D.A. Spandidos, N.M. Siafakas, G. Sourvinos, and K.M. Antoniou. 2013. Self-eating: friend or foe? The emerging role of autophagy in idiopathic pulmonary fibrosis. Biomed Res Int. 2013.

Marks, P.A., and W.S. Xu. 2009. Histone deacetylase inhibitors: Potential in cancer therapy. J Cell Biochem. 107:600-608.

Martin, M.D., and L.M. Matrisian. 2007. The other side of MMPs: protective roles in tumor progression. Cancer metastasis reviews. 26:717-724.

McMahon, M., K. Itoh, M. Yamamoto, S.A. Chanas, C.J. Henderson, L.I. McLellan, C.R. Wolf, C. Cavin, and J.D. Hayes. 2001. The Cap'n'Collar basic leucine zipper transcription factor Nrf2 (NF-E2 p45-related factor 2) controls both constitutive and inducible expression of intestinal detoxification and glutathione biosynthetic enzymes. Cancer Res. 61:3299-3307.

McMahon, M., N. Thomas, K. Itoh, M. Yamamoto, and J.D. Hayes. 2006. Dimerization of substrate adaptors can facilitate cullin-mediated ubiquitylation

134

of proteins by a "tethering" mechanism: a two-site interaction model for the Nrf2-Keap1 complex. The Journal of biological chemistry. 281:24756-24768.

Miao, W., L. Hu, P.J. Scrivens, and G. Batist. 2005. Transcriptional regulation of NF- E2 p45-related factor (NRF2) expression by the aryl hydrocarbon receptor- xenobiotic response element signaling pathway: direct cross-talk between phase I and II drug-metabolizing enzymes. The Journal of biological chemistry. 280:20340-20348.

Miller, W.R. 2010. Aromatase inhibitors: prediction of response and nature of resistance. Expert Opin Pharmacother. 11:1873-1887.

Moi, P., K. Chan, I. Asunis, A. Cao, and Y.W. Kan. 1994. Isolation of NF-E2-related factor 2 (Nrf2), a NF-E2-like basic leucine zipper transcriptional activator that binds to the tandem NF-E2/AP1 repeat of the beta-globin locus control region. Proc Natl Acad Sci U S A. 91:9926-9930.

Mori, S., J.T. Chang, E.R. Andrechek, N. Matsumura, T. Baba, G. Yao, J.W. Kim, M. Gatza, S. Murphy, and J.R. Nevins. 2009. Anchorage-independent cell growth signature identifies tumors with metastatic potential. Oncogene. 28:2796- 2805.

Mulcahy, R.T., M.A. Wartman, H.H. Bailey, and J.J. Gipp. 1997. Constitutive and beta-naphthoflavone-induced expression of the human gamma- glutamylcysteine synthetase heavy subunit gene is regulated by a distal antioxidant response element/TRE sequence. The Journal of biological chemistry. 272:7445-7454.

Nelson, W.J., and R. Nusse. 2004. Convergence of Wnt, beta-catenin, and cadherin pathways. Science. 303:1483-1487.

Nguyen, T., C.S. Yang, and C.B. Pickett. 2004. The pathways and molecular mechanisms regulating Nrf2 activation in response to chemical stress. Free Radic Biol Med. 37:433-441.

Nicolini, A., P. Ferrari, M. Fini, V. Borsari, P. Fallahi, A. Antonelli, P. Berti, A. Carpi, and P. Miccoli. 2011. Stem cells: their role in breast cancer development and resistance to treatment. Curr Pharm Biotechnol. 12:196-205.

Nioi, P., and T. Nguyen. 2007. A mutation of Keap1 found in breast cancer impairs its ability to repress Nrf2 activity. Biochemical and biophysical research communications. 362:816-821.

Niture, S.K., A.K. Jain, and A.K. Jaiswal. 2009. Antioxidant-induced modification of INrf2 cysteine 151 and PKC-delta-mediated phosphorylation of Nrf2 serine 40 are both required for stabilization and nuclear translocation of Nrf2 and increased drug resistance. J Cell Sci. 122:4452-4464.

135

Niture, S.K., and A.K. Jaiswal. 2010. Hsp90 interaction with INrf2(Keap1) mediates stress-induced Nrf2 activation. The Journal of biological chemistry. 285:36865-36875.

Niture, S.K., and A.K. Jaiswal. 2012. Nrf2 protein up-regulates antiapoptotic protein Bcl-2 and prevents cellular apoptosis. The Journal of biological chemistry. 287:9873-9886.

Niture, S.K., and A.K. Jaiswal. 2013. Nrf2-induced antiapoptotic Bcl-xL protein enhances cell survival and drug resistance. Free Radic Biol Med. 57:119-131.

Niture, S.K., R. Khatri, and A.K. Jaiswal. 2013. Regulation of Nrf2-an update. Free Radic Biol Med.

Ohnuma, T., T. Matsumoto, A. Itoi, A. Kawana, T. Nishiyama, K. Ogura, and A. Hiratsuka. 2011. Enhanced sensitivity of A549 cells to the cytotoxic action of anticancer drugs via suppression of Nrf2 by procyanidins from Cinnamomi Cortex extract. Biochemical and biophysical research communications. 413:623-629.

Ohta, T., K. Iijima, M. Miyamoto, I. Nakahara, H. Tanaka, M. Ohtsuji, T. Suzuki, A. Kobayashi, J. Yokota, T. Sakiyama, T. Shibata, M. Yamamoto, and S. Hirohashi. 2008. Loss of Keap1 function activates Nrf2 and provides advantages for lung cancer cell growth. Cancer Res. 68:1303-1309.

Okawa, H., H. Motohashi, A. Kobayashi, H. Aburatani, T.W. Kensler, and M. Yamamoto. 2006. Hepatocyte-specific deletion of the keap1 gene activates Nrf2 and confers potent resistance against acute drug toxicity. Biochemical and biophysical research communications. 339:79-88.

Osburn, W.O., and T.W. Kensler. 2008. Nrf2 signaling: an adaptive response pathway for protection against environmental toxic insults. Mutat Res. 659:31-39.

Padmanabhan, B., K.I. Tong, T. Ohta, Y. Nakamura, M. Scharlock, M. Ohtsuji, M.I. Kang, A. Kobayashi, S. Yokoyama, and M. Yamamoto. 2006. Structural basis for defects of Keap1 activity provoked by its point mutations in lung cancer. Molecular cell. 21:689-700.

Pankiv, S., T.H. Clausen, T. Lamark, A. Brech, J.A. Bruun, H. Outzen, A. Overvatn, G. Bjorkoy, and T. Johansen. 2007. p62/SQSTM1 binds directly to Atg8/LC3 to facilitate degradation of ubiquitinated protein aggregates by autophagy. The Journal of biological chemistry. 282:24131-24145.

Patra, S.K., and S. Bettuzzi. 2009. Epigenetic DNA-(cytosine-5-carbon) modifications: 5-aza-2'-deoxycytidine and DNA-demethylation. Biochemistry. Biokhimiia. 74:613-619.

Pavet, V., M.M. Portal, J.C. Moulin, R. Herbrecht, and H. Gronemeyer. 2011. Towards novel paradigms for cancer therapy. Oncogene. 30:1-20.

136

Pawlowski, K.M., J. Mucha, K. Majchrzak, T. Motyl, and M. Krol. 2013. Expression and role of PGP, BCRP, MRP1 and MRP3 in multidrug resistance of canine mammary cancer cells. BMC Vet Res. 9:119.

Pećina-Slaus, N. 2003. Tumor suppressor gene E-cadherin and its role in normal and malignant cells. Cancer Cell Int. 3:17-17.

Rachakonda, G., K.R. Sekhar, D. Jowhar, P.C. Samson, J.P. Wikswo, R.D. Beauchamp, P.K. Datta, and M.L. Freeman. Increased cell migration and plasticity in Nrf2-deficient cancer cell lines. Oncogene. 29:3703-3714.

Rachakonda, G., K.R. Sekhar, D. Jowhar, P.C. Samson, J.P. Wikswo, R.D. Beauchamp, P.K. Datta, and M.L. Freeman. 2010. Increased cell migration and plasticity in Nrf2-deficient cancer cell lines. Oncogene. 29:3703-3714.

Rada, P., A.I. Rojo, S. Chowdhry, M. McMahon, J.D. Hayes, and A. Cuadrado. 2011. SCF/{beta}-TrCP promotes glycogen synthase kinase 3-dependent degradation of the Nrf2 transcription factor in a Keap1-independent manner. Molecular and cellular biology. 31:1121-1133.

Rada, P., A.I. Rojo, N. Evrard-Todeschi, N.G. Innamorato, A. Cotte, T. Jaworski, J.C. Tobon-Velasco, H. Devijver, M.F. Garcia-Mayoral, F. Van Leuven, J.D. Hayes, G. Bertho, and A. Cuadrado. 2012. Structural and functional characterization of Nrf2 degradation by the glycogen synthase kinase 3/beta- TrCP axis. Molecular and cellular biology. 32:3486-3499.

Raguz, S., and E. Yague. 2008. Resistance to chemotherapy: new treatments and novel insights into an old problem. British journal of cancer. 99:387-391.

Ramos-Gomez, M., M.K. Kwak, P.M. Dolan, K. Itoh, M. Yamamoto, P. Talalay, and T.W. Kensler. 2001. Sensitivity to carcinogenesis is increased and chemoprotective efficacy of enzyme inducers is lost in nrf2 transcription factor-deficient mice. Proceedings of the National Academy of Sciences of the United States of America. 98:3410-3415.

Rassool, F.V., and A.E. Tomkinson. 2010. Targeting abnormal DNA double strand break repair in cancer. Cell Mol Life Sci. 67:3699-3710.

Ronaldson, P.T., T. Ashraf, and R. Bendayan. 2010. Regulation of multidrug resistance protein 1 by tumor necrosis factor alpha in cultured glial cells: involvement of nuclear factor-kappaB and c-Jun N-terminal kinase signaling pathways. Molecular pharmacology. 77:644-659.

Saadin, K., J.M. Burke, N.P. Patel, R.E. Zubajlo, and I.M. White. 2013. Enrichment of tumor-initiating breast cancer cells within a mammosphere-culture microdevice. Biomed Microdevices. 15:645-655.

Sabnis, G., O. Goloubeva, R. Gilani, L. Macedo, and A. Brodie. 2010. Sensitivity to the aromatase inhibitor letrozole is prolonged after a "break" in treatment. Mol Cancer Ther. 9:46-56.

137

Sabnis, G.J., D. Jelovac, B. Long, and A. Brodie. 2005. The role of growth factor receptor pathways in human breast cancer cells adapted to long-term estrogen deprivation. Cancer Res. 65:3903-3910.

Sabnis, G.J., L.F. Macedo, O. Goloubeva, A. Schayowitz, and A.M. Brodie. 2008. Stopping treatment can reverse acquired resistance to letrozole. Cancer Res. 68:4518-4524.

Sakamoto, S., and N. Kyprianou. 2010. Targeting anoikis resistance in prostate cancer metastasis. Mol Aspects Med. 31:205-214.

Saraon, P., N. Musrap, D. Cretu, G.S. Karagiannis, I. Batruch, C. Smith, A.P. Drabovich, D. Trudel, T. van der Kwast, C. Morrissey, K.A. Jarvi, and E.P. Diamandis. 2012. Proteomic profiling of androgen-independent prostate cancer cell lines reveals a role for protein S during the development of high grade and castration-resistant prostate cancer. The Journal of biological chemistry. 287:34019-34031.

Satoh, H., T. Moriguchi, K. Taguchi, J. Takai, J.M. Maher, T. Suzuki, P.T. Winnard, Jr., V. Raman, M. Ebina, T. Nukiwa, and M. Yamamoto. 2010. Nrf2- deficiency creates a responsive microenvironment for metastasis to the lung. Carcinogenesis. 31:1833-1843.

Sebova, K., and I. Fridrichova. 2010. Epigenetic tools in potential anticancer therapy. Anti-cancer drugs. 21:565-577.

Shao, Y., Z. Gao, P.A. Marks, and X. Jiang. 2004. Apoptotic and autophagic cell death induced by histone deacetylase inhibitors. Proc Natl Acad Sci U S A. 101:18030-18035.

Shelton, P., and A.K. Jaiswal. 2012. The transcription factor NF-E2-related factor 2 (Nrf2): a protooncogene? FASEB J. 27:414-423.

Shibata, T., T. Ohta, K.I. Tong, A. Kokubu, R. Odogawa, K. Tsuta, H. Asamura, M. Yamamoto, and S. Hirohashi. 2008. Cancer related mutations in NRF2 impair its recognition by Keap1-Cul3 E3 ligase and promote malignancy. Proc Natl Acad Sci U S A. 105:13568-13573.

Shih, A.Y., S. Imbeault, V. Barakauskas, H. Erb, L. Jiang, P. Li, and T.H. Murphy. 2005. Induction of the Nrf2-driven antioxidant response confers neuroprotection during mitochondrial stress in vivo. The Journal of biological chemistry. 280:22925-22936.

Shim, G.S., S. Manandhar, D.H. Shin, T.H. Kim, and M.K. Kwak. 2009. Acquisition of doxorubicin resistance in ovarian carcinoma cells accompanies activation of the NRF2 pathway. Free Radic Biol Med. 47:1619-1631.

Shu, L., K.L. Cheung, T.O. Khor, C. Chen, and A.N. Kong. 2010. Phytochemicals: cancer chemoprevention and suppression of tumor onset and metastasis. Cancer metastasis reviews. 29:483-502.

138

Siegel, D., C. Yan, and D. Ross. 2012. NAD(P)H:quinone oxidoreductase 1 (NQO1) in the sensitivity and resistance to antitumor quinones. Biochem Pharmacol. 83:1033-1040.

Siegel, R., D. Naishadham, and A. Jemal. 2013. Cancer statistics, 2013. CA: a cancer journal for clinicians. 63:11-30.

Singh, A., V. Misra, R.K. Thimmulappa, H. Lee, S. Ames, M.O. Hoque, J.G. Herman, S.B. Baylin, D. Sidransky, E. Gabrielson, M.V. Brock, and S. Biswal. 2006. Dysfunctional KEAP1-NRF2 interaction in non-small-cell lung cancer. PLoS Med. 3:e420.

Singh, A., H. Wu, P. Zhang, C. Happel, J. Ma, and S. Biswal. 2010. Expression of ABCG2 (BCRP) is regulated by Nrf2 in cancer cells that confers side population and chemoresistance phenotype. Mol Cancer Ther. 9:2365-2376.

Singh, R.P., and R. Agarwal. 2004. Prostate cancer prevention by silibinin. Current cancer drug targets. 4:1-11.

Singh, T., and S.K. Katiyar. 2013. Honokiol inhibits non-small cell lung cancer cell migration by targeting PGE(2)-mediated activation of beta-catenin signaling. PloS one. 8:e60749.

Sinnberg, T., M. Menzel, D. Ewerth, B. Sauer, M. Schwarz, M. Schaller, C. Garbe, and B. Schittek. 2011. β-Catenin signaling increases during melanoma progression and promotes tumor cell survival and chemoresistance. PloS one. 6.

Slack, J.K., R.B. Adams, J.D. Rovin, E.A. Bissonette, C.E. Stoker, and J.T. Parsons. 2001. Alterations in the focal adhesion kinase/Src signal transduction pathway correlate with increased migratory capacity of prostate carcinoma cells. Oncogene. 20:1152-1163.

Slitt, A.L., N.J. Cherrington, M.Z. Dieter, L.M. Aleksunes, G.L. Scheffer, W. Huang, D.D. Moore, and C.D. Klaassen. 2006. trans-Stilbene oxide induces expression of genes involved in metabolism and transport in mouse liver via CAR and Nrf2 transcription factors. Molecular pharmacology. 69:1554-1563.

Song, N.Y., D.H. Kim, E.H. Kim, H.K. Na, and Y.J. Surh. 2009. 15-Deoxy-delta 12, 14-prostaglandin J2 induces upregulation of multidrug resistance-associated protein 1 via Nrf2 activation in human breast cancer cells. Ann N Y Acad Sci. 1171:210-216.

Taguchi, K., N. Fujikawa, M. Komatsu, T. Ishii, M. Unno, T. Akaike, H. Motohashi, and M. Yamamoto. 2012. Keap1 degradation by autophagy for the maintenance of redox homeostasis. Proc Natl Acad Sci U S A. 109:13561- 13566.

Talmadge, J.E., and I.J. Fidler. 2010. AACR centennial series: the biology of cancer metastasis: historical perspective. Cancer Res. 70:5649-5669.

139

Tang, X., H. Wang, L. Fan, X. Wu, A. Xin, H. Ren, and X.J. Wang. 2011. Luteolin inhibits Nrf2 leading to negative regulation of the Nrf2/ARE pathway and sensitization of human lung carcinoma A549 cells to therapeutic drugs. Free Radic Biol Med. 50:1599-1609.

Tang, X., Q. Zhou, S. Zhang, L. Liu, and N. Cheng. 2002. [A study on the relationship between E-cadherin, β-catenin expression and metastasis and prognosis in non-small cell lung cancer]. Zhongguo Fei Ai Za Zhi. 5:263-267.

Thangasamy, A., J. Rogge, N.K. Krishnegowda, J.W. Freeman, and S. Ammanamanchi. 2011. Novel function of transcription factor Nrf2 as an inhibitor of RON tyrosine kinase receptor-mediated cancer cell invasion. The Journal of biological chemistry. 286:32115-32122.

Thompson, E.A., Jr., and P.K. Siiteri. 1974. Utilization of oxygen and reduced nicotinamide adenine dinucleotide phosphate by human placental microsomes during aromatization of androstenedione. The Journal of biological chemistry. 249:5364-5372.

Tian, H., B. Zhang, J. Di, G. Jiang, F. Chen, H. Li, L. Li, D. Pei, and J. Zheng. 2012. Keap1: one stone kills three birds Nrf2, IKKbeta and Bcl-2/Bcl-xL. Cancer Lett. 325:26-34.

Tobin, L.A., C. Robert, P. Nagaria, S. Chumsri, W. Twaddell, O.B. Ioffe, G.E. Greco, A.H. Brodie, A.E. Tomkinson, and F.V. Rassool. 2012. Targeting abnormal DNA repair in therapy-resistant breast cancers. Mol Cancer Res. 10:96-107.

Valko, M., D. Leibfritz, J. Moncol, M.T. Cronin, M. Mazur, and J. Telser. 2006. Free radicals and antioxidants in normal physiological functions and human disease. Int J Biochem Cell Biol. 39:44-84. van Jaarsveld, M.T.M., J. Helleman, A.W.M. Boersma, P.F. van Kuijk, W.F. van Ijcken, E. Despierre, I. Vergote, R.H.J. Mathijssen, E.M.J.J. Berns, J. Verweij, J. Pothof, and E.A.C. Wiemer. 2012. miR-141 regulates KEAP1 and modulates cisplatin sensitivity in ovarian cancer cells. Oncogene.

Venugopal, R., and A.K. Jaiswal. 1996. Nrf1 and Nrf2 positively and c-Fos and Fra1 negatively regulate the human antioxidant response element-mediated expression of NAD(P)H:quinone oxidoreductase1 gene. Proc Natl Acad Sci U S A. 93:14960-14965.

Venugopal, R., and A.K. Jaiswal. 1998. Nrf2 and Nrf1 in association with Jun proteins regulate antioxidant response element-mediated expression and coordinated induction of genes encoding detoxifying enzymes. Oncogene. 17:3145-3156.

Vollrath, V., A.M. Wielandt, M. Iruretagoyena, and J. Chianale. 2006. Role of Nrf2 in the regulation of the Mrp2 (ABCC2) gene. The Biochemical journal. 395:599- 609.

140

Wang, J., L. Liu, H. Qiu, X. Zhang, W. Guo, W. Chen, Y. Tian, L. Fu, D. Shi, J. Cheng, W. Huang, and W. Deng. 2013. Ursolic Acid Simultaneously Targets Multiple Signaling Pathways to Suppress Proliferation and Induce Apoptosis in Colon Cancer Cells. PloS one. 8.

Wang, M.H., W. Lee, Y.L. Luo, M.T. Weis, and H.P. Yao. 2007a. Altered expression of the RON receptor tyrosine kinase in various epithelial cancers and its contribution to tumourigenic phenotypes in thyroid cancer cells. J Pathol. 213:402-411.

Wang, X.J., J.D. Hayes, C.J. Henderson, and C.R. Wolf. 2007b. Identification of retinoic acid as an inhibitor of transcription factor Nrf2 through activation of retinoic acid receptor alpha. Proc Natl Acad Sci U S A. 104:19589-19594.

Wang, X.J., Z. Sun, W. Chen, Y. Li, N.F. Villeneuve, and D.D. Zhang. 2008. Activation of Nrf2 by arsenite and monomethylarsonous acid is independent of Keap1-C151: enhanced Keap1-Cul3 interaction. Toxicol Appl Pharmacol. 230:383-389.

Wondrak, G.T., N.F. Villeneuve, S.D. Lamore, A.S. Bause, T. Jiang, and D.D. Zhang. 2010. The cinnamon-derived dietary factor cinnamic aldehyde activates the Nrf2-dependent antioxidant response in human epithelial colon cells. Molecules. 15:3338-3355.

Wong, C., and S. Chen. 2012. The development, application and limitations of breast cancer cell lines to study tamoxifen and aromatase inhibitor resistance. J Steroid Biochem Mol Biol. 131:83-92.

Wu, K.C., J.Y. Cui, and C.D. Klaassen. 2012. Effect of graded Nrf2 activation on phase-I and -II drug metabolizing enzymes and transporters in mouse liver. PloS one. 7:e39006.

Wu, X., and G. Brewer. 2012. The regulation of mRNA stability in mammalian cells: 2.0. Gene. 500:10-21.

Yang, J., S.A. Mani, J.L. Donaher, S. Ramaswamy, R.A. Itzykson, C. Come, P. Savagner, I. Gitelman, A. Richardson, and R.A. Weinberg. 2004. Twist, a master regulator of morphogenesis, plays an essential role in tumor metastasis. Cell. 117:927-939.

Yang, M., Y. Yao, G. Eades, Y. Zhang, and Q. Zhou. 2011. MiR-28 regulates Nrf2 expression through a Keap1-independent mechanism. Breast Cancer Res Treat. 129:983-991.

Yang, Z., and D.J. Klionsky. 2010. Mammalian autophagy: core molecular machinery and signaling regulation. Current opinion in cell biology. 22:124-131.

Yates, M.S., Q.T. Tran, P.M. Dolan, W.O. Osburn, S. Shin, C.C. McCulloch, J.B. Silkworth, K. Taguchi, M. Yamamoto, C.R. Williams, K.T. Liby, M.B. Sporn, T.R. Sutter, and T.W. Kensler. 2009. Genetic versus chemoprotective

141

activation of Nrf2 signaling: overlapping yet distinct gene expression profiles between Keap1 knockout and triterpenoid-treated mice. Carcinogenesis. 30:1024-1031.

Zhang, D.D., and M. Hannink. 2003. Distinct cysteine residues in Keap1 are required for Keap1-dependent ubiquitination of Nrf2 and for stabilization of Nrf2 by chemopreventive agents and oxidative stress. Molecular and cellular biology. 23:8137-8151.

Zhang, D.D., S.-C. Lo, J.V. Cross, D.J. Templeton, and M. Hannink. 2004. Keap1 is a redox-regulated substrate adaptor protein for a Cul3-dependent ubiquitin ligase complex. Molecular and cellular biology. 24:10941-10953.

Zhang, D.D., S.C. Lo, Z. Sun, G.M. Habib, M.W. Lieberman, and M. Hannink. 2005. Ubiquitination of Keap1, a BTB-Kelch substrate adaptor protein for Cul3, targets Keap1 for degradation by a proteasome-independent pathway. The Journal of biological chemistry. 280:30091-30099.

Zhang, P., A. Singh, S. Yegnasubramanian, D. Esopi, P. Kombairaju, M. Bodas, H.L. Wu, S.G. Bova, and S. Biswal. 2010. Loss of Kelch-Like ECH-Associated Protein 1 Function in Prostate Cancer Cells Causes Chemoresistance and Radioresistance and Promotes Tumor Growth. Molecular Cancer Therapeutics. 9:336-346.

Zhang, Y.S., J.Z. Xie, J.L. Zhong, Y.Y. Li, R.Q. Wang, Y.Z. Qin, H.X. Lou, Z.H. Gao, and X.J. Qu. 2013. Acetyl-11-keto-beta-boswellic acid (AKBA) inhibits human gastric carcinoma growth through modulation of the Wnt/beta-catenin signaling pathway. Biochim Biophys Acta. 1830:3604-3615.

Zhou, D.J., D. Pompon, and S.A. Chen. 1990. Stable expression of human aromatase complementary DNA in mammalian cells: a useful system for aromatase inhibitor screening. Cancer Res. 50:6949-6954.

Zhou, W., S.C. Lo, J.H. Liu, M. Hannink, and D.B. Lubahn. 2007. ERRbeta: a potent inhibitor of Nrf2 transcriptional activity. Mol Cell Endocrinol. 278:52-62.

Zhu, H., and Y. Li. 2012. NAD(P)H: quinone oxidoreductase 1 and its potential protective role in cardiovascular diseases and related conditions. Cardiovasc Toxicol. 12:39-45.

142