CHARACTERIZATION, EPIGENETIC DRUG EFFECT, and GENE DELIVERY to BREAST CANCER CELLS a Dissertation Presented to the Graduate Facu

Home , Cascade effect (ecology), KIF4A, SH3GLB2

CHARACTERIZATION, EPIGENETIC DRUG EFFECT, AND GENE DELIVERY TO

BREAST CANCER CELLS

A Dissertation

Presented to

The Graduate Faculty of The University of Akron

In Partial Fulfillment

of Requirements for the Degree

Doctor of Philosophy

Shan Lu

December, 2015

CHARACTERIZATION, EPIGENETIC DRUG EFFECT, AND GENE DELIVERY TO

BREAST CANCER CELLS

Shan Lu

Dissertation

Approved: Accepted:

Advisor Department Chair Dr. Vinod Labhasetwar Dr. Stephen Weeks

Committee Chair Dean of the College Dr. Coleen Pugh Dr. John Green

Committee Member Dean of Graduate School Dr. Abraham Joy Dr. Chand Midha

Committee Member Dr. Ali Dhinojwala

Committee Member Dr. Anand Ramamurthi

Committee Member Dr. Peter Niewiarowski

ii ABSTRACT

Cancer relapse is strongly associated with the presence of cancer stem cells (CSCs), which drive the development of metastasis and drug resistance.

In human breast cancer, CSCs are identified by the CD44+/CD24- phenotype and characterized by drug resistance, high tumorigenicity and metastatic potential. In this study, I found that MCF-7/Adr cells that are breast cancer cells resistant to doxorubicin

(Dox) uniformly displayed CSC surface markers, possessed CSC proteins, formed in vitro mammospheres, yet retained low migratory rate. They were also able to self-renew and differentiate under floating culture condition and are responsive to epigenetic drug treatment.

High degree of DNA methylation (modifications of the cytosine residues of DNA) and histone deacetylation are major epigenetic landmarks of CSCs. In this work, I showed that MCF-7/Adr cells are sensitive to histone deacetylation inhibitor suberoylanilide hydroxamic acid (SAHA). Through RNA-sequencing technology, I also found that decitabine (DAC) and SAHA similarly affected a large number of the examined pathways, including drug and nanoparticle cellular uptake and transport, lipid metabolism, carcinogenesis and nuclear transport pathways. However, DAC significantly down-regulated several drug metabolism and drug resistant genes whereas SAHA showed no effect on them.

iii siRNA therapy is the only approved genomic drug up to date; and is under clinical trials for cancer treatment. Cationic polymers are effective in complexing nucleotide through electrostatic interactions, forming polyplexes. Therefore, I further investigated the ability of a polyethyleneimine (PEI)-based cationic polymer, polyethylene glycol

(Peg) modified L-arginine oligo (alkylaminosiloxane)-graft PEI ((P(SiDAAr)5Peg3), in delivering siRNA. The delivery system enabled effective siRNA delivery in vitro. In vivo in xenograft tumor model, polyplex showed insignificant reduction in bioluminanance signal following single-dose intravenous administration than other control groups but the data were not statistically significant, suggesting that further dose optimization of polyplex required to achieve a greater effect. In a drug resistant cell line, the delivery system also enabled anti-ABCB1 siRNA to effectively knockdown its target ABCB1 protein and enhance the Dox uptake as well as its therapeutic efficacy. Furthermore, I conducted DNA-siRNA co-delivery using P(SiDAAr)5Peg3, and obtained significantly enhanced transfection levels for both DNA and siRNA using the co-delivery approach.

In summary, this study identified stem-like characteristics in drug resistant cell line

MCF-7/Adr, showed that epigenetic drugs DAC and SAHA are effective in targeting dug resistant cells, investigated the transcriptomic effects of DAC and SAHA through RNA sequencing, and developed a siRNA and DNA-co-siRNA delivery system that effectively transfected both metastatic and drug resistant breast cancer cell lines in vitro.

iv DEDICATION

This dissertation is dedicated to my parents Li Wang and James Halpert, and my grandparents Guimin Wang and Fengxian Yang, for their support and unending encouragement.

-Shan Lu

v ACKNOWLEDGEMENTS

I would like to express my gratitude to all those who have helped and supported me in my graduate work. First and foremost, I would like to thank my advisor Dr. Vinod

Labhasetwar for his generosity in providing his lab for me to conduct my doctoral study and research. Furthermore, thank you especially for providing me the opportunity to conduct research in my area of interest, the freedom to independently construct my project, tackle scientific issues, formulate hypothesis and test them. Thank you for guiding me, helping me to revise manuscript drafts and providing strategic guidance. I appreciate your patience, knowledge and best practices you shared with me through the 5 years of your mentorship. Your investment in my professional development was instrumental and I would not have been where I am without you.

Dr. Pugh Coleen, thank you for serving as my committee chair and for your endless support and guidance. I thank you for the times you invited me to your house for thanksgiving dinner even after I had left Akron, and the times you took me out for dinner and shared your life experience with me. I appreciate your sincerity, openness, and kindness. I will always remember the eggs I got from Mad. They were delicious. It’s an honor to have you as my committee chair.

I also want to express my sincerest appreciation to my committee: Dr. Abraham Joy,

Dr. Ali Dhinojwala, Dr. Anand Ramamurthi and Dr. Peter Niewiarowski. You are the

vi most wonderful committee members I have ever known. Your willingness to help me with any issues, your patience and encouragement has helped me significantly throughout my graduate career. It’s a true privilege to have you all on my committee, and I’m grateful for all your contributions from the deepest part of my heart.

I thank the entire Labhasetwar lab, past and present, for all their generous sharing of expertise and support in completing my graduate studies. They are the most supportive, kind and pleasant people I’ve ever worked with. I want to thank Melinda Stees, Samantha

Cramer, Dr. Yue Gao, Dr. Youzhi Kuang, Dr. Vijay Raghavan, and Dr. Jun Yang. I also thank my former labmates: Dr. Chiranjeevi Peetla, Dr. Sivakumar Vijayraghavalu, Dr.

Isaac M. Adjei, Dr. Viola Morris, Dr. Susan Foy, Dr. Blanka Sharma, and Dr. Jaffer

Hayder for their help and feedback with various experiments.

I also thank the cores and the staff at the Cleveland Clinic Foundation, specifically, I thank Mei Yin, Dr. Judith A. Drazba, Eric Diskin and Sage O'Bryant, for their assistance with my experiments.

I appreciate the Biology Department and the Integrated Bioscience Program at the

University of Akron for providing a supportive environment as well as the financial support and the opportunity afforded to carry out my graduate studies. I deeply value the knowledge I gained from the Department of Polymer Science faculty and staff members.

vii Last but not the least, I want to thank my parents, James Halpert and Li Wang, and grandparents, Guimin Wang and Fengxian Yang, and other family members for supporting me through this significant time of my life. Mom, thank you for always being there for me whenever I needed you. Thank you for providing valuable insights, suggestions, and honest opinions on all the matters we discussed. I would not be who I am without you. Jim, it is a privilege to have you in my family, and thank you for your encouragement through text messages, emails, e-cards which have supported me through the good and bad times.

I also thank all my friends, Yu Pei, Yan Xing, Jinjin Ma, Zhaosheng Gao, and others for sharing the precious moments of life with me.

This study was funded by grant R01 CA149359 (to VL) from the National Cancer

Institute of the National Institutes of Health.

viii TABLE OF CONTENTS

Page

LIST OF FIGURES...... xv

LIST OF SCHEMES...... xviii

LIST OF TABLES...... xix

CHAPTER

I. INTRODUCTION...... 1

1.1 Breast Cancer...... 1

1.1.1 Clinical Hurdles for Breast Cancer Treatment...... 2

1.1.2 Drug Resistance...... 2

1.1.3 Metastasis...... 3

1.2 Mechanisms of Cancer Progression...... 4

1.2.1 Genetic Mutations...... 4

1.2.2 Epigenetic Alterations...... 5

1.2.3 Epithelial-Mesenchymal Transition (EMT) and Cancer Stem Cells (CSCs)...... 6

1.3 Treatment Options for Breast Cancer...... 7

1.3.1 Hormone Therapy...... 7

1.3.2 Chemotherapy...... 8

1.3.3 Epigenetic Drug Therapy...... 8

1.3.4 Gene Therapy...... 13

ix 1.3.4.1 siRNA Therapy...... 14

1.3.4.2 Gene Delivery Nano-carriers...... 15

1.4 Deep-Sequencing...... 16

1.5 Hypothesis...... 17

1.6 Specific Aims...... 17

1.7 General Dissertation Overview...... 17

II. DRUG RESISTANT BREAST CANCER CELL LINE DISPLAYS CANCER STEM CELL PHENOTYPE AND RESPONDS SENSITIVELY TO EPIGENETIC DRUGS SAHA AND DAC...... 19

2.1 Introduction...... 19

2.2 Materials and Methods...... 21

2.2.1 Materials...... 21

2.2.2 Cell Culture...... 22

2.2.3 Flow Cytometric Analysis...... 22

2.2.4 Mammosphere Formation Assay...... 23

2.2.5 Migration/Invasion Assay………...... 24

2.2.6 In vitro Proliferation Assay...... 24

2.2.7 Immunofluorescence Imaging...... 24

2.2.8 Statistical Analysis...... 25

2.3 Results...... 25

2.3.1 Immunophenotype of the Stem-Like Cancer Cells...... 26

2.3.2 Mammosphere Formation Shows Distinct Cellular Properties...... 27

2.3.3 Short-Term SAHA Treatment Reverses Drug Resistance but not BCSC Immunophenotype...... 32

2.3.4 Cells Affected Differently by SAHA in their Migratory Ability...... 37

x 2.3.5 Expression of E-cadherin, fibronectin and vimentin in MCF-7/Adr cells....37

2.4 Discussion...... 38

2.5. Conclusions...... 44

2.6. Acknowledgements...... 44

III: RNA-SEQUENCING REVEALS DRUG RESISTANCE TRANSCRIPTOMIC SIGNATURES AND PATHWAYS REVERSED IN RESISTANT BREAST CANCER CELLS BY EPIGENETIC DRUGS...... 45

3.1 Introduction...... 45

3.2 Materials and Methods...... 47

3.2.1 Materials...... 47

3.2.2 Cell Culture...... 47

3.2.3. Cell Viability...... 48

3.2.4. RNA Sequencing and Analysis of the Significantly Regulated Genes...... 48

3.2.5. Statistical Analysis...... 49

3.3 Results...... 50

3.4 Discussion...... 69

3.5 Conclusions...... 81

3.6 Acknowledgements...... 82

IV: EFFECTIVENESS OF SIRNA DELIVERY VIA ARGININE-RICH PEI-BASED POLYPLEX...... 83

4.1 Introduction...... 83

4.2 Materials and Methods...... 85

4.2.1 Materials...... 85

4.2.2 Synthesis of P(SiDAAr)5Peg3...... 86

xi 4.2.3 TRITC Conjugation of P(SiDAAr)5Peg3...... 86

4.2.4 Polymer Complexation and Polyplexes Characterization...... 87

4.2.5 Determination of the Complete siRNA to P(SiDAAr)5Peg3 Complexation Ratio...... 88

4.2.6 Protection from siRNA RNase Degradation...... 88

4.2.7 Polyplexes Stability Evaluation...... 89

4.2.8 Cell Culture...... 89

4.2.9 Cytotoxicity Evaluation...... 89

4.2.10 Polyplexes Intracellular Accumulation by Flow Cytometry...... 90

4.2.11 siRNA-mediated Gene Silencing in vitro...... 90

4.2.12 Testing Target Protein Knockdown Effectiveness in Enhancing Dox Uptake and as Chemotherapy...... 91

4.3 Results...... 92

4.4 Discussion...... 98

4.5 Conclusions...... 105

4.6 Acknowledgements...... 106

V: CO-DELIVERY OF DNA AND SIRNA VIA ARGININE-RICH PEI-BASED POLYPLEXES...... 107

5.1 Introduction...... 107

5.2 Materials and Methods...... 109

5.2.1 Materials...... 109

5.2.2 pDNA Amplification and Purification...... 110

5.2.3 Polymer Synthesis...... 110

5.2.4 Polymer/Nucleotide Complexation...... 111

5.2.5 Size/Zeta Potential Measurements...... 112

xii 5.2.6 Gel Retardation Assay...... 112

5.2.7 Transmission Electron Microscopy (TEM)...... 113

5.2.8 Polyanion Competition Assay...... 113

5.2.9 Cell Culture...... 113

5.2.10 In Vitro Cytotoxicity Assay...... 114

5.2.11 In Vitro Luciferase Transfection Assay...... 115

5.2.12 DNA Cellular Uptake/Flow Cytometry...... 116

5.2.13 DNA Intracellular Visualization...... 116

5.2.14 Statistical Analysis...... 117

5.3 Results and Discussion...... 117

5.4 Conclusions...... 151

5.5 Acknowledgements...... 151

VI: SUMMARY, CONCLUSIONS AND FUTURE DIRECTION...... 152

6.1 Summary...... 152

6.2 Limitations...... 157

6.3 Final Conclusion and Future Directions...... 159

REFERENCES...... 161

APPENDICES...... 178

APPENDIX A: LIST OF ABBREVIATIONS...... 179

APPENDIX B: THE REDUCTION OF LUCIFERASE PROTEIN EXPRESSION IN VIVO...... 183

APPENDIX C: HYDRODYNAMIC DIAMETERS AND THE ZETA POTENTIALS OF POLYMERS AND POLYPLEXES ………………………...... 186

APPENDIX D: SYNTHESIS OF P(SIDAAR)5...... 193

xiii LIST OF FIGURES

Figure Page

1.1 Mechanisms of drug resistance in cancer cells...... 4

1.2 Epigenetic illustration of gene regulation...... 10

1.3 Concurrent widespread changes in gene expression with epigenetic therapy...... 11

1.4 The overview of epithelial-mesenchymal transition (EMT) and metastasis signaling networks...... 12

2.1 Drug resistant breast cancer cell lines characterization uncovers breast cancer stem cell immunophenotypes...... 30

2.2 Immunophenotype changes after mammosphere culture condition...... 34

2.3 SAHA treatment alterations on cell surface markers...... 36

2.4 Migration Assay assesses the migratory potential of SAHA in 24 hours...... 39

2.5 Comparative illustration of SAHA resistance in different cell lines...... 40

2.6 Immunoﬂuorescence staining examines the presence of E-cadherin, fibronectin and vimentin with or without 5 days of SAHA treatment in MCF-7/Adr...... 41

3.1 The structure of (A) DAC and (B) SAHA...... 52

3.2 The cytotoxicity of epigenetic drugs in cell line MCF-7/Adr...... 53

3.3 Outline of the tools and pipeline used to conduct RNA-Seq transcriptomic analysis...... 54

3.4 Overview of the significantly altered transcriptomes in MCF-7/Adr after SAHA and DAC treatment...... 66

3.5 Overview of alternatively expressed genes in (A) DAC and (B) SAHA treated cells...... 67

3.6 Gene regulation overview in selected molecular functions...... 70

xiv 3.7 The effect of DAC and SAHA on endocytotic regulation...... 71

3.8 Visualization (A) and gene fold changes (B) in PPAR signaling pathway...... 72

3.9 (A) Visualization and (B) list of altered genes in mTOR signaling pathway...... 73

3.10 RAN signaling pathway in SAHA-treated cells...... 74

4.1 Morphological analysis of siRNA polyplexes...... 94

4.2 P(SiDAAr)5Peg3 completely complexed siRNA and protected it from RNAse degradation at N/P ratio of 1/4...... 95 4.3 Polyanion competition assay reveals the polyplex stability...... 96

4.4 Polyplexes cellular uptake ...... 99

4.5 Cytotoxicity and transfection efficiency of siRNA/P(SiDAAr)5P3 polyplex in MDA- MB-231-luc-D3H2LN cell line...... 101

4.6 Effect of ABCB1 siRNA on ABCB1 surface protein, Dox accumulation and cytotoxicity...... 103

5.1 Chemical structures of starting materials for the polymer P(SiDAAr)5Peg3...... 125

5.2 Spectral characterization of P(SiDAAr)5Peg3...... 127

5.3 Polyplex agarose gel electrophoresis identifies the un-complexed nucleotides...... 130

5.4 Polyplexes characterization via TEM...... 131

5.5 Stability of polyplexes measured through the polyanion competition assay...... 134

5.6 Polymer and polyplex cytotoxicity assays in drug-sensitive and resistant cell lines...... 135

5.7 DNA transfection levels of both PEI and P(SiDAAr)5Peg3 luc-DNA-co-s-siRNA polyplexes in MCF-7 and MCF-7/Adr cells ...... 137

5.8 Polyplexes’ siRNA transfections in MDA-MB-231-luc-D3H2LN cell line……….143

5.9 P(SiDAAr)5Peg3 polyplexes’ intracellular DNA quantification in MCF-7 cell line...... 144

xv 5.10 P(SiDAAr)5Peg3 polyplexes’ intracellular DNA quantification in MCF-7/Adr cell line………………………………………………………………………………………145

B.1 Examining the in vivo effectiveness of luc-siRNA P(SiDAAr)5Peg3 polyplex...... 185

xvi LIST OF SCHEMES

Scheme Page

5.1 Synthesis of Polyethylene glycol modified L-arginine oligo (alkylaminosiloxane) grafted poly(ethylenimine) P(SiDAAr)5Peg3...... 126

D.1 Synthesis of (A) oligo(-alkylaminosiloxanes) (SiDA), (B) L-arginine modified oligo (-alkylaminosiloxanes) (SiDAAr5), (C) L-arginine oligo (alkylaminosiloxane) grafted poly(ethylenimine) (P(SiDAAr)5)...... 197

xvii LIST OF TABLES

Table Page

2.1 Human Breast Cancer Cell Lines Differentially Express Solid Breast Cancer Stem Cell Markers (% of Control)...... 28

2.2 Alterations of BCSC Markers in Adherent vs. Mammosphere Culture Conditions ... 31

2.3 BCSC Surface Marker Expression in Cell Lines after 14 days of SAHA Treatment (% of control)...... 35

3.1 The Number of Gene Alterations in Selective Pathways ...... 55

3.2 Genes Significantly Regulated in Selected Pathways Affecting Drug Resistance...... 56

3.3 Gene Fold Changes and Functions ...... 58

4.1 Average Hydrodynamic Diameter (Dh) (nm), Polydispersity Index (PDI) and Zeta Potential (mV) of P(SiDAAr)5 and P(SiDAAr)5Peg3 Polyplexes at Different siRNA to Polymer Weight Ratios. Three samples were measured per polyplex ratio...... 100

5.1 Mean Hydrodynamic Diameter (Dh) (nm) and Zeta Potential (mV) of Polymer Solutions in dH2O. Three measurements were done per polymer...... 128

5.2 Average Hydrodynamic Diameter (Dh) (nm) and Zeta Potential (mV) of PEI and P(SiDAAr)5Peg3 Polyplexes at Different Weight Ratios. Three samples were measured per polyplex ratio...... 131

xviii CHAPTER I

INTRODUCTION

1.1 Breast Cancer

Breast cancer is a leading cause of morbidity and mortality among women with approximately 14 million new cases and 8.2 million deaths in 2012 worldwide [1]. Breast cancer is caused by the mutation of breast tissues. Among all initial diagnosis, about 6%-

10% people have Stage IV breast cancer.

Breast cancer can be classified in several ways. Histologically, breast cancer can be separated into ductal, lobular, nipple, or not otherwise specified (NOS). Based on the presence of estrogen receptors (ER), progesterone receptors (PR), and human epidermal growth factor receptor-2 (HER2), breast cancer can be categorized into luminal A (ER+ and/or PR+; HER2–), luminal B (ER+ and/or PR+; HER2+), triple negative (ER–, PR–, and HER2–), and HER2 (ER–, PR–, and HER2+) type. ER-positive tumors are benign and account for over 70% of breast cancer [2, 3]. ER-negative tumors are more malignant [4]. The over-expression of HER2 has a strong association with poor survival [5].

1 1.1.1 Clinical Hurdles for Breast Cancer Treatment

Difficulties in treating breast cancer largely lie in the heterogeneous nature of the various tumor types, including the diverse morphological and molecular characteristics as well as treatment responses [6]. Current treatment challenges include cancer relapse, the development of drug resistance and metastasis.

1.1.2 Drug Resistance

In spite of the advances in chemotherapy of breast cancer, drug resistance is a major clinical challenge for patient survival [7]. Resistance is the result of both genetic and epigenetic alterations. Mechanisms of resistance to chemotherapy are complex, including increased drug efflux and decreased drug uptake, compartmentalization of drug into intracellular vesicles, impaired anti-apoptotic machinery, inability of drugs to reach the target, and increased cellular mechanism of repair of DNA damage (Fig 1.1) [8].

Therefore, the identification and reversal of drug resistance genes as well as elucidation of mechanisms of resistance is crucial for more effective treatment strategies. Currently, new therapies are under development with the aim to overcome the resistance to chemo- and radio-therapy of tumors. What are these new therapies? Either delete the line or explain

Acquired drug resistance often results from overexpression of energy-dependent membrane transporters, such as P-glycoprotein (P-gp), which pump out anticancer drugs.

Resistance can also occur through reduced intracellular drug uptake. Multidrug resistance protein 1 (MDR-1) and multidrug resistance associated protein 1 (MRP-1) are the two

most common pump-related drug resistant genes in humans [9, 10]. MCF-7/Adr, a widely studied multi-drug resistant breast cancer cell line is derived from its parental MCF-7 cell line through drug-induced transformation. Compared with MCF-7, which is estrogen receptor (ER) positive, MCF-7/Adr is ER negative, and not susceptible to hormone and drug therapy. Therefore, MCF-7/Adr serves as an ideal model for the understanding and conquest of drug resistance.

1.1.3 Metastasis

Metastatic breast cancer (also called stage IV or advanced breast cancer) remains a major cause of morbidity and mortality worldwide. At present, around 155,000

Americans are diagnosed with metastatic breast cancer, which leads to 40,000 deaths annually. The majority (over 80%) of metastatic breast cancer cases are detected months or years after local (stage I, II or III) breast cancer treatment. The common sites of metastasis include lymph nodes, bones and bone marrow, muscle, fatty tissue and skin, liver, lungs and brain.

One of the well-established metastatic adenocarcinoma cell lines extracted from the mammary gland is MDA-MB-231-D3H2LN, which is claudin-low, triple negative (ER–,

PR–, HER2–), and produces metastasis to clinically relevant tissues such as lymph nodes

[11].

1.2 Mechanisms of Cancer Progression

Mechanisms of cancer progress are composed of genetic mutations as well as epigenetic alterations.

1.2.1 Genetic Mutations

Hereditary accounts for between 5% and 10% of all breast cancers. Most inherited breast cancers are associated with BRCA1 (breast cancer gene 1) and BRCA2 (breast cancer gene 2).

Figure 1.1 Mechanisms of drug resistance in cancer cells.

The increase of drug efflux, decrease of drug uptake, compartmentalization and mutation/loss of cellular targets, inactivation of drugs and anti-apoptotic machinery are

some major drug resistance mechanisms of cancer cells. Adapted from Gottesman, M.M., et al., Defeating drug resistance in cancer. Discov Med, 2006. 6(31): p. 18-23. [8].

Other genetic changes in breast cancer include ABC transporters, genes regulated by calcium signaling, genes associated with apoptosis, ion channels (ex. potassium and chloride), Ras family and glutathione S-transferase [12]. Mutations in all these genes contribute to the advancement of breast cancer. Furthermore, poor breast cancer prognosis has been shown to correlate with gene mutations, including PACSIN3, OGG1,

KCNN4, and CALB2 [12].

1.2.2 Epigenetic Alterations

Epigenetic reprogramming is involved in a number of diseases, including neurodegenerative disorders (ex. schizophrenia or Alzheimer’s disease) and cancer [13-

15]. Epigenetics was originally defined by C.H. Waddington as interactions between genes and their products which generate their phenotype [16]. A more modern definition of epigenetics is changes in gene expression independent of DNA changes, linking gene expressions with chromatin structure [15]. Recently, studies have shown an even greater role of epigenetic regulation in cellular malignant transfection and the development of breast cancer [17].

Epigenetic modulations include histone modifications, DNA methylation, and miRNA expression, which may function independently or in concert (Fig 1.2). Chromatin structure determines the accessibility of genes in a cell, thus regulating the expression

level of a gene. An ‘epigenetic landscape’ can be created by altering the local structural dynamics of chromatin, its morphology and accessibility. Histone deacetylation is a crucial mechanism of cancer progression. Histone deacetylase (HDACs) can alter the chromatin conformation and bind to proteins that affect RNA transcription, thus affecting cell growth, differentiation and apoptosis. DNA methylation is another important mechanism for control of gene transcription and is tightly connected to cell differentiation [18]. Most methylation takes place on CpG islands, the regions of high

CpG dinucleotide density located in the promoters of housekeeping genes. The methylation of those areas not normally methylated in normal cells may induce the silencing of tumor suppressor genes such as BRCA1 [19, 20]. Understanding the biomarkers and molecular changes caused by epigenetic regulation will be valuable for effective treatment of breast cancer [21].

1.2.3 Epithelial-Mesenchymal Transition (EMT) and Cancer Stem Cells (CSC)

Two new concepts in the area of breast cancer are: 1) CSC and its influence on tumor initiation, 2) EMT and its effect on the acquisition of malignant characteristics.

Epithelial-mesenchymal transition (EMT) is defined as the transitioning of a cell from an epithelial to a mesenchymal state, which is related to increasing cancer aggressiveness and malignancy [22]. EMT is triggered by hypoxia, growth factor signaling, and tumor– stromal cell interactions, and is a driver for the formation of cancer stem cells (CSC) [23].

Breast CSCs are cancer cells with stem characteristics and distinct immunological markers and may be an important source of tumor recurrence and drug resistance.

In vitro, it has been shown that growth factors such as Notch, transforming growth factor beta (TGF-beta), Wnt, and Hedgehog were able to form CSC through EMT [24], inducing mesenchymal proteins such as vimentin (VIM), tumor growth factor-β (TGF-β),

Wnt, Notch, receptor tyrosine kinases (RTKs), and tumor necrosis factor-β (TNF-β) (Fig

1.3) [25-28]. Adriamycin has been shown to induce the expression of TWIST1, an EMT protein, in MCF-7 cells [29]. Thus, it is essential that new strategies be formulated that could reverse EMT pathways, thus aiming to eliminate CSCs.

1.3 Treatment Options for Breast Cancer

The treatment of breast cancer is multidisciplinary. Common treatment methods include surgery, systemic therapy (chemotherapy, hormonal therapy) and radiation therapy. Beside the traditional treatment methods, biological drugs, epigenetic drugs, and gene therapy are also being actively investigated in their effectiveness in treating cancer.

1.3.1 Hormone Therapy

Hormone therapy is an integral part for breast cancer treatment, but is effective only for hormone-dependent cancer types. For example, tamoxifen and fulvestrant, two hormonal therapies, directly block the estrogen receptor (ER), and anastrazole works to prevent the body from producing estrogen. In spite of the success of those drugs, resistance to those treatments is often developed by patients, especially in cases of tumor metastasis.

1.3.2 Chemotherapy

For treatment of late stage cancer, chemotherapy is most common and has the potential to effectively control and eradicate cancer tissue. Chemotherapy is often used in conjunction with radiation or surgery and has the advantage of reaching cancer cells at remote locations that cannot be accessed through surgery. Chemotherapy can also be used for metastatic cancer, but cytotoxic drugs have a relatively narrow therapeutic index as a result of dose-limiting toxic effects [30]. Chemotherapy is often limited only to certain cell cycle stages, and the frequent emergence of multiple drug resistance is a major disadvantage.

1.3.3 Epigenetic Drug Therapy

Due to the limitations of radiation and traditional chemotherapy, there is increasing interest in epigenetic drugs to overcome drug resistance in cancer through differentiation therapy targeting CSCs. Epigenetic drugs are able to decrease chemoresistance, stem-like behavior, cell proliferation, metastasis, stem-like behavior, and increase immune responsiveness and apoptosis (Fig 1.4) [31]. In T-cell prolymphocytic leukemia, epigenetic therapy has shown to help in overcoming treatment induced drug resistance

[32]. Suberanilohydroxamic acid (SAHA), under the name Zolinza, was used for the treatment of cutaneous T cell lymphoma (CTCL). Currently, epigenetic drugs have shown to overcome drug resistance in cancer. Work from our group also showed that epigenetic drug decitabine (DAC) was effective in enhancing the efficacy of a chemotherapeutic doxorubicin (Dox) in breast cancer drug resistant cell line MCF-7/Adr

[33]. An example of the FDA approved epigenetic therapy is vorinostat, a histone- deacetylation inhibitor for cutaneous T-cell lymphoma.

Despite the success of epigenetic drugs such as decitabine (DAC) in clinical trials against hematologic malignancies and its antiproliferative effect in vitro, its efficacy in solid tumors has been disappointing [34]. This is due to the difference in the pharmacokinetics and pharmacodynamics of DAC between solid and hematologic malignancies. DAC has a very high clearance rate in vivo (half-life = 10-35 min) due to its low plasma protein binding and hence rapid clearance and inactivation by cytidine deaminase (CDA), found principally in the liver, as well as in granulocytes, intestinal epithelium and plasma [35]. The discrepancy between the in vitro efficacy of epigenetic drugs, including DAC, and their in vivo efficacy in solid tumor has been attributed to the slowly dividing nature of cancer cells in vivo vs. in vitro [36, 37]. Since DAC is S-phase specific, its activity in the tumor, when injected intravenously, may not last long enough for all the tumor cells to pass through the S-phase and thus for DAC to be effective. It is estimated that it takes approximately 5-15 d, depending on the tumor type and growth rate, for all the cancer cells in solid tumors to pass through S-phase [37-39]. Since DAC targets rapidly dividing cells such as bone marrow stem cells [40], its repeated and high dosing to achieve therapeutic levels in solid tumors was found to have serious side effects, such as chronic myelosuppression and leukopenia [36]. The toxicity of DAC at higher doses is due to its DNA-damaging activity rather than DNA demethylation activity [41].

Therefore, several modified epigenetic drugs which are more stable or drug delivery systems for delivery of existing epigenetic drugs are under investigations [42].

Figure 1.2 Epigenetic illustration of gene regulation.

In a cell, epigenetic modulations such as DNA methylation and chromatin structuring (e.g. chromatin condensation, histone deacetylation) are major regulators (activators and silencers) of gene transcription. Therefore, the reversal of above actions, which is the demethylation of the CpG islands and histone deacetylation inhibition could activate the otherwise silenced genes in cancer cells. Adapted by permission from Macmillan

Publishers Ltd: [Nat. Rev. Clin. Oncol.] (Azad, N. et al., The future of epigenetic therapy in solid tumors—lessons from the past) [31], copyright (2013).

Figure 1.3 The overview of epithelial-mesenchymal transition (EMT) and metastasis signaling networks.

Tumor factor-β (TGF-β), Wnt, Notch, receptor tyrosine kinases (RTKs), and tumor necrosis factor-β (TNF-β) are major pathways that induce tumorigenesis. Upon activation, they induce EMT transcription factors such as Snail, Slug, and Twist. During EMT, the cancer cell undergoes a transformation to acquire stem –like characteristics. This enables them to possess resistance to chemotherapy and radiation therapy, immunosuppression and senescence. Reproduced from Wang, Y. et al. (2013) [43].

Figure 1.4 Concurrent widespread changes in gene expression with epigenetic therapy.

The anticancer efficacy of epigenetic drugs is strongly connected with widespread changes in gene expression that affect numerous biological processes as listed in the figure. Abbreviation: EMT, epithelial–mesenchymal transition. Reproduced /adapted by permission from Macmillan Publishers Ltd: [Nat. Rev. Clin. Oncol.] (Azad, N. et al.) [31], copyright (2013).

Vorinostat is a HDAC inhibitor, promotes cell-cycle arrest. It is a strong anti-drug resistant and anti-metastatic agent [44, 45]. It acts as a zinc ion chelator to bind to the active site of histone deacetylases, and is in several phase I and II clinical trials for hematological and solid tumors [46, 47]. However, the precise mechanisms of drug action are unclear, such as the upstream transcription regulators and downstream targets by which these inhibitors exert their effect. By gaining a more detailed understanding on the mechanisms of action of the drug and therapeutic effect, better treatment strategies can be devised.

At its infancy in drug discovery, we are just starting to grasp the complexities of histone modification and the consequent changes in gene modulation. Although the epigenetic drugs have promise in reversing drug resistance, knowledge of the mechanisms of drug action and the underlying biology of cellular responses remains fragmentary.

1.3.4 Gene Therapy

Gene therapy is defined as the use of genes to treat or prevent disease. Several ways to utilize gene therapy include: 1) replace a mutated gene; 2) down-regulate an overexpressed gene; and 3) insert a new gene into the body. Gene therapy encompasses DNA, small interfering RNA (siRNA) or short hairpin RNA (shRNA), and microRNA.

There have been over 400 clinical studies in gene therapy under evaluation in the past

15 years, 70% of which are cancer-focused [48]. Breast cancer, which is caused by

multiple genetic mutations, is an ideal candidate for gene therapy. Tumor suppressors including BRCA1 have been studied widely. However, one major obstacle confronting the effective application of gene therapy is identifying suitable targets that are selective for cancer cells as opposed to non-cancerous tissue. [49]

1.3.4.1 siRNA Therapy

RNA interference (RNAi) therapeutics is revolutionary in that it represents a powerful new tool for biological research of gene silencing. Among various means of

RNAi, siRNA is the most commonly used. siRNA is a double-stranded RNA with 21-24 nucleotides derived from mRNAs or viral RNAs [50], having an average size of less than

10 nm. They target, bind and degrade complementary mRNAs through the RNA-induced silencing complexes (RISCs) in the cytoplasm [51, 52]. siRNA has enormous therapeutic potential in repressing cancer cell proliferation as well as the metastasis and drug resistance [53]. One other advantage of siRNA is there is no need to penetrate into the nucleus, which is difficult to achieve due to the tight regulation of the nuclear membrane. siRNA also exerts longer silencing activity compared to antisense oligonucleotides [54].

Upon repeated dosing, siRNA is able to exert gene silencing for a long time without influencing the endogenous miRNA functionality of the cell, due to its target specific property [55]. The effective delivery of siRNA in vivo can be used to study gene functions, validate gene targets, for gene therapy. However, disadvantages include the easy degradation of siRNA in the blood serum and the triggering of the innate immune system in vivo [56-58].

1.3.4.2 Gene Delivery Nano-carriers

A bottleneck in gene therapy lies in its delivery. Disseminated breast cancer is far from accessible. Also, there are intracellular as well as extracellular biological barriers, which are yet to be overcome. Nano-vehicles are effective in delivering gene therapy to its target tissue and assist in gene transfection. By definition, nano-carriers have dimensions in two axes of less than 1000 nm [59]. In order to be effective in delivery, they are typically smaller than 300 nm in diameter. They can be organic or inorganic, forming different classes of carriers. Much research has been dedicated to gene delivery in the past 2 decades, using vehicles such as viruses, biological carriers, polymeric carriers, and metals, etc. Viruses are effective in transfecting cells, but concerns about safety and immunogenicity made scientists shift focus to non-viral carriers [60]. Some of the materials used include chitin, gelatin (natural), and polylactides and poly alkyl cyanoacrylates (synthetic). Among these, cationic polymers, peptides and liposomes were all investigated [61].

Cationic vehicles have the advantage of forming electrostatic complexation with anionic nucleotides. For example, polyethylenimine (PEI) complexes siRNA through cationic functional groups (amines), bind the siRNA phosphate group through electrostatic interactions, and also play an important role in the polyplexes uptake [62].

Although showing good siRNA condensation and uptake [63, 64], the stability of PEI as well as transfection efficiency in vivo are also major hurdles in its clinical applications

[65, 66]. Also its toxicity level and rapid clearance adversely narrowed the dosing capacity and the therapeutic window. PEGylation-the covalent attachment of

polyethylene glycol (PEG), and a number of other chemical modifications have been done to PEI to exert a shielding effect on the positively charged polyplexes with the goal of reducing toxicity and improving complex stability and hydrophilicity [67].

1.4 Deep-Sequencing

The transcriptome is the complete set of cellular transcripts under a specific condition.

“Whole Transcriptome Shotgun Sequencing", also called deep sequencing, is an emerging technology to globally assess the RNA content from the sequencing of cDNA

[68]. Through RNA-seq, we can obtain an unbiased, global understanding of the transcriptome and gain a more precise measurement of transcript levels, which is essential for interpreting the functional elements of the genome and their molecular regulations. This technology is well adapted to cancer studies and is especially advantageous because it provides deep coverage at base-level resolution [69]. Also,

RNA-seq it gives information on genes’ quantitative changes under different conditions, differential splicing, non-coding RNAs, post-transcriptional modifications, and has been seen as "a revolutionary tool for transcriptomics" [70].

Recent advances in high-throughput sequencing have enabled genome-wide mapping of chromatin changes occurring during tumorigenesis, including a global loss of acetylated H4-lysine 16 and H4-lysine 20 trimethylation [71]. Such loss of histone acetylation caused by HDACs is a reason behind repression of genes. The global transcriptomic profiling enables a very high resolution of the intricate inner workings of

the transcriptome, enabling us to explore the distinct nature of each cancer landscape, and provide a new tool for the analysis of therapeutic targets [72].

1.5 Hypothesis

I hypothesize that CSC characteristics can be identified in breast cancer cell lines; epigenetic drugs are capable of reversing stem-like characteristics; through global transcriptomics analysis, I can elucidate new genes and pathways used by epigenetic drugs to reverse drug resistance; finally, effective siRNA and DNA-siRNA co- transfection can be achieved using arginine and polyethylene glycol (PEG) modified cationic polymer.

1.6 Specific Aims

Aim 1. To examine the CSC characteristics in breast cancer cell lines and the effect of epigenetic drug SAHA.

Aim 2. To analyze the transcriptomic alterations in drug resistant breast cancer cells as a result of epigenetic drug treatments.

Aim 3. To generate a cationic construct to facilitate siRNA and DNA-co-siRNA delivery.

1.7 General Dissertation Overview

This project combined innovative and integrative approaches involving molecular and cellular biology, bioinformatics, and gene transfection using nano-vehicles. In this study, I observed significant changes in the metastatic potential, drug resistance level, tumorigenic potential, as well as morphological changes due to the epigenetic drugs DAC

and SAHA treatments. Using high-seq analysis, I unraveled genes and pathways that were differentially regulated by DAC and SAHA. The effectiveness of gene therapy lies in having delivery vehicles that are effective to transfect tissues of interest. By optimizing the construct formulation, I developed a novel delivery system with an effective nucleotide condensation and endocytotic rate. The formed polyplexes were able to transfect both metastatic and drug resistant breast cancer cell lines in vitro and insignificantly reduce luminescence level in vivo in the luc-MDA-MB-231 xenograft tumor model.

In the next 4 chapters, this dissertation tests the overall hypothesis and its specific aims. Chapter 2 identifies CSC phenotype from drug resistant breast cancer cells with induced drug resistance, and illustrates the effects of SAHA in combating those cells.

Chapter 3 focuses on the elucidation of pathways and mechanisms through deep- sequencing technology. Chapter 4 and 5 describe a novel vehicle for siRNA delivery and

DNA-siRNA co-delivery.

CHAPTER II

DRUG RESISTANT BREAST CANCER CELL LINE DISPLAYS CANCER STEM

CELL PHENOTYPE AND RESPONDS SENSITIVELY TO EPIGENETIC DRUGS

SAHA AND DAC

Keywords: surface markers; epithelial mesenchymal transition; epigenetic therapy; cancer therapy

2.1 Introduction

According to cancer stem cell (CSC) theory, CSCs are the only cell population in a tumor that causes tumorigenesis, cancer progression and metastasis, and resistant cancer chemotherapy. They are able to self-renew and differentiate, possessing stem cell characteristics [73]. They are also responsible for tumor re-growth in post-treatment [74].

First identified in acute myeloid leukemia, CSCs and the existence of multiple stem cell classes was evidenced by the fact that only ~1% of leukemic cells (signifying clonogenic progenitors) are able to proliferate, distinguishing them from the majority of cells [75,

76]. Soon afterward, CSCs were reported in solid tumors. Many identification systems and treatments then followed, focusing on conquering CSCs, which are drug resistant

[77-83]. However, gaps exist in the CSC theory, such as the lack of an established CSC hierarchy in a solid tumor and linking cellular lineage with their functions. Also, it is difficult to establish connections between CSCs and normal stem cells in a tissue, and

trace back the cellular origin of the identified CSCs. It has been demonstrated that the selected populations possess sphere-forming ability, able to self-renew (making identical replicates), and express a high level of malignancy which allow them to resist treatment and initiate and re-initiate tumors after chemo and radiation therapy.

The most widely recognized breast CSC (BCSC) markers are CD44+/CD24-/low [84-

89]. Subsequent studies have distinguished CD44-/CD24+ as luminal cells while

CD44+/CD24- possess basal-like phenotype [76]. Heterogeneous (CD44+/CD24+ or

CD44-/CD24-) expressions were also observed in the tumors [90]. The CD44+/CD24- sub-population of a tumor are shown to be able to initiate and propagate a tumor, and has been used to signify its ability to self-renew. The same cell population has also displayed considerable chemo-resistance and distant metastasis [81, 91].

Beside CD44+/CD24-, a number of other markers have also shown connection with tumor-initiation and treatment survival, most of which have to do with cellular adhesion and mobility. Epithelial-specific antigen (ESA), CD49f and CD201 are a few that have been identified for BCSC. ESA has been shown to support migration and invasion in human metastatic cell line MDA-MB-231 [92], while CD49f is connected with cell adhesion and surface signaling. In this research, for the first time, I demonstrated the uniform and prolonged display of CD44+/CD24- phenotype along with the above- mentioned BCSC markers in a well-established multi-drug resistant breast cancer cell line MCF-7/Adr without special supplements or culture conditions, bringing new perspectives on the current knowledge of CD44+/CD24- phenotype, and that of BCSC.

20 Global epigenetic mutation is a significant landmark for cancer. Epigenetic drugs are promising in overcoming and reversing cancer malignancy through cellular differentiation targeting CSCs. Epigenetic drug suberoylanilide hydroxamic acid (SAHA), approved to treat cutaneous T-cell lymphoma, is a strong anti-drug resistant and anti- metastatic agent [44]. Taking into account the transient therapeutic nature of SAHA, I further explored its effect to alter the cell surface markers through short-term drug exposure, the understanding of which could profoundly affect the targeting and selecting strategies toward malignant cancer types especially utilizing epigenetic drugs as the combination therapy.

2.2 Materials and Methods

The materials and the methods to characterize the breast cancer cell lines MCF-7, MCF-

7/Adr, MDA-MB-231, SkBr3 were listed below.

2.2.1 Materials

Human breast cancer cell lines MCF-7, MDA-MB-231, SkBr3 were obtained from

ATCC (American Type Culture Collection). DCIS (ductal carcinoma in situ) was purchased from Celprogen (San Pedro, CA). MCF-7/Adr cell line was established through long-term doxorubicin incubation. Human mesenchymal stem cell (hMSC), hMSC basal medium, MSCGM bullet kit (PT-3238 & PT-4105) and hMSC trypsin/EDTA were purchased from Lonza Inc. (Walkersville, MD). APC Mouse Anti-

Human CD44, PE Mouse Anti-Human CD24, FITC Mouse Anti-Human CD326 (ESA),

PE Mouse Anti-Human CD201, PE-Cy™5 Mouse Anti-Human CD49f was purchased from BD Biosciences and BD Pharmingen (San Jose, CA). Twenty four-well ultra-low binding plates were purchased from Corning (NY). Fetal bovine serum (FBS) was purchased from Gibco (Invitrogen, Milan, Italy). Crystal violet solution (HT90132-1L) and MTT reagents were purchased from Sigma-Aldrich (St. Louis, MO). B-27 Serum-

Free Supplement (50X) (17504-044) and all antibodies were obtained from Invitrogen

Inc. Human recombinant fibroblast growth factor basic (GF003) was purchased from

Millipore (Billerica, MA). Keratinocyte serum-free medium (K-SFM) and accompanying supplements (bovine pituitary extract, final concentration 50 μg/ml, recombinant human epidermal growth factor, final concentration 5 ng/ml) were obtained from

GIBCO. SAHA was purchased from Sigma-Aldrich.

2.2.2 Cell Culture

All cancer cell lines were cultured in DMEM supplemented with 15% heat- inactivated fetal bovine serum and 1% penicillin-streptomycin. Cultures were incubated at 37 °C in a humidified atmosphere containing 5% CO2 using 75T cell culture flask.

Cells were sub-cultured reaching 90% confluency. hMSCs were expanded using hMSC

Basal Medium containing MSCGM Bullet Kits and experiments were conducted in

DMEM containing 10% FBS and 1% penicillin-streptomycin. Medium was changed every 3 days and cells were sub-cultured at 85-90% confluency. hMSCs up to passage 6 were used for experiments.

2.2.3 Flow Cytometric Analysis

Breast cancer cell lines were analyzed for the expression of selected BCSC markers including CD44, CD24, CD201 CD49f and ESA. For all experiments, after trypsinization, cells were washed twice with PBS and re-suspended in staining buffer with fluorochrome-conjugated antibody and incubated on ice in the dark for 30 min with antibody solution in flow butter for 30 min on ice bath. The flow buffer consists of PBS with 10% FBS and 0.1% NaN3 sodium azide. Antibody concentrations were consistent throughout experiments. After staining, all samples were washed thrice with PBS, re- suspended in flow buffer and tested using LSR II. Cells were shielded from light at all times. Cells from each cell line were treated with the same procedure but without antibody to serve as the control sample. The samples were analyzed with FlowJo software. A minimum of 2000 events were recorded for each sample and analyzed.

2.2.4 Mammosphere Formation Assay

Cells were single-suspended at a density of 5000 cells in culture media in six-well ultra-low attachment plates (2.5 mL per plate). 2 culture systems were used: 1) Phenol red free serum free DMEM supplemented with 20 ng/ml human basic fibroblast growth factor, 20ng/ml human epidermal growth factor, 1X B27, 1% penicillin-streptomycin solution; 2) K-SFM with BPituitary extract and 20 ng/ml epidermal growth factor, 1% penicillin-streptomycin solution. Culture media were replenished every 3 days and images were taken at day 7. After 7 days, mammospheres were collected and cells were dissociated for flow cytometry.

2.2.5 Migration/Invasion Assay

Twenty four-well Falcon tissue culture plates with 8 micron cell culture insert was used for migration assay. Cells were first starved overnight and plated on top of the insert and allowed to migrate for 24 hours, using 5% FBS in DMEM as attractant. The assay was terminated by fixing cells with 4% paraformaldehyde on ice. Cells were stained with crystal violet and cells on top of the insert were removed with a cotton swab. Images were taken after staining. For quantification, cells migrated through the insert were lysed with RIPA buffer and lysate were quantified in 96 well-plate using a plate reader.

2.2.6 In vitro Proliferation Assay

MTT assays were conducted according to the manufacturer’s protocol. Cells were plated at 3000 cells/well in 96-well plates and cultured with the drugs for 5 days. After the test was completed, 26 µL of 2 mg/mL of MTT (3-(4,5-dimethyl-2-thiazolyl)-2,5- diphenyl-2H-tetrazolium bromide; (Sigma-Aldrich) was added to each well and incubated at 37 C for 4.5 hrs for crystallization. After crystallization, DMSO (100

µL/well) was added to each well. After the crystals were dissolved completely, 90 µL were transferred to a clean plate and the absorbance was measured using plate reading at absorbance of 570 nm (Molecular Devices, Sunnyvale, CA).

2.2.7 Immunofluorescence Imaging

Immunofluorescence staining was performed on cell culture grown on cover slides.

Cells were allowed to grow on cover-slides with or without SAHA for five days.

Afterward, cells were washed with PBS and fixed with 4% paraformaldehyde (for E-

cadherin and fibronectin). For vimentin staining, cells were fixed using 100% ice-cold methanol on ice for 7 minutes. Unspecific binding sites were blocked with 3% goat serum diluted in PBS for 15 minutes at room temperature. After that, the cells were stained with primary antibodies (mouse anti-human E-cadherin 1:500 v/v, mouse anti- human fibronectin 1:250 v/v, mouse anti-human vimentin 1:250 v/v) diluted in blocking buffer overnight at 4 C, washed three times and replaced with secondary antibody (FITC conjugated goat anti-mouse IgG, IgM) for 45 minutes in the dark at room temperature.

Secondary antibodies were diluted in 1:5000. After ﬁve washes of 10 minutes each, the cells were mounted with vectashield with DAPI (Vector Laboratories, Burlingame, CA) and viewed under the fluorescence microscopy.

2.2.8 Statistical Analysis

Statistical analysis was conducted via student t-test. P ≤0.05 was used to define statistical significance. Cellular density was converted to percentage of the control and represented as mean ± standard error of mean.

2.3 Results

Results of the studies were shown below. In this section, immunophenotype, mammosphere formation, 24 hr and 14 days SAHA treatment, and migration assay of the cells treated with SAHA were examined.

2.3.1 Immunophenotype of the Stem-Like Cancer Cells

Surface markers (CD24, CD44, CD49f, CD201, and ESA) were selected according to current BCSC identification systems. We examined the presence of above markers using flow cytometry in cell lines MCF-7, MCF-7/Adr, MDA-MB-231, SkBr3, and DCIS

(ductal carcinoma in situ cell line) (Fig. 2.1, Table 2.1).

Cell lines were found to have distinct CD44/CD24 immunophenotypes (Fig. 2.1A-B,

Table 2.1). In MCF-7/Adr, differential expressions between passages were observed- passage 6 negatively expresses CD24 while passage 33 displays over 30% of CD24+ cells

(Fig. 2.1A-B, Table 2.1). We suspect this relates to the gradual loss of drug resistance, thus phenotypic reversal at higher passages. MCF-7 displayed a broad range of CD24 expression, representing the phenotypic heterogeneity within an established cell line (Fig.

2.1B, Table 2.1). However, the presence of CD49f, ESA and CD201 seem to be independent of any distinctive CD44/CD24 phenotype (Fig. 2.1, Table 2.1). Compared with its parental cell line MCF-7, MCF-7/Adr’s acquisition of drug resistance leading to its malignant transformation is marked by its expression of CD44, CD49f and CD201 and

E-cadherin to 100% and the complete negative expression of CD24.

The acquisition of CD24-/CD44+ phenotype, along with other BCSC markers, signifies the acquisition of stem-like properties in MCF-7/Adr cells. It has been reported that cells that undergo EMT also display a loss of E-cadherin and a gaining of cellular motility. However, the contrary is observed in MCF-7/Adr when compared to its parental cell line MCF-7 (Fig. 2.1). Our study is the first to report the uniform and stable display

of BCSC markers with an increased expression of E-cadherin in a cell line over a long period of time. The cell line is slow-dividing and highly chemo-resistant, while at the same time low in migration and invasiveness. For comparison purposes DCIS cell line, which negatively expresses all the above mentioned surface markers was included in the study (Fig. 2.1, Table 2.1).

2.3.2 Mammosphere Formation Shows Distinct Cellular Properties

Mammosphere formation capacity to a large extent mirrors the tumor-formation capacity in vivo. We examined both cell line’s mammosphere formation ability under different culture media and the immunophenotypes of the cells dissociated from mammospheres. Mammosphere formation starts after 2-3 days and immunophenotyping were done with cells dissociated from mammospheres at day 7 (Fig. 2.2A, Table 2.3).

Cell lines respond differently to different culture media. It is worth-noting the differential responses of MCF-7 and MCF-7/Adr, MCF-7/Adr, MDAMB-231-D3H2LN and DCIS were observed to have similar responses, while MCF-7 and MCF-7/Adr responds in an opposite manner to culture conditions (Fig. 2.2A, Table 2.3). MCF-7/Adr stayed single cell suspended in DMEM which supports MCF-7 sphere formation, but grew good mammospheres in K-SFM which did not support MCF-7 mammospheres (Fig. 2.2A,

Table 2.3).

Table 2.1 Human Breast Cancer Cell Lines Differentially Express Solid Breast Cancer Stem Cell Markers (% of Control). Cells have >90% of positive expression are counted are positive Cells have <10% of positive expression are counted are negative

CD44 CD24 CD201 CD49f ESA E-Cadherin

MCF-7 53 85 47 26 89 31

MCF-7/Adr + Passage 6: − + + + 61

Passage 33: 34%

MDA-MB

-231-D3H2LN + 75 + + 10 −

SkBr3 19 + 88 − + −

DCIS − − − − − −

Figure 2.1 Drug resistant breast cancer cell lines characterization uncovers breast cancer stem cell immunophenotypes.

Red dots are unstained and blue dots are stained cells. Marker profiling of (A) CD44, (B)

CD24, (C) ESA, (D) CD201, (E) CD49f and (F) E-cadherin were done in cell lines MCF-

7, MCF-7/Adr, MDA-MB-231, SkBr3 and DCIS by flow cytometry (see Materials and

Methods). Different CD44/CD24 combinations were observed across cell lines. MCF-

7/Adr showed near 100% CD44+/CD24- phenotype, commonly recognized for BCSC.

MCF-7, MCF-7/Adr and MDA-MB-231 showed heterogeneity of CD24 expression, demonstrated cellular hierarchy and the existence of sub-populations in cell cultures.

MCF-7/Adr’s CD24+ population varied between passages, indicative of immunophenotypic transition.

Table 2.2 Alterations of BCSC Markers in Adherent vs. Mammosphere Culture Conditions. Mmammosphere-dissociated cells

CD44 CD24 CD49 ESA

M MCF-7/Ar 40 63 34 -

M MDA-MB-231 65 95 -

Another interesting observation was the rapid phenotypic transition during MCF-

7/Adr mammosphere culture condition, making MCF-7/Adr cells partially lose CD44,

CD49f and ESA expressions, and gaining CD24+, acquiring

CD44weak/CD24+/CD49weak/ESA- phenotype (Fig. 2.2B, Table 2.2). Such drastic alteration in immunophenotype was not observed in MDA-MB-231, which also formed mammospheres under K-SFM, but not DMEM culture condition (Fig. 2.2A, Table 2.3).

This observation indicates that the complete loss of multiple surface markers in a short period of time in sphere culture conditions is not a common phenotype, but signifies the cell’s ability to reverse its acquired phenotype and the ability to divide differentially. The fluidity of immunophenotypic conversion observed here would open new doors to the in vitro studies of EMT and MET related processes in breast carcinoma, drawing parallels with their in vivo counterparts. The data also indicate that there is no direct correlation between a particular CD44/CD24 phenotype and a specific mammosphere culture condition (Fig 2.2A, Table 2.3).

2.3.3 Short-Term SAHA Treatment Reverses Drug Resistance but not BCSC

Immunophenotype

SAHA treatment was incorporated because it’s shown to counter cancer cell’s malignancy, reverse multi-drug resistance, invasiveness and metastatic potential. Cell surface marker profiling of MCF-7/Adr was done after 24 hours or 14 days of SAHA treatment, to examine if short-term drug treatment would make alterations on the surface markers. I observed no alteration in surface markers (Fig 2.3A).

Figure 2.2 Immunophenotype changes after mammosphere culture condition.

(A) Mammosphere formations for different cell lines require different culture conditions.

MCF-7 forms mammospheres in DMEM, while MCF-7/Adr and MDA-MB-231 only form mammospheres in K-SFM, which signifies significant differences in cellular properties. Surface marker profiling of MCF-7/Adr (B) and MDA-MB-231 (C) in mammosphere culture conditions showed MCF-7/Adr’s asymmetric division.

Mammosphere culture condition after 7 days showed significant reversal of CSC phenotypes in MCF-7/Adr (the positive expression of CD44, CD49f and ESA, and the negative expression of CD24). MDA-MB-231, on the contrary, did not show drastic alteration in phenotypes under mammosphere culture condition.

Table 2.3 BCSC Surface Marker Expression in Cell Lines after 14 days of SAHA Treatment (% of control).

CD44 CD24 CD201 CD49f ESA

MCF-7 S - + 34 - 31

MCF-7/Adr S + 60 86 + 16

MDA-MB-231 S - 70 34 48 59

Figure 2.3 SAHA treatment alterations on cell surface markers.

(A) 24 hours SAHA treatment, which was shown to alter cell membrane fluidity, reverse

(transiently) drug resistance, and slightly increase MCF-7/Adr’s mobility, showed no effect on the expression of MCF-7/Adr surface markers. (B, C, D) 14 days of SAHA treatment on MCF-7/Adr passage 6, MCF-7 and MDA-MB-231, respectively. (E, F) human mesenchymal stem cell surface marker on hMSC before and after 7 days of

SAHA treatment. SAHA treatment for 7 days (dosing every other day) had no effect on hMSC’s identification markers.

Mesenchymal stem cells were incorporated in the study to examine SAHA’s differentiation ability. The cells were treated with SAHA for 7 days (drug replenished every 2 days) and surface markers were examined using MSC identification marker panel, found to retain the same set of surface markers as before treatment (Fig. 2.3E-F).

However, it is possible that the drug triggers the up or down-expression of other surface markers not tested using our purchased identification kit.

2.3.4 Cells Affected Differently by SAHA in their Migratory Ability

Migration assay was performed on the highly migratory cell lines hMSCs and MDA-

MB-231, and highly adherent MCF-7/Adr with or without SAHA treatment. For both migratory cell lines, SAHA was found to significantly lower migratory ability (Fig 2.4).

On the other hand, MCF-7/Adr remains non-migratory, uninfluenced by SAHA treatment

(Fig 2.4). This observation was possibly due to drug effect on MCF-7/Adr’s membrane properties, increasing membrane fluidity and decreasing cell-cell contact and adhesion ability, thus making them more mobile. MTT assay showed MCF-7/Adr to be much more vulnerable to SAHA than both MDA-MB-231 and hMSC (Fig 2.5).

2.3.5 Expression of E-cadherin, fibronectin and vimentin in MCF-7/Adr cells

MCF-7/Adr positively expresses E-cadherin and fibronectin proteins. After 5 days treatment, E-cadherin and fibronectin expressions were drastically reduced and vimentin was positively expressed (Fig 2.6).

2.4 Discussion

It has been shown that cancer stem cells should be able to self-renew and differentiate due to their being both drug and radiation-resistant as well as being highly metastatic and invasive [73]. CD44/CD24 combinations categorize breast cancer cells into luminal, basal A and basal B subtypes [93]. High CD44+/CD24- expression represents basal/mesenchymal phenotype, while CD44-/CD24+ signatures luminal/epithelial phenotype [93]. I observed all four combinations in our cell lines, having metastatic

MDA-MB-231 cell line positively express CD44 and have 70% CD24+ expression. On the other hand, ductal carcinoma have uniform CD44-/CD24- phenotype (Fig 2.1), whose tumorigenic and metastatic properties we know very little about. For the first time, a cell line with uniform CD44+/CD24- phenotype was reported, which possesses high drug resistance but poor migratory and proliferation potential. A number of other BCSC surface markers, include CD201, CD49f, and ESA, were also positively identified in

MCF-7/Adr (Fig 2.1).

The malignant transformation the parental cell line MCF-7 experiences to acquire its drug resistance are likely through epithelial-mesenchymal transition (EMT). EMT genes in MCF-7/Adr were analysed in detail by Ozlem et al, and was confirmed to be up- regulated comparing to its parental cell line [94]. Similar observations were made by

Arumugam et al. have reported that cells acquire drug resistance through EMT in pancreatic cancer [95]. Contrary to traditional views on EMT, which have been closely related to tumour progression, increased cell motility and invasiveness [96], MCF-7/Adr

demonstrated decreased motility and a stronger cellular adhesion compared to its parental cell line MCF-7 (Fig. 2.5). Also, it showed no E-cadherin repression (Table 2.1).

Figure 2.4 Migration Assay assesses the migratory potential of SAHA in 24 hours.

(A, B) 24 hours SAHA treatment during migration assay dramatically reduced cellular motility for highly migratory cell lines hMSC and MDA-MB-231. MCF-7/Adr, a strongly adherent and non-migratory cell line, did not demonstrate significant change due to SAHA treatment. (C) Migratory rate quantification showed significant change in migration rate in both hMSC and MDA-MB-231. n=4. p≤0.05.

Figure 2.5 Comparative illustration of SAHA resistance in different cell lines.

MTT assay showed hMSC has the greatest resistance to SAHA, while MCF-7/Adr, which expresses the multi-drug resistant phenotype, has the least SAHA resistance.

Figure 2.6 Immunoﬂuorescence staining examines the presence of E-cadherin, fibronectin and vimentin with or without 5 days of SAHA treatment in MCF-7/Adr.

(A) SAHA effect on E-cadherin, fibronectin and vimentin in MCF-7/Adr. There is evident expression of E-cadherin, fibronectin and vimentin both with and without SAHA treatment. (B) Dual color imaging of both untreated and SAHA treated MCF-7/Adr. Blue: nucleus, green: E-cadherin or vimentin, red: fibronectin.

Previous studies have identified CD44+/CD24- sub-population in a number of cell lines including MDA-MB-231, MDA-MB-436, and SUM204, and was thought to have higher invasiveness and metastatic potential than the CD44+/CD24+ population [96]. I have also observed heterogeneity in our MDA-MB-231 cell line, containing roughly 25% of CD24- phenotype (Fig. 2.1). The entire MDA-MB-231 cell line positively expresses

CD201 and CD49f but negatively expresses ESA and E-cadherin, which relates to the cell’s migratory ability. Since many therapies are formulated based on cancer cell’s molecular profiles [97], it is important to precisely correlate immunophenotypes with cellular properties. Fillmore et al. also showed that CD44+/CD24- phenotype in the culture have high metastatic ability and invasiveness [97]. Since CD44+/CD24- phenotype is also evident in our MCF-7/Adr cell line which possesses no metastatic or invasive properties, I suspect either the CD44+/CD24- phenotype represents different properties which are cell-line specific, or there are other markers which are more representative of a greater metastatic potentials than previously reported. Furthermore, no study has yet shown an entire cell-line possessing the CD44+/CD24-/ESA+ phenotype and able to be stably maintained for a long period of time without reversing back to its parental phenotype.

Fibronectin and vimentin are both EMT proteins. Vimentin signifies basal phenotype, while E-cadherin represents luminal phenotype [93]. Our study shows that MCF-7/Adr expresses a high level of E. Cadherin while at the same time positively expressing vimentin and fibronectin (Fig 2.6). This observation is uniquely important in revolutionizing the established perceptions on BCSCs and EMT. It was reported that a

number of genes induced in iPS transformation were well-established oncogenes (c-myc and SOX2, for example) [98-100]. It is therefore reasonable to make the connection between the acquisition of drug resistance (the up-regulation of oncogenes) and the transition into BCSC-like cells.

Mammosphere culture mirrors in vitro tumorigenic capacity. It has been shown that

MCF-7 undergoes EMT through mammosphere culture condition, generating sub- population with CD44+/CD24- mesenchymal phenotype, E-cadherin down-regulation and more tumorigenic cells formation [101]. In this dissertation I report the reverse of EMT in MCF-7/Adr through mammosphere culture conditions which loses its acquired BCSC markers in 7 days (Fig. 2.2 B, Table 2.2). MDA-MB-231 mammospheres were also characterized and examined for comparison purposes, and no drastic phenotypic alterations were observed. It is possible that at in vivo implantation, similar transition takes place resulting in a heterogeneous tumor with partial chemo-resistance.

The difficulties of sorting CSC subpopulation from tumors give rise to the need of a reliable system to study CSCs in vitro. Inconsistency across reported studies, especially in the characterizations of CSCs in primary tumors is not uncommon. Also, it is yet unknown whether the gaining of CD44+/CD24− phenotype is an exclusive indication of

EMT. Nevertheless, this study provided a new perspective on breast CSC phenotypes through the display of uniform CD44+/CD24− phenotype in a cell line, which has great implications on the in vitro studies of transitions between epithelial and mesenchymal phenotypes.

2.5 Conclusions

This study showed a drug resistant breast cancer cell line with BCSC immunophenotype and proteins which can be stably maintained overtime and undergoes environmentally-driven mesenchymal epithelial transition (the reversal of EMT) upon 7 days of mammosphere formation assay or 14 days of epigenetic drug SAHA treatment.

Phenotypically, it is high in drug resistance and low in metastasis. The unique characteristics presented by this cell line opens new doors for the in vitro characterization of BCSC-like phenotypes, and for drawing parallels with similar in vivo observations.

2.6 Acknowledgements

Reprinted (adapted) with permission from Lu et al.[102]. Copyright (2012), Springer.

CHAPTER III

RNA-SEQUENCING REVEALS DRUG RESISTANCE TRANSCRIPTOMIC

SIGNATURES AND PATHWAYS REVERSED IN RESISTANT BREAST CANCER

CELLS BY EPIGENETIC DRUGS

Keywords: demethylation, histone deacetylation inhibitor, DAC, SAHA, drug transport

3.1 Introduction

Acquired drug resistance is a major issue for effective cancer chemotherapy and has often led to cancer relapse and a more aggressive phenotype [7]. Therefore, reversing the mechanisms and pathways involved in drug resistance is critical for developing effective therapeutic strategies to reverse drug resistance. A major change due to drug resistance takes place in the epigenome (the change in the chromosome and DNA landscape) [103].

Epigenetic drugs, which reverse nuclear landscape and reactivate the transcriptionally silenced genes with anti-tumor effect [104, 105], can have significant impact on tumor response to drug treatments [106, 107]. Two important epigenetic indicators of drug resistance are DNA methylation and histone deacetylation [108]. Five-aza-2'- deoxycytidine (DAC) is a DNA demethylation agent, which inhibits DNA methyltransferase and activates the otherwise suppressed genes [109, 110].

Suberoylanilide hydroxamic acid (SAHA) is a histone deacetylase inhibitor, which binds to the active site of the histone deacetylase as a zinc ion chelator and opens up the histone for DNA transcription [111]. Due to the ability of DAC and SAHA to reverse epigenetic characteristics of drug resistant cells, they were selected for this study.

The main mechanisms of acquired drug-resistance include membrane transport alteration, onco- and drug resistant gene up-regulation, and changes in drug metabolism

[7, 112, 113]. In our previous study, I have demonstrated that DAC treatment enhances doxorubicin (Dox) cellular uptake either when treated with Dox as solution or in liposomal formulation, hence exerted synergistic effect in drug resistant breast cancer cells (MCF-7/Adr). This effect was partially caused by changes in the resistant cell membrane lipid composition, particularly the altered membrane fluidity from a rigid to a more fluid state, thus regaining drug transport and endocytic function [114]. In the same study, it showed that SAHA, on the other hand, only produced additive effect with Dox

[33]. How epigenetic drugs reverse acquired drug resistance and what causes the differential drug effect between DAC and SAHA remain to be investigated. In this study,

I conducted massive parallel sequencing, a genome-wide and cell-type-specific approach, to understand the transcriptomic alterations by DAC and SAHA and identified novel genes and mechanisms that reverse drug resistance. Our result showed that despite the similar regulatory pattern in numerous mechanistic pathways, the differential effectiveness of DAC and SAHA to enhance Dox toxicity can be contributed by the down-regulation of drug metabolism genes by DAC and an up-regulation of drug resistant genes Erb-B2 Receptor Tyrosine Kinase 2 Interacting Protein (ERBB2IP),

Transforming Growth Factor, Beta Receptor II (TGFBR2) and E3F7 (E2F Transcription

Factor) by SAHA.

3.2 Materials and Methods

The materials and the methods to conduct RNA sequencing on MCF-7/Adr breast cancer cells were listed below.

3.2.1 Materials

DAC and SAHA were purchased from Sigma-Aldrich (St. Louis, MO). Three-(4, 5-

Dimethylthiazol-2-yl)-2, 5-Diphenyltetrazolium Bromide (MTT) was purchased from

Sigma-Aldrich. Dox was purchased from Drug Source Co. LLC (Westchester, IL).

Dulbecco's Modified Eagle's Medium (DMEM), penicillin and streptomycin, and

Dulbecco's phosphate-buffered saline (DPBS) were obtained from our Cell Services’

Media Core. Fetal bovine serum (FBS) and Trypsin/EDTA were purchased from Gibco

(Grand Island, NY).

3.2.2 Cell Culture

Breast cancer resistant cell line MCF-7/Adr was obtained from Prof. Batra’s laboratory at the University of Nebraska Medical Center, Omaha. Cells were cultured at

37 °C with 5% CO2 in DMEM with 15% FBS and 1% penicillin-streptomycin. Cells were washed twice with DPBS before detachment using Trypsin/EDTA. The resistant status of cells was maintained by regularly treating the cells with 100 ng/mL of Dox. Cells were incubated for two passages in drug-free media before using for experiment.

3.2.3. Cell Viability

Cytotoxicity of SAHA and DAC in MCF-7/Adr was evaluated by MTT assay. Cells

(5×105 cells per well) were seeded in 24 well plates (BD Biosciences, San Jose, CA) in

500 µL of culture media and were allowed to attach for 24 hrs. The culture medium in plates was replaced with freshly prepared DMEM medium with FBS containing different concentrations of DAC or SAHA (from 5 ng/mL to 250 ng/mL). Following 24 hr incubation, the medium in wells was replaced with 150 µL of MTT (0.5 mg/mL in

DMEM) reagent and allowed to incubate overnight at room temperature. The content in each well replaced with 500 µL of dimethyl sulfoxide, plates were incubated at 37 °C for

3.5 hr to allow the substrate to dissolve, and the absorbance (abs) was measured at 575 nm using a plate-reader spectrophotometer (Sunnyvale, CA). Cell viability in % was calculated by (abs of treated group)/ (abs of untreated cells) ×100%.

3.2.4. RNA Sequencing and Analysis of the Significantly Regulated Genes

Five million MCF-7/Adr cells were seeded in 75 cm2 flasks (Corning, Lowell, MA) until reaching 60% confluency. The cells were then treated for 24 hrs with either 250 ng/mL SAHA or DAC as described above. The concentration of epigenetic drugs was selected based on their effect on inducing apoptotic pathway without significant effect on cell viability. The treatments were done in duplicates. Cells were washed twice with

DPBS, trypsinized, and collected by centrifugation at 1000 g for 5 min (Thermo Electron

Coorporation, Waltham, MA). The total RNA of the treated cells was isolated and purified using RNeasy MiniPlus Kit (Qiagen, Austin, TX) according to the instructions provided in the manual. In brief, the cells were lysed using Buffer RLT. Next, the cell

lysate was then homogenized by vortexing for 1 min, and genomic DNA was removed from the total RNA by gDNA Eliminator spin column provided with the kit. Total RNA bound to the column was washed and eluted into RNase-free tubes provided with the kit.

RNA sequencing library was prepared and analyzed using 100 bp pair-end

ILLUMINA HiSeq2500 (San Diego, CA) at the University of Chicago Genomics Core

Facility (Chicago, IL). FastQC (Babraham Bioinformatics, Cambridgeshire, UK) was used to evaluate the quality of raw reads. Raw reads of each sample were mapped directly against the hg19 genome using TopHat, a public GitHub repository originally developed at the Center for Bioinformatics and Computational Biology at the University of

Maryland (College Park, MD). Next, Cufflinks (Trapnell Lab, University of Washington,

Seattle, WA) was used to assemble transcripts and estimate gene expression level, and

Cuffdiff (Trapnell Lab) was used to perform pairwise sample comparison to detect significantly differentially expressed genes between two samples at a strict false discovery rate (FDR) of 0.05. The differentially expressed genes in epigenetic-drug treated cells were explored using the IPA system (Ingenuity Pathways Analysis,

Ingenuity H Systems, www.ingenuity.com) and the analysis reports were generated on significantly associated pathways. The Ingenuity Report was used to generate gene association maps and relevant biological pathways.

3.2.5. Statistical Analysis

Student’s t test was used to conduct statistical analysis. P ≤0.05 was used to define statistical significance.

3.3 Results

A Comprehensive view of the disrupted transcriptome profile

The structures of DAC and SAHA are shown in Fig 3.1. The cytotoxicity data show that DAC and SAHA treatment at a 250 ng/mL concentration has insignificant effect on cell viability during a 24 hr treatment period (Fig 3.2), hence cells treated at this concentration were used for RNA sequencing analysis. The high throughput RNA sequencing analysis of resistant cells (Fig 3.3) showed that 1786 and 1687 genes were differentially expressed following DAC and SAHA treatment, at a strict false discovery rate (FDR) of 0.05. DAC treated cells showed a significantly greater down-regulation compare to SAHA treated cells (Fig 3.4A). The majority of the differential expressed genes have p≤0.01 (fall above 2 on the y-axis), independent of the fold changes (Fig

3.4B). The Volcano Plot of the differential expressed genes further confirmed the independent nature between the degree of change and their statistical significance (Fig

3.4C).

Overall, DAC and SAHA resulted in the alteration of a similar number of genes, pathways, processes and diseases in resistant cells (Table 3.1A). Furthermore, both epigenetic drugs caused similar changes in gene expression patterns (both in quantity and direction) according to cellular location, molecular function, and genes participated in the

DNA transcription (Fig 3.5). Nuclear genes showed the most significant up-regulation and plasma membrane genes had the most significant down-regulation in both treatment groups. The data showed significant regulation in cell functions such as apoptosis, tumor cell lines proliferation, and cell viability of breast cancer cell lines, where a greater

number of genes were down-regulated than up-regulated. Moreover, DAC treatment had a more significant down-regulatory effect than SAHA treatment in majority of the examined pathways, including apoptosis, tumor cell lines proliferation, and cell viability

(Table 3.2, Fig 3.6).

Cytotoxicity Networks Affected by Epigenetic Modulations

According to the number of affected genes, the 5 most perturbed cytotoxicity networks were identical. The majority of the top altered toxicity pathways showed similar number of genes altered by DAC and SAHA treatment. However, G2/M checkpoint was specifically seen in cells treated with DAC but not in cells treated with SAHA (Table

3.1B). Significant regulation of pathways in cell proliferation, fat synthesis/metabolism, cell mobility and inflammation were observed (Table 3.1C).

Figure 3.1 The structures of (A) DAC and (B) SAHA.

Figure 3.2 The cytotoxicity of epigenetic drugs in cell line MCF-7/Adr.

The cytotoxicity of (A) DAC and (B) SAHA were examined in a 24 hr treatment. No significant cytotoxicity was observed at the concentrations measured. n=2. Data were represented as mean ± standard deviation.

Figure 3.3 Outline of the tools and pipeline used to conduct RNA-Seq transcriptomic analysis.

(A) A list of the programs used to map and decode raw reads into transcriptomic expression levels. (B) Illustration of the steps taken in post-transcriptomic analysis to format sequencing data into gene expression levels. The transcriptomic sequencing was done using 100-bp pair-end lllumina HiSeq2500. The transcriptomes were mapped against the hg19 genome.

Table 3.1 The Number of Gene Alterations in Selective Pathways A. Epigenetic effect overview

DAC SAHA

Total Genes Affected 1722 1615

No. of Pathways 481 474

No. of Processes 310 327

No. of Diseases 123 128

B. Number of genes in top toxicity lists

DAC (Genes SAHA (Genes

Top Toxicity Lists changed/Gene changed/Gene

total) total)

Renal Necrosis/Cell Death 164/466 (0.352) 146/466 (0.313)

Cell Cycle: G1/S Checkpoint Regulation 34/65 (0.523) 32/65 (0.492)

p53 Signaling 44/95 (0.463) 38/95 (0.4)

Hepatic Fibrosis 18/96 (0.188) 38/96 (0.396)

Cell Cycle: G2/M DNA Damage 25/48 (0.521) -- Checkpoint Regulation

A/B (C): No. of Genes Changed/Total Genes in the Category (fraction of the total)

C. The major perturbed pathways and number of genes affected in each pathway

Pathways DAC SAHA

Proliferation of Cells 524 489

Tumor Cell Lines 227 221 Proliferation

Necrosis 388 347

Transcription of RNA 227 221 Cell

Proliferation Apoptosis 390 327

and Cell Apoptosis of Tumor Cell Death 203 177 Lines

Cell Cycle Progression 148 140

Cell Viability of Breast 27 28 Cancer Cell Lines

DNA Damage Response 28 23

Synthesis of Lipid 111 95 Fat Synthesis

and Fatty Acid Metabolism 99 74

Metabolism Hydrolysis of Fatty 32 8

Acid/Lipid

Cell Movement 307 289

Microtubule Dynamics 168 159 Cell Motility

Invasion of Cells/Tumor 139/99 124/89 Cells

Inflammation Inflammatory Response 115 112

The number of genes affected in each gene function

Table 3.2 Genes Significantly Regulated in Selected Pathways Affecting Drug Resistance. A. Pathways affecting the nuclear localization of drug molecules and nanoparticles

Pathways DAC (up-regulated/total) SAHA (up-regulated/total)

FGF1, ACTC1, SH3KBP1, DNM1 FGF1, ACTC1, SH3KBP1, Clathrin- L, LDLR, ITGB4, SNAP91, DNM1L, LDLR, ITGB4, mediated DNM1, SH3GL2, SH3GLB2, SNAP91, FGF5, APOL1, Endocytosis CLU, RAC1, FGF18, SERPINA1, PDGFA, VEGFA (5/11) ACTA2 (4/15)

Caveolar- ACTC1, FYN, HLA-A, ITGB4, ACTC1, FYN, HLA-A, ITGB4, mediated ITGAX, CAV1, ITGA6, HLA- ITGAX, FLNC (3/6) Endocytosis B, HLA-C, ACTA2 (4/10)

Macropino- HRAS, PLCG2, NGF, CSF1, ITGB HRAS, NGF, PLCG2, CSF1,

cytosis 4, RAC1, RRAS, PRKCG (1/8) ITGB4, PDGFA, (1/6)

TC10, TNFAIP2, CDC42, Exocytosis TC10, TNFAIP2, CDC42, RAB11 RAB11

ABCA7, ABCD1, SLC47A1, Efflux ABCA2, ABCB4, MFSD12, ABCA7, ABCD1, SLC47A1 Pumps MFSD3

RAN RCC1, KPNA2 (2/2) RCC1, TNPO1, XPO1 (3/3)

B. Cell lipids and drug molecules metabolism

DAC SAHA

FYN, BRCA1, MYO5A, IL18, FAD BRCA1, FYN, IL18, MYO5A, S1, SREBF1, GPC1, PTGES, SREBF1, LDLR, C5AR1, CCL5, CCL5, FAAH, FADS2, PTGIS, CEBPB, CXCL1, FAAH, FADS1, IL6, C5AR1, IL8, PTGDS, LDLR, FADS2, GPC1, HMOX1, IL1B, NFKBIA, HMOX1, CEBPB, FOXO IL6, IL8, NFKBIA, PTGDS, 3, CAV1, LIMK1, MAP3K8, Synthesis PTGES, PTGIS, KITLG, LPL, ASAH1, of Fatty MAP3K14, MAPK8, NRG1, NTN1, ACADVL, DGAT1, ACSL1, Acid OLR1, IL32, FGFR3, VEGFA, FASN, ADA, SLC2A4, NGF, SEMA ACSS2, CASP1, EPHX2, IGF2, 3A, PLA2G4A, TRIB3, CLU, RAC1, INSIG1, MID1IP1, NR1H3, PLA2G6,NR1H3, SIGIRR, IL1R2, SIGIRR, XBP1, C5, FASN, RARB, RELB, SCD, EDNRB, FCG FOXA1, NGF, NTN1, SCD, R2A, C5, ACSS2, CXCL1, STAT5A, TRIB3 (10/48) INSIG1, IL32, (6/53)

PLCB4, PLCD3, PLCG2, PLA2G4

Phospholi PLCB4, HMOX1, PLCG2, PLCD1, C, HMOX1, PLCD1, PLD1, LIPG, pases PLD3, PLA2G4C, PLD1, RARRES3,

Pathway RARRES3, LIPG, PNPLA3 (1/10) PLA2G4A, PLA2G16, PLA2G6, P

LD3 (1/13)

TFEB, CTSF, CTSB, Rab3a TFEB, CTSF, CTSB Lysosoma

l Genes

HRAS, RHOQ, PPP2R5B, HMOX1, HRAS, PPP2R5B, RHOQ,

PLD3, DDIT4, PLD1, ATG13, HMOX1, PLD3, PLD1, DDIT4, mTOR RND2, RAC1, RHOC, PPP2R2A, MAPKAP1, RPS26, pathway RPS17/RPS17L, RRAS, PRKCG, RHOV, RPS24, PRR5 VEGFA

RPS24, PPM1J (2/16) (4/14)

Wnt WISP2 (-12.2 fold), DKK1(3.2 fold) WISP2 (-5.2 fold), DKK1(6.0 fold) Pathway

IL1RAP, HSP90AA1, IL18, HSrc, IL1RAP, IL18, HSP90AA1, NGFR, PPAR IL1B, NGFR, NFKBIA, IKBKB, IL1B, HRAS, NFKBIA, MAP3K14, signaling NR1H3, IL1R21, NFKB2, PPARA, STAT5A, pathway IL1R1, FOS, RRAS (5/13) PDGFA, NR1H3 (5/13)

Cytochro CYP27C1, CYP1A1, CYP1B1, me 450 --- CYP26B1, CYP2J2, CYP39A1 Family

Highlighted genes: genes that are regulated by both DAC and SAHA. Red: up-regulated genes; blue: down-regulated genes

Table 3.3 Gene Fold Changes and Functions A. Nuclear transport Gene Symbol: DAC SAHA Function Name

Interacts with the NLSs of DNA helicase Q1 and

Kpna2: SV40 T antigen, functions in nuclear protein

Karyopherin --- 2.04 import as an adapter protein for nuclear receptor

Alpha 2 KPNB1, and may be involved in the nuclear

transport of proteins

Component of the NPC, a complex required for the

NUP153: trafficking across the nuclear envelope. Functions

Nucleoporin 2.10 --- as a scaffolding element in the nuclear phase of the

153kDa NPC essential for normal nucleocytoplasmic

transport of proteins and mRNAs.

Localized to the nuclear rim, and is part of the NUP35: nuclear pore complex. All molecules entering or Nucleoporin --- 2.25 leaving the nucleus either diffuse through or are 35kDa actively transported by the nuclear pore complex.

Functions in nuclear protein import as nuclear KPNB2 (TNPO1): 2.23 --- transport receptor. Serves as receptor for nuclear Transportin 1 localization signals (NLS) in cargo substrates.

RCC1: Regulator Plays a key role in nucleo-cytoplasmic transport, of Chromosome 2.83 2.77 mitosis and nuclear-envelope assembly Condensation 1

B. Nucleoskeleton network genes Gene

Symbol: DAC SAHA Function

Name

KIF4A: Intracellular transport of membranous organelles; Kinesin 2.09 2.04 chromosome arms condensation and chromosome Family integrity during mitosis. Member 4A

A member of kinesin-like protein family. This

family includes microtubule-dependent molecular KIF23: motors that transport organelles within cells and Kinesin 4.21 3.67 move chromosomes during cell division. This Family protein has been shown to cross-bridge antiparallel Member 23 microtubules and drive microtubule movement in

vitro.

Fibrous proteins providing structural function LAMIN B 2.36 5.06 and transcriptional regulation in the cell nucleus

SYNE1: A nuclear envelope protein involved in the Spectrin maintenance of nuclear organization and structural Repeat 2.29 2.38 integrity, tethering the cell nucleus to Containing, the cytoskeleton by interacting with the nuclear Nuclear envelope and with F-actin in the cytoplasm. Envelope 1

SYNE2:

Spectrin Binds to cytoplasmic F-actin, tethering the nucleus Repeat 14.7 15.9 to the cytoskeleton and maintaining the structural Containing, integrity of the nucleus. Nuclear

Envelope 2

C. Drug resistance, EMT and drug metabolism genes Gene Symbol: DAC SAHA Function Name

TGFBR3L: Multifunctional peptides that regulate transforming proliferation, differentiation, adhesion, migration, growth factor, -3.18 -3.51 and other functions regulates many other growth beta receptor factors III-like

ERBB2IP: Regulates ERBB2 function and localization, is Erb-B2 up-regulated in paclitaxel-resistant cell lines. Receptor 4.1 ERBB2 overexpression has been shown to play an Tyrosine --- important role in the development and Kinase 2 progression of certain aggressive types of breast Interacting cancer. Protein

The encoded protein is a transmembrane protein TGFBR2： that has a protein kinase domain, forms a Transforming 2.5 heterodimeric complex with another receptor Growth --- protein, and binds TGF-beta. TGFBR2 is Factor, Beta connected with breast cancer response to tamoxifen Receptor II

E2F: E2F A family of transcription factors (TF) in higher -3.2 2.1 Transcription eukaryotes involved in the cell cycle regulation (E2F2) (E3F7) Factor and synthesis of DNA in mammalian cells

SNAI3: SNAG zinc-finger proteins are transcriptional

snail family -3.39 -3.55 repressors, an oncogene, implicated in zinc finger 3 carcinogenesis and embryogenesis

Chromatin-binding protein that converts stress NUPR1: signals into a program of gene expression that Nuclear empowers cells with resistance to the stress -2.93 -11.7 Protein, induced by a change in their Transcriptiona microenvironment. Strong connections with p21 l Regulator, 1 and chemoresistance

The major cytoskeletal component VIM: -2.37 -2.61 of mesenchymal cells, a marker of epithelial-to- Vimentin mesenchymal transition (EMT)

GSTA4: Detoxification of lipid peroxidation products. glutathione S- Involved in cellular defense against toxic, -10.5 -2.25 transferase carcinogenic, and pharmacologically active alpha 4 electrophilic compounds.

BRCA-1: A human tumor suppressor gene. Its protein is breast cancer 3.17 2.12 responsible for repairing DNA. 1, early onset

BCLAF1: Tumor-suppressor. Encodes a transcriptional

BCL2- repressor that interacts with several members of associated 4.58 3.03 the BCL2 family of proteins. Overexpression of transcription this protein induces apoptosis, which can be factor 1 suppressed by co-expression of BCL2 proteins.

Gene functions were obtained from Entrez Gene

Figure 3.4 Overview of the significantly altered transcriptomes in MCF-7/Adr after

SAHA and DAC treatment.

(A) Compare to control, DAC up-regulated 558 genes and down-regulated 1228 genes, whereas SAHA up-regulated 661 genes and down-regulated 1026 genes. From this statistics, DAC had a greater gene down-regulatory effect than SAHA. (B) Gene fold changes were plotted against their statistical significance. Most genes had from -5.66 to

5.66 fold changes. There is no linear relationship between the fold changes and their statistical significance. (C) Volcano plot displayed log(fold-change) against -log(p-value).

It compared the magnitude of fold changes (x-axis) against their statistical significance

(y-axis), where higher y value indicates a greater statistical significance. The genes scattered in the far left and right on the x-axis (over 1000 fold changes) are likely to be

those not expressed in the control group, but are activated by the treatment, or are completely suppressed by the treatments. X-axis showed both positive and negative log

(fold changes), and the values nearer to the top of y-axis are the changes with high statistical significance (shown in red dots). In this plot, I observed a similar pattern between DAC and SAHA treated groups. No relationship was observed between the fold changes and their statistical values, indicating that all expressions were valid and that the statistical significance is not affected by the fold changes. Ninety-five percent confidence level was used and p≤0.05 was used to define statistical significance.

Figure 3.5 Overview of alternatively expressed genes in (A) DAC and (B) SAHA treated cells.

The distribution of genes based on cellular location (left most), molecular functions

(middle) and genes participated in DNA transcription. Color range represents the level of fold changes. Based on the cellular location, the nucleus had more up-regulated genes than the extracellular space, cytoplasm, and the plasma membrane in both drug treatments.

Down-regulation of Oncogenes and Up-regulation of Tumor Suppressors

A number of major upstream regulators in epithelial-mesenchymal transition

(EMT) (VIM GSTA4, SNAI3), drug resistance (e.g. NUPR1), and detoxification

(e.g. GSTA4) were down-regulated by both DAC and SAHA. Tumor suppressor genes BRCA-1 and BCLAF1 were up-regulated by both DAC and SAHA. However, we observed drug resistance genes to be up-regulated by SAHA and either down- regulation or not affected by DAC (ERBB2IP, TGFBR2 and E2F) (Table 3.3C).

Alterations in Genes involved in Exocytosis and Intracellular Trafficking

Analysis of lipid-related pathways (e.g. fatty acid synthesis, phospholipases signaling pathway, and mammalian target of rapamycin (mTOR)) showed majority of them to be down-regulated (Fig 3.9, Table 3.2B). Examining the pathways involved in endocytosis and drug transport, we demonstrated that the major endocytosis pathways had a mixed up- and down-regulation of genes. However, all the genes regulating drug degradation and export (e.g. lysosome, exocytosis, efflux pumps) were down-regulated (Fig 3.9,

Table 3.2). All nuclear localization and nucleoskeleton genes were shown to be up-

regulated (Table 3.3). The RAN pathway, which is involved in nuclear transport, was shown to be up-regulated following treatment with epigenetic drugs (Fig 3.10).

Changes in Drug Metabolism Genes and Pathways

The results showed that both DAC and SAHA effectively down-regulated the Wnt pathway through the suppression of WNT1 Inducible Signaling Pathway Protein 2

(WISP2) and the up-regulation of Dickkopf WNT signaling pathway inhibitor 1 (DKK1)

(Table 3.2B). WISP2 is significantly expressed in breast adenocarcinoma and contributes to malignant transformation. DKK1 is a Wnt antagonist and a tumor suppressor [115].

The alterations in WISP2 and DKK1 will effectively reduce the drug metabolism of therapeutic molecules, allowing them to exert their drug effects.

Six genes in the CP450 family were significantly regulated by DAC, with 5 down- regulated and 1 up-regulated. In contrast, SAHA did not cause significant changes in the expression of CP450 genes (Table 3.2B). In the peroxisome proliferator-activated receptors (PPAR) signaling pathway, DAC and SAHA down-regulated the majority of the genes, showing suppression in drug metabolism, thus allowing a greater level of active drug molecules to exert their effect compare to that in untreated cells (Fig 3.8).

3.4 Discussion

This study elucidated the effect of DAC and SAHA on the global transcriptomics of

MCF-7/Adr breast cancer cell line, with focus on the pathways and genes affecting the level of drug resistance in a cell. Demethylation has been shown to reverse drug

resistance in tumor [116]. Studies also showed that histone deacetylase inhibition is able to overcome drug resistance in diffuse large B-cell lymphoma and other tumor cell lines

[117, 118].

Figure 3.6 Gene regulation overview in selected molecular functions.

Number of the genes significantly regulated in (A) apoptosis, (B) tumor cell lines proliferation, and (C) cell viability of breast cancer cell lines. In all pathways, more genes were down-regulated than up-regulated. DAC down-regulated more genes than SAHA did.

Figure 3.7 The effect of DAC and SAHA on endocytotic regulation.

(A) Clathrin-mediated endocytosis signaling pathway in DAC treated cells. (B) The fold change of genes significantly regulated in clathrin-mediated endocytosis pathway. (C)

The fold change of genes significantly regulated in caveolar-mediated endocytosis pathway. In both endocytotic pathways, DAC caused more changes in gene than SAHA does.

Figure 3.8 (A) Visualization and (B) gene fold changes in PPAR signaling pathway.

PPAR pathway affects intracellular lipid levels, fatty acid metabolism, and peroxisome proliferation and showed a greater level of down-regulation than up-regulation in both

DAC and SAHA treated cells.

Figure 3.9 (A) Visualization and (B) list of altered genes in mTOR signaling pathway. mTOR is involved in cellular metabolism and lysosome regulation. There is significantly more down-regulation than up-regulation in both DAC and SAHA treated cells.

Figure 3.10 RAN signaling pathway in SAHA-treated cells.

RCC1, TNPO1, XPO1were all significantly up-regulated.

Our RNA-seq data identified alteration of a number of major genes and pathways that are critical for reversing drug resistance following epigenetic drug treatments. The large number of identical genes in DAC and SAHA treated cells (Table 3.2) indicated that although both epigenetic drugs work by different mechanisms, they have similarities in their regulatory effects.

In Fig 3.4B, it was shown that there are similar patterns in transcriptomic fold changes vs. –log(pValue) between DAC and SAHA treated cells. First, most fold changes in the transcriptomes are from -5.66 to -2, and from 2 to 5.66, using 2 fold changes as cutoffs. We observed –log(pValue) falls between 2 and 4, equals to the p-value of 0.01 to

0.0001. This means most transcriptomic changes were highly statistically significant.

In Fig 3.4C, the volcano plot showed similar patterns in transcriptomic fold changes vs. –log(pValue) between DAC and SAHA treated cells. All genes above 1.3 (P≤0.05) on the y-axis were statistically significant.

The majority of the genes were in the double-filtering window, above 1.3 on the y- axis and ≥0.3 or ≤-0.3 on the x-axis, which had greater than 2 fold changes. Large gene fold-changes were those farther from the center of the x-axis. The genes that were on similar spots in DAC and SAHA treatment groups had similar p-values and fold changes.

The higher on the y-axis, the greater the statistical significance. The genes with over 100 fold changes are those that are not expressed in the control, but are expressed in the treated groups, or vice versa.

Table 3.3 showed that in both treatment groups, similar numbers of genes were significantly regulated in majority of the top affected signaling pathways and cellular functions. This is also a new trend that we observed. We saw that different epigenetic mechanisms could similarly affect different pathways. If we do comparative analysis between the 2 treatment groups, we will be able to see the exact genes that were differentially affected by treatments, and those will allow a more thorough examination on the drug effects.

There are a large number of different genes with similar fold changes and statistical significance (ex. 2-20 fold changes). Here, I focused on how they enhance the therapeutic effect of chemotherapy and nanoparticles through pathway analyses. On Table 3.2 and

3.3, both similarly and differently affected genes were observed in pathways.

Several observations from the Volcano plot were: 1. among the significantly regulated genes, there were both up- and down-regulations, 2. A number of genes that are similarly regulated in the same direction—that would partially contribute to the similar patterns observed in both plots. Among those, part of them were the same genes, also a part were different genes with similar fold changes and statistical significances.

The red dots on the top of the plot were genes that had the highest statistical significances. The genes on the far left and far right of the plot were ones with high fold changes but low statistical significance.

Compare DAC and SAHA treated cells, we observed a large number of overlaps in the genes significantly affected by DAC and SAHA. Second, from the examined pathways, we observed that in majority of the pathways, the pattern of regulation for

DAC and SAHA treated groups were similar. For instance, in cell viability of breast cancer cell lines (Fig 3.6C), DAC caused 9 up-regulated genes and 18 down-regulated genes, whereas SAHA resulted in 11 up-regulated genes and 17 down-regulated genes.

Thirdly, we observed that genes regulated in both pathways mostly have the same direction of alteration (Fig 3.7-3.9). Because of the similarities in pattern of changes in the transcriptome, they resulted in similar patterns shown on the Volcano Plot.

Drug resistance is influenced by the level of drug intracellular localization, lipid quantity (as drug molecules such as Dox enter the cell by diffusion), lysosomal regulation, and nuclear localization, which all contribute to the level of drug sensitivity in the cell.

Our previously published work showed that DAC effectively enhances the uptake of Dox and Doxil [114]. Therefore, the genes and pathways that affected uptake of drug in solution form and in nanoparticle form were examined. The plasma membrane is composed largely of lipids. Therefore, lipids represent the first major barrier to the intracellular localization of drugs. Once enter the cell, it is important for the drug molecules to escape from lysosomal degradation. The intracellular accumulation and retention of the drug in solution form and in nano-carriers are also affected by the decreased anti-cancer drug efflux and exocytosis pathway.

Lipid is a major component of the cell membrane and the intracellular vesicles. Their alterations produce major biophysical changes in the resistant cell membrane, thus influencing the uptake and uptake of drug molecules and nanoparticles [119]. Previous studies from our group showed that resistant cell membrane lipids form more condensed and rigid monolayers than that of sensitive cells, and epigenetic drug DAC was able to reverse such characteristics to restore the trafficking of cancer drugs [120]. In this study, I further showed that the majority of the cellular fat levels and fat synthesis genes were down-regulated (Table 3.2B), evidenced a reduction of fat compositions in the cell. One main pathway for the fatty acid synthesis and metabolism is the phospholipases signaling pathway. Its down-regulation confirms the previous study of the accumulation of phospholipids in the cell [114]. Furthermore, the down-regulation of fatty acid synthase

(FASN) gene and mTOR pathway by both the epigenetic drugs (Table 3.2B) also had major impact on lipid synthesis [121, 122]. FASN causes drug resistance through TNF-α inhibition, which strongly connects with poor breast cancer prognosis and higher risk of recurrence [123, 124]. Therefore FASN presents an effective target gene for reversing of cancer cell drug resistance.

Lysosomes are the main vesicles for drug molecule degradation. They also play a role in plasma membrane repair (Table 3.2) [125]. One of the major lysosomal genes is transcription factor EB (TFEB), which is a major regulator of lysosomal biogenesis and induces both the docking and the fusion of lysosomes with the plasma membrane [126].

The down-regulation of TFEB, together with other lysosomal (e.g. CTSF, CTSB) and exocytotic (e.g. TC10, TNFAIP2) genes, exert major effect on the cellular uptake as well

as recycling of drug molecules [114, 127] (Table 3.2A), enabling a greater amount of drug to stay in the cytoplasm.

Nuclear membrane penetration is the last challenge for the effective localization of nuclear targeting chemotherapies (e.g. Dox). The nuclear pore complex (NPC) allows the passive diffusion of small molecules with diameters less than 9nm [128]. It has been shown that drug resistant and sensitive cells have different nuclear pores and the down- regulation of NUP protein contributes to the resistance of chemotherapy cisplatin [129,

130]. Therefore, the increased NUP and nucleoskeleton genes may be a contributor to a greater amount of drugs entering the nucleus.

In the nucleus, the up-regulation of RAN (RAs-related nuclear protein) plays a major role in transporting molecules in and out of the NUP. RAN also has effect on DNA synthesis and cell cycle, and its mutation has shown to disrupt DNA synthesis [131]. An essential gene in the RAN pathway, RCC1, is located inside the nucleus and is bound to chromatin (both to the nucleosomes and double-stranded DNA). It plays a major role in nucleo-cytoplasmic transport and nuclear-envelope assembly. Through up-regulating the above-mentioned genes (Fig 3.10, Table 3.2A), DAC and SAHA could enhance the therapeutic effect of drugs that require nuclear penetration (e.g., Dox) [132].

One important mechanism of drug resistance is through enhancing the metabolism as well as detoxification of the drugs. Cytochrome P450 (CP 450) family plays an essential role in interacting with the drugs, breaking them down into metabolites, thus inactivating

their effect on the cell [133]. DNA methylation is a common mechanism among currently identified genes in drug-metabolism [134]. Up-regulation of drug metabolizing genes (ex.

CYP1A1, CYP7B1) has been shown in MCF-7/Adr breast cancer cell line to enhance drug resistance [135, 136]. In this study, I noticed a major effect down-regulating effect of DAC, but not of SAHA, on cytochrome P450 enzymes (CP450), which showed a new mechanism of DAC in reversing drug resistance (Table 3.2B). PPARs, which were down-regulated by both epigenetic drugs, are also essential in drug metabolism and elimination of toxic metabolites [137].

The genetic over-expression of drug resistant genes and pathways always play a role in enhancing the level of drug resistance. One of such pathway is mTOR, and when this pathway is inhibited, increased chemo-sensitivity is observed in ovarian carcinoma cell lines [138]. Previous study showed SAHA able to dampen the mTOR pathway [139], consistent with our report (Table 3.2B, Fig 3.9). However, in this study, I uncovered

DAC to also be an mTOR pathway suppressor, signifying yet another mechanism through which DAC and SAHA reverses drug resistance. A well-recognized onco-gene that is down-regulated in both DAC and SAHA treatment is Wnt (Table 3.2B). Wnt signaling pathway is important for drug resistance and associates with cancer stem cell phenotype [140]. The inhibition of Wnt pathway resulted in P-gp down-regulation and the reversal of drug resistance [141]. The up-regulation of TGFBR2, a part of the TGF-β pathway, has been shown to drug resistance and cell heterogeneity [142, 143].

NUPR1, a chromatin-binding protein, is one important drug resistant gene that is down-regulated by both DAC and SAHA (Table 3.3C). When activated by the stress

signals, it empowers the resistant cells to change their microenvironment, leading to resistance of chemotherapeutic agents like doxorubicin in breast cancer cells. Therefore, the NUPR1 down-regulation by both SAHA and DAC treatments showed a novel candidate for reversing drug resistance.

In the examined up-stream drug resistant genes, I found similarly as well as differentially regulated genes by DAC and SAHA treatment. Examples of SAHA up- regulated drug resistant genes are ERBB2IP, TGFBR2 and E2F. Erb-B2 Receptor

Tyrosine Kinase 2 interacting protein (ERBB2IP) is a gene that regulates the function and localization of ERBB2, which mediates breast cancer chemo-resistance via receptor- mediated anti-apoptotic signals and found to be up-regulated in paclitaxel-resistant cell lines [144, 145]. Also, E2F is a distinctive marker of tamoxifen resistant breast tumors, which is down-regulated in DAC and up-regulated in SAHA [146]. The differential regulation of the above 2 genes by DAC and SAHA contribute to the differential effect of

DAC and SAHA on the drug resistance level of the cell (Table 3.3C).

3.5 Conclusions

In this study, I elucidated the effects of SAHA and DAC on drug resistant MCF-

7/Adr cells using RNA-seq. Our study showed that there are major overlaps in genes changed by DAC and SAHA treatments in areas of drug and nanoparticle tracking, lipid metabolism, carcinogenesis and nuclear transport, with consistent direction of regulation, either up- or down-regulated by both drugs. Nevertheless, DAC affects more genes and down-regulated in a larger scale in majorities of the pathways analyzed. Also, a number

of drug metabolism genes essential to render drugs ineffective and drug resistant upstream regulators are down-regulated in DAC, but not in SAHA, and a number of drug resistant genes were only up-regulated by SAHA but not DAC. This study is essential in understanding the differential effects between DAC and SAHA and can be used further to examine their potential as combination therapy with other drugs.

3.6 Acknowledgements

I thank the Bioinformatics Core at the University of Chicago for carrying out RNA sequencing and statistical analysis.

CHAPTER IV

EFFECTIVENESS OF siRNA DELIVERY VIA ARGININE-RICH PEI-BASED

POLYPLEXES

Keywords: siRNA therapy, cationic polymers, metastatic cells, drug resistance, nanocomposite

4.1 Introduction

The therapeutic promise of small interfering RNAs (siRNAs) has gained wide recognition in treating different diseases, such as cancer, viral infections, HIV and non- arteritic ischemic optic neuropathy [147-149]. A major advantage of siRNA lies in the fact that it functions in the cytoplasm, therefore, avoiding the need for nuclear penetration, a major delivery obstacle particularly for gene therapy, due to the tight regulation of the nuclear membrane. It has the potential to repress cancer cell proliferation and metastasis and to reverse drug resistance [150-152]. Despite its therapeutic potentials, translation of siRNA-based therapy is limited by several factors, including poor pharmacokinetic property, possible immune response, off-target effects and poor cellular uptake, due to the anionic and hydrophilic nature of the siRNA to diffuse efficiently across cell membrane [153-155].

For effective siRNA delivery, carrier needs to achieve the capacity to form stable complexes with minimum cytotoxicity, protect siRNA from degradation, and facilitate transfection [156]. To achieve the above objective, lipids (ex. liposomes), polymers (ex. cyclodextrin, chitosan, polyethylenimine), and inorganic (ex. carbon nanotubes, iron oxide) vehicles have been investigated [157, 158]. Cationic polymers have the ability to both complex siRNA and effectively transfect cells. The most commonly used are poly(ethylenimine)s (PEI) and poly(amido amine) [159, 160]. PEI can spontaneously self-assemble with siRNA to form polyplexes, achieving effective transfection rate [161-

164]; however, they faced significant toxicity issues including membrane damage and the activation of apoptotic pathway by mitochondria [165-167]. General strategy is to minimize the toxic effect of cationic polymers while retaining transfection ability.

Arginine has the ability to facilitate intracellular translocation. The arginine has strong affinity with heparan sulfate expressed on mammalian cell membranes. Also, the guanidine groups on arginine are able to form hydrogen bonds with polyhydroxy compounds in cell membrane [168-170]. Studies that conjugated arginine to dendrimers such as polymer amido amine (PAMAM) and polymer (propylene imine) (PPI) have also shown to enhance transfection efficiency [171, 172]. Here, I explored the ability of

P(SiDAAr)5Peg3 to deliver siRNA to both breast cancer metastatic (MDA-MB-231-luc-

D3H2LN) and drug resistant (MCF-7/Adr) cell lines. I hypothesize that P(SiDAAr)5Peg3 will be effective in facilitating siRNA transfection as PEG enhances the biocompatibility of the polymer and arginine facilitates cell membrane penetration. Our data show that

polyplexes at optimized composition are nontoxic to cells, effective in transfecting luciferase and therapeutic siRNA in both metastatic and drug resistant cell lines.

4.2 Materials and Methods

The materials and the methods to synthesize P(SiDAAr)5Peg3 and conduct siRNA transfection were listed below.

4.2.1 Materials

Three-(2-aminoethylamino) propyl-methyl-dimethoxysilane, 25-kDa branched PEI, aspartic acid, L-arginine, O-(2-aminoethyl)-O’-(2-carboxyethyl) polyethylene glycol-3K hydrochloride, 1-Ethyl-3-(3-dimethylaminopropyl) carbodiimide hydrochloride (EDC), heparin sodium, N-hydroxysuccinimide (NHS), and 3-(4,5-dimethylthiazol-2-yl)-2,5- diphenyltetrazolium bromide (MTT) were purchased from Sigma-Aldrich (St. Louis,

MO). Fetal bovine serum (FBS) and Trypsin/EDTA were obtained from Gibco (Grand

Island, NY). RNAse I (cloned) and Silencer Negative Control No.1 siRNA were purchased from Life Technologies (Carlsbad, CA). Sodium bicarbonate powder, hydroxylamine hydrochloride, uranyl acetate, and TrisAcetate-EDTA buffer were purchased from Sigma-Aldrich (St. Louis, MO). Dulbecco's Modified Eagle's Medium

(DMEM), DPBS, penicillin and streptomycin were purchased from the Cell Services’

Media Core at the Cleveland Clinic.

4.2.2 Synthesis of P(SiDAAr)5Peg3

Three molars equivalents of polyethylene glycol and 5 molars equivalents of L- arginine oligo (alkylaminosiloxane) grafted poly(ethylenimine) (P(SiDAAr)5Peg3) were synthesized as shown in appendix D [173]. PEI conjugation of SiDAAr was done by using aspartic acid as a linker. The acid groups of the aspartic acid were activated using

0.1208 g (3.5/5 mmol) of EDC and 0.0725 g (3.5/5 mmol) of NHS at 4 oC for 6 hr.

Thereafter, 0.05712 g (0.06 mmol equivalents) of SiDAAr and 0.3 g (0.012 mmol equivalents) of PEI were added into the reaction mixture and kept for 18 hr at room temperature, followed by dialysis using 12kDa MWCO dialyzer. To conjugate PEG to

P(SiDAAr)5, briefly, the free acid group of 0.1 g (3 mmol) equivalents of

NH2−PEG−COOH was activated using 0.1208 g (3.5/5 mmol) of EDC and 0.0725 g

(3.5/5 mmol) NHS (catalyst) in MES buffer at pH 6 for 3.5 hrs at 4 °C. Thereafter, the activated PEG were added to 0.3 g (1 mmol equivalents) of P(SiDAAr)5 in 2 mL of PBS at pH 7.5 overnight (16 hr). Afterward, the polymer was dialyzed for 2 days in 2 liters of dH2O using 12kDa MWCO dialysis membrane (Spectrum Laboratories, Rancho

Dominguez, CA) to remove unreacted elements. The polymer was then lyophilized and re-dissolved in dH2O and filtered through a sterile filter. Please refer to Appendix D for detailed synthetic procedures of parent polymer P(SiDAAr)5.

4.2.3 TRITC Conjugation of P(SiDAAr)5Peg3

To visualize the polyplexes uptake using flow cytometry, I first conjugated

P(SiDAAr)5Peg3 with TRITC fluorescence dye. Six-tetramethylrhodamine-6- isothiocyanate (TRITC) dye was obtained from Invitrogen (Carlsbad, CA). To conjugate

TRITC to polymer P(SiDAAr)5Peg3, first, 2 mg of polymer was dissolved in 1 mL of 0.1

M sodium bicarbonate (8.4 mg in 1 mL) buffer, and 1 mg of TRITC was dissolved in 500

µL of DMSO. While stirring, 100 µL of reactive TRITC dye was slowly added to the polymer solution dropwise. The reaction was kept for 1 hr at room temperature with continuous stirring. Afterward, the reaction was stopped by adding 0.1 mL of freshly prepared 1.5 M hydroxylamine (pH 8.5), which was prepared by dissolving hydroxylamine hydrochloride at 210 mg/mL in distilled water and adjusting the pH to 8.5 with 5 M (1 g in 5 mL) NaOH. The polymer was dialyzed overnight using 12K MWCO dialysis membrane to remove excess TRITC dye.

4.2.4 Polymer Complexation and Polyplexes Characterization

P(SiDAAr)5 (parent polymer before PEG conjugation) and P(SiDAAr)5Peg3 were dissolved in dH2O at 1 mg/mL and filtered through a 0.22 µm sterile polyethersulfone filter (Millipore, Darmstadt, Germany). siRNA was diluted to 1 mg/mL in RNAse free dH2O (Qiagen, Valencia, CA). Polyplexes were prepared by pipet-mixing siRNA and polymer for 30 times and allowing self-assembly by incubating for 1 hr. Three µg of siRNA were used per sample. Size and zeta potential measurements were conducted with the sample in dH2O using quasi-elastic light scattering (NICOMP 380 ZLS Particle

Sizing System, Santa Barbara, CA). The morphology of the polyplexes was characterized by transmission electron microscopy (TEM) using a Tecnai G2 TEM microscope (FEI,

Hillsboro, Oregon). Polyplexes with a 1/4 siRNA/polymer w/w ratio were prepared in dH2O using the same method described above. After that, 10 µL of the polyplex dispersion was dropped on TEM grids coated with silicon monoxide stabilized formvar

films (Electron Microscopy Sciences (Hatfield, PA)). The samples were then air dried, stained with 2% uranyl acetate for 7 min and imaged.

4.2.5 Determination of the Complete siRNA to P(SiDAAr)5Peg3 Complexation Ratio

Agarose gel electrophoresis was used to analyze the siRNA/P(SiDAAr)5Peg3 complexation ratio. The polyplexes were prepared at nucleotide/polymer (N/P) w/w ratios of 1/0.5, 1/1, 1/2, 1/4 and 1/6, keeping the siRNA amount constant at 1 µg while varying the quantity of the polymer to achieve the desired N/P ratio. Using naked DNA as a reference, naked siRNA and siRNA polyplexes were loaded onto a 1% Gold Agarose gel (Lonza, Allendale, NJ) in TrisAcetate-EDTA buffer. The samples were run for 60 min at 60 V on an electrophoresis system (Bio-Rad, Hercules, CA). The gel was stained using SYBR Green (Thermo Scientific, Rockford, lL) and visualized using a UV imager

(FUJIFILM FLA-5100, FUJIFILM Life Science, Stamford, CT).

4.2.6 Protection from siRNA RNase Degradation

The effectiveness of P(SiDAAr)5Peg3 to protect siRNA from RNAse degradation was measured using a Nanodrop 1000 (Thermo Scientific, Waltham, MA). Briefly, naked siRNA and siRNA polyplexes were incubated in 100IU of RNAse buffer for 30 min.

Thereafter, measurements were carried out using a nanodrop to assess the degradation of the siRNA.

4.2.7 Polyplexes Stability Evaluation

The stability of siRNA polyplexes at 1/4 w/w siRNA/P(SiDAAr)5Peg3 ratio was measured in the presence of heparin polyanions using a polyanion competition assay. The polyplexes, consisting of 5 µg of siRNA and 20 µg of P(SiDAAr)5Peg3 were incubated in

15 µL of various concentrations of heparin sodium solution (Sigma-Aldrich) for 15 min.

The samples were run on a 1% gold agarose gel (Lonza, Allendale, NJ) in TrisAcetate-

EDTA buffer, and the gel was stained and visualized by the same method described in section 2.5.

4.2.8 Cell Culture

The resistant cell line MCF-7/Adr cells was obtained from UNMC, and maintained through long-term Dox treatment [114]. Cells were changed to drug free media for 2 passages prior to use in an experiment. Cell line MDA-MB-231-luc-D3H2LN was purchased from Caliper Life Sciences (Hopkinton, MA). Both cell lines were cultured in

15% FBS enriched DMEM with 1% penicillin-streptomycin under 5% CO2 and 95% humidity at 37 ºC.

4.2.9 Cytotoxicity Evaluation

MDA-MB-231-luc-D3H2LN cells were seeded at 5×104 cells/well in a 24 well-plate

(BD Biosciences, San Jose, CA) in DMEM with 15% FBS. Polyplexes made of 1 µg luciferase siRNA and varying ratios of P(SiDAAr)5Peg3 in 15 µL of DMEM were gently dropped into each well followed by mixing with the culture medium. DMEM containing polyplexes were replaced with fresh media after 16 hr and MTT assay was carried out

after 3 days. After that, cell culture medium was removed and replaced with 150 µL of

MTT (0.5 mg/mL in colorless DMEM) and incubated for 3.5 hr in cell culture conditions.

Thereafter, MTT solution was replaced with 250 µL of DMSO and incubated for another

30 min at 37 ºC. One hundred and fifty µL of the dissolved solution was taken out to a clear 96-well plate and absorbance was measured at 595 nm using a plate reader

(SpectraMax M2, Molecular Devices Inc., Sunnyvale, CA). Cell viability was calculated as (abs of treated cells)/(abs of untreated cells)×100%.

4.2.10 Polyplexes Intracellular Accumulation by Flow Cytometry

The cellular uptake of 1/4 N/P siRNA polyplexes with FITC-conjugated

P(SiDAAr)5Peg3 described in section 2.3 was examined in both MDA-MB-231-luc-

D3H2LN and MCF-7/Adr cell lines. Cells were seeded in 6 well plates in 15% DMEM.

After 24 hrs, polyplexes made of 3 µg of siRNA and 12 µg of TRITC-conjugated

P(SiDAAr)5Peg3 were added to each well, incubated for 4 hrs, and then flow cytometry was carried out to quantify intracellular polyplex accumulation.

4.2.11 siRNA-mediated Gene Silencing in vitro

For ubiquitous gene knockdown, I assayed firefly luciferase gene because it has no apparent effect on the regulatory machinery of the cell. For luciferase gene silencing, cells were seeded at a density of 2×105 cells/well in 24 well plates. DMEM with 15%

FBS and 1% penicillin-streptomycin was used for all transfections. Polyplex solution containing 2 µg of siRNA was added to each well and incubated for 24 hrs. Thereafter, cells were washed with PBS and lysed with 100 µL of Reporter Lysis Buffer (Promega,

Madison, WI). Twenty µL of cell lysate and 50 µL of Luciferase Assay Reagent

(Promega) were added into a Nunc F96-MicroWell White Polystyrene plate (Nunc,

Thermo Scientific). Luciferase activity was then measured using a Luminescence Plate

Reader (Wallac 1420 VICTOR2™, PerkinElmer, Waltham, MA).

Cell protein levels were analyzed via BCA protein assay. The BCA assay kit was purchased from Thermo Scientific (Rockford, lL). To perform the assay, 10 µL of cell lysate and 150 µL of freshly prepared BCA reagent (1 part of reagent A and 50 parts of reagent B) were added to a 96 well-plate and incubated at 37 ºC for 30 min. Absorbance was measured at a wavelength of 562 nm using a Microplate Reader (Molecular Devices,

Sunnyvale, CA).

The in vivo delivery of luciferase siRNA was described in Appendix B.

4.2.12. Testing Target Protein Knockdown Effectiveness in Enhancing Dox Uptake and as Chemotherapy

To examine the effectiveness of functional siRNA in down-regulating its target protein in drug resistant MCF-7/Adr, cells were incubated with anti-ABCB1 siRNA

(Dharmacon, Lafayette, CO) for 24, 48 and 72 hrs in 6-well plate. Two µg of anti-

ABCB1 siRNA were used for each condition with no solution change during incubation time. After incubation time as above, cells were washed, trypsinized and incubated with 3

µL of phycoerythrin-conjugated anti-ABCB1 antibody (eBioscience, San Diego, CA) at

0.1 µg/µL for 30 min on ice bath. The cells were then washed twice and kept in flow buffer (PBS with 10% FBS and 0.1% NaN3 sodium azide) for flow cytometry. All flow

cytometry was done with freshly collected cells using a LSR II System (BD Biosciences,

Franklin Lakes, NJ).

To analyze the effect of anti-ABCB1 siRNA on the cellular uptake of Dox, Dox solution at 2500 ng/mL and the unlabeled anti-ABCB1 siRNA P(SiDAAr)5Peg3 polyplexes were added to cells in a 6-well plate with MCF-7/Adr cells. After 4.5 hrs, the cells were washed, trypsinized and collected for flow cytometry to examine the accumulation of Dox in the cell.

To examine the effect of anti-ABCB1 siRNA in enhancing the Dox cytotoxicity, the cells were seeded in a 96 well plate at 5,000 cells/well. After 24 hrs of cell attachment, anti-ABCB1 siRNA polyplexes and Dox solution were co-dosed to the cells with 0.5µg of siRNA/well and a range of Dox concentrations from 100 to 10,000 ng/mL. Treatments were extended for 3 days and MTT was conducted thereafter to examine the effect of anti-ABCB1 transfection on doxorubicin toxicity.

4.3 Results

Physical characterization of siRNA Polyplexes

Particle size and ζ-potential measured via dynamic light scattering showed that at lower N/P ratios (1/2 and 1/3), P(SiDAAr)5Peg3 polyplexes had in average smaller hydrodynamic size than P(SiDAAr)5 polyplexes. For each polymer, as the polymeric ratio increases, the size of the polyplexes was reduced. As the polymeric ratio increased, both P(SiDAAr)5 and P(SiDAAr)5Peg3 had increased ζ-potentials (Table 4.1). However,

PEG conjugation significant reduced the ζ-potentials at low polymeric ratios but not at higher ones (1/5 and 1.6). The ζ-potential reduced from 34.3 to 8.7 mV at N/P of 1/2, and again from 35.5 to 26.2 mV at N/P of 1/3. In addition, all siRNA polyplexes had positive zeta potentials. Please refer to appendix C for the distribution of size and zeta potential measurements.

TEM images showed round polyplexes well dispersed on the TEM grid (Fig 4.1). Gel electrophoresis data showed complete siRNA complexation at 1/4 N/P ratio as indicated by no observed free siRNA migration on the gel (Fig 4.2 A).

siRNA Stability in the Presence of RNAse

In the presence of RNAse, free siRNA showed significant fragmentation within 30 min; however, there was no degradation of siRNA when complexed as 1/4 N/P ratio

P(SiDAAr)5Peg3 polyplexes (Fig 4.2 B).

Polyplex Stability in the Presence of Competing Polyanions

Polyanion competition assay showed high polyplex integrity and stability in the presence of competing heparin. The polyplexes remained intact up to a heparin concertation of 2500 µg/mL. At 2500 µg/mL, free siRNA migrated down the gel, signifying a partial release of siRNA. When reached an even higher heparin concentration (5000 µg/mL), the near complete release of siRNA took place and a migrating siRNA band with roughly equal band size as that of free siRNA was observed

(Fig 4.3).

Figure 4.1 Morphological analysis of siRNA polyplexes.

TEM images of 1/4 siRNA/P(SiDAAr)5Peg3 polyplexes at (A) low (×23000, bar=1μm) and (B) high (×68000, bar=0.5μm) magnifications. Two images were taken of the polyplexes-1 at low resolution and 1 at high resolution.

Figure 4.2 P(SiDAAr)5Peg3 completely complexed siRNA and protected it from RNAse degradation at N/P ratio of 1/4.

(A) Agarose gel electrophoresis of a range of polyplexes from N/P ratio of 1/0.5 to 1/6.

The un-complexed siRNA was shown as white bands on the gel. Complete complexation took place at N/P of 1/4 and 1/6. At 1/6 siRNA/polymer w/w ratio, there is complete masking of siRNA taking place, shielding it from the interaction with CybrGreen dye, thus showing no band in the well on the gel. (B) siRNA fragmentation is an indicator of its degradation. The siRNA fragmentation will result in an increase of absorbance. siRNA degradation assay showed no increase in siRNA absorbance in 1/4 P(SiDAAr)5Peg3 polyplexes, indicated polymer protection of siRNA from RNAse degradation. Free siRNA without polymer complexation showed significant siRNA fragmentation, thus increased absorbance. *: p<0.0001. n=3.

Figure 4.3 Polyanion competition assay revealed the polyplex stability.

Polyanion competition assay revealed siRNA polyplex stability in the presence of varying concentrations of heparin solutions, as visualized via gel electrophoresis. A partial siRNA release was observed at 2500 µg/mL. At 5000µg/mL of heparin, the total release of siRNA was observed.

Cellular Uptake of siRNA Polyplexes

The intracellular uptake of the polyplexes using TRITC-conjugated P(SiDAAr)5Peg3 polymer showed uniform polyplexes distribution among MDA-MB-231-luc-D3H2LN and MCF-7/Adr cells, evident by a narrow peak in the fluorescence intensity. Uneven distribution of polyplexes among cells would result in a wide fluorescence peak under flow cytometry. Also, at a short polyplex incubation time (4.5 hrs), rapid polyplex internalization was observed (Fig 4.4).

Polyplex Cytotoxicity

Toxicity was observed only with the highest polymeric ratio (N/P of 1/6) among the examined samples (Fig 4.5A). The in vitro transfection depended on the polyplex composition. As shown in Fig 4.5B, the transfection levels significantly increased with increasing polymeric ratio, until reaching a plateau at N/P of 1/4. Because 1/4 polyplexes produces similar target knockdown with the least amount of polymer and the least toxicity (Fig 4.5A), I selected this ratio for the following experiments.

Anti-ABCB1 siRNA Down-regulates ABCB1 Protein and Enhances the Toxicity of

Dox

Next, I determined if high levels of polyplex uptake in hard-to-transfect MCF-7/Adr can be translated into high levels of functional siRNA transfection. At 72 hrs, ~70% of the target anti-ABCB1 protein was down-regulated, based on the shift in the fluorescence intensity curve. Anti-ABCB1 siRNA polyplexes also enabled a significantly greater level of Dox intracellular accumulation, shown by the shift in Dox fluorescence intensity. To

correlate ABCB1 target transcript knockdown with its effect on Dox cytotoxicity, I observed Dox toxicity to increase by around 90% based on the absorbance level following drug incubation for 3 days.

4.4 Discussion

Critical issues for siRNA delivery include effective incorporation of siRNA into the polymeric vector, forming polyplexes that have tolerable toxicities, able to protect siRNA from degradation, have strong polyplex stability and achieving effective transfection. In this study, I show that P(SiDAAr)5Peg3 effectively facilitated siRNA transfection in metastatic and drug resistant breast cancer cell lines, and ABCB1 siRNA transfection effectively enhanced the cytotoxicity of Dox upon co-treatment.

To enhance the effectiveness as well as the biocompatibility of PEI, we utilized PEG decorated L-arginine modified oligo (alkylaminosiloxane) grafted poly (ethylenimine) polymer, P(SiDAAr)5Peg3. The PEG chains on the polymer could undergo detachment through hydrolytic degradation, which enables a better endocytosis of the polyplexes

[174]. Our previous work showed that this polymer had lower toxicity compare to PEI

[173]. It has also been shown that long linear PEG was able to exert balanced steric hindrance and minimize the lysis of RBC [175]. The arginine on our polymer could also play a role in the swift uptake of polyplexes due to its ability to conduct transmembrane localization.

Figure 4.4 The cellular uptake of 1/4 TRITC-conjugated P(SiDAAr)5Peg3 polyplexes examined by flow cytometry.

The level of intracellular uptake of P(SiDAAr)5Peg3 polyplexes in (A) MDA-MB-231- luc-D3H2LN and (B) MCF-7/Adr cell lines. Level of fluorescence quantified by flow cytometry showed that siRNA uptake was rapid and uniform in both MDA-MB-231-luc-

D3H2LN and MCF-7/Adr cell lines at a short incubation time (4 hrs). The level of polyplexes uptake was equivalent in both MDA-MB-231-luc-D3H2LN and MCF-7/Adr cell lines.

Table 4.1 Average Hydrodynamic Diameter (Dh) (nm), Polydispersity Index (PDI) and Zeta Potential (mV) of P(SiDAAr)5 and P(SiDAAr)5Peg3 Polyplexes at Different siRNA to Polymer Weight Ratios. Three samples were measured per polyplex ratio.

siRNA/P(SiDAAr)5 Size, Mean (SD) PDI (SD) Zeta, Mean (SD)

1/2 546 (242) 0.13 (0.1) 34.3 (24.7)

1/3 558 (290) 0.17 (0.1) 35.5 (25.7)

1/4 565 (177) 0.20 (0.02) 28.3 (3.3)

1/5 598 (134) 0.51 (0.4) 21.9 (11.2)

1/6 364 (40) 0.18 (0.01) 34.1 (16.9)

siRNA/P(SiDAAr)5Peg3 Size, Mean (SD) PDI (SD) Zeta, Mean (SD)

1/2 327 (170) 0.21 (0.06) 8.7 (18.3)

1/3 295 (230) 0.64 (0.43) 26.2 (13.7)

1/4 350 (147) 0.34 (0.09) 30.0 (11.4)

1/5 231 (67) 0.52 (0.29) 42.5 (11.1)

1/6 240 (22) 0.46 (0.19) 36.3 (16.0)

100

Figure 4.5 Cytotoxicity and transfection efficiency of siRNA/P(SiDAAr)5Peg3 polyplexes in MDA-MB-231-luc-D3H2LN cell line.

(A) Cell toxicity after 72 hr of polyplexes treatment. *: p<0.05 compared to the control group, n=2. The polyplexes up to N/P of 1/5 showed no toxicity to the cells. However, at

1/6 N/P ratio, cytotoxicity was observed. (B) Luc-siRNA transfection with polyplexes at varying N/P ratios. Transfection level increased at increasing polymeric ratio. Highest transfection levels were achieved at 1/4, 1/5 and 1/6 N/P ratios. The increased transfection level at N/P of 1/4 compared to 1/2 and 1/3 showed the effect of complete siRNA complexation. *: p<0.01 compare to control group, n=2.

The polyplexes underwent self-assembly and were able to protect siRNA from degradation. They achieved over 70 fold siRNA knockdown in vitro in MDA-MB-231- luc-D3H2LN.

101

The efficacy of siRNA polyplexes depended on their hydrodynamic size and surface potential. siRNA polyplexes had a relatively wide range of size and zeta potential (Table

4.1). The smaller hydrodynamic size of P(SiDAAr)5Peg3 polyplexes compared to

P(SiDAAr)5 showed the effectiveness of high molecular weight linear PEG in forming more compact polyplexes. The observation that PEG reduced the zeta potential of the polyplexes at lower N/P ratios (1/2 and 1/3) but not high N/P ratios was likely due to the effect of PEG in masking the cationic charge at low N/P ratios. However, as polymer to siRNA ratio increases, the polyplexes may undergo structure rearrangement causing a greater exposure of PEI to be on the surface, which reduces the PEG exposure to the surface and results in higher polyplex charge. The mild cationic surface charge of the polyplexes at the optimal ratio effectively facilitates transfection without exerting cytotoxicity PEG may also assist in forming well dispersed polyplexes by forming a hydrated shell around the polyplexes.

One method to examine the polyplex integrity and complexation strength in the presence of serum (anionic proteins), is through visualizing its stability in the presence of competing polyanions.

102

Figure 4.6 The effect of anti-ABCB1 siRNA on ABCB1 surface protein expression, Dox accumulation and cell cytotoxicity. (A) ABCB1 surface protein expression was significantly down-regulated after 3 days of anti-ABCB1 siRNA polyplexes incubation as examined by flow cytometry. (B) Co- treatment of anti-ABCB1 siRNA polyplexes and Dox solution enhanced Dox uptake, examined by flow cytometry. (C) Anti-ABCB1 siRNA polyplexes and Dox solution co-

103

treatment for 3 days at varying Dox concentrations showed significantly enhanced cellular cytotoxicity examined by MTT assay.

My results show that the complexation strength between 1/4 w/w siRNA polyplexes and 1/1/4 w/w DNA-siRNA polyplexes (from the previous study) [173] were comparable, signifying that siRNA polyplexes had a loosened structure at the absence of DNA (Fig

4.3) because it takes a higher polymer ratio to complete complex siRNA than DNA- siRNA mixture.

Drug resistance and metastasis are the two most challenging issues in cancer treatment. Therefore, both metastatic MDA-MB-231-luc-D3H2LN cell line and the chemo- and transfection-resistant MCF-7/Adr cell line were selected for this study. The uniform internalization of the polyplexes observed in this study could be caused by the enhanced cell membrane penetration of arginine (Fig 4.4) [176].

Degree of siRNA complexation is an important factor in the transfection level of siRNA polyplexes. The plateau of siRNA transfection from 1/4 onward showed that the significantly enhanced transfection from 1/2 to 1/4 polyplexes was a result of the degree of nucleotide complexation (Fig 4.5).

ABCB1 is a major contributor of multi-drug resistance in MCF-7/Adr. ABCB1 actively pumps out chemo-drugs such as Dox, thus lowers the drug concentration inside the cells. Studies have shown that effective anti-ABCB1 siRNA transfection was able to

104

reverse the multi-drug resistance characteristics and sensitize cells for chemotherapy

[177]. MCF-7/Adr was known to have an ABCB1 efflux pump over-expression. The swift ability of the polyplexes to enhance Dox uptake within 4.5 hrs was in contrast with the effective surface protein knockdown that took place after 3 days. This could due to the fact that when endocytosis of polyplex takes place, free Dox in solution was also taken up through the same mechanism, thus causing an enhanced Dox accumulation in the cell. The half-life and degradation time of ABCB1 protein is a deciding factor for the time it takes for the level of protein to decrease. Anti-ABCB1 siRNA showed its effectiveness as a combinational therapy with Dox. However, at a higher Dox concentration, the cytotoxicity level leveled off, suggested the need to overcome other mechanisms of resistance in order to completely eradicate drug resistant cells (Fig 4.6).

4.5 Conclusions

In conclusion, I have shown an arginine and PEG decorated polymer P(SiDAAr)5Peg3 to effectively deliver siRNA. The overall results suggested that at an optimal composition, the polyplexes were able to achieve a complete siRNA complexation with good cytocompatibility, protect siRNA from degradation, and effectively transfect cells in both metastatic and drug resistant breast cancer cell lines. By effectively targeting ABCB1 in drug resistant cells to enhance Dox cellular uptake as well as cytotoxicity, we further demonstrated the potential of functional siRNA polyplexes as a combination therapy with drugs to treat resistant cancer cells. Future studies are needed to enhance the drug effect in resistant cells using multiple functional siRNAs.

105

4.6 Acknowledgements

TEM was conducted at the Cleveland Clinic Imaging Core with the assistance of Mei

Yin.

106

CHAPTER V

CO-DELIVERY OF DNA AND SIRNA VIA ARGININE-RICH PEI-BASED

POLYPLEXES

Keywords: combination therapy, nonviral gene delivery, cancer therapy, cationic polymers, cellular uptake

5.1 Introduction

The discovery of siRNA has drawn much attention to gene therapy in the last few years, particularly for its usefulness in targeting a number of genetic diseases [178-181].

However, challenges remain in gene therapies, such as low transfection efficiencies, particularly with non-viral vectors, and achieving therapeutic effects at tolerable doses.

Many diseases are the result of a single gene mutation. Yet single gene disorders are not common in the population. Therefore, simultaneously targeting multiple genetic components and regulatory pathways pose the possibility of reaching a greater therapeutic potential for a number of diseases, including cancer [182]. In addition, the different physical characteristics of DNA and siRNA (size, morphology, rigidity and charge) could cause them to have different electrostatic interaction strengths with cationic polymers like PEI, thus changing the binding affinity of the polyplexes [183].

Polyethylenimine (PEI) is a commonly used nucleic acid delivery vehicle [184]. It condenses nucleotides through electrostatic interactions, protects them from nuclease degradation and facilitates endosomal escape [185, 186]. The cationic charges on the polyplex surface also promote cellular uptake through adsorptive endocytosis [187, 188].

Despite its promising features, a number of drawbacks hinder it from achieving its therapeutic potential. For example, interaction of the high cationic charge of PEI with the anionic cell surface can destabilize the plasma-membrane and induce toxicity [189, 190].

PEI could also interact with negatively charged serum proteins (such as albumin) and red blood cells, causing protein precipitation or red blood cell aggregation and lysis.[191]

Therefore, chemical modifications are usually necessary to mitigate the effect of high cationic charge of PEI [192].

In this study, to attenuate toxicity, enhance biocompatibility and improve cellular uptake, we modified PEI with 5 molar equivalents of L-arginine modified oligo

(alkylaminosiloxane) graft and 3 molar equivalents of PEG, forming P(SiDAAr)5Peg3.

The previous study had shown better transfection with P(SiDAAr)5 than with PEI in vitro due to L-arginine residues that assist in adsorptive endocytosis, enhancing membrane permeability, and facilitating nuclear localization [193]. P(SiDAAr)5’s siloxane derivative is biodegradable and is excreted in the urine [194]. Furthermore, P(SiDAAr)5 following PEG and folic acid conjugation showed greater tumor accumulation than the unconjugated polymer [175].

DNA and siRNA delivery systems have many similar characteristics, such as facilitating efficient cellular uptake and rapid endosomal escape. However, their distinctive structural and chemical characteristics determine their different complex formation properties. DNA is usually several kilobase pairs in length while siRNA is only

21-23 base pairs. Topologically, siRNA forms A-form helixes, while DNA forms B-form helixes. The A-form siRNA has a larger diameter and a smaller rise per base pair, and thus is stiffer than the B-form DNA [195]. Therefore, the structural differences between

DNA and siRNA may result in DNA-siRNA polyplexes having different structural properties than DNA- or siRNA-only polyplexes [196]. This study shows that DNA- siRNA co-delivery significantly enhances both nucleotides’ gene transfection levels compared to that of single-nucleotides. Also, comparing polyplexes, P(SiDAAr)5Peg3 polyplexes have significantly better cytocompatibility than PEI polyplexes with the same nucleotide composition.

5.2 Materials and Methods

The materials and the methods to synthesize P(SiDAAr)5Peg3 and conduct siRNA transfection were listed below.

5.2.1 Materials

3-(2-aminoethylamino) propyl-methyl-dimethoxysilane, 25-kDa branched PEI, L- arginine, O-(2-Carboxyethyl)polyethylene glycol PEG-3kDa hydrochloride, L-aspartic acid, 1-Ethyl-3-(3-dimethylaminopropyl) carbodiimide hydrochloride (EDC), N- hydroxysuccinimide (NHS) and 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium

109

bromide (MTT) were purchased from Sigma-Aldrich (St. Louis, MO). Fetal bovine serum (FBS) and Trypsin/EDTA were obtained from Gibco (Grand Island, NY). The luciferase assay system and reporter lysis buffer were purchased from Promega

(Madison, WI). Scrambled siRNA (s-siRNA) and YOYO-1 DNA intercalating dye were purchased from Life Technologies (Carlsbad, CA). The 1% Gold Agarose Reliant Gel

System was purchased from Lonza (Allendale, NJ). Anti-luciferase siRNA (luc-siRNA) was purchased from Dharmacon, Inc. (Lafayette, CO). The BCA Protein Assay Kit was purchased from Thermo Scientific (Rockford, lL). Cell culture media Dulbecco's

Modified Eagle's Medium (DMEM), DPBS, penicillin and streptomycin were purchased from our Cell Services’ Media Preparation Core.

5.2.2 pDNA Amplification and Purification

The Pgl3 (luciferase reporter vector) and control (noncoding) plasmid DNA from

Promega (Madison, WI) were cloned via E. coli (JM109, Promega) according to the manufacture’s technical manual. Pdna was purified using Qiagen EndoFree Plasmid

Mega Kit (Qiagen, Valencia, CA) according to the manufacturer’s instructions and re- suspended in Dnase-free Dh2O. The Pdna concentration were measured with a Nanodrop

1000 (Thermo Scientific, Waltham, MA).

5.2.3 Polymer Synthesis

Oligo-(alkylaminosiloxane) (SiDA), arginine conjugated oligo amino alkyldialkoxymethylsilane (SiDAAr) and arginine modified oligo-(alkylaminosiloxane)-

110

graft-PEI ((P(SiDAAr)5) (5:1 graft: PEI molar ratio) were synthesized as previously described [193, 197] (Appendix D).

Conjugation of PEG to P(SiDAAr)5: Briefly, the free acid group of CH3-PEG-COOH

(0.1 g, 3 mmol equivalents) was dissolved in 2 mL of MES buffer at pH=6. The acid group of PEG was activated using 0.1208 g (3.5/5 mmol) of EDC as a coupling agent and

0.0725 g (3.5/5 mmol) NHS as a catalyst in MES buffer at pH=6 for 3.5 hr at 4 °C. Three molar equivalents of the activated PEG were then reacted with 1 molar equivalent of

P(SiDAAr)5 (0.3 g, 3 mmol equivalents) in 2 mL of PBS (pH=7.5) overnight (16 hours).

The synthesized polymer was dialyzed for 2 days using 12KDa MWCO dialysis membrane (Spectrum Labs, Rancho Dominguez, CA) to remove the un-reacted elements.

The polymer was then filtered through a 0.22 µm sterile filter and lyophilized. The

1 polymer was dissolved in D2O for H NMR analysis (δ, ppm) on a Varian Mercury 300

(300 MHz) spectrometer. The resonances were measured referencing to TSP (4.66 ppm) using TSP. The yield for P(SiDAAr)5Peg3 is 140 mg (35%). Please refer to Appendix D for detailed synthetic procedures of polymer P(SiDAAr)5.

5.2.4 Polymer/Nucleotide Complexation

The polymers were dissolved in MilliQ water (dH2O) and filtered through a 0.22 µm sterile polyethersulfone filter (Millipore, Darmstadt, Germany). SiRNA was dissolved at

1 mg/mL in RNase-free dH2O. For characterization studies, DNA and polymers were dissolved separately at 1 mg/mL in DNase- and RNase-free dH2O. For cell culture experiments, DNA and polymer were dissolved separately at 1 mg/mL in serum-free

111

Dulbecco's Modified Eagle Medium (DMEM). Polyplexes were formed by combining

DNA, siRNA and polymer solutions, pipetting to mix thoroughly, and self-assembling for

30 min. PEI was used in parallel with P(SiDAAr)5Peg3 for comparison purposes.

5.2.5 Size/Zeta Potential Measurements

Polymers PEI and P(SiDAAr)5Peg3, and the corresponding polyplexes with different

D/S/P ratios, were measured for mean hydrodynamic diameters and zeta potentials in dH2O via dynamic light scattering (DLS) using the NICOMP 380 ZLS Particle Sizing

System (Santa Barbara, CA). All polymer and polyplex measurements were done in 3 replicates. Average and standard deviation were calculated for each polyplex composition.

5.2.6 Examine the Complete Nucleotide Complexation via Gel Retardation Assay

Agarose gel electrophoresis was used to determine if there was complete nucleotide complexation in each polyplex formulation. The polyplex solutions were prepared at varying polymer/nucleotide (P/N) weight/weight ratios. Naked non-coding plasmid DNA

(Ø-pDNA), luc-siRNA and polyplex solutions were loaded in a 1% Gold Agarose gel

(Lonza, Allendale, NJ) in TrisAcetate-EDTA buffer at pH 8.0. For control wells, 2 µL of

DNA or siRNA was used per well. For other wells, 2 µL of DNA was used per well, and the quantities of the siRNA and the polymer were adjusted accordingly. The free nucleotides and polyplexes were run on an electrophoresis system (Bio-Rad, Hercules,

CA). The gels were stained using SYBR Green (Thermo Scientific) and visualized under a Carestream Molecular Imaging Gel Logic 112 Imaging System (Carestream Health,

Inc., Rochester, NY).

112

5.2.7 Examine the Polyplexes Morphology via Transmission Electron Microscopy

(TEM)

1/1/4 D/S/P polyplexes were prepared in dH2O by vigorous pipetting and incubating for 30 min for polyplex self-assembly. Thereafter, 10 µL of the polyplex solution was dropped on TEM grids coated with silicon monoxide stabilized formvar films (Ted Pella

Inc., Redding, CA). After drying, the samples were stained with 2% uranyl acetate for 7 minutes and imaged using a Tecnai G2 TEM microscope (FEI, Hillsboro, Oregon).

5.2.8 Polyanion Competition Assay

The relative stability of 1/1/4 D/S/P(SiDAAr)5Peg3 polyplexes was investigated using a heparin polyanion competition assay, where the nucleotides released from polyplexes were measured using gel electrophoresis. Polyplex solutions were first prepared by mixing 5 µg of DNA, 5 µg of siRNA and 20 µg of polymer and incubating in dH2O for

30 minutes. The polyplexes were then exposed to heparin sodium (Sigma-Aldrich, St.

Louis, MO) solution (10 µL) of varying concentrations for 15 min. Then the samples were run on a 1% Gold Agarose gel (Lonza, Allendale, NJ) in TrisAcetate-EDTA buffer, and the gel stained and visualized by the same method described in section 2.6.

5.2.9 Cell Culture

Cell line MCF-7 was purchased from American Type Culture Collection (Manassas,

VA). Luciferase-expressing breast cancer metastatic cell line MDA-MB-231-luc-

D3H2LN was purchased from Caliper Life Sciences (Hopkinton, MA). MCF-7/Adr was

113

obtained from Prof. Batra’s laboratory at the University of Nebraska Medical Center,

Omaha, and maintained by culturing in media with 100 ng/mL doxorubicin. MCF-7/Adr was incubated in drug-free medium for two passages before being used in experiments.

All cell lines were cultured at 37 °C in humidified air containing 5% CO2. DMEM with

15% FBS and 1% penicillin-streptomycin was used in all cell-culture related studies, including MTT, in vitro transfections, flow cytometry and confocal microscopic studies.

Cell detachment was done using trypsin/EDTA at 37 °C.

5.2.10 In Vitro Cytotoxicity Assay

The MTT assay was used to evaluate polymer and polyplex toxicities in vitro in

MCF-7 and MCF-7/Adr cell lines. For the 72-hr polymer toxicity assay, 5×104 cells were seeded per well in a 24-well plate. For the 24-hr polyplex toxicity assay, 5×105 cells were seeded per well in a 24-well plate (BD Biosciences, San Jose, CA) and 1 µg of DNA was used per well. The 72-hr polymer MTT was designed to assess the polymer’s differential growth inhibition effects, while the 24-hr polyplex MTT was designed to assess the polyplex’s toxicity under transfection conditions. After 24 hr of cell attachment, all media was removed, and each well received fresh media containing polymers or polyplexes.

After a designated incubation time, media was removed again and replaced with 150 µL of MTT (2 mg/mL in DMEM). Plates were incubated at room temperature overnight according to the standard MTT protocol. On the next day, the MTT was removed, the wells received 500 µL DMSO, and the plates were incubated for 3.5 hr at 37 °C to allow the substrate to dissolve. One hundred and fifty µL of the dissolved solution was then taken out and absorbance (abs) measured at 575 nm using a clear 96-well plate. Cell

114

viability (as a %) was calculated as (abs of sample-background)/(abs of control- background) ×100%.

5.2.11 In Vitro Luciferase Transfection Assay

Luciferase-DNA (luc-DNA)’s transfection levels in luc-DNA or luc-DNA/s-siRNA polyplexes were measured in MCF-7 and MCF-7/Adr cells. Luc-siRNA’s transfection levels in luc-siRNA or luc-siRNA/Ø-pDNA polyplexes were measured in MDA-MB-

231-luc-D3H2LN cells. One µg of luc-DNA/well was used for MCF-7, 2 µg of luc-

DNA/well for MCF-7/Adr, and 2 µg of luc-siRNA/well for MDA-MB-231-luc-D3H2LN.

MCF-7 and MCF-7/Adr cells were seeded at 5×105 cells/well and MDA-MB-231-luc-

D3H2LN cells at 2×105 cells/well in 24 well plates. After 24 hours, the media was replaced and the polyplex suspensions prepared in DMEM were added drop-wise to each well. After 1 day of transfection, the cells were washed twice with PBS and lysed with

100 µL of Reporter Lysis Buffer (Promega). Luciferase activity in relative luminescence units (RLU) was measured by combining 25 µL of lysate and 50 µL of Luciferase Assay

Reagent and measured in an opaque 96-well plate (Sigma-Aldrich) using a micro plate reader (FLUOstar OPTIMA (BMG Labtech, Durham, NC).

Protein levels were determined using BCA protein assay (Thermo Scientific). Ten µL of each lysate and 150 µL of BCA reagent were combined in clear 96-well plates and incubated at 37 °C for 30 minutes. Absorbance was measured at a wavelength of 562 nm using an Absorbance Microplate Reader (Molecular Devices, Sunnyvale, CA).

115

5.2.12 DNA Cellular Uptake/Flow Cytometry

YOYO-pDNA intercalation was carried out by adding 30 µL of 10 µM YOYO-1 to 1

µg of 1 mg/mL DNA and incubating for 30 minutes at room temperature in the dark.

MCF-7 and MCF-7/Adr cells were seeded in 6-well plates and allowed to grow to 70% confluence. Polyplex solutions were added to each well and flow cytometry was carried out after 24 hr. One µg of DNA per well was used for MCF-7 cells and 2.5 µg of DNA per well were used for MCF-7/Adr cells. Cells were washed twice with PBS, trypsinized and collected for flow cytometry (LSR II System, BD Biosciences). Fluorescence detector 530 ± 15 nm was used for YOYO-1 detection. A population of 10,000 freshly collected cells was used per condition.

5.2.13 DNA Intracellular Visualization

MCF-7 and MCF-7/Adr cells were seeded on coverslips placed in 6 well plates and allowed to reach 60% confluence. At 60% confluence, polyplexes formed with YOYO-1 stained DNA were added to each well in FBS enriched media. One µg of DNA per well was used for MCF-7 cells and 2.5 µg of DNA for MCF-7/Adr cells. After 24 hr incubation, the cells were rinsed with DPBS and fixed with 4% paraformaldehyde

(Sigma-Aldrich). Coverslips were then mounted on clean glass slides using DAPI

ProLong Gold Anti-fade Mounting Media (Invitrogen, Eugene, OR). The cells were then evaluated with a laser scanning confocal microscope (Leica TCS-SP II Spectral Laser

Scanning Confocal Microscope, Leica Microsystems GmbH, Wetzlar, Germany).

116

5.2.14 Statistical Analysis

Data are presented as mean ± standard deviation. Statistical analysis of the differences between any mean values was performed using 1-sided Student’s t-test.

Statistical significance was defined at p≤0.05; and highly significant at p≤0.01 and p≤0.001.

5.3 Results and Discussion

Polymer Synthesis and Characterization.

Polyethylene glycol modified L-arginine oligo(alkylaminosiloxane) grafted poly(ethylenimine) (P(SiDAAr)5Peg3) was synthesized using a route shown in Scheme

5.1 and Appendix D.2. Synthesis was carried out with reference to prior works published on oligo(alkylaminosiloxane) and L-arginine oligo(alkylaminosiloxane) grafted poly(ethylenimine) [193, 197]. SiDA is a mixture of ring and linear oligomers, with more linear architecture than ring architecture. The characterization of linear vs. ring architecture of the oligomer has not been performed. We assumed our synthesis will yield the same architectures as published in prior publications. Following our previously established procedure, the polymer P(SiDAAr)5 was first synthesized by carbodiimide chemistry using EDC/NHS with molar ratio of arginine (Ar) to SiDA (5:1) and SiDAAr to PEI (5:1).

The synthesis was carried out in 4 steps, conjugated SiDA, arginine, PEI and PEG components. The work I conducted is PEG conjugation to the P(SiDAAr)5 polymer. PEI,

SiDA and arginine effectively contributed to the transfection efficiency of the polymer

117

and PEG reduced the cytotoxicity of the polymer. SiDA is biodegradable, and has shown to be able to complex DNA and transfect DNA in HeLa cells [197]. In addition, arginine has been reported to enhance endocytosis and assist in nuclear translocation [193].

The conjugations of arginine, aspartic acid, PEI and PEG were carried out using carbodiimide chemistry. P(SiDAAr)5 polymer may have different numbers of SiDAArs per PEI unit, since the number of SiDAAr conjugated to each PEI cannot be controlled in the coupling process. The same also applies to PEG conjugation. Some P(SiDAAr)5 may be conjugated to more than 3 PEG chains and other less than 3. This gives rise to the higher polydispersity of P(SiDAAr)5Peg3 compared to PEI from which it was synthesized, measured by particle sizer. The Rapid stirring of the solution in the SiDAAr and PEG coupling process can minimize the dispersities in the polymer conjugation. In PEG conjugation to P(SiDAAr)5, since both PEI and arginine possess amide groups, PEG could be conjugated directly to PEI as well as to the to the arginine arms. However, the different conjugation site of PEG does not change the properties of the polymer. See

Appendix D for detailed synthesis procedures of P(SiDAAr)5.

Figure 5.2 represents 1H NMR spectra of L-arginine oligo (alkylaminosiloxane) grafted poly(ethylenimine) (P(SiDAAr)5) (300 MHz) and polyethylene glycol modified

L-arginine oligo (alkylaminosiloxane) grafted poly(ethylenimine) (P(SiDAAr)5Peg3) (300

MHz). PEI with 25kDa has 5 times as much H proton (2906 H protons) as that of

1 (SiDAAr)5 (500 H protons). The integrals of SiDAAr (133.38) and PEI (618.16) on the H

NMR spectrum also showed approximately 5 times more H protons from PEI than from

118

(SiDAAr)5 (Fig 5.1A). The resonance at 0.11 ppm is-(bs, Si-CH3). Chemical shifts at δ =

0.89 and δ = 1.15 are from the two ethylene groups of SiDA. Chemical shifts at δ =1.76 and δ = 2.26 are from SiDA (-Si-CH2CH2CH2NH-); Chemical shifts at δ =3.69-3.09 are

SiDAAr (-HCCH2CH2CH2NH-). Chemical shifts at δ =3.0-2.4 are (-NHCH2CH2-, PEI

1 ethylene). The H NMR spectrums of P(SiDAAr)5 is dominated by the PEI resonances: (-

HNCH2CH2-) at 3.0-2.5 ppm. Nevertheless, the ethylene resonances of the SiDAAr units were clearly visible. The resonances of PEG (-OCH2CH2-) were also visualized at δ

=3.7-3.6 (Fig 5.1).

P(SiDAAr)5 has roughly 3 times as much ethylene groups as 3 equivalents of 3kDa

PEG, as shown by the integrals in Figure 5.2B. Branched PEI has much more abundant amino groups than the hydroxyl groups on PEG. Therefore, from the moles inputted in the reaction mixture, all of the activated PEGs should react with P(SiDAAr)5. However, the large molecular weight of PEG may result in incomplete reaction with P(SiDAAr)5; thus a higher reaction time may be needed to generate a higher polymer yield. The resolution of Figure 5.2B is not ideal; higher polymeric concentration or a longer time of

NMR acquisition could be adapted to enhance the resolution of the spectrum.

The final polymer P(SiDAAr)5Peg3 has a white sponge-like morphology as dried polymer. The polymer was water soluble, and was stable at room temperature. It was stored in vacuum as dry polymer to eliminate the influence of moisture in the air. For

o experiments, it was reconstituted in dH2O and kept in -20 C for storage.

119

In the previously published works on SiDA, it has been shown that the basic hydrolysis condition generates a mixture of linear and cyclic oligomers with linear/cyclic ratio of 3.42/1 [197]. The study also showed that the oligomer interacted effectively with

DNA and was able to bind DNA through electrostatic interactions, as revealed via the electron microscopy and DNA retardation assays. In HeLa cells, the basic hydrolysis of

SiDA showed efficient transfection through proton capture process of endocytosis. Its effect toward DNA transfection should not be affected by the linear vs. ring forms of the oligomer. Since my work focuses primarily on the impact of SiDA, arginine and PEG modification on PEI for siRNA and DNA-siRNA co-delivery, the presence of ring and linear forms of SiDA would not affect the cytotoxicity as well as transfection levels in the cells.

1 In the published work of Morris et al on the synthesis of P(SiDAAr)n [193], the H

NMR (D2O) spectrum showed δ (ppm) of 0.1-(bs, Si-CH3), 1.66-arginine(-

HCCH2CH2CH2NH-), 1.86-arginine (-HCCH2CH2CH2NH-), 3.24-arginine(-

HCCH2CH2CH2NH-), 3.86-arginine (-HCCH2CH2CH2NH-), and 3.3-2.5 (-NHCH2CH2-,

PEI ethylene). The chemical shifts were close to the ones I have identified on the 1H

NMR spectrum in Figure 5.2A. The NMR spectrum showed other peaks than that of the polymer, which could be from the side products of the polymer. Morris et al also demonstrated that among all of the P(SiDAAr)n derivatives, 5 units of (SiDAAr) were the most effective in reducing the PEI cytotoxicity and achieving the highest transfections in

KB cell line.

120

The average hydrodynamic diameter (Dh) of polymers PEI, P(SiDAAr)5 and

P(SiDAAr)5Peg3 measured by DLS increased after each step of modification, from PEI to P(SiDAAr)5 to P(SiDAAr)5Peg3 (Table 5.1). PEI and P(SiDAAr)5 have similar polydispersity indices (PDI) (0.3-0.4), whereas P(SiDAAr)5Peg3 has a greater PDI (0.62).

Zeta potential increased slightly from PEI to P(SiDAAr)5 as a result of the amine groups present on arginine. Compared to P(SiDAAr)5, PEG-conjugation reduced

P(SiDAAr)5Peg3’s zeta potential.

Polyplexes Characterization.

PEI and the lack of control in the conjugation process of SiDAAr and PEG to PEI, resulting in some PEIs having more than 5 (SiDAAr)s and 3 PEGs and others less. The variations in zeta potential among the 3 polymers were not significant. The hydrophilic properties of PEG resulted in P(SiDAAr)5Peg3 to have significantly higher hydrodynamic diameters than PEI and P(SiDAAr)5. First, the linear and hydrophilic nature of PEG gave it a larger hydrodynamic diameter in dH2O compare branched PEI, which is a lot more condensed and hydrophobic. Please refer to appendix C for the distribution of size and zeta potential measurements.

121 P(SiDAAr)5 polyplexes had larger particle sizes and broader standard deviations compared to P(SiDAAr)5Peg3 polyplexes (Table 5.2). This demonstrated that

P(SiDAAr)5Peg3 was able to form polyplexes with more consistent hydrodynamic sizes than P(SiDAAr)5 did. The increase of P(SiDAAr)5Peg3 polymer size compared to PEI and

P(SiDAAr)5 could potentially assist P(SiDAAr)5Peg3 to form more homogeneous polyplexes. PEG could reduce the formation of aggregates.

P(SiDAAr)5Peg3 polyplexes had drastically reduced zeta potentials compare to

P(SiDAAr)5 polyplexes at lower polymeric ratios (1/2 and 1/3) (Table 5.2). This was a result of PEG masking the cationic charges from the polyplexes. As the polymeric ratio increases (1/4 and above), PEG can no longer shield the high cationic charges due to the possible re-structuring and rearrangement of the polyplexes, thus resulting in an increased zeta potentials.

From Table 5.2, several trends were observed: 1) PEI polyplexes had a larger size standard deviation and a larger zeta potential compared to P(SiDAAr)5Peg3 polyplexes, proving P(SiDAAr)5Peg3 was able to form more uniform and neutral polyplexes than PEI.

This can be attributed to the presence of SiDAAr in assisting nucleotide complexation and PEG in reducing the cationic charges in P(SiDAAr)5Peg3.

In DNA, siRNA and DNA-co-siRNA polyplexes (Table 4.1, 5.1 and 5.2), broad polydispersity in H2O was observed. This was contributed by the large PDI among PEI,

P(SiDAAr)5 and P(SiDAAr)5Peg3, the incorporation of both DNA and siRNA, and the

122 possibilities of polyplex aggregates. Also, the self-assembly process makes us unable to modulate the variation in size and zeta potentials. Comparing the morphologies of siRNA,

DNA and DNA-siRNA polyplexes, I found siRNA polyplexes much more dispersed

(forming loose polyplexes), whereas DNA polyplexes were much more coherent. DNA- siRNA polyplexes, on the other hand, are a lot more dispersed. Darker color represents a coherent structure, whereas lighter is an indicator of a looser structure.

A gel retardation assay showed P(SiDAAr)5Peg3’s nucleotide-binding capacity with different DNA and siRNA weight ratios based on the nucleotides’ electrophoretic mobility (Fig 5.3). Compared to the controls (free luc-DNA and s-siRNA), the nucleotides showed no band on the gel, dim band in the wells with low P/N polyplexes and no band in the wells with high P/N polyplexes. The absence of DNA migration signifies complete DNA-polymer complexation. Absent of visible DNA in the wells signifies complete DNA masking by the polymer, therefore giving no exposure to the staining dye (Fig 5.3).

DSL and TEM were both incorporated in the characterization of DNA, siRNA and

DNA-co-siRNA polyplexes (Table 5.1, Table 5.2, Fig 5.4). DLS gave the hydrodynamic diameters of the particles whereas TEM gave the morphologies as well as the estimated diameters of the polyplexes as dry samples. Hydrodynamic diameters revealed the hydrophilic layer around the particle under brownian motion, whereas TEM estimated the size absent of the hydration layer. Often, the hydrodynamic diameter measured by DLS is

123 larger than the size estimated by TEM. TEM, however, reveals the morphologies and possible aggregates of the polyplexes.

Polyplex morphologies were visualized under TEM in both high and low magnifications (Fig 5.4). The polyplexes have an almost spherical shape. Furthermore, polyplex hydrodynamic sizes and zeta potentials were measured via dynamic light scattering. Both nucleotide to polymer ratio and the polymeric material play a role in the size and surface charge of the polyplexes. On average, P(SiDAAr)5Peg3 polyplexes are smaller and have a smaller size range compared to PEI polyplexes (Table 5.2).

Regarding polydispersity, most polyplexes have a PDI ranging from 0.2-0.5. Comparing average zeta potentials, P(SiDAAr)5Peg3 polyplexes have on average lower zeta potentials than PEI polyplexes. For both PEI and P(SiDAAr)5Peg3 polyplexes, surface charges increase as the polymer ratio increases in the polyplexes. P(SiDAAr)5Peg3 polyplexes have both positive and negative surface charges, whereas all PEI polyplexes have positive charges (Table 5.2). Finally, both siRNA polyplexes and siRNA-co-DNA polyplexes made from P(SiDAAr)5Peg3 have similar hydrodynamic sizes.

1/1/4 P(SiDAAr)5Peg3 polyplexes were incubated with heparin solution of varying concentrations, then separated using gel electrophoresis. Polyplexes incubated in 5 mg/mL of heparin released their nucleotides.

124 Polyplexes Stability in the Presence of Competing Polyanions.

A heparin displacement assay was used to examine the integrity and stability of

DNA/siRNA/ P(SiDAAr)5Peg3 polyplexes in the presence of heparin sodium (Fig 5.5).

Different concentrations of heparin sodium solutions were used to gradually induce DNA and siRNA decomplexation within the polyplex. When no heparin was added, neither

DNA or siRNA was released from the polyplexes. This is an indication of the polyplexes’ stability in the absence of competing polyanions. With increasing heparin concentration, polyplexes were sufficiently stable, until reaching 5 mg/mL heparin concentration. At 5 mg/mL, polyplex decomplexation took place, where both DNA and siRNA were replaced by competing polyanions and released from the polyplexes.

Figure 5.1 Chemical structures of starting materials for polymer P(SiDAAr)5Peg3. (A) 3-

(2-aminoethylamino) propyl-methyl-dimethoxysilane, (B) L-arginine, (C) 25-kDa

125

branched PEI, (D) O-(2-Carboxyethyl)polyethylene glycol PEG-3kDa hydrochloride, (E)

L-aspartic acid, (F) 1-Ethyl-3-(3-dimethylaminopropyl) carbodiimide hydrochloride

(EDC), (G) N-hydroxysuccinimide (NHS).

Scheme 5.1 Synthesis of Polyethylene glycol modified L-arginine oligo

(alkylaminosiloxane) grafted poly(ethylenimine) P(SiDAAr)5Peg3.

The final polymer has a white sponge-like morphology after lyophilization. The polymer is water soluble, and is stable in room temperature. However, we stored it in vacuum as dry polymer to eliminate the influence of moisture in the air. It is reconstituted in dH2O and frozen at -20oC for experiments.

126

1 Figure 5.2 Spectral characterization of P(SiDAAr)5Peg3. H NMR spectra of (A) L- arginine oligo (alkylaminosiloxane) grafted poly(ethylenimine) (P(SiDAAr)5) (300 MHz), and (B) Polyethylene glycol modified L-arginine oligo (alkylaminosiloxane) grafted poly(ethylenimine) (P(SiDAAr)5Peg3) (300 MHz) containing 5 mmol equivalents of

127

SiDAAr per 1 mmol equivalent of PEI and 3 mmol equivalents of PEG for 1 mmol of

P(SiDAAr)5, using EDC as a coupling agent and NHS as a catalyst.

Table 5.1 Mean Hydrodynamic Diameter (Dh) (nm) and Zeta Potential (mV) of Polymer Solutions in dH2O. Three measurements were done per polymer. PDI: polydispersity index.

Polymer Average (Stdev) PDI (Stdev) zeta (Stdev)

PEI 9 (1) 0.40 (0.10) 8.1 (6.3)

P(SiDAAr)5 11 (1) 0.38 (0.11) 12.1 (3.7)

P(SiDAAr)5Peg3 41 (11) 0.62 (0.38) 10.4 (5.7)

Cytotoxicity Evaluation.

In the 72-hour dose-dependent polymer toxicity assay of MDA-MB-231-luc-

D3H2LN cells, both P(SiDAAr)5 and P(SiDAAr)5Peg3 had significantly reduced toxicity compared to PEI [(p≤0.05) at 25 and 50 µg/mL] (Fig 5.6A). Compared to P(SiDAAr)5,

P(SiDAAr)5Peg3 showed further cytotoxicity reduction. Cytotoxicity evaluation following 24 hr of PEI and P(SiDAAr)5Peg3 polyplex exposure showed that both MCF-7 and MCF-7/Adr at high polymer concentrations (1/1/4 and 1/2/6 polyplexes), inhibited cell growth (Fig 5.6B, C). However, P(SiDAAr)5Peg3 achieved significantly higher cell viability compared to PEI in MCF-7/Adr in a 1/2/6 D/S/P ratio.

Luciferase DNA Transfection.

Luciferase DNA transfection using luc-DNA-co-s-siRNA polyplexes in MCF-7 and

MCF-7/Adr cell lines showed that MCF-7 was more prone to DNA transfection than

128

MCF-7/Adr (Fig 5.7A, B). With a 2 fold lower dose of DNA, MCF-7 still had higher transfection levels (from 10 to 100-fold) than MCF-7/Adr. In MCF-7, where both P/N ratios of 1 and 2 were used, polyplexes with P/N of 2 showed a greater transfection level than P/N of 1. In polyplexes with both polymers in both cell lines, s-siRNA incorporation resulted in enhanced transfection expressions. In MCF-7, 1/2/3 (D/S/P ratio) PEI polyplexes achieved 26-fold higher transfections than 1/O/1 PEI polyplexes. In

P(SiDAAr)5Peg3 polyplexes, both 1/1/2 and 1/2/3 polyplexes expressed an over 10 fold higher transfection level compared to 1/O/1 polyplexes. In addition, the 2 polymers affected different cell lines differently. In MCF-7, P(SiDAAr)5Peg3 polyplexes showed significantly higher transfections, whereas in MCF-7/Adr, higher transfections were observed with PEI polyplexes.

129

Figure 5.3 Polyplex agarose gel electrophoresis identifies the un-complexed nucleotides.

Complete nucleotide (Ø-pDNA and anti-luc siRNA) and polymer (P(SiDAAr)5Peg3) complexation was shown in all D/S/P w/w ratios.

130

Figure 5.4 Polyplexes characterization via TEM.

TEM images of 1/2 DNA/P(SiDAAr)5Peg3 and 1/1/4 DNA/siRNA/P(SiDAAr)5Peg3 polyplexes at low and high magnifications. To specifically examine the morphological differences between DNA and DNA-siRNA polyplexes, DNA polyplexes showed to be more separated and well-dispersed, while DNA-siRNA polyplexes formed more aggregates. Both polyplexes compositions were roughly spherical in morphology. Darker as well as lighter colored polyplexes were observed. The darker polyplexes probably had a greater amount of DNA supercoils while the lighter polyplexes had a greater amount of siRNA in it. The images for each polyplex composition are representatives of 6 TEM images.

131

Table 5.1 Average Hydrodynamic Diameter (Dh) (nm) and Zeta Potential (mV) of PEI and P(SiDAAr)5Peg3 Polyplexes at Different Weight Ratios. Three samples were measured per polyplex ratio.

DNA/siRNA/PEI Size (stdev) PDI (stdev) Zeta (stdev)

1/O/1 276 (156) 0.44 (0.21) 40.6 (6.0)

1/1/2 301 (148) 0.34 (0.25) 39.1 (8.7)

1/0.5/1.5 454 (160) 0.29 (0.10) 38.3 (9.0)

1/2/3 386 (142) 0.34 (0.21) 38.8 (3.2)

1/O/2 313 (81) 0.37 (0.17) 48.2 (9.4)

1/1/4 439 (127) 0.23 (0.18) 36.9 (10.1)

1/0.5/3 387 (200) 0.33 (0.09) 53.6 (23.1)

1/2/6 356 (85) 0.29 (0.15) 67.2 (18.0)

DNA/siRNA/ Size (stdev) PDI (stdev) Zeta (stdev) P(SiDAAr)5Peg3

1/O/1 289 (23) 0.24 (0.04) 3.9 (2.5)

1/1/2 386 (47) 0.37 (0.15) -2.4 (18.1)

1/0.5/1.5 307 (66) 0.24 (0.14) 14.3 (15.8)

1/2/3 302 (56) 0.29 (0.35) 20.2 (7.3)

1/O/2 280 (17) 0.28 (0.04) 21.2 (14.5)

1/1/4 284 (29) 0.26 (0.10) 22.7 (19.4)

1/0.5/3 307 (14) 0.29 (0.06) 22.6 (6.0)

1/2/6 319 (44) 0.31 (0.07) 17.9 (8.8) siRNA/DNA/ Size (stdev) PDI (stdev) Zeta (stdev)

132

P(SiDAAr)5Peg3

1/O/2 233 (18) 0.25 (0.07) 20.4 (15.0)

1/0.5/3 228 (11) 0.19 (0.06) 11.0 (1.0)

1/1/4 272 (87) 0.31 (0.23) 29.1 (16.0)

1/2/6 297 (85) 0.30 (0.14) 18.3 (7.5)

PDI: Polydispersity Index, D/S/P: DNA/siRNA/polymer, S/D/P: siRNA/DNA/polymer.

133

Figure 5.5 Stability of polyplexes measured through the polyanion competition assay.

1/1/4 P(SiDAAr)5Peg3 polyplexes were incubated for 15 minutes with heparin solutions of varying concentrations from 100 µg/mL to 5000 µg/mL. Gel electrophoresis was used to visualize the migration of free DNA and siRNA as polyanion replacement took place.

Polyanion replacement is defined as the replacement of DNA and siRNA molecules by heparin molecules, causing nucleotide polyplexes to release their encapsulated nucleotides. As shown on the gel, at a concentration of 5000 µg/mL, the polyplexes incubated in heparin released their nucleotides.

134

Figure 5.6 Polymer and polyplex cytotoxicity assays in drug-sensitive and resistant cell lines.

In all 3 experiments (A, B, and C), PEI was used as a reference to compare to

P(SiDAAr)5 and P(SiDAAr)5Peg3. (A) Polymer solution MTT assay was conducted after

72 hr of polymer incubation in MDA-MB-231-luc-D3H2LN cell line. n=8. *: statistically significant compared to PEI polymer at the same concentration. †: statistically significant compared to P(SiDAAr)5 polymer at the same concentration. Compared to PEI, both

P(SiDAAr)5 and P(SiDAAr)5Peg3 had a much more reduced cytotoxicity. Also, compared

135

P(SiDAAr)5 and P(SiDAAr)5Peg3, P(SiDAAr)5Peg3 showed reduced toxicity at higher polymer concentrations (25ng/mL and 50 ng/mL). (B) Polyplex cytotoxicity in MCF-7 after 24 hr polyplexes incubation. In drug sensitive cell line MCF-7, P(SiDAAr)5Peg3 generated reduced polyplexes toxicity at N/P ratios of 1/1/4 and 1/2/3. (C) Polyplex cytotoxicity in MCF-7/Adr after 24 hr of polyplex incubation. In drug resistant cell line

MCF-7/Adr, P(SiDAAr)5Peg3 polyplexes resulted in reduced cytotoxicity compared to

PEI polyplexes. (B)&(C): n=2 *: statistically significant compared to the control group; †: statistically significant compared to PEI polyplexes of the same composition. *, † p≤0.05;

**, †† p≤0.01; ***, ††† p≤0.001. All data were represented as mean±standard deviation.

136

Figure 5.7 DNA transfection levels of both PEI and P(SiDAAr)5Peg3 luc-DNA-co-s- siRNA polyplexes in MCF-7 and MCF-7/Adr cells.

Data represented as mean ± standard deviation. n=2. (A) In MCF-7 cell line,

P(SiDAAr)5Peg3 polyplexes achieved a significantly higher transfection level than PEI polyplexes with the same composition. (B) In MCF-7/Adr cell line, PEI polyplexes

137

achieved significantly higher transfection levels than P(SiDAAr)5Peg3 polyplexes with the same composition. siRNA incorporation significantly enhanced DNA transfection levels in both cell lines. †: statistically significant compared to PEI polyplexes of the same composition. *, † p≤0.05; **, †† p≤0.01; ***, ††† p≤0.001.

Luciferase expression reduction in MDA-MB-231-luc-D3H2LN cells using luc- siRNA and Ø-pDNA polyplexes showed that ØDNA significantly enhanced siRNA’s transfection levels (Fig 5.8). Consistent with MCF-7, P(SiDAAr)5Peg3 showed greater transfections than PEI. The maximum transfection level (97% luciferase expression knock-down) was achieved with 1/2/6 D/S/P polyplexes.

DNA Intracellular Uptake in MCF-7 and MCF-7/Adr Cell Lines.

The intracellular uptake of DNA examined in MCF-7 by flow cytometry showed that

1/1/4 polyplexes had a greater DNA uptake level than 1/O/2 polyplexes (Fig 5.9A).

Confocal microscopy showed higher DNA cellular internalization in 1/2/3 polyplexes than in 1/1/2 polyplexes, and both expressed higher DNA cellular internalization than

1/O/1 polyplexes (Fig 5.9B).

In MCF-7/Adr, both flow cytometry and confocal microscopy showed a higher DNA uptake in 1/1/4 polyplexes compared to 1/O/2 polyplexes. However, 1/2/6 polyplexes showed a reduced DNA uptake, confirmed by both flow cytometry and confocal imaging

(Fig 5.10). In all studies, DNA uptake is affected by polyplex compositions and the presence of siRNA in the polyplex.

138

The PDIs of the polyplexes measured via particle sizer using dynamic light scattering were similar among the commercial branched PEI (Sigma Aldrich), the intermediate polymer P(SiDAAr)5, and P(SiDAAr)5Peg3. The relatively high polydispersity of the polymers was a result of the lack of uniformity of the commercial PEI and the lack of control in the conjugation process of SiDAAr and PEG to PEI, resulting in some PEIs having more than 5 (SiDAAr)s and 3 PEGs and others less. The variations in zeta potential among the 3 polymers were not significant. The hydrophilic properties of PEG resulted in P(SiDAAr)5Peg3 to have significantly higher hydrodynamic diameters than

PEI and P(SiDAAr)5. First, the linear and hydrophilic nature of PEG gave it a larger hydrodynamic diameter in dH2O compare branched PEI, which is a lot more condensed and hydrophobic.

In DNA, siRNA and DNA-co-siRNA polyplexes (Table 4.1, 5.1 and 5.2), broad polydispersity was observed. This was contributed by the large PDI among PEI,

P(SiDAAr)5 and P(SiDAAr)5Peg3, the incorporation of both DNA and siRNA, and the possibilities of polyplex aggregates. Also, the self-assembly process makes us unable to modulate the variation in size and zeta potentials. Compare the morphologies of siRNA,

DNA and DNA-siRNA polyplexes, I found siRNA polyplexes much more dispersed

139

DSL and TEM were both incorporated in the characterization of DNA, siRNA and

DNA-co-siRNA polyplexes. DLS gave the hydrodynamic diameters of the particles whereas TEM gave the morphologies as well as the estimated diameters of the polyplexes as dry samples. Hydrodynamic diameters revealed the hydrophilic layer around the particle under brownian motion, whereas TEM estimated the size absent of the hydration layer. Often, the hydrodynamic diameter measured by DLS is larger than the size estimated by TEM. TEM however, reveals the morphologies and possible aggregates of the polyplexes.

Polyplex sizes and zeta potentials have major effects on the particle interactions with the biological system, including the endocytotic pathways the particles undertake and the ability of the particles to perform endosomal escape and to assist in nucleotide transfections. Zeta potential is the electrokinetic potential of the particles dispersed in colloidal forms. Investigating the size and zeta properties of DNA polyplexes and

DNA/siRNA polyplexes, I found that in the same N/P ratio, there are no major differences in the size of the polyplexes. Moreover, I did not observe significant differences between the polydispersities and zeta potentials of 1/O/2 and 1/1/4 polyplexes.

Comparing PEI and P(SiDAAr)5Peg3 polymers, however, P(SiDAAr)5Peg3 polyplexes had a smaller zeta potential than PEI polyplexes. This is a result of PEG modifications, which reduced the surface charge of the polyplexes. From the above observation, we can conclude that the differences in transfections between DNA and DNA-siRNA polyplexes was not caused by the sizes and zeta potentials of the polyplexes.

140

The polyplexes were formed via self-assembly process through electrostatic interactions. The polymer sizes as well as charges are major factors for the formation of polyplexes, the release of the nucleotides in the cell through endosomal escape, as well as effectiveness in transfections. The level of polydispersity in P(SiDAAr)5Peg3 will also cause formed polyplexes to have different sizes

In our PEI-based delivery systems, we showed that DNA and siRNA enhances each other’s transfection levels. Luc-DNA was combined with s-siRNA, and luc-siRNA combined with Ø-pDNA. S-siRNA and Ø-pDNA had no genetic interferences with functional nucleotides. Polymer characterization demonstrated that polymer size increased significantly from P(SiDAAr)5 to P(SiDAAr)5Peg3, partly due to the hydrodynamic size of the long linear PEG chains (Table 5.1). P(SiDAAr)5 polyplexes had larger particle sizes and broader standard deviations compare to P(SiDAAr)5Peg3 polyplexes (Table 5.2). This demonstrated that P(SiDAAr)5Peg3 was able to form polyplexes with more consistent hydrodynamic sizes than P(SiDAAr)5 did. The increase of P(SiDAAr)5Peg3 polymer size compared to PEI and P(SiDAAr)5 could potentially assist P(SiDAAr)5Peg3 to form more homogeneous polyplexes. PEG could reduce the formation of aggregates.

P(SiDAAr)5Peg3 polyplexes had drastically reduced zeta potentials compare to

141

possible re-structuring and rearrangement of the polyplexes, thus resulting in an increased zeta potentials.

This can be contributed by the presence of SiDAAr in assisting nucleotide complexation and PEG in reducing the cationic charges in P(SiDAAr)5Peg3.

The reduction in zeta potential from P(SiDAAr)5 to P(SiDAAr)5Peg3 indicated the

PEG’s ability to neutralize polymeric charges. PEI and P(SiDAAr)5Peg3 polyplex demonstrated different characteristics as carrier materials (Table 5.2). Polyplex size and zeta potential have a major effect on the particle interactions within the biological system.

Zeta potential is the electrokinetic potential of the particles dispersed in colloidal forms.

The particle size and zeta potential play a major role in the endocytotic pathways that the particles undertake and the ability of the particles to perform endosomal escape and to assist in nucleotide transfections. Investigating the size and zeta properties of DNA polyplexes with DNA/siRNA polyplexes, I found that in the same N/P ratio, there is no major difference in sizes between the compositions, indicating that the siRNA and DNA components do not result in significant differences in the sizes of the polyplexes.

Moreover, comparing the polydispersity of 1/O/2 and 1/1/4 polyplexes in PEI and

P(SiDAAr)5Peg3 polymers, Ifound no significant difference, indicating that DNA and siRNA components do not increase the dispersity nature of the polyplexes. Comparing

142

the zeta potentials, I saw that 1/O/2 and 1/1/4 polyplexes showed no major difference in zeta potentials. However, P(SiDAAr)5Peg3 polyplexes have a smaller zeta potential than

PEI polyplexes. This is a result of PEG conjugation, which reduces the surface charges of the polyplexes.

Figure 5.8 Polyplexes’ siRNA transfections in MDA-MB-231-luc-D3H2LN cell line.

In siRNA/DNA/polymer (PEI or P(SiDAAr)5Peg3) polyplexes, Ø-pDNA significantly enhanced luc-siRNA’s transfection levels in MDA-MB-231-luc-D3H2LN cell line. In this condition, P(SiDAAr)5Peg3 is a more effective transfection agent compared to PEI.

Data as mean ± standard deviation. n=2. †: statistically significant compared to PEI polyplexes of the same weight ratio. *, † p≤0.05; **, †† p≤0.01; ***, ††† p≤0.001.

143

Figure 5.9 P(SiDAAr)5Peg3 polyplexes’ intracellular DNA quantification in MCF-7 cell line.

(A) Flow cytometry revealed that DNA-siRNA polyplexes have a greater DNA cellular uptake than DNA polyplexes. (B) Intracellular DNA was visualized by confocal microscopy showed that polyplexes with siRNA (1/2/3 and 1/1/4) have a stronger DNA internalization than those without siRNA (1/O/1 and 1/O/2). Images were captured using a 40× objective lens. Bar = 100 μm.

144

Figure 5.10 P(SiDAAr)5Peg3 polyplexes’ intracellular DNA quantification in MCF-

7/Adr cell line.

(A) Flow cytometry quantified P(SiDAAr)5Peg3 polyplexes’ DNA uptake levels in MCF-

7/Adr cell line. 1/1/4 polyplexes showed over 2-fold higher DNA uptake than 1/O/2 polyplexes. The level dropped in 1/2/6 polyplexes due to polymer masking. (B) Confocal microscopy visualized YOYO-1-DNA’s cellular internalization. 1/1/4 polyplexes showed a stronger DNA internalization than 1/O/2 polyplexes. Fluorescence intensity dropped in

145

1/2/6 polyplexes, confirming the flow cytometry results. Images were captured using 80× objective lens. Bar=50 μm.

Compared to PEI polyplexes, P(SiDAAr)5Peg3’s PEG groups have charge shielding effects. Furthermore, PEI has more positive charges per weight unit compare to

P(SiDAAr)5Peg3. Both factors result in PEI polyplexes having a higher average zeta potential than P(SiDAAr)5Peg3 polyplexes. Since polyplex formation is a self-assembly process, the nucleotide exposure to the exterior cationic charge results in negative polyplex surface charge due to nucleic acids’ anionic nature. On average,

P(SiDAAr)5Peg3 formed smaller and more uniform sized polyplexes than those formed with PEI polyplexes because P(SiDAAr)5Peg3’s PEG chains could help in preventing polyplexes aggregation.

In all complexation ratios, complete nucleotide conjugation was observed, indicated by the lack of nucleotide migration on the electrophoresis gel (Fig 5.3). Upon complete nucleotide masking by the polymer, SYBR Green dye had no binding with the nucleotide, thus showing no fluorescent band in the well. 1/1/4 polyplexes visualized by TEM showed well dispersion at low magnification. At high magnification, nearly spherical individual polyplexes polyplexes were observed (Fig 5.4).

Heparin is a large polyanion. It is a glycosaminoglycan, linear and negatively charged with sulfonate groups. It can be found inside cells and on the surface of the cell membrane [198]. Heparin is also a major component of the extracellular matrix of some

146

tissues, including connective tissues.[198] Therefore, we chose heparin as a competition anion to evaluate the polyplex stabilities. 1/1/4 D/S/P polyplexes showed no disintegration at concentrations of up to 5 mg/mL heparin solution, demonstrating the polyplex stability in an anionic environment. The data thus suggest the polyplex would remain stable in the presence of serum or plasma proteins which is critical for its in vivo application. Premature release of the cargo in the circulation could make the polyplex ineffective as a transfecting agent.

Polymer toxicities were determined after 3 days of polymer incubation because 3 days incubation allows a more precise examination of the polymer effect on cell proliferations (Fig 5.6A). Polyplex toxicities, however, were examined in transfection conditions (1 day incubation) to more clearly reflect the polyplexes’ growth inhibitory effects during the transfection period (Fig 5.6B, C). Although they contain equal amounts of DNA, 1/1/4 polyplexes weigh twice as much as 1/O/2 polyplexes, and showed higher toxicity. Although it appears that polyplexes have similar toxic effect in MCF-7 and

MCF-7/Adr, the doses of polyplexes used in MCF-7/Adr was twice as much of each polyplex as MCF-7. Therefore MCF-7/Adr is in reality more resistant to polyplex toxicity than MCF-7. Most polyplexes had a mild growth-inhibitory effect. However,

P(SiDAAr)5Peg3 polyplexes showed less toxicity than PEI at higher polyplex concentrations (ex. 1/2/6).

Concerning DNA transfections, PEI and P(SiDAAr)5Peg3 show different transfection levels which is also cell line dependent. For example, P(SiDAAr)5Peg3 shows higher

147

transfection in MCF-7 and PEI in MCF-7/Adr (Fig 5.7-8). In the process of acquiring drug resistance, MCF-7/Adr’s membrane lipid composition undergoes changes, forming more compact cell membranes and increased lysosomes compared to MCF-7. This change affects its cellular endocytosis, membrane trafficking and intracellular processing/transport, which contribute to decreased DNA transfection [119, 199]. Thus

MCF-7/Adr could require a highly cationic delivery vehicle for interaction with cell membrane to trigger cellular uptake process. Drug sensitive cell lines (MCF-7 and MDA-

MB-231-luc-D3H2LN), however, have relatively higher membrane permeabilities due to the fluid nature of their membranes, allowing endocytosis. In drug sensitive cells,

P(SiDAAr)5Peg3 may better facilitate intracellular nucleotide release and nuclear localization due to arginine and siloxane’s aforementioned properties, therefore enhancing endocytosis.

Significantly enhanced luciferase protein knock-down was observed in siRNA-pDNA polyplexes (Fig 5.8), yet their average size and zeta potential do not differ significantly from siRNA polyplexes. Therefore, there may be other mechanisms that assist in transfection, such as the polyplexes’ endosomal escape and intracellular trafficking. DNA uptake studies confirmed that high nucleic acid internalization and nuclear localization rates are not the only influences on transfection rates. For example, free YOYO-DNA’s cellular uptake was significantly higher than that of 1/O/2 polyplexes in MCF-7 (Fig 5.9), yet naked DNA showed insignificant transfections compared to polyplexes (Fig 5.7A).

Also, 1/1/4 P(SiDAAr)5Peg3 polyplexes had approximately 2-fold higher DNA uptake than 1/O/2 polyplexes, yet showed over 6-fold DNA transfection (Fig 5.7A). All these

148

indicate that the polymers not only play a role in DNA’s intracellular uptake, they also affect other parameters that influence gene transfections. In all gene therapies, transfection level is heavily dependent upon the carrier’s ability to overcome extracellular and intracellular barriers, including (1) cellular uptake, (2) endosomal escape and protection from lysosomal degradation, and (3) intracellular polyplex de-complexation for effective gene functioning [200]. Therefore, further studies are needed to understand their intracellular trafficking that could impact transfection.

The decrease in fluorescence from 1/1/4 to 1/2/6 polyplexes (Figure 5.9 and 5.10) can be a result of two factors. One, since we only visualized the fluorescence intensity of

YOYO-DNA, a large quantity of 1/2/6 polyplexes was needed to be ingested in order to internalize the same amount of DNA as that of 1/O/2 or 1/1/4 polyplexes. Therefore, in reality, the quantity of 1/2/6 polyplexes internalized was a lot greater than that of 1/O/2 polyplexes. Another reason is that some of YOYO-1 fluorescence is quenched upon complexation and unquenched upon decomplexation. Therefore I could not conclude if the strong intracellular fluorescence intensity was related exclusively to polyplexes internalization, or if it is due in part by the decomplexation of DNA. One experiment to perform is to use the dye-conjugated polymer instead of YOYO-DNA for the quantification of polyplex internalization.

One observation to note is that 1/2/6 polyplexes showed a lower intracellular DNA uptake, as detected by both flow cytometry and confocal microscopy (Fig 5.10). This may due to YOYO-1-DNA’s fluorescence quenching as a result of polymer and siRNA

149

masking. When the overall ratio of (siRNA and polymer) to DNA increases, DNA masking is likely to result, thus lowering the detected DNA signal.

One possibility for co-delivery to enhance DNA and siRNA’s transfection levels is by alterations in polyplex structures. DNA, which has long and winding strands, and siRNA, which has short and rigid strands, has different PEI binding behaviors. Study has shown differences in hierarchical mechanism of formation between DNA-PEI and siRNA-PEI polyplexes [201]. Further, siRNA undergoes biphasic binding with the polycations and molecular reorganization at low N/P ratios, which is not present with DNA [201]. When combined, they will form different polyplex structures from those formed by single nucleotides, thus potentially affecting the nucleotides’ intracellular release, localization, and interaction with intracellular molecules.

Co-delivery is a promising yet challenging field. siRNA and paclitaxel co-delivery has been achieved by cationic core-shell nanoparticles with enhanced transfection efficiencies [202]. Poly(2-ethyl-2-oxazoline)-PLA-g-PEI, multifunctional chitosan nanoparticles and PEI-functionalized lipid nano-capsules have all been used to explore nucleotide and drug co-delivery to enhance transfection outcomes [203-205]. In this study, DNA and siRNA have been co-delivered in PEI-based polyplexes as helper polyanions to enhance each other’s transfection levels. Mechanistically, both DNA and siRNA are complementary to one other. For example, when siRNA targets one gene and pDNA targets its suppressor gene, the combination may elongate the silencing effects.

150

5.4 Conclusions

This study demonstrated that long, supercoiled pDNA and short, rigid siRNA enhance each other’s gene transfection levels in a cell. Co-delivery of siRNA and pDNA using a PEI-based carrier is a promising approach to maximize gene transfection efficiency. Specifically, 1/1/4 D/S/P polyplexes resulted in higher DNA and siRNA transfection with reduced toxicity than other polyplex compositions tested; thus 1/1/4

D/S/P polyplexes is the best composition for co-delivery of DNA and siRNA among the ones we tested. Such an approach has the potential to achieve synergistic therapeutic effects by simultaneously targeting multiple genetic levels. This study showed that in all of the complexation ratios tested, synergistic effects observed in co-delivery were due to the presence of a helper polyanion (DNA or siRNA). The study also gives incentive for future explorations in designing and implementing multi-nucleic acid delivery strategies, such as functional DNA P53 and siRNA anti-bcl2. A better understanding of co- delivery’s physicochemical, biological and therapeutic characteristics will be essential to future implementation of co-delivery strategies in clinical settings.

5.5 Acknowledgements

Confocal microscopy studies were performed at the Cleveland Clinic Imaging Core with guidance from Dr. John Peterson. TEM was conducted at the Cleveland Clinic

Imaging Core with assistance of Mei Yin.

151

CHAPTER VI

SUMMARY, CONCLUSIONS AND FUTURE DIRECTION

6.1 Summary

The cure for cancer lags despite the best efforts from the medical community. The failure to treat advanced breast cancer can be attributed by the development of drug resistance and metastasis, where the cells undergo epigenetic and genetic mutations to acquire EMT and the CSC phenotype.

The aims of this thesis are 1) to examine the CSC characteristics in breast cancer cell lines and the effects of epigenetic drug SAHA; 2) to analyze the transcriptomic alterations in drug resistant breast cancer cells as a result of epigenetic drug treatments; and 3) To generate a cationic construct to facilitate siRNA and DNA-co-siRNA delivery.

The development of drug resistance and malignant transformation has been correlated with the presence of CSC, which drives metastasis and drug resistance. In chapter 2, I first examined the presence of breast tumor CSC immunophenotype (CD44+/CD24-,

ESA+, CD201+, CD49f+) in drug resistant breast cancer cell line MCF-7/Adr in comparison with metastatic cell line MDA-MB-231 and other breast cancer cell lines, and found MCF-7/Adr to possess the surface markers currently identified for breast CSC.

I further identified the presence of CSC proteins (fibronectin+, vimentin+) as well as

epithelial protein marker E-cadherin using fluorescence microscopy. Next, I evaluated the ability of the cells to form tumors in vitro under different cell culture conditions

(mammosphere formation assay). The migratory rate of MCF-7/Adr in comparison to

MDA-MB-231-D3H2LN in vitro was also examined using migration assay. Despite the fact that MCF-7/Adr possesses all the surface markers of CSC, it showed low migratory ability, where is contributed by the presence of E-cadherin.

Reversing epigenetic patterns can be effective in eradicating drug resistant cells.

Therefore, next, I investigated the effectiveness of histone deacetylation inhibitor SAHA in eliminating MCF-7/Adr. I found that compared to cell lines MDA-MB-231 and hMSC,

SAHA is more effective in killing MCF-7/Adr. Comparing the effect of SAHA on the migratory rate of the cell, the metastatic ability of both hMSC and MDA-MB-231 was hampered significantly by SAHA. Also, I found that 7 days treatment of SAHA did not completely reverse the breast cancer stem-cell immunophenotype possessed by MCF-

7/Adr.

In chapter 3, I further investigated the transcriptomic changes by the epigenetic drugs

DAC and SAHA in reversing drug resistant characteristics. Massively parallel sequencing/RNA-seq was conducted to elucidate the underlying mechanisms by which

DAC and SAHA reverse pathways affecting cellular drug resistance. By quantitative analysis, I identified over 1000 transcripts that exhibited greater than two-fold differential expression (p<0.005) from each drug treatment compare to untreated cells. We found

DAC and SAHA to regulate a large number of the same genes in the above pathways;

however, only DAC significantly down-regulate drug metabolism genes whereas SAHA showed an up-regulation of drug resistant genes ERBB2IP, TGFBR2 and E3F7. This work elucidated the genetic effect of DAC and SAHA that enhances the therapeutic efficacy of chemotherapy in both solution and nano- delivery vehicles.

siRNA has potential as a gene therapy due to its high potency and ability to exert gene knockdown with high specificity. However, poor cellular uptake and rapid degradation by RNAse are major barriers for achieving its therapeutic efficacy. Therefore, it is essential to protect siRNA from RNAse degradation, maintain polyplex stability, and increase transfection efficiency during siRNA delivery. In chapter 4, we investigated the ability of P(SiDAAr)5Peg3, an arginine oligopeptide-grafted and polyethylene glycol modified bPEI, to successfully deliver siRNA to MDA-MB-231-luc-D3H2LN and

MCF7/Adr. First, we examined the physical characteristics (hydrodynamic diameter and zeta potential) of the siRNA polyplexes using dynamic light scattering. Then I examined the polyplexes integrity using a polyanion competition assay. Thirdly, I examined the morphology of the polyplexes using TEM, its ability of the polyplexes to protect siRNA from RNAse degradation, and the polyplexes cytotoxicity using MTT assay. Lastly, I examined the siRNA transfection level in vitro. I showed over 70 fold luciferase expression knockdown by luc-siRNA in metastatic MDA-MB-231-luc-D3H2LN and effective suppression of ABCB1 protein by anti-ABCB1 siRNA in drug resistant MCF-

7/Adr cell line.

154

A wide range of gene therapies are under clinical trials. However, a major challenge is the low transfection level. In chapter 5, I aimed to enhance the transfection levels of

DNA and siRNA through co-delivery approach. I formulated polyplexes with different nucleotide to polymer ratios and examined the effectiveness of each ratio in conducting transfection. Cationic polymer 25 kDa bPEI was used as a standard in comparison with polyethylene glycol modified L-arginine oligo (alkylaminosiloxane)-graft poly

(ethylenimine) P(SiDAAr)5Peg3, the polymer I used for siRNA delivery. Polyplexes at different nucleic acid to polymer weight ratios were characterized and used for transfecting sensitive (MCF-7) and resistant (MCF-7/Adr) breast cancer cell lines. The gene silencing effect of the polyplexes through siRNA transfection was examined in

MDA-MB-231-luc-D3H2LN cell line. The results showed that the polyplexes formed with derivative P(SiDAAr)5Peg3 exhibited lower toxicity compared to polyplexes formed using bPEI. Furthermore, co-delivery using P(SiDAAr)5Peg3 resulted in a maximum of

~20-fold higher DNA transfection and 2-fold higher siRNA transfection as compared to the single nucleotide delivery with the same N/P w/w ratio. Confocal microscopy imaging and flow cytometry analysis demonstrated that the enhanced transfection does not solely depend on the cellular uptake of DNA, suggesting that we need a combination of different therapeutic agents to overcome drug resistance. DNA and siRNA co-delivery could be a promising therapeutic approach to achieve synergistic effects, as it has demonstrated the ability to enhance each nucleotide’s transfection rate in both drug sensitive and resistant breast cancer cell lines. DNA-siRNA co-delivery also has the advantage of simultaneously targeting multiple regulatory pathways in a single formulation to halt and reverse disease progression.

155

In this work, there are a number of interesting findings. First, CSC characteristics were observed in the drug resistant MCF-7/Adr cell line but not in the metastatic MDA-

MB-231 cell line, even though both cell lines possess triple negative and basal phenotypes. This suggests that the existing breast CSC immunophenotype has a strong connection with the level of drug resistance, but not with metastatic ability. This point has never been addressed in the literature. Also, for the first time, this study identified uniform breast CSC surface marker expression (CD44+/CD24-) in a cell line.

In RNA sequencing study, a large number of genes and pathways, including PPAR, mTOR and lysosomal signaling pathways, were similarly affected by DAC and SAHA.

By analyzing the gene alteration patterns based on cellular locations, it was shown that nuclear genes were mostly up-regulated while plasma membrane genes were mostly down-regulated by both DAC and SAHA. The differential expression pattern could possibly be due to the cascade effect of the nuclear genes on the down-stream genes and pathways.

In DNA siRNA co-delivery using PEI-based cationic polymer, it was surprising to find DNA and siRNA enhanced each other’s transfection levels. In the past, co-delivery of functional DNA and siRNA was studied. However, their ability to enhance each other’s transfections as helper anions has not been investigated. This study provides incentives for the future works using functional DNA and siRNA.

156

6.2 Limitations

A number of experimental limitations were present in this work. First, all of the

MCF/7/Adr cell-related works were conducted using the cell line obtained from Dr.

Batra’s group at the University of Nebraska Medical Center, Omaha, and maintained by culturing with 100 ng/mL of doxorubicin in culture media. For CSC characterization, uniform CSC immunophenotype in early MCF-7/Adr passages was observed. However, there was a reversal of the CSC characteristics at a later passage (P36), with increasing

CD24+ sub-population in the culture. Due to the potential differences in the cell line sources and the variations in culture conditions, it is possible that MCF-7/Adr from another source with different culture conditions may not possess 100% of the

CD44+/CD24- immunophenotype. Also, in the study, the possible variations between in vitro and in vivo conditions have not been ddressed, although the cells showed alterations in immunophenotype when cultured in floating mammosphere conditions. As there are different cytokines and growth factors in in vivo growth environments, there was a possibility of obtaining a different immunophenotype when MCF-7/Adr was grown in an in vivo condition. A future study could be to form tumors in mice with stem-like cancer cells, treating with drugs or epigenetic drugs, followed by tumor dissociation and surface immunophenotype characterization to examine in parallel the different states the cells are as cell culture in vitro and as tumors in vivo.

For RNA sequencing, due to the large number of gene outputs, only pathways based on the current knowledge of endocytosis, drug transport and metabolism were analyzed.

DAC and SAHA affected a large number of genes (ex. KIF23, LAMIN B, SYNE1,

157

SYNE2) in a similar way. However, there are still genes and pathways that were not comprehensively examined. The gene distribution patterns on the Volcano Plot had many similarities. However, with the software limitations, it was difficult to identify the exact gene represented by each dot on the Volcano Plot. Therefore, it was not feasible to comparatively analyze the gene fold changes with their statistical significance on the

Volcano plot. One way to better correlate the genes differentially regulated by DAC and

SAHA was to directly compare these two groups without first comparing with the control.

Another way was to manually select the genes with high fold changes from the 2 groups, and conduct analyses on similarly as well as differently affected genes.

The third limitation of this study was the presence of polymer side products, for instance, the mixture of linear and ring forms of SiDA in our polymer P(SiDAAr)5Peg3.

However, I have not further pursued the side-product separations for two reasons. First, the synthetic procedure allows the presence of side products such as P(SiDAAr) with different PEG chains, and the incomplete reaction due to the large molecular weight of

PEI and PEG. However, the different side-products could be separated through column chromatography due to their different architectures, where the ring formed oligomers will be eluted first. However, since I focused on the overall effect of the polymer in assisting nucleotide transfections, structural variations of the polymer should not greatly affect the transfection efficiency.

158

6.3 Final Conclusion and Future Directions

In this study, CSC surface markers in drug resistant breast cancer cells MCF-7/Adr were identified, which could undergo surface marker reversal under different culture conditions (mammosphere or epigenetic drug SAHA treatment). The effectiveness of epigenetic drugs DAC and SAHA in altering the transcriptomes of MCF-7/Adr was also examined, and was found to effectively reverse the genes and pathways affecting drug resistance. Finally, siRNA delivery and DNA-siRNA co-delivery polyplexes were formulated to target breast cancer cells, especially in metastatic and drug resistant cell lines.

There are a number of future studies that could be carried out as a continuation of this work. One is to use functional DNA-siRNA combination for future studies. In this work,

DNA and siRNA have demonstrated to effectively enhance each other’s transfections, therefore upon effective combination, DNA-siRNA co-treatment using P(SiDAAr)5Peg3 could result in superior effect than both DNA and siRNA delivered individually. On the other hand, if functional DNA and siRNA achieve a greater synergistic effect than what was observed with the incorporation of non-coding helper DNA and siRNA, the combination would be highly synergistic in their therapeutic effects. Another area of exploration is to analyze other genes and pathways identified by RNA-seq. This work focused on the different aspects of gene regulations associated with drug resistance.

However, other areas such as resistance to immunotherapy and free radical generation could also be examined in future works.

159

It has been shown that epigenetic drugs were less effective in solid tumor treatment than in blood cancer [206]. However, they have strong synergistic effects with chemotherapy, such as doxorubicin [33]. Much study is needed to find out the optimal drug combinations to use and to formulate treatment strategies. For future studies, epigenetic drugs could be combined with other drug and gene therapies to test their combination effects.

Future in vivo works could be carried out in addition to the studies detailed in the appendix, including using animal model to study the in vivo efficacy of luc-siRNA after multiple injections as well as using functional DNA and siRNA co-delivery strategies.

Specifically, on anti-luciferase siRNA delivery, single dosage of luc-siRNA injection resulted in insignificant luciferase expression knockdowns. However, further analysis on different types of treatment conditions, such as multiple systemic injections of siRNA polyplexes throughout a few days could be conducted to examine the differential effects under different treatment conditions. This will allow a more comprehensive understanding of the clinical value of P(SiDAAr)5Peg3 for gene therapy. Up to this point, no study has shown the clinical effectiveness of the polymer P(SiDAAr)5Peg3.

Finally, this project investigated the presence of breast CSC immunophenotypes in breast cancer cell lines, analyzed the effect of epigenetic drugs on the transcriptomic level of drug resistant cells MCF-7/Adr, and formulated siRNA and DNA-siRNA co-delivery systems to transfect drug resistant and metastatic breast cancer cell lines.

160

REFERENCES

1. Cancer IAfRo: Latest world cancer statistics. Global cancer burden rises to 14.1 million new cases in 2012: Marked increase in breast cancers must be addressed World Health Organization: 2013.

2. Masood S: Estrogen and progesterone receptors in cytology: a comprehensive review. Diagn Cytopathol 8: 475-91, 1992.

3. Dunnwald LK, Rossing MA and Li CI: Hormone receptor status, tumor characteristics, and prognosis: a prospective cohort of breast cancer patients. Breast Cancer Res 9: R6, 2007.

4. Putti TC, El-Rehim DM, Rakha EA, Paish CE, Lee AH, Pinder SE and Ellis IO: Estrogen receptor-negative breast carcinomas: a review of morphology and immunophenotypical analysis. Mod Pathol 18: 26-35, 2005.

5. Walker RA: Immunohistochemical markers as predictive tools for breast cancer. J Clin Pathol 61: 689-96, 2008.

6. Dick JE: Breast cancer stem cells revealed. Proc Natl Acad Sci U S A 100: 3547-9, 2003.

7. Housman G, Byler S, Heerboth S, Lapinska K, Longacre M, Snyder N and Sarkar S: Drug resistance in cancer: an overview. Cancers (Basel) 6: 1769-92, 2014.

8. Gottesman MM, Ludwig J, Xia D and Szakacs G: Defeating drug resistance in cancer. Discov Med 6: 18-23, 2006.

9. Fattal E and Barratt G: Nanotechnologies and controlled release systems for the delivery of antisense oligonucleotides and small interfering RNA. Br J Pharmacol 157: 179-94, 2009.

10. Yang X, Grailer JJ, Rowland IJ, Javadi A, Hurley SA, Matson VZ, Steeber DA and Gong S: Multifunctional stable and pH-responsive polymer vesicles formed by heterofunctional triblock copolymer for targeted anticancer drug delivery and ultrasensitive MR imaging. ACS Nano 4: 6805-17, 2010.

11. Jenkins DE, Hornig YS, Oei Y, Dusich J and Purchio T: Bioluminescent human breast cancer cell lines that permit rapid and sensitive in vivo detection of mammary tumors and multiple metastases in immune deficient mice. Breast Cancer Res 7: R444-54, 2005.

12. van 't Veer LJ, Dai H, van de Vijver MJ, He YD, Hart AA, Mao M, Peterse HL, van der Kooy K, Marton MJ, Witteveen AT, Schreiber GJ, Kerkhoven RM, Roberts C, Linsley PS, Bernards R and Friend SH: Gene expression profiling predicts clinical outcome of breast cancer. Nature 415: 530-6, 2002.

13. Zawia NH, Lahiri DK and Cardozo-Pelaez F: Epigenetics, oxidative stress, and Alzheimer disease. Free Radic Biol Med 46: 1241-9, 2009.

14. Lyko F and Brown R: DNA methyltransferase inhibitors and the development of epigenetic cancer therapies. J Natl Cancer Inst 97: 1498-506, 2005.

15. Sharma S, Kelly TK and Jones PA: Epigenetics in cancer. Carcinogenesis 31: 27-36, 2010.

16. Waddington CH: Towards a theoretical biology. Nature 218: 525-7, 1968.

17. Li Y, Chen H, Hardy TM and Tollefsbol TO: Epigenetic regulation of multiple tumor- related genes leads to suppression of breast tumorigenesis by dietary genistein. PLoS One 8: e54369, 2013.

18. Bird AP: The relationship of DNA methylation to cancer. Cancer Surv 28: 87-101, 1996.

19. Bird A: DNA methylation patterns and epigenetic memory. Genes Dev 16: 6-21, 2002.

20. Press JZ, De Luca A, Boyd N, Young S, Troussard A, Ridge Y, Kaurah P, Kalloger SE, Blood KA, Smith M, Spellman PT, Wang Y, Miller DM, Horsman D, Faham M, Gilks CB, Gray J and Huntsman DG: Ovarian carcinomas with genetic and epigenetic BRCA1 loss have distinct molecular abnormalities. BMC Cancer 8: 17, 2008.

21. Byler S, Goldgar S, Heerboth S, Leary M, Housman G, Moulton K and Sarkar S: Genetic and epigenetic aspects of breast cancer progression and therapy. Anticancer Res 34: 1071-7, 2014.

22. Sarrio D, Rodriguez-Pinilla SM, Hardisson D, Cano A, Moreno-Bueno G and Palacios J: Epithelial-mesenchymal transition in breast cancer relates to the basal-like phenotype. Cancer Res 68: 989-97, 2008.

23. Makki J, Myint O, Wynn AA, Samsudin AT and John DV: Expression distribution of cancer stem cells, epithelial to mesenchymal transition, and telomerase activity in breast cancer and their association with clinicopathologic characteristics. Clin Med Insights Pathol 8: 1-16, 2015.

24. Hill JJ, Tremblay TL, Cantin C, O'Connor-McCourt M, Kelly JF and Lenferink AE: Glycoproteomic analysis of two mouse mammary cell lines during transforming growth factor (TGF)-beta induced epithelial to mesenchymal transition. Proteome Sci 7: 2, 2009.

25. Chaffer CL, Thompson EW and Williams ED: Mesenchymal to epithelial transition in development and disease. Cells Tissues Organs 185: 7-19, 2007.

162

26. Thiery JP: Epithelial-mesenchymal transitions in tumour progression. Nat Rev Cancer 2: 442-54, 2002.

27. Thiery JP: Epithelial-mesenchymal transitions in development and pathologies. Curr Opin Cell Biol 15: 740-6, 2003.

28. Moreno-Bueno G, Portillo F and Cano A: Transcriptional regulation of cell polarity in EMT and cancer. Oncogene 27: 6958-69, 2008.

29. Li QQ, Xu JD, Wang WJ, Cao XX, Chen Q, Tang F, Chen ZQ, Liu XP and Xu ZD: Twist1-mediated adriamycin-induced epithelial-mesenchymal transition relates to multidrug resistance and invasive potential in breast cancer cells. Clin Cancer Res 15: 2657-65, 2009.

30. Chatterjee D, Roy S, Hazra A, Dasgupta P, Ganguly S and Das AK: Variation of adverse drug reaction profile of platinum-based chemotherapy with body mass index in patients with solid tumors: an observational study. Indian J Pharmacol 46: 222-4, 2014.

31. Azad N, Zahnow CA, Rudin CM and Baylin SB: The future of epigenetic therapy in solid tumours--lessons from the past. Nat Rev Clin Oncol 10: 256-66, 2013.

32. Hasanali ZS, Saroya BS, Stuart A, Shimko S, Evans J, Vinod Shah M, Sharma K, Leshchenko VV, Parekh S, Loughran TP, Jr. and Epner EM: Epigenetic therapy overcomes treatment resistance in T cell prolymphocytic leukemia. Sci Transl Med 7: 293ra102, 2015.

33. Vijayaraghavalu S, Dermawan JK, Cheriyath V and Labhasetwar V: Highly synergistic effect of sequential treatment with epigenetic and anticancer drugs to overcome drug resistance in breast cancer cells is mediated via activation of p21 gene expression leading to G2/M cycle arrest. Mol Pharm 10: 337-52, 2013.

34. Cherblanc F, Chapman-Rothe N, Brown R and Fuchter MJ: Current limitations and future opportunities for epigenetic therapies. Future Medicinal Chemistry 4: 425-46, 2012.

35. Jabbour E, Issa JP, Garcia-Manero G and Kantarjian H: Evolution of decitabine development: accomplishments, ongoing investigations, and future strategies. Cancer 112: 2341-51, 2008.

36. Momparler RL, Bouffard DY, Momparler LF, Dionne J, Belanger K and Ayoub J: Pilot phase I-II study on 5-aza-2'-deoxycytidine (Decitabine) in patients with metastatic lung cancer. Anti-cancer drugs 8: 358-68, 1997.

37. Brunner N, Bronzert D, Vindelov LL, Rygaard K, Spang-Thomsen M and Lippman ME: Effect on growth and cell cycle kinetics of estradiol and tamoxifen on MCF-7 human breast cancer cells grown in vitro and in nude mice. Cancer Research 49: 1515-20, 1989.

38. Terry NH, Meistrich ML, Roubein LD, Lynch PM, Dubrow RA and Rich TA: Cellular kinetics in rectal cancer. British Journal of Cancer 72: 435-41, 1995.

163

39. Oloumi A, Maidan M, Lock FE, Tearle H, McKinney S, Muller WJ, Aparicio SA and Dedhar S: Cooperative signaling between Wnt1 and integrin-linked kinase induces accelerated breast tumor development. Breast Cancer Research : BCR 12: R38, 2010.

40. Hollenbach PW, Nguyen AN, Brady H, Williams M, Ning Y, Richard N, Krushel L, Aukerman SL, Heise C and MacBeth KJ: A comparison of azacitidine and decitabine activities in acute myeloid leukemia cell lines. PLoS ONE 5: e9001, 2010.

41. Brown R and Plumb JA: Demethylation of DNA by decitabine in cancer chemotherapy. Expert Review of Anticancer Therapy 4: 501-10, 2004.

42. Cramer SA, Adjei IM and Labhasetwar V: Advancements in the delivery of epigenetic drugs. Expert Opin Drug Deliv 12: 1501-12, 2015.

43. Wang Y and Zhou BP: Epithelial-mesenchymal transition in breast cancer progression and metastasis. Chin J Cancer 30: 603-11, 2011.

44. An Z, Gluck CB, Choy ML and Kaufman LJ: Suberoylanilide hydroxamic acid limits migration and invasion of glioma cells in two and three dimensional culture. Cancer Lett 292: 215-27, 2010.

45. Duong V, Bret C, Altucci L, Mai A, Duraffourd C, Loubersac J, Harmand PO, Bonnet S, Valente S, Maudelonde T, Cavailles V and Boulle N: Specific activity of class II histone deacetylases in human breast cancer cells. Mol Cancer Res 6: 1908-19, 2008.

46. Sarfstein R, Bruchim I, Fishman A and Werner H: The mechanism of action of the histone deacetylase inhibitor vorinostat involves interaction with the insulin-like growth factor signaling pathway. PLoS One 6: e24468, 2011.

47. Siegel D, Hussein M, Belani C, Robert F, Galanis E, Richon VM, Garcia-Vargas J, Sanz- Rodriguez C and Rizvi S: Vorinostat in solid and hematologic malignancies. J Hematol Oncol 2: 31, 2009.

48. Breyer B, Jiang W, Cheng H, Zhou L, Paul R, Feng T and He TC: Adenoviral vector- mediated gene transfer for human gene therapy. Curr Gene Ther 1: 149-62, 2001.

49. Rosenberg SA: Finding suitable targets is the major obstacle to cancer gene therapy. Cancer Gene Ther 21: 45-7, 2014.

50. Hamilton AJ and Baulcombe DC: A species of small antisense RNA in posttranscriptional gene silencing in plants. Science 286: 950-2, 1999.

51. Tseng YH, Hsu HT, Chou YL, Hu CC, Lin NS, Hsu YH and Chang BY: The two conserved cysteine residues of the triple gene block protein 2 are critical for both cell-to- cell and systemic movement of Bamboo mosaic virus. Mol Plant Microbe Interact 22: 1379-88, 2009.

164

52. Filipowicz W, Bhattacharyya SN and Sonenberg N: Mechanisms of post-transcriptional regulation by microRNAs: are the answers in sight? Nat Rev Genet 9: 102-14, 2008.

53. Schiffelers RM, Ansari A, Xu J, Zhou Q, Tang Q, Storm G, Molema G, Lu PY, Scaria PV and Woodle MC: Cancer siRNA therapy by tumor selective delivery with ligand- targeted sterically stabilized nanoparticle. Nucleic Acids Res 32: e149, 2004.

54. Lei XY, Zhong M, Feng LF, Zhu BY, Tang SS and Liao DF: siRNA-mediated Bcl-2 and Bcl-xl gene silencing sensitizes human hepatoblastoma cells to chemotherapeutic drugs. Clin Exp Pharmacol Physiol 34: 450-6, 2007.

55. Bardita C, Predescu D and Predescu S: Long-term silencing of intersectin-1s in mouse lungs by repeated delivery of a specific siRNA via cationic liposomes. Evaluation of knockdown effects by electron microscopy. J Vis Exp: 2013.

56. Layzer JM, McCaffrey AP, Tanner AK, Huang Z, Kay MA and Sullenger BA: In vivo activity of nuclease-resistant siRNAs. RNA 10: 766-71, 2004.

57. Nguyen DN, Mahon KP, Chikh G, Kim P, Chung H, Vicari AP, Love KT, Goldberg M, Chen S, Krieg AM, Chen J, Langer R and Anderson DG: Lipid-derived nanoparticles for immunostimulatory RNA adjuvant delivery. Proc Natl Acad Sci U S A 109: E797-803, 2012.

58. Jackson AL and Linsley PS: Recognizing and avoiding siRNA off-target effects for target identification and therapeutic application. Nat Rev Drug Discov 9: 57-67, 2010.

59. Schroeder A, Heller DA, Winslow MM, Dahlman JE, Pratt GW, Langer R, Jacks T and Anderson DG: Treating metastatic cancer with nanotechnology. Nat. Rev. Cancer 12: 39- 50, 2012.

60. McCrudden CM and McCarthy HO: Current status of gene therapy for breast cancer: progress and challenges. Appl Clin Genet 7: 209-20, 2014.

61. Simoes S, Filipe A, Faneca H, Mano M, Penacho N, Duzgunes N and de Lima MP: Cationic liposomes for gene delivery. Expert Opin Drug Deliv 2: 237-54, 2005.

62. Green JJ, Langer R and Anderson DG: A combinatorial polymer library approach yields insight into nonviral gene delivery. Accounts of Chemical Research 41: 749-759, 2008.

63. Arnold AS, Tang YL, Qian K, Shen L, Valencia V, Phillips MI and Zhang YC: Specific beta1-adrenergic receptor silencing with small interfering RNA lowers high blood pressure and improves cardiac function in myocardial ischemia. J Hypertens 25: 197-205, 2007.

64. Urban-Klein B, Werth S, Abuharbeid S, Czubayko F and Aigner A: RNAi-mediated gene-targeting through systemic application of polyethylenimine (PEI)-complexed siRNA in vivo. Gene Therapy 12: 461-466, 2005.

165

65. Fischer D, Bieber T, Li YX, Elsasser HP and Kissel T: A novel non-viral vector for DNA delivery based on low molecular weight, branched polyethylenimine: Effect of molecular weight on transfection efficiency and cytotoxicity. Pharmaceutical Research 16: 1273- 1279, 1999.

66. Wilczewska AZ, Niemirowicz K, Markiewicz KH and Car H: Nanoparticles as drug delivery systems. Pharmacological Reports 64: 1020-1037, 2012.

67. Pack DW, Hoffman AS, Pun S and Stayton PS: Design and development of polymers for gene delivery. Nature Reviews Drug Discovery 4: 581-593, 2005.

68. Morin R, Bainbridge M, Fejes A, Hirst M, Krzywinski M, Pugh T, McDonald H, Varhol R, Jones S and Marra M: Profiling the HeLa S3 transcriptome using randomly primed cDNA and massively parallel short-read sequencing. Biotechniques 45: 81-94, 2008.

69. Maher CA, Kumar-Sinha C, Cao X, Kalyana-Sundaram S, Han B, Jing X, Sam L, Barrette T, Palanisamy N and Chinnaiyan AM: Transcriptome sequencing to detect gene fusions in cancer. Nature 458: 97-101, 2009.

70. Wang Z, Gerstein M and Snyder M: RNA-Seq: a revolutionary tool for transcriptomics. Nat Rev Genet 10: 57-63, 2009.

71. Kapoor-Vazirani P, Kagey JD, Powell DR and Vertino PM: Role of hMOF-dependent histone H4 lysine 16 acetylation in the maintenance of TMS1/ASC gene activity. Cancer Res 68: 6810-21, 2008.

72. Eswaran J, Cyanam D, Mudvari P, Reddy SD, Pakala SB, Nair SS, Florea L, Fuqua SA, Godbole S and Kumar R: Transcriptomic landscape of breast cancers through mRNA sequencing. Sci Rep 2: 264, 2012.

73. Dick JE: Stem cells: Self-renewal writ in blood. Nature 423: 231-3, 2003.

74. Li W, Liu F, Lei T, Xu X, Liu B, Cui L, Wei J, Guo X, Lang R, Fan Y, Gu F, Tang P, Zhang X and Fu L: The clinicopathological significance of CD44+/CD24-/low and CD24+ tumor cells in invasive micropapillary carcinoma of the breast. Pathol Res Pract 206: 828-34, 2010.

75. Griffin JD and Lowenberg B: Clonogenic cells in acute myeloblastic leukemia. Blood 68: 1185-95, 1986.

76. McCulloch EA: Stem cells in normal and leukemic hemopoiesis (Henry Stratton Lecture, 1982). Blood 62: 1-13, 1983.

77. Charafe-Jauffret E, Ginestier C and Birnbaum D: Breast cancer stem cells: tools and models to rely on. BMC Cancer 9: 202, 2009.

78. Chuthapisith S, Eremin J, El-Sheemey M and Eremin O: Breast cancer chemoresistance: emerging importance of cancer stem cells. Surg Oncol 19: 27-32, 2010.

166

79. Dontu G: Breast cancer stem cell markers - the rocky road to clinical applications. Breast Cancer Res 10: 110, 2008.

80. Kakarala M and Wicha MS: Implications of the cancer stem-cell hypothesis for breast cancer prevention and therapy. J Clin Oncol 26: 2813-20, 2008.

81. Li X, Lewis MT, Huang J, Gutierrez C, Osborne CK, Wu MF, Hilsenbeck SG, Pavlick A, Zhang X, Chamness GC, Wong H, Rosen J and Chang JC: Intrinsic resistance of tumorigenic breast cancer cells to chemotherapy. J Natl Cancer Inst 100: 672-9, 2008.

82. Morrison BJ, Schmidt CW, Lakhani SR, Reynolds BA and Lopez JA: Breast cancer stem cells: implications for therapy of breast cancer. Breast Cancer Res 10: 210, 2008.

83. Phillips TM, McBride WH and Pajonk F: The response of CD24(-/low)/CD44+ breast cancer-initiating cells to radiation. J Natl Cancer Inst 98: 1777-85, 2006.

84. Abraham BK, Fritz P, McClellan M, Hauptvogel P, Athelogou M and Brauch H: Prevalence of CD44+/CD24-/low cells in breast cancer may not be associated with clinical outcome but may favor distant metastasis. Clin Cancer Res 11: 1154-9, 2005.

85. Al-Hajj M, Wicha MS, Benito-Hernandez A, Morrison SJ and Clarke MF: Prospective identification of tumorigenic breast cancer cells. Proc Natl Acad Sci U S A 100: 3983-8, 2003.

86. Bauerschmitz GJ, Ranki T, Kangasniemi L, Ribacka C, Eriksson M, Porten M, Herrmann I, Ristimaki A, Virkkunen P, Tarkkanen M, Hakkarainen T, Kanerva A, Rein D, Pesonen S and Hemminki A: Tissue-specific promoters active in CD44+CD24-/low breast cancer cells. Cancer Res 68: 5533-9, 2008.

87. Hill RP and Perris R: "Destemming" cancer stem cells. J Natl Cancer Inst 99: 1435-40, 2007.

88. Kelly PN, Dakic A, Adams JM, Nutt SL and Strasser A: Tumor growth need not be driven by rare cancer stem cells. Science 317: 337, 2007.

89. Snyder EL, Bailey D, Shipitsin M, Polyak K and Loda M: Identification of CD44v6(+)/CD24- breast carcinoma cells in primary human tumors by quantum dot- conjugated antibodies. Lab Invest 89: 857-66, 2009.

90.Honeth G, Bendahl PO, Ringner M, Saal LH, Gruvberger-Saal SK, Lovgren K, Grabau D, Ferno M, Borg A and Hegardt C: The CD44+/CD24- phenotype is enriched in basal-like breast tumors. Breast Cancer Res 10: R53, 2008.

91. Mylona E, Giannopoulou I, Fasomytakis E, Nomikos A, Magkou C, Bakarakos P and Nakopoulou L: The clinicopathologic and prognostic significance of CD44+/CD24(-/low) and CD44-/CD24+ tumor cells in invasive breast carcinomas. Hum Pathol 39: 1096-102, 2008.

167

92. Osta WA, Chen Y, Mikhitarian K, Mitas M, Salem M, Hannun YA, Cole DJ and Gillanders WE: EpCAM is overexpressed in breast cancer and is a potential target for breast cancer gene therapy. Cancer Res 64: 5818-24, 2004.

93. Blick T, Widodo E, Hugo H, Waltham M, Lenburg ME, Neve RM and Thompson EW: Epithelial mesenchymal transition traits in human breast cancer cell lines. Clin Exp Metastasis 25: 629-42, 2008.

94. Iseri OD, Kars MD, Arpaci F, Atalay C, Pak I and Gunduz U: Drug resistant MCF-7 cells exhibit epithelial-mesenchymal transition gene expression pattern. Biomed Pharmacother 65: 40-5, 2011.

95. Arumugam T, Ramachandran V, Fournier KF, Wang H, Marquis L, Abbruzzese JL, Gallick GE, Logsdon CD, McConkey DJ and Choi W: Epithelial to mesenchymal transition contributes to drug resistance in pancreatic cancer. Cancer Res 69: 5820-8, 2009.

96. Sheridan C, Kishimoto H, Fuchs RK, Mehrotra S, Bhat-Nakshatri P, Turner CH, Goulet R, Jr., Badve S and Nakshatri H: CD44+/CD24- breast cancer cells exhibit enhanced invasive properties: an early step necessary for metastasis. Breast Cancer Res 8: R59, 2006.

97. Fillmore CM and Kuperwasser C: Human breast cancer cell lines contain stem-like cells that self-renew, give rise to phenotypically diverse progeny and survive chemotherapy. Breast Cancer Res 10: R25, 2008.

98. Bass AJ, Watanabe H, Mermel CH, Yu S, Perner S, Verhaak RG, Kim SY, Wardwell L, Tamayo P, Gat-Viks I, Ramos AH, Woo MS, Weir BA, Getz G, Beroukhim R, O'Kelly M, Dutt A, Rozenblatt-Rosen O, Dziunycz P, Komisarof J, Chirieac LR, Lafargue CJ, Scheble V, Wilbertz T, Ma C, Rao S, Nakagawa H, Stairs DB, Lin L, Giordano TJ, Wagner P, Minna JD, Gazdar AF, Zhu CQ, Brose MS, Cecconello I, Jr UR, Marie SK, Dahl O, Shivdasani RA, Tsao MS, Rubin MA, Wong KK, Regev A, Hahn WC, Beer DG, Rustgi AK and Meyerson M: SOX2 is an amplified lineage-survival oncogene in lung and esophageal squamous cell carcinomas. Nat Genet 41: 1238-42, 2009.

99. Takahashi K, Tanabe K, Ohnuki M, Narita M, Ichisaka T, Tomoda K and Yamanaka S: Induction of pluripotent stem cells from adult human fibroblasts by defined factors. Cell 131: 861-72, 2007.

100.Takahashi K and Yamanaka S: Induction of pluripotent stem cells from mouse embryonic and adult fibroblast cultures by defined factors. Cell 126: 663-76, 2006.

101.Guttilla IK, Phoenix KN, Hong X, Tirnauer JS, Claffey KP and White BA: Prolonged mammosphere culture of MCF-7 cells induces an EMT and repression of the estrogen receptor by microRNAs. Breast Cancer Res Treat 132: 75-85, 2012.

168

102.Lu S and Labhasetwar V: Drug Resistant Breast Cancer Cell Line Displays Cancer Stem Cell Phenotype and Responds Sensitively to Epigenetic Drug SAHA. Drug Deliv Transl Res 3: 183-94, 2013.

103.Vojta A and Zoldos V: Adaptation or malignant transformation: the two faces of epigenetically mediated response to stress. Biomed Res Int 2013: 954060, 2013.

104.Daskalakis M, Blagitko-Dorfs N and Hackanson B: Decitabine. Recent Results Cancer Res 184: 131-57, 2010.

105.Sato A: Vorinostat approved in Japan for treatment of cutaneous T-cell lymphomas: status and prospects. Onco Targets Ther 5: 67-76, 2012.

106.Zoncu R, Efeyan A and Sabatini DM: mTOR: from growth signal integration to cancer, diabetes and ageing. Nat Rev Mol Cell Biol 12: 21-35, 2011.

107.Martinez R: Beyond Genetics in Glioma Pathways: The Ever-Increasing Crosstalk between Epigenomic and Genomic Events. J Signal Transduct 2012: 519807, 2012.

108.Feinberg AP and Tycko B: The history of cancer epigenetics. Nat Rev Cancer 4: 143-53, 2004.

109.Kantarjian H, Issa JP, Rosenfeld CS, Bennett JM, Albitar M, DiPersio J, Klimek V, Slack J, de Castro C, Ravandi F, Helmer R, 3rd, Shen L, Nimer SD, Leavitt R, Raza A and Saba H: Decitabine improves patient outcomes in myelodysplastic syndromes: results of a phase III randomized study. Cancer 106: 1794-803, 2006.

110.Kantarjian HM, O'Brien S, Cortes J, Giles FJ, Faderl S, Issa JP, Garcia-Manero G, Rios MB, Shan J, Andreeff M, Keating M and Talpaz M: Results of decitabine (5-aza- 2'deoxycytidine) therapy in 130 patients with chronic myelogenous leukemia. Cancer 98: 522-8, 2003.

111.Richon VM: Cancer biology: mechanism of antitumour action of vorinostat (suberoylanilide hydroxamic acid), a novel histone deacetylase inhibitor. British Journal of Cancer 95: S2-S6, 2006.

112.Luqmani YA: Mechanisms of drug resistance in cancer chemotherapy. Med Princ Pract 14 Suppl 1: 35-48, 2005.

113.Kerbel RS: Molecular and physiologic mechanisms of drug resistance in cancer: an overview. Cancer Metastasis Rev 20: 1-2, 2001.

114.Vijayaraghavalu S, Peetla C, Lu S and Labhasetwar V: Epigenetic modulation of the biophysical properties of drug-resistant cell lipids to restore drug transport and endocytic functions. Mol Pharm 9: 2730-42, 2012.

115.Hirata H, Hinoda Y, Nakajima K, Kawamoto K, Kikuno N, Ueno K, Yamamura S, Zaman MS, Khatri G, Chen Y, Saini S, Majid S, Deng G, Ishii N and Dahiya R: Wnt

169

antagonist DKK1 acts as a tumor suppressor gene that induces apoptosis and inhibits proliferation in human renal cell carcinoma. Int J Cancer 128: 1793-803, 2011.

116.Plumb JA, Strathdee G, Sludden J, Kaye SB and Brown R: Reversal of drug resistance in human tumor xenografts by 2'-deoxy-5-azacytidine-induced demethylation of the hMLH1 gene promoter. Cancer Res 60: 6039-44, 2000.

117.Murray-Stewart T, Hanigan CL, Woster PM, Marton LJ and Casero RA, Jr.: Histone deacetylase inhibition overcomes drug resistance through a miRNA-dependent mechanism. Mol Cancer Ther 12: 2088-99, 2013.

118.Gupta M, Ansell SM, Novak AJ, Kumar S, Kaufmann SH and Witzig TE: Inhibition of histone deacetylase overcomes rapamycin-mediated resistance in diffuse large B-cell lymphoma by inhibiting Akt signaling through mTORC2. Blood 114: 2926-35, 2009.

119.Peetla C, Vijayaraghavalu S and Labhasetwar V: Biophysics of cell membrane lipids in cancer drug resistance: Implications for drug transport and drug delivery with nanoparticles. Adv Drug Deliv Rev 65: 1686-98, 2013.

120.Peetla C, Bhave R, Vijayaraghavalu S, Stine A, Kooijman E and Labhasetwar V: Drug resistance in breast cancer cells: biophysical characterization of and doxorubicin interactions with membrane lipids. Mol Pharm 7: 2334-48, 2010.

121.Luong A, Hannah VC, Brown MS and Goldstein JL: Molecular characterization of human acetyl-CoA synthetase, an enzyme regulated by sterol regulatory element-binding proteins. J Biol Chem 275: 26458-66, 2000.

122.Laplante M and Sabatini DM: An emerging role of mTOR in lipid biosynthesis. Curr Biol 19: R1046-52, 2009.

123.Liu H, Wu X, Dong Z, Luo Z, Zhao Z, Xu Y and Zhang JT: Fatty acid synthase causes drug resistance by inhibiting TNF-alpha and ceramide production. J Lipid Res 54: 776-85, 2013.

124.Alo PL, Visca P, Trombetta G, Mangoni A, Lenti L, Monaco S, Botti C, Serpieri DE and Di Tondo U: Fatty acid synthase (FAS) predictive strength in poorly differentiated early breast carcinomas. Tumori 85: 35-40, 1999.

125.Nagler DK, Sulea T and Menard R: Full-length cDNA of human cathepsin F predicts the presence of a cystatin domain at the N-terminus of the cysteine protease zymogen. Biochem Biophys Res Commun 257: 313-8, 1999.

126.Settembre C, Fraldi A, Medina DL and Ballabio A: Signals from the lysosome: a control centre for cellular clearance and energy metabolism. Nature Reviews Molecular Cell Biology 14: 283-296, 2013.

127.Park JS, Han TH, Lee KY, Han SS, Hwang JJ, Moon DH, Kim SY and Cho YW: N- acetyl histidine-conjugated glycol chitosan self-assembled nanoparticles for

170

intracytoplasmic delivery of drugs: endocytosis, exocytosis and drug release. J Control Release 115: 37-45, 2006.

128.GM. C: The Nuclear Envelope and Traffic between the Nucleus and Cytoplasm. In: The Cell: A Molecular Approach. 2nd edition. Sunderland (MA): Sinauer Associates, 2000.

129.Kinoshita Y, Kalir T, Rahaman J, Dottino P and Kohtz DS: Alterations in nuclear pore architecture allow cancer cell entry into or exit from drug-resistant dormancy. Am J Pathol 180: 375-89, 2012.

130.Lewin JM, Lwaleed BA, Cooper AJ and Birch BR: The direct effect of nuclear pores on nuclear chemotherapeutic concentration in multidrug resistant bladder cancer: the nuclear sparing phenomenon. J Urol 177: 1526-30, 2007.

131.Sazer S and Dasso M: The ran decathlon: multiple roles of Ran. J Cell Sci 113 ( Pt 7): 1111-8, 2000.

132.Bischoff FR and Ponstingl H: Catalysis of guanine nucleotide exchange on Ran by the mitotic regulator RCC1. Nature 354: 80-2, 1991.

133.Cupp MJ and Tracy TS: Cytochrome P450: new nomenclature and clinical implications. Am Fam Physician 57: 107-16, 1998.

134.Lai Peng XZ: Epigenetic regulation of drug metabolism and transport. Acta Pharmaceutica Sinica B 5(2):106-112: 2015.

135.AbuHammad S and Zihlif M: Gene expression alterations in doxorubicin resistant MCF7 breast cancer cell line. Genomics 101: 213-20, 2013.

136.Dias V and Ribeiro V: Ethnic differences in the prevalence of polymorphisms in CYP7A1, CYP7B1 AND CYP27A1 enzymes involved in cholesterol metabolism. J Pharm Bioallied Sci 3: 453-9, 2011.

137.Hardwick JP and Chiang JY: PPARs, RXRs, and Drug-Metabolizing Enzymes. PPAR Res 2009: 589626, 2009.

138.Foster H, Coley HM, Goumenou A, Pados G, Harvey A and Karteris E: Differential expression of mTOR signalling components in drug resistance in ovarian cancer. Anticancer Res 30: 3529-34, 2010.

139.Kurundkar D, Srivastava RK, Chaudhary SC, Ballestas ME, Kopelovich L, Elmets CA and Athar M: Vorinostat, an HDAC inhibitor attenuates epidermoid squamous cell carcinoma growth by dampening mTOR signaling pathway in a human xenograft murine model. Toxicol Appl Pharmacol 266: 233-44, 2013.

140.Jang GB, Kim JY, Cho SD, Park KS, Jung JY, Lee HY, Hong IS and Nam JS: Blockade of Wnt/beta-catenin signaling suppresses breast cancer metastasis by inhibiting CSC-like phenotype. Sci Rep 5: 12465, 2015.

171

141.Shen DY, Zhang W, Zeng X and Liu CQ: Inhibition of Wnt/beta-catenin signaling downregulates P-glycoprotein and reverses multi-drug resistance of cholangiocarcinoma. Cancer Sci 104: 1303-8, 2013.

142.Brunen D, Willems SM, Kellner U, Midgley R, Simon I and Bernards R: TGF-beta: an emerging player in drug resistance. Cell Cycle 12: 2960-8, 2013.

143.Oshimori N, Oristian D and Fuchs E: TGF-beta promotes heterogeneity and drug resistance in squamous cell carcinoma. Cell 160: 963-76, 2015.

144.Yu D and Hung MC: Overexpression of ErbB2 in cancer and ErbB2-targeting strategies. Oncogene 19: 6115-21, 2000.

145.Murakami H, Ito S, Tanaka H, Kondo E, Kodera Y and Nakanishi H: Establishment of new intraperitoneal paclitaxel-resistant gastric cancer cell lines and comprehensive gene expression analysis. Anticancer Res 33: 4299-307, 2013.

146.Higgins MJ and Stearns V: Understanding resistance to tamoxifen in hormone receptor- positive breast cancer. Clin Chem 55: 1453-5, 2009.

147.Dykxhoorn DM, Novina CD and Sharp PA: Killing the messenger: short RNAs that silence gene expression. Nat Rev Mol Cell Biol 4: 457-67, 2003.

148.Lan H, Chen H, Liu Y, Jiang C, Mao Q, Jia D, Chen Q and Wei T: Small interfering RNA pathway modulates initial viral infection in the midgut epithelium of insect after ingestion of virus. J Virol: 2015.

149.Burnett JC, Rossi JJ and Tiemann K: Current progress of siRNA/shRNA therapeutics in clinical trials. Biotechnol J 6: 1130-46, 2011.

150.Zi Y, Zhao W, Zhou J, He H and Xie M: Silencing of TMSG1 enhances metastasis capacity by targeting V-ATPase in breast cancer. Int J Clin Exp Pathol 8: 1312-20, 2015.

151.Yang X, Iyer AK, Singh A, Choy E, Hornicek FJ, Amiji MM and Duan Z: MDR1 siRNA loaded hyaluronic acid-based CD44 targeted nanoparticle systems circumvent paclitaxel resistance in ovarian cancer. Sci Rep 5: 8509, 2015.

152.Scamuffa N, Siegfried G, Bontemps Y, Ma L, Basak A, Cherel G, Calvo F, Seidah NG and Khatib AM: Selective inhibition of proprotein convertases represses the metastatic potential of human colorectal tumor cells. J Clin Invest 118: 352-63, 2008.

153.Shegokar R, Al Shaal L and Mishra PR: SiRNA delivery: challenges and role of carrier systems. Pharmazie 66: 313-8, 2011.

154.Whitehead KA, Langer R and Anderson DG: Knocking down barriers: advances in siRNA delivery. Nat Rev Drug Discov 8: 129-38, 2009.

172

155.de Fougerolles A, Vornlocher HP, Maraganore J and Lieberman J: Interfering with disease: a progress report on siRNA-based therapeutics. Nat Rev Drug Discov 6: 443-53, 2007.

156.Jackson AL, Bartz SR, Schelter J, Kobayashi SV, Burchard J, Mao M, Li B, Cavet G and Linsley PS: Expression profiling reveals off-target gene regulation by RNAi. Nat Biotechnol 21: 635-7, 2003.

157.Lee JM, Yoon TJ and Cho YS: Recent developments in nanoparticle-based siRNA delivery for cancer therapy. Biomed Res Int 2013: 782041, 2013.

158.Shen H, Sun T and Ferrari M: Nanovector delivery of siRNA for cancer therapy. Cancer Gene Ther 19: 367-73, 2012.

159.Kawakami S, Ito Y, Charoensit P, Yamashita F and Hashida M: Evaluation of proinflammatory cytokine production induced by linear and branched polyethylenimine/plasmid DNA complexes in mice. J Pharmacol Exp Ther 317: 1382-90, 2006.

160.Hobel S and Aigner A: Polyethylenimine (PEI)/siRNA-mediated gene knockdown in vitro and in vivo. Methods Mol Biol 623: 283-97, 2010.

161.Yuan X, Naguib S and Wu Z: Recent advances of siRNA delivery by nanoparticles. Expert Opin Drug Deliv 8: 521-36, 2011.

162.Wang Y, Li Z, Han Y, Liang LH and Ji A: Nanoparticle-based delivery system for application of siRNA in vivo. Curr Drug Metab 11: 182-96, 2010.

163.Wu ZW, Chien CT, Liu CY, Yan JY and Lin SY: Recent progress in copolymer- mediated siRNA delivery. J Drug Target 20: 551-60, 2012.

164.Merkel OM, Beyerle A, Beckmann BM, Zheng M, Hartmann RK, Stoger T and Kissel TH: Polymer-related off-target effects in non-viral siRNA delivery. Biomaterials 32: 2388-98, 2011.

165.Alameh M, Dejesus D, Jean M, Darras V, Thibault M, Lavertu M, Buschmann MD and Merzouki A: Low molecular weight chitosan nanoparticulate system at low N:P ratio for nontoxic polynucleotide delivery. Int J Nanomedicine 7: 1399-414, 2012.

166.Moghimi SM, Symonds P, Murray JC, Hunter AC, Debska G and Szewczyk A: A two- stage poly(ethylenimine)-mediated cytotoxicity: implications for gene transfer/therapy. Mol Ther 11: 990-5, 2005.

167.Hunter AC and Moghimi SM: Cationic carriers of genetic material and cell death: a mitochondrial tale. Biochim Biophys Acta 1797: 1203-9, 2010.

173

168.Richard JP, Melikov K, Vives E, Ramos C, Verbeure B, Gait MJ, Chernomordik LV and Lebleu B: Cell-penetrating peptides. A reevaluation of the mechanism of cellular uptake. J Biol Chem 278: 585-90, 2003.

169.Calnan BJ, Tidor B, Biancalana S, Hudson D and Frankel AD: Arginine-mediated RNA recognition: the arginine fork. Science 252: 1167-71, 1991.

170.Wang S, Amornwittawat N, Juwita V, Kao Y, Duman JG, Pascal TA, Goddard WA and Wen X: Arginine, a key residue for the enhancing ability of an antifreeze protein of the beetle Dendroides canadensis. Biochemistry 48: 9696-703, 2009.

171.Choi JS, Nam K, Park J, Kim JB, Lee JK and Park J: Enhanced transfection efficiency of PAMAM dendrimer by surface modification with L-arginine. Journal of Controlled Release 99: 445-456, 2004.

172.Kim TI, Baek JU, Bai CZ and Park JS: Arginine-conjugated polypropylenimine dendrimer as a non-toxic and efficient gene delivery carrier. Biomaterials 28: 2061-2067, 2007.

173.Lu S, Morris VB and Labhasetwar V: Codelivery of DNA and siRNA via arginine-rich PEI-based polyplexes. Mol Pharm 12: 621-9, 2015.

174.Liu Y, Steele T and Kissel T: Degradation of Hyper-Branched Poly(ethylenimine)-graft- poly(caprolactone)-block-monomethoxyl-poly(ethylene glycol) as a Potential Gene Delivery Vector. Macromol Rapid Commun 31: 1509-15, 2010.

175.Morris VB and Sharma CP: Folate mediated l-arginine modified oligo (alkylaminosiloxane) graft poly (ethyleneimine) for tumor targeted gene delivery. Biomaterials 32: 3030-41, 2011.

176.Kim TI, Ou M, Lee M and Kim SW: Arginine-grafted bioreducible poly(disulfide amine) for gene delivery systems. Biomaterials 30: 658-64, 2009.

177.Susa M, Iyer AK, Ryu K, Choy E, Hornicek FJ, Mankin H, Milane L, Amiji MM and Duan Z: Inhibition of ABCB1 (MDR1) expression by an siRNA nanoparticulate delivery system to overcome drug resistance in osteosarcoma. PLoS One 5: e10764, 2010.

178.Wolfram JA and Donahue JK: Gene therapy to treat cardiovascular disease. J Am Heart Assoc 2: e000119, 2013.

179.van der Marel S, Majowicz A, van Deventer S, Petry H, Hommes DW and Ferreira V: Gene and cell therapy based treatment strategies for inflammatory bowel diseases. World J Gastrointest Pathophysiol 2: 114-22, 2011.

180.Yacyshyn BR, Bowen-Yacyshyn MB, Jewell L, Tami JA, Bennett CF, Kisner DL and Shanahan WR, Jr.: A placebo-controlled trial of ICAM-1 antisense oligonucleotide in the treatment of Crohn's disease. Gastroenterology 114: 1133-42, 1998.

174

181.Merdan T, Kopecek J and Kissel T: Prospects for cationic polymers in gene and oligonucleotide therapy against cancer. Adv Drug Deliv Rev 54: 715-58, 2002.

182.Chen SH, Chen XH, Wang Y, Kosai K, Finegold MJ, Rich SS and Woo SL: Combination gene therapy for liver metastasis of colon carcinoma in vivo. Proc Natl Acad Sci U S A 92: 2577-81, 1995.

183.Deleavey GF, Watts JK, Alain T, Robert F, Kalota A, Aishwarya V, Pelletier J, Gewirtz AM, Sonenberg N and Damha MJ: Synergistic effects between analogs of DNA and RNA improve the potency of siRNA-mediated gene silencing. Nucleic Acids Res 38: 4547-57, 2010.

184.Read ML, Singh S, Ahmed Z, Stevenson M, Briggs SS, Oupicky D, Barrett LB, Spice R, Kendall M, Berry M, Preece JA, Logan A and Seymour LW: A versatile reducible polycation-based system for efficient delivery of a broad range of nucleic acids. Nucleic Acids Res 33: e86, 2005.

185.Tripathi SK, Goyal R, Kumar P and Gupta KC: Linear polyethylenimine-graft-chitosan copolymers as efficient DNA/siRNA delivery vectors in vitro and in vivo. Nanomedicine-Nanotechnology Biology and Medicine 8: 337-345, 2012.

186.Lungwitz U, Breunig M, Blunk T and Gopferich A: Polyethylenimine-based non-viral gene delivery systems. Eur J Pharm Biopharm 60: 247-66, 2005.

187.LabatMoleur F, Steffan AM, Brisson C, Perron H, Feugeas O, Furstenberger P, Oberling F, Brambilla E and Behr JP: An electron microscopy study into the mechanism of gene transfer with lipopolyamines. Gene Ther 3: 1010-1017, 1996.

188.Godbey WT, Wu KK and Mikos AG: Tracking the intracellular path of poly(ethylenimine)/DNA complexes for gene delivery. Proc Natl Acad Sci U S A 96: 5177-81, 1999.

189.Tiera MJ, Shi Q, Winnik FM and Fernandes JC: Polycation-based gene therapy: current knowledge and new perspectives. Curr Gene Ther 11: 288-306, 2011.

190.Rekha MR and Sharma CP: Polymers for gene delivery: current status and future perspectives. Recent Pat DNA Gene Seq 6: 98-107, 2012.

191.Fischer D, Bieber T, Li Y, Elsasser HP and Kissel T: A novel non-viral vector for DNA delivery based on low molecular weight, branched polyethylenimine: effect of molecular weight on transfection efficiency and cytotoxicity. Pharm Res 16: 1273-9, 1999.

192.Peng YS, Lai PL, Peng S, Wu HC, Yu S, Tseng TY, Wang LF and Chu IM: Glial cell line-derived neurotrophic factor gene delivery via a polyethylene imine grafted chitosan carrier. Int J Nanomedicine 9: 3163-74, 2014.

175

193.Morris VB and Sharma CP: Enhanced in-vitro transfection and biocompatibility of L- arginine modified oligo (-alkylaminosiloxanes)-graft-polyethylenimine. Biomaterials 31: 8759-69, 2010.

194.Rosenholm JM, Meinander A, Peuhu E, Niemi R, Eriksson JE, Sahlgren C and Linden M: Targeting of porous hybrid silica nanoparticles to cancer cells. ACS Nano 3: 197-206, 2009.

195.Kwok A: A Comparison of the Properties of Plasmid DNA and siRNA. In: Volker A. Erdmann JB (ed) DNA and RNA Nanobiotechnologies in Medicine: Diagnosis and Treatment of Diseases. Heidelberg : Springer, Berlin, Germany, 2013, pp 215-216.

196.Wong LS: Elucidation of Design Criteria for siRNA Delivery in Mammalian Cells Using Polyethylenimine. In: Chemical Engineering. University of Illinois at Urbana-Champaign, Urbana-Champaign, 2009, p 209.

197.Kichler A, Sabourault N, Decor R, Leborgne C, Schmutz M, Valleix A, Danos O, Wagner A and Mioskowski C: Preparation and evaluation of a new class of gene transfer reagents: poly(-alkylaminosiloxanes). J Control Release 93: 403-14, 2003.

198.Ruponen M, Ronkko S, Honkakoski P, Pelkonen J, Tammi M and Urtti A: Extracellular glycosaminoglycans modify cellular trafficking of lipoplexes and polyplexes. J Biol Chem 276: 33875-80, 2001.

199.Kang HC, Samsonova O and Bae YH: Trafficking microenvironmental pHs of polycationic gene vectors in drug-sensitive and multidrug-resistant MCF7 breast cancer cells. Biomaterials 31: 3071-8, 2010.

200.Kim WJ and Kim SW: Efficient siRNA delivery with non-viral polymeric vehicles. Pharm Res 26: 657-66, 2009.

201.Zheng M, Pavan GM, Neeb M, Schaper AK, Danani A, Klebe G, Merkel OM and Kissel T: Targeting the blind spot of polycationic nanocarrier-based siRNA delivery. ACS Nano 6: 9447-54, 2012.

202.Wang Y, Gao S, Ye WH, Yoon HS and Yang YY: Co-delivery of drugs and DNA from cationic core-shell nanoparticles self-assembled from a biodegradable copolymer. Nat Mater 5: 791-6, 2006.

203.Gaspar VM, Goncalves C, de Melo-Diogo D, Costa EC, Queiroz JA, Pichon C, Sousa F and Correia IJ: Poly(2-ethyl-2-oxazoline)-PLA-g-PEI amphiphilic triblock micelles for co-delivery of minicircle DNA and chemotherapeutics. J Control Release 189: 90-104, 2014.

204.Wang C, Ravi S, Garapati US, Das M, Howell M, MallelaMallela J, Alwarapan S, Mohapatra SS and Mohapatra S: Multifunctional Chitosan Magnetic-Graphene (CMG) Nanoparticles: a Theranostic Platform for Tumor-targeted Co-delivery of Drugs, Genes and MRI Contrast Agents. J Mater Chem B Mater Biol Med 1: 4396-4405, 2013.

176

205.Skandrani N, Barras A, Legrand D, Gharbi T, Boulahdour H and Boukherroub R: Lipid nanocapsules functionalized with polyethyleneimine for plasmid DNA and drug co- delivery and cell imaging. Nanoscale 6: 7379-90, 2014.

206.Ramachandran S, Ient J, Gottgens EL, Krieg AJ and Hammond EM: Epigenetic Therapy for Solid Tumors: Highlighting the Impact of Tumor Hypoxia. Genes (Basel) 6: 935-56, 2015.

207.Pecot CV, Calin GA, Coleman RL, Lopez-Berestein G and Sood AK: RNA interference in the clinic: challenges and future directions. Nat Rev Cancer 11: 59-67, 2011.

208.Knight M, Miller A, Liu Y, Scaria P, Woodle M and Ittiprasert W: Polyethyleneimine (PEI) mediated siRNA gene silencing in the Schistosoma mansoni snail host, Biomphalaria glabrata. PLoS Negl Trop Dis 5: e1212, 2011.

209.Takei Y, Kadomatsu K, Yuzawa Y, Matsuo S and Muramatsu T: A small interfering RNA targeting vascular endothelial growth factor as cancer therapeutics. Cancer Res 64: 3365-70, 2004.

210.Filleur S, Courtin A, Ait-Si-Ali S, Guglielmi J, Merle C, Harel-Bellan A, Clezardin P and Cabon F: SiRNA-mediated inhibition of vascular endothelial growth factor severely limits tumor resistance to antiangiogenic thrombospondin-1 and slows tumor vascularization and growth. Cancer Res 63: 3919-22, 2003.

177

APPENDICES

178

APPENDIX A

LIST OF ABBREVIATIONS

ABCB1 – multidrug resistance protein 1

Ab – Antibody

Abs – absorbance

BCSC – breast cancer stem cell bPEI – branched polyethylenimine

BRCA1 – breast cancer gene 1

BRCA2 – breast cancer gene 2

BSA – bovine serum albumin

CSC – cancer stem cell

DAC – decitabine

DCC – dicyclohexylcarbodiimide

Dh – hydrodynamic diameters

DLS – dynamic light scattering

DMEM – Dulbecco’s modified eagle’s medium

DOTAP – dioleyltrimethylammoniumpropane

DOX (Dox) – doxorubicin

179

Doxil – trade name of a liposome form of doxorubicin

DPBS – Dulbecco’s phosphate-buffered saline

D/S/P ratio – DNA/siRNA/polymer weight/weight ratio

EDC – 1-ethyl-3-(3-dimethylaminopropyl)carbodiimide hydrochloride

EMT – epithelial-mesenchymal transition

ER- estrogen receptor

ESA – epithelial-specific antigen

FASN – fatty acid synthase

HDAC – histone deacetylase hMSC – human mesenchymal stem cell

HPLC – high-performance liquid chromatography

HNRNPU – heterogeneous nuclear ribonucleoprotin U (scaffold attachment factor A)

IV – intravenous luc-DNA – luciferase-DNA luc-siRNA – anti-luciferase siRNA mTOR – mammalian target of rapamycin

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

MWCO – molecular weight cutoff

N/P ratio – siRNA/polymer weight/weight ratio;

NPC – nuclear pore complex

NUP35 – nucleoporin 35kDa

NUP153 – nucleoporin 153kDa

NUPR1 – nuclear protein, transcriptional regulator, 1

180

PAMAM – poly(amidoamine)

PDI – Polydispersity Index

PEG – poly(ethylene glycol)

Pgp – P-glycoprotein

PLGA – poly(lactide co-glycolic acid)

P/N – polymer/nucleotide weight/weight ratio

PPARs – peroxisome proliferator-activated receptors

P(SiDAAr)5–5 siloxane-arginine modified bPEI polymer

P(SiDAAr)5Peg3 – 3 PEG-conjugated 5 siloxane-arginine modified bPEI polymer

Ø-pDNA – noncoding plasmid DNA index

RAN – RAs-related Nuclear protein

RCC1 – regulator of chromosome condensation 1

RISCs – RNA-induced silencing complexes

RNAi – RNA interference

SAHA – suberanilohydroxamic acid siRNA – small interfering RNA s-siRNA – scramble siRNA

SNAI1 – zinc finger protein snail 1

TEM – transmission electron microscopy

TGFbeta – transforming growth factor beta

TFEB: Transcription Factor EB

TNPO1: transportin 1

V/v – volume/volume

181

VIM – vimentin

182

APPENDIX B

THE REDUCTION OF LUCIFERASE PROTEIN EXPRESSION IN VIVO

Introduction:

Although siRNA has shown significant therapeutic promises and to effectively transfect cell lines in vitro, its clinical applications are still in their infancy due to the challenges presented in the in vivo delivery of siRNA. Since naked siRNA is too large and hydrophilic to diffuse across the cell membrane, the poor cellular uptake of siRNA in vivo remains a significant barrier for the advancement of the technology [154, 155] Some other limitations include poor pharmacokinetic property, siRNA nucleases degradation after systemic administration, the triggering of the innate immune system in vivo, and a lack of effective systemic delivery methods that are safe and nontoxic [56-58, 207].

P(SiDAAr)5Peg3 has shown to effectively deliver siRNA for in vitro transfections, with reduced toxicity compare to PEI, therefore its effectiveness in vivo was also examined to explore its potential therapeutic value in clinical applications.

Methods:

Nude/nude athymic mice (5-6 weeks-old, female) obtained from Charles River

Laboratories (Willmington, MA) were housed in isolated cages at room temperatures

183

with access to food and water. MDA-MB-231-luc-D3H2LN cells were cultured in 15%

DMEM in 150 cm2 flasks till reaching confluency. Cells were then washed, detached and mixed with matrigel matrix (1:1 v/v ratio), and 3 million cells were inoculated subcutaneously into the mammary complex of each mouse. After tumors of >100 mm3 were formed, luc-siRNA polyplexes (150 µL) with 1:4 siRNA to polymer w/w ratio were injected into the animal via tail vein injection. One hundred and fifty µL of free luc- siRNA solution in PBS and scramble siRNA polyplexes were used as negative controls.

Animals were imaged for luciferase expressions on the following day after treatments.

Prior to imaging, 100 µL of VivoGlo Luciferin (Promega) at 10 mg/mL were injected intraperitoneal (IP) and images were taken at 10 min using IVIS® Lumina II

(PerkinElmer, Hopkinton, MA) post luciferin injection.

2.510 7 p=0.13 p=0.21 2.010 7

1.510 7 p=0.52

1.010 7

5.010 6

Luminescence (RLU) Average Radiance (RLU) Average Luminescence C

Luc-siRNA Polyplexes Free Luc-siRNA in Saline Scramble siRNA Polyplexes

184

Figure B.1 Examining the in vivo effectiveness of luc-siRNA P(SiDAAr)5Peg3 polyplex.

One dose of the polyplexes formulated with the anti-luc siRNA was injected through the tail vein in mice, and images were taken after 1 day post treatment. Control is mice without any treatments. Animals with multiple tumors were used for this study. Data are shown as mean (of the number of tumors) ± SEM. Five mice were injected with scramble siRNA NP, 4 mice were used with free luc-siRNA in saline treatment, and 4 with luc- siRNA polyplexes.

Results and Discussion:

Compare the effect of free luc-siRNA and luc-siRNA polyplexes via tail vein injection, a single-dose of the polyplexes showed no significant reduction in the luminescence expression (Fig A1). This could due to a number of reasons. First, multiple doses may be needed to achieve a more significant knock-down effect in vivo. Second, a longer time period may be required for the polyplexes to achieve effectiveness in vivo. In the past, studies have shown PEI-siRNA polyplexes able to transfect different animal models [208]. In mice model, effective HER-2 gene knock-down has been reported by single-dose of siRNA polyplexes via IP injection [64]. However, majority of the studies use either intra-tumoral injection [209] or have relatively poor effect [210]. Future works are needed to elucidate the effectiveness of this P(SiDAAr)5Peg3 in siRNA delivery.

185

APPENDIX C

HYDRODYNAMIC DIAMETERS AND THE ZETA POTENTIALS OF POLYMERS

AND POLYPLEXES

PEI polymer: mean diameter: 7.8nm, Stdev (nm): 4.487, variance: 0.331

186

P(SiAAr)5 polymer: mean diameter: 10.5 nm, stdev (nm): 6.557, variance: 0.389

P(SiAAr)5Peg3 polymer: mean diameter: 45.8 nm, stdev (nm): 29.58, variance: 0.41

187

1/O/2 DNA/siRNA/PEI Polyplexes: mean diameter: 249.9 nm, stdev (nm): 129.2, variance: 0.267

1/2 siRNA/ P(SiDAAr)5 Polyplexes: mean diameter: 571.7 nm, stdev (nm): 290.42, variance: 0.258

188

1/1/4 DNA/siRNA/P(SiDAAr)5Peg3 Polyplexes: mean diameter: 273.2 nm, stdev (nm): 115.8, variance: 0.180

1/4 siRNA/ P(SiDAAr)5 Polyplexes: mean diameter: 429.2 nm, stdev (nm): 230.075, variance: 0.287

189

Zeta Potential:

Polymer: PEI

Polymer: P(SiDAAr)5

190

Polymer: P(SiDAAr)5Peg3

1/1/2 DNA/siRNA/P(SiDAAr)5Peg3 Polyplexes

191

1/1/4 DNA/siRNA/P(SiDAAr)5Peg3 Polyplexes

192

APPENDIX D

SYNTHESIS OF P(SIDAAR)5

Materials

MO). Fetal bovine serum (FBS) and Trypsin/EDTA were obtained from Gibco (Grand

Island, NY). RNAse I (cloned) and Silencer Negative Control No.1 siRNA were purchased from Life Technologies (Carlsbad, CA). Sodium bicarbonate powder, hydroxylamine hydrochloride, uranyl acetate, and TrisAcetate-EDTA buffer were purchased from Sigma Aldrich (St. Louis, MO). Dulbecco's Modified Eagle's Medium

(DMEM), DPBS, penicillin and streptomycin were purchased from the Cell Services’

Media Core at the Cleveland Clinic.

Methods

Oligo(alkylaminosiloxane) (SiDA), arginine conjugated oligo amino alkyldialkoxymethylsilane (SiDAAr) and arginine modified

193

oligo(alkylaminosiloxane)graft-PEI ((P(SiDAAr)5) (5:1 SiDAAr: PEI mol:mol ratio) was synthesized as previously described [193, 197].

For oligo(alkylaminosiloxane) (SiDA) synthesis, 32 μL of 1 N NaOH solution was added to 1.8 mmol (0.371 g) of 3-(2-aminoethylamino)propyl-methyldimethoxysilane.

The reaction was carried on for 20 hours via stirring at room temperature. After that, reduced pressure was applied to the oligomers to remove the volatiles. Following that, 5 ml of dH2O was added to the reaction mixture and 1 N HCl was used to neutralize the mixture to pH of 7. The product of the reaction was analyzed by 1H NMR and mass spectrometry. Ratio linear/cyclic: 3.42/1. Mass: m/z = 641, 143, 302, 321, 481, 802

(Kichler et al) [197].

Arginine conjugated oligo amino alkyldialkoxymethylsilane (SiDAAr) at 5:1 molar ratio was prepared by EDC/NHS chemistry. Briefly, 0.785 g of L-arginine (5 mmol equivalents) and NHS (7 mmol equivalents, 0.725 g) was dissolved in 2.5 mL of PBS of pH 8. The solution was cooled to 4 oC and EDC (7 mmol equivalents, 1.2 g) was slowly added into the solution to activate the carboxylic acid of L-arginine. The reaction was continued for 4 hours at 4 oC and the reaction mixture was then added into alkylaminosiloxane oligomer (1 mmol equivalent, 0.8 g) dissolved in 1.5 mL of PBS. The reaction was carried out for 24 hours at room temperature. The product was then dialyzed

(MWCO=1000 Da) using deionized water for 48 hours to remove un-reacted substrates and lyophilized.

194

Conjugation of N-(tert-Butoxycarbonyl)-L-Aspartic acid (Boc-Asp) to SiDAAr:

SiDAAr coupling to PEI using 5 molar equivalents of SiDAAr (P(SiDAAr)5) was performed by EDC/NHS chemistry using Boc-Asp as a linker. Briefly, one of the acid groups of the carboxylic acid of BOC-Aspartic acid (1/5 mmol equivalents, 0.042g) was activated with 3.5/5 mmol equivalents of NHS (0.0725 g) and EDC (0.1208 g) in 0.5 mL

MES of pH 8 for 4 hours at 4 oC. The reaction mixture was then added into the 1.53 g

(1mmol equivalent) of SiDAAr solution and the reaction was continued for 18 hours.

Afterward, the reaction mixture was kept for dialysis (MWCO=1000 Da) for 48 hours to remove the un-reacted substrates.

Synthesis of P(SiDAAr)5: The other carboxylic acid of BOC-Aspartic acid (1/5 mmol equivalents, 0.042 g) was activated using 3.5/5 mmol equivalents of NHS (0.0725 g) and

EDC (0.1208 g) and the reaction was carried out for 6 hours at 4 oC. After the activation of aspartic acid, 0.05712 g (0.06 mmol equivalents) of SiDAAr with the activated BOC- aspartic acid was added to PEI (0.012 mmol equivalents; 0.3 g) dissolved in 6 mL (5% w/v) of PBS at pH of 8. Reaction was carried on for 18 hours at room temperature. The resultant mixture was dialyzed (MWCO=12k Da) for 48 hours and then lyophilized. The yield of P(SiDAAr)5 was approximately 30%. Above polymer conjugations were carried out by Viola Morris, a former postdoc in Dr. Labhasetwar group.

195

196

Scheme D.1 Synthesis of (A) oligo(-alkylaminosiloxanes) (SiDA), (B) L-arginine modified oligo (-alkylaminosiloxanes) (SiDAAr5), (C) L-arginine oligo

(alkylaminosiloxane) grafted poly(ethylenimine) (P(SiDAAr)5).

197