Chemoprevention of Oral Squamous Cell Carcinoma: Extending Therapeutic Parameters of Fenretinide

Chemoprevention of Oral Squamous Cell Carcinoma: Extending Therapeutic

Parameters of Fenretinide

DISSERTATION

Presented in Partial Fulfillment of the Requirements for the Degree Doctorate of Philosophy in the Graduate School of The Ohio State University

Byungdo Brian Han, B.S.

Graduate Program in Oral Biology

The Ohio State University

2015

Dissertation Committee:

Susan R. Mallery, D.D.S., Ph.D., Advisor

James C. Lang, Ph.D.

John D. Walters, D.D.S., M.M.Sc.

Ann L. Griffen, D.D.S., M.S.

Byungdo Brian Han, B.S.

2015

Abstract

Oral squamous cell carcinoma (OSCC) represents the primary cancer of the oral cavity and poses a world-wide health problem. Notably, OSCC does not arise de novo, but from malignant transformation of clinically detectable premalignant epithelial lesions known as oral intraepithelial neoplasia (OIN). OIN’s visible nature and clinical accessibility render this process a prime target for local chemopreventive strategies. OIN chemoprevention is a patient-centric, optimal strategy to prevent OSCC’s co-morbidities and mortality. The membrane-associated protein, focal adhesion kinase (FAK), modulates cell-extracellular matrix interactions and also conveys pro-survival and proliferative signals. Notably, increased intraepithelial FAK levels accompany transformation of premalignant oral intraepithelial neoplasia (OIN) to oral squamous cell carcinoma (OSCC). The cancer chemopreventive and synthetic vitamin A derivative, fenretinide, has demonstrated protein-binding capacities e.g. mTOR and retinol binding protein interactions. These studies employed a continuum of human oral keratinocytes

(normal-HPV E6/E7-transduced-OSCC) to assess potential fenretinide-FAK drug protein interactions and functional consequences on cellular growth regulation and motility.

Molecular modeling studies demonstrated fenretinide has ~200-fold greater binding

ii affinity relative to the natural ligand (ATP) at FAK’s kinase domain. Fenretinide also shows intermediate binding at FAK’s FERM domain and interacts at the ATP-binding site of the closest FAK analogue, Pyk2. Fenretinide significantly suppressed proliferation via induction of apoptosis and G2/M cell cycle blockade. Fenretinide-treated cells also demonstrated F-actin disruption, significant inhibition of both directed migration and invasion of a synthetic basement membrane, and decreased phosphorylation of growth- promoting kinases. A commercially available FAK inhibitor did not suppress cell invasion. Notably, while FAK’s FERM domain directs cell invasion, FAK inhibitors target the kinase domain. In addition, FAK-specific siRNA treated cells showed an intermediate cell migration capacity; data which suggest co-contribution of the established migrating-enhancing Pyk2. Our data imply that fenretinide is uniquely capable of disrupting FAK’s and Pyk2’s pro-survival and mobility-enhancing effects and further extend fenretinide’s chemopreventive contributions beyond induction of apoptosis and differentiation.

Intra-/inter-tumor heterogeneity of oral squamous cell carcinoma (OSCC) poses a conundrum to conventional therapeutic management, which assumes the presence of therapeutic targets within OSCC precursor lesions or tumors. Notably, cetuximab, a

FDA-approved EGFR-targeting monoclonal antibody for locally advanced HNSCC, has encountered chemoresistance, lowering its clinical efficacy; OSCC cells activate compensatory signaling pathways to ensure continual proliferation and survival.

Personalized tumor profiling, therefore, enables identification of aberrant molecular and

iii biochemical phenotypes and customization of therapeutic strategy to fit individual patient’s OSCC profiles. Characterization of specific tumorigenic signaling pathways and cytokine secretion profiles revealed significant heterogeneity among recently isolated

OSCC cell lines (JSCC-1, JSCC-2, and JSCC-3), leading to varied response to specific inhibition of key tumorigenic pathways (e.g. EGFR and STAT3); Afatinib, specific

EGFR inhibitor, was able to inhibit STAT3 phosphorylation in JSCC-3 but not in JSCC-1 and JSCC-2. Notably, STAT3 inhibition-refractory cells lines were responsive to fenretinide, a vitamin A-derived chemopreventive. Previous studies have demonstrated fenretinide’s capacity to modulate cellular phosphorylation status toward antitumorigenic phenotypes. Fenretinide combined with specific small molecule inhibitors

(i.e. afatinib and vargatef) were able to prevent activation of targeted tumorigenic pathway as well as redundant signaling. Our data demonstrated that fenretinide is capable of regulating key tumorigenic pathways both as a single agent and an adjunct drug.

Dedicated to my wife, Heather, and my three children, Katie, Joey, and Ella.

Acknowledgments

First of all, I would like to express my most sincere gratitude to my mentor, Dr. Susan

Mallery, for her tireless dedication to my success as a clinician scientist. Her guidance, sage advice, and infectious positivity helped me to grow both personally and professionally.

I would also like to thank my committee members: Dr. James Lang for intellectually stimulating conversations on diverse topics ranging from running to head and neck cancer biology; Dr. John Walters for being my advocate in the oral biology program and teaching me crucial scientific and life lessons both in and out of the clinic; Dr. Ann

Griffen for providing invaluable guidance for my pursuit of becoming an academic pediatric dentist. My dual degree training would not certainly have been possible without

Dr. John Sheridan, who has been facilitating this incredible program and providing me continual encouragements throughout the program, and Dr. David Padgett, my first mentor in the program, for trusting me and positively affirming my decision to pursue this career path.

I would like to thank my friends and colleagues who shared many unforgettable moments throughout the past 8 years: Drew Holpuch, Meng Tong, Blake Warner, Matt Mason,

Heidi Steinkamp, Sukirth Ganesan, Shareef Dabdoub, Amy Mackos, Scott Wray, Ping

Pei. You have shared my joy and sorrow and taught me so much about life. I am also thankful for all the Mallery lab members with whom I have had the privilege of working together.

Finally, I owe so much to my family for unconditionally supporting my educational endeavors: My mom for her incredible faith and unceasing prayers for me; my three precious children who turned my world upside down and brought to my life more love than I ever thought possible; and Heather, my wife and the love of my life, for her support, encouragement, and unyielding devotion.

This research was funded by the National Institutes of Health grants: T32 DE14320, R01

CA129609, RC2 CA148099, R21 CA132138.

vii

VITA

May 2002…………………....Bob Jones Academy, Greenville, SC

May 2006……………………B.S. in Pre-dentistry, Bob Jones University

2007 to Present……...………D.D.S./Ph.D. Graduate Fellow,

Comprehensive Training in Oral and Craniofacial Sciences,

The Ohio State University College of Dentistry

PUBLICATIONS

Han BB, Li S, Tong M, Holpuch AS, Spinney R, Wang D, Border MB, Liu Z, Sarode S, Pei P, Schwendeman, SP, Mallery SR. Fenretinide Perturbs Focal Adhesion Kinase in Premalignant and Malignant Human Oral Keratinocytes. Fenretinide's chemopreventive mechanisms include ECM interactions. Cancer Prev Res (Phila). 2015 Feb 24. pii: canprevres.0418.2014. [Epub ahead of print]

Mallery SR, Tong M, Shumway BS, Curran AE, Larsen PE, Ness GM, Kennedy KS, Blakey GH, Kushner GM, Vickers AM, Han B, Pei P, Stoner GD. Topical application of a mucoadhesive freeze-dried black raspberry gel induces clinical and histologic regression and reduces loss of heterozygosity events in premalignant oral intraepithelial lesions: results from a multicentered, placebo-controlled clinical trial. Clin Cancer Res. 2014 Apr 1;20(7):1910-24.

Tong M, Han BB, Holpuch AS, Pei P, He L, Mallery SR. Inherent phenotypic plasticity facilitates progression of head and neck cancer: endotheliod characteristics enable angiogenesis and invasion. Exp Cell Res. 2013 Apr 15; 319(7): 1028-42.

Holpuch AS, Phelps MP, Desai KG, Chen W, Koutras GM, Han BB, Warner BM, Pei P, Seghi GA, Tong M, Border MB, Fields HW, Stoner GD, Larsen PE, Liu Z, Schwendeman SP, Mallery SR. Evaluation of a mucoadhesive fenretinide patch for local intraoral delivery: a strategy to

viii reintroduce fenretinide for oral cancer chemoprevention. Carcinogenesis. 2012 May; 33(5):1098- 105.

FIELDS OF STUDY

Major Field: Oral Biology

TABLE OF CONTENTS

Abstract ...... ii

Dedication ...... v

Acknowledgements ...... vi

Vita ...... viii

List of Tables ...... xiii

List of Figures ...... xiv

Chapter 1: Introduction ...... 1 1.1 Statement of the clinical problem: Oral squamous cell carcinoma ...... 1 1.2 Premalignant lesions of oral squamous cell carcinoma ...... 3 1.3 Chemoprevention: Origin and History ...... 8 1.4 Chemopreventive targets ...... 10 1.4.a Apoptosis ...... 11 1.4.b Terminal Differentiation ...... 12 1.4.c Cell Migration and Invasion ...... 15 1.4.d Inflammation, Angiogenesis, and Tumor Microenvironment ...... 17 1.5 Delivery Strategy for Chemoprevention of Oral Cancer: Local Delivery ...... 19 1.6 Fenretinide ...... 21 1.7 References ...... 24

Chapter 2: : Fenretinide Perturbs Focal Adhesion Kinase in Premalignant and Malignant Human Oral Keratinocytes. Fenretinide’s chemopreventive mechanisms include ECM interactions...... 33 2.1 Abstract ...... 33 2.2 Introduction ...... 34 2.3 Materials and Methods ...... 36 2.4 Results ...... 43 2.5 Discussion ...... 48 2.6 References ...... 55

Chapter 3: Personalized Tumor Profiling: Utilizing Fenretinide (4-HPR) as a Biological Modifier to Optimize Therapeutic Outcome ...... 80 3.1 Abstract ...... 80 3.2 Introduction ...... 81 3.3 Materials and Methods ...... 84 3.4 Results ...... 90 3.5 Discussion ...... 96 3.6 References ...... 101

Chapter 4: Executive Summary and Future Research Directions ...... 117 4.1 Fenretinide Perturbs Focal Adhesion Kinase in Premalignant and Malignant Human Oral Keratinocytes. Fenretinide’s chemopreventive mechanisms include ECM interactions...... 117 4.2 Personalized Tumor Profiling: Utilizing Fenretinide (4-HPR) as a Biological Modifier to Optimize Therapeutic Outcome ...... 118 4.3 Future Research Directions ...... 120

Bibliography ...... 121

Appendix A: Inherent phenotypic plasticity facilitates progression of head and neck cancer: endotheliod characteristics enable angiogenesis and invasion...... 137 A.1 Abstract ...... 137 A.2 Introduction ...... 138 A.3 Materials and Methods ...... 141 A.4 Results ...... 147 A.5 Disccusion ...... 155 A.6 Conclusions ...... 162 A.7 References ...... 163

Appendix B: Evaluation of a Mucoadhesive Fenretinide Patch for Local Intraoral Delivery: A Strategy to Re-introduce Fenretinide for Oral Cancer Chemoprevention ...... 180 B.1 Abstract ...... 180 B.2 Introduction ...... 181 B.3 Materials and Methods ...... 183 B.4 Results and Discussion ...... 193 B.5 Conclusions ...... 202 B.6 References ...... 210

xii

LIST OF TABLES

Table 2.1. Levels of 4-HPR in OSCC cells and complete medium during incubation...... 78

2.2: 4-HPR binding Interactions with Pyk2 ...... 79

xiii

LIST OF FIGURES

Figure 2.1. Activation of caspase 3/7 and cell cycle modulations by 4-HPR ...... 61

2.2. 4-HPR disrupts actin cytoskeleton organization ...... 62

2.3. 4-HPR interacts with FAK’s kinase and FERM domains ...... 63

2.4. Evaluation of 4-HPR’s effects on directed cell migration ...... 65

2.5. Evaluation of the effects of 4-HPR and freeze dried blackraspberries (BRB) on cell invasion...... 67

2.6. Modulation of phosphorylation status of kinases associated with cell migration, proliferation, survival, and apoptosis by 4-HPR and BRB ...... 69

2.7: JSCC cell lines contain cytokeratin and vimentin, suggestive of epithelial-to- mesenchymal transition in subpopulations ...... 71

2.8: FAK inhibitor II demonstrated negligible invasion-inhibitory effects on EPI and SCC2095sc cell lines...... 72

2.9: Fenretinide (4-HPR) inhibits cell migration in transwell migration assay...... 73

2.10: Evaluation of fenretinide (4-HPR)’s effects on cell viability ...... 74

2.11: Fenretinide inhibits tubulin polymerization...... 75

2.12: Evaluation of fenretinide (4-HPR)’s effects on tubulin polymerization...... 76

3.1. Determination of optimal dosing levels for multi-agent chemopreventive strategy...... 105

xiv

3.2. Extensive inter-tumor heterogeneity in specific growth factor and treatment response profile in OSCC tumor-derived cells...... 106

3.3. Inter-tumor heterogeneity extends to unique cytokine profiles among OSCC cell lines ...... 107

3.4. Determination of autologous IL-6 and EGF levels in OSCC cells ...... 109

3.5. 4-HPR abrogates EGFR and STAT3 phosphorylation...... 111

3.6.: Fenretinide differentially modulates STAT3 activation pathways ...... 112

3.7. 4-HPR-mediated inhibition of STAT3 signaling...... 113

3.8. Evaluation of freeze-dried black raspberry extract (BRB) and 4-HPR ‘s combined treatment effects...... 115

A.1. HNSCC cells express endothelial-associated proteins...... 168

A.2. HNSCC cells express vimentin and CD31 proteins...... 169

A.3. HUVECs and HNSCC cells form tubular networks on Matrigel...... 170

A.4. HNSCC cells internalize AcLDL, which is enhanced following TGF-β1 challenge...... 171

A.5. TGF-β1 induces an endotheliod phenotype in HNSCC cells...... 173

A.6. Endothelial associated proteins are also present in HNSCC tumors...... 175

A.7. VEGF activates comparable signaling pathways in HNSCC and endothelial cells...... 177

A.8. Densitometric analyses results of Western blot data presented in Figure A.7...... 179

B.1. Mucoadhesive patch attachment on rabbit oral mucosa...... 185

B.2. Intra-oral fenretinide patch application delivered pharmacological intra-oral levels and did not induce any deleterious side-effects...... 195

B.3. Quantified immunohistochemical analyses showed fenretinide’s dose-specific effects on treated rabbit oral epithelia growth state...... 197

B.4. Patch-delivered fenretinide effects intra-epithelial metabolic enzyme profile...... 198

B.5. Immunoblots of normal human oral mucosa revealed considerable inter-donor heterogeneity in a fenretinide bioactivating enzyme, CYP3A4...... 204

B.6. Fenretinide and its bioactive metabolite 4-oxo-4-HPR demonstrated preferential growth suppression towards premalignant (dysplastic) oral keratinocytes...... 205

xvi

Chapter 1: Introduction

1.1. Statement of the clinical problem: Oral squamous cell carcinoma

Cancer of the oral cavity and the pharynx is a worldwide health issue that is associated with significant morbidity, mortality, and an economic cost that rival or exceed other solid tumors [1-3]. The American Cancer Society estimates that approximately 45,780 new cases will be diagnosed in 2015 with estimated 8,650 deaths as a result of this disease in the United States alone [4]. Oral cancer most frequently occurs in middle-age or older individuals. Epidemiological studies demonstrate gender predilection favoring male over female (over 2:1) [4,5]. Oral cancer constitutes approximately 3% of all malignancies, and about 90% of oral cancer manifests as oral squamous cell carcinoma (OSCC) [4,6]. The World Health Organization reports the incidence of more than 300,000 OSCC cases every year worldwide [6].

The most common intraoral sites for oral squamous cell carcinoma in the United

States include the ventrolateral tongue and the floor of the mouth [6]. Commonly affected sites may vary in other regions of the world where population experiences prolonged exposure to different risk factors (e.g. buccal mucosa due to betel quid and/or smokeless tobacco use in India). Tobacco and alcohol are significant risk factors to OSCC development, and together they synergistically increase the risks of OSCC [5]. 1

OSCC carcinogenesis is a multistep, multifocal process which involves exposure to carcinogens, resultant genetic mutations, alteration of cellular metabolic profiles and tumor microenvironments, and migration and invasion [7,8]. Highly complex nature of

OSCC development confounds identification of clearly defined etiologic agents. It is suspected that both extrinsic and intrinsic factors play a role in OSCC development [5].

In addition to the aforementioned extrinsic factors (e.g. tobacco, alcohol, and betel quid),

OSCC lesions located in the oropharynx are strongly associated with high risk human papilloma viruses (HPV16 and 18) [9]. HPV16 and 18 utilize their oncoproteins E6 and

E7 to inactivate p53 and retinoblastoma protein (pRb), respectively, thereby disrupting cell cycle regulation [10,11]. In addition, genetic factors may play a significant role in

OSCC development. Notably, a small group of individuals who exhibit genetic polymorphism of genes responsible for xenobiotic metabolism, detoxification, and DNA repair may show greater susceptibility to carcinogens [12,13]. Emerging evidence suggests associations between specific bacterial species, e.g., Porphymonas gingivalis and Fusobacterium nucleatum, and oral cancer [14]. This intriguing concept of bacteria as etiologic agents or co-factors, however, warrants further investigation.

Continual carcinogenic onslaughts on epithelial cells lead to accumulation of genetic mutations in proto-oncogenes and tumor suppressor genes. OSCC can take several years to develop, and often multiple intraoral sites concurrently undergo preneoplastic changes, albeit at different rates. This phenomenon is consistent with the notion “field cancerization” [15]. Genetic analysis using microsatellite markers (e.g. chromosomes 3p, 9p, 17p) revealed that tumor and genetically altered fields share certain

2 common events, which likely occurred during the initiation stage [16-18]. They, however, also possess diverging genetic alterations mutually exclusive of each other [16-

18]. Once altered, these allelic imbalances at key genetic sites, e.g., p53, p16, FHIT, set the stage for eventual malignant transformation [19].

Advances in OSCC management have brought modest improvements in 5-year survival rates (up to 66% from mid-50%) [4]. Management of oral squamous cell carcinoma depends on the stage and the location of the lesion. Surgery remains the primary treatment modality for early-staged lesions, whereas larger and/or metastatic lesions may necessitate surgical resection along with chemoradiation therapy. Even those patients who receive a surgical cure encounter significant morbidity due to gross facial disfigurement and compromised function. Clearly, early detection of premalignant oral lesions in conjunction with effective chemopreventive measures could improve clinical outcomes.

1.2. Premalignant lesions of oral squamous cell carcinoma

Oral premalignant lesions are classically classified as leukoplakia, erythroplakia, and erythroleukoplakia. These lesions clinically appear as well-defined areas of white, red, and mixed colors, respectively. The majority of leukoplakia occurs on the buccal mucosa, gingiva, or lower lip. Dysplasia, however, is most often detected in leukoplakia located on the ventrolateral tongue, the floor of the mouth, and lower lip [5], where prolonged exposure to carcinogen-rich saliva occurs. Leukoplakia, erythroplakia, and

3 erythroleukoplakia represent clinical diagnoses from exclusion; they describe lesions that cannot be characterized as any other pathologic entity. The clinical practice of using these terms to refer to oral precancerous lesions have resulted in much confusion and controversy. In recent years, the term, “oral intraepithelial neoplasia (OIN), was proposed to provide a more concrete and unifying notion for premalignant oral lesions [20]. This new terminology utilizes histological criteria to establish premalignancy of clinically suspected lesions.

The oral mucosa consists of oral epithelium (stratified squamous non-keratinized epithelium) and underlying connective tissue called lamina propria. Oral epithelium is further divided into four layers: stratum basale, spinosum, granulosum, and corneum. In normal epithelium, only the deepest layer (i.e. stratum basale) contains dividing cells which would progressively undergo terminal differentiation as they migrate through more superficial layers. Normal epithelial cells possess intricately balanced mechanisms to carefully guard cells from entering into unwarranted cell division. P53, colloquially termed “the guardian of the genome,” exemplifies one such powerful molecular switch that prevents aberrant proliferation [21]. In oral dysplasia, a series of these safeguards are dismantled, and keratinocytes gain an uncontrolled ability to proliferate throughout the partial- or full-thickness epithelium [22,23].

By definition, malignant transformation of oral dysplasia to overt oral squamous cell carcinoma (OSCC) occurs when premalignant cells breach the basement membrane and invade into the stroma. This capacity to invade and metastasize represents one of the key biological traits of cancer cells [24,25]. Premalignant epithelial cells must acquire the

4 capacity to detach themselves from neighboring keratinocytes (E-cadherin, and N- cadherin), resist apoptosis (FAK, Bax, Bcl-2), migrate (FAK), and invade (MMP-2/9).

Epithelial-to-mesenchymal transition (EMT) is now well established as one of the most critical processes to enable such cellular plasticity during carcinogenesis [26,27].

Notably, focal adhesions kinase (FAK), which is implicated in proliferation, cell survival, and migration/invasion, also plays an integral role in EMT [28]. Targeting such key signaling mediator, therefore, could provide distinct chemopreventive benefits (See

Chapter 2).

Silverman et al and others demonstrated that microscopically confirmed oral premalignant lesions undergo malignant transformation up to 36% [29,30].

Histopathologic grades of dysplastic lesions, albeit useful, do not provide accurate means of predicting malignant transformation potential. Premalignant keratinocytes show a wide range of aberrant phenotypes including ploidy status, allelic imbalance, oncogene expressions or tumor suppressor gene inactivations [31]. Numerous studies have investigated individual biomarkers as predictive indicators for malignant transformation

[30,32-33]. Premalignant keratinocytes undergo allelic imbalance over time and exhibit loss of heterozygosity (LOH) at specific chromosomal regions (e.g. 3p, 9p, 17p). These

LOH sites often involve key tumor suppressor genes (e.g. p53, p16, FHIT) that become inactivated due to their microsatellite instability [34-35]. Notably, studies have demonstrated specific sites of allelic imbalance may serve as a risk indicator [36].

Whereas premalignant lesions with LOH occurring only at 3p and/or 9p have a 3.8-fold risk increase, those with additional site involvement (e.g. 17p) on top of 3p/9p have

5 greater than a 30-fold increase in relative risk [36]. In addition, DNA content (i.e. ploidy status) may serve as predictive indicator for malignant transformation. Aneuploid cells show a higher probability of malignant transformation than tetraploid and diploid counterparts [37,38]. A significant number of oral cancer cases, however, exhibit diploidy and would require additional means of determining cancer risk [39].

Furthermore, a great deal of interpatient heterogeneity exists within premalignant lesions as well as OSCC [31]. A collective panel of biomarkers, therefore, would provide more reliable methods of predicting malignant transformation potential [31].

This heterogeneous nature of cancer should come as no surprise, given the significant role that random mutations play in carcinogenesis. This is especially true for cancers that are associated with environmental carcinogens; the rate of somatic mutations of head and neck squamous cell carcinoma (HNSCC) is significantly higher than pediatric and hematological cancers [42]. As cancer represents a microcosm of evolution, each neoplasm undergoes selection process and develops a unique set of survival advantages to cope with the hostile host milieu [40]. This inevitable diversity poses a tremendous challenge to the development of efficacious therapeutic strategy. The complexity of the issue is exacerbated by genetic and metabolic heterogeneity of individual patients. For example, our clinical trial results show a significant inter-patient variability in anthocyanin-related metabolic enzymes [41]. Predicting high risk patients through reliable biomarkers, patient-specific tumor profiling, and personalizing treatment according to patient’s genetic and metabolic make-up would open up a new era of personalized medicine and generate optimal treatment strategy for every patient.

In addition to the monumental task of identifying high risk lesions and devising personalized treatments to account for inter-patient heterogeneity, it is crucial to design a delivery strategy enabling drugs of choice to reach the effective therapeutic levels to the target tissue (see Appendix B). Furthermore, additional research is warranted to understand the exact nature and extent of the premalignant lesional tissue. While clinical detection of suspected premalignant lesions (e.g. leukoplakia) often shows well-defined, focal areas of involvement, the aforementioned concept of field cancerization must be considered during management of premalignant OSCC lesions.

The concept of “field cancerization” was first introduced in 1953 by Slaughter et al to explain frequent post-treatment loco-regional recurrences of oral squamous cell carcinoma and emergence of multi-focal tumors in the oral cavity [15]. This term denotes a process by which one or more areas of mucosal epithelium undergo genetic or epigenetic alterations and becomes pre-neoplastic, albeit clinically undetectable (i.e. cancer initiation). These lesions would require additional molecular changes to reach malignancy and may or may not exhibit histological signs of dysplasia. Tabor et al analyzed microsatellite markers at chromosomal loci frequently involved in preneoplastic changes (3p, 9p, 17p) using tumor-adjacent normal mucosa in 28 HNSCC patients. The results showed that about 36% of the patients possessed genetically altered fields [17]. This concept of field cancerization, therefore, bears profound therapeutic implications for patient management. The current standard of care, which involves resection of malignant lesions with confirmation of a clean margin, would potentially leave the “cancerized field” intact, risking locoregional recurrence or second primary

7 tumor development. Therapeutic strategy, therefore, should target both lesion-specific areas as well as general area coverage to provide optimal clinical benefits to the patient.

1.3. Chemoprevention: Origin and History

Michael B. Sporn, in 1976, first coined the term chemoprevention [43].

Chemoprevention can be defined as the use of natural, synthetically derived, or biologic agents to prevent, suppress, or reverse carcinogenic progression to invasive cancer [6,

44]. The wisdom behind chemoprevention is not limited to cancer. Medical community has applied this preventive strategy to cardiovascular disease over the past three decades.

This pharmacologic intervention of lowering serum cholesterol levels (e.g. simvastatin), inhibiting platelet aggregation (e.g. aspirin), and controlling blood pressure (e.g. thiazide diuretics) has resulted in significant reduction of cardiovascular-related morbidity and mortality [44].

The ideal cancer chemopreventive and chemotherapeutic treatment shares the same vision with other chemotherapeutic interventions: target offending cells/organisms with a minimal damage to non-offending, normal cells. Paul Ehrlich, who discovered

“magic bullets” for syphilis and diphtheria treatment, coined the term chemotherapy and demonstrated the effective use of animal models for the purpose of drug screening [45].

This major advancement in scientific thinking reverberates to this day with numerous drugs being successfully used to treat various human diseases including cancer. Unlike targeting virulent microbial organisms, cancer therapy poses a unique challenge as

8 genotypic and phenotypic differences between the host (i.e. normal cells) and the target

(i.e. premalignant and malignant cells) are greatly minimized. Despite molecular perturbations resulting in differential expression patterns of cell surface receptors and cytokines, cancer cells essentially share the same set of cellular machinery as normal cells. Key differences, however, do exist: (1) rate of cell proliferation, (2) metabolic profile, (3) cell signaling, and (4) tumor microenvironment [24, 25]. These key differences laid conceptual framework for the development of innovative treatment strategies, such as (1) chemotherapy targeting DNA replication, (2) tumor detection using radioactive glucose during Positron Emission Tomography (PET) and Computed

Tomography (CT) imaging, (3) tyrosine kinase inhibitors, and (4) anti-angiogenic agents.

Despite the aforementioned scientific advances, cancer remains a serious health threat to our society as approximately ½ males and 1/3 females on average will develop cancer in their lifetime [46]. Notably, oral squamous cell carcinoma represents a debilitating disease which accompanies devastating esthetic and financial ramifications even with complete remission. Successful chemopreventive therapy, therefore, provides not only a greater chance for survival but also an opportunity to avoid battling a horrific disease head-on. Steward and Brown explain three subsets of chemopreventive strategy:

(1) primary chemoprevention to target healthy patients with established risk factors, (2) secondary chemoprevention as pharmacologic intervention to deter premalignant lesions from progressing to overt cancer, (3) tertiary chemoprevention to prevent recurrence or development of second primary cancers [44].

Successful implementation of chemoprevention, particularly primary chemoprevention, predicates on our knowledge of risk factors and quantifiable biomarkers. Increased risks of morbidity and mortality linked to rigorously tested biomarkers would provide credence to patients and increase patient compliance. In addition, these biomarkers would allow more targeted approach to chemoprevention and could serve as measurable therapeutic endpoints [44]. Given the genetic and metabolic interpatient heterogeneity, a panel of genetic and molecular biomarkers will likely be required to accurately assess risks and generate useful prognostic information for patients.

1.4. Chemopreventive targets

Conventional chemotherapy relies on its ability to exploit genotypic and phenotypic differences between normal and premalignant/malignant cells. A similar paradigm applies to OSCC chemoprevention. Common chemopreventive targets manipulate key cellular and molecular differences existing between premalignant and normal epithelial cells (e.g. intracellular redox state and aberrations in signaling pathways). In addition, increased understanding of tumor microenvironment’s role in carcinogenesis warrants inclusion of non-epithelial components as chemopreventive targets (e.g. angiogenesis).

1.4.a. Apoptosis

Apoptosis, also known as programmed cell death, is an essential part of homeostatic mechanisms to eliminate unwanted or deleterious cells [47]. During human embryonic development, cells located between digits undergo apoptosis to achieve digital separation [48]. Natural killer cells can induce apoptosis via Fas ligand engagement or cytotoxic granule secretion during tumor surveillance [49]. In addition, cells have mechanisms to "self-detonate" when genomic integrity is compromised at cell cycle checkpoints. P53 represents one such crucial molecular switch to trigger apoptotic cascades upon detection of genomic instability [50]. Premalignant cells, therefore, must not only acquire an ability to bypass cell cycle checkpoints but also to evade detection by or disable apoptotic machineries [51,52]. Indeed, approximately 40% of primary oral squamous cell carcinoma possesses mutated p53 [53]. In addition, tumor cells’ evasion of apoptosis and sustained proliferation could occur via deregulation of key proteins involved in cell cycle checkpoints (e.g. Rb, p16ink4a) or cellular mediators responsible for regulating such key proteins (e.g. Mdm2).

Apoptosis represents one of the prominent mechanisms of chemotherapy [54].

Several FDA-approved chemotherapeutic agents induce genotoxic damage, leading to apoptosis. Mechanisms of action include intercalating into DNA (e.g. doxorubicin), DNA strand breaks (e.g. bleomycin), and crosslinking DNA (e.g. cisplatin) [54]. Since the inception of chemopreventive strategy, it has been demonstrated that most chemopreventive measures result in cytostatic effects (e.g. inhibiting proliferation, inducing differentiation, halting migration and invasion), implying life-long

11 administration of chemopreventive agents [55]. Interestingly, studies demonstrated the possibility that premalignant lesions may be more susceptible to reactive oxygen species

(ROS)-generating chemopreventive agents, providing a potential therapeutic point of intervention [56]. This, in part, could be explained that premalignant lesions may show increased levels of intracellular reactive oxygen species with poor normal glutathione levels, leading to increased sensitivity to ROS-generating chemopreventive agents (e.g. fenretinide) [56,57]. Collectively, apoptosis represents a viable chemopreventive mechanism. Notably, many chemopreventive agents produce pleiotropic effects. Black raspberries induce terminal differentiation and inhibit proliferation [58]. Epigallocatechin gallate (EGCG) reduces cell proliferation, decreases migration, and inhibits angiogenesis

[59,60]. Finally, fenretinide is capable of inducing both terminal differentiation and apoptosis in a dose-dependent manner [61].

1.4.b. Terminal differentiation:

Keratinocytes could undergo another form of programmed cell death called terminal differentiation [62]. Differentiation represents a normal part of cellular maturation process, beginning from a zygote to terminally differentiated organ and tissue specific cells. In oral mucosa, adult stem cells residing in the basal layer of oral epithelium undergo either self-renewal or generate sister cells destined for final maturation as part of stratified squamous epithelium. Keratinocytes in epithelium are tightly bound together by adhesion junctions and desmosomes to facilitate intercellular signaling and function as a cohesive unit [62,63]. In addition, basal keratinocytes are 12 anchored to the basement membrane via integrin family members (e.g. alpha5beta1, fibronectin receptor) [63]. This tightly orchestrated spatial arrangements between basal keratinocytes and the base membrane enable fine control of terminal differentiation process as well as provides necessary proliferative and survival signals for basal keratinocytes [62,63].

While apoptosis achieves complete and orderly removal of cellular components via phagocytosis, terminal differentiation of keratinocytes results in utilization of dead cells as insoluble protein matrix serving as barrier (i.e. cornification) to environmental assaults (e.g. carcinogens) [62]. Cornification involves transglutaminase-mediated crosslinking of cytokeratins to other terminal differentiation markers, such as loricrin and involucrin [62]. While apoptosis and terminal differentiation share several common intracellular events including nuclear degradation, organelle destruction, and cytoskeletal rearrangements, both processes employ distinct signaling mediators and pathways. For example, p63, a p53 family member, plays a key role in epithelial differentiation as evidenced by p63-deficient mice exhibiting developmental skin defect [62]. In contrast, p53-deficient mice do not impede epithelium formation, suggesting a lack of p53 involvement in cornification [62]. In addition, transglutaminases (i.e. I, III, V) essential for cornified envelope formation do not seem to participate in apoptosis [62].

In oral dysplasia, basal keratinocyte proliferation extends into the lower third

(mild), middle third (moderate), upper third (severe), or the full-thickness epithelium

(carcinoma in situ). Under normal conditions, detachment of normal keratinocytes triggers terminal differentiation process and results in attenuation of survival and

13 proliferative pathways, such as phosphoinositide 3-kinase (PI3K)-Akt, epidermal growth factor receptor (EGFR), and signal transducer and activator 3 (STAT3) [62,64,65,68].

Not only all three markers show significantly increased expressions in oral epithelial dysplasia, but also EGFR and STAT3 correlate positively with worse clinical outcomes

[66-68]. Furthermore, inhibition of proliferative phenotypes by blocking PI3K or EGFR could initiate keratinocyte differentiation [69].

In addition to maintaining the survival and proliferative potential, premalignant keratinocytes “de-differentiate” from their terminal state by down-regulating epithelial markers (e.g. E-cadherin) and up-regulating mesenchymal markers (e.g. N-cadherin, vimentin). This phenotypic plasticity is termed epithelial-to-mesenchymal transition

(EMT) and plays a pivotal role in cell migration, invasion, and metastasis [27]. During

EMT, epithelial cells undergo cytoskeletal remodeling, loses apico-basal polarity, and gain mobile phenotype [27]. Notably, activity of focal adhesion kinase (FAK), which acts a modulator of extracellular signals, plays an integral role in EMT [70]. In addition, a recent study demonstrated FAK’s intimate involvement in epithelial differentiation by rescuing keratinocytes from “detachment-induced cell death” [71]. Given the cellular and molecular changes associated with deregulated terminal differentiation, chemopreventive strategy targeting PI3K, EGFR, STAT3, and/or FAK could inhibit aberrant cell proliferation, abrogate unchecked cell survival, and resume keratinocyte differentiation.

Retinoids, collectively referring to vitamin A metabolites and their synthetic analogs, have been a subject of active investigation over half a century for its potent ability to induce cellular differentiation and regulate cell proliferation [72]. Wolbach and

Howe’s seminal work using vitamin A deprived rat model demonstrated epithelial cells’ failure to achieve physiologic differentiation [73]. In addition, some of vitamin-A deficient epithelia showed an increase in cell proliferation [73]. As cancer fundamentally involves disruption of cellular differentiation, investigations in the past 50 years have focused on (1) identification of potent retinods for therapeutic use and (2) elucidation of retinoids’ mechanisms of action on cellular differentiation and profileration. As a result, various retinoids have been accepted for clinical use. 13-cis retinoic acids (Isotretinoin, trade name: Accutane) is used to treat cystic acne, and has shown potential in chemoprevention clinical trials for oral squamous cell carcinoma [74]. Isotretinoin’s sister compound, tretinoin (all-trans retinoic acid) is currently approved for treatment of acute promyelocytic leukemia for its ability to induce differentiation in leukemic promyelocytes [75]. Notably, fenretinide, a synthetic analog of all-trans retinoic acid, is capable of inducing terminal differentiation as well as triggering apoptosis [76].

1.4.c. Cell Migration and Invasion

Acquisition of migratory and invasive capabilities by premalignant cells is an important step for malignant transformation and metastasis. Keratinocyte migration depends on the ability to detach from the basement membrane, evade apoptosis, sustain survival, and assume migratory phenotypes by activating EMT. This migratory behavior requires complex interactions involving cytoskeletal re-organizations, focal adhesion turnovers, and transcriptional and epigenetic modulations. During cell invasion, keratinocytes must degrade extracellular matrix (ECM) and basement membrane (BM) 15 by secreting proteolytic enzymes (e.g. matrix metalloproteases (MMP) 2/3/9- type IV collagen, plasminogen activator-laminin, fibronectin) [77, 78]. Increasing understanding of tumor-stroma interactions demonstrates that stromal cells, such as macrophages, also contribute to ECM and BM degradation process by secreting various MMPs, cysteine cathepsins, and serine proteases [79].

Studies have demonstrated that highly invasive cells are capable of forming actin- rich membrane protrusions, called invadopodia. Invadopodia contain several actin reorganization proteins (Arp2/3, Wiscott-Aldrich syndrome proteins) as well as ECM- degrading proteins (e.g. MMPs) [80]. Epidermal growth factor receptor (EGFR), a receptor tyrosine kinase, plays a key role in growth regulation and cell migration/invasion

[81]. Epidermal growth factor (EGF), a ligand for EGFR, can serve as both proliferative signal and chemoattractant [81]. Notably, Hwang et al demonstrated that EGFR signaling is crucial for invadopodia formation as evidenced by treatment with EGFR inhibitors resulting in abrogation of invadopodia [82]. Another key player in invadopodia formation is focal adhesion kinase (FAK), which regulates interactions between actin cytoskeleton and ECM via integrins and growth factor receptors [80]. Only wild-type FAK with phosphorylation-capable tyrosine 397 residue (Y397) could restore formation and invasive function of invadopodia [80]. C-Src, a downstream target of EGFR kinase activity, mediates crosstalk between EGFR and FAK by binding to Y397 residue of integrin-activated FAK via Src Homology 2 (SH2) domain [83]. Notably, intraepithelial levels of EGFR and FAK significantly correlate with progression to overt carcinoma, suggesting their potential use in the panel of predictive indicators for OSCC malignant

16 transformation [84,85]. As such, targeting both FAK and EGFR may curtail migratory and invasive phenotypes of premalignant and malignant cells. Our studies have demonstrated that fenretinide is capable of interfering with both FAK (See Chapter 2) and EGFR-Src signaling (See Chapter 3). Functional analyses of using various cell migration and invasion assays demonstrated that fenretinide indeed inhibited cell migration and invasion in a continuum of oral keratinocytes ranging from normal-to- immortalized-to-malignant. As the majority of cancer mortality results from metastasis, chemopreventive strategy to inhibit cell migration and invasion could greatly improve clinical outcomes by reducing not only primary tumor occurrence but also potential metastatic lesions.

1.4.d. Inflammation, Angiogenesis, and Tumor Microenvironment

For many years, research focused on epithelial dysfunction as the main culprit of carcinoma progression. Emerging evidence, however, emphasizes the crucial role that tumor microenvironment plays in the carcinogenic process [86]. Interactions between keratinocytes and stromal components (e.g. fibroblasts, endothelial cells, and immune cells) begin well before the breach of the basement membrane [86]. This crosstalk prepares the ideal “soil” for invading cells and promote survival by inducing angiogenesis and immune tolerance [86]. Once successfully invaded, cancer cells are nestled in the tumor microenvironment that sustains continued tumor growth and facilitates further invasion/metastasis to distant sites. Chemopreventive strategies should, therefore, target not only epithelial components (e.g. differentiation and apoptosis) but 17 also various cells involved in creating pro-tumorigenic microenvironment by sustaining inflammation and angiogenesis (e.g. macrophages, neutrophils, endothelial cells, etc).

Challenges remain even if chemopreventive therapies can control stromal cells from supplying cancer cells with new vascular beds, growth signals, and escape routes from immune surveillance. The issue of vasculogenic mimicry, another form of cell plasticity related to EMT, illustrates this difficulty [87]. Actively proliferating population of cells require continual supply of nutrients to support their growth. Cancer cells typically accomplish this by recruiting endothelial cells (angiogenesis) and progenitor endothelial cells (vasculogenesis) to establish new vasculature. VM-capable cancer cells, however, can trigger a phenotypic switch to assume certain endothelial characteristics and form vascular structures without the aid of endothelial cells. In fact, head and neck squamous cell carcinoma (HNSCC) cells display endotheliod functional features (e.g. low density lipoprotein uptake, tube formation, VEGF responsiveness) and facilitate angiogenesis and cell invasion (See Appendix A). As cancer cells can develop compensatory mechanisms, such as vasculogenic mimicry, one way to circumvent this issue is to devise a chemopreventive strategy which takes advantage of common molecular targets present in both epithelium and stroma. Albini and Sporn list key transcriptional factors, such as signal transducer and activators (STATs) and nuclear factor kappaB (NF-kB), as potential chemopreventive targets in the tumor microenvironment due to their association with inflammation and angiogenesis [86].

Previous studies demonstrate that many premalignant and malignant oral squamous cell carcinoma cells express constitutive activation of JAK/STAT3 and NF-kB pathways [88].

As such, JAK/STAT3 and NF-kB may serve as promising chemopreventive targets (See

Chapter 3).

1.5. Delivery Strategy for Chemoprevention of Oral Cancer: Local Delivery

Delivery strategies for chemotherapeutic and chemopreventive agents can largely be divided into systemic and local delivery [74]. Systemic administration of chemotherapeutic agents has remained a mainstay for chemotherapy since its inception.

Many cancers, which are either located in inaccessible locations (e.g. liver) or frequently metastatic (e.g. lung, breast), warrant systemic delivery of chemotherapeutic agents.

Furthermore, studies demonstrated that approximately 30% of primary breast cancer patients without any signs of overt metastases possessed disseminated tumor cells at the time of diagnosis, further supporting the use of systemic chemotherapeutic agents for advanced stage cancers [89].

The same rationale (i.e. inaccessible locations) necessitates the use of systemic delivery for chemoprevention of certain cancers (e.g. breast, prostate, and lung).

Successful clinical trials have led to Food and Drug Administration (FDA) approval of several chemopreventive drugs (e.g. tamoxifen and raloxifene). The results, however, accompanied moderate toxicity concerns as systemic delivery usually results in exposing non-target cells/organs to the compound of choice [44].

Systemic delivery becomes no longer the only feasible option for cancers whose involved sites exhibit greater accessibility (e.g. oral, skin, cervical). The oral cavity is

19 particularly well-suited for local delivery of chemopreventive compounds as most oral premalignant lesions appear as well-defined, easily identified white, red, or mixed-color lesions. The ease of identification and visualization facilitate biopsy and application of chemopreventive agents. Despite the advantages of local delivery strategy for oral cancer chemoprevention, several clinical trials maintained the standard systemic delivery model

[74]. These clinical trials, which used vitamin A derivatives, Cox-2 inhibitors, and natural products (e.g. green tea extract), achieved only modest clinical efficacy [74]. Holpuch et al attributed these limited clinical outcomes to inefficient delivery of chemopreventive compounds to the target sites [74]. In addition, trial design suffered from inconsistent inclusion criteria and failure to determine intraepithelial concentrations of chemopreventive drugs, rendering the trial results difficult to decipher [74]. Finally, prolonged administration of systemic chemopreventives often accompanied dose-limiting systemic toxicities resulting in the early termination of the clinical trials [74].

Systemic delivery of oral cancer chemopreventive requires the drug to enter into circulation, travel to the target tissue, perfuse through the stroma, and be absorbed by the target epithelial cells. In order to achieve this, chemopreventive drugs will need to retain sufficient concentrations to discount metabolic inactivation and excretion which often occurs with systemically absorbed drugs. In contrast, locally delivered compounds bypass the first pass liver metabolism and are deposited directly into the target tissue. Although similar issues regarding systemic delivery clinical trials are detected in local delivery clinical trials, fewer toxicities were associated with the clinical trials investigating local delivery of chemopreventive compounds (e.g. isotretinoin, fenretinide). Notably, our

20 laboratories conducted Phase I and II oral cancer chemoprevention studies to evaluate safety and efficacy of a mucoadhesive freeze-dried black raspberry (BRB) gel in healthy volunteers and in patients with histologically confirmed premalignant oral epithelial lesions [90-94]. Phase I clinical trial results show that the locally delivered BRB gel was well tolerated and void of any deleterious side effects. Phase II studies showed the BRB gel significantly reduced lesional sizes, histologic grades, and loss of heterozygosity

(LOH) indices unlike the placebo gel group which showed an increase in lesional size and no change associated with histologic grades and LOH indices [90]. Furthermore, the pharmacokinetic study conducted in our lab using the rabbit model demonstrated that a mucoadhesive fenretinide patch is capable of delivering therapeutically relevant fenretinide concentrations (~5uM) to the target sites without systemic exposure and toxicities (See Appendix B) [61]. Collectively, local delivery provides a pharmacologic advantage over systemic delivery for chemoprevention of oral cancer.

1.6. Fenretinide

Vitamin A plays an essential role in a variety of physiologic functions including growth and development, vision, and modulation of the immune system [95]. For example, enzymatic recycling of 11-cis retinal, a natural vitamin A metabolite, enables vision in low-light conditions by forming rhodopsin [96]. All-trans retinoic acid, another metabolite of vitamin A, is capable of inducing cell differentiation and modulating the adaptive immune response by controlling proliferation rates and cytokine releases of B and T lymphocytes [97]. As cancer often involves de-differentiation from the terminal 21 state in cellular lineage and acquisition of stem-cell like characteristics, scientists have recognized and investigated therapeutic potentials of vitamin A metabolites and their synthetic analogs. As a result, vitamin A derivatives are currently used as treatments for various dermatologic diseases (e.g. psoriasis and acnes) and as cancer chemopreventives

(e.g. breast cancer and acute promyelocytic leukemia).

Intense research efforts have been undertaken over the years to assess clinical efficacy of various retinoids. Despite their promising therapeutic effects, toxicity emerged as the key concern for clinical use of retinoids [95]. Notably, retinyl acetate was found to accumulate and cause severe liver toxicity [98]. In an attempt to design a retinoid with comparable therapeutic parameters with low toxicity profile, fenretinide, a synthetic analog of all trans retinoic acid, was first reported by Gander and Gurney by their 1978 patent [99]. Moon et al synthesized fenretinide by substituting the carboxylic acid group of all trans retinoic acid with hydroxyphenyl attached to an amid group [98].

This pharmacokinetic study also demonstrated that prolonged p.o. administration of fenretinide (up to 6 months) resulted in accumulation in rat breast tissue without causing overt toxicity. In contrast, up to 50% of the retinyl acetate treated rats experienced severe hepatotoxicity and perished within 2 weeks [98]. Notably, fenretinide retained antiproliferative and chemopreventive effects in the rat breast cancer model, albeit less potent than the retinyl acetate [98].

In addition to the cell differentiation-inducing effects associated with retinoids, fenretinide possesses capacity to induce apoptosis [100, 101]. Two main pathways mediate apoptosis: (1) extrinsic pathway via death receptors (e.g. Fas receptors) and (2)

22 intrinsic pathway via mitochondria [101]. Mechanistic studies demonstrated that fenretinide exerts its pro-apoptotic effects by disrupting mitochondrial membrane potential (MMP) and causing release of cytochrome c and pro-apoptotic proteins (e.g.

Bak and Bax) [100, 102]. Fenretinide-induced apoptosis does not involve p53, another key player in cell cycle regulation and apoptosis, as p53 inhibtion does not abrogate fenretinide-induced apoptosis [100].

Recent studies have demonstrated that fenretinide’s chemopreventive effects may be extended beyond apoptosis and terminal differentiation. Ferrari et al demonstrated that fenretinide suppressed migration of Kaposi’s sarcoma and exerted anti-proliferative/antiangiogenic effects [104]. Additional studies showed that fenretinide inhibited migration and invasion of ovarian and prostate cancer cells, implicating the focal adhesion kinase

(FAK) signaling pathway as fenretinide’s potential molecular target [105,106]. Notably, our laboratories recently demonstrated that fenretinide is capable of inhibiting oral squamous cell carcinoma (OSCC) cell migration and invasion by competitively binding to FAK kinase and FERM (4.1 protein, ezrin, radixin, moesin) domain (See Chapter 2)

[103].

Collectively, fenretinide represents a promising chemopreventive that warrants further investigation. Previous oral cancer clinical trials using fenretinide produced only modest results due to fenretinide’s low bioavailability and inappropriate selection of delivery mechanisms [74]. Specific targeting of premalignant oral epithelial lesions utilizing local delivery strategy, such as, mucoadhesive patches (See Appendix B), would facilitate distribution of therapeutically relevant drug concentrations in the target

23 tissue [61]. Notably, our Phase I/II chemoprevention trials revealed “high” responders to freeze-dried black raspberry (BRB) gel [90-94]. Further investigation demonstrated great interpatient heterogeneity in oral metabolism of BRB gel [91]. In addition, considerable interpatient differences exists for the enzyme (i.e. CYP3A4) responsible for bioactivating fenretinide (See Appendix B) [61]. These findings support accumulating evidence to implement personalized approach to medicine. Fenretinide, therefore, would likely achieve the greatest therapeutic benefits when administered in local delivery format with the surrounding field coverage to patients with appropriate metabolic profiles. In addition, a chemopreventive cocktail containing fenretinide and other chemopreventive compounds should be considered as it is unlikely that a single chemopreventive agent would sufficiently manage all oral epithelial dysplasia. Studies to investigate clinical efficacy of such combination treatment (i.e. fenretinide and BRB) are ongoing in our laboratories.

1.7. Bibliography 1. Menzin J, Lines LM, Manning LN. The economics of squamous cell carcinoma of the head and neck. Curr Opin Otolaryngol Head Neck Surg. 15: 68-73, 2007. PMID: 17413405 2. Lefebvre JL. Current clinical outcomes demand new treatment options for SCCHN. Ann Oncol. 16: vi7-vi12, 2005. PMID: 15987995 3. Lang K, Menzin J, Earle CC, Jacobson J, Hsu M-A. The Economic Cost of Squamous Cell Cancer of the Head and Neck. Findings from Linked SEER-Medicare Data. Arch Otolaryngol Head Neck Surg. 130: 1269-1275, 2004. PMID: 15545580 4. Siegel RL, Miller KD, Jemal A. Cancer Statistics, 2015. CA Cancer J Clin. 2015, 65:5-29. 5. Neville BW, Day TA. Oral Cancer and Precancerous Lesions. CA Cancer J Clin. 2002; 52: 195- 215.

6. Tanaka T, Tanaka M, Tanaka T. Oral carcinogenesis and oral cancer chemopreventive: a review. Patholog Res Int. 2011; 22;2011:431246. 7. Leemans CR, Boudewijn JMB, Brakenhoff RH. The molecular biology of head and neck cancer. Nat Rev Cancer. 2011; 11(1):9-22. 8. Wong DT, Todd R, Tsuji T, Donoff RB. Molecular biology of human oral cancer. Crit Rev Oral Biol Med. 1996; 7(4):319-28. 9. Benson E, Li R, Elsele D, Fakhry C. The clinical impact of HPV tumor status upon head and neck squamous cell carcinomas. Oral Oncol. 2014; 50(6): 565-74. 10. Munger K, Scheffner M, Huibregtse JM, Howley PM. Interactions of HPV E6 and E7 oncoproteins with tumour suppressor gene products. Cancer Surv. 1992; 12:197-217. 11. Wiest T, Schwarz E, Enders C, Flechtenmacher C, Bosch FX. Involvement of intact HPV16 E6/E7 gene expression in head and neck cancers with unaltered p53 status and perturbed pRb cell cycle control. Oncogene. 2002; 21(10):1510-17. 12. Sugimura T, Kumimoto H, Tohnai I, Fukui T, Matsuo K, Tsurusako S, et al. Gene-environment interaction involved in oral carcinogenesis: molecular epidemiological study for metabolic and DNA repair gene polymorphisms. J Oral Pathol Med. 2006; 35(1):11-18. 13. Chang KW, Lee TC, Yeh WI, Chung MY, Liu CJ, Chi LY, et al. Polymorphism in heme oxygenase-1 (HO-1) promoter is related to the risk of oral squamous cell carcinoma occurring on male areca chewers. Br J Cancer. 2004; 91(8):1551-5. 14. Whitmore S, Lamont RJ. Oral Bacteria and Cancer. PLoS Pathog. 2014; 10(3):e1003933. 15. Slaughter DP, Southwick HW, Smejkal W. Field cancerization in oral stratified squamous epithelium. Cancer (Phila.). 1953; 6:963-68. 16. Scholes AGM, Woolgar JA, Boyle MA, Brown JS, Vaughan ED, Hart CA, et al. Synchronous oral carcinomas: independent or common clonal origin? Cancer Res. 1998; 58:2003-6. 17. Tabor MP, Brakenhoff RH, van Houten VMM, Kummer JA, Snel MHJ, Snijders PJF. Persistence of genetically altered fields in head and neck cancer patients: biological and clinical implications. Clin Cancer Res. 2001; 7(6):1523-32. 18. Braakhuis BJM, Tabor MP, Kummer A, Leemans CR, Brakenhoff RH. A genetic explanation of Slaughter’s concept of field cancerization: evidence and clinical implications. Cancer Res. 2003; 63:1727-1730. 19. Chen Y, Chen C. DNA copy number variation and loss of heterzygosity in relation to recurrence of and survival from head and neck squamous cell carcinoma: a review. Head Neck. 2008; 30(10):1361-83. 20. Kuffer R, Lombardi T. Premalignant lesions of the oral mucosa. A discussion about the place of oral intraepithelial neoplasia (OIN). Oral Oncol. 2002; 38(2): 125-30. 21. Lane DP. Cancer. p53, guardian of the genome. Nature. 1992; 358(6381):15-6. 25

22. Rowley H, Sherrington P, Helliwell TR, Kinsella A, Jones AS. P53 expression and p53 gene mutation in oral cancer and dysplasia. Otolaryngol Head Neck Surg. 1998; 118(1):115-23. 23. Ha PK, Califano JA. Promoter methylation and inactivation or tumour-suppressor genes in oral squamous-cell carcinoma. Lancet Oncol. 2006; 7(1): 77-82. 24. Hanahan D, Weinberg RA. The hallmarks of cancer. Cell. 2000; 100(1):57-70. 25. Hanahan D, Weinberg RA. Hallmarks of cancer: the next generation. 2011; 144(5):646-74. 26. Lamouille S, Xu J, Derynck R. Molecular mechanisms of epithelial-mesenchymal transition. Nat Rev Mol Cell Biol. 2014; 15:178-196. 27. Thiery JP. Epithelial-mesenchymal transitions in tumour progression. Nat Rev Cancer. 2002; 2:442-54. 28. Cicchini C, Laudadio I, Citarella F, Corazzari M, Steindler C, Conigliaro A, et al. TGFbeta- induced EMT requires focal adhesion kinase (FAK) signaling. Exp Cell Res. 2008; 314(1):143-52. 29. Silverman Jr S, Gorsky M, Lozada F. Oral leukoplakia and malignant transformation. A follow- up study of 257 patients. Cancer 1984;53:563-69. 30. Lee JJ, Honk WK, Hittelman WN, Mao L, Lotan R, Shin DM, Benner SE, et al. Prediciting cancer development in oral leukoplakia: ten years in translational research. Clin Cancer Res, 2000; 6:1702-1210. 31. Lippman SM, Hong WK. Molecular markers of the risk of oral cancer. N Engl J Med. 2001; 344(17):1323-26. 32. Von Zeidler SV, Botelho TdS, Mendonca EF, Batista AC. E-cadherin as a potential biomarker of malignant transformation in oral leukoplakia: a retrospective cohort study. BMC Cancer. 2014; 14:972. 33. Reibel J. Prognosis of oral pre-malignant lesions: significance of clinical, histopathological, and molecular biological characteristics. Crit Rev Oral Biol Med. 2003; 14(1):47-62. 34. Tsui IF, Rosin MP, Zhang L, Ng RT, Lam WL. Multiple aberrations of chromosome 3p detected in oral premalignant lesions. Cancer Prev Res (Phila). 2008; 1(6):424-9. 35. Galipeau PC, Prevo LJ, Sanchez CA. Clonal expansion and loss of heterozygosity at chromsomes 9p and 17p in premaligant esophageal (Barrett’s) tissue. J Natl Cancer Inst. 1999; 91(24):2087- 95. 36. Guillaud M, Zhang L, Poh C, Rosin MP, MacAulay C. Potential use of quantitative tissue phenotype to predict malignant risk for oral premalignant lesions. Cancer Res. 2008; 68(9):3099- 3107. 37. van Zyl AW, van Heerden MB, Langenegger E, van Heerden WFP. Correlation between dysplasia and ploidy status in oral leukoplakia. Head Neck Pathol. 2012; 6(3):322-327.

38. Sperandio M, Brown Al, Lock C, Morgan PR, Coupland, VH, Madden PB, et al. Predictive value of dysplasia grading and DNA ploidy in malignant transformation or oral potentially malignant disorders. Cancer Pre Res (Phila). 2013; 6(8): 822-31. 39. Hemmer J, Kreidler J. Flow cytometric DNA ploidy analysis of squamous cell carcinoma of the oral cavity. Comparison with clinical staging and histologic grading. Cancer. 1990; 66(2): 317- 20. 40. Casas-Selves M, DeGregor J. How cancer shapes evolution, and how evolution shapes cancer. Evolution (N Y). 2011; 4(4): 624-634. 41. Mallery SR, Budendorf DE, Larsen MP, Pei P, Tong M, Holpuch AS, Larsen PE, Stoner GD, Fields HW, Chan KK, Ling Y, and Liu Z. Effects of Human Oral Mucosal Tissue, Saliva, and Oral Microflora on Intraoral Metabolism and Bioactivation of Black Raspberry Anthocyanins. Cancer Prev Res. 2011, 4(8): 1209-1221. 42. Watson IR, Takahashi K, Futreal PA, Chin L. Emerging patterns of somatic mutations in cancer. Nat Rev Genet. 2013 Oct;14(10):703-18. 43. Sporn MB, Dunlop NM, Newton DL, Smith JM. Fed Proc. Prevention of chemical carcinogenesis by vitamin A and its synthetic analogs (retinoids). 1976 May 1;35(6):1332-8. 44. Steward WP, Brown K. Cancer chemoprevention: a rapidly evolving field. Br J Cancer. 2013 Jul 9;109(1):1-7. 45. DeVita VT Jr, Chu E. A history of cancer chemotherapy. Cancer Res. 2008 Nov 1;68(21):8643-53 46. Howlader N, Noone AM, Krapcho M, Garshell J, Miller D, Altekruse SF, et al. SEER Cancer Statistics Review, 1975-2011, National Cancer Institute. Bethesda, MD, http://seer.cancer.gov/csr/1975_2011/, based on November 2013 SEER data submission, posted to the SEER website, April 2014. Updated September 10, 2014. 47. Cotter TG. Nat Rev Cancer. Nat Rev Cancer. 2009 Jul;9(7):501-7. 2009 Jul;9(7):501-7. 48. Chimal-Monroy J, Abarca-Buis RF, Cuervo R, Díaz-Hernández M, Bustamante M, Rios-Flores JA, et al. Molecular control of cell differentiation and programmed cell death during digit development. IUBMB Life. 2011 Oct;63(10):922-9. 49. Warren HS, Smyth MJ. NK cells and apoptosis. Immunol Cell Biol. 1999 Feb;77(1):64-75. 50. Fridman JS, Lowe SW. Control of apoptosis by p53. Oncogene. 2003 Dec 8;22(56):9030-40. 51. Schoelch ML, Regezi JA, Dekker NP, Ng IO, McMillan A, Ziober BL,et al.Cell cycle proteins and the development of oral squamous cell carcinoma. Oral Oncol. 1999 May;35(3):333-42. 52. Crew KD, Gammon MD, Terry MB, Zhang FF, Agrawal M, Eng SM, et al. Genetic polymorphisms in the apoptosis-associated genes FAS and FASL and breast cancer risk. Carcinogenesis. 2007 Dec;28(12):2548-51. Epub 2007 Oct 24.

53. Yamazaki Y, Chiba I, Hirai A, Sugiura C, Notani K, Kashiwazaki H, Tei K, Totsuka Y, Fukuda H. Specific p53 mutations predict poor prognosis in oral squamous cell carcinoma. Oral Oncol. 2003 Feb;39(2):163-9. 54. Fulda S, Debatin KM. Extrinsic versus intrinsic apoptosis pathways in anticancer chemotherapy. Oncogene. 2006 Aug 7;25(34):4798-811. 55. Surh YJ. Cancer chemoprevention with dietary phytochemicals. Nat Rev Cancer.2003 Oct;3(10):768-80. Review. PubMed PMID: 14570043. 56. Gurudath S, Ganapathy K, D S, Pai A, Ballal S, Ml A. Estimation of superoxide dismutase and glutathione peroxidase in oral submucous fibrosis, oral leukoplakiaand oral cancer--a comparative study. Asian Pac J Cancer Prev.2012;13(9):4409-12. PubMed PMID: 23167351. 57. Bose SC, Singh M, Vyas P, Singh M. Plasma zinc antioxidant vitamins, glutathione levels and total antioxidant activity in oral leukoplakia. Dent Res J(Isfahan). 2012 Mar;9(2):158-61. 58. Rodrigo KA, Rawal Y, Renner RJ, Schwartz SJ, Tian Q, Larsen PE, Mallery SR. Suppression of the tumorigenic phenotype in human oral squamous cell carcinoma cells by an ethanol extract derived from freeze-dried black raspberries. Nutr Cancer. 2006;54(1):58-68. 59. Brusselmans K, De Schrijver E, Heyns W, Verhoeven G, Swinnen JV.Epigallocatechin-3-gallate is a potent natural inhibitor of fatty acid synthase in intact cells and selectively induces apoptosis in prostate cancer cells. Int J Cancer. 2003 Oct 10;106(6):856-62. 60. Shankar S, Ganapathy S, Hingorani SR, Srivastava RK. EGCG inhibits growth,invasion, angiogenesis and metastasis of pancreatic cancer. Front Biosci. 2008 Jan 1;13:440-52. 61. Holpuch AS, Phelps MP, Desai KG, Chen W, Koutras GM, Han BB, et al. Evaluation of a mucoadhesive fenretinide patch for local intraoral delivery: a strategy to reintroduce fenretinide for oral cancer chemoprevention. Carcinogenesis. 2012 May;33(5):1098-105. 62. Lippens S, Denecker G, Ovaere P, Vandenabeele P, Declercq W. Death penalty forkeratinocytes: apoptosis versus cornification. Cell Death Differ. 2005 Nov;12 Suppl 2:1497-508. Review. PubMed PMID: 16247497. 63. Watt FM, Kubler MD, Hotchin NA, Nicholson LJ, Adams JC. Regulation of keratinocyte terminal differentiation by integrin-extracellular matrix interactions. J Cell Sci. 1993 Sep;106 ( Pt 1):175- 82 64. Calautti E, Li J, Saoncella S, Brissette JL, Goetinck PF. Phosphoinositide 3-kinase signaling to Akt promotes keratinocyte differentiation versus death. J Biol Chem. 2005 Sep 3;280(38):32856- 65. 65. Eckert RL, Efimova T, Dashti SR, Balasubramanian S, Deucher A, Crish JF, Sturniolo M, Bone F. Keratinocyte survival, differentiation, and death: many roads lead to mitogen-activated protein kinase. J Investig Dermatol Symp Proc. 2002 Dec;7(1):36-40.

66. Macha MA, Matta A, Kaur J, Chauhan SS, Thakar A, Shukla NK, Gupta SD, Ralhan R. Prognostic significance of nuclear pSTAT3 in oral cancer. Head Neck. 2011 Apr;33(4):482-9. 67. Watanabe S, Sato K, Okazaki Y, Tonogi M, Tanaka Y, Yamane GY. Activation of PI3K-AKT pathway in oral epithelial dysplasia and early cancer of tongue. Bull Tokyo Dent Coll. 2009 Aug;50(3):125-33. 68. Ribeiro DC, Gleber-Netto FO, Sousa SF, Bernardes VD, Guimarães-Abreu MH, Aguiar MC. Immunohistochemical expression of EGFR in oral leukoplakia: association with clinicopathological features and cellular proliferation. Med Oral Patol Oral Cir Bucal. 2012 Sep 1;17(5):e739-44. 69. Sayama K, Hanakawa Y, Shirakata Y, Yamasaki K, Sawada Y, Sun L, Yamanishi K, Ichijo H, Hashimoto K. Apoptosis signal-regulating kinase 1 (ASK1) is an intracellular inducer of keratinocyte differentiation. J Biol Chem. 2001 Jan 12;276(2):999-1004. 70. Mitra SK, Hanson DA, Schlaepfer DD. Focal adhesion kinase: in command and control of cell motility. Nat Rev Mol Cell Biol. 2005 Jan;6(1):56-68. 71. Ishikawa F, Ushida K, Mori K, Shibanuma M. Loss of anchorage primarily induces non-apoptotic cell death in a human mammary epithelial cell line under atypical focal adhesion kinase signaling. Cell Death Dis. 2015 Jan 22. 72. Sporn MB, Roberts AB. Role of retinoids in differentiation and carcinogenesis. Cancer Res. 1983 Jul;43(7):3034-40. 73. Wolbach, S. B., and Howe, P. R. Tissue changes following deprivation of fat soluble A vitamin. J. Exp. Med., 42: 753-777,1925. 74. Holpuch AS, Desai KG, Schwendeman SP, Mallery SR. Optimizing therapeutic efficacy of chemopreventive agents: A critical review of delivery strategies in oral cancer chemoprevention clinical trials. J Carcinog. 2011;10:23. 75. Hansen LA, Sigman CC, Andreola F, Ross SA, Kelloff GJ, De Luca LM. Retinoids in chemoprevention and differentiation therapy. Carcinogenesis. 2000 Jul;21(7):1271-9. 76. Chen S, Samuel W, Fariss RN, Duncan T, Kutty RK, Wiggert B. Differentiation of human retinal pigment epithelial cells into neuronal phenotype by N-(4-hydroxyphenyl)retinamide. J Neurochem. 2003 Mar;84(5):972-81. 77. Vernon E. Steele, Gary D. Stoner, Gary J. Kelloff, Charles W. Boone. Cellular and Molecular Targets for Chemoprevention. CRC Press. 1992. Textbook 78. Andreasen PA, Egelund R, Petersen HH. The plasminogen activation system in tumor growth, invasion, and metastasis. Cell Mol Life Sci. 2000 Jan 20;57(1):25-40. 79. Joyce JA, Pollard JW. Microenvironmental regulation of metastasis. Nat Rev Cancer. 2009 Apr;9(4):239-52. doi: 10.1038/nrc2618. Epub 2008 Mar 12.

80. Chan KT, Cortesio CL, Huttenlocher A. FAK alters invadopodia and focal adhesion composition and dynamics to regulate breast cancer invasion. J Cell Biol. 2009 Apr 20;185(2):357-70. doi: 10.1083/jcb.200809110. Epub 2009 Apr 13. 81. Smirnova T, Adomako A, Locker J, Van Rooijen N, Prystowsky MB, Segall JE. In vivo invasion of head and neck squamous cell carcinoma cells does not require macrophages. Am J Pathol. 2011 Jun;178(6):2857-65. 82. Hwang YS, Park KK, Chung WY. Invadopodia formation in oral squamous cell carcinoma: the role of epidermal growth factor receptor signalling. Arch Oral Biol. 2012 Apr;57(4):335-43. 83. Ray RM, Li C, Bhattacharya S, Naren AP, Johnson LR. Spermine, a molecular switch regulating EGFR, integrin β3, Src, and FAK scaffolding. Cell Signal. 2012, Apr;24(4):931-42. 84. de Vicente JC, Rodrigo JP, Rodriguez-Santamarta T, Lequerica-Fernández P, Allonca E, García- Pedrero JM. Cortactin and focal adhesion kinase as predictors of cancer risk in patients with premalignant oral epithelial lesions. Oral Oncol. 2012 Jul;48(7):641-6. 85. Ries J, Vairaktaris E, Agaimy A, Bechtold M, Gorecki P, Neukam FW, Nkenke E. The relevance of EGFR overexpression for the prediction of the malignant transformation of oral leukoplakia. Oncol Rep. 2013 Sep;30(3):1149-56. 86. Albini A, Sporn MB. The tumour microenvironment as a target for chemoprevention. Nat Rev Cancer. 2007 Feb;7(2):139-47. Review. PubMed PMID: 17218951. 87. Fan YL, Zheng M, Tang YL, Liang XH. A new perspective of vasculogenic mimicry: EMT and cancer stem cells (Review). Oncol Lett. 2013 Nov;6(5):1174-1180. 88. Squarize CH, Castilho RM, Sriuranpong V, Pinto DS Jr, Gutkind JS. Molecular cross-talk between the NFkappaB and STAT3 signaling pathways in head and neck squamous cell carcinoma. Neoplasia. 2006 Sep;8(9):733-46. 89. Pantel K, Alix-Panabières C, Riethdorf S. Cancer micrometastases. Nat Rev Clin Oncol. 2009 Jun;6(6):339-51. 90. Mallery SR, Tong M, Shumway BS, Curran AE, Larsen PE, Ness GM, et al. Topical application of a mucoadhesive freeze-dried black raspberry gel induces clinical and histologic regression and reduces loss of heterozygosity events in premalignant oral intraepithelial lesions: results from a multicentered, placebo-controlled clinical trial. Clin Cancer Res. 2014 Apr 1;20(7):1910-24. 91. Mallery SR, Budendorf DE, Larsen MP, Pei P, Tong M, Holpuch AS, et al. Effects of human oral mucosal tissue, saliva, and oral microflora on intraoral metabolism and bioactivation of black raspberry anthocyanins. Cancer Prev Res (Phila). 2011 Aug;4(8):1209-21. 92. Ugalde CM, Liu Z, Ren C, Chan KK, Rodrigo KA, Ling Y, et al. Distribution of anthocyanins delivered from a bioadhesive black raspberry gel following topical intraoral application in normal healthy volunteers. Pharm Res. 2009 Apr;26(4):977-86.

93. Mallery SR, Zwick JC, Pei P, Tong M, Larsen PE, Shumway BS, et al. Topical application of a bioadhesive black raspberry gel modulates gene expression and reduces cyclooxygenase 2 protein in human premalignant oral lesions. Cancer Res. 2008 Jun 15;68(12):4945-57. 94. Shumway BS, Kresty LA, Larsen PE, Zwick JC, Lu B, Fields HW, et al. Effects of a topically applied bioadhesive berry gel on loss of heterozygosity indices in premalignant oral lesions. Clin Cancer Res. 2008 Apr 15;14(8):2421-30. 95. Formelli F, Barua AB, Olson JA. Bioactivities of N-(4-hydroxyphenyl)retinamide and retinoyl beta-glucuronide. FASEB J. 1996 Jul;10(9):1014-24. 96. Miyazono S, Isayama T, Delori FC, Makino CL. Vitamin A activates rhodopsin and sensitizes it to ultraviolet light. Vis Neurosci. 2011 Nov;28(6):485-97. 97. Mora JR, Iwata M, von Andrian UH. Vitamin effects on the immune system: vitamins A and D take centre stage. Nat Rev Immunol. 2008 Sep;8(9):685-98. 98. Moon RC, Thompson HJ, Becci PJ, Grubbs CJ, Gander RJ, Newton DL, Smith JM, Phillips SL, Henderson WR, Mullen LT, Brown CC, Sporn MB. N-(4-Hydroxyphenyl)retinamide, a new retinoid for prevention of breast cancer in the rat. Cancer Res. 1979 Apr;39(4):1339-46. 99. Gander, RJ, and Gurney, JA. All-trans relinoic acid esters and amides. U.S. patent 4,108,880, August 22, 1978;Chem. Abstr. 88,89892 100. Boya P, Morales MC, Gonzalez-Polo RA, Andreau K, Gourdier I, Perfettini JL, et al. The chemopreventive agent N-(4-hydroxyphenyl)retinamide induces apoptosis through a mitochondrial pathway regulated by proteins from the Bcl-2 family. Oncogene. 2003 Sep 18;22(40):6220-30. 101. Hail Jr N, Kin HJ, Lotan R. Mechanisms of fenretinide-induced apoptosis. Apoptosis 11: 1677-1694, 2006. 102. Ulukaya E, Pirianov G, Kurt MA, Wood EJ, Mehmet H. Fenretinide induces cytochrome c release, caspase 9 activation and apoptosis in the absence of mitochondrial membrane depolarisation. Cell Death Differ. 2003 Jul;10(7):856-9. 103. Han BB, Li S, Tong M, Holpuch AS, Spinney R, Wang D, et al. Fenretinide Perturbs Focal Adhesion Kinase in Premalignant and Malignant Human Oral Keratinocytes. Fenretinide's chemopreventive mechanisms include ECM interactions. Cancer Prev Res (Phila). 2015 Feb 24. pii: canprevres.0418.2014. [Epub ahead of print] PubMed PMID: 25712051. 104. Ferrari N, Morini M, Pfeffer U, Minghelli S, Noonan DM, Albini A. Inhibition of Kaposi’s Sarcoma in Vivo by Fenretinide. Clin Cancer Res 2003; 9: 6020-6029. 105. Golubkov V, Garcia A, Markland FS. Action of Fenretinide (4-HPR) on Ovarian Cancer and Endothelial Cells. Anticancer Res 2005; 25: 249-253.

106. Benelli R, Monteghirfo S, Vene R, Tosetti F, Ferrari N. The chemopreventive retinoid 4HPR impairs prostate cancer cell migration and invasion by interfering with FAK/AKT/GSK3β pathway and β-catenin stability. Mol Cancer 2010; 9: 1-13.

Chapter 2: Fenretinide Perturbs Focal Adhesion Kinase in Premalignant and

Malignant Human Oral Keratinocytes. Fenretinide’s chemopreventive mechanisms

include ECM interactions.

2.1. Abstract

The membrane-associated protein, focal adhesion kinase (FAK), modulates cell- extracellular matrix interactions and also conveys pro-survival and proliferative signals.

Notably, increased intraepithelial FAK levels accompany transformation of premalignant oral intraepithelial neoplasia (OIN) to oral squamous cell carcinoma (OSCC). OIN chemoprevention is a patient-centric, optimal strategy to prevent OSCC’s co-morbidities and mortality. The cancer chemopreventive and synthetic vitamin A derivative, fenretinide, has demonstrated protein-binding capacities e.g. mTOR and retinol binding protein interactions. These studies employed a continuum of human oral keratinocytes

(normal-HPV E6/E7-transduced-OSCC) to assess potential fenretinide-FAK drug protein interactions and functional consequences on cellular growth regulation and motility.

Molecular modeling studies demonstrated fenretinide has ~200-fold greater binding affinity relative to the natural ligand (ATP) at FAK’s kinase domain. Fenretinide also shows intermediate binding at FAK’s FERM domain and interacts at the ATP-binding

33 site of the closest FAK analogue, Pyk2. Fenretinide significantly suppressed proliferation via induction of apoptosis and G2/M cell cycle blockade. Fenretinide-treated cells also demonstrated F-actin disruption, significant inhibition of both directed migration and invasion of a synthetic basement membrane, and decreased phosphorylation of growth- promoting kinases. A commercially available FAK inhibitor did not suppress cell invasion. Notably, while FAK’s FERM domain directs cell invasion, FAK inhibitors target the kinase domain. In addition, FAK-specific siRNA treated cells showed an intermediate cell migration capacity; data which suggest co-contribution of the established migrating-enhancing Pyk2. Our data imply that fenretinide is uniquely capable of disrupting FAK’s and Pyk2’s pro-survival and mobility-enhancing effects and further extend fenretinide’s chemopreventive contributions beyond induction of apoptosis and differentiation.

2.2. Introduction

Focal adhesion kinase (FAK) was originally identified as a Src oncogene substrate [1]. FAK is now known to also be activated by the SRC family kinases i.e.

PLC, SOCS, GRB7, PI3K as well as bioactivated lipids such as lysophosphatidic acid

[2]. In its mechanosensor capacity, FAK mediates cytoskeletal adaptations in response to extracellular matrix (ECM) interactions, regulates formation of cell membrane protrusions e.g. actin and matrix metalloproteinase-rich, extracellular matrix-degrading invadopodia and ultimately directs cell migration and invasion [1]. In a related role,

FAK’s functions also extend to translocation of lipid raft components to the leading edge 34 of motile cells thereby enabling the microtubule-cortical receptor stabilization that is essential for directed cell movement [1]. Furthermore, via its FERM domain, the membrane-spanning protein FAK serves as a chemosensor that links membrane-bound growth factor receptors such as EGFR and PDGFR, provides receptor cross-talk and ultimately signal transduction to the nucleus [3]. FAK’s FERM domain also directs FAK nuclear translocation enabling FAK-mediated p53 degradation and resultant increased cell survival and proliferation [4]. These abilities to promote cell survival/proliferation/angiogenesis while concurrently modulating ECM interactions and assisting invadopodia formation, make FAK an attractive cancer prevention therapeutic target [1, 5]. Notably, premalignant lesions that arise at visibly accessible sites, such as the mouth, are particularly well-suited for chemoprevention as treatment effects can be directly monitored.

Oral squamous cell carcinoma (OSCC) is a worldwide health problem that conveys significant socioeconomic impact [6]. Analogous to other surface origin cancers,

OSCCs arise from malignant transformation of a precursor lesion i.e. oral intraepithelial neoplasia (OIN i.e. a white, red or mixed adherent lesion that possesses microscopically confirmed cytologic and maturational perturbations superior to the basement membrane).

The poor prognosis of higher stage OSCCs, co-morbidities associated with vital tissue loss during surgical treatment, and visually accessible premalignant lesions combine to make chemoprevention the optimal OSCC treatment strategy [7].

Vitamin A and its derivatives have been regarded as promising OSCC chemopreventive agents for many years [8]. More recently, the synthetic analogue of all- trans retinoic acid, fenretinide (4-HPR), gained attention due to its reduced toxicity profile and its strong proapoptotic and prodifferentiation effects [9, 10, 11]. Additional studies demonstrated that 4-HPR disrupted cytoskeletal networks and suppressed migration of Kaposi’s sarcoma, ovarian cancer, and endothelial cells [12, 13]. Another investigation showed 4-HPR inhibited directed migration and invasion of prostate cancer cells; findings speculated by the investigators to reflect disruption of the

FAK/AKT/GSK3β pathway and β-catenin stability [14].

This study investigated a spectrum of 4-HPR-FAK interactions including drug- protein interactions and functional consequences of these interactions on cellular growth state and motility. The final series of experiments introduced an additional chemopreventive shown in be clinically effective in OIN lesions i.e. freeze dried black raspberries (BRB) [15]. Concurrent 4-HPR + BRB administration provided additive invasion-inhibitory effects.

2.3. Materials and Methods

Cell Culture. OSCC cell lines CRL-2095, SCC-15 (American Type Cell Culture, human tongue primary tumor) and JSCC-1, JSCC-2 and JSCC-3 derived from human OSCC tumors of tonsil (JSCC-1), tongue (JSCC-2) and floor of mouth (JSCC3), a normal oral

36 keratinocyte cell strain [ScienCell, Carlsbad, CA] HOK3437, and two immortalized cell lines [HPV E6/E7 transduced normal oral keratinocytes (HOK3437 E6/E7) and ethanol- treated HPV E6/E7-transduced normal oral keratinocytes (EPI)]. [16]. were used. All immortalized cells were cultured in Advanced DMEM supplemented with 1X Glutamax and 5% heat-inactivated fetal bovine serum (GIBCO, Life Technologies Grand Island,

NY, “Complete Medium”) while normal oral keratinocytes were cultured in Keratinocyte

Sera Free Medium + supplements (Gibco). Cells were cultured in a sera or growth factor free “Base” medium for chemoattractant-based experiments. Cell lines were authenticated by genomic analyses conducted by John Hopkins’ Genetic Resources Core

Facility. With the exception of the HOK3437 E6/E7 strain and the EPI cell line (both transduced with HPV16 E6 and E7) cell lines were negative for HPV, as determined by

PCR.

Cell Line Characterization. Formalin fixed cells were incubated with vimentin (1:200,

Abcam, Cambridge,MA) or a pancytokeratin cocktail (AE1/AE3 + 5D3, 1:100, Abcam,) antibodies, followed by incubation with FITC or Texas Red conjugated secondary antibodies (Abcam, Cambridge, MA) [17]. Nuclei were stained with 4’,6’-Diaminidino-

2-phenylindole dihydrochloride (DAPI, Abcam). Fluorescence microscopy images were obtained by using an Olympus BX51 microscope (Olympus, Japan), NikonDS-Fi1 digital camera (Nikon, Japan) and ImagePro 6.0 (Media-Cybernetics, Bethesda, MD).

Immunoblot analyses were conducted to determine presence or absence of 4-HPR metabolizing enzymes (CYPs 3A4, 2C8, 26A1) and UDP glucuronosyl transferase 1A1

(UGT1A1) in accordance with our previously published method [18]. Additional characterization studies entailed a time-course assessment of intracellular levels of 4-

HPR during 4-HPR treatment with concurrent 4-HPR medium evaluation using LC-

MS/MS analyses as previously described [11].

4-HPR’s induction of the execution phase of apoptosis. Cultured cells were treated with

1, 5, 10uM 4-HPR (0.1% DMSO, control cells received DMSO only) for 24 hours.

Functional caspase 3 and 7 activities were determined by Caspase-Glo® 3/7 Assay

(Promega, Madison, WI) according to the manufacturer’s protocol. Concurrent studies evaluated the effects of 4-HPR treatment on cell proliferation (CyQuant Assay,

Invitrogen, Carlsbad, CA). Complementary FACS analyses, which used propridum iodide labelled DNA, were conducted to identify cell cycle distribution during 4-HPR challenge.

Immunocytochemical characterization of 4-HPR’s effects on F-actin and microtubules. Adherent cells were wounded with a sterile pipette tip, washed with PBS, followed by 5uM or 10uM 4-HPR for 24 hours in Complete Medium. Post-treatment, cells were extracted with 0.1% Triton X-100/PBS and fixed with 4% paraformaldehyde, permeabilized, blocked, and probed with Alexa Fluoro 488-conjugated phalloidin

(Invitrogen, Carlsbad, CA) and DAPI (Vector Laboratories, Inc, Burlingame, CA). For co-localization studies, cells were first incubated with anti-tubulin antibody (1:500,

Abcam, Cambridge, MA) and its Texas Red conjugated secondary antibody (1:1000,

Abcam) for 1 hour at room temperature, and subsequently incubated with phalloidin and

DAPI. Fluorescence microscopy images were obtained by using an Apotome

Fluorescence Microscope (Carl Zeiss), and AxioVision software (Carl Zeiss).

Molecular Modeling of 4-HPR-FAK Interactions. Molecular modeling studies were conducted using AutoDock Vina software [19]. Initial 4-HPR binding studies used retinol binding protein (1BRP) [20] as a model binding protein.

A number of crystal structures exist for ligands bound to the FAK region of the protein. 2J0L was obtained and used for the AutoDock Vina binding study as it had the most complete structure (least disorder). An initial survey of the entire FAK protein surface for all ligands (retinol, ATP, an ATP-Mg complex, fenretinide) and ligands from the 2ETM and 2JKK crystal structures, as well as known agonists (PF228, TAE-226 and

A18) revealed that all ligands could bind at the kinase domain ATP binding site. The calculations were then rerun to focus on the kinase (ATP) binding site while allowing for flexible amino acid side chains at the binding site (GLY 431, GLN 432, VAL 436, LYS

454, GLU 500, LEU 501, CYS 502, GLU 506, LEU 553, ASP 564 and LYS 583). Use of the flexible amino acid side chains resulted in marked improvement in calculated binding energies.

Modeling studies were also conducted to evaluate 4-HPR-FERM domain interactions. FAK’s FERM domain structure was acquired from the Protein Data bank

(2AEH) [20]. All ligands were minimized using MMFF in Spartan 10 [21] while the protein structure was optimized via the default minimization protocol in Yasara [22].

Each ligand was run three times on a global search for the entire protein structure. 4-HPR interactions with the Fak family enzyme, protein tyrosine kinase 2 (PYK2) were also assessed. Analyses were conducted at the “closed” DFG and “DFG out” configurations which used 3FZR and 3FZT, respectively. All AutoDock Vina calculations were again repeated three times with an “exhaustiveness” of 100.

Assessment of 4-HPR’s effects on cell migration. Three complementary migration assays were used to assess 4-HPR’s effects on the diverse aspects of directed cell migration.

Scratch wound assay. Confluent cells were wounded by gently scratching the well surface with a sterile, cotton-tipped applicator, washed with PBS and treated with 1uM or

5uM 4-HPR for 24, 48, or 72 hours in Complete Medium with freshly prepared treatment supplied every 24 hours. At the end of each treatment period, three pictures were obtained for each well (left, middle, right) by using an Apotome Fluorescence

Microscope (Carl Zeiss, Dublin, CA), and AxioVision software (Carl Zeiss, Dublin, CA).

Immediately following the image capture, cell viability and proliferation were determined by using a hemocytometer. Quantitative image analysis of wound closure/cell migration was performed by utilizing ImagePro software (Media Cybernetics, Inc., Rockville, MD).

Cells with a high migration rate (EPI) underwent FAK siRNA (5-

AGCCAGUGAACCUCCUCUGACCGCAGG-3) (Integrated DNA Technologies Inc) treatment in accordance with standard procedures, with confirmation by immunoblotting

[23].

Cell-free Zone Exclusion Assay. The cell-free zone exclusion assay was conducted using the Oris Cell Migration assay (Platypus Technologies, Madison, WI). Briefly, following gel plug removal, cells were treated with freshly prepared 1, 5, or 10uM 4-HPR (0.1%

DMSO) or 0.1% DMSO, no 4-HPR (control) for 24 and 48 hours. Cells were stained with

0.5ug/mL Calcein AM in 1X PBS (Molecular Probes-Life Technologies, Grand Island,

NY) for 30 minutes, followed by flurostar microplate reader (485nmEx/528nmEm) analyses.

Chemoattractant-initiated Transwell migration assays. 96-well plates and 8µm pore membrane inserts were purchased from Trevigen (Gaithersburg, MD). JSCC-3 conditioned medium was determined to be the optimal chemoattactant relative to complete medium, or conditioned media from JSCC-1, JSCC-2 or 2095sc cells. 24 hour sera-starved cells were seeded into the top chamber with vehicle (0.1% DMSO), 1µM 4-

HPR, or 5µM 4-HPR and were incubated for 16 hours. The bottom chamber contained either (1) sterile-filtered, conditioned JSCC-3 media or (2) base medium. Formalin fixed cells were stained with 0.1% v/v crystal violet solution, followed by removal of cells

41 remaining in top chamber. Nikon DS-Ri1 using NIS Elements (Nikon, Melville, NY) was used to capture images, followed by target pixelation analyses by image segmentation

[ImagePro software (Media Cybernetics, Inc., Rockville, MD)].

Assessment of 4-HPR, FAK Inhibitor II and freeze dried black raspberries’ effects on

OSCC invasion of a synthetic basement membrane comprised of collagen type IV.

Preliminary studies determined that only the JSCC-1, JSCC-2, EPI and SCC2095 cell lines successfully invaded the collagen type IV layer and the optimal chemoattractant was

JSCC-3 conditioned medium. Fifty thousand 24-hour sera-starved cells/well were seeded onto type IV collagen-coated microporous polyester membrane (InnoCyte cell invasion kit, Calbiochem, San Diego, CA) with treatments (0, 5 µM 4-HPR, FAK II inhibitor

(Calbiochem, San Diego, CA., CAS 869288-64-2, 500 nM and 2.5 µM), or freeze dried black raspberries (10 µM cyanidin 3-rutinoside equivalent in base medium [15]. After 16

o hours of invasion (37 C, 5% CO2), cells were fixed and analyzed as described in migration assay.

Evaluation of treatment effects on phosphorylation status of proproliferative intracellular kinases. Cell lines with the greatest invasive capacities i.e. JSCC-2, EPI and

2095sc cells were pre-treated in sera-free media for 24 hours prior to 24-hour treatment in

JSCC-3 conditioned medium. Experimental groups were: (1) Vehicle (0.1% DMSO, determined to have no deleterious effects on cell viabilities), (2) 5uM 4-HPR, (3) BRB

(10uM cyanidin rutinoside equivalent), and (4) 4-HPR and BRB. Cells were harvested and analyzed in accordance with instructions (R & D Systems, Minneapolis, MN.

Phospho-MAPK array kit #ARY002B) was used to extract proteins, which were quantified by a BCA assay (Pierce, Rockford, IL.). Equivalent input proteins [BCA assay, (Pierce, Rockford, IL.)] were incubated, images obtained with the Li-Cor Odyssey imager (Li-Cor Biosciences, Lincoln, NE) and analyzed by ImagePro software (Media

Cybernetics, Inc., Rockville, MD).

Statistical analyses: Initial analyses confirmed that all data sets demonstrated a Gaussian distribution. A one-way ANOVA followed by Bonferroni’s multiple comparisons post hoc test was used to assess 4-HPR’s effects on caspase 3/7 activation and accompanying

FACS analyses, and also to determine the effects of 4-HPR, BRB, or combined treatments on cell invasion. 4-HPR’s effects on cell migration in the cell-free zone exclusion assay and the scratch wound assay were evaluated by the two-way ANOVA followed by Bonferroni’s multiple comparisons post hoc test.

2.4. Results.

Cell lines co-express cytokeratin and vimentin and possess 4-HPR metabolizing enzymes. Similar to our previous ATCC OSCC cell characterization studies [17], JSCC1,

JSCC2 and JSCC3 cell cultures uniformly demonstrated strong cytokeratin staining along with coexpression of cytokeratin and vimentin in cellular subpopulations (Figure 2.7).

Time course cell-4-HPR incubation studies revealed intracellular 4-HPR levels were higher than media levels during both the single and multiple dosing experiments (Table

2.1).

Furthermore, two of the three enzymes responsible for oxidative bioactivation of

4-HPR to 4-oxo-HPR i.e. cytochrome P450 (CYP) CYP3A4 and CYP26A1 were present in all the cell lines evaluated i.e. EPI, 2095sc, JSCC1, JSCC2, and JSCC3 cell lines.

CYP2C8 and the Phase II enzyme capable of 4-HPR glucuronidation (UGT1A1) were not present.

4-HPR treatment activated caspases 3 and 7 and perturbed F-actin organization. 4-

HPR treatment activated caspases 3 and 7 in a dose-dependent fashion in 6 of the 8 evaluated cell lines. While 1 µM 4-HPR significantly increased caspase activity in the

HOK3437E6/E7, JSCC1, and SCC15 cell lines, the HOK3437, EPI, and JSCC-2 cells only showed caspase induction with higher (5 µM) 4-HPR treatment (Figure 2.1.A.).

The JSCC3 and 2095sc cells were refractory to 4-HPR mediated caspase activation. Cell viabilities were comparable in all treatment groups. Corresponding FACS analyses, conducted in caspase-responsive (EPI) and caspase-refractory (2095sc) cell lines revealed increases in the sub-G1 (EPI) and G2/M (EPI and 2095sc) cell populations, respectively, during 4-HPR treatment (5 µM, 24 h treatment) (Figure 2.1.B). Somewhat paradoxically,

44 the 2095sc cells showed a proapoptotic DNA profile with the lower 1 µM 4-HPR dose

(Figure 2.1.B.).

4-HPR treatment also elicited distinct qualitative effects. 4-HPR challenge disrupted actin filament polymerization and intercellular adhesion as shown by loss of cellular polarity and dissipation of F-actin-cell membrane interactions (Figure 2.2).

4-HPR interacts with FAK’s kinase and FERM domains and also FAK’s closest homologue, Pyk2. 4-HPR demonstrates the highest binding affinity of all ligands at

FAK’s kinase ATP-binding site (Figure 2.3.A.). While a direct comparison between the binding affinity and an IC50 is not possible, these data imply that 4-HPR has a lower IC50 than any of the other compounds including the natural ligand ATP (See Figure 2.3.A.).

Three distinct binding pockets are located in the FERM domain. Pocket 1 is in a deep cleft between the F1, F2 and F3 domains of FERM, pocket 2 is in a deep cleft in the

F2 domain and pocket 3 is on the surface on the “backside” to the other two pockets and spans F1 and F2 (Figure 2.1.B.). 4-HPR shows an intermediate binding affinity with pockets 1 and 2 (3rd of 6 and 3rd of 5, respectively) relative to the other ligands evaluated

(Figure 2.3.B.). Retinol and 4-HPR were the exclusive ligands capable of binding in

FERM pocket 3 and 4-HPR demonstrated a slightly higher binding affinity. ATP binding at the FERM domain was restricted to pockets 1 and 2, while 2ETM binds only in pocket

1 (Figure 2.3.B.).

Pyk2 modeling studies evaluated 4-HPR’s interactions with its kinase catalytic site using the “closed” DFG (3FZR) and “DFG out” (3FZT) configurations (Table 2.2).

All compounds were determined to bind with a higher affinity to the DFG out conformer

(3FZT). Results indicated that 4-HPR binds to both conformers of Pyk2 with affinities comparable to recognized Pyk2 inhibitors such as PF-4618433 (Table 2.2).

4-HPR significantly inhibits cell migration. 4-HPR inhibited scratch wound healing in a cell line, dose and time dependent fashion (Figure 2.4.A). 5μM 4-HPR significantly suppressed both SCC15 and SCC2095sc cell line migration (p<0.0001, n=6, at 24, 48 and

72 h time points). In contrast, cell migration in normal oral keratinocytes (HOK3437) and the transduced HOK3437E6/E7 cells was only significantly affected at the 48 and 72 h time point when using the 5μM 4-HPR dose (p<0.0001, n=6). Scratch wound cell viabilities were comparable among all cell lines and treatment groups at every time point.

Furthermore, FAK-targeted siRNA treated cells demonstrated wound healing that was intermediate between control and 4-HPR treated cultures (Figure 2.4.B.). Corresponding

Western immunoblotting confirmed FAK siRNA treatment reduced endogenous cellular

FAK levels while Pyk2 protein levels remained unchanged or slightly increased. Also apparent was a distinct transition in cellular morphology from a flattened shape to a more rounded, less adherent phenotype in 4-HPR-treated cultures (Figure 2.4.B.).

Zone exclusion assays demonstrated that 5 µM 4-HPR significantly inhibited every cell line relative to its matched control cultures at 24h with the exception of normal keratinocytes (HOK3437) (Figure 2.4.C.). By 48 h 4-HPR significantly inhibited 46 migration of all cell lines (normal HOK 3437, HOKE6/E7 cells, SCC15, and SCC2095sc cells [p<0.001, n=8].

Preliminary studies confirmed that JSCC-3 conditioned media was the optimal chemoattractant relative to 10% FBS or conditioned media from any other cell lines.

Media protein array analyses revealed that JSCC3 conditioned medium contained appreciably higher levels of the established chemoattractant, IL-8, relative to either conditioned media or 10% FBS. As the JSCC-2, EPI, and SCC2095sc cells demonstrated the greatest motility, these lines were selected for the Boyden chamber assays. The chemotaxis-directed migration study results were comparable to our other migration data as 4-HPR suppressed cell migration in a dose-dependent fashion (Figure

2.4.D.).

While 4-HPR suppresses invasion, concurrent 4-HPR + BRB treatment provides additional invasion-suppressive effects. Pilot studies revealed that only the EPI, 2095sc and JSCC2 cells were reproducibly invasion-competent. Treatment with 5 µM 4-HPR significantly suppressed collagen type IV membrane invasion all three tested cell lines

(Figure 2.5.A. and 2.5.B.). Furthermore, although solitary BRB treatment produced modest anti-invasion effects, concurrent 5µM 4-HPR+ BRB (10µM cyanidin-3- rutinoside equivalent) treatment of all cell lines demonstrated an additive anti-invasive effect in the EPI, 2095sc cells and synergistic effects in the JSCC2 line (Figure 2.5.B.).

Inclusion of the FAK inhibitor II (0.5 and 2.5 µM final concentrations) had no anti- invasive effects on any cell lines. 47

Treatment with 4-HPR and BRB, singularly and in combination, reduced phosphorylation status of kinases associated with cell proliferation, survival, and apoptosis. Singular and combined treatment with 4-HPR and BRB affected kinase phosphorylation status (Figure 2.6). The JSCC2 cells experienced the greatest therapeutic effects relative to the other invasion-competent 2095sc or EPI cells. Of the 24 proteins evaluated in the JSCC2 cells, 4-HPR suppressed phosphorylation in 10, and had no effect on 10. Notably, the two proteins that showed increased phosphorylation in

JSCC2 cells i.e. c-Jun and Chk-2 are associated with stress-induced apoptosis and cell cycle arrest, respectively. p53 (S46) and p27 (T198) were not detected. While singular

BRB treatment also decreased phosphorylation levels, its effects were not as pervasive as

4-HPR. Notably, only combination 4-HPR + BRB treatment was able to decrease phosphorylation of the proliferation and migration enabling EGFR and the transcriptional activator MSK1/2.

2.5. Discussion.

Clinical evidence implicates FAK in the development and progression of OSCC

[24]. While FAK expression is restricted to the proliferative basal cell layer in healthy human oral epithelia, full-thickness FAK protein is present in premalignant OIN lesions

[24]. Notably, FAK contributes to essential aspects of OIN malignant transformation by facilitating basement membrane invasion and inappropriately sustaining proliferation

[25]. Our data demonstrate that local delivery achievable levels of 4-HPR [11] inhibit

FAK’s prosurvival, mobility-enhancing functions in a spectrum of cultured oral human 48 keratinocytes that range from normal to HPV E6/E7-transduced to malignant to metastatic.

All cell lines used in this study contained subpopulations that co-expressed cytokeratin and vimentin; findings consistent with the epithelial-to mesenchymal transition [26]. Our migration and invasion data show 4-HPR suppressed this mobile phenotype. Also, 4-HPR treatment resulted in an intracellular gradient that was appreciably higher than 4-HPR media levels. These findings suggest that intracellular 4-

HPR retention is sustainable and at least energetically neutral-potentially via phospholipid and protein binding. Our previous in vivo studies, which showed a time- dependent increase in target tissue 4-HPR levels following sequential 4-HPR topical dosing, support this premise [11].

4-HPR, at levels comparable to those used in this study, induced apoptosis in a variety of cultured human cancer cells including head and neck, ovary, and small cell lung carcinomas [27-29]. Following 4-HPR treatment in the current studies, execution phase caspase induction occurred in half of the cell lines. Treated cell DNA content showed increases in the sub-G1 and G2/M populations, even in those cell lines that did not show 4-HPR mediated caspase induction. These findings are consistent with 4-HPR and 4-oxo-HPR’s pro-apoptotic effects and 4-oxo-HPR’s mitotic arrest capabilities, respectively [30]. This premise is substantiated by the intracellular presence of cytochrome P450s (CYPs) capable of oxidative bioactivation of 4-HPR to 4-oxo-HPR i.e.

3A4 (consistent with human oral epithelia) and CYP26A1 [18]. Further, cell-ECM interactions are integral for both cell survival and induction of apoptosis [31]. FAK’s

49 dual capacity as a signaling kinase and adaptor/scaffold protein enables modulation of cell-ECM interactions and ultimately cell survival [1]. Our data, which showed disruption of actin filaments and transition to tall, rounded cells, confirmed 4-HPR disrupted cytoskeletal-ECM interactions [30]. Although cell-ECM disruptions generally trigger apoptosis, upregulated FAK activates constitutive cell survival pathways and apoptosis- resistance [31]. Notably, concurrent upregulation of FAK and oncogenic transformation of formerly cell adhesion-based survival signaling pathways occurs in a variety of human cancers [31]. We speculate that transformation of ECM-associated survival pathways was at least partially responsible for the failure of caspase activation in some of the

OSCC cell lines.

4-HPR demonstrated the highest binding affinity-including the endogenous ligand

ATP-at the FAK-kinase domain ATP-binding site. These findings recapitulate another

4-HPR-natural ligand interaction i.e. nyctalopia induced by 4-HPR’s displacement of vitamin A on retinol binding protein [32]. 4-HPR also interacted, albeit at a reduced affinity, with FAK-FERM’s 1, 2 and 3 pockets. FAK’s FERM domain links FAK to plasma membrane-associated growth factors, regulates FAK’s tyrosine kinase activity and facilitates FAK nuclear translocation [3]. In addition, the FERM domain binds to the

Arp2/3 complex, a key mediator in actin nucleation, and regulates lamellipodia formation, cell spreading, and ultimately cell movement [33]. Consequently, 4-HPR-

FERM interactions could significantly abate FAK’s proliferative, survival and pro- migratory functions [3]. Finally, an additional therapeutic effect is achieved via 4-HPR’s interaction with FERM’s pocket 2 [1, 3]. 4-HPR’s occupancy of pocket 2 will block its

50 associated Lys152, prevent a key FAK post-translational modification i.e. sumoylation and subsequently suppress FAK autophosphorylation at Tyr397 (integral in FAK kinase activation) and inhibit FAK nuclear translocation [1, 3].

4-HPR also interacts with FAK’s closest homologue, proline rich tyrosine kinase

2 (Pyk2), at its kinase catalytic site. Because Pyk2 can also contribute to p53 degradation and enable invasion and migration, it is regarded as a “FAK-alternative enzyme” [34].

Consequently, exclusive reliance on a FAK-only blockade can be at least partially overcome by Pyk2 [35]. FAK’s and Pyk2’s kinase sites contain a uniquely conserved glycine residue immediately adjacent their N terminals; a feature speculated to convey compound binding specificity [35]. In addition, 4-HPR’s capacity to bind more efficiently to Pyk2’s “out” DFG conformation corresponds to more selective kinase inhibitors [5]. Previous modeling studies by Xie et al. demonstrated 4-HPR interactions at the ATP-binding pocket of mammalian target of rapamycin (mTOR) [36]. The reduced phosphorylation of mTOR’s downstream target proteins following 4-HPR treatment supported these modeling studies [36]. Collectively, these data along with our kinase profiling results imply a predilection for 4-HPR binding at kinase ATP binding pockets which perturbs kinase function.

All migratory functions were significantly inhibited by 4-HPR. These findings likely reflect a dual mechanism of action i.e. disrupted actin microtubule assembly with concurrent reduction in cell proliferation via apoptosis or mitotic blockade. Our F-actin, caspase activation and flow cytometry data all support this mechanistic combination.

Our migration inhibition results compare favorably to other studies that demonstrated

51 comparable 4-HPR levels inhibited migration of cultured Kaposi’s sarcoma cells and androgen-independent prostate cancer cells [12, 14]. Also, FAK siRNA treatment intermediately suppressed scratch wound closure. These results are consistent with the co-contribution of Pyk2 in directed cell migration and support the modeling studies that implied 4-HPR perturbs both FAK and Pyk2 functions [34]. Notably, FAK translocates to the lipid raft components (an ideal milieu for retention of lipophilic 4-HPR) of migrating cells’ leading edges. This intracellular proximity increases prospects for 4-HPR-FAK interactions.

Basement membrane invasion by transformed keratinocytes defines OIN malignant transformation to OSCC. Invading cancer cells generate actin rich cellular protrusions “invadopodia” that contain a variety of proteins including cortactin, β1 integrin and matrix metalloproteinases (MMPs) [37]. The coordinated efforts of proteins such as FAK that modulate signaling, cytoskeletal-ECM interactions and actin stabilization are integral for invadopodia formation [38]. 4-HPR, putatively via perturbations in FAK and Pyk2 functions, significantly inhibited invasion in all 3 invasion-competent cell lines. The ineffectiveness of the FAK kinase targeted FAK inhibitor II to suppress invasion suggests that 4-HPR exerts its anti-migratory/-invasive effects via interference with the FERM domain. Furthermore, concurrent treatment with

BRB + 4-HPR augmented the inhibitory effects. Intracellular reactive species levels, which are elevated in many cancers, provide a plausible mechanism for these observations [39]. Reactive species mobilize MMPs via zymogen pro-domain cleavage and protease catalytic domain activation [39, 40]. BRB contain numerous redox-active

52 compounds e.g. anthocyanins proficient in reactive species scavenging [41].

Complementary proteome profiling analyses revealed 4-HPR singularly and in combination with BRB reduced phosphorylation status of 8 proteins which are integral for pro-proliferative signaling, cell adhesion and mobility. As reactive species also contribute to activation of kinase signaling cascades, these findings are consistent with the established redox-active functions of both BRB and 4-HPR [42, 43].

FAK dysregulation-such as observed in some premalignant oral lesions-can promote progression to OSCC. Our data, which imply 4-HPR is uniquely capable of perturbing FAK and Pyk2 survival and mobility enhancing effects, expand 4-HPR’s chemopreventive range beyond induction of apoptosis and differentiation. Although this investigation focused on 4-HPR-FAK interactions, virtually every bioactive compound elicits multiple cellular effects. As previously mentioned, 4-HPR’s proapoptotic effects likely co-contributed to inhibition of cell migration. To preserve the 4-HPR-FAK emphasis, this study concentrated on experimental parameters that were FAK-function based i.e. F-actin organization, cell-ECM interaction-based migration assays, formation of invadopodia, digestion of type IV collagen and invasion.

While this study focused on 4-HPR and to a lesser extent BRB, a variety of other

OSCC chemopreventives, with varied mechanisms of action, have been identified.

Among natural products, green tea extract (GTE), whose bioactive constituents include polyphenols (including epigallocatechin-3-gallate) and alkaloids (caffeine, theophylline and theobromine) has shown promising chemopreventive effects at both the in vitro and in vivo levels [44]. Although expression of the high output cyclooxygenase isoform

COX-2 had been implicated in OSCC development, negative results from a celecoxib oral premalignant lesion trial [45] combined with associated adverse cardiac events have eliminated COX-2 inhibitors from further OSCC chemopreventive considerations.

Recently, signaling pathway monoclonal antibodies and small molecule growth factor inhibitors that target either the receptor or associated tyrosine kinases have been introduced as therapeutic agents with chemopreventive potential [46]. Notably, the

“bench” chemopreventive success of 4-HPR has not translated to clinical oral cancer prevention [45]. This disconnect likely reflects poor bioavailability and significant first pass metabolism of systemically administered 4-HPR. To address this challenge, our labs developed a 4-HPR releasing mucoadhesive patch for direct application to OIN lesions

[11]. In vivo studies confirmed patch-released 4-HPR provided therapeutically-relevant levels to the treatment site, did not elicit any local or systemic toxicity, increased enzymes associated with keratinocyte differentiation and Phase II drug detoxification and also increased apoptosis [11]. We are, therefore, optimistic that targeted local delivery will enable 4-HPR to fulfill its chemopreventive potential.

Although selectively targeting pathways that are overexpressed in cancer cells is a compelling treatment concept, clinical use has revealed a range of side effects and eventual development of redundant signaling pathways in treated cancers [46, 47]. As recently discussed by a well-recognized oral cancer chemoprevention researcher, despite extensive efforts we still do not have an effective oral cancer chemoprevention strategy

[48]. Provided the extensive inter-patient heterogeneity of premalignant oral epithelial lesions [15], agent combinations based on complementary mechanisms of actions may be

54 necessary. Therefore, continued elucidation of agent(s)’ chemopreventive mechanisms combined with development of refined delivery formulations to address bioavailability issues appears timely and warranted.

2.6. References

1. Mitra AK, Hanson DA, Schlaepfer, DD. Focal Adhesion kinase: In Command and Control of Cell Motility. Nature Rev 2005; 6:56-68.

2. Liao Y, Mu G, Zhang L, Zhou W, Zhang J, Yu H. Lysophosphatidic Acid Stimulate Activation of Focal Adhesion Kinase and Paxillin and Promotes Cell Motility, via LPA1-3, in Human Pancreatic Cancer. Dig Dis Sci 2013; 58:3524- 3533.

3. Frame MC, Patel H, Serrels B, Lietha D, Eck MJ. The FERM domain: organizing the structure and function of FAK. Nature Rev 2010; 11:802-814.

4. Lim S-TS. Nuclear FAK: a New Mode of Gene Regulation from Cellular Adhesions. Mol. Cells 2013; 36:1-6.

5. Schultze A and Fiedler W. Therapeutic potential and limitations of new FAK inhibitors in the treatment of cancer. Expert Opin Investig Drugs 2010; 19:777- 788

6. Jacobson JJ, Epstein JB, Eichmiller FC, Gibson TB, Carls GS, Vogtmann E, et al. The cost burden of oral, oral pharyngeal and salivary gland cancers in three groups: commercial insurance, Medicare and Medicaid. Head & Neck Oncology 2012; 4:1-17.

7. Ho MW, Field EA, Field JK, Risk JM, Rajlawat BP, Rogers SN, et al. Outcomes of oral squamous cell carcinoma arising from oral epithelial dysplasia: rationale for monitoring premalignant oral lesions in a multidisciplinary clinic. Br J Oral Maxillofac Surg 2013; 51:594-599.

8. Hong WK, Endicott J, Itri LM, Doos W, Batsakis JG, Bell R, et al. 13-cis- retinoic acid in the treatment of oral leukoplakia. N Engl J Med 1986; 315:1501-5.

9. Chiesa F, Tradati N, Grigolato R, Boracchi P, Biganzoli E, Crose N, et al. Randomized trial of fenretinide (4-HPR) to prevent recurrences, new 55

localizations and carcinomas in patients operated on for oral leukoplakia: long- term results. Int J Cancer 2005; 115:625-29.

10. Ulukay E, Pirianov G, Kurt MA, Wood EJ and Mehmet H. Fenretinide induces cytochrome c release, caspase 9 activation and apoptosis in the absence of mitochondrial membrane depolarization. Cell Death and Differentiation 2003; 10, 856–859.

11. Holpuch AS, Phelps MP, Desai K-GD, Chen W, Koutras GM, Han BB et al. Evaluation of a mucoadhesive fenretinide patch for local intraoral delivery: a strategy to reintroduce fenretinide for oral cancer chemoprevention. Carcinogenesis 2012; 33: 1098-1105.

12. Ferrari N, Morini M, Pfeffer U, Minghelli S, Noonan DM, Albini A. Inhibition of Kaposi’s Sarcoma in Vivo by Fenretinide. Clin Cancer Res 2003; 9: 6020- 6029.

13. Golubkov V, Garcia A, Markland FS. Action of Fenretinide (4-HPR) on Ovarian Cancer and Endothelial Cells. Anticancer Res 2005; 25: 249-253.

14. Benelli R, Monteghirfo S, Vene R, Tosetti F, Ferrari N. The chemopreventive retinoid 4HPR impairs prostate cancer cell migration and invasion by interfering with FAK/AKT/GSK3β pathway and β-catenin stability. Mol Cancer 2010; 9: 1-13.

15. Mallery SR, Tong M, Shumway BS, Curran AE, Larsen PI, Ness GM et al. Topical Application of a Mucoadhesive Freeze-Dried Black Raspberry Gel Induces Clinical and Histologic Regression and Reduces Loss of Heterozygosity Events in Premalignant Oral Intraepithelial Lesions: Results from a Multicentered, Placebo-Controlled Clinical Trial. Clin Cancer Res 2014; 20:1910-1924.

16. Müssig E, Steinberg T, Kohl A, Chamulitrat W, Komposch G, Tomakidi P. Discrimination of epithelium-like and fibroblast-like phenotypes derived from ethanol-treated immortalized human gingival keratinocytes in Epithelial equivalents. Cell Tissue Res 2008; 332:57–71.

17. Tong M, Han BB, Holpuch AS, Pei P, He L, Mallery SR. Inherent phenotypic plasticity facilitates progression of head and neck cancer: Endotheliod characteristics enable angiogenesis and invasion. Exp Cell Res 2013; 319: 1028-1042.

18. Mallery SR, Tong M, Michaels GC, Kiyani AR, Hecht SS. Clinical and Biochemical Studies Support Smokeless Tobacco’s Carcinogenic Potential in the Human Oral Cavity. Cancer Prev Res 2014; 7:23-32.

19. Trott O, Olson AJ. J. AutoDock Vina: Improving the speed and accuracy of docking with a new scoring function, efficient optimization, and multithreading. Computational Chem 2010; 31:455-461.

20. http://www.rcsb.org/pdb/home/home.do Helen M. Berman, John Westbrook, Zukang Feng, Gary Gilliland, T. N. Bhat, Helge Weissig, et al. The Protein Data Bank, Nucl. Acids Res. (2000) 18 (1): 235-242. Doi: 10.1093/nar/28.1.235

21. Spartan’10, Wavefunction, Inc. Irvine, CA Y. Shao, L.F. Molnar, Y. Jung, J. Kussmann, C. Ochsenfeld, S.T. Brown, A.T.B. Gilbert, et al. Phys. Chem. Chem. Phys., 8, 3172 (2006).

22. Krieger E, Koraimann G, Vriend G. Increasing the precision of comparative models with YASARA NOVA—a self-parameterizing force field. Proteins 2002; 47:393–402.

23. Wang D, Zhang H, Li M, Frid MG, Flockton AR, McKeon BA, et al. MicroRNA-124 controls the proliferative, migratory, and inflammatory phenotype of pulmonary vascular fibroblasts. Circ Res 2014; 114:67-78.

24. Canel M, Secades P, Rodrigo JP, Cabanillas R, Herrero A, Suarez C et al. Overexpression of Focal Adhesion Kinase in Head and Neck Squamous Cell Carcinoma Is Independent of fak Gene Copy Number. Clin Cancer Res 2006; 12: 3272-3279.

25. Siesser PM, Hanks SK. The Signaling and Biological Implications of FAK Overxpression in Cancer. Clin Cancer Res 2006; 12: 3233-3237.

26. Kalluri R and Weinberg RA. The basics of epithelial-mesenchymal transition. J Clin Invest 2009; 119: 1420-1428.

27. Scher RL, Saito W, Dodge RK, Richtsmeier WJ, Fine RL. Fenretinide-induced apoptosis of human head and neck squamous carcinoma cell lines. Otolaryngology 1998; 118: 464-471.

28. Supino R. Crosti M, Clerici M, Warlters A, Cleris L. Induction of apoptosis by fenretinide (4HPR) in human ovarian carcinoma cells and its association with retinoic acid receptor expression. Int J Cancer 1996; 491-497.

29. Kalemkerian GP, Slusher R, Ramalingam S, Gadgeel S, Mabry M. Growth Inhibition and Induction of Apoptosis by Fenretinide in Small-Cell Lung Cancer Cell Lines. J Natl Cancer Inst 1995; 22: 1674-1680.

30. Tiberio P, Cavadini E. Abolafio G, Formelli F, Appierto V. 4-oxo-N-(4- hydroxyphenyl)retinamide: two independent ways to kill cancer cells. PLoS One 2010; 5: 1-11.

31. van Nimwegen MF, Huigsloot M, Camier Am, Tijdens IB, Van de Water B. Focal Adhesion Kinase and Protein Kinase B Cooperate to Suppress Doxorubicin-Induced Apoptosis of Breast Tumor Cells.. Mol Pharmacol. 2006; 70: 1330-1339.

32. Garaventa A, Luksch R, LoPicolo MS, Cavadini E, Montaldo PG, Pizzitola MR et al. Phase I Trial and Pharmacokinetics of Fenretinide in Children with Neuroblastoma. Clin Cancer Res 2003; 9:2032-2039.

33. Serrels B, Serrels A, Brunton VG, Holt M, McLean GW, Gray CH, Jones GE, Frame MC. Focal adhesion kinase controls actin assembly via a FERM- mediated interaction with the Arp2/3 complex. Nature Cell Biology. 2007; 9:1046-1056.

34. Lipinski CA, Tran NL, Menashi E, Rohl C, Kloss J, Bay RC et al. The Tyrosine Kinase Pyk2 Promotes Migration and Invasion of Glioma cells. Neoplasia 2005; 7: 435-445.

35. Lipinski CA, Loftus JC. Targeting Pyk2 for therapeutic intervention. Expert Opin Ther Targets 2010; 14: 95-108.

36. Xie H, Zju F, Huang Z, Lee M-H, Kin DJ, Li X et al. Identification of mammalian target of rapamycin as a direct target of fenretinide both in vitro and in vivo. Carcinogenesis 2012; 33:1814-1821.

37. Yamaguchi H. Pathological roles of invadopodia in cancer invasion and metastasis. Eur J Cell Biol 2012; 91: 902-907.

38. Wang Y, McNiven MA. Invasive matrix degradation at focal adhesions occurs via protease recruitment by a FAK-p130Cas complex. J Cell Biol 2012; 196: 375-385.

39. Storz P. Reactive Oxygen Species in Tumor Progression. Front Biosci 2005; 10:1881-1896.

40. Chiu WT, Shen SC, Chow JM, Lin CW, Shia LT, Chen TC. Contribution of reactive oxygen species to migration/invasion of human glioblastoma cells U87 via ERK-dependent COX-2/PGE2 activation. Neurobiol Dis 2010; 37:118-129.

41. Stoner GD, Wang LS, Casto BC. Laboratory and clinical studies of cancer chemoprevention by antioxidants in berries. Carcinogenesis 2008; 29:1665- 1674.

42. McCubrey JA, LaHair MM, Franklin RA. Reactive Oxygen Species-Induced Activation of the MAP Kinase Signaling Pathways. Antioxid Redox Signal 2006; 8: 1775-1789.

43. Wang K, Fang H, Xiao D, Zhu X, He M, Pan X et al. Converting Redox Signaling to Apoptotic Activities by Stress-Responsive Regulators HSF1 and NRF2 in Fenretinide Treated Cancer Cells. PLoS ONE 2009; 4: e7538 1-13.

44. Chen D, Wan SB, Yang H, Yuan J, Chan TH, Dou QP. EGCG, Green Tea Polyphenols and their Synthetic Analogs and Prodrugs for Human Cancer Prevention and Treatment. Adv Clin Chem 2011; 53: 155-177.

45. Papadimitrakopoulou VA, William WN Jr, Dannenberg AJ, Lippman SM, Lee JJ, Ondrey FG, et al. Pilot randomized phase II study of celecoxib in oral premalignant lesions. Clin Cancer Res 2008; 14: 2095–2101.

46. Widakowich C, De Castro G, De Azambuja E, Dinh P, Awada A. Review: Side Effects of Approved Molecular Targeted Therapies in Solid Cancers. The Oncologist 2007; 12: 1443-1455.

47. Widakowich; Logue JS and Morrison DK. Complexity in the signaling network: insights from the use of targeted inhibitors in cancer therapy. Genes Dev 2012; 26: 641-650.

48. William AN Jr. Oral premalignant lesions: any progress with systemic therapies? Curr Opin Oncol 2012; 24: 205-210.

Figure 2.1. Activation of caspase 3/7 and cell cycle modulations by 4-HPR. (A) 4HPR induced activation of the execution phase apoptotic enzymes, caspase 3/7, in HOK3437

(a), HOK3437 E6/E7 (b), EPI (c), JSCC1 (d), and JSCC2 (e), and SCC15 (g) cell lines.

JSCC-3 (f) and SCC2095sc (h) did not show caspase 3/7 activation during treatment with any of the 4-HPR doses. Cells were seeded at 1x105/well in 96 well plates and treated in serum-free media for 24 hours before measurement. Data are represented as means ±

SEM of 7 replicates (c,d,e,f) or of 4 replicates (a,b,g,h). Asterisks indicate a significant difference from cell line matched vehicle control (B) FACS analyses demonstrated 4-

HPR treatment perturbed cell cycle kinetics by increasing sub-G1 and G2/M DNA distribution in both a caspase induced (EPI) and caspase refractory (2095sc) cell lines

(n=2). (*p<0.05, **p<0.01, ****p<0.0001).

Figure 2.1

Figure 2.2. 4-HPR disrupts actin cytoskeleton organization. Actin filaments (F-actin) were labeled with a rhodamine fluorescent probe and visualized under 400X image scale via fluorescent microscopy. 4-HPR-treated samples showed (1) loss of cellular polarity,

(2) dissipation of cortical actin networks, and (3) loosening of intercellular junctions.

HOK3437, HOK3437 E6/E7 and SCC2095sc all exhibited 4-HPR’s dose-escalating effects manifesting as cytoskeletal rearrangement and/or destabilization of actin filaments. SCC15 appeared to be largely unaffected by 4-HPR. A DAPI counterstain identifies the nuclei.

Figure 2.3. 4-HPR interacts with FAK’s kinase and FERM domains. Molecular modeling studies were conducted using AutoDock Vina software [19]. Initial 4-HPR binding studies used retinol binding protein as a model binding protein. (A) Molecular modeling image depicting 4-HPR (blue and white) interacting with the FAK kinase ATP binding site (orange, green and red). The accompanying table compares ligand binding affinities for the kinase domain of FAK. [* 7-PYRIDIN-2-YL-N-(3,4,5-

TRIMETHOXYPHENYL)- 7H-PYRROLO[2,3-D]PYRIMIDIN-2-AMINE; ** 2-({5-

CHLORO-2-[(2-METHOXY-4-MORPHOLIN-4- YLPHENYL)AMINO]PYRIMIDIN-

4-YL}AMINO)-N-METHYLBENZAMIDE; *** 1,4-bis(diethylamino)-5,8- dihydroxyanthraquinone; **** 6-(4-(3-(methylsulfonyl)benzylamino)-5-

(trifluoromethyl)pyrimidin-2-ylamino-3,4-dihydroquinolin-2(1H)-one] (B) 4-HPR (blue and white) depicted binding in FERM domain’s pocket 1. Ligand binding affinities at

FAK’s FERM domain pockets are listed in the table below.

Footnotes: a) The Protein Data Bank: http://www.rcsb.org/pdb/home/home.do :2ETM b) The Protein Data Bank: http://www.rcsb.org/pdb/home/home.do : 2JKK c) Shi, Q, et al. Molecular Carcinogenesis, 46: 488-96 (2007) d) Slack-Davis, J.K. et al J. Biol. Chem. 2007, 282: 14845-52. doi:

10.1074/jbc.M606695200

Figure 2.3

Figure 2.4. Evaluation of 4-HPR’s effects on directed cell migration. (A) HOK3437,

HOK3437 E6/E7, SCC15, and SCC2095sc cells were treated with 1 and 5uM 4-HPR over 72 hours with fresh drug and media replenished every 24 hours. 4-HPR significantly attenuated directed cell migration in a dose- and time-dependent manner. (n=6). (B)

FAK-targeted siRNA treated cells showed an intermediate cell migration capacity between control and 4-HPR-treated samples. Western blot analyses demonstrate decreased levels of FAK protein in EPI and SCC2095sc cell after FAK siRNA transfection while Pyk2 levels remained constant or were slightly higher in the FAK siRNA-treated samples. (C) Normal, HPV E6/E7 immortalized, and oral squamous cell carcinoma cell lines were treated with 5uM 4-HPR for 24 hours in a cell-free zone exclusion cell migration assay. While 4-HPR significantly inhibited cell migration in

HOK3437 E6/E7 (n=8), SCC15 (n=8), JSCC-3 (n=7), and SCC2095sc (n=7), no anti- migratory effects were noted on normal HOK3437 cells. (*p<0.05, **p<0.01,

****p<0.0001) (D) Directed migration of EPI cells was evaluated by using Boyden chambers. 1uM and 5uM 4-HPR treatment for 24 hours resulted in a dose dependent inhibition of directed cell migration. Comparable results were obtained in SCC2095sc.

These migration assays showed 4-HPR’s consistent inhibitory effects on all aspects of migration

Figure 2.4

Figure 2.5. Evaluation of the effects of 4-HPR and freeze dried blackraspberries (BRB) on cell invasion. (A) The effects of 4-HPR and BRB of directed cell invasion were evaluated by using collagen IV-coated transwell membrane (8 micron pores). Cells were stained with 0.1% v/v crystal violet after 4% paraformaldehyde fixation. 5uM 4-HPR significantly inhibited invasion in all cell lines with 2095sc cells showing most 4-HPR responsiveness. While single agent treatment with BRB significantly suppressed invasion in the JSCC2 and 2095sc cell lines, concurrent 4-HPR + BRB demonstrated additive (EPI and 2095sc) or synergistic (JSCC-2) anti-invasive effects. (B) Histogram depiction of 4-HPR and BRB’s effects on cell invasion of a synthetic basement membrane. (n=8, error bars represent SEM, * p<0.0001). Data were normalized to the total area of the field. Introduction of the FAK inhibitor II (0.5 and 2.5 µM), which inhibits FAK’s kinase function, had no effects on cell invasion (Figure 2.8).

Figure 2.5

Figure 2.6. Modulation of phosphorylation status of kinases associated with cell migration, proliferation, survival, and apoptosis by 4-HPR and BRB. Proteome profiles were currently conducted on the 3 highly invasive cell lines depicted in Figure 2.5. Of the

24 proteins evaluated in the panel of serine/threonine/tyrosine phosphorylation residues,

4-HPR consistently suppressed phosphorylation at 5 target residues (i.e. β-catenin,

STAT3 Y705, STAT3 S727, PYK2, and Erk1/2; orange-colored cells) in all cell lines tested. Furthermore, 4-HPR treatment increased phosphorylation of 2 (c-Jun, which is associated with stress-induced apoptosis and Chk-2 which leads to cell cycle arrest) in the

JSCC-2 cells.

Figure 2.6

Figure 2.7: JSCC cell lines contain cytokeratin and vimentin, suggestive of epithelial-to-mesenchymal transition in subpopulations. JSCC cells were successfully isolated from primary oral cancer lesions and cultured. The cells were fixed and probed for both epithelial (i.e. pan cytokeratin) and mesenchymal (i.e. vimentin) markers.

JSCC1, JSCC2, and JSCC3 co-express both pan-cytokeratin and vimentin, indicating that the cells have undergone epithelial-to-mesenchymal transitions (EMT). Abbreviations

(PNI: perineural invasion, LVI: lymphovascular invasion, HPV: human papillomavirus)

JSCC1 - Tonsil/Hypopharynx - T4bN2bMx

PNI: Yes LVI: Yes Nodal Status: 2 of 104 HPV: Negative

JSCC2 - Tongue - T3N0Mx

PNI: Yes LVI: Yes

Nodal Status: 0 of 50 HPV: Negative

JSCC3 - Floor of Mouth - T2N1Mx

PNI: Yes LVI: No Nodal Status: 1 of 27 HPV: Negative

Figure 2.8: FAK inhibitor II demonstrated negligible invasion-inhibitory effects on

EPI and SCC2095sc cell lines. EPI and SCC2095sc were sera-starved for 24 hours prior

to seeding into the invasion chamber coated with type IV collagen. Cells were treated for

16 hours with a vehicle (0.1% DMSO), 5uM 4-HPR, 500nM or 2.5uM FAK inhibitor,

followed by 4% paraformaldehyde fixation and 0.1% (v/v) crystal violet staining. The

invading cells were then visualized using light microscopy. Treatment of EPI and

SCC2095sc cell lines with 500nM and 2.5uM FAK inhibitor II, designed to target FAK’s

kinase domain, did not result in inhibition of cell invasion. 5uM fenretinide was used as a

positive control.

EPI

SCC2095sc VEH 5uM 4-HPR 500nM FAK 2.5uM FAK Inhibitor II Inhibitor II

Figure 2.9: Fenretinide (4-HPR) inhibits cell migration in transwell migration assay.

This data supplements Figure 2.4.D. and followed the same experimental protocol

outlined in the methods section. SCC2095sc showed comparably less migratory behavior

to EPI in the transwell migration system. The qualitative data shows that fenretinide

treatment results in inhibition of SCC2095sc migration.

SCC2095sc

VEH 1uM 4-HPR 5uM 4-HPR

Figure 2.10: Evaluation of fenretinide (4-HPR)’s effects on cell viability. HOK3437,

HOK3437 E6/E7, SCC15, and SCC2095sc cells were treated with vehicle (DMSO 0.1% v/v), 1uM 4-HPR, or 5uM 4-HPR for 24 hours. Cells were detached with trypsin

(GIBCO, Life Technologies Grand Island, NY), stained with trypan blue (Sigma St.

Louis, MO), and counted using the hemacytometer. 4-HPR treatment demonstrated no significant alteration in cell viability.

C e ll V ia b ility

1 5 0 C o n tro l 1 u M 4 -H P R

5 u M 4 -H P R y

t 1 0 0

5 0 %

7 7 5 c 3 E 1 s 4 / C 5 3 6 9 E C 0 K S 2 O 7 3 C H 4 C 3 S K O H

Figure 2.11: Fenretinide inhibits tubulin polymerization. Fenretinide and its active

(i.e. 4-oxo) and inactive (i.e. 4-MPR) along with positive (i.e. paclitaxel and vinblastine) and negative (vehicle control) controls were co-incubated with tubulin polymerization reaction mixture containing tubulin monomers (2mg/mL), 10% glycerol buffer, and 1mM

GTP. The polymerization reaction was continuously monitored in kinetic mode for 60 minutes by reading excitation/emission wavelengths at 350/440nm. The degree of tubulin polymerization was quantified by subtracting the blank-normalized relative fluorescence unit (RFU) value at the starting point (0 minute) from the final RFU value (60 minutes).

The data demonstrated that fenretinide and its active metabolite, 4-oxo, are capable of inhibiting tubulin polymerization in the cell-free environment. As expected, paclitaxel and vinblastine (positive controls) destabilized microtubule formation, and 4-MPR, an inactive metabolite of fenretinide, did not significantly inhibit tubulin polymerization.

F e n re tin id e In h ib its T u b u lin P o ly m e riz a tio n

) 1 0 0 0 0 0

e l

c 5 0 0 0 0

(

F R

e d

l l e o e R o R o r x in P x x t t o P o ta s H - - n i - 4 -M o l la 4 /4 c c 4 a b M R e n M u M P l P i u 5 u ic V 5 5 -H h 4 e V

Figure 2.12: Evaluation of fenretinide (4-HPR)’s effects on tubulin polymerization.

HOK3437 E6/E7 cells were treated with vehicle (0.1% DMSO), 1uM paclitaxel, 1uM vinblastine, 10uM 4-HPR, 10mM N-acetylcysteine (NAC), and 10uM 4-HPR/10mM

NAC for 24 hours, and proteins were harvested via differential detergent fractionation.

Protein samples containing soluble or insoluble fractions of tubulins were probed with anti-beta-tubulin antibody (Abcam) and processed for Western blotting. Densitometry was utilized to determine soluble/insoluble tubulin fraction ratios. As expected, paclitaxel caused accumulation of insoluble fractions by stabilizing microtubule, whereas vinblastine caused an increase in soluble fraction by destabilizing tubulin polymerization.

Surprisingly, fenretinide stabilized the microtubule formation, contradicting the data shown in Figure 2.11. 10mM NAC, a known anti-oxidant capable of scavenging reactive oxygen species, restored fenretinide’s microtubule stabilizing effects. Collectively, fenretinide’s effects on tubulin polymerization are equivocal and warrant further investigation.

Figure 2.12

Table 2.1. Levels of 4-HPR in OSCC cells and complete medium during incubation.

Log growth SCC2095sc cells (cultured in DMEM/F12 medium + 10% heat inactivated

FBS “complete medium”) were treated for sequential days with 5 µM 4-HPR (5 µM 4-

HPR in fresh medium q 24 h-samples 2-5, or additional 5µM 4-HPR q 24 h added to conditioned medium-samples 7-10), followed by cell lysis and sample preparation for

LC/MS/MS analyses [11]. Accompanying studies to assess effects of 37OC incubation on

4-HPR’s stability in complete medium (while protected from light) revealed a time and 4-

HPR dose dependent effect. Single dose 5µM 4-HPR showed 35% degradation over the first 24h, and with degradation increasing to 49% at 48 h and then remaining stable up to the 96 h time point. Fresh supplementation of 4-HPR q 24h in medium resulted in an overall 25% reduction over the 96 h time course (data shown above). Cell and media incubations showed an approximate 80% decrease in medium 4-HPR levels accompanied by appreciable increases in intracellular 4-HPR levels. These data equate to an overall

45% reduction in 4-HPR medium levels, attributable to intracellular retention, after accounting for incubation-mediated degradation. N.D.=Not Determined

Table 2.2: 4-HPR binding Interactions with Pyk2

Compound E (kcal/mol) Kd flex 2 – E (kcal/mol) Kd flex 2 IC50 (nM) 3FZR 3FZR 3FZT – 3FZT Retinol -9.1 2.12 x 10-7 -12.7 4.85 x 10- 10 2ETM-ligand -8.8 3.52 x 10-7 -9.7 7.69 x 10- * 8 ATP -8.0 1.36 x 10-6 -9.4 1.28 x 10- 7 ATP-Mg -8.7 4.16 x 10-7 -10.2 3.30 x 10- complex 8 4-HPR -10.8 4.16 x 10-8 -13.1 2.47 x 10- 10 4-oxo-4HPR -9.1 2.12 x 10-7 -12.4 8.05 x 10- 10 A18 *** -7.6 2.67 x 10-6 -9.3 1.51 x 10- 7 PF228 **** -11.1 7.23 x 10-9 -14.2 3.85 x 10- >1000a 11 TAE-226 ** -9.8 6.48 x 10-8 -11.0 8.56 x 10- 5.5b 9 BIRB796 -9.4 1.28 x 10-7 -12.3 9.53 x 10- 31c 10 PF-431396 -10.3 2.79 x 10-8 -13.1 2.47 x 10- 11d 10 PF-4618433 -10.7 1.42 x 10-8 -11.7 2.62 x 10- 9 Compound-9 -9.9 5.49 x 10-8 -11.5 3.68 x 10- 170 c 9 References: a) Slack-Davis, J.K. et al J. Biol. Chem. 2007, 282: 14845-52. doi: 10.1074/jbc.M606695200 b) Shi, Q, et al. Molecular Carcinogenesis, 46: 488-96 (2007) c) Samit K. Bhattacherya, et al. “Identification of novel series of pyrazole and indole-urea based DFG-out PYK2 inhibitors” Bioorganic & Medicinal Chemistry Letters, 22 (2012) 7523-29. doi: 10.1016/j.bmcl.2012.10.039. d) Seungil Han, et al. “Structural Characterization of Proline-rich Tyrosine Kinase 2 (PYK2) Reveals a Unique (DFG-out) Conformation and Enables Inhibitor Design” J. Biol. Chem. 2009, 284: 13193-201. doi: 10.1074/jbc.M809038200.

Chapter 3: Personalized Tumor Profiling: Utilizing Fenretinide (4-HPR) as a

Biological Modifier to Optimize Therapeutic Outcome

3.1. Abstract

Personalized tumor profiling, therefore, enables identification of aberrant molecular and biochemical phenotypes and customization of therapeutic strategy to fit individual patient’s OSCC profiles. Characterization of specific tumorigenic signaling pathways and cytokine secretion profiles revealed significant heterogeneity among recently isolated

OSCC cell lines (JSCC-1, JSCC-2, and JSCC-3), leading to varied response to specific inhibition of key tumorigenic pathways (e.g. EGFR and STAT3); Afatinib, specific

80 fenretinide’s capacity to modulate cellular phosphorylation status toward antitumorigenic phenotypes. Fenretinide combined with specific small molecule inhibitors

3.2. Introduction

Oral squamous cell carcinoma represents a debilitating disease associated with a significant morbidity and mortality [1-4]. Despite extensive research, 5-year survival rates have improved only marginally over the past 40 years, and surgical resection remains as the primary treatment modality. Patients who are fortunate enough to enter into complete remission are often left with severely compromised esthetics and psychosocial dysfunction. Chemoprevention, which suppresses, reverses, and prevents

OSCC carcinogenesis, therefore, constitutes an important facet of OSCC management and can provide life-saving interventions for OSCC patients.

Accumulating evidence demonstrates that a great deal of molecular and biochemical heterogeneity exists among OSCC tumors. Stransky et al’s global exome sequencing study revealed a glimpse of OSCC’s genomic inter-tumor heterogeneity exemplified by a diverse array of mutated genes in individual tumors [5]. Another study demonstrated differential gene expressions of OSCC tumors according to their anatomic locations [6]. In addition, evaluation of loss of heterozygosity (LOH) and microsatellite

81 instability (MI) within the same tumors revealed intra-tumor heterogeneity in OSCC [7].

Customization of therapeutic strategy to tailor to specific patient needs could, therefore, account for intra-/inter-tumor heterogeneity and offer a distinct pharmacologic advantage.

Premalignant and malignant OSCC cells employ growth factor and cytokine signaling to facilitate cell proliferation, survival, and migration/invasion. These oncogenic pathways are capable of receptor cross-talk and utilize redundant signaling mediators to increase survival advantage. Notably, receptor tyrosine kinases (RTKs), the most common targets of FDA-approved therapies, often evade specific inhibitors by activating compensatory pathways sharing common downstream effectors [8].

Cetuximab, a small molecule inhibitor of epidermal growth factor receptor (EGFR) for

OSCC chemotherapy, has demonstrated limited clinical efficacy [9]. Studies suggest that resistance to EGFR inhibition may develop via activation of signal transducer and activator of transcription 3 (STAT3) pathways [9,10].

STAT3, a downstream target of EGFR, plays a crucial role in cell proliferation, survival, malignant transformation, and even metastatic niche formation [11]. STAT3 activation occurs via multiple signaling pathways including EGFR and IL-6R/gp130 [11];

(1) EGFR activation leads to increased c-Src kinase activity resulting in STAT3 activation and (2) IL-6R/gp130 complex transduces signal to JAK2, which in turn, activates STAT3 by phosphorylating Y705 residue, a key event required for STAT3 dimerization [11]. As such, cancer cells exhibit resilience in maintaining STAT3 activation by activating the alternative IL-6R pathway upon EGFR inhibition [12]. In addition, STAT3 inhibition enhanced sensitivity to EGFR inhibition by afatinib [12].

Despite the investigation of several STAT3-specific small molecule inhibitors over the years, no FDA-approved STAT3 inhibitor currently exists for OSCC treatment

[13]. In addition, studies have demonstrated that STAT3 inhibitors alone produce only modest therapeutic effects [14]. This may be, in part, due to (1) STAT3 inhibitors’ limited scope of action targeting only phosphorylation and dimerization of STAT3 and

(2) existence of compensatory pathways capable of activating a similar set of downstream genes (e.g. NF-kB) [15]. As such, signaling redundancy and development of resistance via compensatory pathways continue to pose a challenge for therapeutic use of specific inhibitors. Biological modifiers, capable of modulating multiple signaling pathways and

“priming” cells for targeted therapeutic drugs (e.g. afatinib and vargatef), could significantly enhance therapeutic outcomes. Such therapeutic strategy, also known as chemosensitization, utilizes natural or synthetically derived products to overcome chemoresistance [16]. Biological modifiers are capable of exerting pleotropic effects on key signaling pathways without causing overt toxicity [16].

Fenretinide (4-HPR), a synthetic analog of all-trans retinoic acid, possesses potential to serve as a biological modifier. Our study, which evaluated 4-HPR’s effects on over 40 key tyrosine/serine/threonine phosphorylation residues, demonstrated that fenretinide negatively modulated phosphorylation status associated with cell proliferation, survival, and apoptosis (e.g. STAT3 and FAK) (See Chapter 2) [17]. In addition, fenretinide, at therapeutically relevant dosing levels, did not cause overt cytotoxicity (See Chapter 2). This study, therefore, aim to determine fenretinide’s

83 capacity to augment therapeutic efficacy of small molecule inhibitors (e.g. afatinib/vargatef) based on tumor profiling.

3.3. Materials and Methods

Cell culture. JSCC-1, JSCC-2, and JSCC-3 were isolated from tumor tissues resected during surgery. Each JSCC cell line has a tumor tissue match that was subsequently utilized for immunohistochemistry studies. In order to ensure authenticity of these cell lines and demonstrate lack of cross-contamination, DNA was isolated (Genomic DNA

Purification Kit, Promega, Madison, WI) from the entire panel of cell lines that are currently being used in our laboratories and authenticated via short tandem repeats (STR) profiling analysis at the Genetic Resources Core Facility (Johns Hopkins University,

Baltimore, MD). SCC2095sc represents a highly malignant cell line isolated and cultured from the tumor cell xenotransplantation experiment using SCC2095 cell line acquired from American Type Culture Collection (ATCC, Manassas, VA).

Cell Line Characterization. Formalin fixed cells were incubated with vimentin (1:200,

Abcam, Cambridge,MA) or a pancytokeratin cocktail (AE1/AE3 + 5D3, 1:100, Abcam,) antibodies, followed by incubation with FITC or Texas Red conjugated secondary antibodies (Abcam, Cambridge, MA) [18]. Nuclei were stained with 4’,6’-Diaminidino-

2-phenylindole dihydrochloride (DAPI, Abcam). Fluorescence microscopy images were 84 obtained by using an Olympus BX51 microscope (Olympus, Japan), NikonDS-Fi1 digital camera (Nikon, Japan) and ImagePro 6.0 (Media-Cybernetics, Bethesda, MD).

Evaluation of cytokine secretion profile from OSCC cell lines. JSCC-1, JSCC-2, and

JSCC-3 were seeded and sera-starved for 24 hours prior to collecting conditioned media samples. The conditioned media samples were processed and analyzed according to the manufacturer’s instructions provided in Proteome Profiler Human Cytokine XL Array

(R&D Systems, Minneapolis, MN). Briefly, the samples were incubated with nitrocellulose membrane imbedded with capture antibodies against a panel of cytokines, chemokines, and growth factors. The membrane was subsequently washed and probed with biotinylated detection antibodies followed by incubation with streptavidin- conjugated fluorophores. Fluorescent images were obtained with the Li-Cor Odyssey imager (Li-Cor Biosciences, Lincoln, NE) and analyzed by Image Studio software (Li-

Cor Biosciences, Lincoln, NE).

Cell Viability and Proliferation Assay. Sera-starved JSCC-2 cells were treated for 24 and

48 hours with 0.1% DMSO, 100nM afatinib+100nM vargatef (A+V), A+V with 1, 3,

5uM 4-HPR. Cells received freshly prepared media containing the drugs every 24 hours.

Cells were trypsinized, and their viability and number were quatified using Trypan Blue

Exclusion assay.

Immunohistochemistry. Twelve primary or metastatic human OSCC tumor samples were obtained from the Ohio State University Comprehensive Cancer Center Tissue

Procurement Services (in accordance with Ohio State University IRB approval) and were immediately placed in 10% neutral buffered formalin (24-48 h). The histopathology reports containing information of patients’ gender, age, clinical sites, etc. were provided by the Tissue Procurement Services. One histologically normal and one ulcerated nonneoplastic oral mucosal tissues were randomly chosen from achieved oral pathology biopsy samples (IRB approved). Paraffin embedded tissue sections were deparaffinized, rehydrated. Endogenous peroxidase activity was blocked by pretreatment with 3% H2O2, followed by microwaved in 10mM citrate buffer (pH 9.0) for antigen retrieval. Sections were then blocked with 5% normal serum, 1% bovine serum albumin (Sigma-Aldrich),

0.05% Tween® 20 (Fisher Scientific, Pittsburgh, PA) in PBS for 1 h and incubated with

Phospho-Stat3 rabbit monoclonal antibody (1:25, Cell Signaling Tec., Danvers, MA) ,

Phospho-EGF receptor rabbit monoclonal antibody (1:200, Cell Signaling Tec., Danvers,

MA) antibodies or rabbit IgG isotype control (negative control) at 4oC overnight, followed by incubation with biotinylated secondary antibodies and Vectastain ABC reagent (Vector Laboratories, Burlingame, CA). Immunoreactions were visualized using

DAB substrate, followed by hematoxylin counterstaining. Images were obtained by using an Olympus BX51 microscope (Olympus, Japan) and Nikon DS-Fi1 digital camera

(Nikon, Japan).

Western Blotting. HOK3437 E6/E7 and OSCC cells were harvested by using trypsin

(GIBCO, Life Technologies Grand Island, NY). Total cellular protein was extracted from the cell pellets by RIPA protein extraction reagent (Pierce, Rockford, IL) and 1x Halt protease/phosphatase inhibitor (Pierce, Rockford, IL). Western blot analyses were then conducted using the previously described method [19]. The antibodies and working dilutions were as follows: p-EGFR rabbit monoclonal antibody (1:1000 Cell Signaling

Technologies), EGFR rabbit monoclongal antibody (1:2000 Cell Signaling

Technologies), p-STAT3 rabbit monoclongal antibody (1:1000 Cell Signaling

Technologies), STAT3 rabbit monoclongal antibody (1:2000 Cell Signaling

Technologies), Erk1/2 mouse monoclonal antibody (1:2000, Cell Signaling Tec.), and phospho-Erk1/2 rabbit polyclonal antibody (1:1000, Cell Signaling Tec.), β-actin mouse monoclonal antibody (1:10000 Santa Cruz Biotechnology), GAPDH rabbit monoclongal antibody (1:10000 Cell Signaling Technologies). Kodak 1D3 image analysis software

(Kodak) was employed to perform densitometric analyses. Data were normalized relative to protein levels of β-actin or GAPDH.

Survey of OSCC signaling pathways. OSCC Cell lines were pretreated for 1 hour with 0.1% DMSO (vehicle control), 100nM afatinib, 100nM vargatef, or 100nM afatinib

+ 100nM vargatef. The cells in each treatment group were subsequently stimulated for

20 minutes with vehicle (1μl ddH2O), 50ng/ml EGF, 50ng/ml VEGF, or 50ng/ml EGF +

50ng/ml VEGF. The standard Western Blot protocol was followed to determine levels of p-EGFR, pERK1/2, and STAT3. The data were normalized to GAPDH.

Fenretinide (4-HPR)’s effects on EGFR and STAT3 signaling. SCC2095sc cells were pre-treated for 24 hours prior to treatment. The cells were then treated with (1) vehicle (0.1% DMSO), (2) 5uM 4-HPR, or (3) 100nM afatinib for 24, 48, or 72 hours.

The standard Western Blot protocol was followed to determine intracellular levels of p-

EGFR, EGFR, p-STAT3, and STAT3. The data were normalized to GAPDH.

Evaluation of fenretinide (4-HPR)’s effects on EGFR and IL-6 signaling pathways. JSCC-3 cells were sera-starved for 24 hours. JSCC-3 cells were either stimulated with 50ng/mL EGF or 50ng/mL IL-6 for 0, 5, 15, 30, 60, or 120 minutes with vehicle (0.1% DMSO) or 5uM 4-HPR treatment. Western Blot protocol was followed to determine intracellular levels of p-STAT3 and STAT3.

Evaluation of combined therapeutic effects of fenretinide (4-HPR) and small molecule EGFR and triple angiogenic kinase inhibitors (i.e. afatinib and vargatef).

JSCC-2 cells were seeded and sera-starved for 24 hours prior to 24-hour treatment with

0.1% DMSO, 1, 3, or 5uM 4-HPR, 100nM afatinib+100nM vargatef (A+V), A+V with 1,

3, 5uM 4-HPR. Western Blot protocol was followed to determine intracellular levels of p-STAT3.

Enzyme-Linked ImmunoSorbent Assay (ELISA)

Determination of autogenous IL-6 and EGF levels. OSCC cells were seeded and sera-starved for 24 hours. At the end of the incubation period, IL-6 or EGF levels in the

88 conditioned media were detected using respective DuoSet ELISA Development Kits

(R&D Systems, Minneapolis, MN). Protein levels were normalized as pg/104 cells.

Evaluation of fenretinide (4-HPR)’s effects on IL-6 production. JSCC-1 and

JSCC-2 cells were seeded and sera-starved for 24 hours prior to fenretinide treatment.

Treatment groups include No-treatment control, vehicle control (0.1% DMSO), 1/3/5/

7/10uM 4-HPR, 100nM afatinib, 100nM vargatef, or 100nM afatinib + 100nM vargatef.

The cells were treated for 24 hours. IL-6 levels in the culture media were detected using

IL-6 DuoSet ELISA Development Kits (R&D Systems, Minneapolis, MN). Protein levels were normalized as pg/104 cells.

Evaluation of fenretinide (4-HPR) and freeze-dried black raspberry extract

(BRB)’s effects on vacular endothelial growth factor (VEGF) production. HOK3437

E6/E7 and SCC2095sc were seeded and sera-starved for 24 hours prior to fenretinide treatment. Treatment groups include vehicle control (0.1% DMSO), 1uM and 5uM 4-

HPR, BRB (10uM cyanidin rutinoside equivalent), and 1uM/5uM 4-HPR and BRB. The cells were treated for 24 hours. VEGF levels in the culture media were detected using

VEGF DuoSet ELISA Development Kits (R&D Systems, Minneapolis, MN). Protein levels were normalized as pg/104 cells.

Statistical analyses: Initial analyses confirmed that all data sets demonstrated a Gaussian distribution. One-way ANOVA followed by Bonferroni’s multiple comparisons post hoc test was used for cell viability, proliferation, and ELISA experiments.

3.4. Results:

Authentication and phenotypic characterization. Short tandem repeats (STR) profiling analyses show that all cell lines (i.e. HOK3437 E6/E7, JSCC-1, JSCC-2, JSCC-3, and

SCC2095sc) are human in origin and void of any cross contamination. JSCC-1, JSCC-2, and JSCC-3 co-express pan-cytokeratin and vimentin, the features indicative of epithelial-to-mesenchymal transition (See Figure 2.7).

Optimizing in vitro dosing levels for 4-HPR, afatinib, and vargatef combination treatments. Previous cell viability studies have demonstrated that 1uM and 5uM 4-HPR alone does not reduce cell viability in OSCC cell lines (See Figure 2.10). In addition, preliminary proliferation (BrdU) and cell viability (WST-1) studies established the optimal dosing level for combined afatinib (100nM)/vargatef (100nM) treatments (data not shown). The current study showed that the fenretinide/afatinib/vargatef combination did not decrease cell viability except when treated with the highest dosing combination

(i.e. 100nM afatinib+100nM vargatef+5uM) for 48 hours (Figure 3.1.). Such dosing, in fact, results in cytotoxic effects, as evidenced by the cell number reduction below the starting cell number. Afatinib and vargatef cocktail with 3uM fenretinide appeared to exert anti-proliferative effects without causing overt toxicity.

Extensive inter-tumor heterogeneity in specific growth factors and treatment response profile in OSCC tumor-derived cells. The inhibitor response profile and baseline phosphorylation of EGFR, ERK1/2, and STAT3 was cell-line dependent (see Figure 3.2. for overview of responsiveness to growth factor and inhibitor treatments). The JSCC1 line exhibited negligible levels of pEGFR and pERK1/2 when stimulated with vehicle or

VEGF, but responded to EGF by phosphorylating EGFR and ERK1/2. The moderate intensity of pSTAT3 at baseline remained unchanged with all stimulations parameters.

Vargatef treatment did not affect the activation of downstream signaling in EGF- stimulated samples, while afatinib completely blocked phosphorylation of EGFR and

ERK1/2 but did not modulate STAT3 activation. Interestingly, VEGF stimulation in combination with vargatef resulted in phosphorylation of EGFR, suggesting the possibility of compensatory signaling pathways. This result was also observed (data not shown) in SCC2095sc, which is a highly pro-angiogenic, VEGF- responsive OSCC cell line [19].

The JSCC2 line exhibited high autologous EGFR and ERK1/2 activation, which were increased in intensity when stimulated with EGF. The high intensity of pSTAT3 at baseline remained unchanged with all stimulation and inhibitor treatments. Interestingly, vargatef treatment alone did not affect EGF-stimulated samples, but decreased pERK1/2 in VEGF-stimulated samples (i.e., again, suggesting compensatory signaling pathways).

Furthermore, in EGF-stimulated samples, afatinib treatment alone decreased the intensity of pEGFR and pERK1/2, but combination afatinib and vargatef treatment had an additive

91 effect leading to substantial inhibition of pEGFR and complete inhibition of pERK1/2.

These collective treatments, however, did not affect pSTAT3 levels.

The JSCC3 line exhibited high autologous levels of pEGFR, minimal pERK1/2, and negligible pSTAT3; all of which were increased in intensity when stimulated with

EGF, but non-responsive to VEGF-stimulation. Afatinib treatment alone lowered the levels of pEGFR and pERK1/2, while pSTAT3 was completely inhibited by afatinib.

Combination treatment with afatinib and vargatef again had an additive effect in blocking pEGFR, pERK1/2, and pSTAT3.

Inter-tumor heterogeneity extends to unique cytokine profiles among OSCC cell lines.

Similar to the signaling profiles, the cytokine and growth factor secretion profile of each cell line was unique. Of the 102 screened soluble proteins, only 4 (i.e. dickkopf-1, interleukin-8, macrophage migration inhibitory factor, and plasminogen activator inhibitor-1/serpin E1) were shared by JSCC1, JSCC2, and JSCC3 (Figure 3.3.).

Interestingly, 3 pro-angiogenic cytokines (i.e. angiogenin, CXCL1, and PDGF-AA) were secreted by JSCC1 and JSCC2 but not JSCC3. ELISA assays are ongoing to validate these data. Additional proteome profiling assays are underway to assess phosphorylation state of tyrosine kinase receptors and downstream signaling.

Determination of autologous IL-6 and EGF levels in OSCC cell lines. JSCC-1, JSCC-2, and SCC2095sc cells autogenously produce appreciable levels of IL-6. The highest level

92 of IL-6 is secreted by JSCC-2 cells, followed by SCC2095sc and JSCC-1. These OSCC cell lines produce IL-6 levels comparable to those reported in the literature [20]. JSCC-1

IL-6 levels determined by ELISA (Figure 3.4) seemingly contradict a lack of JSCC-1 IL-

6 production from the cytokine array (Figure 3.3). ELISA provides quantitative measures of IL-6 levels, and the IL-6 level in JSCC-1 may have been under the detection threshold in the cytokine array. All OSCC lines tested (i.e. JSCC-1, JSCC-2, JSCC-3, and

SCC2095sc) produce negligible amounts of EGF.

4-HPR abrogates EGFR and STAT3 activation. SCC2095sc show constitutive activation of both EGFR and STAT3 (Figure 3.5.). 72-hour time course study revealed that 4-HPR inhibited STAT3 phosphorylation (tyrosine 705) in SCC2095sc in a time- dependent manner. Afatinib, a small molecule EGFR inhibitor, did not block STAT3 phosphorylation despite its complete inhibition of EGFR phosphorylation. 4-HPR attenuated EGFR phosphorylation in a time-dependent manner. 4-HPR and afatinib treatment did not result in alteration of intracellular EGFR and STAT3 levels. phosphorylation status rather than modulating intracellular levels of EGFR.

4-HPR blocks STAT3 signaling via EGFR signaling pathway. Immunoblotting data demonstrates the absence of constitutive STAT3 activation in JSCC-3 cells (Figure 3.2 and Figure 3.7.). This cellular feature enabled testing of 4-HPR’s capacity to modulate inducible STAT3 activation. Both EGF and IL-6 signaling play a key role in OSCC

93 carcinogenesis and bind to their putative receptors to activate STAT3. EGF (50ng/mL) and IL-6 (50ng/mL) stimulation resulted in robust induction of STAT3 phosphorylation at Y705. Whereas IL-6-induced STAT3 phosphorylation virtually disappeared by the end of 2-hour time period, EGF-induced phospho-STAT3 signal remained strong (Figure

3.7.). Notably, 5uM fenretinide treatment significantly attenuated STAT3 activation in

EGF-stimulated samples. Fenretinide, however, exerted minimal inhibitory effects on IL-

6-induced STAT3 phosphorylation. Densitometric analyses of phospho-STAT3 normalized to total STAT3 confirmed this observation.

Inhibition of IL-6 production by interfering with STAT3 signaling. Consistent with 4-

HPR’s inhibitory effects on STAT3 phosphorylation in JSCC-3 and SCC2095sc, 5uM 4-

HPR significantly inhibits STAT3 phosphorylation. Notably, the combined treatment of

100nM afatinib/100nM vargatef/fenretinide (4-HPR+A+V) achieves the comparable level of STAT3 inhibition at a lower fenretinide dose (3uM). Given the cell viability result (See Figure 3.1.), 3uM 4-HPR+A+V dosing level shows the best potential of inhibiting STAT3 activation without causing overt toxicity. Notably, STAT3 activation results in homo-/hetero-dimerization of STAT3, nuclear translocation, and transactivation of pro-tumorigenic genes, such as IL-6 and VEGF [11-13]. As such, IL-6 levels can serve as a surrogate marker for STAT3 activation. In fact, 4-HPR is capable of attenuating IL-6 production in a dose-dependent manner. Afatinib, despite its inability to block EGFR signaling in JSCC-2 cells, inhibited IL-6 production at a level comparable to 5uM 4-

HPR, whereas vargatef did not. The combined afatinib and vargatef treatment without

94 fenretinide produced the same inhibitory effects as afatinib alone, confirming vargatef’s lack of effects on IL-6 production.

Freeze-dried black raspberry extract (BRB) abrogates fenretinide-induced VEGF production. Inhibition of STAT3 phosphorylation and IL-6 production by fenretinide prompted an investigation on another key signaling mediator under the control of

STAT3. Unexpectedly, fenretinide treatment increased VEGF production in both

HOK3437 E6/E7 (HPV E6/E7-transduced) and SCC2095sc cell lines (Figure 3.8.A).

Given that reactive oxygen species, well established signaling mediators employed by fenretinide [21], upregulates VEGF expression [22], HOK3437 E6/E7 and SCC2095sc cell lines were concurrently treated with fenretinide and BRB, an promising anti-oxidant chemopreventive that has been under investigation in our laboratory. The results demonstrate that BRB alone can abrogate VEGF production in both cell lines, which is consistent with the previous studies [23]. In addition, BRB is capable of suppressing 4-

HPR-induced VEGF production to the level achieved by BRB alone. Interestingly, BRB by and large does not inhibit 4-HPR’s capacity to induce apoptosis via caspase 3/7 activation (Figure 3.8.B).

Disclaimer: “Extensive inter-tumor heterogeneity in specific growth factors and treatment response profile in OSCC tumor-derived cells (Figure 3.2)” and “Inter-tumor heterogeneity extends to unique cytokine profiles among OSCC cell lines (Figure 3.3)”

95 have been previously included in the dissertation of Dr. Andrew S. Holpuch. The author of this dissertation has significantly contributed to conceptual development and interpretation of these initial data, which are integral part of the manuscript in preparation. As such, they are included as part of this dissertation chapter to facilitate the logical flow of the story.

3.5. Discussion:

Extensive intra-/inter-lesional heterogeneity of OSCC tumors provides the rationale for biochemical and molecular profiling of OSCC. Tumor profiling would allow specific targeting of key tumorigenic signaling pathways during inductive chemotherapy to increase clinical efficacy and enable tissue-sparing of vital structures. Assuming secondary tumors retain molecular and biochemical characteristics of the primary, tumor profiling could also contribute to secondary chemoprevention by targeting progenitor cells that have undergone tumor initiation.

Prolonged serial passage of in vitro cell cultures introduce genetic (e.g. mutations, translocation) and epigenetic (e.g. CpG methylation patterns) instability [24-25].

Recently isolated, low-passage (less than 20) JSCC cells lines, therefore, provide excellent cell culture models for in vitro evaluation of therapeutic efficacy as they closely recapitulate original tumor genotypes and phenotypes. Heterogeneity in signaling pathways and cytokine secretion profiles reflect inter-tumor heterogeneity observed in

OSCC (Figure 3.2 and 3.3). In addition, inter-tumor homogeneity among OSCC cell

96 lines (i.e. Dickkopf-1, interleukin-8, macrophage migration inhibitory factor, and serpin

E1) provide potential generic predictive markers or therapeutic targets [26-29]. The ongoing study of immunohistochemical analyses demonstrated that JSCC cell lines share activation of key tumorigenic signaling pathways (i.e. EGFR and STAT3) found in matched tumor tissues (data not shown).

A majority of OSCC tumors over-express epidermal growth factor receptors

(EGFRs), and persistently activated EGFR signaling pathways contribute to OSCC carcinogenesis [30-31]. Epidermal growth factors are mainly supplied by stromal cells

(e.g. endothelial cells and macrophages) [32] although some carcinoma cells are known to produce their own EGF [33]. Notably, none of the OSCC cell lines tested showed autogenous EGF levels (See Figure 3.4). JSCC-2, 3, and SCC2095sc, however, show robust constitutive activation of EGFR (See Figure 3.2 and 3.5). As EGFR signaling plays a key role in OSCC development, many OSCC tumors circumvent co-dependent tumor-stroma relationships by acquiring constitutively activated EGF receptors (via mutations, exon deletions, or receptor truncation) and/or increasing EGFR expression

[30]. High prevalence and importance of EGFR signaling led to the development of small molecule inhibitors of EGFR , such as cetuximab [34].

FDA-approved EGFR inhibitor (i.e. Cetuximab), however, demonstrated limited clinical efficacy due to acquired drug resistance caused by compensatory signaling pathways and EGFR mutations [9, 35, 36]. Figure 3.2 and 3.5 clearly recapitulate this issue in JSCC-1, 2, and SCC2095sc. Afatinib, another EGFR inhibitor, consistently inhibits EGFR phosphorylation in all 4 cell lines (i.e. JSCC-1, 2, 3, SCC2095sc).

Afatinib, however, was not able to inhibit phosphorylation of STAT3, one of the cellular mediators implicated in EGFR chemoresistance, in JSCC-1, 2, and SCC2095sc (Figure

3.2 and 3.5) [9].

STAT3 activation can also occur via IL-6R/gp130/JAK [11]. Interleukin 6, a putative ligand for IL-6R, is a potential prognostic marker for head and neck squamous cell carcinoma (HNSCC) [37]. Interestingly, only JSCC-1, 2, and SCC2095sc cell lines, which show resistance to EGFR inhibition-mediated STAT3 dephosphorylation, produced high levels of autogenous IL-6 (See Figure 3.4). We speculate JSCC-1, JSCC-

2, and SCC2095sc are capable of IL-6R-mediated JAK/STAT3 activation via autologous

IL-6 secretion. In contrast, STAT3 activation largely depends on EGFR signaling in

JSCC-3 as JSCC-3 cells shows the comparatively negligible capacity to autogenously produce IL-6 (See Figure 3.4).

JSCC-2, one of the chemoresistant cell lines which exhibited the most robust

EGFR and STAT3 activation profile, therefore, was used as the model cell to test 4-

HPR’s chemosensitizing capacity (See Figure 3.6). 5uM 4-HPR abrogated STAT3 phosphorylation in JSCC-2 as a single agent or in combination with afatinib/vargatef

(Figure 3.6.A). Consistent with this result, 5uM 4-HPR attenuated STAT3 phosphorylation in a time-dependent manner over 72 hours (Figure 3.5). Collectively, 4-

HPR demonstrated its capacity to inhibit STAT3 activation in JSCC-2 and SCC2095sc cells, which are high IL-6 producers and capable of constitutively activating the STAT3 pathway. 3uM 4-HPR/afatinib/vargatef combination treatment achieves the similar

STAT3 inhibition with the higher single 4-HPR dose (5uM) in JSCC-2 (Figure 3.6.A),

98 potentially suggesting lower individual dosing levels to achieve comparable therapeutic efficacy.

STAT3 forms homo-/hetero-dimers upon activation via Y705 phosphorylation, translocates into the nucleus, and regulates transcription of several pro-inflammatory/pro- angiogenic genes (e.g. IL-6 and VEGF) [11, 38]. Inhibition of STAT3 activation, therefore, should result in a decrease in IL-6 production. Consistent with this prediction,

4-HPR reduces extracellular IL-6 production in a dose-dependent manner (See Figure

3.6.B). Interestingly, afatinib, which did not decrease STAT3 phosphorylation in JSCC-2 cells, also appreciably decreased IL-6 levels, suggesting regulation of IL-6 expression via

EGFR-dependent/STAT-3-independent mechanism.

Figure 3.2 showed that JSCC-3 does not exhibit constitutive p-STAT3 phosphorylation and can be induced by exogenous stimulant, such as EGF. JSCC-3 was, therefore, used as a model cell line to determine the molecular mechanism behind 4-

HPR’s p-STAT3 inhibitory effects (Figure 3.7). EGF and IL-6 stimulation initially led to comparable levels of p-STAT3 activation (at 5 minutes) regardless of 4-HPR treatment

(Figure 3.7.). 5uM fenretinide treatment subsequently attenuated STAT3 phosphorylation resulting from EGF, whereas fenretinide exerts virtually no inhibitory effects on IL-6 stimulated STAT3 phosphorylation. Several plausible explanations exist:

(1) 4-HPR blocks STAT3 autophosphorylation, (2) 4-HPR augments phosphatase activity

(e.g. protein tyrosine phosphatase receptor T) responsible for STAT3 dephosphorylation at Y705, (3) inhibiting c-Src mediated phosphorylation of STAT3. These potential mechanisms are not mutually exclusive. Inhibition of c-Src kinase activity, however,

99 provides the most compelling explanation since EGFR-mediated activation of STAT3 necessitates c-Src kinase activity [39], and our data supports 4-HPR’s STAT3 inhibition by blocking EGFR signaling (Figure 3.7).

Studies have demonstrated that vascular endothelial growth factor (VEGF), a key angiogenic molecule, is under the regulation of STAT3 trasncriptional activity. Contrary to our hypothesis, 4-HPR resulted in increased production of VEGF (Figure 3.8.A).

Previous studies demonstrated an increase in intracellular reactive oxygen species (ROS) could induce VEGF expression by increasing hypoxia inducible factor 1alpha (HIF-

1alpha) activity [22]. VEGF level increase, therefore, could be attributed to 4-HPR’s potent pro-oxidant potential [40]. Notably, fenretinide’s another pro-apoptotic effects involve uncoupling of mitochondrial oxidative phosphorylation chain via reactive oxygen species generation [40]. Interestingly, concurrent treatment with 4-HPR and BRB, an antioxidant chemopreventive, largely did not compromise fenretinide’s ability to induce apoptosis (Figure 3.8.B), whereas the combined treatment significantly inhibited extracellular VEGF production (Figure 3.8.A). Two plausible explanations that are not mutually exclusive can be given: (1) Evidence shows that BRB is capable of downregulating VEGF expression by blocking the PI3K/Akt/AP-1 pathway [23]; (2)

Target sites for BRB’s anti-oxidant effects (i.e. cytosol) differ from the primary location of 4-HPR’s pro-oxidant effects (i.e. mitochondria).

Tumor profiling of OSCC cells reveals inter-tumor heterogeneity reflected in signaling pathways and unique cytokine secretion profiles. In addition, varied therapeutic response to small molecule inhibitors confirm that OSCC cells are capable of maintaining

100 pro-growth/pro-survivals via compensatory pathways despite inhibition of key pro- tumorigenic pathways. Our data supports the premise of using 4-HPR, either as a single agent or in combination with other small molecule inhibitors, to abrogate compensatory pathways and potentially re-sensitize cells to specific inhibitors.

3.6. References 1. Menzin J, Lines LM, Manning LN. The economics of squamous cell carcinoma of the head and neck. Curr Opin Otolaryngol Head Neck Surg. 15: 68-73, 2007. PMID: 17413405 2. Lefebvre JL. Current clinical outcomes demand new treatment options for SCCHN. Ann Oncol. 16: vi7-vi12, 2005. PMID: 15987995 3. Lang K, Menzin J, Earle CC, Jacobson J, Hsu M-A. The Economic Cost of Squamous Cell Cancer of the Head and Neck. Findings from Linked SEER-Medicare Data. Arch Otolaryngol Head Neck Surg. 130: 1269-1275, 2004. PMID: 15545580 4. Siegel RL, Miller KD, Jemal A. Cancer Statistics, 2015. CA Cancer J Clin. 2015, 65:5-29. 5. Stransky N, Egloff AM, Tward AD, Kostic AD, Cibulskis K, Sivachenko A, et al. The mutational landscape of head and neck squamous cell carcinoma. Science. 2011 Aug 26;333(6046):1157-60. 6. Severino P, Alvares AM, Michaluart P Jr, Okamoto OK, Nunes FD, Moreira-Filho CA, Tajara EH; Head and Neck Genome Project GENCAPO. Global gene expression profiling of oral cavity cancers suggests molecular heterogeneity within anatomic subsites. BMC Res Notes. 2008 Nov 13;1:113. 7. Wang X, Fan M, Chen X, Wang S, Alsharif MJ, Wang L, Liu L, Deng H. Intratumor genomic heterogeneity correlates with histological grade of advanced oral squamous cell carcinoma. Oral Oncol. 2006 Aug;42(7):740-4. 8. Sun C, Bernards R. Feedback and redundancy in receptor tyrosine kinase signaling: relevance to cancer therapies. Trends Biochem Sci. 2014 Oct;39(10):465-74. 9. Sen M, Joyce S, Panahandeh M, Li C, Thomas SM, Maxwell J, Wang L, Gooding WE, Johnson DE, Grandis JR. Targeting Stat3 abrogates EGFR inhibitor resistance in cancer. Clin Cancer Res. 2012 Sep 15;18(18):4986-96. 10. Kijima T, Niwa H, Steinman RA, Drenning SD, Gooding WE, Wentzel AL, Xi S, Grandis JR. STAT3 activation abrogates growth factor dependence and contributes to head and neck squamous cell carcinoma tumor growth in vivo. Cell Growth Differ. 2002 Aug;13(8):355-62.

101

11. Yu H, Lee H, Herrmann A, Buettner R, Jove R. Revisiting STAT3 signalling in cancer: new and unexpected biological functions. Nat Rev Cancer. 2014 Nov;14(11):736-46. 12. Kim SM, Kwon OJ, Hong YK, Kim JH, Solca F, Ha SJ, Soo RA, Christensen JG, Lee JH, Cho BC. Activation of IL-6R/JAK1/STAT3 signaling induces de novo resistance to irreversible EGFR inhibitors in non-small cell lung cancer with T790M resistance mutation. Mol Cancer Ther. 2012 Oct;11(10):2254-64. 13. Xiong A, Yang Z, Shen Y, Zhou J, Shen Q. Transcription Factor STAT3 as a Novel Molecular Target for Cancer Prevention. Cancers (Basel). 2014 Apr 16;6(2):926-57. 14. Yue P, Turkson J. Targeting STAT3 in cancer: how successful are we? Expert Opin Investig Drugs. 2009 Jan;18(1):45-56.\ 15. Grivennikov SI, Karin M. Dangerous liaisons: STAT3 and NF-kappaB collaboration and crosstalk in cancer. Cytokine Growth Factor Rev. 2010 Feb;21(1):11-9. 16. Gupta SC, Kannappan R, Reuter S, Kim JH, Aggarwal BB. Chemosensitization of tumors by resveratrol. Ann N Y Acad Sci. 2011 Jan;1215:150-60. 17. Han BB, Li S, Tong M, Holpuch AS, Spinney R, Wang D, et al. Fenretinide Perturbs Focal Adhesion Kinase in Premalignant and Malignant Human Oral Keratinocytes. Fenretinide's chemopreventive mechanisms include ECM interactions. Cancer Prev Res (Phila). 2015 Feb 24. pii: canprevres.0418.2014. [Epub ahead of print] PubMed PMID: 25712051. 18. Tong M, Han BB, Holpuch AS, Pei P, He L, Mallery SR. Inherent phenotypic plasticity facilitates progression of head and neck cancer: Endotheliod characteristics enable angiogenesis and invasion. Exp Cell Res 2013; 319: 1028-1042. 19. M. Tong, B. Lloyd, P. Pei, S.R. Mallery. Human head and neck squamous cell carcinoma cells are both targets and effectors for the angiogenic cytokine, VEGF. J. Cell Biochem. 105 (2008) 1202- 1210. 20. Chen MF, Chen PT, Lu MS, Lin PY, Chen WC, Lee KD. IL-6 expression predicts treatment response and outcome in squamous cell carcinoma of the esophagus. Mol Cancer. 2013 Apr 5;12:26. doi: 10.1186/1476-4598-12-26. 21. Oridate N, Suzuki S, Higuchi M, Mitchell MF, Hong WK, Lotan R. Involvement of reactive oxygen species in N-(4-ydroxyphenyl)retinamide-induced apoptosis in cervical carcinoma cells. J Natl Cancer Inst. 1997 Aug 20;89(16):1191-8. 22. Xia C, Meng Q, Liu LZ, Rojanasakul Y, Wang XR, Jiang BH. Reactive oxygen species regulate angiogenesis and tumor growth through vascular endothelial growth factor. Cancer Res. 2007 Nov 15;67(22):10823-30. 23. Huang C, Li J, Song L, Zhang D, Tong Q, Ding M, Bowman L, Aziz R, Stoner GD. Black raspberry extracts inhibit benzo(a)pyrene diol-epoxide-induced activator protein 1 activation and

102

VEGF transcription by targeting the phosphatidylinositol 3-kinase/Akt pathway. Cancer Res. 2006 Jan 1;66(1):581-7. 24. Geraghty RJ, Capes-Davis A, Davis JM, Downward J, Freshney RI, Knezevic I, et al. Cancer Research UK. Guidelines for the use of cell lines in biomedical research. Br J Cancer. 2014 Sep 9;111(6):1021-46. doi: 10.1038/bjc.2014.166. Epub 2014 Aug 12. 25. Wenger SL, Senft JR, Sargent LM, Bamezai R, Bairwa N, Grant SG. Comparison of established cell lines at different passages by karyotype and comparative genomic hybridization. Biosci Rep. 2004 Dec;24(6):631-9. 26. Watanabe H, Iwase M, Ohashi M, Nagumo M. Role of interleukin-8 secreted from human oral squamous cell carcinoma cell lines. Oral Oncol. 2002 Oct;38(7):670-9. 27. Magnussen S, Rikardsen OG, Hadler-Olsen E, Uhlin-Hansen L, Steigen SE, Svineng G. Urokinase plasminogen activator receptor (uPAR) and plasminogen activator inhibitor-1 (PAI-1) are potential predictive biomarkers in early stage oral squamous cell carcinomas (OSCC). PLoS One. 2014 Jul 7;9(7):e101895. 28. He XX, Chen K, Yang J, Li XY, Gan HY, Liu CY, Coleman TR, Al-Abed Y. Macrophage migration inhibitory factor promotes colorectal cancer. Mol Med. 2009 Jan-Feb;15(1-2):1-10. 29. Wang J, Shou J, Chen X. Dickkopf-1, an inhibitor of the Wnt signaling pathway, is induced by p53. Oncogene. 2000 Mar 30;19(14):1843-8. 30. Kalyankrishna S, Grandis JR. Epidermal growth factor receptor biology in head and neck cancer. J Clin Oncol. 2006 Jun 10;24(17):2666-72. Review. 31. Grandis JR, Tweardy DJ. Elevated levels of transforming growth factor alpha and epidermal growth factor receptor messenger RNA are early markers of carcinogenesis in head and neck cancer. Cancer Res. 1993 Aug 1;53(15):3579-84. 32. Zhang Z, Dong Z, Lauxen IS, Filho MS, Nör JE. Endothelial cell-secreted EGF induces epithelial to mesenchymal transition and endows head and neck cancer cells with stem-like phenotype. Cancer Res. 2014 May 15;74(10):2869-81. 33. Rajagopal S, Huang S, Moskal TL, Lee BN, el-Naggar AK, Chakrabarty S. Epidermal growth factor expression in human colon and colon carcinomas: anti-sense epidermal growth factor receptor RNA down-regulates the proliferation of human colon cancer cells. Int J Cancer. 1995 Sep 15;62(6):661-7. 34. Mehra R, Cohen RB, Burtness BA. The role of cetuximab for the treatment of squamous cell carcinoma of the head and neck. Clin Adv Hematol Oncol. 2008 Oct;6(10):742-50. 35. Rebucci M, Peixoto P, Dewitte A, Wattez N, De Nuncques MA, Rezvoy N, Vautravers-Dewas C, Buisine MP, Guerin E, Peyrat JP, Lartigau E, Lansiaux A. Mechanisms underlying resistance to cetuximab in the HNSCC cell line: role of AKT inhibition in bypassing this resistance. Int J Oncol. 2011 Jan;38(1):189-200. 103

36. Montagut C, Dalmases A, Bellosillo B, Crespo M, Pairet S, Iglesias M, et al. Identification of a mutation in the extracellular domain of the Epidermal Growth Factor Receptor conferring cetuximab resistance in colorectal cancer. Nat Med. 2012 Jan 22;18(2):221-3. doi: 10.1038/nm.2609. Erratum in: Nat Med. 2012 Sep;18(9):1445.Somasekar, Seshagiri [corrected to Seshagiri, Somasekar]. PubMed PMID: 22270724. 37. Riedel F, Zaiss I, Herzog D, Götte K, Naim R, Hörmann K. Serum levels of interleukin-6 in patients with primary head and neck squamous cell carcinoma. Anticancer Res. 2005 Jul- Aug;25(4):2761-5. 38. Niu G, Wright KL, Huang M, Song L, Haura E, Turkson J, Zhang S, Wang T, Sinibaldi D, Coppola D, Heller R, Ellis LM, Karras J, Bromberg J, ardoll D, Jove R, Yu H. Constitutive Stat3 activity up-regulates VEGF expression and tumor angiogenesis. Oncogene. 2002 Mar 27;21(13):2000-8. 39. Garcia R, Bowman TL, Niu G, Yu H, Minton S, Muro-Cacho CA, Cox CE, Falcone R, Fairclough R, Parsons S, Laudano A, Gazit A, Levitzki A, Kraker A, Jove R. Constitutive activation of Stat3 by the Src and JAK tyrosine kinases participates in growth regulation of human breast carcinoma cells. Oncogene. 2001 May 3;20(20):2499-513. 40. Hail Jr N, Kin HJ, Lotan R. Mechanisms of fenretinide-induced apoptosis. Apoptosis 11: 1677- 1694, 2006.

104

Figure 3.1. Determination of optimal dosing levels for multi-agent chemopreventive strategy. 4-HPR/afatinib/vargatef combination treatment significantly inhibits cell proliferation at both 24 and 48 hours (n=3). At its highest dose, the combination treatment appears to exert cytotoxic effects. The treatment does not affect cell viability except the cytotoxic dose.

* * * * * *

105

Figure 3.2. Extensive inter-tumor heterogeneity in specific growth factor and treatment response profile in OSCC tumor-derived cells. Responsiveness to epidermal growth factor and vascular endothelial growth factor activation of downstream kinase signaling and susceptibility to EGFR inhibition (afatinib) and/or triple angiokinase inhibition (VEGFR, PDFR, bFGFR – vargatef).

106

Figure 3.3. Inter-tumor heterogeneity extends to unique cytokine profiles among

OSCC cell lines. These semi-quantitative arrays demonstrate the vast heterogeneity with regard to protein secretion by each cell line, and provide insight to their varied responses to EGF/VEGF stimulation and subsequent inhibition (as shown in Figure 5.3). As the

JSCC3 cells were completely responsive to afatinib and vargatef inhibition (i.e. p-STAT3 was completely inhibited following inhibitor treatment), the production of growth factors/cytokines by both JSCC1 and JSCC2 lines was of particular interest (i.e. possible autocrine/intracrine signaling responsible for refractory afatinib/vargatef treatment).

Interestingly, only three proteins (angiogenin, CXCL1, and PDGF-AA) were produced by

JSCC1 and JSCC2, but not JSCC3. Studies are ongoing to determine the effect of these pro-angiogenic proteins on the EGFR and VEGFR signaling pathways.

107

Figure 3.3. Inter-tumor heterogeneity extends to unique cytokine profiles among OSCC cell lines.

108

Figure 3.4. Determination of autologous IL-6 and EGF levels in OSCC cells. 24-hour Sera-starved OSCC cells were incubated at 5% CO2/37oC for 24 hours before quantification of extracellular IL-6 and EGF by ELISA. (A) JSCC-2 cells produced the highest IL-6 levels. JSCC-3 secreted the lowest amount of IL-6. JSCC-1 and SCC2095sc produced intermediate IL-6 levels. These JSCC-1, JSCC-2, and SCC2095sc IL-6 levels in cell density dependent terms are comparable to or higher than the IL-6 levels reported in other solid tumor cell lines. (B) None of the OSCC cell lines tested secreted appreciable levels of EGF. Both IL-6 and EGF levels were normalized by the cell number and expressed as pg/104 cells (n=6). Compare to other human solid tumor, preferably carcinomas, in cell density dependent terms.

109

Figure 3.4. Determination of autologous IL-6 and EGF levels in OSCC cells.

110

Figure 3.5. 4-HPR abrogates EGFR and STAT3 phosphorylation. 24-hour Sera- starved SCC2095sc were treated with 0.1% DMSO (V), 5uM 4-HPR (H), or 100nM afatinib (A) over 72 hours. EGFR phosphorylation signal became more intense over time, and 4-HPR began to attenuate p-EGFR at 48 hours and significantly inhibit p-EGFR by 72 hours. The total EGFR levels remained consistent throughout the experiment. 4-HPR treatment inhibited STAT3 phosphorylation in a time-dependent manner without altering the total STAT3 level. Afatinib blocked EGFR phosphorylation as expected but showed no inhibition on STAT3 phosphorylation. The data were normalized to intracellular GAPDH levels.

111

Figure 3.6.: Fenretinide differentially modulates STAT3 activation pathways. JSCC- 3, pSTAT3-inducible cell line, was stimulated with key cytokine (IL-6) and growth factor (EGF) involved in OSCC development. Both IL-6 and EGF generated robust STAT3 phosphorylation at 5 minutes. JSCC-3 maintained EGF-induced STAT3 phosphorylation over the course of 2 hours, whereas IL-6-induced STAT3 phosphorylation was mostly lost after 60 minutes. 5uM fenretinide treatment attenuated STAT3 phosphorylation resulting from EGF, whereas fenretinide exerts virtually no inhibitory effects on IL-6 stimulated STAT3 phosphorylation. 4-HPR treatment did not significantly alter total STAT3 levels. Densitometric analyses were used to quantify the signal intensities, and the data were normalized to either beta-actin or GAPDH.

112

Figure 3.7. 4-HPR-mediated inhibition of STAT3 signaling. 5uM 4-HPR alone or 4- HPR combined with afatinib/vargatef (A+V) inhibited STAT3 phosphorylation in JSCC- 2 cells. Whereas 3uM 4-HPR alone did not inhibit phosphorylation, 3uM 4-HPR with A+V exerted significant inhibitory effects on STAT3 phosphorylation. In addition, 4- HPR was capable of inhibiting production of IL-6, one of the genes under STAT3 transcriptional regulation, in a dose-dependent manner. 100nM afatinib also inhibits IL-6 production comparably to 5uM 4-HPR. 100nM vargatef did not alter IL-6 production. Error bars represent standard deviation (n=6).

113

Figure 3.7. 4-HPR-mediated inhibition of STAT3 signaling.

3uM 4-HPR A+V+1uMA+V+3uM 4-HPRA+V+5uM 4-HPR 4-HPR 1uM 4-HPR 5uM 4-HPR A+V A VEH

P-Stat3

GAPDH

114

Figure 3.8. Evaluation of freeze-dried black raspberry extract (BRB) and 4-HPR ‘s combined treatment effects. (A) Freeze-dried black raspberry extract (BRB) abrogates 4-HPR-induced VEGF production. 4-HPR induced a dose-dependent increase in VEGF production in both HOK3437 E6/E7 and SCC2095sc cell lines. BRB, either as a single agent or with 4-HPR, suppressed VEGF production. Error bars represent standard deviation (n=6). (B) 4-HPR induces dose-dependent caspase 3/7 activation in HOK normal and HOK3437 E6/E7 cell lines. BRB, a chemopreventive with well established antioxidant capacity, largely does not interfere with 4-HPR’s caspase 3/7 induction except in HOK3437 E6/E7, where BRB significantly, albeit only slightly, inhibits 4- HPR’s pro-apoptotic effects. Error bars represent standard deviation (n=4).

115

Figure 3.8: Evaluation of freeze-dried black raspberry extract (BRB) and 4-HPR ‘s combined treatment effects.

116

Chapter 4: Executive summary and Future Research Directions

4.1. Fenretinide Perturbs Focal Adhesion Kinase in Premalignant and Malignant Human

Oral Keratinocytes. Fenretinide’s chemopreventive mechanisms include ECM interactions.

FAK’s prosurvival, mobility-enhancing functions in a spectrum of cultured oral human keratinocytes that range from normal to HPV E6/E7-transduced to malignant to metastatic. Molecular modeling studies predict fenretinide’s binding strong binding affinity to FAK’s kinase domain, FERM domain, and PYK2, FAK’s closely related member in the FAK family. Our experimental results confirm fenretinide’s pleiotropic effects on cellular milieu. Fenretinide significantly suppressed proliferation via induction of apoptosis and G2/M cell cycle blockade. Fenretinide-treated cells also demonstrated F- actin disruption and significant inhibition of both directed migration and invasion of a synthetic basement membrane. and decreased phosphorylation of growth-promoting kinases. In addition, the proteome profiler array, which demonstrated broad modulation 117 of growth-promoting kinases, supports the notion that fenretinide acts as a broad biological modifier rather than targeting specific pathways. Collectively, this study demonstrated fenretinide’s binding affinity to intracellular proteins and its capacity to modulate multiple signaling pathways involved in tumorigenic phenotypes and provide the rationale for using fenretinide as a broad biological modifier for OSCC therapeutic management (Chapter 3).

Interestingly, combining with fenretinide an additional chemopreventive shown to be clinically effective in oral intraepithelial neoplasia (OIN) lesions, i.e. freeze dried black raspberries (BRB) provided additive invasion-inhibitory effects on cells that were less responsive to fenretinide treatment alone. Although the preliminary data accumulated so far support the therapeutic strategy of combining these two promising chemopreventives, further experiments are warranted to (1) determine any synergistic, additive, or antagonistic interactions both in vitro and in vivo and (2) optimize 4-HPR/BRB combined local delivery formulations.

4.2. Personalized Tumor Profiling: Utilizing Fenretinide (4-HPR) as a Biological

Modifier to Optimize Therapeutic Outcome

This study investigated intra-/inter-tumor heterogeneity of oral squamous cell carcinoma by characterizing specific signaling pathways and cytokine secretion profiles in OSCC 118 cells lines, which were recently isolated from primary OSCC tumors. Characterization of specific tumorigenic signaling pathways and cytokine secretion profiles revealed significant heterogeneity among recently isolated OSCC cell lines (JSCC-1, JSCC-2, and

JSCC-3). This heterogeneity, in part, manifested as variable therapeutic response to targeted inhibition of key tumorigenic pathways (e.g. EGFR). For example, afatinib,

EGFR tyrosine kinase inhibitor, was able to inhibit STAT3 phosphorylation in JSCC-3 but not in JSCC-1 and JSCC-2. Notably, fenretinide was capable of inhibiting STAT3 phosphorylation in these refractory cell lines at local delivery-achievable dose (5uM). In addition, combined treatment between fenretinide and afatinib/vargatef decreased fenretinide dose to as low as 1uM to achieve significant STAT3 inhibition. Future studies will need to examine different dosing combinations between fenretinide and afatinib/vargatef to determine if lower dosing levels for individual drugs can achieve significant therapeutic efficacy.

Finally, our data further supports therapeutic benefits of combining fenretinide with freeze dried black raspberries (BRB), as evidence by BRB’s capacity to retain fenretinide’s pro-apoptotoic effects while reducing VEGF levels induced by fenretinide.

As mentioned above, further studies, therefore, are warranted to examine drug interactions and design optimal delivery formulations.

119

4.3. Future Research Directions

Evidence presented in this dissertation collectively supports fenretinide as a promising chemopreventive and a biological modifier that could enhance therapeutic efficacy and provide an pharmacologic advantage. Future research will elucidate (1) additional mechanistic details regarding fenretinide’s effects on tumorigenic signaling pathways and interactions with the cellular/extracellular environments. The data presented in chapter 3 provide a compelling reason to investigate (1) 4-HPR’s binding to STAT3 autophosphorylation sites as well as upstream kinases of STAT3 and (2) functional ramifications of STAT3 inhibition by 4-HPR. Molecular modeling studies are ongoing to delineate the potential mechanisms for 4-HPR inhibition of STAT3. In addition, gene expression studies and chromatin immunoprecipitation studies are currently being conducted to verify the immediate consequences of STAT3 inhibition.

The ultimate goal of our laboratories’ research is to deliver clinically efficacious drug formulations to OSCC patients. As such, future research will also entail drug formulation, pharmacokinetic, and pharmacodynamics studies involving fenretinide as a single or adjunct drug. These studies, which showed fenretinide’s therapeutic capacity as a potent chemopreventive and biological modifier, affirm the need for further research to develop a safe and efficacious therapeutic compound to help our patients.

120

Bibliography

Albini A, Sporn MB. The tumour microenvironment as a target for chemoprevention. Nat Rev Cancer. 2007 Feb;7(2):139-47. Review. PubMed PMID: 17218951.

Andreasen PA, Egelund R, Petersen HH. The plasminogen activation system in tumor growth, invasion, and metastasis. Cell Mol Life Sci. 2000 Jan 20;57(1):25-40.

Benelli R, Monteghirfo S, Vene R, Tosetti F, Ferrari N. The chemopreventive retinoid 4HPR impairs prostate cancer cell migration and invasion by interfering with FAK/AKT/GSK3β pathway and β-catenin stability. Mol Cancer 2010; 9: 1-13.

Benson E, Li R, Elsele D, Fakhry C. The clinical impact of HPV tumor status upon head and neck squamous cell carcinomas. Oral Oncol. 2014; 50(6): 565-74.

Bose SC, Singh M, Vyas P, Singh M. Plasma zinc antioxidant vitamins, glutathione levels and total antioxidant activity in oral leukoplakia. Dent Res J(Isfahan). 2012 Mar;9(2):158-61.

Boya P, Morales MC, Gonzalez-Polo RA, Andreau K, Gourdier I, Perfettini JL, et al. The chemopreventive agent N-(4-hydroxyphenyl)retinamide induces apoptosis through a mitochondrial pathway regulated by proteins from the Bcl-2 family. Oncogene. 2003 Sep 18;22(40):6220-30.

Braakhuis BJM, Tabor MP, Kummer A, Leemans CR, Brakenhoff RH. A genetic explanation of Slaughter’s concept of field cancerization: evidence and clinical implications. Cancer Res. 2003; 63:1727-1730.

Brusselmans K, De Schrijver E, Heyns W, Verhoeven G, Swinnen JV.Epigallocatechin-3-gallate is a potent natural inhibitor of fatty acid synthase in intact cells and selectively induces apoptosis in prostate cancer cells. Int J Cancer. 2003 Oct 10;106(6):856-62.

Calautti E, Li J, Saoncella S, Brissette JL, Goetinck PF. Phosphoinositide 3-kinase signaling to Akt promotes keratinocyte differentiation versus death. J Biol Chem. 2005 Sep 3;280(38):32856-65.

121

Canel M, Secades P, Rodrigo JP, Cabanillas R, Herrero A, Suarez C et al. Overexpression of Focal Adhesion Kinase in Head and Neck Squamous Cell Carcinoma Is Independent of fak Gene Copy Number. Clin Cancer Res 2006; 12: 3272-3279.

Casas-Selves M, DeGregor J. How cancer shapes evolution, and how evolution shapes cancer. Evolution (N Y). 2011; 4(4): 624-634.

Chan KT, Cortesio CL, Huttenlocher A. FAK alters invadopodia and focal adhesion composition and dynamics to regulate breast cancer invasion. J Cell Biol. 2009 Apr 20;185(2):357-70. doi: 10.1083/jcb.200809110. Epub 2009 Apr 13.

Chang KW, Lee TC, Yeh WI, Chung MY, Liu CJ, Chi LY, et al. Polymorphism in heme oxygenase-1 (HO-1) promoter is related to the risk of oral squamous cell carcinoma occurring on male areca chewers. Br J Cancer. 2004; 91(8):1551-5.

Chen D, Wan SB, Yang H, Yuan J, Chan TH, Dou QP. EGCG, Green Tea Polyphenols and their Synthetic Analogs and Prodrugs for Human Cancer Prevention and Treatment. Adv Clin Chem 2011; 53: 155-177.

Chen MF, Chen PT, Lu MS, Lin PY, Chen WC, Lee KD. IL-6 expression predicts treatment response and outcome in squamous cell carcinoma of the esophagus. Mol Cancer. 2013 Apr 5;12:26. doi: 10.1186/1476-4598-12-26.

Chen Y, Chen C. DNA copy number variation and loss of heterzygosity in relation to recurrence of and survival from head and neck squamous cell carcinoma: a review. Head Neck. 2008; 30(10):1361-83.

Chen S, Samuel W, Fariss RN, Duncan T, Kutty RK, Wiggert B. Differentiation of human retinal pigment epithelial cells into neuronal phenotype by N-(4- hydroxyphenyl)retinamide. J Neurochem. 2003 Mar;84(5):972-81.

Chiesa F, Tradati N, Grigolato R, Boracchi P, Biganzoli E, Crose N, et al. Randomized trial of fenretinide (4-HPR) to prevent recurrences, new localizations and carcinomas in patients operated on for oral leukoplakia: long-term results. Int J Cancer 2005; 115:625-29.

Chimal-Monroy J, Abarca-Buis RF, Cuervo R, Díaz-Hernández M, Bustamante M, Rios- Flores JA, et al. Molecular control of cell differentiation and programmed cell death during digit development. IUBMB Life. 2011 Oct;63(10):922-9.

Chiu WT, Shen SC, Chow JM, Lin CW, Shia LT, Chen TC. Contribution of reactive oxygen species to migration/invasion of human glioblastoma cells U87 via ERK- dependent COX-2/PGE2 activation. Neurobiol Dis 2010; 37:118-129.

122

Cicchini C, Laudadio I, Citarella F, Corazzari M, Steindler C, Conigliaro A, et al. TGFbeta-induced EMT requires focal adhesion kinase (FAK) signaling. Exp Cell Res. 2008; 314(1):143-52.

Cotter TG. Nat Rev Cancer. Nat Rev Cancer. 2009 Jul;9(7):501-7. 2009 Jul;9(7):501-7.

Crew KD, Gammon MD, Terry MB, Zhang FF, Agrawal M, Eng SM, et al. Genetic polymorphisms in the apoptosis-associated genes FAS and FASL and breast cancer risk. Carcinogenesis. 2007 Dec;28(12):2548-51. Epub 2007 Oct 24. de Vicente JC, Rodrigo JP, Rodriguez-Santamarta T, Lequerica-Fernández P, Allonca E, García-Pedrero JM. Cortactin and focal adhesion kinase as predictors of cancer risk in patients with premalignant oral epithelial lesions. Oral Oncol. 2012 Jul;48(7):641-6.

DeVita VT Jr, Chu E. A history of cancer chemotherapy. Cancer Res. 2008 Nov 1;68(21):8643-53.

Eckert RL, Efimova T, Dashti SR, Balasubramanian S, Deucher A, Crish JF, Sturniolo M, Bone F. Keratinocyte survival, differentiation, and death: many roads lead to mitogen-activated protein kinase. J Investig Dermatol Symp Proc. 2002 Dec;7(1):36-40.

Fan YL, Zheng M, Tang YL, Liang XH. A new perspective of vasculogenic mimicry: EMT and cancer stem cells (Review). Oncol Lett. 2013 Nov;6(5):1174-1180.

Ferrari N, Morini M, Pfeffer U, Minghelli S, Noonan DM, Albini A. Inhibition of Kaposi’s Sarcoma in Vivo by Fenretinide. Clin Cancer Res 2003; 9: 6020-6029.

Formelli F, Barua AB, Olson JA. Bioactivities of N-(4-hydroxyphenyl)retinamide and retinoyl beta-glucuronide. FASEB J. 1996 Jul;10(9):1014-24.

Frame MC, Patel H, Serrels B, Lietha D, Eck MJ. The FERM domain: organizing the structure and function of FAK. Nature Rev 2010; 11:802-814.

Fridman JS, Lowe SW. Control of apoptosis by p53. Oncogene. 2003 Dec 8;22(56):9030- 40.

Fulda S, Debatin KM. Extrinsic versus intrinsic apoptosis pathways in anticancer chemotherapy. Oncogene. 2006 Aug 7;25(34):4798-811.

Galipeau PC, Prevo LJ, Sanchez CA. Clonal expansion and loss of heterozygosity at chromsomes 9p and 17p in premaligant esophageal (Barrett’s) tissue. J Natl Cancer Inst. 1999; 91(24):2087-95.

123

Gander, RJ, and Gurney, JA. All-trans relinoic acid esters and amides. U.S. patent 4,108,880, August 22, 1978;Chem. Abstr. 88,89892

Garaventa A, Luksch R, LoPicolo MS, Cavadini E, Montaldo PG, Pizzitola MR et al. Phase I Trial and Pharmacokinetics of Fenretinide in Children with Neuroblastoma. Clin Cancer Res 2003; 9:2032-2039.

Garcia R, Bowman TL, Niu G, Yu H, Minton S, Muro-Cacho CA, Cox CE, Falcone R, Fairclough R, Parsons S, Laudano A, Gazit A, Levitzki A, Kraker A, Jove R. Constitutive activation of Stat3 by the Src and JAK tyrosine kinases participates in growth regulation of human breast carcinoma cells. Oncogene. 2001 May 3;20(20):2499-513.

Geraghty RJ, Capes-Davis A, Davis JM, Downward J, Freshney RI, Knezevic I, et al. Cancer Research UK. Guidelines for the use of cell lines in biomedical research. Br J Cancer. 2014 Sep 9;111(6):1021-46. doi: 10.1038/bjc.2014.166. Epub 2014 Aug 12.

Golubkov V, Garcia A, Markland FS. Action of Fenretinide (4-HPR) on Ovarian Cancer and Endothelial Cells. Anticancer Res 2005; 25: 249-253.

Grandis JR, Tweardy DJ. Elevated levels of transforming growth factor alpha and epidermal growth factor receptor messenger RNA are early markers of carcinogenesis in head and neck cancer. Cancer Res. 1993 Aug 1;53(15):3579-84.

Grivennikov SI, Karin M. Dangerous liaisons: STAT3 and NF-kappaB collaboration and crosstalk in cancer. Cytokine Growth Factor Rev. 2010 Feb;21(1):11-9.

Guillaud M, Zhang L, Poh C, Rosin MP, MacAulay C. Potential use of quantitative tissue phenotype to predict malignant risk for oral premalignant lesions. Cancer Res. 2008; 68(9):3099-3107.

Gupta SC, Kannappan R, Reuter S, Kim JH, Aggarwal BB. Chemosensitization of tumors by resveratrol. Ann N Y Acad Sci. 2011 Jan;1215:150-60.

Gurudath S, Ganapathy K, D S, Pai A, Ballal S, Ml A. Estimation of superoxide dismutase and glutathione peroxidase in oral submucous fibrosis, oral leukoplakiaand oral cancer--a comparative study. Asian Pac J Cancer Prev.2012;13(9):4409-12. PubMed PMID: 23167351.

Ha PK, Califano JA. Promoter methylation and inactivation or tumour-suppressor genes in oral squamous-cell carcinoma. Lancet Oncol. 2006; 7(1): 77-82.

Hail Jr N, Kin HJ, Lotan R. Mechanisms of fenretinide-induced apoptosis. Apoptosis 11: 1677-1694, 2006.

124

Han BB, Li S, Tong M, Holpuch AS, Spinney R, Wang D, et al. Fenretinide Perturbs Focal Adhesion Kinase in Premalignant and Malignant Human Oral Keratinocytes. Fenretinide's chemopreventive mechanisms include ECM interactions. Cancer Prev Res (Phila). 2015 Feb 24. pii: canprevres.0418.2014. [Epub ahead of print] PubMed PMID: 25712051.

Hanahan D, Weinberg RA. The hallmarks of cancer. Cell. 2000; 100(1):57-70.

Hanahan D, Weinberg RA. Hallmarks of cancer: the next generation. 2011; 144(5):646-74.

Hansen LA, Sigman CC, Andreola F, Ross SA, Kelloff GJ, De Luca LM. Retinoids in chemoprevention and differentiation therapy. Carcinogenesis. 2000 Jul;21(7):1271-9.

He XX, Chen K, Yang J, Li XY, Gan HY, Liu CY, Coleman TR, Al-Abed Y. Macrophage migration inhibitory factor promotes colorectal cancer. Mol Med. 2009 Jan-Feb;15(1-2):1-10.

Helen M. Berman, John Westbrook, Zukang Feng, Gary Gilliland, T. N. Bhat, Helge Weissig, et al. http://www.rcsb.org/pdb/home/home.do The Protein Data Bank, Nucl. Acids Res. (2000) 18 (1): 235-242. Doi: 10.1093/nar/28.1.235

Hemmer J, Kreidler J. Flow cytometric DNA ploidy analysis of squamous cell carcinoma of the oral cavity. Comparison with clinical staging and histologic grading. Cancer. 1990; 66(2): 317-20.

Ho MW, Field EA, Field JK, Risk JM, Rajlawat BP, Rogers SN, et al. Outcomes of oral squamous cell carcinoma arising from oral epithelial dysplasia: rationale for monitoring premalignant oral lesions in a multidisciplinary clinic. Br J Oral Maxillofac Surg 2013; 51:594-599.

Holpuch AS, Phelps MP, Desai KG, Chen W, Koutras GM, Han BB, et al. Evaluation of a mucoadhesive fenretinide patch for local intraoral delivery: a strategy to reintroduce fenretinide for oral cancer chemoprevention. Carcinogenesis. 2012 May;33(5):1098-105.

Holpuch AS, Desai KG, Schwendeman SP, Mallery SR. Optimizing therapeutic efficacy of chemopreventive agents: A critical review of delivery strategies in oral cancer chemoprevention clinical trials. J Carcinog. 2011;10:23.

Hong WK, Endicott J, Itri LM, Doos W, Batsakis JG, Bell R, et al. 13-cis-retinoic acid in the treatment of oral leukoplakia. N Engl J Med 1986; 315:1501-5.

Howlader N, Noone AM, Krapcho M, Garshell J, Miller D, Altekruse SF, et al. SEER Cancer Statistics Review, 1975-2011, National Cancer Institute. Bethesda, MD, 125 http://seer.cancer.gov/csr/1975_2011/, based on November 2013 SEER data submission, posted to the SEER website, April 2014. Updated September 10, 2014.

Huang C, Li J, Song L, Zhang D, Tong Q, Ding M, Bowman L, Aziz R, Stoner GD. Black raspberry extracts inhibit benzo(a)pyrene diol-epoxide-induced activator protein 1 activation and VEGF transcription by targeting the phosphatidylinositol 3- kinase/Akt pathway. Cancer Res. 2006 Jan 1;66(1):581-7.

Hwang YS, Park KK, Chung WY. Invadopodia formation in oral squamous cell carcinoma: the role of epidermal growth factor receptor signalling. Arch Oral Biol. 2012 Apr;57(4):335-43.

Ishikawa F, Ushida K, Mori K, Shibanuma M. Loss of anchorage primarily induces non-apoptotic cell death in a human mammary epithelial cell line under atypical focal adhesion kinase signaling. Cell Death Dis. 2015 Jan 22.

Jacobson JJ, Epstein JB, Eichmiller FC, Gibson TB, Carls GS, Vogtmann E, et al. The cost burden of oral, oral pharyngeal and salivary gland cancers in three groups: commercial insurance, Medicare and Medicaid. Head & Neck Oncology 2012; 4:1- 17.

Joyce JA, Pollard JW. Microenvironmental regulation of metastasis. Nat Rev Cancer. 2009 Apr;9(4):239-52. doi: 10.1038/nrc2618. Epub 2008 Mar 12.

Kalemkerian GP, Slusher R, Ramalingam S, Gadgeel S, Mabry M. Growth Inhibition and Induction of Apoptosis by Fenretinide in Small-Cell Lung Cancer Cell Lines. J Natl Cancer Inst 1995; 22: 1674-1680.

Kalluri R and Weinberg RA. The basics of epithelial-mesenchymal transition. J Clin Invest 2009; 119: 1420-1428.

Kalyankrishna S, Grandis JR. Epidermal growth factor receptor biology in head and neck cancer. J Clin Oncol. 2006 Jun 10;24(17):2666-72. Review.

Kijima T, Niwa H, Steinman RA, Drenning SD, Gooding WE, Wentzel AL, Xi S, Grandis JR. STAT3 activation abrogates growth factor dependence and contributes to head and neck squamous cell carcinoma tumor growth in vivo. Cell Growth Differ. 2002 Aug;13(8):355-62.

Kim SM, Kwon OJ, Hong YK, Kim JH, Solca F, Ha SJ, Soo RA, Christensen JG, Lee JH, Cho BC. Activation of IL-6R/JAK1/STAT3 signaling induces de novo resistance to irreversible EGFR inhibitors in non-small cell lung cancer with T790M resistance mutation. Mol Cancer Ther. 2012 Oct;11(10):2254-64.

126

Krieger E, Koraimann G, Vriend G. Increasing the precision of comparative models with YASARA NOVA—a self-parameterizing force field. Proteins 2002; 47:393– 402.

Kuffer R, Lombardi T. Premalignant lesions of the oral mucosa. A discussion about the place of oral intraepithelial neoplasia (OIN). Oral Oncol. 2002; 38(2): 125-30.

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

Lane DP. Cancer. p53, guardian of the genome. Nature. 1992; 358(6381):15-6.

Lang K, Menzin J, Earle CC, Jacobson J, Hsu M-A. The Economic Cost of Squamous Cell Cancer of the Head and Neck. Findings from Linked SEER-Medicare Data. Arch Otolaryngol Head Neck Surg. 130: 1269-1275, 2004. PMID: 15545580

Lee JJ, Honk WK, Hittelman WN, Mao L, Lotan R, Shin DM, Benner SE, et al. Prediciting cancer development in oral leukoplakia: ten years in translational research. Clin Cancer Res, 2000; 6:1702-1210.

Leemans CR, Boudewijn JMB, Brakenhoff RH. The molecular biology of head and neck cancer. Nat Rev Cancer. 2011; 11(1):9-22.

Lefebvre JL. Current clinical outcomes demand new treatment options for SCCHN. Ann Oncol. 16: vi7-vi12, 2005. PMID: 15987995

Liao Y, Mu G, Zhang L, Zhou W, Zhang J, Yu H. Lysophosphatidic Acid Stimulate Activation of Focal Adhesion Kinase and Paxillin and Promotes Cell Motility, via LPA1-3, in Human Pancreatic Cancer. Dig Dis Sci 2013; 58:3524-3533.

Lim S-TS. Nuclear FAK: a New Mode of Gene Regulation from Cellular Adhesions. Mol. Cells 2013; 36:1-6.

Lipinski CA, Loftus JC. Targeting Pyk2 for therapeutic intervention. Expert Opin Ther Targets 2010; 14: 95-108.

Lipinski CA, Tran NL, Menashi E, Rohl C, Kloss J, Bay RC et al. The Tyrosine Kinase Pyk2 Promotes Migration and Invasion of Glioma cells. Neoplasia 2005; 7: 435-445.

Lippens S, Denecker G, Ovaere P, Vandenabeele P, Declercq W. Death penalty forkeratinocytes: apoptosis versus cornification. Cell Death Differ. 2005 Nov;12 Suppl 2:1497-508. Review. PubMed PMID: 16247497.

127

Lippman SM, Hong WK. Molecular markers of the risk of oral cancer. N Engl J Med. 2001; 344(17):1323-26.

Macha MA, Matta A, Kaur J, Chauhan SS, Thakar A, Shukla NK, Gupta SD, Ralhan R. Prognostic significance of nuclear pSTAT3 in oral cancer. Head Neck. 2011 Apr;33(4):482-9.

Mehra R, Cohen RB, Burtness BA. The role of cetuximab for the treatment of squamous cell carcinoma of the head and neck. Clin Adv Hematol Oncol. 2008 Oct;6(10):742-50.

Magnussen S, Rikardsen OG, Hadler-Olsen E, Uhlin-Hansen L, Steigen SE, Svineng G. Urokinase plasminogen activator receptor (uPAR) and plasminogen activator inhibitor-1 (PAI-1) are potential predictive biomarkers in early stage oral squamous cell carcinomas (OSCC). PLoS One. 2014 Jul 7;9(7):e101895.

McCubrey JA, LaHair MM, Franklin RA. Reactive Oxygen Species-Induced Activation of the MAP Kinase Signaling Pathways. Antioxid Redox Signal 2006; 8: 1775-1789.

Montagut C, Dalmases A, Bellosillo B, Crespo M, Pairet S, Iglesias M, et al. Identification of a mutation in the extracellular domain of the Epidermal Growth Factor Receptor conferring cetuximab resistance in colorectal cancer. Nat Med. 2012 Jan 22;18(2):221-3. doi: 10.1038/nm.2609. Erratum in: Nat Med. 2012 Sep;18(9):1445.Somasekar, Seshagiri [corrected to Seshagiri, Somasekar]. PubMed PMID: 22270724.

Mallery SR, Tong M, Michaels GC, Kiyani AR, Hecht SS. Clinical and Biochemical Studies Support Smokeless Tobacco’s Carcinogenic Potential in the Human Oral Cavity. Cancer Prev Res 2014; 7:23-32.

Mallery SR, Tong M, Shumway BS, Curran AE, Larsen PE, Ness GM, et al. Topical application of a mucoadhesive freeze-dried black raspberry gel induces clinical and histologic regression and reduces loss of heterozygosity events in premalignant oral intraepithelial lesions: results from a multicentered, placebo-controlled clinical trial. Clin Cancer Res. 2014 Apr 1;20(7):1910-24.

Mallery SR, Budendorf DE, Larsen MP, Pei P, Tong M, Holpuch AS, et al. Effects of human oral mucosal tissue, saliva, and oral microflora on intraoral metabolism and bioactivation of black raspberry anthocyanins. Cancer Prev Res (Phila). 2011 Aug;4(8):1209-21.

Mallery SR, Zwick JC, Pei P, Tong M, Larsen PE, Shumway BS, et al. Topical application of a bioadhesive black raspberry gel modulates gene expression and reduces cyclooxygenase 2 protein in human premalignant oral lesions. Cancer Res. 2008 Jun 15;68(12):4945-57. 128

Menzin J, Lines LM, Manning LN. The economics of squamous cell carcinoma of the head and neck. Curr Opin Otolaryngol Head Neck Surg. 15: 68-73, 2007.

Mitra SK, Hanson DA, Schlaepfer DD. Focal adhesion kinase: in command and control of cell motility. Nat Rev Mol Cell Biol. 2005 Jan;6(1):56-68.

Miyazono S, Isayama T, Delori FC, Makino CL. Vitamin A activates rhodopsin and sensitizes it to ultraviolet light. Vis Neurosci. 2011 Nov;28(6):485-97.

Moon RC, Thompson HJ, Becci PJ, Grubbs CJ, Gander RJ, Newton DL, Smith JM, Phillips SL, Henderson WR, Mullen LT, Brown CC, Sporn MB. N-(4- Hydroxyphenyl)retinamide, a new retinoid for prevention of breast cancer in the rat. Cancer Res. 1979 Apr;39(4):1339-46.

Mora JR, Iwata M, von Andrian UH. Vitamin effects on the immune system: vitamins A and D take centre stage. Nat Rev Immunol. 2008 Sep;8(9):685-98.

Munger K, Scheffner M, Huibregtse JM, Howley PM. Interactions of HPV E6 and E7 oncoproteins with tumour suppressor gene products. Cancer Surv. 1992; 12:197-217.

Müssig E, Steinberg T, Kohl A, Chamulitrat W, Komposch G, Tomakidi P. Discrimination of epithelium-like and fibroblast-like phenotypes derived from ethanol-treated immortalized human gingival keratinocytes in Epithelial equivalents. Cell Tissue Res 2008; 332:57–71.

Neville BW, Day TA. Oral Cancer and Precancerous Lesions. CA Cancer J Clin. 2002; 52: 195-215.

Niu G, Wright KL, Huang M, Song L, Haura E, Turkson J, Zhang S, Wang T, Sinibaldi D, Coppola D, Heller R, Ellis LM, Karras J, Bromberg J, ardoll D, Jove R, Yu H. Constitutive Stat3 activity up-regulates VEGF expression and tumor angiogenesis. Oncogene. 2002 Mar 27;21(13):2000-8.

Oridate N, Suzuki S, Higuchi M, Mitchell MF, Hong WK, Lotan R. Involvement of reactive oxygen species in N-(4-ydroxyphenyl)retinamide-induced apoptosis in cervical carcinoma cells. J Natl Cancer Inst. 1997 Aug 20;89(16):1191-8.

Pantel K, Alix-Panabières C, Riethdorf S. Cancer micrometastases. Nat Rev Clin Oncol. 2009 Jun;6(6):339-51.

Papadimitrakopoulou VA, William WN Jr, Dannenberg AJ, Lippman SM, Lee JJ, Ondrey FG, et al. Pilot randomized phase II study of celecoxib in oral premalignant lesions. Clin Cancer Res 2008; 14: 2095–2101.

129

Ray RM, Li C, Bhattacharya S, Naren AP, Johnson LR. Spermine, a molecular switch regulating EGFR, integrin β3, Src, and FAK scaffolding. Cell Signal. 2012, Apr;24(4):931-42.

Rajagopal S, Huang S, Moskal TL, Lee BN, el-Naggar AK, Chakrabarty S. Epidermal growth factor expression in human colon and colon carcinomas: anti-sense epidermal growth factor receptor RNA down-regulates the proliferation of human colon cancer cells. Int J Cancer. 1995 Sep 15;62(6):661-7.

Rebucci M, Peixoto P, Dewitte A, Wattez N, De Nuncques MA, Rezvoy N, Vautravers-Dewas C, Buisine MP, Guerin E, Peyrat JP, Lartigau E, Lansiaux A. Mechanisms underlying resistance to cetuximab in the HNSCC cell line: role of AKT inhibition in bypassing this resistance. Int J Oncol. 2011 Jan;38(1):189-200.

Reibel J. Prognosis of oral pre-malignant lesions: significance of clinical, histopathological, and molecular biological characteristics. Crit Rev Oral Biol Med. 2003; 14(1):47-62.

Ribeiro DC, Gleber-Netto FO, Sousa SF, Bernardes VD, Guimarães-Abreu MH, Aguiar MC. Immunohistochemical expression of EGFR in oral leukoplakia: association with clinicopathological features and cellular proliferation. Med Oral Patol Oral Cir Bucal. 2012 Sep 1;17(5):e739-44.

Riedel F, Zaiss I, Herzog D, Götte K, Naim R, Hörmann K. Serum levels of interleukin-6 in patients with primary head and neck squamous cell carcinoma. Anticancer Res. 2005 Jul-Aug;25(4):2761-5.

Ries J, Vairaktaris E, Agaimy A, Bechtold M, Gorecki P, Neukam FW, Nkenke E. The relevance of EGFR overexpression for the prediction of the malignant transformation of oral leukoplakia. Oncol Rep. 2013 Sep;30(3):1149-56.

Rodrigo KA, Rawal Y, Renner RJ, Schwartz SJ, Tian Q, Larsen PE, Mallery SR. Suppression of the tumorigenic phenotype in human oral squamous cell carcinoma cells by an ethanol extract derived from freeze-dried black raspberries. Nutr Cancer. 2006;54(1):58-68.

Rowley H, Sherrington P, Helliwell TR, Kinsella A, Jones AS. P53 expression and p53 gene mutation in oral cancer and dysplasia. Otolaryngol Head Neck Surg. 1998; 118(1):115-23.

Sayama K, Hanakawa Y, Shirakata Y, Yamasaki K, Sawada Y, Sun L, Yamanishi K, Ichijo H, Hashimoto K. Apoptosis signal-regulating kinase 1 (ASK1) is an intracellular inducer of keratinocyte differentiation. J Biol Chem. 2001 Jan 12;276(2):999-1004.

130

Schoelch ML, Regezi JA, Dekker NP, Ng IO, McMillan A, Ziober BL,et al.Cell cycle proteins and the development of oral squamous cell carcinoma. Oral Oncol. 1999 May;35(3):333-42.

Scher RL, Saito W, Dodge RK, Richtsmeier WJ, Fine RL. Fenretinide-induced apoptosis of human head and neck squamous carcinoma cell lines. Otolaryngology 1998; 118: 464-471.

Scholes AGM, Woolgar JA, Boyle MA, Brown JS, Vaughan ED, Hart CA, et al. Synchronous oral carcinomas: independent or common clonal origin? Cancer Res. 1998; 58:2003-6.

Schultze A and Fiedler W. Therapeutic potential and limitations of new FAK inhibitors in the treatment of cancer. Expert Opin Investig Drugs 2010; 19:777-788

Sen M, Joyce S, Panahandeh M, Li C, Thomas SM, Maxwell J, Wang L, Gooding WE, Johnson DE, Grandis JR. Targeting Stat3 abrogates EGFR inhibitor resistance in cancer. Clin Cancer Res. 2012 Sep 15;18(18):4986-96.

Serrels B, Serrels A, Brunton VG, Holt M, McLean GW, Gray CH, Jones GE, Frame MC. Focal adhesion kinase controls actin assembly via a FERM-mediated interaction with the Arp2/3 complex. Nature Cell Biology. 2007; 9:1046-1056.

Severino P, Alvares AM, Michaluart P Jr, Okamoto OK, Nunes FD, Moreira-Filho CA, Tajara EH; Head and Neck Genome Project GENCAPO. Global gene expression profiling of oral cavity cancers suggests molecular heterogeneity within anatomic subsites. BMC Res Notes. 2008 Nov 13;1:113.

Shankar S, Ganapathy S, Hingorani SR, Srivastava RK. EGCG inhibits growth,invasion, angiogenesis and metastasis of pancreatic cancer. Front Biosci. 2008 Jan 1;13:440-52.

Shao Y., L.F. Molnar, Y. Jung, J. Kussmann, C. Ochsenfeld, S.T. Brown, A.T.B. Gilbert, et al. Spartan’10, Wavefunction, Inc. Irvine, CA Phys. Chem. Chem. Phys., 8, 3172 (2006).

Shumway BS, Kresty LA, Larsen PE, Zwick JC, Lu B, Fields HW, et al. Effects of a topically applied bioadhesive berry gel on loss of heterozygosity indices in premalignant oral lesions. Clin Cancer Res. 2008 Apr 15;14(8):2421-30.

Siegel RL, Miller KD, Jemal A. Cancer Statistics, 2015. CA Cancer J Clin. 2015, 65:5-29.

Siesser PM, Hanks SK. The Signaling and Biological Implications of FAK Overxpression in Cancer. Clin Cancer Res 2006; 12: 3233-3237.

131

Silverman Jr S, Gorsky M, Lozada F. Oral leukoplakia and malignant transformation. A follow-up study of 257 patients. Cancer 1984;53:563-69.

Slaughter DP, Southwick HW, Smejkal W. Field cancerization in oral stratified squamous epithelium. Cancer (Phila.). 1953; 6:963-68.

Smirnova T, Adomako A, Locker J, Van Rooijen N, Prystowsky MB, Segall JE. In vivo invasion of head and neck squamous cell carcinoma cells does not require macrophages. Am J Pathol. 2011 Jun;178(6):2857-65.

Sperandio M, Brown Al, Lock C, Morgan PR, Coupland, VH, Madden PB, et al. Predictive value of dysplasia grading and DNA ploidy in malignant transformation or oral potentially malignant disorders. Cancer Pre Res (Phila). 2013; 6(8): 822-31.

Sporn MB, Dunlop NM, Newton DL, Smith JM. Fed Proc. Prevention of chemical carcinogenesis by vitamin A and its synthetic analogs (retinoids). 1976 May 1;35(6):1332-8.

Sporn MB, Roberts AB. Role of retinoids in differentiation and carcinogenesis. Cancer Res. 1983 Jul;43(7):3034-40.

Squarize CH, Castilho RM, Sriuranpong V, Pinto DS Jr, Gutkind JS. Molecular cross-talk between the NFkappaB and STAT3 signaling pathways in head and neck squamous cell carcinoma. Neoplasia. 2006 Sep;8(9):733-46.

Steward WP, Brown K. Cancer chemoprevention: a rapidly evolving field. Br J Cancer. 2013 Jul 9;109(1):1-7.

Stoner GD, Wang LS, Casto BC. Laboratory and clinical studies of cancer chemoprevention by antioxidants in berries. Carcinogenesis 2008; 29:1665-1674.

Storz P. Reactive Oxygen Species in Tumor Progression. Front Biosci 2005; 10:1881-1896.

Stransky N, Egloff AM, Tward AD, Kostic AD, Cibulskis K, Sivachenko A, et al. The mutational landscape of head and neck squamous cell carcinoma. Science. 2011 Aug 26;333(6046):1157-60.

Sugimura T, Kumimoto H, Tohnai I, Fukui T, Matsuo K, Tsurusako S, et al. Gene- environment interaction involved in oral carcinogenesis: molecular epidemiological study for metabolic and DNA repair gene polymorphisms. J Oral Pathol Med. 2006; 35(1):11-18.

Sun C, Bernards R. Feedback and redundancy in receptor tyrosine kinase signaling: relevance to cancer therapies. Trends Biochem Sci. 2014 Oct;39(10):465-74.

132

Supino R. Crosti M, Clerici M, Warlters A, Cleris L. Induction of apoptosis by fenretinide (4HPR) in human ovarian carcinoma cells and its association with retinoic acid receptor expression. Int J Cancer 1996; 491-497.

Surh YJ. Cancer chemoprevention with dietary phytochemicals. Nat Rev Cancer.2003 Oct;3(10):768-80. Review. PubMed PMID: 14570043.

Tabor MP, Brakenhoff RH, van Houten VMM, Kummer JA, Snel MHJ, Snijders PJF. Persistence of genetically altered fields in head and neck cancer patients: biological and clinical implications. Clin Cancer Res. 2001; 7(6):1523-32.

Tanaka T, Tanaka M, Tanaka T. Oral carcinogenesis and oral cancer chemopreventive: a review. Patholog Res Int. 2011; 22;2011:431246.

Thiery JP. Epithelial-mesenchymal transitions in tumour progression. Nat Rev Cancer. 2002; 2:442-54.

Tiberio P, Cavadini E. Abolafio G, Formelli F, Appierto V. 4-oxo-N-(4- hydroxyphenyl)retinamide: two independent ways to kill cancer cells. PLoS One 2010; 5: 1- 11.

Tong M, Han BB, Holpuch AS, Pei P, He L, Mallery SR. Inherent phenotypic plasticity facilitates progression of head and neck cancer: Endotheliod characteristics enable angiogenesis and invasion. Exp Cell Res 2013; 319: 1028-1042.

Tong M, B. Lloyd, P. Pei, S.R. Mallery. Human head and neck squamous cell carcinoma cells are both targets and effectors for the angiogenic cytokine, VEGF. J. Cell Biochem. 105 (2008) 1202-1210.

Trott O, Olson AJ. J. AutoDock Vina: Improving the speed and accuracy of docking with a new scoring function, efficient optimization, and multithreading. Computational Chem 2010; 31:455-461.

Tsui IF, Rosin MP, Zhang L, Ng RT, Lam WL. Multiple aberrations of chromosome 3p detected in oral premalignant lesions. Cancer Prev Res (Phila). 2008; 1(6):424-9.

Ugalde CM, Liu Z, Ren C, Chan KK, Rodrigo KA, Ling Y, et al. Distribution of anthocyanins delivered from a bioadhesive black raspberry gel following topical intraoral application in normal healthy volunteers. Pharm Res. 2009 Apr;26(4):977- 86.

Ulukay E, Pirianov G, Kurt MA, Wood EJ and Mehmet H. Fenretinide induces cytochrome c release, caspase 9 activation and apoptosis in the absence of mitochondrial membrane depolarization. Cell Death and Differentiation 2003; 10, 856–859.

133 van Nimwegen MF, Huigsloot M, Camier Am, Tijdens IB, Van de Water B. Focal Adhesion Kinase and Protein Kinase B Cooperate to Suppress Doxorubicin-Induced Apoptosis of Breast Tumor Cells.. Mol Pharmacol. 2006; 70: 1330-1339. van Zyl AW, van Heerden MB, Langenegger E, van Heerden WFP. Correlation between dysplasia and ploidy status in oral leukoplakia. Head Neck Pathol. 2012; 6(3):322-327.

Vernon E. Steele, Gary D. Stoner, Gary J. Kelloff, Charles W. Boone. Cellular and Molecular Targets for Chemoprevention. CRC Press. 1992. Textbook

Von Zeidler SV, Botelho TdS, Mendonca EF, Batista AC. E-cadherin as a potential biomarker of malignant transformation in oral leukoplakia: a retrospective cohort study. BMC Cancer. 2014; 14:972.

Wang D, Zhang H, Li M, Frid MG, Flockton AR, McKeon BA, et al. MicroRNA-124 controls the proliferative, migratory, and inflammatory phenotype of pulmonary vascular fibroblasts. Circ Res 2014; 114:67-78.

Wang K, Fang H, Xiao D, Zhu X, He M, Pan X et al. Converting Redox Signaling to Apoptotic Activities by Stress-Responsive Regulators HSF1 and NRF2 in Fenretinide Treated Cancer Cells. PLoS ONE 2009; 4: e7538 1-13.

Wang J, Shou J, Chen X. Dickkopf-1, an inhibitor of the Wnt signaling pathway, is induced by p53. Oncogene. 2000 Mar 30;19(14):1843-8.

Wang X, Fan M, Chen X, Wang S, Alsharif MJ, Wang L, Liu L, Deng H. Intratumor genomic heterogeneity correlates with histological grade of advanced oral squamous cell carcinoma. Oral Oncol. 2006 Aug;42(7):740-4.

Wang Y, McNiven MA. Invasive matrix degradation at focal adhesions occurs via protease recruitment by a FAK-p130Cas complex. J Cell Biol 2012; 196: 375-385.

Warren HS, Smyth MJ. NK cells and apoptosis. Immunol Cell Biol. 1999 Feb;77(1):64- 75.

Watanabe H, Iwase M, Ohashi M, Nagumo M. Role of interleukin-8 secreted from human oral squamous cell carcinoma cell lines. Oral Oncol. 2002 Oct;38(7):670-9.

Watanabe S, Sato K, Okazaki Y, Tonogi M, Tanaka Y, Yamane GY. Activation of PI3K- AKT pathway in oral epithelial dysplasia and early cancer of tongue. Bull Tokyo Dent Coll. 2009 Aug;50(3):125-33.

134

Watson IR, Takahashi K, Futreal PA, Chin L. Emerging patterns of somatic mutations in cancer. Nat Rev Genet. 2013 Oct;14(10):703-18.

Watt FM, Kubler MD, Hotchin NA, Nicholson LJ, Adams JC. Regulation of keratinocyte terminal differentiation by integrin-extracellular matrix interactions. J Cell Sci. 1993 Sep;106 ( Pt 1):175-82

Wenger SL, Senft JR, Sargent LM, Bamezai R, Bairwa N, Grant SG. Comparison of established cell lines at different passages by karyotype and comparative genomic hybridization. Biosci Rep. 2004 Dec;24(6):631-9.

Whitmore S, Lamont RJ. Oral Bacteria and Cancer. PLoS Pathog. 2014; 10(3):e1003933.

Widakowich C, De Castro G, De Azambuja E, Dinh P, Awada A. Review: Side Effects of Approved Molecular Targeted Therapies in Solid Cancers. The Oncologist 2007; 12: 1443- 1455.

Widakowich; Logue JS and Morrison DK. Complexity in the signaling network: insights from the use of targeted inhibitors in cancer therapy. Genes Dev 2012; 26: 641-650.

Wiest T, Schwarz E, Enders C, Flechtenmacher C, Bosch FX. Involvement of intact HPV16 E6/E7 gene expression in head and neck cancers with unaltered p53 status and perturbed pRb cell cycle control. Oncogene. 2002; 21(10):1510-17.

William AN Jr. Oral premalignant lesions: any progress with systemic therapies? Curr Opin Oncol 2012; 24: 205-210.

Wolbach, S. B., and Howe, P. R. Tissue changes following deprivation of fat soluble A vitamin. J. Exp. Med., 42: 753-777,1925.

Wong DT, Todd R, Tsuji T, Donoff RB. Molecular biology of human oral cancer. Crit Rev Oral Biol Med. 1996; 7(4):319-28.

Xia C, Meng Q, Liu LZ, Rojanasakul Y, Wang XR, Jiang BH. Reactive oxygen species regulate angiogenesis and tumor growth through vascular endothelial growth factor. Cancer Res. 2007 Nov 15;67(22):10823-30.

Xie H, Zju F, Huang Z, Lee M-H, Kin DJ, Li X et al. Identification of mammalian target of rapamycin as a direct target of fenretinide both in vitro and in vivo. Carcinogenesis 2012; 33:1814-1821.

Xiong A, Yang Z, Shen Y, Zhou J, Shen Q. Transcription Factor STAT3 as a Novel Molecular Target for Cancer Prevention. Cancers (Basel). 2014 Apr 16;6(2):926-57.

135

Yamaguchi H. Pathological roles of invadopodia in cancer invasion and metastasis. Eur J Cell Biol 2012; 91: 902-907.

Yamazaki Y, Chiba I, Hirai A, Sugiura C, Notani K, Kashiwazaki H, Tei K, Totsuka Y, Fukuda H. Specific p53 mutations predict poor prognosis in oral squamous cell carcinoma. Oral Oncol. 2003 Feb;39(2):163-9.

Yu H, Lee H, Herrmann A, Buettner R, Jove R. Revisiting STAT3 signalling in cancer: new and unexpected biological functions. Nat Rev Cancer. 2014 Nov;14(11):736-46.

Yue P, Turkson J. Targeting STAT3 in cancer: how successful are we? Expert Opin Investig Drugs. 2009 Jan;18(1):45-56.

Zhang Z, Dong Z, Lauxen IS, Filho MS, Nör JE. Endothelial cell-secreted EGF induces epithelial to mesenchymal transition and endows head and neck cancer cells with stem-like phenotype. Cancer Res. 2014 May 15;74(10):2869-81.

136

Appendix A: Inherent phenotypic plasticity facilitates progression of head and neck

cancer: endotheliod characteristics enable angiogenesis and invasion.

A.1. Abstract

The presence of the EMT (epithelial-mesenchymal transition), EndMT (endothelial- mesenchymal transition) and VM (vasculogenic mimicry) demonstrates the multidirectional extent of phenotypic plasticity in cancers. Previous findings demonstrating the crosstalk between head and neck squamous cell carcinoma (HNSCC) and vascular endothelial growth factor (VEGF) imply that HNSCC cells share some functional commonalities with endothelial cells. Our current results reveal that cultured

HNSCC cells not only possess endothelial-specific markers, but also display endotheliod functional features including low density lipoprotein uptake, formation of tube-like structures on Matrigel and growth state responsiveness to VEGF and endostatin. HNSCC cell subpopulations are also highly responsive to transforming growth factor-β1 and express its auxiliary receptor, endoglin. Furthermore, the endotheliod characteristics observed in vitro recapitulate phenotypic features observed in human HNSCC tumors.

Conversely, cultured normal human oral keratinocytes and intact or ulcerated human oral epithelia do not express comparable endotheliod characteristics, which imply that assumption of endotheliod features is restricted to transformed keratinocytes. In addition, 137 this phenotypic state reciprocity facilitates HNSCC progression by increasing production of factors that are concurrently pro-proliferative and pro-angiogenic, conserving cell energy stores by LDL internalization and enhancing cell mobility. Finally, recognition of this endotheliod phenotypic transition provides a solid rationale to evaluate the antitumorigenic potential of therapeutic agents formerly regarded as exclusively angiostatic in scope.

A.2. Introduction

Head and neck squamous cell carcinoma (HNSCC) represents the sixth most common human cancer worldwide and the eighth leading cause of cancer death with a five year survival rate as low as 50% [1, 2]. Substantial research has focused on identification of the underlying molecular and biochemical mechanisms associated with HNSCC tumorigenesis in order to facilitate development of disease-specific treatments [3-5].

Such mechanistic insights have enabled development of more targeted chemotherapeutic agents, including a spectrum of growth factor specific antagonists, e.g. cetuximab and bevacizumab, to suppress EGF and VEGF signaling, respectively [6, 7]. Previous investigations, by our lab and others, demonstrated that the angiostatic agent, endostatin, inhibited migration and invasion of HNSCC cells, implying that HNSCC cells share some functional characteristics with endothelial cells [8, 9]. In subsequent studies, we demonstrated that HNSCC cells produce exceptionally high levels of VEGF, express both

VEGFR1 and VEGFR2, and proliferate in response to autologously produced VEGF

[10]. This autocrine-paracrine VEGF growth loop can promote HNSCC tumorigenesis in 138 a biphasic fashion via its proangiogenic and growth promotion roles [10]. Importantly, these VEGF-HNSCC in vitro interactions recapitulate components of the premalignant lesion transformation to overt cancer including the increased vascular density that accompanies development and subsequent progression of HNSCC [10, 11]. Collectively, these findings suggest an inherent phenotypic plasticity toward endotheliod features in a subset, if not the entire population, of HNSCC cells.

The presence of variable phenotypic modulations, including epithelial to mesenchymal transition (EMT), mesenchymal to epithelial transition (MET), endothelial-mesenchymal transition (EndMT) and vasculogenic mimicry (VM), indicate a vast extent of multidirectional phenotypic plasticity among tumor cells and their corresponding stroma

[12-16]. Most of these transitions provide tumor-promoting effects such as assumption of an invasive phenotype during the EMT [12, 17], release of the tumor promoting factors

TGF-β and VEGF by cancer associated fibroblasts which originate from transformed endothelial cells that have undergone the EndMT [14, 18], and acquisition of de novo microcirculation and blood supply by forming functional vascular-like channels during cancer cell VM [15, 19, 20]. While VM describes one potential aspect of the functional significances underlying the assumption of endothelial characteristics by malignant tumor cells [19-21], based on our previous findings regarding the HNSCC-VEGF crosstalk [10] and additional basic and clinical research papers [6, 8, 9], we hypothesized that the endotheliod features of HNSCC cells extend beyond formation of vessel-like VM networks.

139

The purpose of this study was to explore and compare the extent of endotheliod functional characteristics present in both in vitro cultured HNSCC cell lines as well as primary and metastatic HNSCC tumors relative to corresponding normal oral keratinocytes and oral mucosa. Our results demonstrate that the extent of phenotypic plasticity is reduced and much more transient in normal keratinocytes relative to fully transformed HNSCC cells and tissues. These data also imply that the epithelial- endotheliod transition is highly supportive for tumorigenesis by virtue of the corresponding increased production of growth factors that are both pro-proliferative and angiogenic, assumption of a more mobile phenotype, and conservation of cellular energy stores. Recognition of this unique phenotypic state reciprocity encourages evaluation of the antitumorigenic potential of therapeutic agents formally regarded as exclusively angiostatic in scope.

140

A.3. Materials and Methods

Due to the acknowledged inter- and intra-HNSCC tumor heterogeneity, these studies employed three well characterized HNSCC cell lines and evaluated 24 HNSCC tumor samples.

Cell Culture

Three HNSCC cell lines (CAL27/CRL-2095, SCC9/CRL-1629, and SCC15/CRL-1623), derived from human tongue cancers were obtained from the American Type Cell Culture

(ATCC, Manassas, VA). For expansion, cells were cultured in DMEM/F-12 medium supplemented with 10% fetal bovine serum (GIBCO, Grand Island, NY) at 37oC, 5%

CO2. In order to avoid the interfering effects from serum components, cells were cultured in serum free medium for selected experiments. Human oral keratinocytes

(HOK) were purchased from ScienCell (Carlsbad, CA) and cultured in oral keratinocyte medium (OKM+OKGS+PIS, ScienCell). U937 monocytes (ATCC) were incubated in

RPMI-1640 medium supplemented with 10% FBS (GIBCO, Grand Island, NY) at 37oC,

5% CO2. Human umbilical vein endothelial cells (HUVECs, ScienCell) were grown in endothelial cell medium (ECM) supplemented with 5% FBS, 1% endothelial cell growth

o supplement (ECGS) and 1% penicillin/streptomycin (P/S) at 37 C, 5% CO2 (ECM,

ECGS and antibiotics were obtained from ScienCell). Transforming growth factor beta 1

(TGF-β1) is a recognized acute phase reactant associated with cell phenotypic changes,

141 e.g., EMT. For selected experiments, cells were exposed to recombinant human TGF-β1

(BioLegend, San Diego, CA) to assess its effects on the endotheliod phenotypic transition.

Immunocytochemistrty

HUVEC, HOK and HNSCC cells (1x104 cells/chamber) were seeded on chamber slides followed by 24 h incubation in serum free base medium with or without 10 ng/ml

o recombinant human TGF-β1 (BioLegend) at 37 C, 5% CO2. Following routine wash, fixation steps, cell monolayers were then blocked with 1% BSA for 30 min and incubated with VE-Cadherin (1:50, Santa Cruz, Biotechnology, Santa Cruz, CA), CD31 (1:100,

Cell Signaling Tec., Boston, MA), vimentin (1:200, Abcam, Cambridge, MA) or cytokeratin 6 (1:200, Santa Cruz Biotechnology) antibodies at 4oC overnight. Cells were then incubated with FITC or Texas Red conjugated secondary antibodies (Abcam,

Cambridge, MA) for 1 h at room temperature. DAPI was used to visualize the nuclei.

Fluorescence microscopy images were obtained by using an Olympus BX51 microscope

(Olympus, Japan), Nikon DS-Fi1 digital camera (Nikon, Japan) and ImagePro 6.0 software (MediaCybernetics, Bethesda, MD).

142

Western Blot analysis

HNSCC, HUVEC or HOK cells were harvested by using AccutaseTM cell detachment reagent (Chemicon, Temecula, CA). Total cytoplasmic and nuclear protein was extracted from the cell pellets by M-PER mammalian protein extraction reagent (Pierce, Rockford,

IL). Western blot analyses were then conducted using the previously described method

[10]. The antibodies and working dilutions were as follows: Erk1/2 mouse monoclonal antibody (1:2000, Cell Signaling Tec.), and phospho-Erk1/2 rabbit polyclonal antibody

(1:1000, Cell Signaling Tec.), Src rabbit monoclonal antibody (1:1000 dilution, Cell

Signaling Tec.), phospho-Src rabbit polyclonal antibody (1:1000 dilution, Cell Signaling

Tec.), VE-cadherin mouse monoclonal antibody (1:200, Santa Cruz Biotechnology).

Kodak 1D3 image analysis software (Kodak) was employed to perform densitometric analyses. Data were normalized relative to protein levels of β-actin, which was probed by a mouse monoclonal antibody (1:1000, Santa Cruz Biotechnology).

Qualitative and quantitative assessment of acetylated low density lipoprotein

(AcLDL) internalization

Log-growth HNSCC cells, HOKs and HUVECs were seeded on chamber slides

(qualitative assay) or 96-well plates (quantitative assay) and pretreated with or without 10 ng/ml recombinant human TGF-β1 (BioLegend) in serum free base medium for 24 h.

Conditioned media were then discarded and cells were exposed to 10μg/ml Alexa Fluor

488 conjugated AcLDL (Molecular Probes, Eugene, OR) for 4 h in 10% FBS

143

o supplemented growth media at 37 C, 5%CO2. Extracellular Alexa Fluor 488 signals were quenched with 0.08% Trypan Blue (Sigma-Aldrich, St Louis, MO), followed by two washes with phosphate buffered saline (PBS). Intracellular Alexa Fluor 488-AcLDL uptake was qualitatively observed using an Olympus BX51 fluorescence microscope. In the qualitative assay, E-Cadherin (mouse monoclonal antibody, 1:50, Abcam Inc) and

CD31 (1:100, Cell Signaling Tec.) staining were used to delineate HNSCC, HOKs and

HUVECs intercellular boundaries, respectively. DAPI was used to visualize the nuclei.

Cellular fluorescent images were acquired using a Nikon DS-Fi1 digital camera and the

ImagePro 6.0 software. The FLUOstar Omega microplate reader (BMG Labtech,

Germany) was employed to quantify cellular AcLDL uptake. An AcLDL standard curve was conducted with each assay. U973 monocytes were also included in the quantitative assay as a positive control. Quantitative data were normalized to AcLDL (ng) per 105 cells. Cell numbers were determined by CellTiter 96 Aqueous One Solution Cell

Proliferation Assay (Promega, Madison, WI) incorporating cell line-specific standard curves.

Enzyme-Linked ImmunoSorbent Assay (ELISA)

Twenty-four hour serum deprived HUVECs and HNSCC cell cultures were incubated with or without 10 ng/ml recombinant human TGF-β1 (BioLegend, San Diego, CA) for

48 h in serum free base medium. Conditioned media were collected and concentrated with Amicon Ultra-15 centrifugal filter devices (Millipore, Billerica, MA). Total cellular proteins were then extracted from the cell monolayers by using M-PER. Protein levels of 144

Endoglin present in both conditioned media and total cell lysates were analyzed by protein specific DuoSet ELISA Development Kits (R&D Systems, Minneapolis, MN).

Concentrations of target proteins were normalized to pg/mg total protein.

Cell Invasion Assay

Following 24 h serum deprivation, conditioned media were removed and HNSCC cells were pretreated with 0 (negative control) or 10 ng/ml recombinant human TGF-β1

(BioLegend) in fresh, serum free, base medium for an additional 24 h prior to the cell invasion assay (InnoCyte cell invasion kit, Calbiochem, San Diego, CA). Cells (3.5 x105/well) were then seeded in the upper chambers with or without 10 ng/ml TGF-β1 in fresh, serum free, base medium. Media supplemented with 10% FBS were added to the lower chambers as chemoattractants. The invasion plates were then incubated at 37oC,

5% CO2 for 48 h with fresh TGF-β1 (10 ng/ml) added every 24 h. Standard curves consisting of cell line matched control cells were run concurrently with each invasion assay.

Immunohistochemistry

Twenty-four primary or metastatic human HNSCC tumor samples were obtained from the Ohio State University Comprehensive Cancer Center Tissue Procurement Services (in accordance with Ohio State University IRB approval) and were immediately placed in

10% neutral buffered formalin (24-48 h). The histopathology reports containing 145 information of patients’ gender, age, clinical sites, etc. were provided by the Tissue

Procurement Services. Five histologically normal and ten ulcerated nonneoplastic oral mucosal tissues were randomly chosen from achieved oral pathology biopsy samples

(IRB approved). Paraffin embedded tissue sections were deparaffinized, rehydrated and microwaved in citrate buffer (pH 6.0) for antigen retrieval. Sections were then blocked with 5% normal serum, 1% bovine serum albumin (Sigma-Aldrich), 0.05% Tween® 20

(Fisher Scientific, Pittsburgh, PA) in PBS for 1 h and incubated with primary antibodies or PBS (negative control) at 4oC overnight, followed by incubation with biotinylated secondary antibodies and Vectastain ABC reagent (Vector Laboratories, Burlingame,

CA). Antibodies employed in this study were: VE-cadherin (1:10, Millipore), CD31

(1:100, Cell Signaling Tec.), vimentin (1:100, Abcam), Pan-cytokeratin (1:75, Abcam).

Three dimensional cell culture and tube formation assay

Log growth HUVEC, HOK and HNSCC cells were trypsonized and resuspended to

4x105cells/ml with 10% FBS supplemented growth media. Cell suspensions (300 μl/well) were then added to Matrigel (289 μl/well, BD Biosciences, Bedford, MA) pre-coated 24-

o well plates and incubated at 37 C, 5% CO2 for 17 hrs. Formation of tube-like structures was observed using a Zeiss Axiovert 200 phase contract inverted microscope (Zeiss,

Germany). Digital images were acquired using a Carl Zeiss AxioCam MRC5 digital camera (Zeiss, Germany).

146

Statistical analyses

TGF-β1’s effects on AcLDL uptake and endoglin protein expression levels in individual

HNSCC cell lines were evaluated using the Mann Whitney U two-tailed test. Results of cell invasion assay were analyzed using the Wilcoxon one-tailed signed rank test.

Comparison of inter-cell line AcLDL uptake was conducted using the Kruskal-Wallis

Analysis of Variance, followed by a Dunn’s Multiple Comparisons post-test. The data distribution determined whether parametric or non-parametric analyses were conducted.

A.4. Results

Cultured HNSCC cells express endothelial-associated proteins VE-cadherin, CD31, and Vimentin

Consistent with their endothelial lineage, cultured HUVECs showed the presence of three

“endothelial markers”: 1) the vascular endothelial specific intercellular adhesion molecule VE-cadherin (Figure A.1A), 2) the platelet endothelial cell adhesion molecule

PECAM-1 (CD31, see Figure A.2B and Figure A.3A) and 3) vimentin, the intermediate filament which is typically found in mesenchymal and endothelial cells (Figure A.2A).

Corresponding immunocytochemistry (ICC) analyses of HNSCC cells confirmed the presence of VE-cadherin (Figure A.1A), vimentin, and CD31 in all three HNSCC cell lines evaluated (Figure A.2 C-H). The presence of VE-cadherin protein was also verified by immunoblotting in all three HNSCC cell lines (Figure A.1B). Furthermore, the extent 147 of expression was both marker and HNSCC cell line dependent, with highest VE- cadherin levels detected in SCC15 cells (Figure A.1B). HOKs, in contrast, did not express VE-cadherin (Figure A.1A). The epithelial origin of HNSCC cells and HOKs was confirmed by E-cadherin staining shown on Figure A.4A and Figure A.5A.

HNSCC cells form tube-like structures on Matrigel 3D cultures.

The capacity to form tube-like structures on Matrigel is a recognized feature of vasculogenic mimicry [15, 23]. Similar to HUVECs, all 3 HNSCC cell lines demonstrated formation of tube-like structures following 17h three-dimensional culture on Matrigel (Figure A.3). Notably, the tube-like structures formed in HUVEC cultures were lined by a single layer of HUVEC cells. Their counterparts in HNSCC cultures, however, were surrounded with irregularly arranged HNSCC cell aggregates (Figure

A.3). The tube-like structure morphology was dynamic and the tubules themselves were obliterated within 3 days due to expansion of cell population. HOKs failed to show the similar capacity of tube formation following 3D culture on Matrigel (Figure A.3).

HNSCC cells internalize AcLDL, which is augmented following exposure to TGF-

β1.

As AcLDL uptake is an established characteristic of endothelial cells and macrophages

[24-26], HUVECs and U937 monocytes served as the respective control cell populations for these studies. Both qualitative and quantitative analyses confirmed the ability of 148

HNSCC cell lines to internalize AcLDL. While qualitative AcLDL levels were highest in

HUVECs (Figure A.4A.), all three HNSCC cell lines evaluated (CAL27, SCC9 and

SCC15) demonstrated cytoplasmic internalization of Alexa Fluor 488 conjugated AcLDL

(Figure A.4A). Normal HOKs, however, failed to show any evident AcLDL uptake

(Figure A.4A). Due to their lack of adherent growth, U937 cells were excluded from these qualitative studies. Quantitative analyses showed that among the HNSCC cells,

CAL27 cells exhibited the greatest internalization of Alexa Fluor 488-AcLDL, which was comparable to the control HUVECs population (46.7 ± 2.3%, P > 0.05, n = 8) (Figure

A.4B). Internalized AcLDL levels achieved in SCC9 and SCC15 cells were significantly lower than control HUVECs [(33.5 ± 0.8%, P < 0.01, n = 8, SCC9) and (16.8 ± 1.0%, P

< 0.001, n = 8, SCC15)] (Figure A.4B). Finally, 24 h treatment of HNSCC cells with 10 ng/ml of the known phenotypic modulating agent, TGF-β1, significantly increased cellular capacity to internalize AcLDL in all three HNSCC cell lines (Figure A.4B).

While TGF-β1 treated SCC15 cells demonstrated a 29.4 ± 9.0% (P = 0.0207, n = 8) increase of AcLDL uptake relative to the non-treated SCC15 cells, the total internalized

AcLDL levels in treated SCC15 cells were still significantly lower relative to HUVECs

(P < 0.001, n = 8) and treated CAL27 cells (P < 0.01, n = 8). Although AcLDL uptake is known to be mediated by scavenger receptors in macrophages and endothelial cells [27,

28], we did not identify the presence of the scavenger receptors (SR-A, CD36) in cultured

HNSCC cells in this study (data not shown). These findings imply HNSCC internalization of AcLDL occurred by another mechanism.

149

Introduction of TGF-β1 redirects expression of HNSCC cell adhesion molecules towards a less adherent, more mobile “endotheliod” phenotype.

A 24 h exposure of SCC15 cells to 10 ng/ml TGF-β1 elicited striking modulations in cell adhesion molecules. Notably, TGF-β1 treated SCC15 cells demonstrated “cadherin switching” as manifested by a marked decrease in E-cadherin with a concurrent increase in VE-cadherin. In addition, levels of CD31 and vimentin were also increased in treated

SCC15 cells (Figure A.5A). Interestingly, CAL27 and SCC9 cultures’ cell adhesion molecules were refractory to TGF-β1 challenge (data not shown). Repeated experiments demonstrated comparable findings.

TGF-β1 treatment increases levels of a component of the TGF-β1 receptor complex, endoglin, in HNSCC cells.

The three HNSCC cell lines demonstrated very low constitutive levels of endoglin in cell lysates, which corresponded to 0.81 ± 0.09% (CAL27), 0.22 ± 0.14% (SCC9) and 0.42 ±

0.18% (SCC15) of the endoglin level in positive control HUVECs. Forty-eight hour treatment (10ng/ml TGF-β1) significantly increased endoglin levels in CAL27 and

SCC15 cells by 210.0 ± 55.6% (P = 0.0286, n = 4) and 575.3 ± 85.2% (P = 0.0286, n =

4), respectively (Figure A.5B). SCC9 cells, which were essentially refractory to TGF-β1 challenge, showed an insignificant decrease (-34.2 ± 57.2%, P = 0.8834, n = 4) of endoglin compared to their matched control cells (Figure A.5B). Similar to others’ findings which demonstrated that endoglin could be shed and released into stroma as a

150 soluble form [29, 30] , extracellular endoglin levels were detected in the conditioned media of cultured HUVECs (380.0 ± 75.3 pg/ml, n = 2) and only in one SCC cell line-

SCC15 (16.2 ± 5.3 pg/ml, n = 2). TGF-β1 treatment increased endoglin levels in conditioned medium of SCC15 cells to 29.3 ± 1.2 pg/ml (n = 2).

TGF-β1 treatment augments invasiveness of HNSCC cells.

We have previously confirmed that all three HNSCC cell lines readily invade synthetic basement membranes in response to chemoattractants [10]. For these current studies, we employed the most TGF-β1 responsive (SCC15) and TGF-β1 refractory (SCC9) cell lines. Our data show TGF-β1 (10 ng/ml, 24 h) treatment significantly increased SCC15 invasion by 36.3 ± 17.5% (P = 0.0313, n = 5) (Figure A.5C). Similar to the endoglin experiments, SCC9 cells were unaffected by treatment, resulting in an insignificant decrease (-11.1 ± 2.6%, P = 0.2188, n = 5) in invasion following TGF-β1 challenge

(Figure A.5C).

151

Surface epithelial cells from biopsies of ulcerated, otherwise normal oral mucosa display enhanced levels of the “mesenchymal” microfilament, vimentin, which is not expressed in intact, normal oral epithelium.

Five histologically normal oral mucosal samples with intact epithelium and ten ulcerated, non-neoplastic oral biopsies were evaluated for the presence of VE-cadherin, vimentin, and CD31. Intact oral epithelium was negative for all of these markers. Focal positivity was observed for CD31 and vimentin in the resident antigen presenting cells (Langerhans cells) interspersed among the surface epithelial cells (Figure A.6, Column 1). In contrast, biopsies of ulcerated mucosa demonstrated the presence of vimentin in surface epithelial cells, with greater intensity of staining (as red arrows indicate) observed in the basal l/3 of epithelium adjacent to the ulcerated sites (Figure A.6, Column 2). CD31 (Figure A.6,

Column 2 Row 4) and VE-cadherin (Figure A.6, Column 2 Row 5, as the red rectangle indicates) were shown on stromal vascular endothelial cells which served as an internal control. These two endothelial markers, however, were not detected in the ulcerated epithelial cells.

VE-cadherin, Vimentin, and CD31 proteins are also present in lesional cells of primary and metastatic HNSCC tumors.

Twenty-four HNSCC tumors, which corresponded to HNSCC clinical stages II through

IV, were evaluated for the presence of pan-cytokeratin, VE-cadherin, vimentin, and

CD31 by immunohistochemistry (See Figure A.6 and Table I). Positive pan-cytokeratin

152 staining confirmed the epithelial origin of the tumor nests (Figure A.6, Row 1).

Endothelial and connective tissue stromal cells served as the internal controls for VE- cahderin, CD31 (endothelial cells) and vimentin (both endothelial and stromal cells) staining (Figure A.6). Tumor nests that approximated regions of tumor necrosis demonstrated the highest level of CD31 and vimentin (Figure A.6). VE-cadherin was most prevalent in the regions that contained advancing tumor islands (Figure A.6,

Column 4 Row 5, as the red rectangle indicates). Depiction of the distribution (as indicated by percent positive staining) of these three markers relative to clinical stage and nodal involvement were summarized in Table II. These data demonstrate that higher percentages of VE-cadherin and vimentin were detected in lesional cells of HNSCC tumors obtained from patients with lymph node metastasis (N > 0, i.e., presence of 1 or more metastatic nodes) and more advanced clinical stages (> II, i.e., primary HNSCC tumor >2cm). While these data imply a positive correlation between higher clinical stage and extent of endotheliod features, due to the small number (n = 3) of lower stage i.e. < Stage II tumors, correlative analyses were not possible. Relevant to the VM, a number of tube-like structures with lumen exclusively lined with cancer cells were identified in many HNSCC specimens. No blood cells, however, were observed within those structures (data not shown).

153

VEGF elicits phosphorylation of downstream signaling mediators Erk1/2 and Src in

HNSCC cells.

We have previously demonstrated that HNSCC cells possess VEGFR1 and VEGFR2, which are receptors that are traditionally regarded as endothelial specific [10]. In this current study, Western blot results demonstrated time-dependent modulation of phospho-

Erk1/2 and phospho-Src in CAL27 cells following 50 ng/ml VEGF challenge (see Figure

A.7A). Phosphorylation of Erk1/2 and Src occurred rapidly (within 1 min) after addition of VEGF. The phosphorylated levels remained elevated for approximately 20 min and gradually decreased to the basal levels over the subsequent 40 min. As the Src and

Phospho-Src antibodies react with all kinase members in Src family [31], multiple bands were observed in Src and Phospho-Src blots (Figure A.7A).

Endostatin attenuates VEGF-induced downstream signaling in HNSCC cells.

While endostatin is a recognized angiostatic agent and is presumed to primarily target endothelial cells, previous studies by our lab and others, demonstrated that endostatin also inhibited migration and invasion of HNSCC cells [8, 9]. Our current data showed that pretreatment with endostatin (10μg/ml) prior to VEGF challenge (50ng/ml, 2 min) attenuated VEGF-induced phosphorylation of Erk1/2 and Src in a time-dependent fashion in CAL27 cells (Figure A.7B). Sixty-minute endostatin pretreatment markedly diminished the phosphorylation of Erk1/2 and Src to the baseline levels (Figure A.7B).

154

A.5. Discussion

Expression of cell adhesion molecules and the presence of specific intermediate filaments have traditionally been regarded as cell-lineage specific, e.g., cytokeratins-epithelial cells, vimentin-endothelial or mesenchymal origins [32, 33]. Consistent with this concept, identification of these “lineage specific” molecules by techniques such as IHC is currently a standard diagnostic tool used by pathologists to aid in the diagnosis of lesions of uncertain histogenesis. The recognition of the EMT, MET, EndMT and VM, however, transformed this traditional cell-lineage specific paradigm [12-15]. Our data, which demonstrate the presence of molecules typically associated with endothelial cells, i.e.,

VE-cadherin, CD31, and vimentin, in cultured HNSCC cell lines and solid HNSCC tumors, expand the cell plasticity continuum to include the endothelial-like phenotype which fulfills a distinct role from the EMT. Acquisition of endotheliod characteristics by cancer cells enhances the VEGF-based autocrine-paracrine growth loop [10], facilitates cell invasion and promotes tumor-associated angiogenesis. As a component of the phenotypic state reciprocity, assumption of endotheliod phenotype may occur concurrently with other phenotypic transitions to sustain a pro-proliferative, pro-invasive growth state for tumor progression through their combined effects.

While the expression pattern of vimentin in HNSCC cells is analogous to that in

HUVECs, only a subpopulation of HNSCC cells express vimentin. In contrast, the

155 location of CD31 staining was cell origin specific. The endothelial lineage cells

(HUVEC) demonstrate classic membrane staining. Similar to the findings in human uterine squamous cell carcinomas [51], human HNSCC cells showed a cytosolic CD31 distribution. These data are consistent with assumption of other endotheliod characteristics in HNSCC cell subpopulations. Furthermore, human HNSCC tumor samples also demonstrated both inter-patient and intra-tumor variations in expression of endotheliod markers. For example, CD31 expression was highest in HNSCC cells adjacent to necrotic tumor whereas VE-cadherin was most prominent at the advancing invasive tumor margin. The heterogeneity present in HNSCC tumor cells e.g. high VEGF production by some cells with assumption of a mobile endotheliod phenotype by others could benefit progression of the tumor as whole. These HNSCC phenotypic modulations appear to reflect tumor cell responses to changes in its dynamic microenvironment.

Similar to our findings, Sajithlal et al. [34] successfully identified and isolated tumor- derived endothelial-like cells (TDEC) from a variety of human cancer xenografts (breast, prostate, and lung cancers). Those TDEC cells contain only human chromosomes and share variable commonalities with endothelial cells [34]. Sajitheal et al. speculate that the transition from cancer cells to TDEC may be due to a transdifferentiation-like process and/or VM [34]. While our data demonstrated assumption of immunological and functional endotheliod characteristics including tube-like structure formation in HNSCC cells and tumors, we failed to identify any blood cell-containing VM structures in our 24 separate HNSCC tumor specimens. These findings imply that the functional

156 consequences following assumption of the endotheliod phenotype may not be restricted to formation of VM network per se. The acquisition of enhanced mobility via cadherin switching in conjunction with facilitation of tumor progression by augmenting the

HNSCC-VEGF growth loop provides a solid rationale for development of this epithelial to endotheliod phenotypic transition.

In addition to tube-like structure formation on 3D cultures, internalization of low density lipoprotein is another unique functional characteristic attributed to endothelial cells [24,

25]. Perturbations in this “scavenger” function are speculated to contribute to the pathogenesis of atherosclerosis [24]. Accordingly, endocytosis of AcLDL is widely used as a screening assay to identify endothelial cells [25]. Our qualitative and quantitative results confirmed that HNSCC cells internalize AcLDL and that there are cell-line related differences in uptake capacity. Furthermore, while the scavenger receptors (SR-A and

CD36) are associated with LDL uptake in macrophages and endothelial cells [27, 28], such receptors were not detected in HNSCC cells. These data imply that HNSCC cells employ receptor-independent mechanisms such as pinocytosis to internalize AcLDL.

Previous studies from our lab, which demonstrate the uptake of nanoparticles by HNSCC cells, support this premise [35]. As an important cholesterol carrier, LDL is crucial to maintain the cholesterol homeostasis in both normal endothelial cells and transformed tumor cells [36, 37]. Consequently, new cholesterol delivered into tumor cells via LDL internalization could preserve cell energy reserves by suppression of endogenous cholesterol production [38]. Furthermore, LDL delivered cholesterol could contribute an

157 essential cell membrane component which is vital to support rapid tumor cell growth and division [38].

TGF-β1 is known to induce the EMT via Smad and non-Smad pathways during both normal growth and carcinogenesis [39]. Relevant to HNSCC progression, previous studies have demonstrated that TGF-β1 is overexpressed in human HNSCC tumors and adjacent tissues [40], and is associated with MMP-9 induction in HNSCC cells [41].

Furthermore, while TGF-β1 upregulates VEGF production in human cancer cells [42], additional studies failed to show an association between VEGF secretion and microvessel density in HNSCC tumors [43]. These data suggest that TGF-β1 stimulated HNSCC production of VEGF may fulfill other functions in addition to angiogenic induction e.g. an intracrine growth factor. Our data revealed that TGF-β1 treatment significantly augmented AcLDL uptake in every HNSCC cell line, despite their baseline differences in

AcLDL uptake. Interestingly, the cell line that showed the lowest baseline and TGF-β1 stimulated AcLDL uptake (SCC15) was also the most TGF-β1 responsive with regard to cadherin switching and invasion. These data indicate that TGF-β1 affects multiple pathways and that the relative extent of transition to an endotheliod phenotype is contingent on the responsiveness of the heterogeneous HNSCC cell subpopulations.

In order to clarify the underlying mechanism responsible for TGF-β1 induced endotheliod phenotypic transition, TGF-β1’s effect on expression of endoglin protein in

158

HNSCC cells was investigated. Endoglin, which is a homodimeric membrane glycoprotein that functions as an auxiliary receptor for the TGF-β family, binds a number of TGF-β superfamily members including TGF-β1, TGF-β3, activin-A, BMP-7 and

BMP-2 in association with TGF-β type I and type II receptors [44, 45]. While endoglin is typically expressed in vascular endothelial cells, macrophages, fibroblasts and stromal cells, endoglin expression has also been identified in transformed epithelial lineage cells such as prostate cancer cells [46] and metastatic breast cancer cells [47]. Endoglin overexpression was found to be associated with enhanced invasive phenotype of human breast cancer cells [47]. Also, membrane-bound endoglin could be shed and released into stroma as a soluble form which is associated with the development of poorly differentiated carcinoma [29, 30]. Our data show that the baseline levels of endoglin in

HNSCC cells are relatively low compared to HUVECs. Exposure to TGF-β1, however, significantly increased endoglin levels in CAL27 and SCC15 cell lysates. While other studies have detected endoglin in human cancer cells [46, 47], these data were reported as relative amounts, making direct comparisons to HNSCC endoglin levels impossible.

Notably, SCC15 cells are most responsive to TGF-β1 with regard to cadherin switching and cell invasion. In contrast, endoglin levels in SCC9 cells, which are refractory to TGF-

β1 modulation of cadherins and invasion, remained negligible regardless of TGF-β1 treatment.

It is well established that, in addition to its crucial role in cancer progression, the EMT is also an essential process for normal growth and development [17]. Our results, which

159 showed the expression of vimentin in ulcerated epithelium, support the role of EMT in facilitating epithelial cell migration necessary for ulcer healing. The absence of CD31 and VE-cadherin in normal and ulcerated oral epithelium and in vitro cultured HOKs together with the negative results of tube-like structure formation and AcLDL uptake shown by HOKs, indicate that the most extensive phenotypic plasticity is more associated with transformed malignant cells. Our in vitro cell data was validated by IHC studies on

HNSCC tumor tissues which showed endotheliod HNSCC cells were predominantly located at the advancing end (including metastases) of the tumor. While normal physiological process such as wound healing may recruit a transient pro-migratory phenotypic transition like EMT, a sustained pro-proliferative and pro-migratory growth state which involves multidirectional phenotypic transitions (including epithelial- endotheliod phenotype reciprocity) appears essential for malignant tumor progression.

We have previously reported that VEGFR1 and VEGFR2 are present in cultured HNSCC cells and VEGF enhanced HNSCC cell proliferation and invasion [10]. Binding of VEGF to its cognate tyrosine kinase receptors (VEGFRs) elicits activation of a series of intracellular signaling cascades including Erk and Src pathways, resulting in expression of variable downstream genes associated with proliferation and migration [48]. VEGF- mediated Erk and Src activation thus indicates the endothelial or endothelial-like phenotype of target cells. These current data, which show that VEGF challenge initiates

VEGF-receptor mediated intracellular signaling in CAL27 cells, substantiate our previous

HNSCC-VEGF observations. While similar studies were conducted on the all three

160

HNSCC lines, the CAL27 cells were the most responsive to exogenous VEGF. Previous studies [10, 49] have shown that most HNSCC cell lines produce exceptionally high levels of VEGF resulting in these cells being refractory to exogenous VEGF. In contrast,

CAL27 cells (also referred to as SCC2095) produce relatively lower VEGF levels and thereby retain responsiveness to exogenous VEGF without the need for siRNA blocking of endogenous VEGF. The fact that HNSCC cells not only serve as target of VEGF, but also respond to angiostatic agent endostatin substantiates the premise that HNSCC tumors and isolated cells contain at least a cellular subpopulation with endotheliod characteristics.

Previous and current data from our lab and others demonstrated that the angiostatic agent, endostatin, interfered with HNSCC cell migration and HNSCC-VEGF crosstalk in an analogous fashion to its effects on endothelial cells [8, 9]. Without the potential for clinical implications, the recognition of an endotheliod transition in HNSCC cells is merely an interesting observation. The clinical relevance of these findings, however, lies with regard to implementation of future treatments. Notably, introduction of compounds that target angiogenesis-associated pathways, e.g., bevacizumab, sorafenib, have improved the prognosis and survival of HNSCC patients [6, 50]. These findings imply that the positive clinical effects of these agents likely reflect both their angiostatic and antitumorigenic properties and establishes the precedent to identify additional, more selective agents capable of concurrently inhibiting both the activated endothelial cells as

161 well as the endotheliod HNSCC cells necessary for tumor associated angiogenesis and metastasis, respectively.

A.6. Conclusions:

Our data demonstrate that HNSCC tumor cells express markers traditionally regarded as endothelial cell specific with the highest expression detected at the leading edges of the tumor nests (VE-cadherin) and peri-necrosis regions (CD31 and vimentin). Cultured

HNSCC cells recapitulate these findings as demonstrated by expression of endotheliod markers, and also show functional endotheliod characteristics such as AcLDL uptake and responsiveness to VEGF and endostatin. Furthermore, the well-established tumor phenotypic modulator, TGF-β1, enhances these endotheliod characteristics and augments invasiveness in HNSCC cell subpopulations. Collectively, these findings imply that assumption of an endotheliod phenotype by at least a subpopulation of cells facilitates

HNSCC progression by enhancing cell mobility and establishing a pro-angiogenic and pro-proliferative VEGF-mediated intracrine growth loop.

A.7. Acknowledgements:

This work was supported by the National Institution of Health (NCI R01CA129609 NCI

RC2CA148099 and NCI R01CA171329 to Dr. Susan R. Mallery, F30 DE02992 to

Andrew S. Holpuch and T32 DE14320 to Byungdo B. Han). We thank Mary Marin and

Mary Lloyd, our histotechnologists, for their assistance with tissue specimen preparation. 162

A.8. References:

1. K. Shibuya, C. Mathers, C. Boschi-Pinto, A. Lopez, C. Murray. Global and regional estimates of cancer mortality and incidence by site: II. results for the global burden of disease 2000. BMC Cancer 2 (2002) 37-62.

2. D. Chin, G.M. Boyle, S. Porceddu, D.R. Theile, P.G. Parsons, W.B. Coman. Head and neck cancer: Past, present and future. Expert. Rev. Anticancer. Ther. 6 (2006) 1111-1118.

3. A. Bozec, F. Peyrade, J. Fischel, G. Milano. Emerging molecular targeted therapies in the treatment of head and neck cancer. Expert. Opin. Emerg. Drugs. 14 (2009) 299-310.

4. M. Montag, G. Dyckhoff, J. Lohr, B.M. Helmke, E. Herrmann, P.K. Plinkert, C. Herold-Mende. Angiogenic growth factors in tissue homogenates of HNSCC: Expression pattern, prognostic relevance, and interrelationships. Cancer Sci. 100 (2009) 1210-1218.

5. A. Matta, R. Ralhan. Overview of current and future biologically based targeted therapies in head and neck squamous cell carcinoma. Head Neck Oncol. 1 (2009) 6- 13.

6. K. Fujita, D. Sano, M. Kimura, Y. Yamashita, M. Kawakami, Y. Ishiguro, G. Nishimura, H. Matsuda, M. Tsukuda. Anti-tumor effects of bevacizumab in combination with paclitaxel on head and neck squamous cell carcinoma. Oncol. Rep. 18 (2007) 47-51.

7. P. Kabolizadeh, G.J. Kubicek, D.E. Heron, R.L. Ferris, M.K. Gibson. The role of cetuximab in the management of head and neck cancers. Expert. Opin. Biol. Ther. 12 (2012) 517-528.

8. R.F. Wilson, M.A. Morse, P. Pei, R.J. Renner, D.E. Schuller, F.M. Robertson, S.R. Mallery. Endostatin inhibits migration and invasion of head and neck squamous cell carcinoma cells. Anticancer Res. 23 (2003) 1289-1295.

9. M. Li, C. Ye, C. Feng, F. Riedel, X. Liu, Q. Zeng, J.R. Grandis. Enhanced antiangiogenic therapy of squamous cell carcinoma by combined endostatin and epidermal growth factor receptor-antisense therapy. Clin. Cancer Res. 8 (2002) 3570-3578.

10. M. Tong, B. Lloyd, P. Pei, S.R. Mallery. Human head and neck squamous cell carcinoma cells are both targets and effectors for the angiogenic cytokine, VEGF. J. Cell Biochem. 105 (2008) 1202-1210.

163

11. E.R. Sauter, M. Nesbit, J.C. Watson, A. Klein-Szanto, S. Litwin, M. Herlyn. Vascular endothelial growth factor is a marker of tumor invasion and metastasis in squamous cell carcinomas of the head and neck. Clin. Cancer Res. 5 (1999) 775- 782.

12. M. Iwatsuki, K. Mimori, T. Yokobori, H. Ishi, T. Beppu, S. Nakamori, H. Baba, M. Mori. Epithelial–mesenchymal transition in cancer development and its clinical significance. Cancer Sci. 101 (2009) 293-299.

13. A.A. Leontovich, S. Zhang, C. Quatraro, I. Iankov, P.F. Veroux, M.W. Gambino, A. Degnim, J. McCubrey, J. Ingle, E. Galanis, A.B. D'Assoro. Raf-1 oncogenic signaling is linked to activation of mesenchymal to epithelial transition pathway in metastatic breast cancer cells. Int. J. Oncol. 40 (2012) 1858-1864.

14. S. Potenta, E. Zeisberg, R. Kalluri. The role of endothelial-to-mesenchymal transition in cancer progression. Br. J. Cancer 99 (2008) 1375-1379.

15. A.J. Maniotis, R. Folberg, A. Hess, E.A. Seftor, L.M. Gardner, J. Pe'er, J.M. Trent, P.S. Meltzer, M.J. Hendrix. Vascular channel formation by human melanoma cells in vivo and in vitro: vasculogenic mimicry. Am. J. Pathol. 155 (1999) 739-752.

16. J.P. Thiery, D. Chopin. Epithelial cell plasticity in development and tumor progression. Cancer Metastasis Rev. 18 (1999) 31-42.

17. J.P. Thiery, H. Acloque, R.Y.J. Huang, M.A. Nieto. Epithelial-mesenchymal transitions in development and disease. Cell 139 (2009) 871-890.

18. E.M. Zeisberg, S. Potenta, L. Xie, M. Zeisberg, R. Kalluri. Discovery of endothelial to mesenchymal transition as a source for carcinoma-associated fibroblasts. Cancer Res. 67 (2007) 10123-10128

19. T. Upile, W. Jerjes, H. Radhi, M. Al-Khawalde, P. Kafas, S. Nouraei, H. Sudhoff. Vascular mimicry in cultured head and neck tumour cell lines. Head Neck Oncol. 23 (2011) 55.

20. S. Zhang, D. Zhang, B. Sun. Vasculogenic mimicry: current status and future prospects. Cancer Lett. 254 (2007) 157-164.

21. Y.W. Paulis, P.M. Soetekouw, H.M. Verheul, V.C. Tjan-Heijnen, A.W. Griffioen. Signalling pathways in vasculogenic mimicry. Biochim. Biophys. Acta. 1806 (2010) 18-28.

22. B. Herren, B. Levkau, E.W. Raines, R. Ross. Cleavage of beta-catenin and plakoglobin and shedding of VE-cadherin during endothelial apoptosis: evidence for a role for caspases and metalloproteinases. Mol. Biol. Cell 9 (1998) 1589-1601.

164

23. N. Sharma, R.E. Seftor, E.A. Seftor, L.M. Gruman, P.M. Jr Heidger, M.B. Cohen, D.M. Lubaroff, M.J. Hendrix. Prostatic tumor cell plasticity involves cooperative interactions of distinct phenotypic subpopulations: role in vasculogenic mimicry. Prostate 50 (2002) 189-201.

24. A. Liew, F. Barry, T. O'Brien. Endothelial progenitor cells: Diagnostic and therapeutic considerations. Bioessays 28 (2006) 261-270.

25. J.C. Voyta, D.P. Via, C.E. Butterfield, B.R. Zetter. Identification and isolation of endothelial cells based on their increased uptake of acetylated-low density lipoprotein. J. Cell Biol. 99 (1984) 2034-2340.

26. A.M. Fogelman, B.J. Van Lenten, C. Warden, M.E. Haberland, P.A. Edwards. Macrophage lipoprotein receptors. J. Cell Sci. Suppl 9 (1988) 135-149.

27. T.J. Van Berkel, M. Van Eck, N. Herijgers, K. Fluiter, S. Nion. Scavenger receptor classes A and B. Their roles in atherogenesis and the metabolism of modified LDL and HDL. Ann. N. Y. Acad. Sci. 902 (2000) 113-126.

28. V.V. Kunjathoor, M. Febbraio, E.A. Podrez, K.J. Moore, L. Andersson, S. Koehn, J.S. Rhee, R. Silverstein, H.F. Hoff, M.W. Freeman. Scavenger receptors class AI/II and CD36 are the principal receptors responsible for the uptake of modified low density lipoprotein leading to lipid loading in macrophages. J. Biol. Chem. 277 (2002) 49982-49988.

29. E. Pérez-Gómez, M. Villa-Morales, J. Santos, J. Fernández-Piqueras, C. Gamallo, J. Dotor, C. Bernabéu, M. Quintanilla. A role for endoglin as a suppressor of malignancy during mouse skin carcinogenesis. Cancer Res. 67 (2007) 10268-10277.

30. S. Venkatesha, M. Toporsian, C. Lam, J. Hanai, T. Mammoto, Y.M. Kim, Y. Bdolah, K.H. Lim, H.T. Yuan, T.A. Libermann. Soluble endoglin contributes to the pathogenesis of preeclampsia. Nat. Med. 12 (2006) 642-649.

31. J. Sai, D. Raman, Y. Liu, J. Wikswo, A. Richmond. Parallel phosphatidylinositol 3- kinase (PI3K)-dependent and src-dependent pathways lead to CXCL8-mediated Rac2 activation and chemotaxis. J. Biol. Chem. 283 (2008) 26538-26547.

32. M.J. Wheelock, Y. Shintani, M. Maeda, Y. Fukumoto, K.R. Johnson. Cadherin switching. J. Cell Sci. 121 (2008) 727-735.

33. R.B. Hazan, R. Qiao, R. Keren, I. Badano, K. Suyama. Cadherin switch in tumor progression. Ann. N. Y. Acad. Sci. 1014 (2004) 155-163.

34. G.B. Sajithlal, T.F. McGuire, J. Lu, D. Beer-Stolz, E.V. Prochownik. Endothelial- like cells derived directly from human tumor xenografts. Int. J. Cancer 127 (2010) 2268-2278.

165

35. A.S. Holpuch, G.J. Hummel, M. Tong, G.A. Seghi, P. Pei, P. Ma, R.J. Mumper, S.R. Mallery. Nanoparticles for local drug delivery to the oral mucosa: Proof of principle studies. Pharm. Res. 27 (2010) 1224-1236.

36. F. Antohe, M.J. Poznansky, M. Simionescu. Low density lipoprotein binding induces asymmetric redistribution of the low density lipoprotein receptors in endothelial cells. Eur. J. Cell. Bio. 78 (1999) 407-415.

37. J.R. Krycer, I. Kristiana, A.J. Brown. Cholesterol homeostasis in two commonly used human prostate cancer cell-lines, LNCaP and PC-3. PLoS One 4 (2009) e8496.

38. D.E. Sadava. Cell biology: organelle structure and function. Jones and Bartlett Publishers, Boston, 1993, pp 66.

39. A. Nawshad, D. Lagamba, A. Polad, E.D. Hay. Transforming growth factor-β signaling during epithelial-mesenchymal transformation: Implications for embryogenesis and tumor metastasis. Cells Tissues Organs 179 (2005) 11-23.

40. S.L. Lu, D. Reh, A.G. Li, J. Woods, C.L. Corless, M. Kulesz-Martin, X.J. Wang. Overexpression of transforming growth factor beta1 in head and neck epithelia results in inflammation, angiogenesis, and epithelial hyperproliferation. Cancer Res. 64 (2004) 4405-4410.

41. S.N. Sinpitaksakul, A. Pimkhaokham, N. Sanchavanakit, P. Pavasant. TGF-beta1 induced MMP-9 expression in HNSCC cell lines via Smad/MLCK pathway. Biochem. Biophys. Res. Commun. 371 (2008) 713-718.

42. H. Teraoka, T. Sawada, T. Nishihara, M. Yashiro, M. Ohira, T. Ishikawa, H. Nishino, K. Hirakawa. Enhanced VEGF production and decreased immunogenicity induced by TGF-beta 1 promote liver metastasis of pancreatic cancer. Br. J. Cancer. 85 (2001) 612-617.

43. L. Artese, C. Rubini, G. Ferrero, M. Fioroni, A. Santinelli, A. Piattelli. Microvessel density (MVD) and vascular endothelial growth factor expression (VEGF) in human oral squamous cell carcinoma. Anticancer Res. 21 (2001) 689-695.

44. C. Bernabeu, J.M. Lopez-Novoa, M. Quintanilla. The emerging role of TGF-[beta] superfamily coreceptors in cancer. Biochimica. Biophysica. Acta. 1792 (2009) 954- 973.

45. E. Fonsatti, M. Altomonte, M.R. Nicotra, P.G. Natali, M. Maio. Endoglin (CD105): A powerful therapeutic target on tumor-associated angiogenetic blood vessels. Oncogene 22 (2003) 6557-6563.

166

46. Y. Liu, B. Jovanovic, M. Pins, C. Lee, R.C. Bergan. Over expression of endoglin in human prostate cancer suppresses cell detachment, migration and invasion. Oncogene 21 (2002) 8272-8281.

47. D. Oxmann, J. Held-Feindt, A.M. Stark, K. Hattermann, T. Yoneda, R. Mentlein. Endoglin expression in metastatic breast cancer cells enhances their invasive phenotype. Oncogene 27 (2008) 3567-3575.

48. A.K. Olsson, A. Dimberg, J. Kreuger, L. Claesson-Welsh. VEGF receptor signalling? in control of vascular function. Nat Rev Mol Cell Biol 7 (2006) 359-371.

49. K.A. Rodrigo, Y. Rawal, R.J. Renner, S.J. Schwartz, Q. Tian, P.E. Larsen, S.R. Mallery. Suppression of the tumorigenic phenotype in human oral squamous cell carcinoma cells by an ethanol extract derived from freeze-dried black raspberries. Nutr. Cancer. 54 (2006) 58-68.

50. S.K. Williamson, J. Moon, C.H. Huang, P.P. Guaglianone, M. LeBlanc, G.T. Wolf, S.G. Urba. Phase II evaluation of sorafenib in advanced and metastatic squamous cell carcinoma of the head and neck: Southwest oncology group study S0420. J. Clin. Oncol. 28 (2010) 3330-3335.

51. B. Davidson, I. Goldberg, W.H. Gotlieb, L. Lerner-Geva, G. Ben-Baruch, J. Kopolovic. Ulex Europaeus lectin and anti-CD31 staining in squamous cell carcinoma of the uterine cervix: potential prognostic markers. Int. J. Gynecol. Pathol. 17 (1998) 205-210.

167

Figure A.1. HNSCC cells express endothelial-associated proteins. (A)

Immunocytochemistry staining of the vascular endothelial intercellular adhesion molecule VE-cadherin on HUVECs and HNSCC cells. HUVECs served as the positive control. (B) Western Blots demonstrated the presence of VE-cadherin protein in HNSCC cells and also revealed a cell-line dependent level of expression. The additional lower molecular weight band shown in HUVEC lysate likely represents a proteolytic fragment of VE-cadherin during apoptosis of a subpopulation of HUVECs [22]. Ratios of VE- cadherin band density versus matched β-actin band density are presented at the bottom.

Normal HOKs did not express VE-cadherin by either ICC staining (data not shown) or

Western Blot (analyzed on a separate gel).

168

Figure A.2. HNSCC cells express vimentin and CD31 proteins. Left four panels: Double staining of endothelial/mesenchymal intracellular intermediate filament vimentin (red) and its epithelial counterpart cytokeratin 6 (green). While HUVECs showed universal expression of vimentin (A), only a subpopulation of cells in each HNSCC cell line demonstrated vimentin positive staining (C, E, G). Cytokeratin 6 (green) was exclusively expressed in HNSCC cells(C, E, G). Right four panels: Immunocytochemistry staining of

CD31. HUVECs exhibited classic membrane expression of CD31 (B). A subpopulation of HNSCC cells demonstrated cytosolic staining of CD31 (D, F, H).

169

Figure A.3. HUVECs and HNSCC cells form tubular networks on Matrigel. Log growth

HUVEC, HOK and HNSCC cells were added to Matrigel pre-coated 24-well plates and then incubated for 17 hrs. The HUVEC cultures generated delicate tubular structures whereas the HNSCC cells aligned in a coarser network. HOK cells, however, remained in a dispersed monolayer and did not display the same cell interaction patterns.

170

Figure A.4. HNSCC cells internalize AcLDL, which is enhanced following TGF-β1 challenge. (A) Qualitative studies. Subconfluent HUVECs, which served as the positive control, HOKs, or HNSCC cells were exposed to 10 μg/ml Alexa Fluor 488 conjugated

AcLDL (green dots) for 4 h. Extracellular Alexa Fluor 488 signals were quenched with

0.08% Trypan Blue followed by two washes with PBS. CD31 and E-cadherin were stained (red) to delineate HUVECs or HOKs and HNSCC intercellular boundaries respectively. Nuclei were counterstained with DAPI (blue). Normal oral mucosa was included as a positive control for E-cadherin. (B) Quantitative studies. HNSCC cells were treated with 0 (control) or 10 ng/ml TGF-β1 (24 h) followed by Alexa Fluor 488 conjugated AcLDL incubation and Trypan Blue quenching. AcLDL internalization data were then obtained and quantified relative to an assay specific standard curve. Data were normalized to ng LDL per 105 cells. HUVECs (n = 8) and the transformed monocyte cell line U937 (n = 3) were employed as positive control populations.

171

Figure A.4

172

Figure A.5. TGF-β1 induces an endotheliod phenotype in HNSCC cells. (A) SCC15 cells were treated with 0 (control) or 10 ng/ml TGF-β1 for 24 h prior to immnunocytochemistry staining of epithelial phenotypic intercellular adhesion molecule

E-cadherin and three endotheliod markers i.e. CD31, vimentin and VE-cadherin.

Microscopic field-matched images of nuclei (labeled with DAPI) were included to demonstrate the comparable cell densities between the control and TGF-β1 treated groups. (B) Endoglin ELISAs were conducted on cell lysates of HUVECs (positive control) and three HNSCC cell lines either pretreated with 10 ng/ml TGF-β1 or control for 48 h. Endoglin levels were normalized to pg per mg total protein, data expressed as mean ± S.E.M. Y axis was sectioned due to the great difference of endoglin levels between HUVECs and HNSCC cells. (C) SCC9 and SCC15 cells were exposed to 0

(control) or 10 ng/ml TGF-β1 for 24 h prior to the 48 h cell invasion assay. Invaded cell numbers were determined relative to the cell line specific standard curves. Error bars represent standard error of mean, n = 5.

173

Figure A.5

174

Figure A.6. Endothelial associated proteins are also present in HNSCC tumors. Five intact normal oral epithelium samples, 10 ulcerated, non-neoplastic oral mucosal biopsies and 24 HNSCC tumor tissues respectively underwent IHC staining for pancytokeratin and three traditionally mesenchymal and/or endothelial markers i.e. vimentin, CD31 and

VE-cadherin. Images demonstrated in each column represent the same microscopic field of a same tissue specimen. Normal, intact oral epithelium is characterized by extensive cytokeratin positivity, with vimentin, CD31 and VE-cadherin positivity restricted to connective tissue cells (vimentin) and endothelium (CD31 and VE-cadherin). In contrast, oral epithelium adjacent to an ulcer demonstrates vimentin positivity. Primary HNSCC tumors showed intense keratin positivity, moderate to high vimentin and CD31 expression and modest VE-cadherin. Metastatic tumor cells and advancing tumor edges demonstrated higher levels of “endotheliod” markers. Interestingly, many “endotheliod marking” tumor cells retained dual expression of cytokeratin. In the peri-necrotic tissue zones, CD31 positivity was observed in endothelial cells, some infiltrating inflammatory cells and the HNSCC cells.

175

Figure A.6

176

Figure A.7. VEGF activates comparable signaling pathways in HNSCC and endothelial cells. (A) VEGF induces a rapid phosphorylation of Erk1/2 and Src. 24 h serum deprived

CAL27 cells were challenged with 50 ng/ml VEGF for 0, 1, 5, 10, 20, 30 or 60 min. Cell lysates were then obtained for Western Blot analyses of both total and phosphorylated levels of Erk1/2 and Src. (B) Endostatin (10 μg/ml) pretreatment diminished CAL27 intracellular signaling, as manifested by a reduction in VEGF-mediated phosphorylation of Erk1/2 and Src relative to control cultures. Notably, a 60 min endostatin pretreatment ablated CAL27 responsiveness to VEGF-induced phosphorylation. Intensity ratios of phosphorylated and matched total proteins were presented at the bottom of each section.

177

Figure A.7

178

Figure A.8. Densitometric analyses results of Western blot data presented in Figure A.7.

(A) and (B) VEGF induces a rapid phosphorylation of Erk1/2 and Src in CAL27 cells.

Both phosphorylated and total protein levels of Erk1/2 and Src were normalized relative to the levels of beta-actin, which serves as the loading control. Ratios of phosphorylated protein versus total protein levels were then calculated for both Erk1/2 and Src. (C) and

(D) Endostatin pretreatment (10 μg/ml, for 10, 30, or 60 min) diminished VEGF-induced

CAL27 intracellular signaling, as manifested by a reduction in VEGF-mediated phosphorylation of Erk1/2 and Src relative to control cultures.

179

Appendix B: Evaluation of a Mucoadhesive Fenretinide Patch for Local Intraoral

Delivery: A Strategy to Re-introduce Fenretinide for Oral Cancer Chemoprevention

B.1 Abstract

Systemic delivery of fenretinide in oral cancer chemoprevention trials has been largely unsuccessful due to dose-limiting toxicities and sub-therapeutic intra-oral drug levels. Local drug delivery, however, provides site-specific therapeutically-relevant levels while minimizing systemic exposure. These studies evaluated the pharmacokinetic and growth-modulatory parameters of fenretinide mucoadhesive patch application on rabbit buccal mucosa. Fenretinide and blank-control patches were placed on right/left buccal mucosa, respectively, in 8 rabbits (30 minutes, q.d., 10-days). No clinical or histological deleterious effects occurred. LC-MS/MS analyses of post-treatment samples revealed a delivery gradient with highest fenretinide levels achieved at the patch-mucosal interface (no metabolites), pharmacologically-active levels in fenretinide-treated oral mucosa (mean: 5.65M; trace amounts of 4-oxo-4-HPR), and undetectable sera levels.

Epithelial markers for cell proliferation (Ki-67), terminal differentiation

(transglutaminase 1 – TGase1), and glucuronidation (UDP-glucuronosyltransferase1A1 –

UGT1A1) exhibited fenretinide concentration-specific relationships (elevated TGase1 180 and UGT1A1 levels <5M, reduced Ki-67 indices >5M) relative to blank-treated epithelium. All fenretinide-treated tissues showed significantly increased intraepithelial apoptosis (TUNEL) positivity, implying activation of intersecting apoptotic and differentiation pathways. Human oral mucosal correlative studies showed substantial inter-donor variations in levels of the enzyme (cytochrome P450 3A4) responsible for conversion of fenretinide to its highly active metabolite, 4-oxo-4-HPR. Complementary in vitro assays in human oral keratinocytes revealed fenretinide and 4-oxo-4-HPR’s preferential suppression of DNA-synthesis in dysplastic as opposed to normal oral keratinocytes. Collectively, these data showed that mucoadhesive patch-mediated fenretinide delivery is a viable strategy to re-introduce a compound known to induce keratinocyte differentiation to human oral cancer chemoprevention trials.

B.2 Introduction

Over the past several decades, many promising cancer-preventing compounds have been evaluated in patients with oral premalignant lesions [reviewed in 1]. Systemic delivery trials which predominantly relied on oral capsule administration induced dose- limiting systemic toxicities and lacked initial and/or sustained efficacy [1]. Formulations used in local delivery trials, e.g., mucoadhesive gels [2-4] and rinses, demonstrated a range of therapeutic efficacies largely without deleterious side effects [1]. Notably, only three of the nineteen reviewed trials (i.e., two local and one systemic) quantified compound levels achieved at the target site [1]. This lack of data precludes determination of whether the compounds evaluated were pharmacologically ineffective or failed to reach therapeutic levels in lesional tissues (i.e., oral epithelium). 181

Many of these oral cancer chemopreventive studies evaluated vitamin A, its precursors, and analogs (retinoids) [1]. In vitro, fenretinide (a synthetic analog of all- trans retinoic acid) has shown exceptional capacity to promote keratinocyte terminal differentiation or apoptosis in a dose-dependent fashion [5]. Furthermore, fenretinide exhibits a reduced toxicity-induction profile (i.e., decreased gastrointestinal distress and nyctalopia), and thus has been an agent of choice for recent oral cancer chemoprevention clinical trials [6-9]. Fenretinide trials in patients with oral dysplastic lesions, however, have been largely unsuccessful [7-9]. Notably, these studies investigated systemic delivery of fenretinide (oral capsule) at varying concentrations (low-dose: 100 mg b.i.d. or 200 mg q.d., high-dose: 900 mg b.i.d.), which resulted in minimal therapeutic efficacy accompanied by dose-limiting toxicities [7-9]. Sera levels, which were used as a surrogate marker for target tissue levels, never achieved therapeutic concentrations [7-9].

Issues such as first pass metabolism in the liver and reliance upon perfusion from the underlying vasculature to overlying target surface epithelium likely compromised levels achieved from systemic delivery. In contrast, a pilot local delivery trial in which patients placed fenretinide capsule contents on a variety of reactive and preneoplastic oral lesions did not demonstrate any local or systemic deleterious effects [10].

Our labs obtained promising results from a local delivery strategy that evaluated the effects of a 10% freeze-dried black raspberry (BRB) gel in patients with oral dysplastic lesions [2,3]. Results from these and additional recent studies showed: local gel delivery provided a pharmacologic advantage [4], a subset of patients responded favorably to local BRB gel application [2], and differential bioactivation and retention of chemopreventive compounds in human oral mucosa [11]. These data, which implied that 182

BRB was insufficient to induce regression in some patients’ dysplastic oral lesions, served as the impetus for the current study. Considering these clinical implications, our laboratories recently developed a novel mucoadhesive patch which provided improved site-specific intraoral delivery of fenretinide [12]. The goal of the current study was to determine if mucoadhesive patches delivered therapeutically relevant fenretinide levels to oral mucosa. Subsequent evaluations of fenretinide-treated and blank-treated rabbit oral tissues assessed the modulation of chemopreventive and metabolic parameters.

Corresponding studies of human oral tissues and cultured oral keratinocytes recapitulated the rabbit data and provided human clinical relevance.

B.3 Materials and Methods

Intraoral mucoadhesive fenretinide patch pharmacokinetic analyses in New

Zealand white rabbits. Rabbit studies were conducted with approval from The Ohio

State University’s Institutional Animal Care and Use Committee. Eight female New

Zealand white rabbits (Harlan Laboratories, Indianapolis, IN, USA) weighing 2.710.18 kg were used for analysis of intraoral fenretinide delivery from mucoadhesive patches

[12]. Upon arrival, rabbits were acclimated for 7 days and monitored for signs of distress. Following acclimation, each rabbit was sedated with a subcutaneous injection

(0.2cc) of acepromazine (2 mg/ml, Butler Schein Animal Health, Dublin, OH, USA) and placed under general anesthesia via isoflurane inhalation (2-3 v/v%) for the entire 30-

183 minute patch application procedure. Once unconscious and prior to patch placement,

0.5cc of blood was collected from the central ear artery. A fenretinide-loaded

184

Figure B.1. Mucoadhesive patch attachment on rabbit oral mucosa. Fenretinide mucoadhesive patches were attached (q.d., 30 minutes for 10 consecutive days) to the right buccal mucosa (blank patches on left buccal mucosa) immediately posterior to the intra-oral commissure of the upper and lower lips. Pre- and post-treatment blood specimens and patch-mucosal interface samples were collected daily, and fenretinide- treated and blank-treated oral mucosal biopsies were harvested upon completion of the 10-day study. All samples were analyzed via LC-MS/MS to detect the levels of fenretinide and its metabolites (4-oxo-4-HPR and 4-MPR).

185 mucoadhesive patch (0.5 mg fenretinide/patch, i.e., 400- to 3,600-fold less than daily systemic administrations in recent clinical trials [7-9]) was attached to the right buccal mucosa and a blank control patch attached to the left buccal mucosa, both immediately posterior to the intraoral commissure of the upper and lower lips (Figure B.1). Patches were left in place for 30 minutes and an additional blood sample collected at 30 minutes.

The patches were then removed, fenretinide-treated underlying surface epithelium lavaged with 200 l of 1x phosphate buffered saline (PBS), and patch-mucosa interface sample immediately collected. This procedure was conducted q.d. for 10 consecutive days.

Following treatment on day 10, rabbits were sacrificed via intravenous potassium chloride injection, and oral biopsies of the fenretinide-treated and blank-treated mucosal sites were harvested. Modulation of chemopreventive and metabolic parameters relative to intra-mucosal fenretinide concentrations were compared in intra-rabbit blank-treated versus fenretinide-treated mucosal specimens, while patch-mucosal interface samples provided inter-rabbit comparisons of daily patch delivery efficacy. Notably, these oral biopsies were cut in two pieces: half for LC-MS/MS analysis (stored in stabilizing buffer: pH 6.5, 1x PBS + 9 mM EDTA + 25 mM sodium ascorbate + 21 mM sodium sulfate) and the other half for protein analysis (half for immunohistochemistry: 10% formalin, half for

Western blot: RNALater). All tissue and patch-mucosa interface LC-MS/MS samples were immediately frozen and stored at -80C until analysis. Blood samples were clotted, centrifuged at 10,000xg, serum supernatant collected and stored at -80C until analysis.

186

Preparation of rabbit patch-mucosa interface, oral mucosa homogenate and serum samples for LC-MS/MS analysis. Due to the high concentration of fenretinide present in the patch-mucosa interface samples, they were pre-diluted prior to analysis

(i.e., serially diluted in 1x PBS 200- and 2,000-fold). 100 l of the 200- or 2,000-fold diluted samples were mixed with 10 l of internal standard (hesperetin, 10 g/l).

Samples were extracted with 800 l ethyl acetate for 30 minutes by mechanical shaking.

For mucosa homogenate samples, both fenretinide and blank-treated samples were grossly dissected to remove excess amounts of underlying connective tissue. Oral epithelium was homogenized in 1x PBS for 3x5 minutes on a bead mill, and 260 l of the homogenate was spiked with 10 l of internal standard and 300 l of lysis buffer.

Samples were incubated on ice for 5 minutes, centrifuged at 11,000×g for 1 minute and the supernatant was extracted with 1 ml ethyl acetate for 30 minutes. To process rabbit serum samples, 100 µl of each rabbit serum sample was spiked with 10 l of internal standard followed by extraction using 800 l ethyl acetate for 30 minutes. After the extractions, all the samples were centrifuged at 11,000×g for 1 minute and ethyl acetate fraction was collected and evaporated to dryness under nitrogen stream. The residue was reconstituted in 100 µl of 50% methanol containing 0.2% formic acid and subsequently subjected to LC-MS/MS for analysis.

LC-MS/MS analysis. The LC-MS/MS analysis was conducted on a Thermo TSQ

Quantum triple quadrupole mass spectrometer equipped with an electrospray ionization source, Shimadzu LC-20AD HPLC pump, and LC-20AC autosampler. The samples were analyzed on a Thermo Betabasic C8 column (50×2.1 mm, 5 µm), of which the mobile

187 phase (85% methanol with 0.2% formic acid) was used in isocratic mode with 0.2 ml/min flow rate. The analysis was conducted in positive mode. The transfer line temperature was 325ºC. The multiple reaction monitoring (MRM) transitions selected (based on parameters established by fenretinide/metabolite-spiked blank rabbit oral mucosal tissues and sera) for the analysis of fenretinide, 4-oxo-HPR, 4-MPR and the internal standard were 392>283, 406>297, 406>283 and 303>177 m/z, respectively.

Conversion of fenretinide mass quantities to molar equivalents. In order to compare oral mucosal fenretinide levels (ng fenretinide/gm tissue) to previously published therapeutic values (i.e., 1-10 M), mass quantities were converted to micromolar concentrations, by standards previously published [13].

Evaluation and quantification of chemoprevention-associated markers and metabolic enzyme distribution within the rabbit surface epithelium by immunoblotting and immunohistochemical staining. The presence/absence of deleterious side effects within the fenretinide-treated and blank-treated rabbit oral mucosa was assessed by standard hematoxylin and eosin staining of paraffin-embedded tissues. Rabbit mucosal tissues were first screened by immunoblotting to determine the presence or absence of proteins associated with or responsive to fenretinide. Matched rabbit oral mucosal specimens were also analyzed by immunohistochemistry (IHC) to elucidate levels of immunoblot-confirmed metabolic enzymes and chemopreventive endpoints within the target oral epithelia. The antibodies employed for Western blot and IHC staining were: goat polyclonal UGT1A1 (WB: 1:100, IHC: 1:300), goat polyclonal UGT1A6 (WB:

1:100, IHC: not conducted), mouse monoclonal CYP3A4 (WB: 1:250, IHC: not

188 conducted), mouse monoclonal CYP26A1 (WB: 1:250, IHC: not conducted), goat polyclonal CYP2C8/9/18/19 (WB: 1:500, IHC: not conducted), mouse monoclonal Ki-67

(WB: not conducted – highly specific in IHC applications [14], IHC: 1:150), mouse monoclonal keratinocyte-specific transglutaminase 1 (TGase 1; WB: 1:100, IHC: 1:50 and 1:300) antibodies (Santa Cruz Biotechnology, Santa Cruz, CA, USA). Blocking buffer (negative control) was used in place of the primary antibody, and all tissue sections (including negative control) were incubated with their respective biotinylated secondary antibody (1:200, Vector Laboratories, Burlingame, CA, USA). Vectastain

ABC reagent (Vector Laboratories) was applied, and identically timed immunoreactions were conducted with DAB substrate and hematoxylin counterstain. Evaluation of apoptotic indices in fenretinide-treated versus blank-treated rabbit epithelium was assessed via an immunofluorescent terminal deoxynucleotidyl transferase dUTP nick end labeling (TUNEL) assay (Roche Diagnostics, Indianapolis, IN, USA).

Immunohistochemical and immunofluorescent images were obtained with identical magnification and exposure limits using a Nikon DS-Fi-1 high-resolution digital camera and analyzed via Image-Pro Plus 6.2 software (Media Cybernetics, Bethesda, MD, USA).

Image analysis and quantification was conducted on the positive reacted antibodies: Ki-67, TGase 1, TUNEL, and UGT1A1 tissue sections. Quantification of positive Ki-67 nuclear staining was conducted by selecting the basal epithelial layer of cells as the area of interest, setting limits of staining intensity considered to be positive

(identical for intra-rabbit tissues), and quantifying the positive nuclear indices relative to basement membrane length (i.e., Ki-67 positive nuclei/length of epithelial basement

189 membrane). Similarly, quantification of the TGase 1, TUNEL, and UGT1A1 tissue specimens involved selection of the positive staining intensity limits (identical for intra- rabbit tissues) within the area of interest (i.e., entire oral epithelium excluding the non- staining superficial orthokeratin layer) and quantification by area of positive staining relative to total epithelial area of interest. Notably, these methods of immunohistochemistry quantification are restrictive to intra-rabbit comparisons (i.e., inter-rabbit comparisons are not appropriate due to variation of the selected inter-rabbit positivity limits).

Evaluation of metabolic enzyme heterogeneity in human oral mucosa by Western blot. Eight clinically and histologically normal human oral mucosal tissues were collected from patients undergoing elective oral surgical procedures. Human subject participation was approved by Ohio State University’s Institutional Review Board and followed the tenets of the Declaration of Helsinki 1964. Tissues were collected and immediately bisected for Western blot (mammalian protein extraction reagent; subsequent homogenization, protein extraction, and Bradford protein quantification) or immunohistochemical (10% formalin; subsequent standard tissue processing) analyses.

Western blot analyses were conducted using the iBlot Western blot system (Invitrogen,

Carlsbad, CA, USA) on 4-12% Bis-Tris gels (Invitrogen). Thirty micrograms of human oral mucosal protein was loaded in each well, pooled human liver microsomes (BD

Biosciences, San Jose, CA, USA) were loaded as a positive control, separated by SDS-

PAGE, and blotted by the standard iBlot protocol (i.e., nitrocellulose stacks and 7 minute transfer time). Blots were incubated in 5% non-fat milk blocking buffer overnight at 4C

190 with the following primary antibodies: CYP3A4 (1:250), CYP26A1 (1:250), CYP2C8

(1:500), UGT1A1 (1:250), UGT1A6 (1:250), indolethylamine N-methyltransferase

(rabbit polyclonal; INMT, 1:1000; Sigma-Aldrich, Saint Louis, MO, USA), and Beta- actin loading control (1:20,000, Santa Cruz). Each blot was subsequently incubated at room temperature for 1 hour with the respective horseradish peroxidase secondary antibody: goat anti-mouse, goat anti-rabbit, or donkey anti-goat (1:1000, Santa Cruz) in blocking buffer. Following secondary antibody incubation blots were incubated with the

ECL Plus Western blot detection system (Amersham GE Healthcare Life Sciences,

Buckinghamshire, UK) and exposed on CL-Xposure films (Kodak, Rochester, NY,

USA). Quantification of positive immunoreactive bands (i.e., CYP3A4) were assessed via densitometry analysis (Kodak 1D3 image analysis software; Kodak) and results normalized relative to the expression of endogenous β-actin.

Determination of site-specificity of metabolic enzymes in human oral mucosa by immunohistochemical analyses. Although Western blots demonstrate relative levels of the protein of interest within the full-thickness oral mucosa specimens, they do not specify relative levels of tissue distribution in the targeted treatment site, i.e., oral epithelium. To identify the site-specific distribution of protein levels in oral mucosa, immunohistochemical analyses were therefore conducted for CYP3A4 (1:100) by the methods previously described.

Assessment of cell proliferation following fenretinide and 4-oxo-4-HPR treatments of human oral keratinocytes in vitro. Normal human oral keratinocytes

(HOK) were cultured in oral keratinocyte growth medium with oral keratinocyte growth

191 supplement and penicillin/streptomycin (ScienCell, Carlsbad, CA, USA). A dysplastic cell strain was developed through stable transduction of the human papillomavirus E6/E7 genes in the HOKs, as per previously published methods [15]. HOK E6/E7 transduction was confirmed via PCR and Western blot (data not shown).

Evaluation of the effects of fenretinide (4-HPR) and its bioactive metabolite 4- oxo-4-HPR on cell proliferation was conducted using the BrdU cell proliferation ELISA

(Roche). Briefly, log growth HOK or HOK E6/E7 cells were seeded at 1x104 cells per well in 24-well plates. Cells were treated daily with vehicle (0.01% DMSO), 1, 5, or 10

M 4-HPR or 4-oxo-4-HPR for 24, 48, and 72-hour time points. BrdU (10 M) was added 24-hours prior to the respective cell harvesting time point, where subsequent detection (absorbance: 370 nm) of BrdU incorporation followed the standard protocol

(Roche) on a FLUOstar Omega plate reader (BMG Labtech, Cary, NC, USA).

Statistical analysis. Normality of data was assessed using the Shapiro-Wilks test to determine the appropriate use of parametric or nonparametric statistical tests.

Fenretinide-treated versus blank-treated rabbit mucosal specimens were compared using the Wilcoxon matched-pairs signed-ranks test. Inter-rabbit variations of mean 10-day fenretinide levels at the patch-mucosal interface were evaluated with the Kruskal-Wallis nonparametric analysis of variance. Fenretinide-treated and blank-treated immunohistochemical sections were compared using the Wilcoxon matched-pairs signed- ranks test (TGase 1 and Ki-67) and paired t-test (TUNEL and UGT1A1). BrdU incorporation in HOK and HOK E6/E7 cells following 4-HPR and 4-oxo-4-HPR was

192 evaluated using a Kruskal-Wallis nonparametric analysis of variance followed by a

Dunn’s multiple comparisons post-test.

B.4 Results

Therapeutically relevant fenretinide levels are achieved in rabbit oral mucosa without deleterious side effects. Intraoral fenretinide mucoadhesive patch application delivered a fenretinide gradient with the highest levels achieved at the patch-mucosal interface, pharmacologically active levels in treated oral mucosa, and undetectable sera levels (Figure B.2A). Levels of fenretinide achieved at the fenretinide-treated patch- mucosal interface (mean±SEM: 11513.9±391.2 ng/mg protein) were comparable in all animals throughout the 10-day treatment period (p>0.5, n=8). Furthermore, levels in fenretinide-treated oral mucosa were significantly greater than their rabbit-matched blank-treated sites (fenretinide-treated mean±SEM: 2210.3±986.2 ng/gm tissue

[5.652.52M] versus blank-treated mean±SEM: 9.0±0.08 ng/gm tissue [0.020.00M], p<0.01, n=8). Despite achieving intra-mucosal fenretinide concentrations greater than the in vitro-established therapeutic range (i.e., 1-10 M) in rabbits VII (12.98 M) and VIII

(19.96 M) and the presence of a natural fenretinide reservoir, i.e., buccal fat pad, both clinical and histopathological assessments of treatment sites revealed normal oral mucosa, Figure B.2B. In addition, the inactive fenretinide metabolite 4-MPR was not detected in the patch-mucosal interface, oral mucosa, or serum samples, while trace

193 amounts (i.e., below the linear level of quantification) of 4-oxo-4-HPR were observed in the fenretinide-treated oral mucosal samples of rabbits III through VIII.

Fenretinide modulates oral epithelial growth state parameters.

Immunohistochemical analyses of fenretinide-treated tissue relative to matched blank- treated tissue exhibited fenretinide concentration-dependent patterns with regard to cell

194

Figure B.2. Intra-oral fenretinide patch application delivered pharmacological intraoral levels and did not induce any deleterious side-effects. A. Patch application delivered a fenretinide gradient with highest levels achieved at the patch-mucosal interface, therapeutically-relevant levels in the treated oral mucosa (meanSEM: 5.652.52 M, n=8), and undetectable levels in sera (linear limit of quantification: 1 ng/ml analyte). B. Histopathological evaluation of hematoxylin and eosin stained blank- treated and fenretinide-treated rabbit tissues revealed histologically normal oral mucosa. No evidence of either intra-epithelial or superficial connective tissue contact mucositis or inflammation was observed.

195 proliferation (Ki-67), terminal differentiation (TGase 1), and apoptosis (TUNEL).

Although the Ki-67 findings were not statistically significant, moderate decreases in cell proliferation were observed in oral mucosal tissues containing greater than 5 M fenretinide (i.e., rabbits VII and VIII, Figure B.3A-B). Similarly, TGase 1 levels were increased in oral mucosal tissues with levels of fenretinide in the range of 0.1-5 M (i.e., rabbits I-VI), while TGase 1 levels were decreased in tissues containing greater than 5

M (i.e., rabbits VII and VIII, Figure B.3A). This relationship was further demonstrated in the representative photomicrographs in Figure B.3B, in which rabbit IV (2.22 M) exhibited a 54.0% increase of TGase 1 levels in the fenretinide-treated relative to the blank-treated epithelium and rabbit VIII (19.96 M) exhibited a 24.0% decrease of

TGase 1 levels in the fenretinide-treated relative to blank-treated epithelium. Notably, the TGase 1 staining pattern in rabbit IV (differentiation-inducing levels) extended from the basal to the granular layers and was strongest in the keratinocyte cytosol. In contrast, the fenretinide-treated tissue from rabbit VIII (apoptosis-inducing levels) exhibited strong nuclear staining in the basal layer, which also extended to the spinous and granular layers. Furthermore, all of the fenretinide-treated oral epithelial tissues showed significantly increased apoptotic indices (Figure B.3A, p<0.01), which tended to be elevated in tissues with greater than 3 M fenretinide, as demonstrated in Figure B.3B.

Notably, inset photomicrographs in Figure B.3B represent positive indices (red highlighted areas – Ki-67 and TGase 1) or negative control (TUNEL – due to inherent auto-fluorescence in paraffin-embedded tissue sections).

196

197

Figure B.3. Quantified immunohistochemical analyses showed fenretinide’s dose-specific effects on treated rabbit oral epithelia growth state. A. Results of these studies showed that lower fenretinide levels (<5 M) elicited differentiation effects. While Ki-67 positivity was modestly reduced in epithelia of rabbits I through VI (<5 M), epithelial proliferation was markedly decreased at higher fenretinide levels (>5 M, rabbits VII and VIII). Similarly, TGase 1 levels (indicator of terminal differentiation) showed higher induction at lower fenretinide levels (<5 M). Apoptosis (TUNEL indices) was significantly increased in fenretinide-treated relative to blank-treated mucosal epithelia (n=8). These data, however, did not demonstrate an apparent dose-dependent relationship. B. IHC images were quantified (blank-treated relative to fenretinide-treated epithelia) by designating the stratum basale (Ki-67) or full-thickness epithelia (excluding orthokeratinized layer - TGase 1 and TUNEL) as the area of interest. Positive indices (denoted by red highlighted areas in the inset photomicrographs – Ki-67 and TGase 1) were divided by the total area of interest (TGase 1 and TUNEL) or by the basement membrane length (Ki-67; yellow line in the inset photomicrograph).

197

Figure B.4. Patch-delivered fenretinide effects intra-epithelial metabolic enzyme profile.

A. Human oral mucosa fenretinide metabolic pathways can include generation of the inactive metabolite (4-MPR), active metabolites (4-oxo-4-HPR and 4-HPR-O- glucuronide), and the functionally unknown metabolite (4-HPR-O-sulfate). Pathways that can contribute to local oral recycling of metabolites back to the parent compound are similarly indicated. Abbreviations and intra-oral locations are as follows: LPH, lactase phlorizin hydrolase (oral microflora); -Gluc, -Glucuronidase (human saliva); SULT, sulfotransferase (absent in human oral cavity); UGT1A, UDP-glucuronosyltransferase 1A

(human oral epithelium); CYP3A4, cytochrome P450 3A4 (human oral epithelium);

INMT, indolethylamine N-methyltransferase (absent in human oral cavity). B.

Differentiation-inducing levels of fenretinide (<5 M) increased intra-epithelial levels of

UGT1A1 relative to blank-treated tissues. C. Positive staining in the red highlighted area

(area of interest) was used to quantify UGT1A1 levels.

198

Figure B.4.

199

Figure B.4., cont’d

200

Induction of UGT1A1 levels in fenretinide-treated rabbit oral epithelium.

Immunohistochemical analysis of fenretinide-related phase I/II metabolic enzymes identified significantly increased UGT1A1 expression in fenretinide-treated relative to blank-treated rabbit oral mucosal tissues (p<0.01, Figure B.4B). Notably, the percent of

UGT1A1 induction was greatest in the mucosal samples with less than 5 M fenretinide

(i.e., rabbit I-VI, Figure B.4B). Representative photomicrographs demonstrated a dose- dependent effect of fenretinide on phase II enzyme induction (Figure B.4C). Additional fenretinide-specific metabolic enzymes (i.e., CYP3A4, CYP2C8, CYP26A1, and INMT) were not detected in rabbit oral mucosa (data not shown).

Human oral mucosa contained fenretinide-related metabolic enzymes capable of generating the bioactive 4-oxo-4-HPR metabolite and also exhibits appreciable inter- individual heterogeneity. In contrast to rabbit oral mucosa, human oral epithelia possessed CYP3A4 (Figure B.5). As demonstrated by immunoblotting, considerable inter-patient heterogeneity of protein levels was exhibited (i.e., 16-fold difference of

CYP3A4 in samples 2 and 5; Figure B.5). These levels of protein expression were not associated with the amount of epithelium in each sample, and therefore, reflected actual inter-patient variations. In addition, the fenretinide-related metabolic enzymes CYP2C8,

CYP26A1, and INMT were not detected in human oral mucosa, while results from

UGT1A1 and UGT1A6 immunoblots were inconclusive (data not shown).

Fenretinide and its bioactive metabolite 4-oxo-4-HPR modulated the growth state of cultured normal human oral keratinocytes. Evaluation of 4-HPR and its oxidized, bioactive metabolite 4-oxo-4-HPR treatment effects on BrdU incorporation in normal and

201 dysplastic oral keratinocytes showed a preferential suppression of DNA synthesis in dysplastic as opposed to normal keratinocytes. Dose, duration and treatment compound all impacted BrdU incorporation (p<0.05, Figure B.6). Although 4-HPR treatment did not significantly decrease DNA synthesis in dysplastic relative to normal cells at 24- hours, subsequent 48- and 72-hour time points showed significant inhibition in the 5 M treated oral dysplastic cells (p<0.05). 4-oxo-4-HPR treatment resulted in significantly decreased BrdU incorporation in dysplastic cells relative to normal at all time points (24- hours: 10 M, 48-hours: 10 M, 72-hours: 5 M; p<0.05).

B.5 Discussion

Despite its favorable chemopreventive profile in vitro, systemic administration of fenretinide in oral cancer chemoprevention clinical trials has demonstrated limited therapeutic efficacy and dose-limiting side effects [7-9]. In contrast, local delivery strategies can circumvent first pass liver metabolism, minimize systemic exposure, and deliver therapeutically-relevant drug levels to the target tissues. Location, however, is paramount when considering use of local delivery formulations. The oral cavity is an optimal site for local delivery as it is amenable to direct visualization, which facilitates both agent placement and clinical monitoring. The basis for this current study arose from our familiarity with local delivery formulations in conjunction with our enthusiasm to reintroduce fenretinide for clinical oral cancer chemoprevention trials.

202

These pharmacokinetic studies confirmed the therapeutic advantage imparted by intra-oral fenretinide mucoadhesive patch application, i.e., delivery of pharmacologically active levels of fenretinide to the rabbit oral mucosa (i.e., 5.65 M average) while

203

Figure B.5. Immunoblots of normal human oral mucosa revealed considerable inter- donor heterogeneity in a fenretinide bioactivating enzyme, CYP3A4. Similar to other intra-epithelial metabolic enzymes, CYP3A4 was primarily distributed in the lower epithelial layers (i.e., basilar and spinous), with decreased levels in the increasingly differentiated granular and cornified layers. Notably, levels of additional fenretinide- activating enzymes (i.e., CYP2C8 and CYP26A1) and the fenretinide- inactivating/eliminating enzyme (i.e., INMT) were not detected in the human oral epithelia. Although previous studies have demonstrated the presence of UGT1A enzymes in human oral epithelia [11], UGT1A1 and UGT1A6 immunoblots were indeterminate (data not shown).

204

Figure B.6. Fenretinide and its bioactive metabolite 4-oxo-4-HPR demonstrated preferential growth suppression towards premalignant (dysplastic) oral keratinocytes. Treatment of normal and dysplastic oral keratinocytes with doses of fenretinide or 4-oxo-4-HPR that are achievable via mucoadhesive patch delivery significantly decreased DNA synthesis in a dose, cell strain, and time-dependent fashion. Dysplastic keratinocytes were markedly more susceptible to suppression of DNA synthesis. These data convey the promise for mucoadhesive fenretinide patch application in oral cancer chemoprevention.

205 negating systemic exposure and toxicities. Our data also showed undetectable levels of the inactive metabolite 4-MPR and trace amounts of the potent metabolite 4-oxo-4-HPR in the fenretinide-treated rabbit oral mucosal samples. Accompanying Western blots did not show the presence of the major metabolic enzymes responsible for conversion of fenretinide to 4-MPR (amine N-methyltransferases, such as INMT) or 4-oxo-4-HPR

(CYP3A4, CYP2C8, CYP26A1) in rabbit oral mucosa [16,17]. The trace amount of 4- oxo-4-HPR detected could therefore reflect 4-HPR metabolism by other cytochrome

P450 enzymes or oral microflora. While technical challenges prevented LC-MS/MS detection of 4-HPR-O-glucuronide, the marked induction of UGT1A1 levels in rabbits I through VI suggested concomitant fenretinide metabolism to the more potent 4-HPR-O- glucuronide [18]. This finding was limited to the percent increase relative to the blank- treated control tissues (i.e., not total epithelial levels of UGT1A1), and therefore assumed increased metabolism with enzyme induction. Notably, the marked induction of

UGT1A1 in rabbits I through VI was coupled with low baseline enzyme levels in the blank-treated control tissues, while rabbits VII and VIII demonstrated elevated baseline levels and minimal induction. Similar to previous studies demonstrating maximal Phase

II enzyme expression [19], these findings also suggested that UGT1A1 is maximally expressed in rabbits VII and VIII.

Also observed in these studies was a delivery gradient, in which highest fenretinide levels were consistently achieved at the patch-mucosal interface, followed by second highest, variable levels within the targeted epithelium. Ideally, highest fenretinide levels should be achieved at the pivotal basal layer cells that direct keratinocyte growth and differentiation. The constant inter-rabbit levels at the patch-mucosal interface

206 suggested effective patch delivery, while the range of intra-mucosal fenretinide concentrations was likely attributed to inter-rabbit epithelial permeability variations.

This permeability issue was addressed in a recent study by our laboratories, which evaluated a fenretinide mucoadhesive patch formulated with the permeability enhancers propylene glycol and menthol [20]. This permeability-enhanced patch demonstrated consistent fenretinide penetration of rabbit oral mucosa following a single 30-minute patch application, which achieved comparable intra-mucosal levels to those observed in the current 10-day study [20]. Furthermore, these collective findings (i.e., similar intra- mucosal fenretinide levels after a single 30-minute patch application and 30-minute application q.d. for 10-days) suggested that therapeutic levels of fenretinide were achieved after each 30-minute patch application and were subsequently metabolized and cleared from the treatment site prior to subsequent patch application 24-hours later.

Similarly, these observations indicated that transient fenretinide levels affect keratinocyte protein translation in a therapeutic fashion.

A distinct rabbit versus human species variation was seen with regard to oral mucosal tissues’ fenretinide metabolic profile. The present and previous studies by our laboratory confirm the presence of highly variable levels of UGT1A and CYP3A4 enzymes in normal human oral epithelium, implying the probable metabolism of fenretinide to its active metabolites 4-HPR-O-glucuronide and 4-oxo-4-HPR in human applications [11]. Additionally, while the absence of INMT in human oral mucosa suggests the inability to form the inactive metabolite 4-MPR, other members of the amine

N-methyltransferase family could also inactivate fenretinide [16]. Collectively, the human fenretinide-related metabolic enzyme profile favors production of bioactive

207 metabolites, and exhibits the capacity for increased fenretinide retention at the treatment site via local enteric recycling (i.e., 4-HPR-O-glucuronide recycling by bacterial - glucuronidases [11]). The large inter-patient heterogeneity, however, will likely necessitate individual metabolic enzyme profiling to determine optimal duration and frequency of patch placement to achieve the desired therapeutic effect (i.e., patients with an elevated metabolic capacity could benefit from multiple doses throughout the day to effectively increase the drug levels at the target site).

The observed dose-dependent modulation of the keratinocyte growth state recapitulated previously published in vitro differentiation-associated concentrations and also confirmed that patch-delivered fenretinide retained its bioactive properties [5].

Although not statistically significant, levels of the enzyme responsible for cornified envelope formation, i.e., TGase 1, increased in all oral mucosa tissues containing 0.1 to 5

M fenretinide (i.e., previously established differentiation range). In contrast, tissues with fenretinide levels greater than 10 M (i.e., apoptotic range [5]) contained reduced

TGase 1 levels. Interestingly, apoptotic indices were not drastically increased in the oral tissues with greater than 10 M fenretinide, and did not inversely correlate with the

TGase 1 data. These preliminary findings could reflect the complexity and interaction of the pathways responsible for the keratinocyte transitioning from a proliferative growth state. These somewhat paradoxical results are supported by the concept that epithelial differentiation is a specialized form of apoptosis [21,22] and, key to the chemopreventive aspect, both pathways result in the keratinocyte leaving the proliferative pool.

Furthermore, several studies have shown nuclear translocation of the other transglutaminase isoform (TGase 2) resulted in cross-linking of transcription factors and

208 subsequent induction of apoptosis [23,24]. Although the relationship between TGase 1 nuclear staining and TUNEL positivity was not definitive, the prospect that TGase 1 fulfills a similar role in keratinocytes is highly probable. Ongoing studies in our labs are investigating these interactions.

Our data show that the Ki-67 proliferation indices in normal rabbit oral epithelia were not affected by fenretinide patch application. As preservation of an intact oral mucosal surface is essential for defense, these findings are favorable. Similarly, normal oral keratinocyte proliferation, as assessed by BrdU incorporation, was not perturbed by the addition of either fenretinide or 4-oxo-4-HPR. Treatment of dysplastic oral keratinocytes, however, showed both fenretinide and 4-oxo-4-HPR significantly suppressed DNA synthesis; suggesting the prospect for preferential targeting of dysplastic relative to normal keratinocytes. Similarly, if future studies demonstrate that

4-oxo-4-HPR provides greater therapeutic efficacy than fenretinide, subsequent patch formulations would deliver the bioactive metabolite, 4-oxo-4-HPR.

Lesions of oral epithelial dysplasia are molecularly and biochemically diverse

[25]. It is therefore reasonable to predict that combinations as opposed to a single category agent may be necessary for chemoprevention of some dysplastic lesions.

Indeed, our previous BRB gel chemoprevention trial, in which a subset of patients’ lesions did not respond to topical gel application, emphasizes this point. In addition, complimentary local delivery strategies (e.g., mucoadhesive patch and rinse) would provide both site-specific and field coverage components to aid in the prevention of initial and second primary oral dysplastic lesions. A locally deliverable fenretinide formulation re-introduces a potent keratinocyte differentiation-inducing agent to the oral

209 chemoprevention battery and initiates the prospect for anthocyanin-retinoid based combination therapy.

B.6 References

1. Holpuch A.S., Desai K.G., Schwendeman S.P., and Mallery S.R. (2011) Optimizing therapeutic efficacy of chemopreventive agents: A critical review of delivery strategies in oral cancer chemoprevention clinical trials. J Carcinog, 10:23.

2. Mallery S.R., Zwick J.C., Pei P., Tong M., Larsen P.E., Shumway B.S., Lu B., Fields H.W., Mumper R.J., and Stoner G.D. (2008) Topical application of bioadhesive black raspberry gel modulates gene expression and reduces cyclooxygenase 2 protein in human premalignant oral lesions. Cancer Res, 68, 4945-4957.

3. Shumway B.S., Kresty L.A., Larsen P.E., Zwick J.C., Lu B., Fields H.W., Mumper R.J., Stoner G.D., and Mallery S.R. (2008) Effects of topically applied bioadhesive berry gel on loss of heterozygosity indices in premalignant oral lesions. Clin Cancer Res, 14, 2421-2430.

4. Ugalde C.M., Liu Z., Ren C., Chan K.K., Rodrigo K.A., Ling Y., Larsen P.E., Chacon G.E., Stoner G.D., Mumper R.J., Fields H.W., and Mallery S.R. (2009) Distribution of anthocyanins delivered from a bioadhesive black raspberry gel following topical intraoral application in normal healthy volunteers. Pharm Res, 26, 977-986.

5. Clifford J.L., Menter D.G., Wang M., Lotan R., and Lippman S.M. (1999) Retinoid receptor-dependent and -independent effects of N-(4- hydroxyphenyl)retinamide in F9 embryonal carcinoma cells. Cancer Res, 59, 14- 18.

6. Formelli F., Barua A.B., and Olson J.A. (1996) Bioactivities of N-(4- hydroxyphenyl)retinamide and retinoyl beta-glucuronide. FASEB J, 10, 1014- 1024.

210

7. Chiesa F., Tradati N., Grigolato R., Boracchi P., Biganzoli E., Crose N., Cavadini E., Formelli F., Costa L., Giardini R., Zurrida S., Costa A., de Palo G., and Veronesi U. (2005) Randomized trial of fenretinide (4-HPR) to prevent recurrences, new localizations and carcinomas in patients operated on for oral leukoplakia: long-term results. Int J Cancer, 115, 625-629.

8. Lippman S.M., Lee J.J., Martin J.W., El-Naggar A.K., Xu X., Shin D.M., Thomas M., Mao L., Fritsche H.A., Zhou X., Papadimitrakopoulou V.A., Khuri F.R., Tran H., Clayman G.L., Hittelman W.N., Hong W.K., and Lotan R. (2006) Fenretinide activity in retinoid-resistant oral leukoplakia. Clin Cancer Res, 12, 3109-3114.

9. William W.N. Jr, Lee J.J., Lippman S.M., Martin J.W., Chakravarti N., Tran H.T., Sabichi A.L., Kim E.S., Feng L., Lotan R., and Papadimitrakopoulou V.A. (2009) High-dose fenretinide in oral leukoplakia. Cancer Prev Res (Phila) 2, 22-26.

10. Tradati N., Chiesa F., Rossi N., Grigolato R., Formelli F., Costa A., and de Palo G. (1994) Successful topical treatment of oral lichen planus and leukoplakias with fenretinide (4-HPR). Cancer Lett, 76, 109-111.

11. Mallery S.R., Budendorf D.E., Larsen M.P., Pei P., Tong M., Holpuch A.S., Larsen P.E., Stoner G.D., Fields H.W., Chan K.K., Ling Y., and Liu Z. (2011) Effects of human oral mucosal tissue, saliva and oral microflora on intraoral metabolism and bioactivation of black raspberry anthocyanins. Cancer Prev Res (Phila), 4, 1209-1221.

12. Desai K.G., Mallery S.R., Holpuch A.S., and Schwendeman S.P. (2011) Development and in vitro-in vivo evaluation of fenretinide-loaded oral mucoadhesive patches for site-specific chemoprevention of oral cancer. Pharm Res, 28, 2599-2609.

13. Vaishampayan U., Heilbrun L.K., Parchment R.E., Jain V., Zwiebel J., Boinpally R.R., LoRusso P., and Hussain M. (2005) Phase II trial of fenretinide in advanced renal carcinoma. Invest New Drugs, 23, 179-185.

14. Ren F., Wu H., Lei Y., Zhang H., Liu R., Zhao Y., Chen X., Zheng D., Tong A., Chen L., Wei Y., and Huang C. (2010) Quantitative proteomics identification of phosphoglycerate mutase 1 as a novel therapeutic target in hepatocellular carcinoma. Mol Cancer, 9:81.

211

15. Halbert C.L., Demers G.W., and Galloway D.A. (1992) The E6 and E7 genes of human papillomavirus type 6 have weak immortalizing activity in human epithelial cells. J Virol, 66, 2125-2134.

16. Illingworth N.A., Boddy A.V., Daly A.K., and Veal G.J. (2010) Characterization of the metabolism of fenretinide by human liver microsomes, cytochrome P450 enzymes and UDP-glucuronosyltransferases. Br J Pharmacol, 162, 989-999.

17. Villani M.G., Appierto V., Cavadini E., Valsecchi M., Sonnino S., Curley R.W., and Formelli F. (2004) Identification of the fenretinide metabolite 4-oxo- fenretinide present in human plasma and formed in human ovarian carcinoma cells through induction of cytochrome P450 26A1. Clin Cancer Res, 10, 6265- 6275.

18. Abou-Issa H., Curley R.W., Panigot M.J., Tanagho S.N., Sidhu B.S., and Alshafie G.A. (1997) Chemotherapeutic evaluation of N-(4-hydroxyphenyl)retainamide-O- glucuronide in the rat mammary tumor model. Anticancer Res, 17, 3335-3339.

19. Rinaldi A.L., Morse M.A., Fields H.W., Rothas D.A., Pei P., Rodrigo K.A., Renner R.J., and Mallery S.R. (2002) Curcumin activates the aryl hydrocarbon receptor yet significantly inhibits (-)-benzo(a)pyrene-7R-trans-7,8-dihydrodiol bioactivation in oral squamous cell carcinoma and oral mucosa. Cancer Res, 62, 5451-5456.

20. Wu X., Desai K.G., Mallery S.R., Holpuch A.S., Phelps M.P., Schwendeman S.P. (2012) Mucoadhesive fenretinide patches for site-specific chemoprevention of oral cancer: enhancement of oral mucosal permeation of fenretinide by co- incorporation of propylene glycol and menthol. Mol Pharm, In press.

21. Lippens S., Denecker G., Ovaere P., Vandenabeele P., and Declercq W. (2005) Death penalty for keratinocytes: apoptosis versus cornification. Cell Death and Differ, 12, 1497-1508.

22. Candi E., Schmidt R., and Melino G. (2005) The cornified envelope: a model of cell death in the skin. Nat Rev Mol Cell Biol, 6, 328-340.

23. Lesort M., Attanavanich K., Zhang J., and Johnson G.V.W. (1998) Distinct nuclear localization and activity of tissue transglutaminase. J Biol Chem, 273, 11991-11994.

212

24. Tatsukawa H., Sano T., Fukaya Y., Ishibashi N., Watanabe M., Okuno M., Moriwaki H., and Kojima S. (2011) Dual induction of caspase 3- and transglutaminase-dependent apoptosis by acyclic retinoid in hepatocellular carcinoma cells. Mol Cancer, 10:4.

25. Smith J., Rattay T., McConkey C., Helliwll T., and Mehanna H. (2009) Biomarkers in dysplasia of the oral cavity: a systematic review. Oral Oncol, 45, 647-653.

213