Id3 Induces an Elk-1- and Caspase-8-Dependent Apoptotic Pathway In

ID3 INDUCES AN ELK-1- AND CASPASE-8-DEPENDENT APOPTOTIC PATHWAY IN

SQUAMOUS CARCINOMA CELLS

A Dissertation submitted to the Faculty of the Graduate School of Arts and Sciences of Georgetown University in partial fulfillment of the requirements for the degree of Doctor of Philosophy in Biochemistry and Molecular & Cellular Biology

You-shin Chen, M.S.

Washington, DC

November 18, 2014

ID3 INDUCES AN ELK-1- AND CASPASE-8-DEPENDENT APOPTOTIC PATHWAY IN

SQUAMOUS CARCINOMA CELLS

You-shin Chen, M.S.

Thesis Advisor: Dean S. Rosenthal, Ph.D.

ABSTRACT

Inhibitors of differentiation/DNA binding (Id) proteins are helix-loop-helix (HLH) transcription factors. The Id protein family (Id1-Id4) mediates tissue homeostasis by regulating cellular processes including differentiation, proliferation, and apoptosis. Previously, we found that Id3 induced apoptosis in immortalized human keratinocytes (Simbulan-Rosenthal et al.,

2006), consistent with its role as a tumor suppressor (Richter et al., 2012; Schmitz et al., 2012).

To investigate the role of Id3 in malignant SCC cells (A431), a tetracycline-regulated inducible system was used to induce Id3 in cell culture and mouse xenograft models. We found that upon

Id3 induction, there was a decrease in cell number under low serum conditions, as well as in soft agar. Microarray, RT-PCR, immunoblot, siRNA, and inhibitor studies revealed that Id3 induced expression of Elk-1, an ETS-domain transcription factor, inducing procaspase-8 expression and activation. Id3 deletion mutants revealed that 80 C-terminal amino acids, including the HLH, are important for Id3-induced apoptosis. In a mouse xenograft model, Id3 induction decreased tumor size by 30%. Using immunofluorescence analysis, we determined that the tumor size decrease was also mediated through apoptosis. Further, we show that Id3 synergizes with 5-FU and cisplatin therapies for non-melanoma skin cancer cells. Our studies have shown that Id3 induces apoptosis in SCC cells via an Elk-1- and caspase-8-dependent pathway, and this information can provide some strategic insights for new SCC treatments.

iii

Acknowledgements

First, I would like to sincerely thank my thesis advisor Dr. Dean Rosenthal for giving me the opportunity to work in his lab. Without his guidance, I couldn’t have completed the research and thesis writing. Besides, I feel fortunate to be given the opportunities to learn how to lead a research team comprised of people from different scientific and cultural background. The past years in the Rosenthal lab have been very productive, and I feel blessed to have this kind of training experience. I want to thank my committee members: Drs. Albanese, Chirikjian, Kasid, and Vasudevan and the Department Chair Dr. Crooke for their advice and feedback throughout the years. Special thanks to Dr. Cynthia Rosenthal, from whose life experience I find inspiring and for giving me the opportunities to develop teaching skills.

I would also like to thank my colleagues along the way: Valerie, Kyle, Lucia, Virginia,

Anirudh, and Maryam, for their constant encouragement and help in the lab. The good memories with them will always be remembered.

Last but not least, I would like to thank my family and friends. My father Chin-tu, mother Tung-chin, sister You-jen, and brother Yu-fu form strongest support team to help me in any possible ways in the past years to complete my degree. My boyfriend Derrick has been my best friend, top cheerleader, and supporter. Derrick has offered me his warmest hands during my down times, without whom, the completion of my research would not be possible. I also want to thank my best friends, Lynn, Peiyu, and Tina who never abandon me even when I miss their important life events while pursuing this degree.

Many thanks,

You-shin

Table of Contents

Table of Abbreviations and Definitions ...... viii List of Figures ...... xv List of Tables ...... xviii Chapter 1. Introduction ...... 1 1.1. Hypothesis, Rationale, and Research Aims ...... 1 1.2. Introduction to Squamous Cell Carcinoma (SCC) ...... 2 1.2.1. Skin Cancer Statistics ...... 2 1.2.2. Risk Factors for Squamous Cell Carcinoma ...... 3 1.2.3. SCC Progression ...... 4 1.2.4. Current Treatment Options for SCC ...... 5 1.3. Introduction to Id Proteins ...... 9 1.3.1. Id Genetic and Protein Information ...... 9 1.3.2. Ids as HLH Transcription Factors ...... 11 1.3.3. Ids as Regulators of Differentiation and Proliferation ...... 17 1.3.4. Ids as Regulators of Apoptosis ...... 21 1.3.5. Ids as Potential Tumor Suppressors ...... 22 1.3.6. Ids and SCC ...... 25 1.3.7. Ids and TCF Transcription Factors ...... 27 1.4. Introduction to Apoptosis ...... 34 1.4.1. Caspases ...... 34 1.4.2. Bcl-2 Family Proteins ...... 37 1.4.3. Apoptotic Signaling Pathways ...... 39 1.4.4. Ids and Apoptosis ...... 41 Chapter 2. Materials and Methods ...... 43 2.1. Cell Culture ...... 43 2.2. Tet-inducible System ...... 44 2.3. Plasmids ...... 47 2.4. Inducible Cell Lines ...... 48 2.5. Fluorescently-labeled Cell Lines ...... 49

2.6. Growth Curve ...... 49 2.7. Cell Cycle Analysis ...... 50 2.8. Cell Invasion Assay ...... 51 2.9. Microarray Analysis ...... 51 2.10. siRNA Transfection ...... 51 2.11. Tumor Xenografts ...... 52 2.12. Immunoblot Analysis ...... 52 2.13. Immunostaining ...... 53 2.14. RT-PCR/qRT-PCR ...... 53 2.15. Soft Agar Growth Assay ...... 54 2.16. 5-FU and Cisplatin Treatment ...... 54 2.17. Statistical Analysis ...... 55 Chapter 3. Results ...... 56 3.1. Establishment of A431/Id3 and A431/Vc Cell Lines ...... 56 3.1.1. Endogenous Levels of Id3 are Down-regulated in A431 Cells ...... 56 3.1.2. Verification of Inducible A431/Id3 Clones ...... 57 3.2. Study of the Effects of Id3 in vitro ...... 59 3.2.1. Id3 Does Not Affect Cell Growth in Normal Serum, in Matrigel, or Invasion ...... 59 3.2.2. Id3 Reduces Viable Cell Number and Induces sub-G1 Population in Low Serum ...... 64 3.2.3. Id3-induced Cell Death is not through a Paracrine Signaling Cascade ...... 66 3.2.4. Id3 Induces Caspase-3 and Caspase-9 Proteolytic Cleavages ...... 68 3.2.5. Microarray Analysis Revealed Elk-1 as a Downstream Target of Id3...... 68 3.2.6. Id3 Induces Apoptosis through Elk-1-Caspase-8-Dependent Pathway ...... 72 3.2.7. Caspase Inhibitors Abolished Id3-induced Decreases in Cell Numbers in Low Serum, and Colony Formation in Soft Agar ...... 74 3.2.8. Down-regulation of Endogenous Id3 Repressed Caspase-8 Activation in Other SCC Lines ...... 77 3.2.9. Id3-induced Apoptosis is C-terminus and HLH-dependent ...... 78

3.2.10. Id3 Sensitizes SCC Cells to Chemotherapeutic Drugs ...... 81 3.3. Effects of Id3 in vivo ...... 82 3.3.1. Id3 induction decreased xenograft sizes in vivo ...... 82 Chapter 4. Discussion ...... 86 Appendix ...... 98 A431 Mycoplasma Detection Test and IMPACT Study ...... 98 A431 Cell Line Authentication Test ...... 101 References ...... 110

vii

Table of Abbreviations and Definitions

Abbreviation Full Name and/or Definition

17-AAG 17-N-allylamino-17-demethoxygeldanamycin; also known as Tanespimycin

5-FU 5-Flurouracil

A431 a vulvar SCC cell line from a 85-year-old female

ACCM an adenoid cystic carcinoma cell line

AHR Aromatic Hydrocarbon Receptor

AK Actinic Keratosis

ALA Porphyrin 5-aminolevulinic Acid

Apaf-1 Apoptosis Protease Activating Factor-1

ARNT AHR Nuclear Translocator

BCC Basal Cell Carcinoma

Bcl-2 B-cell Lymphoma-2

Bcl-6 B-cell Lymphoma-6

Bcl-2 Homology Domains; they are conserved domains among Bcl-2 family BH domains proteins

BMP Bone Morphogenetic Protein

bHLH Basic Helix-Loop-Helix

bHLH-LZ bHLH-Leucine Zipper

bHLH-PAS bHLH-Period Circadian, Ah Receptor Nuclear Translocator, Single-Minded

BL Burkitt’s Lymphoma

CARD Caspase Activation or Recruitment Domain

Caspase Cysteine-dependent Aspartate-specific Protease

Castleman's a lymphoproliferative disorder disease

viii

Abbreviation Full Name and/or Definition

CBP CREB-binding Protein

CDDP Cis-diamminedichloroplatinum II; also known as cisplatin

CDK Cyclin Dependent Kinase

CDKI Cyclin Dependent Kinase Inhibitor

ChIP Chromatin Immunoprecipitation

CLIC4 Chloride Intracellular Channel 4

CREB cAMP Response Element-binding Protein

Da Daughterless; E protein homolog in Drosophila

DD Death Domain

DED Death Effector Domain

DHL Double Hit Lymphoma

DISC Death-inducing Signaling Complex

DLBCL Diffuse Large B Cell Lymphoma

DMEM Dulbecco's Modified Eagle Medium

DMSO Dimethyl Sulfoxide

DOX Doxycycline

DR5 Death Receptor-5

DU145 a prostate cancer cell line

Immunoglobulin Enhancer-binding Factors E12/E47; a gene encoded for E12 E2A and E47 proteins

ED&C Electrodessication And Curettage

EGFR Epidermal Growth Factor Receptor

Egr-1 Early Growth Response Gene-1

Elk-1 Ets-like Transcription factor-1

Abbreviation Full Name and/or Definition

Elk-3 Ets-Like Transcription Factor-3

Elk-4 Ets-Like Transcription Factor-4

Emc Extramacrochaetae

EMT Epithelial to Mesenchymal Transition

ERK Extracellular-signal Related Kinase

ETS E-Twenty-Six

E-Twenty-Six Transformation Specific-1; a helix-turn-helix transcription Ets-1 factor

E-Twenty-Six Transformation Specific-2; a helix-turn-helix transcription Ets-2 factor

FACS Fluorescence-activated Cell Sorting

FADD Fas-associated Death Domain Adaptor

FBS Fetal Bovine Serum

GAPDH Glyceraldehyde 3-phosphate Dehydrogenase

GFP Green Fluorescent Protein

GOF Gain-Of-Function

HaCaT Spontaneously immortalized human keratinocytes from a 62-year-old male

HAT Histone Acetyltransferase

HD Homeodomain

HDAC Histone Deacetylase

HEK 293 cells Human Embryonic Kidney 293 cells

Her2 Human Epidermal Growth Factor Receptor 2; also known as Erbb2

HES Hairy And Enhancer Of Split

HESR Hairy/Enhancer Of Split Related

Abbreviation Full Name and/or Definition

HFK Human Foreskin Keratinocyte

Immortalized Human Foreskin Keratinocytes, immortalized by HPV E6 and HFK E6/E7 E7 viral proteins

HLH Helix-Loop-Helix

HOC313 an upper aerodigestive track SCC cell line

HPV Human Papillomavirus

HSP Heat Shock Protein hTERT Human Telomerase Reverse Transcriptase

Id Inhibitor of Differentiation/DNA Binding

IEGs Immediate Early Genes

IFN Interferon

IG Immunoglobulin gene

IKK IκB Kinase

IκB Inhibitor of NFκB

IL-3 Interleukin-3

Infectious Microbe PCR Amplification Test; also known as DNA IMPACT fingerprinting test

JNK c-Jun N-terminal Kinase

KS Kaposi’s Sarcoma; a cancer of lymphatic endothelial cells

an androgen-dependent prostate cancer cell line, derived from a 50-year-old LNCaP male from lymph node metastasis

LT23 a lung adenocarcinoma cell line; also known as NIH-LT23

MAPK Mitogen-activated Protein Kinase

MAX Myc-associated Factor X

Abbreviation Full Name and/or Definition

Michigan Cancer Foundation-7; a ER+, PR+ and HER2- breast cancer cell MCF-7 line derived from a 69-year-old female from pleural effusion metastasis

a triple negative breast cancer cell line derived from a 51-year-old female MDA-MB-231 from pleural effusion metastasis mES Mouse Embryonic Stem (Cells)

MG-63 an osteosarcoma cell line derived from a 14-year-old male

MKN-45 a gastric cancer cell line derived from stomach

MMP-2 Matrix Metelloproteinase-2

MMS Mohs Micrographic Surgery

MXI Max-interacting Protein 1

Myc Myelocytomatosis Viral Oncogene Homolog

MyoD Myogenic Differentiation

NES Nuclear Export Signal

NeuroD Neuronal Differentiation

NFκB Nuclear Factor Kappa-light-chain-enhancer of Activated B cells

NID Net Inhibitory Domain

NLS Nuclear Localization Signal

NMSC Non-Melanoma Skin Cancer

Oct3/4 Octamer-binding Transcription Factor ¾

PAS Period Circadian, Ah Receptor Nuclear Translocator, Single-Minded

PAX5 Paired Box-5

an androgen-independent prostate cancer cell line derived from a 62-year-old PC3 male from bone metastasis

PD Paired Domain

Pen/Strep Penicillin and Streptomycin

xii

Abbreviation Full Name and/or Definition

Phoenix a retroviral packaging cell line; also known as ΦNX

PIASxα Protein Inhibitors of Activated STAT; a SUMO E3-ligase

PIRL Preclinical Imaging Research Laboratory at Georgetown University

PP2B Protein Phosphatase 2B; also known as Calcineurin pRB Retinoblastoma Protein

PTHrP Parathyroid Hormone-Related Protein

PTP Permeability Transition Pore Complex qRT-PCR Quantitative Reverse Transcriptase Polymerase Chain Reaction

ROS Reactive Oxygen Species

RT-PCR Reverse Transcriptase Polymerase Chain Reaction

SAP-1 SRF accessory protein 1; also known as Elk-4

SAP-2 SRF accessory protein 2; also known as Elk-3

SCC Squamous Cell Carcinoma

SCC25 a tongue SCC cell line from a 70-year-old male

SCC4 a tongue SCC cell line from a 55-year-old male

SCC9 a tongue SCC cell line from a 25-year-old male

SD Standard Deviation siRNA Small Interfering RNA

SNP Single Nucleotide Polymorphism

Sox2 Sex Determining Region Y (SRY)-box 2

Sp-1 Specificity Protein-1

SPARC Secreted Protein, Acidic and Rich in Cysteine

SRE Serum Response Element

xiii

Abbreviation Full Name and/or Definition

SRF Serum Response Factor

STR Short Tandem Repeat

SUMO Small Ubiquitin-like Modifier

TAD Transcriptional Activation Domain

TBK-BP1 Tank Binding Kinase 1 Binding Protein 1

TCF Ternary Complex Factor; comprised of Elk-1, Elk-3 and Elk-4

TCF 3 Transcription factor 3; also known as E2A

Tet Tetracycline

TGF-β Transforming Growth Factor β

TNF Tumor Necrotic Factor

TNFR Tumor Necrotic Factor Receptor

TR Tet Repressor

TSA Trichostatin A

TUNEL Terminal Deoxynucleotidyl Transferase-dUTP Nick end Labeling

Ub Ubiquitin

Ubc9 Ubiquitin-Conjugating Enzyme 9; a SUMO E2-conjugating enzyme

University of Michigan Squamous Cell Carcinoma 1; a SCC line derived UM-SCC1 from mouth

UVB Ultraviolet B

VEGF Vascular Endothelial Growth Factor

XP Xeroderma Pigmentosum

Z-IETD-FMK Carboxybenzyl Ile-Glu-Thr-Asp Fluoromethylketone; a caspase-8 inhibitor

Z-VAD-FMK Carboxybenzyl Val-Ala-Asp Fluoromethylketone; a pan caspase inhibitor

xiv

List of Figures

Figure 1. Clinical and histological presentation of SCC progression...... 5

Figure 2. The alignment of the amino acid sequences of human Id1-4 show four conserved

regions...... 10

Figure 3. Phenotypes of Id knockout mice...... 11

Figure 4. Binding of E47 homodimers to DNA...... 12

Figure 5. Structural illustration of different HLH proteins...... 13

Figure 6. Ids are dominant-negative regulators of bHLH transcription factors...... 15

Figure 7. Model for Id2-induced S-phase entry...... 19

Figure 8. Molecular pathways by which Ids modulate proliferation, invasion, and apoptosis. ... 20

Figure 9. Frequent Id3 mutations in lymphoma patients...... 23

Figure 10. A proposed pathogenic mechanism for Burkitt’s lymphoma...... 24

Figure 11. Epidermis structure...... 25

Figure 12. Conserved domains in TCF transcription factors...... 28

Figure 13. Post-translational modifications of Elk-1 protein...... 30

Figure 14. Negative feedback loop established between immediate response genes and Ids. .... 31

Figure 15. Pax/TCF transcriptional activator complex is disrupted by Id protein...... 32

Figure 16. Caspase domains and proteolytic cleavage sites...... 36

Figure 17. Caspase substrates...... 37

Figure 18. Apoptotic signaling pathways...... 39

Figure 19. Schematic illustration for procedure to establish inducible cell lines...... 45

Figure 20. pcDNA4/TO vector map...... 45

Figure 21. pcDNA6/TR vector map...... 46

Figure 22. Addition of tetracycline induces transcription of Id3...... 46

Figure 23. Primer binding sites for cloning Id3 mutants...... 48

Figure 24. Endogenous Id3 and Id1 protein levels in different keratinocytes examined via

immunoblot...... 57

Figure 25. Id3 protein inducibility in A431/Id3 cell lines...... 58

Figure 26. Verification of Id3 mRNA induction in A431/Id3 clone 5 cells via RT-PCR and qRT-

PCR...... 58

Figure 27. A431/Id3 and A431/Vc Tet-induction time course...... 59

Figure 28. Id3 does not affect cell growth and death under 10% FBS condition...... 60

Figure 29. Id3 does not affect cell cycle distribution under 10% FBS condition...... 61

Figure 30. Id3 does not affect cell growth on Matrigel under 10% FBS condition...... 62

Figure 31. Id3 does not affect invasion...... 63

Figure 32. Id3 does not significantly alter invasion and stem cell markers...... 63

Figure 33. A431/Id3 growth curves in low serum condition...... 65

Figure 34. A431/Vc growth curves in low serum condition...... 65

Figure 35. Cell cycle analysis for A431/Id3 cells in low serum...... 66

Figure 36. A431/Id3 and A431/Vc co-culture growth curves in low serum...... 67

Figure 37. Caspase-3 and -9 proteolytic cleavage under Id3 induction...... 68

Figure 38. Heatmap showing genes that are significantly up- and down-regulated by Id3...... 69

Figure 39. Verification of genes that are significantly altered by Id3 via RT-PCR...... 71

Figure 40. Verification of genes that are significantly altered by Id3 via immunoblot...... 71

Figure 41. Id3 induces Elk-1 and Caspase-8 expression...... 73

Figure 42. Induction of Fas expression by Id3 in vitro...... 74

xvi

Figure 43. A431/Id3 cell growth curves in the presence of DMSO...... 75

Figure 44. A431/Id3 cell growth curves in the presence of caspase inhibitors...... 75

Figure 45. Id3-reduced colony formation in soft agar is caspase-dependent...... 76

Figure 46. Knocking down endogenous Id3 in other SCC reduced proteolytic cleavage of Casp-

8...... 77

Figure 47. Schematic representation of Id3 deletions of the C-terminus (ΔC38), N-terminus

(ΔN39), C+HLH (ΔC80) domain, or N+HLH (ΔN81) domain...... 78

Figure 48. Verification of Id3 mutant induction by Tet (1 µg/ml for 24 hours) in A431 cells. .. 78

Figure 49. Growth curves for A431/Id3 mutant cell lines in low serum...... 79

Figure 50. Cell cycle analyses for A431/Id3 mutant cell lines in low serum...... 80

Figure 51. Id3 sensitizes A431 cells to CDDP and 5-FU treatments...... 81

Figure 52. Mouse xenografts formed from GFP-expressing A431/Id3 cells...... 82

Figure 53. Id3 reduces xenograft sizes in vivo...... 83

Figure 54. A431/Vc xenografts do not show difference in size under ±Dox conditions...... 83

Figure 55. Id3 induces Elk-1 expression in vivo...... 84

Figure 56. Active Caspase-3 is upregulated in A431/Id3 xenograft sections...... 85

Figure 57. Id3 induces Fas expression in vivo...... 85

Figure 58. Overlap of Id3 and MyoD protein structures...... 90

Figure 59. Oncomine database showing Id3 expression levels in normal vs. cancer samples. ... 95

xvii

List of Tables

Table 1. Incidence of skin cancer cases and deaths in the US in 2013...... 3

Table 2. The 7 families in HLH transcription factor superfamily...... 14

Table 3. Caspases are classified into initiator and effector caspases...... 35

Table 4. Bcl-2 family protein classification...... 38

Table 5. Combination index for Id3 and CDDP or 5-FU dual treatments ...... 92

xviii

Chapter 1. Introduction

1.1. Hypothesis, Rationale, and Research Aims

Our previous studies revealed that ultraviolet B (UVB) exposure induces endogenous levels of Inhibitor of DNA binding 3 (Id3) protein in immortalized human keratinocytes but not in primary keratinocytes. Ectopic expression of Id3 induces apoptosis in these cells via a bax-

Caspase-9-Caspase-3-mediated pathway (Simbulan-Rosenthal et al., 2006). This represents a transient stage in the progression of skin cancer where cells are very sensitive to UVB, at least in part, due to reactive oxygen species (ROS)-induced apoptosis (Trabosh et al., 2009a). To better understand the progression of skin cancer, our current study aims to investigate the role of Id3 in

Squamous Cell Carcinoma (SCC). We hypothesize that Id3 induces apoptosis in malignant squamous carcinoma cells and causes tumor regression in xenografts.

The rationale for our hypothesis is based on three findings. First, Id3 has been identified as tumor suppressor gene in Burkitt’s lymphoma (Campo, 2012; Love et al., 2012; Richter et al.,

2012; Schmitz et al., 2012). Secondly, Id3 has been shown to induce apoptosis in other cancer types, including osteosarcoma (Koyama et al., 2004). Third, reduced endogenous Id3 levels are observed in SCCs (Figure 24) compared to immortalized keratinocytes, indicating that loss of

Id3 is preferred in cancer cells.

To investigate the role of Id3 in SCC the research aims are: 1) to ectopically express Id3 using a tetracycline (Tet)-inducible system, or to downregulate Id3 using siRNA in A431 squamous carcinoma cells, 2) investigate the effects of Id3 induction or suppression using cell culture and xenograft models, and 3) to investigate the mechanism by which Id3 modulates tumor regression in A431 cells.

1.2. Introduction to Squamous Cell Carcinoma (SCC)

1.2.1. Skin Cancer Statistics

Skin cancer is the most common type of cancer; the incidence of which is greater than all other cancers combined in the United States (Martinez and Otley, 2001). Each year, there are more than 2 million new skin cancer cases diagnosed (Rogers et al., 2010). As a result, it is estimated that one in five Americans will develop skin cancer in their lifetimes (Neville et al.,

2007). Further, the direct medical cost associated with skin cancer treatment is more than 1.7 billion dollars in the United States annually (Bickers et al., 2006). Skin cancer is therefore a major public health problem in the US.

Different types of skin cancers are identified based on the cell types from which they originate. The three most common types of skin cancer are: Basal Cell Carcinoma (BCC; 72%),

Squamous Cell Carcinoma (SCC; 25%), and Melanoma (3%). BCC and SCC are epithelial Non-

Melanoma Skin Cancers (NMSC), and they account for 97% of all skin cancer cases. Non- epithelial skin cancers are more rare, and include Kaposi’s sarcoma (KS; cancer of lymphatic endothelial cells), Sebaceous Carcinoma, and Merkel Cell Carcinoma. Even though these tumors are rare, they are usually aggressive (Kanitakis, 2009). The skin cancer incidence and mortality rates are shown in Table 1.

Table 1. Incidence of skin cancer cases and deaths in the US in 2013.

New Cases Diagnosed Deaths reported Skin Cancer Type (% total) (% total)

BCC4 2,000,000 (72%)1 670 (4%)2,3

SCC4 700,000 (25%)1 2,500 (16%)3

Melanoma 76,690 (3%)2 9,480 (60%)2

Non-epithelial skin cancer 6,080 (0.2%)2 3,170 (20%)2

Total 2,782,770 15,820 1. Data source: (American Academy of Dermatology, 2011) 2. Data source: (American Cancer Society, 2013) 3. Data source: (Skin Cancer Foundation, 2014) 4. Note that BCC and SCC cases are not always reported to cancer registries.

1.2.2. Risk Factors for Squamous Cell Carcinoma

UV light is the major risk factor for most skin cancers. As a result, SCC occurs mostly on areas exposed to sun, including head, lips, ears, neck, and hands. In addition, tumor recurrence and metastasis is also frequently seen in these areas (Martinez and Otley, 2001).

Immune suppression is another risk factor for SCC. People who are receiving immunosuppressant drugs, including organ transplant patients, have an increased chance of skin cancer in general, and metastatic skin cancers specifically. Another risk factor for SCC is exposure to carcinogens. Tobacco, alcohol, chemical toxins, ionizing radiation, chronic skin inflammation, and human papillomavirus (HPV) infection all increase SCC risk (Schmitz et al.,

2014). Smoking and drinking can both cause DNA damage and increase chance of head and neck SCC. Chemical toxins like arsenic, ionizing radiation like X-rays, as well as skin inflammation caused by psoriasis increases skin cancer risk. Moreover, many vulvar SCC cases are due to HPV infection (Hietanen et al., 1995).

In addition, skin color, age, and sex are also related to skin cancer risk. Individuals with fair skin color are at greater risk for SCC. SCC risk also increases with age. Men are more likely to get SCC, most likely because of their occupations.

People with personal and/or family histories of SCC, and some genetic disorders render individuals more susceptible. For example, people with Xeroderma Pigmentosum are very sensitive to sunlight. Their DNA repair mechanism is compromised, and therefore they are 1000 times more likely to get SCC (Kraemer et al., 1994; Kraemer et al., 1984).

1.2.3. SCC Progression

Most SCC cases begin as actinic keratosis (AK). These precancerous lesions are mostly found in areas frequently exposed to the sun. AK dysplasia is characterized by disorganized cell growth and perturbed differentiation of skin layers, containing keratinocytes with enlarged and irregular nuclei. Further, parakeratosis (retention of nuclei in stratum corneum) alternates with orthokeratosis (normal loss of nuclei in stratum corneum) (Ratushny et al., 2012).

AK progresses to SCC when atypical keratinocytes invade to the dermis layer.

Depending on the number of AKs as well as the time they persist, the risk of AK progression to

SCC varies. About 26% of AK regresses spontaneously per year. Patients with less than 5 AKs have less than 1% chance to develop SCC. However, patients with ≥20 AK lesions have a 20% chance of developing SCC (Ratushny et al., 2012).

SCC progresses to metastatic SCC when malignant SCC cells invade lymph nodes or other organs. Overall, around 2% to 6% of SCCs become metastatic (Martinez and Otley, 2001).

However, some areas of the body have a higher risk for tumor recurrence; for example the tumor metastatic rate for SCC of the face, ears, and scalp is 30%, resulting in a median survival time of

6-9 months in patients (Martinez and Otley, 2001; O'Bryan and Ratner, 2011). For skin cancer progression from normal to metastatic SCC, please refer to Figure 1.

Figure 1. Clinical (top row) and histological (bottom row) presentation of SCC progression. Normal skin (A) develops into AK (B) typically due to chronic exposure to sun. Atypical keratinocytes in AK lesions display enlarged nuclei and disorganized growth and differentiation. Parakeratosis is found in the stratum corneum layer. AK progresses cutaneous “cSCC” (C) when atypical keratinocytes spread to the dermis. Cutaneous SCC progresses to metastatic SCC (D) when malignant cells spread to lymph nodes or other organs. Figure from (Ratushny et al., 2012).

1.2.4. Current Treatment Options for SCC

Currently, there are many treatment options for most SCC patients with cure rates >90%.

However, the median survival time is only 6-9 months after treatment of recurrent or metastatic

SCC (Argiris et al., 2004; Colevas, 2006), and more therapeutic options are needed as tumor cells gain resistance to chemotherapeutic agents (Sawada et al., 2003). Standard methods for treating SCC in the clinic include surgery, including cryosurgery, electrodessication and curettage (scraping; ED&C), surgical excision, and Mohs micrographic surgery (MMS).

Photodynamic therapy, radiation, retinoid therapy, chemotherapy, immune therapy, and targeted molecular inhibitors are also treatment options.

Cryosurgery uses liquid nitrogen to damage tumor cells through ice crystal formation.

This procedure is painful therefore local anesthesia is used. This technique does not offer histologic analysis of the tumor margin. Side effects from this treatment include vesiculation, exudation, and sloughing of the treated area, as well as hypopigmentation due to damage of melanocytes during the treatment. This is a good option to treat BCC and SCC with defined borders. The 5-year cure rate for NMSC using cryosurgery is between 93 and 96% (Neville et al., 2007).

ED&C technique is used on superficial BCC or SCC. This treatment involves first electrodessication to trigger necrosis and then a sharp curette to scrape away dead cells until friable tumor tissue is removed. Side effect of this treatment is scarring of the treated site. This technique does not offer histologic analysis on tumor margin. Usage of this option on sites like the nose, ear, and chin is reported to have higher recurrence rates (4.5-17.6%) compared to extremities (3.3%) (Neville et al., 2007). The 5-year cure rate using ED&C is between 92% to

96% for NMSC (Neville et al., 2007).

MMS is an effective way of treating non-melanoma skin cancer. It is currently used on high-risk BCC or SCC like recurrent tumors and tumors greater than 2 cm in diameter. In MMS, serial tumor sections are removed and processed to frozen sections for an oncologist to examine microscopically; the procedure is repeated until the area is clear of tumor cells. MMS is performed under local anesthesia by a trained dermatologist. This procedure greatly improves the removal of all tumor tissue and preservation of normal tissue comparing to the vertical excision. While it is a more expensive option, MMS remains a popular treatment as it reduces

6 the 5-year tumor recurrence rate from 8% to 3% compared to standard excision (Neville et al.,

2007).

Photodynamic therapy requires the usage of a photosensitizing agent (e.g., porphyrin 5- aminolevulinic acid, ALA) on the skin. This compound is activated by a light source with wavelengths between 450-750 nm. The accumulation of reactive oxygen species in cells triggers apoptosis. The recurrence rate after treatment is relatively high (up to 52% in some studies) even though the clearance rate is up to 100% for superficial BCC and SCC (Neville et al., 2007).

Ionizing radiation is also used to treat NMSC as a mono- or adjuvant therapy. X-rays or electron-beams are delivered in fractionated doses to patients to ensure the recovery of normal cells and damage of tumor cells. This option is painless and recommended for older debilitated patients. However, the treatment is not recommended for tumors without defined borders, tumors from lower extremities, recurrent tumors from irradiated areas, or patients with

Xeroderma Pigmentosum. Side effects from radiation therapy can include alopecia, radionecrosis in surrounding skin, depigmentation, and atrophy. The 5-year cure rate provided by this treatment is around 90% for small NMSC (Neville et al., 2007).

Retinoids (derivatives of vitamin A) are chemopreventive agents for SCC and BCC that inhibit tumor cell growth through their anti-proliferative, apoptotic, and differentiation- promoting properties. They are prescribed to patients who may suffer from SCC as a result of immunosuppression after organ transplantation. This option is for high-risk patients since it is necessary to monitor them throughout the treatment to prevent tumor re-growth once the drug is discontinued (Neville et al., 2007).

Chemotherapeutic agents like 5-fluorouracil (5-FU) and cisplatin or cis- diamminedichloroplatinum(II) (CDDP) are used to treat mainly superficial BCC or SCC (Neville et al., 2007). 5-FU is a pyrimidine analog, which acts as a suicide inhibitor against thymidylate synthase and in turn interferes with DNA synthesis. Cisplatin, on the other hand, cross-links

DNA and interferes with the cell cycle. Both drugs cause death in actively dividing cells. It is worth mentioning that many cancer cells develop resistance to 5-FU and cisplatin treatments, and new therapeutic strategies are necessary to target these cells.

Interferons (IFNs) act as cytokines through interactions with cell surface receptors to exert their anti-tumor effects. Interferons can inhibit growth of tumor cells, as well as inducing apoptosis through Fas ligand- Fas receptor interactions (Chakrabarty and Geisse, 2004). IFNs are usually given through intralesional injections. The cure rate using this method is relatively low (50-80%) compared to other treatment options, yet it is used when surgery is not an option for patients (Neville et al., 2007).

Imiquimod is an immune-response modulator, which binds cell surface receptors (e.g.,

Toll-like receptor-7 and -8) and promotes secretion of a variety of cytokines such as IFNα and

TNFα, which in turn activates innate and acquired immune systems. Imiquimod also helps proliferation of B-lymphocytes and activates natural killer cells and Langerhans cells (skin macrophages). Imiquimod is primarily used on superficial NMSC as a single agent or in conjunction with MMS or ED&C (Neville et al., 2007).

Currently, molecular inhibitors used to treat SCC target molecules including Epidermal

Growth Factor Receptor (EGFR), Vascular Endothelial Growth Factor (VEGF), VEGF receptor, and other tyrosine kinases (O'Bryan and Ratner, 2011). Overexpression of these proteins has been shown to be associated with more aggressive SCC phenotype. These targeted molecular

8 inhibitors are under investigation for their use as single or combined treatment options with radiation and chemotherapy. EFGR inhibitors include chemotherapeutic drugs cisplatin and 5-

FU (O'Bryan and Ratner, 2011), as well as antibody therapies including cetuximab, zalutumumab, and panitumumab, that latter two of which are still in clinical trials (O'Bryan and

Ratner, 2011; Schmitz et al., 2014).

More treatment options are needed as the current options have limited efficacy, especially with advanced SCC. Investigation of molecular mechanisms by which SCC can be eliminated can provide insights in designing new therapeutic strategies. Some of these new strategies include aptamer inhibitors of Id1 and Id3, and have shown some efficacy in cancer cells, although the pleiotropic pro- and anti-tumorigenic roles of Ids make their effects hard to predict

(Mern et al., 2010a; Mern et al., 2010b).

1.3. Introduction to Id Proteins

1.3.1. Id Genetic and Protein Information

The inhibitors of differentiation/DNA binding (Id) genes encode helix-loop-helix (HLH) transcription factors, which mediate tissue homeostasis by regulating a variety of cellular processes, including differentiation, proliferation, and apoptosis. Ids therefore play important roles in keratinocyte maturation as well as SCC progression.

Id protein homologs can be found in organisms ranging from Drosophila to human, including extramacrochaetae (emc; Drosophila), XId2 and XIdx/XIdI (Xenopus), as well as TId1 and TId2 (trout fish) (Campuzano, 2001; Rescan, 1997), and the four Id members (Id1-Id4) identified in humans and mice (Benezra et al., 1990; Biggs et al., 1992; Christy et al., 1991;

Deed et al., 1993; Ellmeier et al., 1992; Hara et al., 1994; Pagliuca et al., 1995; Riechmann et al.,

1994; Sun et al., 1991).

Id1-Id4 are small proteins ranging from 119 to 159 amino acids in length. They have both overlapping and distinct functions. They are encoded by genes located on different chromosomes: Id1 (20q1), Id2 (2p25), Id3 (1p36), and Id4 (6p22-21; (Deed et al., 1994; Mathew et al., 1995). In terms of their amino acid sequence homology, there are four conserved domains among Ids, in particular the HLH domain. The amino acid sequence alignment for Id1-Id4 is illustrated in Figure 2.

Figure 2. The alignment of the amino acid sequences of human Id1-4 show four conserved regions: box1-box4. Box 2 shows the most conserved helix-loop-helix domain. Box 1 shows the consensus sequence, SPVR, as a target for serine phosphorylation by cyclin-dependent kinases. Box 3 contains the destruction box (D box, RTPLTTLN for Id2), a target site for ubiquitin- mediated proteolysis. The function of Box 4 is currently under investigation. Figure from (Hara et al., 1997).

Studies with Id knockout mice show that homozygous deletion of both Id1 and Id3 genes is embryonically lethal (Lyden et al., 1999). Further, Ids are important for normal developmental processes in B and T lymphocytes and other immune cells (Id1-Id3), mammary tissue (Id2), neural cells (Id4), as well as vascular formation (Id1 and Id3) (Perk et al., 2005). For phenotypes related to Id knockout mice, please refer to Figure 3.

Figure 3. Phenotypes of Id knockout mice. Double knockout of Id1 and Id3 is embryonically lethal. Ids are important in development of cells in the immune, neural, and cardiovascular systems. Figure from (Perk et al., 2005).

1.3.2. Ids as HLH Transcription Factors

The HLH transcription factor superfamily is comprised of more than 240 proteins that play important roles in a variety of cellular processes, including embryogenesis, cell lineage determination, and cell cycle control (Ledent et al., 2002; Massari and Murre, 2000). The HLH domain is highly conserved among the different family members, and is comprised of two amphipathic alpha helices about 17 amino acids long, linked by a short loop of varying size. The

HLH domain is the dimerization domain, allowing for either homo- or heterodimerization between HLH proteins.

All of these HLH proteins, with the exception of the Ids, contain a basic DNA binding region located to the N-terminal side of the HLH domain and are thus termed bHLH proteins.

This positively charged domain allows binding of bHLH transcription factors to the E box

(CANNTG) or N box (CACNAG) consensus sequence in the promoter region of responsive genes (Massari and Murre, 2000). DNA binding is achieved through the interaction between a conserved glutamate residue in the basic DNA binding region of the bHLH proteins and the cytosine and adenine nucleotides in the E box region. A conserved arginine residue adjacent to the glutamate also interacts with the phosphodiester backbone of DNA in the E box region.

Further, the helices of HLH transcription factors sit in the major groove of DNA (Figure 4).

Figure 4. Binding of E47 homodimers to DNA. Crystal structure showing basic region of E47 homodimers interacts directly with DNA. E47 dimers interact through HLH domain. Figure from (Ellenberger et al., 1994)

HLH proteins are important in transcriptional networks, and they can act as both activators and repressors of transcription. They regulate transcription of different cell lineages, including muscle cells, neurons, B and T lymphocytes, and pancreatic cells (Desprez et al., 2003;

Massari and Murre, 2000), and therefore regulate a variety of developmental processes.

Based on their sequence homology, HLH proteins can be classified into 7 families

(Desprez et al., 2003; Massari and Murre, 2000), including (I) E proteins (ubiquitous bHLH), (II) tissue specific bHLH, (III) Myc family (bHLH-leucine zipper, LZ), (IV) Mad/Max/Mxi family

(bHLH-LZ), (V) Id1-Id4 (HLH), (VI) HES family (bHLH-proline), and (VII) AHR/ARNT family (bHLH-PAS). Thus, within the HLH superfamily, three of the families contain a bHLH domain, two families contain a bHLH followed by a leucine zipper (bHLH-LZ), and one family contains a PAS (Period circadian, Ah receptor nuclear translocator, Single-minded) domain

(bHLH-PAS). For structures of different HLH proteins, please refer to Figure 5. For protein members in each of the 7 HLH families, please refer to Table 2.

Figure 5. Structural illustration of different HLH proteins. Helix-loop-helix domain is the dimerization domain with other HLH proteins. All HLH proteins have a basic (b) DNA-binding domain N-terminus to HLH domain. LZ and PAS domains are both binding sites for other members in the same HLH family. Figure from (Norton, 2000).

Table 2. The 7 families in HLH transcription factor superfamily. Representative family members in humans and Drosophila, as well as structural features, are shown.

HLH proteins Family Structural features Human Drosophila

I E2A (E12/E47), E2-2, HEB Da bHLH

II MyoD, NeuroD Achaete, Scute bHLH

III Myc dMyc bHLH-zip

IV Mad/Max/Mxi dMax bHLH-zip

V Id1-Id4 Emc HLH

VI HES1, HESR1 Hairy bHLH-proline

VII AHR/ARNT Spineless bHLH-PAS

Family I proteins include Daughterless (Da, in Drosophila), and the four E proteins derived from three different genes in mammals: E12 and E47 (alternative spliced isoforms of the

E2A gene), E2-2, and HEB. They are ubiquitously expressed in different tissues. On the other hand, expression of Family II HLHs (e.g., Achaete-Scute complex, MyoD, and NeuroD) are restricted temporally and spatially. During differentiation, family I proteins form homodimers within their own family members (e.g., E47 homodimers in B cells (Bhattacharya and Baker,

2011) or heterodimers with Family II members, including MyoD or NeuroD, through the HLH domain. The protein complex then binds to E boxes located in DNA promoter regions of differentiation-related genes and activate their transcription.

The family III proteins, including c-Myc, can form heterodimers with family IV proteins, including Max, and recruit histone acetyltransferases (HATs) to chromatin in the regions of proliferation-associated genes, inducing their transcription. However, when Mad forms

14 complexes with other class IV proteins, such as Max or Mxi, histone deacetylase (HDAC) is recruited and transcription of the target genes is suppressed (Massari and Murre, 2000).

The Id family belongs to family V, which is the only family that does not contain a basic

DNA binding domain. Similar to E proteins, Id proteins are broadly expressed in different tissues (Bhattacharya and Baker, 2011). Id proteins function primarily as dominant-negative transcriptional regulators by dimerizing with family I or II members, and inhibit them from binding to DNA. For an illustration of Ids acting as dominant negative regulators of bHLH proteins, please refer to Figure 6. Ids can act as transcriptional repressors as well as activators, depending upon the role of the proteins they inhibit (Higashikawa et al., 2009).

Figure 6. Ids are dominant-negative regulators of bHLH transcription factors. (a) Id protein binds E protein and sequesters it away from the E box region. (b) During differentiation, heterodimers formed between E proteins and tissue-specific bHLH bind DNA promoters and activate transcription of differentiation-promoting genes. Figure from (Perk et al., 2005).

E proteins can also act as activators of their own transcription, as well as that of Id genes.

For example, ectopic expression of E47 in HEK293T cells induces its own E2A gene as well as

Id1 transcripts (Bhattacharya and Baker, 2011). Since Family V proteins function as dominant negative inhibitors of E proteins, a negative feedback loop is generated. During differentiation, extracellular signals block Id transcription, elevating free E protein levels, which can promote

15 their own transcription in a positive feedback loop (Bhattacharya and Baker, 2011). In addition, extracellular signals induce Family II protein expression to create more differentiation-promoting heterodimers of Families I and II (Bhattacharya and Baker, 2011).

Family VI members include Hairy and Enhancer of Split-1 (HES1) and Hairy/Enhancer of Split Related with YRPW motif protein 1 (HESR1). Family VI proteins have a proline in the basic DNA binding domain and therefore act as a transcriptional repressors, preventing premature differentiation in cells as in the case of neuronal progenitor cells (Kageyama et al.,

2007). Unlike other bHLH proteins, which bind E boxes, these proteins bind N boxes.

Therefore through heterodimerization, family VI proteins can prevent other bHLH from binding

DNA and in turn prevent transcriptional activity of E-box-binding transcription factors

(Kageyama et al., 2007).

Family VII proteins include Aromatic Hydrocarbon Receptor (AHR), and AHR Nuclear

Translocator (ARNT). AHR remains inactive when bound to its chaperone, heat shock protein

(HSP) 90. Upon binding of polycyclic aromatic hydrocarbon ligands such as plant flavonoids,

Hsp90 dissociates from AHR, allowing AHR to translocate to the nucleus, where it dimerizes with ARNT and induces transcription of xenobiotic-metabolizing enzymes to eliminate toxic chemicals (Massari and Murre, 2000). Family VII proteins are also important for developmental processes, including hematopoiesis and neurogenesis (Latchney et al., 2011).

HLH transcription factors are involved in skin cancer progression. Id2 has been shown to suppress p27Kip1 expression through inhibiting E2A activity, leading to fibrosarcoma (Trabosh et al., 2009b). Id1 has been shown to suppress p16Ink4A expression in keratinocytes and melanocytes to delay senescence, which can lead to malignant transformation (Cummings et al.,

2008; Nickoloff et al., 2000). UVB exposure induces Id2 expression and inhibition of

16 differentiation of human primary keratinocytes, which might predispose cells to SCC (Simbulan-

Rosenthal et al., 2005).

1.3.3. Ids as Regulators of Differentiation and Proliferation

Ids play important roles in differentiation and proliferation. As dominant negative regulators of bHLH transcription factors, Ids function to ensure temporal and spatial differentiation in different lineages. Further, Ids typically function to promote proliferation through suppression of E protein function and therefore suppress transcriptions of Cyclin- dependent Kinase Inhibitors (CDKIs). Ids also function to suppress pRb and therefore allow for

S-phase progression. The classic role of Ids in promoting cell proliferation spawns therapeutic strategies to block Id proteins. However, more and more research, including ours, deciphers other cellular processes regulated by Ids.

Usually, Id proteins orchestrate the timing of differentiation with cessation of proliferation. However, their role in inhibiting differentiation can be separate from that in inducing proliferation. Consistent with this idea, Ids are expressed in non-differentiated quiescent cells, including progenitor cells of different cell types such as muscle satellite cells

(Kumar et al., 2009), and prevent premature differentiation. Enforced expression of Id proteins inhibits differentiation of many cell types, including that of muscle cells, myeloid, and mammary epithelial cells (Zebedee and Hara, 2001). Further, Ids are important in determining cell fate in the germ layers. Ectopic expression of Id2 in chick embryo converts ectoderm cells into neural crest cells, rather than epidermal cells (Martinsen and Bronner-Fraser, 1998).

In general, however, Id proteins are usually upregulated in proliferating cells and downregulated in differentiating cells. Indeed, stimulating fibroblast cells with serum or growth factors rapidly induces Id1-Id3 expression. On the other hand, inhibiting Id expression prevents fibroblasts from re-entering cell cycle from G0 stage (Zebedee and Hara, 2001). Ids are immediate early response genes whose expression pattern is biphasic throughout the cell cycle, with the first peak occurring in early G1 and the second peak in the G1/S transition (Zebedee and

Hara, 2001). The first peak of Id induction is believed to ensure inhibition of differentiation, through sequestering E proteins away from binding DNA (Deed et al., 1997). Further, the biphasic induction of Id2 is important for its inhibitory effect on pRb (Retinoblastoma protein) function.

Id2 has been shown to promote S-phase entry through interaction with pRb family of proteins proteins (Rb/p105, RbL1/p107 and Rb2/p130) (Iavarone et al., 1994). The HLH domain and pocket domain of pRb allows for the physical interaction between the two proteins. Id2 binds hypophosphorylated pRb, which then releases E2F transcription activators; E2F in turn expression of genes promoting S-phase entry (Iavarone et al., 1994).

Figure 7 shows the model how Id2 induces cell proliferation through inhibiting pRb function.

Figure 7. Model for Id2-induced S-phase entry. Id2 promotes cell proliferation by releasing E2F from pRb and therefore inducing genes necessary to promote S-phase entry.

Further, Cdk2 phosphorylates Id2-Id4 at serine 5 in late G1 (Deed et al., 1997; Hara et al., 1997; Norton, 2000) (Figure 2). This post-translational modification alters the ability of Ids to disrupt bHLH protein complexes. It has been shown that phosphorylated Id3 has decreased ability to sequester E12 homodimers but increased ability to sequester E12/MyoD away from

DNA promoter region (Deed et al., 1997). This phosphorylation event is necessary for S-phase transition in some cases. For example, phosphorylation of Id2 is important for p21Cip1 inhibition and promotion of proliferation (Matsumura et al., 2002). An illustration of roles of Ids in promoting proliferation can be seen in Figure 8A.

Figure 8. Molecular pathways by which Ids modulate proliferation, invasion, and apoptosis. (A) Role of Id3 in promoting proliferation. Ids induce cell proliferation through inhibition of E proteins or pRb, and in turn regulate CDKIs or E2F proteins, respectively. (B) Role of Id3 in promoting apoptosis and inducing tumor suppression. Genes marked in blue are shown to be significantly up-regulated by Id3 in microarray analysis. Genes marked in red are apoptosis- promoting pathway proposed in our study.

1.3.4. Ids as Regulators of Apoptosis

As revealed by more and more studies on Ids, the effects of Id expression is cell-context specific. In addition to their roles in differentiation and proliferation, Id proteins have been shown to suppress invasion and promote apoptosis in both normal and malignant cells (Figure

8B). Overexpression of Id1-3 proteins induces apoptosis in primary rat cells (Norton and

Atherton, 1998). Further, overexpression of Id3 promotes apoptosis of B lymphocyte progenitor cells in response to TGF-β (Kee et al., 2001). Id3 also sensitizes immortalized human keratinocytes to UVB- and X-ray- induced apoptosis (Lee et al., 2010; Simbulan-Rosenthal et al., 2006). In addition, Id proteins induce apoptosis in tumor cells, including osteosarcoma (Id2)

(Florio et al., 1998) and Burkitt’s lymphoma (Id3) (Richter et al., 2012). Id3 also sensitizes the sarcoma cell line MG-63 to cisplatin (Koyama et al., 2004).

Similar to other proteins important for cell cycle progression (such as c-Myc), ectopic Id expression induces apoptosis in cells depleted of mitogenic factors (Israel et al., 1999). Growth- promoting proteins induce cell proliferation, while growth factor withdrawal induces cell death.

The conflicting signals received by cells therefore trigger a cell death response (Israel et al.,

1999). Id1 and Id2 are found to enhance apoptosis in 32D.3 murine myeloid progenitor cells, in response to withdrawal of interleukin-3 (IL-3) (Florio et al., 1998). Id1, Id2, and Id3 all induce both proliferation and apoptosis in rat embryonic fibroblasts in response to serum depletion

(Norton and Atherton, 1998).

Ectopic expression of Id3 in cells with no detectable endogenous Id3 (e.g., A431 vulvar carcinoma cells) or with high endogenous Id3 (e.g., HaCaT keratinocytes) can both induce cell death (Langlands et al., 2000). This indicates that the amount of Id3 needed to maintain homeostasis is cell-type and context specific.

1.3.5. Ids as Potential Tumor Suppressors

Tumorigenesis is usually driven by collaborating mutations leading to both over activation of oncogenes and inactivation of tumor suppressor genes. Overexpression of the oncogenic protein Myc is a genetic hallmark for both neuroblastoma and Burkitt’s lymphoma, and Id3, as a downstream target of Myc (Seitz et al., 2011), has been shown to acquire inactivating mutations in these tumor types, suggesting that Id3 can act as a tumor suppressor gene.

Gene amplification of the N-myc gene and deletion of chromosome 1p36 are frequently seen in neuroblastoma tumors (Weith et al., 1989). Human Id3 was identified in this neuroblastoma consensus deletion region, and further analysis determined that 7 out of 11 neuroblastoma patients carried this Id3 deletion (Ellmeier et al., 1992), suggesting that Id3 can act as a tumor suppressor in neuroblastomas.

Further, three independent groups identified Id3 as a tumor suppressor gene for Burkitt’s lymphoma using high throughput techniques, including whole genome and exome sequencing

(Love et al., 2012; Richter et al., 2012; Schmitz et al., 2012).

Using functional genomic studies, c-myc, E2A, and Id3 were found to be frequently mutated in Burkitt’s lymphoma. Dysregulation of c-myc due to chromosome translocation t(8:14) adjacent to the immunoglobulin genes (IGs), found in almost all Burkitt’s lymphomas, induces proliferation of tumor cells (Campo, 2012). Similarly, the increased activation of E2A due to mutations in the bHLH domain is believed to contribute to the increased proliferation and survival of Burkitt’s lymphoma cells (Schmitz et al., 2012).

On the other hand, Id3 mutations (point mutations and deletions) occurring in 68% of

Burkitt’s lymphomas are mostly inactivating mutations, and most of them localize to the HLH and C terminal domains, suggesting the requirement of these regions for Id3 function (Figure

9A; (Richter et al., 2012).

Further, another research group concluded that Id3 is recurrently mutated in double-hit B cell lymphomas (DHL) (Gebauer et al., 2013). These tumors have morphological features between Burkitt’s lymphoma and diffuse large B-cell lymphoma (DLBCL). This aggressive type of lymphoma harbors two or more chromosomal rearrangements, involving cMyc and Bcl-2 or Bcl-6 loci, for example, IGH/MYC t: (8: 14) and IGH/BCL2 t: (18: 14) (Snuderl et al., 2010).

Sequencing results from patient samples indicate that Id3 inactivating mutations are frequently seen, especially in HLH and C-terminal domains (Figure 9B), which is consistent with Burkitt’s lymphoma.

Figure 9. Frequent Id3 mutations in lymphoma patients. (A) Id3 mutations in IG-MYC-positive Burkitt’s lymphomas. Most of Id3 mutations are inactivating mutations, resulting in downregulation of Id3 protein as well as reduced binding to other proteins through HLH domain in cancer cells. Figure from (Richter et al., 2012). (B) Id3 mutations in double-hit B- cell lymphomas. Most mutations are seen in HLH and C-terminal domains. Figure from (Gebauer et al., 2013).

In normal cells, Id3 is a dominant negative regulator of E12/47, and its E2A gene is shown to induce the expression of Id3, forming a negative feedback loop. In Burkitt’s lymphoma cells, Id3 mutations cause downregulation of Id3 protein, as well as decreased binding to E12/47, leading to attenuated inhibitory effects on E12/47 (Schmitz et al., 2012).

Tumor cells with Id3 mutations therefore gain growth advantages over wild-type Id3 (Schmitz et al., 2012). In addition to its growth-inhibitory effect, ectopic expression of wild-type Id3 in

Burkitt’s lymphoma cells results in induction of apoptosis (Richter et al., 2012). Id3 is therefore identified as a tumor suppressor gene in Burkitt’s lymphoma. An illustration of a proposed pathogenic mechanism involving Myc, E2A gene products, and Id3 in Burkitt’s lymphoma can be found in Figure 10.

Figure 10. A proposed pathogenic mechanism for Burkitt’s lymphoma. Myc, E2A and Id3 are frequently mutated in Burkitt’s lymphomas. Myc and E12/47 proteins both promote cancer cell survival and therefore act as oncogene products. Id3 on the other hand, represses E2A function, acting as a tumor suppressor gene. Figure modified from (Campo, 2012).

1.3.6. Ids and SCC

Keratinocyte maturation in the skin epidermis is a dynamic process. Initially, basal keratinocytes of the epidermis proliferate. These cells then withdraw from the cell cycle and accumulate differentiation markers as they move upward to the spinous layer (stratum spinosum) at the early differentiation stage. Later stages of keratinocyte differentiation are represented in the stratum granulosum and stratum lucidum. Eventually, these keratinocytes will continue to migrate upward and differentiate into anucleate cells called corneocytes (stratum corneum). The different epidermal layers can be observed upon histological staining with hematoxylin and eosin

(H&E; Figure 11)

Figure 11. Epidermis structure. Epidermis is comprised of different layers (from bottom to top): stratum basale, stratum spinosum, stratum granulosum, stratum lucidum (not shown), and stratum corneum. Highly proliferative keratinocytes are in stratum basale. These cells migrate upward as they undergoing differentiation. Figure from (Simpson et al., 2011).

As regulators of proliferation, differentiation, and apoptosis, Id proteins play important roles in keratinocyte maturation as well as in progression to SCC. It has been shown that Id proteins are expressed primarily in the proliferative basal layer of epidermis in the normal human skin. A decrease in Id expression is observed in the spinous and granular layers (Langlands et al., 2000), consistent with the role of Ids as inhibitors of differentiation. Id proteins have also been shown to play an important role in primary human keratinocytes in culture. Overexpression of Id proteins has been shown to increase life span of primary human keratinocytes either through immortalization or delaying senescence (Alani et al., 1999; Nickoloff et al., 2000). This indicates dysregulation of Id proteins can induce characteristics similar to those of AK or pre-

SCC.

Investigators have speculated that dysregulation of Ids contributes to progression of SCC

(Nishimine et al., 2003). Id proteins are expressed at much higher levels in poorly- vs. well- differentiated SCC of the head and neck. In addition, Ids are not expressed in the keratinized foci in well-differentiated SCC (Langlands et al., 2000; Nishimine et al., 2003), consistent with the role of Ids as inhibitors of differentiation. Further, higher Id expression levels are usually, but not always, correlated with metastasis and prognosis of SCC (Luo et al., 2012; Nishimine et al., 2003). Thus, studying the mechanism by which Ids determine SCC prognosis can help evaluate the potential of using Ids as therapeutic targets for skin cancer.

It is worth mentioning that many of the studies used immunofluorescent staining to examine Id protein levels in tumors. Many research teams, including our own, have found that most commercially available antibodies might not be specific for Id protein staining; therefore, data generated through immunofluorescent staining requires careful evaluation (Perk et al., 2005;

Sikder et al., 2003).

Id3 has been shown to mediate suppression of SCC invasion by ΔNp63α, the dominant isoform of p63 in both normal and malignant squamous cells. ΔNp63α is important in formation of stratified squamous epithelial structures (to ensure asymmetric division and maintain cell polarity). Investigators have discovered a Snail-p63-Id3-Ets1-MMP2 axis in SCC (Figure 8B).

Snail is a transcription factor that promotes epithelial to mesenchymal transition (EMT) leading to cancer invasion. It has been shown that Snail-induced repression of ΔNp63α induces invasion, and overexpression of ΔNp63α represses invasion by inducing Id3 expression in SCC cell line HOC313. Id3 is demonstrated to repress E-Twenty-Six Transformation Specific (Ets-1) transcriptional activity, therefore blocking Ets-1-induced MMP-2 expression (Higashikawa et al.,

2009).

Id2 has been shown to promote E-cadherin expression in mouse mammary epithelial cells

(Kowanetz et al., 2004). Id3 has been shown to induce E-cadherin promoter activity in HEK293 cells (Higashikawa et al., 2009). Further, in prostate cancer cell lines DU145 and PC3, knocking down Id1-Id3 significantly downregulates E-cadherin mRNA expression (Asirvatham et al.,

2007). It is believed that the promotion of E-cadherin expression by Id proteins, at least in part, is through inhibiting E proteins from suppressing E-cadherin expression, and this induction of E- cadherin is important for suppressing invasion in SCC (Asirvatham et al., 2007; Higashikawa et al., 2009). For an illustration of regulation of invasion by Ids, please refer to Figure 8B.

1.3.7. Ids and TCF Transcription Factors

Besides HLH transcription factors, Ids also bind and regulate ETS domain transcription factors (Yates et al., 1999). For example, Id1 binds Ets1 and Ets2 to downregulate p16Ink4a expression (Ohtani et al., 2001). Id3 also inhibits Ets-1 induction of MMP-2 expression

(Higashikawa et al., 2009), and Id1-Id3 all bind Ets-like transcription factor-1 (Elk-1), a member of TCF (below), to inhibit its DNA-binding ability (Yates et al., 1999).

Ternary Complex Factors (TCFs) belong to the ETS-domain transcription factor superfamily. TCFs are comprised of three genes located on different chromosomes that share conserved domains as shown in Figure 12 (Besnard et al., 2011; Sharrocks, 2002). The ETS domain is the DNA-binding domain (DBD), which allows for binding of TCFs to serum response element (SRE) residues CAGGA (Besnard et al., 2011) on DNA promoters of responsive genes. In addition, the ETS domain contains the nuclear localization signal (NLS) and the nuclear export signal (NES), allowing for nuclear translocation. TCF ETS-domain transcription factors include Elk-1, Elk-3 (also known as serum response factor (SRF) accessory protein 2, SAP-2), and Elk-4 (as known as SRF accessory protein 1, SAP-1). They play important roles in many cellular processes, including embryogenesis, proliferation, and apoptosis.

Figure 12. Conserved domains in TCF transcription factors. ETS is the DNA-binding domain. B box domain is SRF-binding site. R domain is the repression domain. D domain is the MAPK- binding site. TAD domain is MAPK phosphorylation site. F domain is ERK-docking site. Figure from (Sharrocks, 2002).

The biological functions of TCFs are investigated using knockout mice. Elk-1-deficient mice are viable and have normal life spans. These mice show elevated c-Fos expression in the brain tissues following severe seizures. Elk-3-deficient mice are viable but die shortly after birth due to respiratory failure. Elk-4-deficient mice are viable but fail to generate single positive T cells and also exhibit symptoms similar to Castleman’s disease (lymphoproliferative disorder)

(Cesari et al., 2004; Sharrocks, 2002).

The four conserved domains are important for functions of TCFs. Binding of Elk-1 to

SRF is through the B box domain. The R domain of Elk-1 is the repression domain. In the case of Elk-1, this domain contains several lysine residues that can be modified by Small Ubiquitin- like MOdifier (SUMO)ylation, which represses transcriptional activity of Elk-1. On the other hand, the R domain in SAP1 and R2 in SAP2 is termed Net inhibitory domain (NID). This is a cis-acting repression domain with an HLH-like motif.

Elk-1 is activated by three MAPKs: extracellular-signal related kinase (ERK), c-Jun N- terminal kinase (JNK), and p38. ERK mostly responds to growth factors and mitogens, while

JNK and p38 mostly response to cytokines and stress signals (Besnard et al., 2011).

Phosphorylation of Elk-1 facilitates its binding to SRF and SRE, as well as recruitment of transcription coactivator complex, including cAMP response element-binding protein (CREB) binding protein (CBP), p300, and histone acetyltransferase (HAT). The D domain contains the

MAPK-binding site. The transcriptional activation domain (TAD) domain contains a

Serine/Threonine (S/T) phosphorylation motif targeted by MAPKs. The F domain is the ERK docking site. It is required for Elk-1 activation by ERK (Besnard et al., 2011). TCF activities are shown be regulated through post-translational modification, including SUMOylation,

29 phosphorylation, and ubiquitylation (Figure 13). Through these regulatory mechanisms, TCFs mediate a variety of cellular pathways.

Figure 13. Post-translational modifications of Elk-1 protein. SUMOylation (SUMO) represses Elk-1 activity while phosphorylation (P) activates Elk-1 activity. When Elk-1 is ubiquitylated (Ub), its protein is degraded in the proteasome. Ubc-9 is a SUMO E2-conjugating enzyme, which adds SUMO to lysines 230, 249 and 254 residues in R domain. On the other hand, PIASxa is a SUMO E3 ligase, which removes SUMO molecule from Elk-1. MAPKs, including ERK, p38, and JNK, phosphorylate Elk-1 (on serine 383 and 389 in TAD domain) while PP2B phosphatase removes phosphate from Elk-1.

TCF transcription factors regulate expression of immediate early genes (IEGs, e.g., c-fos and early growth response gene-1, egr-1) in response to mitogenic stimuli or stress. Mitogenic stimuli activate MAPK signaling, and TCF is activated through phosphorylation and form a protein complex with a dimer of SRF. This protein complex binds SRE site on the promoter region of immediate early gene and induces gene transcription. For an illustration of ternary complex formation on DNA promoter, please refer to Figure 14A.

Id proteins have been shown to inhibit ternary complex formation between TCF and

SRFs on SRE (Yates et al., 1999). Ids prevent TCFs from binding to SRE. By inhibiting ternary

30 complex formation, Ids inhibit promoter activity of IEGs (Yates et al., 1999); see Figure 14B for an illustration). Further, early response gene, egr-1 has been shown to induce Id transcription

(Tournay and Benezra, 1996). Since expression of both IEGs and Ids are induced rapidly by serum, this can form a negative feedback loop to regulate the rapid induction followed by a rapid return to basal expression level of IEGs in response to serum stimulation (Yates et al., 1999).

For an illustration of the negative feedback loop established between IEGs and Ids, please refer to Figure 14A.

Figure 14. Negative feedback loop established between immediate response genes and Ids. (A) Upon mitogenic stimulation, MAPK signaling-induced ternary complex formation between TCF, SRF and SRE induces transcription of immediate early genes (IEGs, e.g., c-fos) and induce cell proliferation. Immediate early gene induces Id expression B) Id proteins disrupt ternary complex formation through direct interaction with TCF protein and therefore repress transcription of IEGs.

In addition, Ids and TCFs can both regulate paired box (Pax) transcription factors. Other than ETS-domain transcription factors, Elk-1 has been shown to bind Pax5 transcription factor through interaction between ETS domain and paired domain, without involvement of B box domain (Fitzsimmons et al., 1996). This transcriptional activator complex binds the B-cell specific mb-1 gene promoter, encoding B cell antigen receptor for Ig-α signaling, important for early B cell development (Hobeika et al., 2006). An illustration of transcriptional activation of mb-1 via the Pax/TCF complex is shown in Figure 15A.

Ids can disrupt the binding of this transcriptional activator complex to DNA promoter region. Id3 has been shown to bind Pax-5, via the interaction between HLH and paired domain, preventing Pax-5 alone and Elk-1/Pax-5 complex from binding to mb-1 (Roberts et al., 2001).

An illustration of how Id protein disrupts DNA binding by Pax/TCF complex can be found in

Figure 15B.

Figure 15. Pax/TCF transcriptional activator complex is disrupted by Id protein. (A) TCF and Pax protein complex binds and induces mb-1 promoter during early B-cell lymphopoiesis. (B) Id protein disrupts transcriptional activation by Pax/TCF, via preventing free Pax and TCF proteins from binding to DNA (Roberts et al., 2001).

Conversely, Pax proteins can regulate Id proteins. Id2 and Id3 have been shown to be downstream targets of Pax3 and Pax7. Upstream of Id3 promoter, there are binding sites (paired domain, PD and homeodomain, HD) for Pax3 and Pax7 binding. Induction of Id2 and Id3 by

Pax3 and Pax7 in quiescent satellite (skeletal muscle progenitor) cells is believed to prevent premature differentiation. Induction of Ids by Pax proteins can also serve as a negative feedback loop, regulating transcriptional activities of TCFs.

In addition to proliferation and differentiation, Elk1 has been shown to induce apoptosis.

It has been shown that Elk-1 induces bax expression through an Egr-1-mediated pathway in immortalized human keratinocytes, HaCaT cells in response to sodium arsenite exposure. Egr-1, a downstream target of Elk-1, binds to the bax promoter and induces its expression. In addition,

Elk-1 was shown to bind to the p21Cip1 promoter and induce its expression, triggering apoptosis in HaCaT cells (Shin et al., 2011).

Overexpression of Elk-1 in dendrites has also been shown to induce apoptosis in rat neurons (Barrett et al., 2006; Besnard et al., 2011). The mechanism is believed to be related to its association with mitochondrial permeability transition pore complex (PTP). Overexpression of Elk-1 also induces apoptosis in breast cancer cells MCF7 as well as in rat fibroblast cells in response to calcium ionophore treatment (Shao et al., 1998).

The established molecular mechanisms by which Ids regulates TCFs (and vice versa) are related to proliferation, survival, and development. Id and TCF family proteins play important roles in apoptosis; however, the death-signal induced cascades mediated through Ids and TCFs have not yet been explored. In our current study, we aim to investigate the Id- and TCF- dependent apoptosis and the therapeutic potential of these proteins in skin cancer.

1.4. Introduction to Apoptosis

Cell death is an essential cellular process to maintain tissue homeostasis. Perturbed regulation in cell death leads to many diseases, including cancers. The Greek term “apoptosis”

(dropping off leaves from trees) was introduced in 1972 by Kerr, Wyllie, and Currie to describe a form of cell death that involves specific cellular morphology changes (Degterev et al., 2003;

Kerr et al., 1972). Cells that are undergoing apoptosis have the following distinct features: cell shrinkage, membrane blebbing, chromatin condensation, DNA and nuclear fragmentation, and formation of apoptotic bodies and phagocytosis by macrophages (Kerr et al., 1972).

1.4.1. Caspases

Apoptosis-regulating proteins are evolutionally conserved. In C. elegans, they are Ced-3,

Ced-4, and Ced-9 (Yuan et al., 1993). Ced-9 is an anti-apoptosis protein, and Ced-3 and Ced-4 both are pro-apoptotic proteins. Ced-3 is a cysteine protease, and its homologue in human is cysteine-dependent aspartate-specific protease (caspase) protein. Ced-4 is an adaptor protein, and its homologue is human is apoptosis protease activating factor-1 (Apaf-1). Ced-9 homologue in human is the B-cell lymphoma 2 (Bcl-2) protein family.

Caspases are important initiators and effectors of apoptosis. There are 11-12 caspases in humans (Degterev et al., 2003), including caspase 1-10 and caspase 14 (Degterev et al., 2003).

Interestingly, caspase 12 is a pseudo gene but activated in some populations. Caspases are classified as either initiator or effector caspases (Table 3). All caspases have a cysteine residue at the enzyme active site and cleave substrates after an aspartate residues. In addition, all caspases are initially synthesized as inactive zymogens. Each zymogen contains a prodomain, a large subunit, and then a small subunit (Figure 16). To activate the zymogen, a caspase is first

34 cleaved between large and small subunits, and then cleaved between the large catalytic unit and prodomain, as shown in Figure 16 (Degterev et al., 2003).

Initiator caspases have longer prodomains compared to effector caspases (Figure 16).

Within the prodomain of initiator caspases, protein-protein interaction motifs can be found, including the death effector domain (DED) and caspase activation or recruitment domain

(CARD), which are responsible for binding to adaptor proteins and propagate death signals to effector caspases by activating effector caspases through proteolytic cleavage (Degterev et al.,

2003).

Table 3. Caspases are classified into initiator and effector caspases.

Initiator Caspase Initiator Caspase Effector Caspase Apoptosis/Inflammation Apoptosis Apoptosis

Casp-1 Casp-2 Casp-3

Casp-4 Casp-8 Casp-6

Casp-5 Casp-9 Casp-7

Casp-12 Casp-10 Casp-14

Figure 16. Caspase domains and proteolytic cleavage sites. Caspase structure includes prodomain, large and small subunits. Initiator caspases have longer prodomain than effector caspases. Within prodomains, CARD and DED domains can be found; they are binding domains to adaptor proteins. Proteolytic cleavage sites are between large and small subunits and between prodomain and large subunit.

Effector or executioner caspases are responsible for cleaving cellular substrates. There are close to 1000 caspase substrates so far identified and they are comprised of a wide range of proteins, including structural proteins actin and nuclear lamins, as well as signaling proteins

Wee1 and Akt kinases (Degterev et al., 2003). For more caspase substrates, please see Figure

17. Cleavage of these important cellular proteins leads to morphological changes including chromatin condensation, membrane blebbing, and formation of apoptotic bodies, eventually leading to endocytosis. This process can be used to design therapeutics to eliminate cancer cells.

Figure 17. Caspase substrates. Caspases cleave substrates at specific sites, causing loss of normal cell functions and leading to cell death. Figure from (Degterev et al., 2003).

1.4.2. Bcl-2 Family Proteins

In addition to caspases, Bcl-2 family proteins are also very important in regulating apoptosis. There are two categories of Bcl-2 family proteins: the anti-apoptotic and the pro- apoptotic proteins. Bcl-2 family proteins have conserved protein-protein interaction domains,

Bcl-2 homology (BH) 1-4 domains. BH1, 2, and 4 are important domains to repress death while the BH3 domain is important to promote death. For Bcl-2 family protein classification, please refer to Table 4.

Table 4. Bcl-2 family protein classification.

Category Domain Present Molecule

Anti-apoptotic BH1, BH2, BH3, BH4 Bcl-2, Bcl-xL, Mcl-1

BH1, BH2, BH3 Bax, Bak

Pro-apoptotic BH3 (Activator) Bid, Bim, PUMA

BH3 (Sensitizer) Bad, NOXA, Bmf

Anti-apoptotic Bcl-2 proteins have all 4 BH domains, while pro-apoptotic Bch-2 proteins have either BH1-BH3 domains (multi-domain proteins) or BH3 alone. BH3-only pro-apoptotic proteins, including Bid, Bim, and Bad, receive death signals and activate multi-domain pro- apoptotic proteins, such as Bax and Bak, and activate apoptosis through regulating mitochondria outer membrane permeability.

Bcl-2 proteins regulate apoptosis mainly through regulating mitochondria (either through controlling release of pro-apoptotic protein from mitochondria or through regulation of mitochondrial function). Release of pro-apoptotic proteins from mitochondria promotes apoptosis. Mitochondrial transition pore complex proteins Bax and Bak induce outer membrane permeability. Bax is located in cytosol in healthy cells and oligomerizes and translocates to mitochondria of apoptotic cells. Mitochondrial depolarization and ROS generation leads to mitochondrial dysfunction, due to the eventual loss of cytochrome C. Cytochrome C, along with other proteins, then forms an apoptosome protein complex and triggers apoptosis.

BH3-only pro-apoptotic molecules can activate multi-domain pro-apoptotic proteins either by direct activation (activator) or indirect activation (sensitizer). Activators are Bid, Bim, and PUMA. Sensitizers are Bad, Noxa, and Bmf. Activators activate Bax and Bak directly

38 while sensitizers compete with Bid, Bim, and PUMA from anti-apoptotic protein binding allowing Bax or Bak to oligomerize.

1.4.3. Apoptotic Signaling Pathways

There are two types of apoptotic signaling pathways: the extrinsic and the intrinsic pathways. The extrinsic pathway is mediated through cell surface death receptors, while the intrinsic pathway is mediated through mitochondria. An illustration of apoptotic signaling pathway is shown in Figure 18.

Figure 18. Apoptotic signaling pathways. Apoptosis is mediated through either death receptor- mediated extrinsic pathway or mitochondria-mediated intrinsic pathway. Figure from (Igney and Krammer, 2002).

The extrinsic apoptotic signal cascade relays the extracellular death signal to the cytosol through the death receptor family of eight proteins in humans, including Fas, TNFR1, TNFR2, and DR5. Upon death ligand binding, pre-assembled trimeric death receptors interact with adaptor proteins through interaction between their death domains (DD), forming the death- inducing signaling complex (DISC). Adaptor proteins (e.g., Fas-Associated Death Domain adaptors FADD) are recruited to and bind to death receptors (e.g., Fas) through homotypic interactions via the C-terminal DD motifs, and interact with initiator caspase (e.g., Caspase-8) through the N-terminal DED motif (Degterev et al., 2003). Initiator caspases are usually recruited to the DISC as homodimers; once recruited, they are activated through autocatalytic cleavage. Autocatalytic cleavage of initiator caspases can occur by “induced proximity” by overexpression in the absence of death receptors. The DISC assembly is completed as ligand- bound receptor, adaptor and initiator caspases protein complex is internalized through clathrin- mediated endocytosis. Without receptor internalization, pro-survival pathways mediated by nuclear factor kappa-light-chain-enhancer of activated B cells (NFκB) and ERK1/2 pathways can be activated instead.

Death signaling downstream of initiator caspases can be cell-type specific. In type I cells, in which DISC assembly and internalization is rapid and sufficient, active caspase-8 levels are able to activate apoptosis directly through proteolytic cleavage of effector caspases. In type

II cells, on the other hand, DISC assembly and internalization is slow, and active caspase-8 levels are insufficient to directly cleave effector caspases. Rather, death receptor-mediated apoptotic signals are amplified through a mitochondrial-mediated apoptotic pathway (Igney and

Krammer, 2002). In this case, caspase-8 cleaves the pro-apoptotic Bcl-2 protein member Bid,

40 causing Bid translocation to the mitochondria. Cytochrome C release, apoptosome formation, and cleavage of caspase-3 are described in the next paragraph.

Death signals inside the cell (e.g., DNA damage) triggers intrinsic apoptotic signaling cascade, mediated by mitochondria. This signaling cascade is mediated through pro-apoptotic protein Bax or Bak interactions, causing mitochondrial outer membrane depolarization and cytochrome C release. Cytochrome C, Apaf-1 adaptor proteins, and caspase-9 dimers together form apoptosome complex, which leads to caspase-9 activation, followed by caspase-3 activation

(Degterev et al., 2003), leading to cell death (Figure 18).

1.4.4. Ids and Apoptosis

Id proteins have been shown to promote apoptosis in both normal and malignant cells.

Overexpression of Id1-3 proteins induced apoptosis in primary rat cells (Norton and Atherton,

1998). Overexpression of Id3 promoted apoptosis of B lymphocyte progenitor cells in response to TGF-β (Kee et al., 2001). Id3 also sensitizes immortalized human keratinocytes to X-ray- and

UVB-induced apoptosis (Lee et al., 2010; Simbulan-Rosenthal et al., 2006). In addition, Id proteins induced apoptosis in tumor cells, including osteosarcoma (Id2; (Florio et al., 1998) and

Burkitt’s lymphoma (Id3; (Richter et al., 2012). Id3 also sensitizes the sarcoma cell line MG-63 to cisplatin treatment (Koyama et al., 2004).

Similar to other proteins important for cell cycle progression (such as c-Myc), ectopic Id expression induces apoptosis in cells depleted with mitogenic factors (Israel et al., 1999).

Growth-promoting proteins induces cell proliferation, while growth factor withdrawal induces cell death. The conflicting signals received by cells therefore trigger a cell death response (Israel et al., 1999). Id1 and Id2 are found to enhance apoptosis in 32D.3 murine myeloid progenitor

41 cells, in response to withdrawal of interleukin-3 (IL-3) (Florio et al., 1998). Id1, Id2, and Id3 all induce both proliferation and apoptosis in rat embryonic fibroblasts in response to serum depletion (Norton and Atherton, 1998).

In this study, we investigate the involvement of Id3 in squamous carcinoma cells in cell culture and xenograft models. We hypothesize that Id3 induces apoptosis in SCC, which can be a target when designing new therapeutic strategies.

Chapter 2. Materials and Methods

2.1. Cell Culture

Human squamous carcinoma cell lines A431, SCC4, SCC9, and SCC25 cells were obtained from Georgetown University Tissue Culture Shared Resources. A431 cells were maintained in Dulbecco’s Modified Eagle Medium (DMEM, #10-013-CV, Corning) with 1%

Penicillin (10,000 IU/ml), streptomycin (10,000 µg/ml), and 10% fetal bovine serum (FBS).

Tet-approved FBS (#631107, Clontech, lot numbers listed below) was used for inducible cell lines. Other SCC cell lines were maintained in DMEM/F-12 50:50 (#10-092-CV, Corning), 1%

Penicillin (10,000 IU/ml), streptomycin (10,000 µg/ml), and 10% fetal bovine serum (FBS, FB-

11, Omega Scientific). Cells were passaged in 1:8 split every 3 days.

Lot numbers of Tet-approved FBS used for each experiment are shown as follows:

Experiment Lot No.

1) Cell growth on Matrigel (Figure 30) A301117007

2) Invasion assays (Figure 31) A301117007

3) Immunoblots checking invasion and stem cell markers (Figure 32) A301117007

4) Growth curves (Figure 33-Figure 34) A301117007

5) Cell cycle analyses (Figure 35) A301117007

6) Growth curves under co-culture conditions (Figure 36) A301117007

7) RT-PCR verifying microarray candidate genes (Figure 39) A13001

8) Immunoblots verifying microarray candidate genes (Figure 40) A301117007

6) Immunoblots showing Elk-1 and caspase-8 induction by Id3 (Figure 41) A13001

7) Growth curves under caspase inhibitors (Figure 44) A13001

8) Soft agar assays (Figure 45) A13001

9) Immunoblot showing caspase-8 levels under Id3 knockdown conditions in FEB11 other SCC lines (Figure 46)

10) Growth curves for Id3 truncation mutants (Figure 49) A13001

11) Cell cycle analyses for Id3 truncation mutants (Figure 50) A13001

12) Cell viability after Id3 and 5-FU or CDDP treatment (Figure 51) A13001

13) Growth curves for A431 xenografts after Id3 induction (Figure 52) A301117007

2.2. Tet-inducible System

To investigate the role of ectopic expression of Id3 in SCC, an inducible system is selected due to failed attempts to establish constitutive cell lines expressing Id1-Id3 reported by another research group (Langlands et al., 2000).

A tetracycline-regulated inducible system (Tet-on) is used in our study (Figure 19). In this system, Id3 is cloned into Tet operator vector (pcDNA4/TO; Figure 20). Subsequently, cells are transfected with both Tet repressor (pcDNA6/TR; Figure 21) and Tet operator constructs, followed by selection with antibiotics for stable clones expressing both Tet repressor and operator constructs. In these cells, Tet-repressor proteins bind to Tet-operator sequences, leading to repression of Id3 transcription. Upon addition of tetracycline molecules, they bind Tet- repressors and trigger a conformational change to release repressor from operator sequences.

Therefore, Id3 transcription can take place. This is illustrated in Figure 22.

Figure 19. Schematic illustration for procedure to establish inducible cell lines. Figure from (Life Technologies, 2011).

Figure 20. pcDNA4/TO vector map. Map from (Life Technologies, 2011).

Figure 21. pcDNA6/TR vector map. Map from (Life Technologies, 2011).

Figure 22. Addition of tetracycline induces transcription of Id3. Figure from (Life Technologies, 2011).

2.3. Plasmids

Full-length Id3 as well as Id3 deletion mutants were cloned into pcDNA4/TO

(Invitrogen™) using Xho I and Xba I restriction sites (Figure 20). Full-length Id3 was subcloned from pCMS/EGFP-Id3 construct. pCMS/EGFP-Id3 was cloned as previously described

(Simbulan-Rosenthal et al., 2006). Id3 truncations were engineered with a Flag tag for detection and cloned using the following primer sequences:

Primer 1: 5’-ACTGCTCGAGGCCGCCACCATGAACATGTTGCTGGACGACATGAACCACTG-3’

Primer 2: 5’-ACTGTCTAGATCACTTATCGTCGTCATCCTTGTAATCGTGGCAAAAGCTCCTTTTGTCG-3’

Primer 3: 5’-ACTGCTCGAGGCCGCCACCATGAAGGCGCTGAGCCCGG-3’

Primer 4: 5’-ACTGTCTAGATCACTTATCGTCGTCATCCTTGTAATCCTGCAGGTCGAG AATGTAGTC-3’

Primer 5: 5’-ACTGCTCGAGGCCGCCACCATGGTAGTCCTGGCCGAGC-3’

Primer 6: 5’-ACTGTCTAGATCACTTATCGTCGTCATCCTTGTAATCGCTCAGCGGCTCCTCAGC-3’

Figure 23. Primer binding sites for cloning Id3 mutants. ∆N39 was amplified using primers 1 and 2. ∆C38 was amplified using primers 3 and 4. ∆N81 was amplified using primers 5 and 2. ∆C80 was amplified using primers 3 and 6. Each forward primer contains a Kozak sequence (GCCGCCACC) and an Xho I restriction site (CTCGAG). Each reverse primer contains an Xba I restriction site (TCTAGA) and flag tag sequence (CTTATCGTCGTCATCCTTGTAATC) encoding DYKDDDDK.

2.4. Inducible Cell Lines

A431/Id3 inducible lines were established first by transfecting cells with pcDNA6/TR

(Figure 21; Invitrogen™) using Lipofectamine-LTX (Invitrogen™). After selection with blasticidin (7.5 µg/ml) for 10 days, pooled clones of A431 cells stably expressing Tet repressor in the pcDNA6 vector (A431/TR) were established. Then, A431/TR cells were transfected with pCDNA4/TO-Id3 or pCDNA4/TO-Id3 mutant constructs, followed by selection with ZeocinTM

(250 µg/ml) for 10 days. Individual clones from each cell line were isolated and Tet inducibility

(1 µg/ml) was verified by RT-PCR and immunoblot analysis. A431 vector control (A431/Vc) cells were established by transfecting A431/TR with pCDNA4/TO (Invitrogen™) followed by selection with ZeocinTM (250 µg/ml) for 10 days.

2.5. Fluorescently-labeled Cell Lines

pLHCX/GFP and pLHCX/Ds-Red constructs were used to transduce A431/Id3 and

A431/Vc cells, respectively. The retroviral packaging cell line ΦNX (Phoenix) were transfected with pLHCX/GFP and pLHCX/Ds-Red retroviral constructs separately using Lipofectamine-

LTX (Invitrogen™). 48 h after transfection, viral supernatant was filtered and used to transduce

A431/Id3 and A431/Vc cells. Subsequently, transduced cells were cultured in medium with hygromycin (80 µg/ ml) for 10 days to select for stable green or red fluorescent clones.

2.6. Growth Curve

Growth Curve under Regular Serum. 3x105 GFP-expressing A431/Id3 cells were plated per well in 6-well plates in DMEM/Pen/Strep supplemented with 10% Tet-approved FBS.

24 hours after plating, cells in three wells were induced with Tet (1 µg/ml, #T3383, Sigma).

Cells were counted every day after Trypan Blue staining using a hemocytometer.

Growth Curve on Matrigel. 3x104 GFP-expressing A431/Id3 and Ds-Red-expressing

A431/Vc cells were plated in Matrigel-coated (250 µg/ml, #BD354248) 6-well plates in

DMEM/Pen/Strep supplemented with 10% Tet-approved FBS. 24 hours after plating, cells in three wells were induced with Tet (1 µg/ml). Cell growth was determined every day measuring fluorescence intensity by VICTOR2TM Fluorometer, using 485/535 filter for GFP and 570/615 filter for Ds-Red.

Growth Curve under Low Serum. 3x104 GFP-expressing A431/Id3 cells were plated per well in 6-well plates in DMEM/Pen/Strep supplemented with 0.1% Tet-approved FBS.

Immediately after cells were split into 6-well plates, three wells were induced with Tet (1 µg/ml) for 24 hours. Every two days, cells were counted after Trypan Blue staining using a

49 hemocytometer. The same procedure was also repeated with DsRed-expressing A431/Vc cells as well as all A431/Id3 mutant cell lines.

Growth Curve under Co-culture Conditions. For co-cultures, 1.5 x 104 GFP A431/Id3 and DsRed A431/Vc cells were cultured together in a 6-well plate in DMEM/Pen/Strep supplemented with 0.1% Tet-approved FBS. Three wells were induced with Tet (1µg/ml) and the other three were uninduced. 20X pictures (5 per well) were taken every two days using an

Olympus epifluorescence microscope. Percentages of areas occupied by either green or red cells were quantified using ImageJ software.

Growth Curve under Caspase Inhibitors. For the growth of GFP A431/Id3 cells under

20 µM of caspase inhibitors, Carboxybenzyl Ile-Glu-Thr-Asp Fluoromethylketone (Z-IETD- fmk; #550380, BD Biosciences) and Carboxybenzyl Val-Ala-Asp Fluoromethylketone (Z-VAD- fmk; #550377, BD Biosciences), 3x104 cells were plated per well in a 6-well plate (under 0.1%

FBS). The same quantification method procedure was performed using ImageJ software for determining growth under both induced and uninduced conditions.

2.7. Cell Cycle Analysis

3x104 A431/Id3 cells were plated per well in 6-well plates in DMEM/Pen/Strep supplemented with 0.1% Tet-approved FBS. Immediately after cells were split into 6-well plates, three wells were induced with Tet (1 µg/ml) for 24 hours. Every other day, cells were collected, fixed in ethanol, stained with propidium iodide to determine DNA content, and analyzed by Flow Cytometry (FACStar). The same procedure was repeated for A431/Id3 mutant cell lines.

2.8. Cell Invasion Assay

5x105 cells A431/Id3 cells plated in 6-well Matrigel Invasion chambers (#354480, BD

Biosciences). Immediately after cells were split into 6-well plates, three wells were induced with

Tet (1 µg/ml) for 24 hours. Cells invaded through basement membrane was stained with Giemsa stain and areas occupied by invaded cells were quantified by ImageJ software.

2.9. Microarray Analysis

Affymetrix HG-U133A 2.0 gene chips (containing 14,500 human genes) were used. The uninduced and induced RNA samples in both A431/Id3 and A431/Vc cell lines were extracted and subjected to cDNA production, fluorescent dye labeling, sample hybridization to gene chip, data scanning, and analysis. Georgetown University Genomics and Epigenomics Shared

Resource generated the microarray data. Gene expression raw data was normalized using dChip software and analyzed with MetaCore software. The induced A431/Id3 group was compared to

A431/Id3 uninduced, A431/Vc uninduced, and A431/Vc induced individually. A heat map was generated for genes that were up- or down-regulated by Id3 by at least 1.5 fold.

2.10. siRNA Transfection

To transfect A431/Id3 cells with siRNA, 4.5x105 cells per well were plated to a 6-well plate to medium supplemented with 0.1% FBS. To transfect SCC cell lines (SCC4, SCC25 and

SCC9) with siRNA, 1.5x105 cells per well were plated to a 6-well plate to medium supplemented with 10% FBS. 2 µl of RNAi Max (#13778-030, Invitrogen) per well was used to transfect cells with different siRNAs (30 pmol). Control siRNA (#36869, Santa Cruz Biotechnology), Id3 siRNA (#38002, Santa Cruz Biotechnology), and Elk-1 siRNA (#35290, Santa Cruz

Biotechnology) were used to knock down transcripts.

2.11. Tumor Xenografts

106 GFP-labeled A431/Id3 cells were injected subcutaneously along with 0.5 x 106

DsRed-labeled A431/Vc cells into the flanks of male athymic nude mice. Mice were divided into two groups: one fed with regular diet (provided by Division of Comparative Medicine at

Georgetown University) and the other fed with doxycycline (Dox)-containing diet (200 mg/kg;

Bio-Serv #S3888). Tumor sizes were monitored using the CRi Maestro™ In-Vivo Imaging

System at the Preclinical Imaging Research Laboratory (PIRL) of Georgetown University.

2.12. Immunoblot Analysis

Protein lysates were prepared using lysis buffer containing 5% SDS and 5 mM DTT in

125 mM Tris-HCl (pH 6.8), followed by sonication. Tris-glycine SDS PAGE gel electrophoresis was performed followed by transfer to nitrocellulose membranes. Blocking was done using 5% non-fat milk for 1 hour at room temperature. Primary antibody incubation was achieved overnight at 4 °C. The primary antibodies used included: anti-Id3 (#SC490, 1:200, Santa Cruz

Biotechnology), anti-GAPDH (#MAB374, 1:8000, Millipore), anti-Id31 (#SC488, 1:200, Santa

Cruz Biotechnology), anti-active caspase-3 (#SA-320, 1:2000, Biomol), anti-caspase-9 (#PC680,

1:200, Calbiochem), anti-Elk-1 (#SC355, 1:200, Santa Cruz Biotechnology), anti-TBK-BP1

(#IMG6172A, IMGENEX, 1:500), anti-MXD3 (#H00083463, 1:1000, Abnova), anti-caspase-8

(#MA1-41280,1:500, Thermo Scientific), anti-β-actin (#SC47778, 1:200, Santa Cruz

Biotechnology),anti-Bid (#2002, 1:1000, Cell Signaling), anti-Fas (#SC715, 1:200, Santa Cruz

Biotechnology), anti-E-cadherin (#3195, 1:1000, Cell Signaling), anti-Sox2 (#SC17320, 1:200,

Cell Signaling), anti-Oct3/4 (#SC5279, 1:200, Santa Cruz Biotechnology), anti-MMP2

(#SC13594, 1:200, Santa Cruz Biotechnology). Each membrane was then incubated with

52 secondary antibody conjugated to horseradish peroxidase for 1 hour at RT. Chemiluminescent substrate (#34075, Thermo Scientific) was used for imaging.

2.13. Immunostaining

Paraffin-embedded xenograft tumor sections were deparaffinized with xylene followed by antigen retrieval in 6.5 mM sodium citrate, pH 6.0. Sections were incubated with antibodies specific for active caspase-3 (#9661, 1:200, Cell Signaling) and Ki-67 (#NB110-90592, 1:3200,

Novus Biologicals) overnight at 4°C. The following day, sections were stained with Alexa 594- conjugated secondary antibodies (#1:500, Invitrogen) for 1 hour at RT and visualized under fluorescence microscopy. Cell scoring was performed using MetaMorph software at Microscopy and Imaging Shared Resource of Georgetown University.

2.14. RT-PCR/qRT-PCR

RNA was extracted using TRIzol® reagent (#15596018, Life Technologies). For Id3,

TBKBP-1, ELK-1, MXD3, and GAPDH, 1 µg of RNA was used for RT-PCR reaction (30 cycles). RT-PCR was performed using Gene Amp EZ rTth RNA PCR kit (#N808-0179, Applied

Biosystems). For qRT-PCR, 100ng of RNA was used to detect Id3 and GAPDH. The two-step qRT-PCR was performed: cDNA synthesis (#K1651, Thermo Scientific) followed by PCR

(#172-5201, BioRad). Relative Id3 expression levels were normalized to GAPDH levels using 2-

ΔΔCt quantification method.

Primer sequences for genes we tested are as follows:

Elk-1 F: 5’-TGGGATGAGCCTGACCTAAC-3’

Elk-1 R: 5’-AAGGTCGTTCTCAAGCCAGA-3’

TBK-BP1 F: 5’-TCAACTTGCTGATGGAGACG-3’

TBK-BP-1 R: 5’-CTCCAGGTGGGAGTCAATGT-3’

MXD3 F: 5’-GCCTGGCTATGATGTTCCTC-3’

MXD3 R: 5’-CCATTCCAACAGGTGACTCC-3’

GAPDH F: 5’-GAAGGTGAAGGTCGGAGTC-3’

GAPDH R: 5’-GAAGATGGTGATGGGATTTC-3’

2.15. Soft Agar Growth Assay

A431/Id3 GFP cells were pre-treated with vehicle (Dimethyl Sulfoxide, DMSO), 20 µM caspase-8 inhibitor (Z-IETD-fmk), or pan-caspase inhibitor (Z-VAD-fmk) overnight. As a base layer, 500 µl of 0.6% low melting point agarose was added to 24-well plate. Subsequently, 103

A431/Id3 GFP cells were mixed with 250 µl of 0.3% agarose and added to the top of 0.6% agarose (DiVito et al., 2014). Four days after plating, colonies formed in soft agarose were visualized under Olympus epifluorescence microscope and counted manually in each group

(N>200 per group).

2.16. 5-FU and Cisplatin Treatment

3x104 A431/Id3 cells per well were plated into 6-well plate with 0.1% FBS-containing medium. The top 3 wells remained uninduced, while the bottom 3 wells were induced with tetracycline (1 µg/ml). 10 µM of 5-FU (#F6627, Sigma) and 20 µM of cisplatin (CDDP)

(#479306, Sigma) were used to treat A431 cells. 72 hours after drug treatment, Trypan Blue exclusion assay was used to quantify cell viability.

2.17. Statistical Analysis

Student’s 2-tail t tests were performed to examine differences between control and experimental groups. For all experiments, p values ≤0.05 are considered statistically significant, and are represented by one asterisk. P values ≤0.01 or p≤0.001 are represented by two or three asterisks, respectively. Error bars show standard deviation unless otherwise specified.

Chapter 3. Results

3.1. Establishment of A431/Id3 and A431/Vc Cell Lines

3.1.1. Endogenous Levels of Id3 are Down-regulated in A431 Cells

To understand the biological roles of Id3, we first examine endogenous Id3 protein levels in different stages of human keratinocytes in the progression of SCC, including primary human foreskin keratinocyte (HFK), two immortalized keratinocyte cell lines HFK/E6E7 and HaCaT, and four different malignant SCC cell lines: A431, SCC9, SCC25, and SCC4. HFK/E6E6 cells are immortalized by HPV 16 E6 and E7 viral proteins, resulting in constant activation of human telomerase reverse transcriptase (hTERT) and inactivation of pRb and p53 proteins. HaCaT is a spontaneously immortalized human cell line. These two are used as model cell lines to study precancerous AK lesions (Simbulan-Rosenthal et al., 2006; Simbulan-Rosenthal et al., 2005).

When comparing endogenous Id3 levels, there is an increase in Id3 expression when keratinocytes progress from normal cells (HFK) to pre-malignant cells (HaCaT). Id3 protein is then downregulated in most of the malignant SCC lines (Figure 24). Further, Id1 levels were also examined in these cells, since Id1 and Id3 have been shown to have overlapping functions

(Lyden et al., 1999). Primary keratinocytes have undetectable Id1. Id1 is upregulated in immortalized cells, especially in HaCaT. SCC4 and A431 both express comparable levels of

Id1, while SCC9 and SCC25 both have much lower levels of Id1.

Endogenous Id3 levels increase as primary keratinocytes progress to AK-mimicking cell line HaCaT, then decrease in 3/4 of the malignant SCC cells examined. Consistent with its role in inducing apoptosis in immortalized keratinocytes (Simbulan-Rosenthal et al., 2006) and as

56 tumor suppressor gene in Burkitt’s lymphoma, Id3 downregulation in malignant SCC cells is postulated to be necessary to support tumor development.

Figure 24. Endogenous Id3 and Id1 protein levels in different keratinocytes examined via immunoblot. All cells here were examined when grown to ~80% confluency. HFKs are normal keratinocytes. HFK/E6E7 and HaCaT cells are immortalized keratinocytes. SCC4, SCC25, SCC9 and A431 are malignant SCC cells. GAPDH serves as a loading control.

3.1.2. Verification of Inducible A431/Id3 Clones

Since endogenous Id3 is downregulated in some malignant SCC, to study effects of enforced Id3 expression, we created Tet-inducible cell lines in A431 cells. An inducible system was chosen to control for any genetic alterations that may occur during clonal selection. After transfection and selection, stable cell lines, pooled and individual clones, were extracted and subjected to immunoblot analysis using antisera specific for Id3 to determine protein expression in response to tetracycline (Tet). Id3 was induced by Tet (1 µg/ml) in all cell lines tested (Figure

25), and a robust induction of Id3 mRNA (7-8 fold) was confirmed by semi-quantitative reverse transcriptase-mediated PCR (RT-PCR) as well as by q-RT-PCR (Figure 26).

Figure 25. Id3 protein inducibility in A431/Id3 cell lines. Several stable inducible cell lines were created and subjected to immunoblots for Tet-inducible Id3 expression. One pool and 4 representative clones are shown. Cells were induced by Tet (1 µg/ml) for 24 hours. GAPDH serves as a loading control. Clone 5 was chosen to perform the subsequent experiments.

Figure 26. Verification of Id3 mRNA induction in A431/Id3 clone 5 cells via RT-PCR (left) and qRT-PCR (right). A431/Id3 cells were induced by Tet (1 µg/ml) for 24 hours and total RNA samples were subjected to RT-PCR (left panel) assays to test Id3 levels. Tet-induction time course (right panel) was tested via qRT-PCR to check for Id3 mRNA levels. Relative Id3 expression was normalized to GAPDH using Ct values. Asterisks represent statistically significant Id3 induction compared to that of uninduced sample (t=0).

Further, a Tet-induction time course was tested in both A431/Id3 clone 5 (cl5) and A431 vector control (A431/Vc) cell lines. There was no detectable Id3 expression in A431/Vc cells

58 throughout the time period tested (Figure 27). The induction of Id3 in the A431/Id3cl5 cell line can be seen as early as 2 hours after addition of Tet and a decrease after 16 hours. Normalized

Id3 protein expression levels relative to GAPDH loading control are shown in Figure 27.

Figure 27. A431/Id3 and A431/Vc Tet-induction time course. Id3 protein inducibility in A431/Id3 clone 5 (cl5) and A431/Vc cells were tested at different time points (0, 2, 4, 8, 16, and 24h) after addition of Tet (1 µg/ml) via immunoblot analysis (top panel). After cells were induced for 24h, Tet was removed for another 24h (48a) to assay the effect of Tet removal. GAPDH serves as a loading control, and relative Id3 expression to GAPDH is shown in the bar graph (bottom panel).

3.2. Study of the Effects of Id3 in vitro

3.2.1. Id3 Does Not Affect Cell Growth in Normal Serum, in Matrigel, or Invasion

Id3 has been postulated to alter growth in monolayers on Matrigel (by inducing stem cell markers and altering cadherin profiles) and to alter genes associated with invasion, such as matrix metalloproteinase-2 (MMP-2; Figure 8). We first determined the ability of cells to grow under each of these conditions, with no differences observed due to Id3. We examined cell

59 growth under normal serum (10% FBS) condition in response to Id3 induction. There is no significant change in either viable or dead cell numbers when Id3 is induced in A431 cells

(Figure 28A-B).

Figure 28. Id3 does not affect cell growth and death under 10% FBS condition. A431/Id3 cells were cultured in normal serum ±Tet (1 µg/ml) for 8 days. Viable (A) and dead (B) cells were counted after Trypan Blue staining to generate growth curves.

We also checked if Id3 induction alters distribution of cell population in different phases of cell cyle under normal serum conditions. Id3 does not appear to affect percentages of cells in any phases of the cell cycle (G1, S and G2/M) throughout the time points tested (Figure 29).

Figure 29. Id3 does not affect cell cycle distribution under 10% FBS condition. A431/Id3 cells were cultured in normal serum under ±Tet conditions (1 µg/ml) and harvested at different time points: 12, 24 and 48h for cell cycle analysis. Cell population distributions in G1, S, and G2/M phases of the cell cycle are shown.

Assessment of cell growth on Matrigel gel can be a way to determine the “stemness” of cancer cells. Id3 has been shown to promote colon cancer-initiating cell growth through downregulation of p21Cip1 (O'Brien et al., 2012). Therefore, we checked the ability of cells to grow on Matrigel in response to Id3 induction. We did not see a difference in cell growth on

Matrigel when Id3 is induced in A431 (Figure 30). Further, immunoblot analysis checking stem cell markers octamer-binding transcription factor 3/4 (Oct3/4) and sex determining region Y

(SRY)-box 2 (Sox2) showed no significant change in protein levels when Id3 is induced in

A431/Id3 cells (Figure 32), although we did observe some clonal difference in endogenous

Oct3/4 and Sox2 expression between A431/Vc and A431/Id3 cell lines.

Figure 30. Id3 does not affect cell growth on Matrigel under 10% FBS condition. GFP- expressing A431/Id3 (left panel) and Ds-Red-expressing A431/Vc (right panel) cells were cultured in normal serum ±Tet (1 µg/ml). Blue and red lines represent uninduced and induced cells, respectively. Cell growth was determined by fluorescent intensity using different filters: 485/535 (GFP) and 570/615 (Ds-Red).

We then examined the involvement of Id3 in invasion in A431 cells. Although Id3 is reported to repress invasion in oral SCC cells (Higashikawa et al., 2009), our data did not show difference in invasive ability when Id3 is induced in both clones tested (Figure 31). As we examined invasion markers E-cadherin and MMP-2, we observed only a slight increase in E- cadherin level, and MMP-2 was not significantly altered when Id3 is induced in A431/Id3 cell line (Figure 32).

Figure 31. Id3 does not affect invasion. A431/Id3 clones 5 (left panel) and 10 (right panel) cell lines were subjected to invasion assays ±Tet (1 µg/ml). Blue and red bars represent uninduced and induced conditions, respectively. Percentage areas occupied by invasive cells were quantified after Giemsa staining by ImageJ software.

Figure 32. Id3 does not significantly alter invasion and stem cell markers. A431/Vc and A431/Id3 cells were induced with Tet (1 µg/ml) for 0, 4, and 24 hours. In addition, Tet was removed after induction for 24 hours and cells stayed uninduced for another 24 hours (48a). SCC4 and mouse embryonic stem (mES) cells serves as positive controls for invasion markers and stem cell markers, respectively. GAPDH serves as a loading control. 63

3.2.2. Id3 Reduces Viable Cell Number and Induces sub-G1 Population in Low

Serum

Since Id3 did not affect cell growth in normal serum (Figure 28), we next checked cell growth in low serum since it has been reported that Ids affects growth under growth factor- deprived conditions (Florio et al., 1998; Norton and Atherton, 1998). A431/Id3 cells were cultured in low serum (0.1% FBS) conditions, and viable cell numbers were determined via

Trypan Blue exclusion assay. We observed that Tet induction of Id3 resulted in significantly fewer cells compared to uninduced cells (31% decrease in day 9, p=.02, and 56% decrease in day

13, p=.003; Figure 33). Cell cycle analysis was then performed to determine if the increase in cell number was concomitant with changes in the cell cycle. We found an increase in sub-G1 cell population in Tet-induced cells (40% by day 13, p=.02; Figure 35), but no changes in S phase (Figure 35), nor in other phases of the cell cycle before and after Id3 induction (Figure 36).

These results indicate that Id3 contributes to the sub-G1 population and A431 cell death under low serum conditions. The growth curve of A431 Vc cells was also assayed in the same manner and no difference in cell number was observed, indicating that tetracycline did not alter cell cycle or cell death (Figure 34).

Figure 33. A431/Id3 growth curves in low serum condition. A431/Id3 cells were grown under 0.1% FBS and viable cell numbers were counted on different days (1, 3, 5, 7, 9, 11, and 13) under ±Tet conditions (1 µg/ml) via Trypan Blue exclusion assay. Asterisks represent statistically significant differences between uninduced cells and Tet-induced cells expressing Id3.

Figure 34. A431/Vc growth curves in low serum condition. A431/Vc cells were grown under 0.1% FBS and viable cell numbers were counted on different days (1, 3, 5, 7, 9, 11, and 13) under ±Tet conditions (1 µg/ml) via Trypan Blue exclusion assay.

Figure 35. Cell cycle analysis for A431/Id3 cells in low serum. A431/Id3 cells in 0.1% FBS were assayed via FACS for percentages of cells in different phases of cell cycle (sub-G1, G1, S and G2/M) under ±Tet (1 µg/ml) conditions. Asterisks represent statistically significant differences between ±Tet conditions.

3.2.3. Id3-induced Cell Death is not through a Paracrine Signaling Cascade

To investigate if Id3 induced apoptosis of A431 through a paracrine signaling cascade, we co-cultured GFP-expressing A431/Id3 and Ds-Red-expressing A431/Vc cells in the same wells and examined cell growth under uninduced and induced conditions (Materials and

Methods; Figure 36A). Student’s t tests revealed no statistically significant difference in cell number between the uninduced A431/Vc, induced A431/Vc, and uninduced A431/Id3 cells under co-culture conditions. Thus, co-culture with induced A431/Id3 cells did not alter proliferation of A431/Vc cells (Figure 36B), indicating that changes in cell numbers were not due to cell-cell communication or secreted factors.

Figure 36. A431/Id3 and A431/Vc co-culture growth curves in low serum. (A) GFP-expressing A431/Id3 cells were co-cultured with Ds-Red-expressing A431/Vc cells under ±Tet (1 µg/ml) conditions. (B) Cells amount were quantified on different days (1, 3, 5, 7, and 9) through measuring percentages of area occupied by fluorescent cells using ImageJ software. Statistical analysis shown here compares A431/Id3 ±Tet (1 µg/ml) conditions.

3.2.4. Id3 Induces Caspase-3 and Caspase-9 Proteolytic Cleavages

Since Id3 induces a Caspase-3- and -9-dependent apoptosis in HaCaT cells (Simbulan-

Rosenthal et al., 2006), to identify the form of A431 cell death induced by Id3, we performed immunoblot analysis to determine whether Caspase-3 and -9 were proteolytically processed from procaspases into their smaller, active forms. We observed that upon Id3 induction, active

Caspase-3 and -9 were clearly observed (Figure 37). This result indicates that Id3 induction is sufficient to induce apoptosis in A431 cells involving Caspase-3 and -9.

Figure 37. Caspase-3 and -9 proteolytic cleavage under Id3 induction. A431/Id3 ±Tet (1 µg/ml) cell lysate was subjected to immunoblot to detect Caspase-3 and -9 proteolytic products. GAPDH serves as a loading control.

3.2.5. Microarray Analysis Revealed Elk-1 as a Downstream Target of Id3

To investigate the molecular pathways by which Id3 mediates caspase-3 activation and apoptosis in A431 cells, microarray analysis was performed. Uninduced and induced A431/Id3 and A431/Vc cells were subjected to microarray analysis. A431/Id3 altered expression levels of

16 genes greater than 1.5 fold when induced, compared with the other 3 groups (Figure 38).

Figure 38. Heatmap showing genes that are significantly up- and down-regulated by Id3. A431/Id3 and A431/Vc cells under ±Tet (1 µg/ml for 24h) conditions were subjected to microarray analysis. Heatmap generated shows genes that are significantly altered by Id3 by at least 1.5 fold. Blue represents down-regulation and red represents up-regulation.

Based on our data, TANK binding kinase 1 binding protein 1 (TBK-BP1) is one of the genes that is altered by Id3 in microarray. This adaptor protein binds to TBK1 and is part of the

NFκB network. TBK1 is an IκB kinase (IKK) (Pomerantz and Baltimore, 1999). IκB kinase has been shown to participate in TGF-β pathway to activate ΔNp63 expression and therefore suppress SCC invasion through down regulation of E-cadherin or MMP-2 (Higashikawa et al.,

2009). IκB kinase α acts as a co-Smad in TGF-β pathway in keratinocytes (Fukunishi et al.,

2010). ΔNp63 transcription is activated by TGF-β pathway, through Smad2 and IκB kinase α.

Therefore, TBK-BP1 might also play a role in SCC progression and therefore is considered a good candidate to investigate further.

Chloride intracellular channel 4 (CLIC4) is also significantly altered by Id3 based on microarray analysis. CLIC4 is suggested to act as a tumor suppressor for SCC (Suh et al., 2012). p53- and c-Myc-induced CLIC4 protein expression has been shown to induce apoptosis in primary mouse keratinocytes and rat fibroblasts (Fernandez-Salas et al., 2002; Shiio et al., 2006).

CLIC4 expression is reduced and excluded from nucleus in SCC cell lines, consistent with results in skin cancer tissue microarrays (Suh et al., 2012). Enforced expression of CLIC4 results in smaller mouse xenografts (most likely through induction of apoptosis) and re- sensitizing SCC cells to TGF-β growth inhibitory signaling (Suh et al., 2012).

Elk-1 and MXD3 are also upregulated by Id3 based on microarray analysis. Elk-1 has been shown to interact with Id3 and promote apoptosis in immortalized human keratinocytes.

Max-associated protein 3 (MXD3 or MAD3) is a family IV bHLH, which competes with transcription factor Max to bind Myc and represses transcriptional activation activity of Myc.

MXD3 can therefore be involved in proliferation and apoptosis.

We verified expression levels of candidate genes by analyzing mRNA and protein levels via RT-PCR and immunoblot analysis, respectively. TBK-BP1, Elk-1, and MXD3 were upregulated at the mRNA (Figure 39) and protein (Figure 40) levels in induced samples compared with those that were uninduced.

Figure 39. Verification of genes that are significantly altered by Id3 via RT-PCR. Genes that were shown to be significantly altered by Id3 in microarray analysis were verified for their mRNA expression in response to Id3 induction. A431/Id3 cells were induced by Tet (1 µg/ml) for different times (0, 4, 8, 12, and 24h) and total RNAs were extracted and assayed. GAPDH serves as a loading control.

Figure 40. Verification of genes that are significantly altered by Id3 via immunoblot. Genes that were shown to be significantly altered by Id3 in microarray analysis were verified for their protein expression in response to Id3 induction. A431/Id3 cells were induced by Tet (1 µg/ml) for different times (0, 4, 8, 16, and 24h) and cell lysate was extracted and assayed. GAPDH serves as a loading control.

3.2.6. Id3 Induces Apoptosis through Elk-1-Caspase-8-Dependent Pathway

Of the identified candidate genes, Elk-1 was the only gene product that has been shown to physically interact with Id proteins (Yates et al., 1999). Furthermore, Elk-1 mediates sodium arsenite-induced apoptosis in HaCaT cells, at least in part, through induction of induction of bax

(Shin et al., 2011), consistent with the role of Id3 in HaCaT (Trabosh et al., 2009a). Elk-1 is also involved in expression of Death Receptor 5 (DR5) in response to Celecoxib in lung cancer cells

(Oh et al., 2010). Supporting its role in apoptosis, promoter analysis identifies Sp-1- and ETS- binding sites within caspase-8 promoter region, which are essential sites for caspase-8 transcription (Liedtke et al., 2003). In addition, Sp-1 and Elk-1 have been shown to bind to the caspase-8 promoter and induce apoptosis in colorectal cancer cells in response to Secreted

Protein, Acidic and Rich in Cysteine (SPARC) treatment (Tang et al., 2009). We therefore investigated the potential involvement of Elk-1 and caspase-8 in Id3-induced apoptosis in A431 cells.

Upon Id3 induction with Tet, A431 cells treated with scrambled siRNA showed elevated levels of Elk-1 protein compared to the Tet- uninduced cells (Figure 41). Id3 also induced procaspase-8 as well as active caspase-8 protein expression (Figure 41). When Tet-induced

A431 cells were transfected with siRNA specific for the Elk-1 transcript for 48 and 72 hours, Id3 induction of Elk-1, procaspase-8, and active caspase-8 were all suppressed (Figure 41). This suggests an apoptotic pathway leading from Id3 to Elk1 to procaspase-8 induction and activation in A431 cells. Furthermore, Bid1 is cleaved in the presence of scrambled, but not Elk1siRNA, suggesting that caspase-8 may activate Bid and a mitochondrial pathway of apoptosis, characteristic of type II apoptotic cells. Interestingly, unlike in HaCaT cells, this Id3-induced apoptotic pathway in A431 is not due to elevated levels of Bax protein expression (Figure 41).

Figure 41. Id3 induces Elk-1 and Caspase-8 expression. A431/Id3 cells were treated with scrambled (left panel) or Elk-1 siRNAs (right panel) for 48h and 72h under ±Tet (1 µg/ml) conditions. Cell lysate was extracted from each conditions and subjected to immunoblots to check for Id3, Elk-1, Pro-caspase-8, cleaved caspase-8 Bax and Bid proteins. GAPDH serves as a loading control.

Since Caspase-8 is the initiator caspase for death receptor-mediated apoptosis pathway, we next examined the level of death receptor protein in response to Id3 induction. We found that in response to Id3 induction, Fas receptor protein level increased in A431 cells (Figure 42). This indicates that Id3 may induce apoptosis in A431 cells through both increasing Fas receptor protein and Caspase-8 protein expression (Figure 8).

Figure 42. Induction of Fas expression by Id3 in vitro. A431/Id3 cells were induced with Tet (1 µg/ml) for different time points (4, 8, and 24 hours). 24 hours after induction, Tet was removed for 24 hours (48a). β-actin serves as a loading control.

3.2.7. Caspase Inhibitors Abolished Id3-induced Decreases in Cell Numbers in

Low Serum, and Colony Formation in Soft Agar

To further investigate the involvement of Id3 in apoptosis, we treated cells with caspase inhibitors, including the caspase-8 inhibitor, Z-IETD-fmk, and pan-caspase inhibitor, Z-VAD- fmk, and examined cell growth under different conditions. While treating A431/Id3 cells with vehicle only (DMSO) showed decreased cell numbers (Figure 43), both caspase inhibitors abolished the difference in cell number induced by Id3 (Figure 44). This indicates that Id3- induced decrease in cell number under low serum is caspase-dependent.

Figure 43. A431/Id3 cell growth curves in the presence of DMSO. A431/Id3 ±Tet (1 µg/ml) cells were cultured under DMSO in 0.1% FBS and cell amount was quantified via measuring percentages of area occupied by cells using ImageJ software. Asterisks represent statistically significant differences between ±Tet groups.

Figure 44. A431/Id3 cell growth curves in the presence of caspase inhibitors. A431/Id3 ±Tet (1 µg/ml) cells were cultured under two caspase inhibitors (A)Z-IETD-fmk and (B)Z-VAD-fmk in 0.1% FBS and cell amount was quantified via measuring percentages of area occupied by cells using ImageJ software.

We next investigated the ability of A431 cells to form colonies in soft agar in the presence and absence of Id3 to examine cell anchorage-independence growth. We observed a significant decrease (33%, p=.003) in soft agar colonies when Id3 was induced (Figure 45A).

However, after cells were treated with caspase inhibitors, either caspase-8 inhibitor (Z-IETD- fmk) or pan-caspase inhibitor (Z-VAD-fmk), the decrease in colony formation due to Id3 induction no longer existed (Figure 45A), indicating that Id3 reduction of soft-agar colonies is caspase-dependent. Representative pictures of soft agar colonies are shown in Figure 45B.

Figure 45. Id3-reduced colony formation in soft agar is caspase-dependent. GFP-expressing A431/Id3 cells were grown in soft agar under ±Tet conditions. Cells were treated either with vehicle (DMSO), caspase-8 inhibitor (Z-IETD-fmk) or pan caspase inhibitor (Z-VAD-fmk). (A) 4 days after Tet induction, percentage of cells formed colony were counted manually (N>200 in each group). (B) Representative colonies formed in each group are shown under ±Tet conditions.

3.2.8. Down-regulation of Endogenous Id3 Repressed Caspase-8 Activation in

Other SCC Lines

To further investigate the induction of caspase-8 activation by Id3, we transfected other

SCC lines, including SCC4, SCC9, and SCC25, with either scrambled siRNA as control or siRNA against the endogenous Id3 transcript. 24 hours after transfection, cells were harvested and subjected to immunoblot analysis. We observed that endogenous Id3 was successfully knocked down in all SCC lines tested (Figure 46). We also found that levels of cleaved caspase-

8 decreased in SCC cells (Figure 46) especially in SCC25 (78% decrease); SCC4 cells also show a much smaller (17%) but significant and reproducible decrease. Our results suggest that Id3 can induce apoptosis via a caspase-8-dependent pathway in other SCC lines but not all.

Figure 46. Knocking down endogenous Id3 in other SCC reduced proteolytic cleavage of Casp- 8. Endogenous Id3 levels were knocked down in SCC4, SCC9, and SCC25 cells (left to right) and levels of cleaved caspase-8 decreased by: 74% in SCC25, 38% in SCC9 and 17% in SCC4, when quantified by densitometry. GAPDH serves as a loading control.

3.2.9. Id3-induced Apoptosis is C-terminus and HLH-dependent

To investigate the domains of Id3 that are responsible for apoptosis induction, four inducible cell lines expressing Id3 deletion mutants were created (Figure 47). Id3 mutant cloning (sequence verification) was performed using PCR amplification employing primers as described in Materials and Methods. Induction of Id3 mutants by tetracycline was verified by

RT-PCR (Figure 48).

Figure 47. Schematic representation of Id3 deletions of the C-terminus (ΔC38), N-terminus (ΔN39), C+HLH (ΔC80) domain, or N+HLH (ΔN81) domain.

Figure 48. Verification of Id3 mutant induction by Tet (1 µg/ml for 24 hours) in A431 cells. 4 cell lines were verified for inducibility of Id3 mutants by Tet. Representative clones are shown here. GAPDH serves as a loading control.

A431 cells expressing different Id3 mutants were then cultured under low serum condition, and their growth curves were examined via Trypan Blue exclusion assays (Figure 49). 78

We observed that deleting the N-terminus of Id3 (ΔN39) did not abolish Id3-induced apoptosis

(Figure 49). Other Id3 mutants corresponding to deletions of the HLH, C-terminus, or both, including ΔC38 (Figure 49), ΔC80 (Figure 49), and ΔN81 (Figure 49), all lost their ability to reduce cell numbers in the growth curve experiments. These correspond to the deletion profile observed for Id3 in Burkitt’s lymphoma (Figure 9), suggesting that similar regions of Id3 contribute to tumor suppression in both SCC and Burkitt’s lymphoma.

Figure 49. Growth curves for A431/Id3 mutant cell lines in low serum. 4 different A431/Id3 mutant cell lines were grown under low serum in ±Tet (1 µg/ml) conditions. Viable cell numbers were counted after Trypan Blue staining very two days. Asterisks represent statistically significant differences between ±Tet conditions. Cell cycle analyses were then performed to verify these results. Id3 ΔN39 showed a drastic increase in sub-G1 population (Figure 50, top right), while other mutants showed no

79 significant difference in sub-G1 population (Figure 50). These results indicate that HLH and C- terminal domain of Id3 gene is important for induction of apoptosis in A431 cells. It is worth mentioning that ΔN39 mutant alters all phases of the cell cycle.

Figure 50. Cell cycle analyses for A431/Id3 mutant cell lines in low serum. 4 different A431/Id3 mutant cell lines were grown under low serum in ±Tet (1 µg/ml) conditions and assayed via FACS for percentages of cells in different phases of cell cycle: sub-G1, G1, S and G2/M. Asterisks represent statistically significant differences between ±Tet conditions.

3.2.10. Id3 Sensitizes SCC Cells to Chemotherapeutic Drugs

We then investigated if Id3 can be used in combination with other chemotherapeutic drugs to cause apoptosis of cancer cells. Id3 levels are elevated in different cancer cell lines following cisplatin (CDDP) and 5-FU treatments (Koyama et al., 2004; Troester et al., 2004), and Id3 has further been shown to mediate cisplatin-induced apoptosis in sarcoma cells (Koyama et al., 2004). Further, 5-FU and cisplatin are shown to be EGFR inhibitors (O'Bryan and Ratner,

2011), and A431 cells express high levels of EGFR, contributing to their highly proliferative property (Licitra et al., 2011). We therefore investigated the effect of Id3 induction on cell death in A431 cells in response to 5-FU and CDDP. We found that Id3 sensitized A431 cells to CDDP

(by 41%, p<.001) and 5-FU (by 34%, p=.02) treatments (Figure 51). These results indicate that

Id3 can be used in combination with other drugs to increase treatment efficacy.

Figure 51. Id3 sensitizes A431 cells to CDDP and 5-FU treatments. A431/Id3 and A431/Vc cells were treated with either CDDP (left panel) or 5-FU (right panel) under Tet-induced condition in low serum. 72 hours after drug treatment, cell viability was determined in each group via Trypan Blue exclusion assay.

3.3. Effects of Id3 in vivo

3.3.1. Id3 induction decreased xenograft sizes in vivo

To investigate the ability of Id3 to suppress tumor growth in vivo, we injected GFP- expressing A431 cells subcutaneously into mice and examined tumor growth over time.

Induction of Id3 in vivo was achieved through doxycycline (Dox)-containing feed and tumor growth was tracked via in vivo imaging of GFP-expressing cells (Materials and Methods; Figure

52). Id3 significantly reduces tumor size (≈33%, p=.002) (Figure 53). Ds-Red-expressing A431

Vc cells were also injected into mice, and tumor sizes were monitored for 30 days. There was no difference in the size of tumors fed with Dox diet compared to those fed with regular diet (Figure

54).

Figure 52. Mouse xenografts formed from GFP-expressing A431/Id3 cells. Tumors from A431/Id3 cells under ±Dox conditions were imaged in vivo on different days.

Figure 53. Id3 reduces xenograft sizes in vivo. Size of xenograft formed from A431/Id3 ±Dox cells were determined using in-vivo imaging system (Materials and Methods) based tumor area covered by GFP fluorescent signal. Results are expressed as mean ± SE. –Dox: N=10; +Dox: N=14.

Figure 54. A431/Vc xenografts do not show difference in size under ±Dox conditions. Size of xenograft formed from A431/Vc ±Dox cells were determined using in-vivo imaging system (Materials and Methods) based tumor area covered by Ds-Red fluorescent signal. Results are expressed as mean ± SE. –Dox: N=5; +Dox: N=5.

Since Id3 induces Elk-1 in cultured cells (Figure 40), we investigated whether Id3 also induces Elk-1 in xenografts. Immunoblot analysis verified Id3 induction by Dox-containing feed and showed that Elk-1 expression is induced in tumors with Id3 induction (Figure 55), and Elk-1 is upregulated in tumors with high level of endogenous Id3 (Figure 55, lanes 5 and 6).

Figure 55. Id3 induces Elk-1 expression in vivo. Id3 and Elk-1 protein levels were checked in xenografts formed from A431/Id3 ±Dox conditions. GAPDH serves as a loading control.

Immunofluorescent staining was then performed to investigate if the reduced sizes of Id3 xenografts are due to a decreased proliferation or increased apoptosis. We observed a 50% increase in apoptosis marker active caspase-3 in tumor sections expressing high amount of Id3

(p=.01; Figure 56). On the other hand, expression of the proliferation marker Ki-67 did not change upon Id3 induction (Figure 56). Our data suggest that Id3 induces apoptosis in A431 xenografts. Further, Fas was also upregulated in the Id3-induced tumors (Figure 57) as it was in culture cells (Figure 42), suggesting one mechanism for the activation of procaspase-8, in the

Id3-Elk1-caspase-8 axis demonstrated in vivo.

Figure 56. Active Caspase-3 is upregulated in A431/Id3 xenograft sections. A431/Id3 ±Dox tumor sections were stained with active caspase-3 (top row) and Ki-67 antibodies (bottom rows). Active caspase-3 staining can be seen in areas with sparse cells. Ki-67 reduction can be observed, although is not statistically significant. Low Id3: N=6; high Id3: N=3.

Figure 57. Id3 induces Fas expression in vivo. Fas protein levels were checked in xenografts formed from A431/Id3 ±Dox conditions.

Chapter 4. Discussion

Using a Tet-inducible system, we have demonstrated that Id3 induces apoptosis in the malignant SCC cell line A431 both in vitro and in vivo. In cell culture, we showed that Id3 induction reduced cell number under low serum conditions (Figure 33), accompanied by an increase in sub-G1 population, while no change was observed in S phase population (Figure 35).

Cell colonies were also diminished in soft agar (Figure 45). Under co-culture conditions, it is also shown that the decrease in cell number after Id3 induction is mostly through a cell autonomous signaling cascade (Figure 36). In the mouse xenograft model, we successfully induced Id3 by feeding animals with Dox and showed that Id3 induction significantly decreased tumor size by more than 30 percent (Figure 53). This is due to an increase in cell death rather than a decrease in cell proliferation, as determined by immunofluorescent analysis of active caspase-3 and Ki67, respectively (Figure 56). Furthermore, we have shown that Id3 induced the stress response gene Elk-1 in response to low serum in culture (Figure 41), as well as in xenografts at the mRNA and protein levels (Figure 55). This signaling cascade results in the proteolytic activation of caspases-8 and -3, leading to cell death (Figure 41). As revealed by our

Id3 mutant analyses, this pathway is dependent upon both C-terminal and HLH domains of Id3

(Figure 49).

The mechanism by which Id3 modulates apoptosis in malignant keratinocytes appears to be different from that of non-transformed keratinocytes. In premalignant cells, we have previously shown that Id3 induces bax expression at the promoter level (Simbulan-Rosenthal et al., 2006) . On the other hand, ectopic expression of Id3 alone did not elevate Bax protein levels in A431 cells (Figure 32). Id3-mediated upregulation in Bax level is therefore cell-context

86 dependent. We further investigated other potential apoptosis-inducing pathways to understand the mechanism by which Id3 mediates cell death in A431 cells.

Our current study demonstrates that Id3-induced apoptosis is Elk-1-dependent. It has been shown that Ternary Complex Factor, including Elk-1, and Serum Response Factor form a protein complex that binds to Serum Response Element at promoter region of c-fos in response to MAPK signaling to induce G1-S progression (Yates et al., 1999). Further, these investigators found that Id3 inhibits Elk-1 by sequestering it from the SRE and disrupted the G1-S transition

(Stinson et al., 2003; Yates et al., 1999). Interaction between TCFs and SRFs greatly increase the binding affinity to c-fos promoter. Ids can compete with SRFs to bind TCFs and therefore inhibit transcriptional activation of c-fos. Conversely, Id proteins do not inhibit TCFs from binding to ETS-motif of other transcription factors, such as E74 (Yates et al., 1999). This can be due to the difference in TCF binding affinity to ETS motif. TCFs bind with higher affinity to

ETS motif of E74 promoter than to c-fos promoter. This indicates that Ids can exert different effects on TCFs under different circumstances. For the first time, the role of Id3 in the induction of elk-1 was examined, and we demonstrated that Id3 induces Elk-1 expression in vitro and in vivo. Further, this induction of Elk-1 protein induced apoptosis under stress condition in A431 cells.

We further demonstrated that deletion of 39 amino acids from the N-terminus of Id3 renders it more potent as an inducer of apoptosis (Figure 50). The mechanism remains to be investigated, although this might be due to the loss of the Cdk2 phosphorylation site serine 5 at the N-terminus of this mutant. Phosphorylation of Id2 and Id3 by Cyclin A/E/Cdk2 complexes has been shown to abolish the interaction between Id and TCF proteins (Stinson et al., 2003).

Deleting this phosphorylation site from Id3 protein may allow Id3 to escape Cdk2 regulation and

87 thus interact with Elk-1 with greater affinity. Replacing serine 5 with alanine prevents Id3 phosphorylation at this site and ectopic expression of this mutant Id3 leads to apoptosis (Deed et al., 1997), consistent with our results on ∆N39 mutant.

On the other hand, Id3 mutants lacking either the C terminus, the HLH, or both, failed to induce apoptosis. The HLH and C-terminal regions have been shown to have very important functions for Id3. HLH is a protein-protein interaction domain, which allows for binding of Ids to other transcription factors and therefore plays a critical role for Id proteins. In the C-terminus, there is a single nucleotide polymorphism (SNP) site, A105T, which gives rise to an atherosclerosis risk allele (Doran et al., 2010). The Id3A105T protein has reduced binding affinity to the E12 protein as well as reduced ability to modulate E12 transcriptional activity. It is possible that deleting only amino acid 105 will lead to induction in E12 transcriptional activity and therefore affect the role of Id3 in apoptosis. A similar mechanism is seen in Burkitt’s lymphoma where tumor cells expressing inactivating Id3 mutants (mostly HLH and C-terminal mutants) fail to inhibit E proteins and gain a growth advantage (Schmitz et al., 2012).

Given the similarity between the inactivated engineered mutants, and the Id3 deletions found to inactivate Id3 in Burkitt’s lymphomagenesis (Figure 9), we postulate a similar mechanism has the potential to play a role in SCC progression. Verification of this hypothesis is pending in future in-vivo studies, as well as analysis of additional SCC patient DNA. Further, in- vitro investigation into binding partners of each Id3 truncated mutant via pull-down assays (Co-

IP) can provide mechanistic explanation of Id3-induced apoptosis. Proteins binding only to C- terminus and HLH regions can be the important players for Id3-induced apoptosis. Proteins bound only to N-terminus of Id3 can help us understand why Id3 N-terminal deletion mutants

88 render Id3 a more potent apoptosis inducer. Point mutations can also be generated to localize the position(s) of Id3 that are crucial for its tumor suppressor role.

In the current study, different Id3 deletion mutants are used to examine regions of Id3 important for inducing apoptosis. Since Id3 is a small protein, its truncation mutants might not fold properly, in spite of the presence of the Flag tag that increases protein size. The structures of these deletion mutants are currently unsolved; however, partial sequence of Id3, between amino acids 27 to 83 (partial N-terminus plus HLH domain), have been shown to fold properly by Nuclear Magnetic Resonance (NMR) structure solution (PDB: 2LFH). To gain a better understanding of potential structures of Id3 mutants, an overlap of partial Id3 and MyoD structures (PDB: 1MDY) is generated using Pymol software (

Figure 58). Based on our prediction, N-terminal domain of Id3 folds into loop and alpha helix structures, and the C-terminal regions folds into alpha helix structure. Whether these Id3 mutant proteins can fold properly in vivo remains to be investigated.

Figure 58. Overlap of Id3 and MyoD protein structures. MyoD (molecule in cyan) structure is overlapped with partial Id3 structure (molecule in blue, orange and red). N-terminal region of Id3 is labeled in blue, its HLH region is labeled in orange, and its C-terminal region is labeled in red.

Id3 siRNA reduced caspase-8 levels in SCC4, SCC9, and SCC25, which all harbor p53 mutations in their DNA binding domains. SCC4 and SCC25 have one p53 allele each (Caamano et al., 1993), which are a p53P151S gain-of-function (GOF; SCC4) that promotes cell proliferation and xenograft growth in prostate cancer line LNCaP, and the head and neck SCC line, UM-SCC-

1 (Sano et al., 2011; Tepper et al., 2005), and a p53R209fs 2-bp frameshift (SCC25; (Caamano et al., 1993). SCC9 has a p53274fs 32-bp frameshift. SCC4 overexpresses the GOF p53, while p53 is undetectable in SCC9 and SCC25 (Caamano et al., 1993; Min et al., 1994; Skvortsov et al.,

2007; St John et al., 2000). The loss of p53 expression did not prevent Id3 siRNA-mediated

90 reduction in caspase-8 activation, especially in SCC25 cells, suggesting that Id3-mediated activation of caspase-8 is most likely p53-independent (Figure 46).

A431 also has only one p53 allele harboring a GOF p53R273H missense mutation in the

DNA binding domain (Reiss et al., 1992). This GOF mutation has been shown to affect gene transcription of Id2 and Id4 but not that of Id1 or Id3. The Id2 promoter has been shown to be p53R273H– but not p53WT-responsive in SW480 colon cancer cells (Yan et al., 2008). Knocking down of endogenous p53R273 protein induces Id2 expression, but not Id3 (personal communication with Dr. Yan), and knocking down of Id2 restores the proliferative phenotype of these p53 mutant-null cells (Yan et al., 2008). It is therefore concluded that suppression of Id2 is required for the gain-of-function property of p53R273H mutant. This indicates that Id2 plays a role to inhibit tumor progression. Id4 is also induced by p53R273H, but not wild-type p53 in lung carcinoma cell line H1299 (Fontemaggi et al., 2009). Further, induction of Id4 by p53R273H induces angiogenesis-promoting cytokines and proliferation of endothelial cells, indicating that

Id4 promotes tumor progression in these cells. In the same study, Id1, Id2, and Id3 expression levels were not altered by p53R273H in H1299 cells. In A431 cells, p53R273H down-regulates procaspase-3 and increases resistance to chemotherapeutics (Wong et al., 2007). In contrast to p53R273H, Id3 induces apoptosis in A431 cells and sensitized these cells to cisplatin and 5-FU.

Therefore, Id3-induced apoptosis in A431 is most likely p53-independent; yet knockdown of mutant p53 is necessary to answer the question definitively.

To test if CDDP or 5-FU and Id3 has synergistic effect on A431 cells, CompuSyn software was used to calculate the combination index (CI) based on the Chou-Talalay drug synergy quantification model. We tested different ratios of Id3/CDDP or Id3/5-FU, ranging from 0.01 to 20000, and calculated CI values for each ratio. For both drugs, the CI values for

90% population death and 50% population death are all smaller than 1, which is an indication of synergistic effects (Table 5).

Table 5. Combination index for Id3 and CDDP or 5-FU dual treatments

Id3/CDDP CI for 90% CI for 50% Id3/5-FU CI for 90% CI for 50% Ratio Death Death Ratio Death Death 1/2000 1.20E-4 0.00279 1/1000 Infinity Infinity 1/20 1.22E-04 0.00278 1/10 2.99E-7 5.11E-4 ½ 1.39E-04 0.0027 1 4.36E-7 4.69E-4 1 1.54E-04 0.00263 2 5.64E-7 4.42E-4 5/2 1.88E-4 0.00252 5 8.74E-7 3.99E-4 10 2.84E-4 0.00229 20 1.93E-6 3.32E-4 100 6.28E-4 0.00192 200 8.03E-6 2.39E-4 1000 0.00143 0.00159 2000 8.00E-6 2.39E-4

Moving this research forward, some experiments can be performed to strengthen our findings. First, in Figure 36, GFP-expressing A431/Id3 and Ds-Red-expressing A431/Vc cells were co-cultured together and cell growth curves were generated to study the cell signaling cascade mediated by Id3 induction. In the 9-day growth curves, cell lines without Id3 induction

(Vc no Tet, Vc Tet, and Id3 no Tet) were not showing statistically significant differences.

However, there was a non-significant decrease in the Vc Tet group compared to that of the other two groups (day 9). This might be an indication that although Id3-mediated apoptotic pathways are mostly cell autonomous, extrinsic cellular signaling pathways can also be involved. To further understand the involvement of extrinsic apoptotic pathways, we can perform assays to detect release of cytochrome c from the mitochondrial via immunoblot analysis. In addition, since Fas receptor protein levels were up-regulated when Id3 was induced (Figure 42), an extrinsic apoptosis pathway might be involved through Fas. To test this hypothesis, anti-Fas neutralizing antibody can be used to determine if Id3 apoptosis is dependent on this death receptor pathway. 92

We demonstrated that Id3 induces Elk-1 and caspase-8 protein expression in A431 cells via immunoblot analysis (Figure 41). We can further examine whether Id3 binds Elk-1 and caspase-8 promoters via chromatin immunoprecipitation (ChIP). The regions of promoters that are bound by Id3 can also be investigated by generating different promoter constructs.

Moreover, in our immunoblot analysis, there is a reduction of full-length Bid protein when Id3 is induced (Figure 41). The cleavage of Bid can be verified using the antibody that is specific for cleaved Bid (tBid).

We have shown that Id3 induces the proteolytic cleavage of caspase-3 (Figure 37). To verify caspase-3 activity induced by Id3, in-vitro kinetic caspase-3 activity assays can be performed by incubating caspase-3 substrate with A431 cell lysate. To further investigate the involvement of caspases in Id3-mediated cell death, we performed growth curve analyses after cells were treated with caspase inhibitors, as shown in Figure 44. To strengthen the data, we can further perform additional apoptosis assays, including Annexin V staining, cell cycle analysis of the sub-G1 population, and terminal deoxynucleotidyl transferase nick end labeling (TUNEL) staining.

Other Id proteins show 40% sequence similarity to Id3 (Figure 2); therefore, they might have overlapping functions in induction of apoptosis in SCC cells. In addition, Id1, Id2, and Id3 were all shown to bind Elk-1 (Yates et al., 1999). To further study if other Id proteins are also involved in this apoptosis pathway in SCC, RNA interference techniques can be used to knock down other Ids one at a time and study apoptosis. Our Tet-inducible system can also be used to ectopically express other Id proteins in A431 cells and see if they can function as tumor suppressors.

Tet-inducible systems will allow us to remove tetracycline or doxycycline and observe if the apoptotic phenotypes can be reverted in the future. It is also possible to investigate to effects of the timing of Id3 induction on apoptosis. Whether Id3 induction is necessary to kill tumor cells only at the initial stage of xenograft formation or throughout the experiment can be tested by feeding animals with Dox at different times. Levels of Id3 induction can also be adjusted by using different amount of tetracycline of doxycycline to examine the necessary amount of Id3 needed to function as a tumor suppressor in SCC cells.

In addition, different SCC lines can also be tested in vitro and in vivo to see if Id3 induces apoptosis in other SCC cell lines. For example, in SCC that have higher endogenous levels of

Id3 (e.g., SCC25), Id3 can be knocked down and cell culture apoptosis as well as xenografts assays can be performed. Microarray analyses can also be performed on a panel of SCC lines to find common genes that are regulated by Id3 and establish a common mechanism by which Id3 modulates apoptosis in SCC.

Id3 levels in cancer have been shown to be study- and/or cell type-specific. Based on cBioPortal database looking at Id3 alterations in different types of cancers, Id3 is amplified in 19 studies (40%) deleted in 17 studies (36%), and mutated in 11 studies (23%). Analysis of the

Oncomine® database revealed that Id3 was significantly upregulated in 83 studies and downregulated in 84 studies in which cancers were compared to normal samples (Figure 59). In skin cancers, Id3 mRNA level is significantly downregulated 2 to 5-fold in all melanomas studied

(Riker et al., 2008; Talantov et al., 2005). In Basal Cell Carcinoma, Id3 mRNA is significantly lower than normal skin by 2.3 fold (Riker et al., 2008). However, in Head and Neck and

Hypopharyngeal SCC, Id3 mRNA is upregulated 3-fold compared to normal skin (Ginos et al.,

2004; Schlingemann et al., 2005). In cutaneous SCC, it is reported that Id3 mRNA is

94 downregulated by 1.1-1.5 fold, compared to normal skin; however, the differences were not statistically significant (Nindl et al., 2006; Riker et al., 2008).

Figure 59. Oncomine database showing Id3 expression levels in normal vs. cancer samples. Red represents upregulation of Id3 and blue represents downregulation of Id3. Different color intensity represents the best gene rank percentile for the analyses in each square (for example, blue 10% means Id3 is the top 10% genes to be significantly downregulated in cancer). Number in each square shows number of studies meet thresholds: p-values≤0.05, fold change≥1.5. Figure from (OncomineTM, 2014); access date: 06-18-14.

In spite of the fact that Id3 levels vary in cancer cells, many studies have designed therapeutic strategies using Id3 as a target. In this study, we have demonstrated that Id3 sensitizes A431 cells to the chemotherapeutic agents cisplatin and 5-FU. 5-FU has been reported

95 to induce Id1 and Id3 mRNA levels in breast cancer cells (Barrett, 2004). Id3 has also been shown to sensitize sarcoma cells to cisplatin (Koyama et al., 2004). Treating lung adenocarcinoma cells, LT23, with Trichostatin A (TSA), a histone deacetylase inhibitor, induces

Id1, Id2, and Id3 expression, contributing to cell death (Eickhoff et al., 2000).

However, the Id1/Id3 peptide aptamer (Id1/3-PA7) has been shown to upregulate p21Cip1 and p27Kip1 expression and induce cell cycle arrest and apoptosis in breast cancer cell lines

MCF7 and MDA-MB-231 (Mern et al., 2010b). Knocking down Id expression using RNA interference techniques in gastric cancer cells (MKN45) and adenoid cystic carcinoma cells

(ACCM) has been shown to reduce xenograft sizes, reduce invasion, and induce apoptosis (Chen et al., 2011; Tsuchiya et al., 2005). In combination with Hsp90 inhibitor 17-AAG (also known as Tanespimycin), Id deficiency (Id1 and/or Id3) in human epidermal growth factor receptor 2

(Her2)-dependent breast tumors cells significantly suppresses xenograft tumor formation (de

Candia et al., 2003).

Id3 can function as both a tumor suppressor gene and an oncogene. Given the pleiotropic effects of Id3 in different tumor types, employing Id3 as a therapeutic target requires careful evaluation, and context-specific effects should be used to inform design of therapeutics when making decisions to treat cancer patients. Most likely, our research will benefit SCC patients with genetic signatures similar to those of A431 cells. In addition, based on our molecular model in Figure 8, patients with high levels of Myc may benefit from Id3 overexpression since Id3 induces Mxd3, which disrupts the Myc proliferative signaling. It may also be useful to modulate

Id3 expression in tumors with high ETS and MMP-2 levels, since Id3 blocks ETS proteins, which has been shown to suppress SCC invasion (Higashikawa et al., 2009). Similarly, manipulation of Id3 in tumors with elevated levels of anti-apoptotic Bcl-2 family proteins may

96 be useful since this study shows that Id3 is able to induce apoptosis pathway via a Bax- independent pathway.

It is postulated that E protein levels in the cells are a determining factor for the role of

Id3. Thus, levels of E proteins may be screened in future cancer patients to decide if Id3 will play a role as a tumor suppressor. In cells where E proteins normally function to promote differentiation and inhibit proliferation (e.g., B lymphocytes), Id3 can function as an oncogene via disruption of E protein functions. However, if E proteins are expressed at low levels, Id3 may not influence cell proliferation and its apoptosis-promoting role will be more potent.

The potential of exploiting Id3 as an apoptosis inducer requires more investigation in patients. Inducing endogenous levels of Id3 may be achieved through Bone morphogenetic protein-4 (BMP-4) infusion (Shepherd et al., 2008). However, careful evaluation of its side effects and patients’ genetic makeups are critical to the success of this approach. Our research demonstrates the apoptosis-inducing side of the story for Id3 protein, yet we are aware of the potential of Id3 as an oncogene. Induction and suppression of Id3 may both be life-savers depending on each individual’s situation. Therefore, we are cautiously optimistic about the broad application of Id3 as a therapeutic target.

Appendix

A431 Mycoplasma Detection Test and IMPACT Study

This appendix contains pathogen test results for A431 parental cells. Microbe PCR

Amplification Test (IMPACT) was used to detect 17 commonly seen pathogens in cell lines.

A431 cells were tested negative for all of them.

100

A431 Cell Line Authentication Test

This appendix contains cell line authentication results for GFP-expressing A431/Id3 clone 5 cells comparing o A431 parental (P) cells. This cell line is validated using short tandem repeat (STR) profiling on 8 genetic sites. The DNA fingerprints between the two lines match.

101

A431(p)

A431/Id3 (cl5) A431/Id3

102

A431(p)

A431/Id3 (cl5) A431/Id3

103

A431(p)

A431/Id3 (cl5) A431/Id3

104

A431(p)

A431/Id3 (cl5) A431/Id3

105

A431(p)

A431/Id3 (cl5) A431/Id3

106

A431(p)

A431/Id3 (cl5) A431/Id3

107

A431(p)

A431/Id3 (cl5) A431/Id3

108

A431(p)

A431/Id3 (cl5) A431/Id3

109

References

Alani, R. M., Hasskarl, J., Grace, M., Hernandez, M. C., Israel, M. A., and Munger, K. (1999). Immortalization of primary human keratinocytes by the helix-loop-helix protein, Id-1. Proceedings of the National Academy of Sciences of the United States of America 96, 9637-9641. American Academy of Dermatology (2011). Dermatology A to Z: Diseases and Treatments. In. American Cancer Society (2013). Cancer Facts & Figures 2013. In. Argiris, A., Li, Y., and Forastiere, A. (2004). Prognostic factors and long-term survivorship in patients with recurrent or metastatic carcinoma of the head and neck. Cancer 101, 2222-2229. Asirvatham, A. J., Carey, J. P., and Chaudhary, J. (2007). ID1-, ID2-, and ID3-regulated gene expression in E2A positive or negative prostate cancer cells. The Prostate 67, 1411-1420. Barrett, J. R. (2004). Template for toxicants: gene expression varies by cell type. Environ Health Perspect 112, A944. Barrett, L. E., Sul, J. Y., Takano, H., Van Bockstaele, E. J., Haydon, P. G., and Eberwine, J. H. (2006). Region-directed phototransfection reveals the functional significance of a dendritically synthesized transcription factor. Nature methods 3, 455-460. Benezra, R., Davis, R. L., Lockshon, D., Turner, D. L., and Weintraub, H. (1990). The protein Id: a negative regulator of helix-loop-helix DNA binding proteins. Cell 61, 49-59. Besnard, A., Galan-Rodriguez, B., Vanhoutte, P., and Caboche, J. (2011). Elk-1 a transcription factor with multiple facets in the brain. Frontiers in neuroscience 5, 35. Bhattacharya, A., and Baker, N. E. (2011). A network of broadly expressed HLH genes regulates tissue-specific cell fates. Cell 147, 881-892. Bickers, D. R., Lim, H. W., Margolis, D., Weinstock, M. A., Goodman, C., Faulkner, E., Gould, C., Gemmen, E., Dall, T., American Academy of Dermatology, A., and Society for Investigative, D. (2006). The burden of skin diseases: 2004 a joint project of the American Academy of Dermatology Association and the Society for Investigative Dermatology. Journal of the American Academy of Dermatology 55, 490-500. Biggs, J., Murphy, E. V., and Israel, M. A. (1992). A human Id-like helix-loop-helix protein expressed during early development. Proceedings of the National Academy of Sciences of the United States of America 89, 1512-1516. Caamano, J., Zhang, S. Y., Rosvold, E. A., Bauer, B., and Klein-Szanto, A. J. (1993). p53 alterations in human squamous cell carcinomas and carcinoma cell lines. The American journal of pathology 142, 1131-1139. Campo, E. (2012). New pathogenic mechanisms in Burkitt lymphoma. Nature genetics 44, 1288- 1289. Campuzano, S. (2001). Emc, a negative HLH regulator with multiple functions in Drosophila development. Oncogene 20, 8299-8307. Cesari, F., Brecht, S., Vintersten, K., Vuong, L. G., Hofmann, M., Klingel, K., Schnorr, J. J., Arsenian, S., Schild, H., Herdegen, T., et al. (2004). Mice deficient for the ets transcription factor elk-1 show normal immune responses and mildly impaired neuronal gene activation. Molecular and cellular biology 24, 294-305. Chakrabarty, A., and Geisse, J. K. (2004). Medical therapies for non-melanoma skin cancer. Clinics in dermatology 22, 183-188. Chen, Z., Liu, S., Sumida, T., Sun, S., Wei, Y., Liu, M., Dong, Z., Zhang, F., Hamakawa, H., and Wei, F. (2011). Silencing Id-1 with RNA interference inhibits adenoid cystic carcinoma in mice. The Journal of surgical research 169, 57-66.

110

Christy, B. A., Sanders, L. K., Lau, L. F., Copeland, N. G., Jenkins, N. A., and Nathans, D. (1991). An Id-related helix-loop-helix protein encoded by a growth factor-inducible gene. Proceedings of the National Academy of Sciences of the United States of America 88, 1815-1819. Colevas, A. D. (2006). Chemotherapy options for patients with metastatic or recurrent squamous cell carcinoma of the head and neck. Journal of clinical oncology : official journal of the American Society of Clinical Oncology 24, 2644-2652. Cummings, S. D., Ryu, B., Samuels, M. A., Yu, X., Meeker, A. K., Healey, M. A., and Alani, R. M. (2008). Id1 delays senescence of primary human melanocytes. Molecular carcinogenesis 47, 653-659. de Candia, P., Solit, D. B., Giri, D., Brogi, E., Siegel, P. M., Olshen, A. B., Muller, W. J., Rosen, N., and Benezra, R. (2003). Angiogenesis impairment in Id-deficient mice cooperates with an Hsp90 inhibitor to completely suppress HER2/neu-dependent breast tumors. Proceedings of the National Academy of Sciences of the United States of America 100, 12337-12342. Deed, R. W., Bianchi, S. M., Atherton, G. T., Johnston, D., Santibanez-Koref, M., Murphy, J. J., and Norton, J. D. (1993). An immediate early human gene encodes an Id-like helix-loop-helix protein and is regulated by protein kinase C activation in diverse cell types. Oncogene 8, 599-607. Deed, R. W., Hara, E., Atherton, G. T., Peters, G., and Norton, J. D. (1997). Regulation of Id3 cell cycle function by Cdk-2-dependent phosphorylation. Molecular and cellular biology 17, 6815-6821. Deed, R. W., Hirose, T., Mitchell, E. L., Santibanez-Koref, M. F., and Norton, J. D. (1994). Structural organisation and chromosomal mapping of the human Id-3 gene. Gene 151, 309-314. Degterev, A., Boyce, M., and Yuan, J. (2003). A decade of caspases. Oncogene 22, 8543-8567. Desprez, P. Y., Sumida, T., and Coppe, J. P. (2003). Helix-loop-helix proteins in mammary gland development and breast cancer. Journal of mammary gland biology and neoplasia 8, 225-239. DiVito, K. A., Simbulan-Rosenthal, C. M., Chen, Y. S., Trabosh, V. A., and Rosenthal, D. S. (2014). Id2, Id3 and Id4 overcome a Smad7-mediated block in tumorigenesis, generating TGF-beta-independent melanoma. Carcinogenesis 35, 951-958. Doran, A. C., Lehtinen, A. B., Meller, N., Lipinski, M. J., Slayton, R. P., Oldham, S. N., Skaflen, M. D., Yeboah, J., Rich, S. S., Bowden, D. W., and McNamara, C. A. (2010). Id3 is a novel atheroprotective factor containing a functionally significant single-nucleotide polymorphism associated with intima-media thickness in humans. Circulation research 106, 1303-1311. Eickhoff, B., Ruller, S., Laue, T., Kohler, G., Stahl, C., Schlaak, M., and van der Bosch, J. (2000). Trichostatin A modulates expression of p21waf1/cip1, Bcl-xL, ID1, ID2, ID3, CRAB2, GATA-2, hsp86 and TFIID/TAFII31 mRNA in human lung adenocarcinoma cells. Biological chemistry 381, 107-112. Ellenberger, T., Fass, D., Arnaud, M., and Harrison, S. C. (1994). Crystal structure of transcription factor E47: E-box recognition by a basic region helix-loop-helix dimer. Genes & development 8, 970-980. Ellmeier, W., Aguzzi, A., Kleiner, E., Kurzbauer, R., and Weith, A. (1992). Mutually exclusive expression of a helix-loop-helix gene and N-myc in human neuroblastomas and in normal development. The EMBO journal 11, 2563-2571. Fernandez-Salas, E., Suh, K. S., Speransky, V. V., Bowers, W. L., Levy, J. M., Adams, T., Pathak, K. R., Edwards, L. E., Hayes, D. D., Cheng, C., et al. (2002). mtCLIC/CLIC4, an organellular chloride channel protein, is increased by DNA damage and participates in the apoptotic response to p53. Molecular and cellular biology 22, 3610-3620. Fitzsimmons, D., Hodsdon, W., Wheat, W., Maira, S. M., Wasylyk, B., and Hagman, J. (1996). Pax- 5 (BSAP) recruits Ets proto-oncogene family proteins to form functional ternary complexes on a B-cell- specific promoter. Genes & development 10, 2198-2211. Florio, M., Hernandez, M. C., Yang, H., Shu, H. K., Cleveland, J. L., and Israel, M. A. (1998). Id2 promotes apoptosis by a novel mechanism independent of dimerization to basic helix-loop-helix factors. Molecular and cellular biology 18, 5435-5444.

111

Fontemaggi, G., Dell'Orso, S., Trisciuoglio, D., Shay, T., Melucci, E., Fazi, F., Terrenato, I., Mottolese, M., Muti, P., Domany, E., et al. (2009). The execution of the transcriptional axis mutant p53, E2F1 and ID4 promotes tumor neo-angiogenesis. Nature structural & molecular biology 16, 1086-1093. Fukunishi, N., Katoh, I., Tomimori, Y., Tsukinoki, K., Hata, R., Nakao, A., Ikawa, Y., and Kurata, S. (2010). Induction of DeltaNp63 by the newly identified keratinocyte-specific transforming growth factor beta Signaling Pathway with Smad2 and IkappaB Kinase alpha in squamous cell carcinoma. Neoplasia 12, 969-979. Gebauer, N., Bernard, V., Feller, A. C., and Merz, H. (2013). ID3 mutations are recurrent events in double-hit B-cell lymphomas. Anticancer research 33, 4771-4778. Ginos, M. A., Page, G. P., Michalowicz, B. S., Patel, K. J., Volker, S. E., Pambuccian, S. E., Ondrey, F. G., Adams, G. L., and Gaffney, P. M. (2004). Identification of a gene expression signature associated with recurrent disease in squamous cell carcinoma of the head and neck. Cancer research 64, 55-63. Hara, E., Hall, M., and Peters, G. (1997). Cdk2-dependent phosphorylation of Id2 modulates activity of E2A-related transcription factors. The EMBO journal 16, 332-342. Hara, E., Yamaguchi, T., Nojima, H., Ide, T., Campisi, J., Okayama, H., and Oda, K. (1994). Id- related genes encoding helix-loop-helix proteins are required for G1 progression and are repressed in senescent human fibroblasts. The Journal of biological chemistry 269, 2139-2145. Hietanen, S., Grenman, S., Syrjanen, K., Lappalainen, K., Kauppinen, J., Carey, T., and Syrjanen, S. (1995). Human papillomavirus in vulvar and vaginal carcinoma cell lines. British journal of cancer 72, 134-139. Higashikawa, K., Yoneda, S., Tobiume, K., Saitoh, M., Taki, M., Mitani, Y., Shigeishi, H., Ono, S., and Kamata, N. (2009). DeltaNp63alpha-dependent expression of Id-3 distinctively suppresses the invasiveness of human squamous cell carcinoma. International journal of cancer Journal international du cancer 124, 2837-2844. Hobeika, E., Thiemann, S., Storch, B., Jumaa, H., Nielsen, P. J., Pelanda, R., and Reth, M. (2006). Testing gene function early in the B cell lineage in mb1-cre mice. Proceedings of the National Academy of Sciences of the United States of America 103, 13789-13794. Iavarone, A., Garg, P., Lasorella, A., Hsu, J., and Israel, M. A. (1994). The helix-loop-helix protein Id-2 enhances cell proliferation and binds to the retinoblastoma protein. Genes & development 8, 1270- 1284. Igney, F. H., and Krammer, P. H. (2002). Death and anti-death: tumour resistance to apoptosis. Nature reviews Cancer 2, 277-288. Israel, M. A., Hernandez, M. C., Florio, M., Andres-Barquin, P. J., Mantani, A., Carter, J. H., and Julin, C. M. (1999). Id gene expression as a key mediator of tumor cell biology. Cancer research 59, 1726s-1730s. Kageyama, R., Ohtsuka, T., and Kobayashi, T. (2007). The Hes gene family: repressors and oscillators that orchestrate embryogenesis. Development 134, 1243-1251. Kanitakis, J. (2009). Rare skin cancers. Cancer treatment and research 146, 323-328. Kee, B. L., Rivera, R. R., and Murre, C. (2001). Id3 inhibits B lymphocyte progenitor growth and survival in response to TGF-beta. Nature immunology 2, 242-247. Kerr, J. F., Wyllie, A. H., and Currie, A. R. (1972). Apoptosis: a basic biological phenomenon with wide-ranging implications in tissue kinetics. British journal of cancer 26, 239-257. Kowanetz, M., Valcourt, U., Bergstrom, R., Heldin, C. H., and Moustakas, A. (2004). Id2 and Id3 define the potency of cell proliferation and differentiation responses to transforming growth factor beta and bone morphogenetic protein. Molecular and cellular biology 24, 4241-4254. Koyama, T., Suzuki, H., Imakiire, A., Yanase, N., Hata, K., and Mizuguchi, J. (2004). Id3-mediated enhancement of cisplatin-induced apoptosis in a sarcoma cell line MG-63. Anticancer Research 24, 1519- 1524.

112

Kraemer, K. H., Lee, M. M., Andrews, A. D., and Lambert, W. C. (1994). The role of sunlight and DNA repair in melanoma and nonmelanoma skin cancer. The xeroderma pigmentosum paradigm. Archives of dermatology 130, 1018-1021. Kraemer, K. H., Lee, M. M., and Scotto, J. (1984). DNA repair protects against cutaneous and internal neoplasia: evidence from xeroderma pigmentosum. Carcinogenesis 5, 511-514. Kumar, D., Shadrach, J. L., Wagers, A. J., and Lassar, A. B. (2009). Id3 is a direct transcriptional target of Pax7 in quiescent satellite cells. Molecular biology of the cell 20, 3170-3177. Langlands, K., Down, G. A., and Kealey, T. (2000). Id proteins are dynamically expressed in normal epidermis and dysregulated in squamous cell carcinoma. Cancer research 60, 5929-5933. Latchney, S. E., Lioy, D. T., Henry, E. C., Gasiewicz, T. A., Strathmann, F. G., Mayer-Proschel, M., and Opanashuk, L. A. (2011). Neural precursor cell proliferation is disrupted through activation of the aryl hydrocarbon receptor by 2,3,7,8-tetrachlorodibenzo-p-dioxin. Stem cells and development 20, 313- 326. Ledent, V., Paquet, O., and Vervoort, M. (2002). Phylogenetic analysis of the human basic helix- loop-helix proteins. Genome biology 3, RESEARCH0030. Lee, Y. S., Mollah, M. L., Sohn, K. C., Shi, G., Kim, D. H., Kim, K. H., Cho, M. J., Kim, S., Lee, Y. H., Kim, C. D., and Lee, J. H. (2010). ID3 mediates X-ray-induced apoptosis of keratinocytes through the regulation of beta-catenin. J Dermatol Sci 60, 138-142. Licitra, L., Perrone, F., Tamborini, E., Bertola, L., Ghirelli, C., Negri, T., Orsenigo, M., Filipazzi, P., Pastore, E., Pompilio, M., et al. (2011). Role of EGFR family receptors in proliferation of squamous carcinoma cells induced by wound healing fluids of head and neck cancer patients. Annals of oncology : official journal of the European Society for Medical Oncology / ESMO 22, 1886-1893. Liedtke, C., Groger, N., Manns, M. P., and Trautwein, C. (2003). The human caspase-8 promoter sustains basal activity through SP1 and ETS-like transcription factors and can be up-regulated by a p53- dependent mechanism. The Journal of biological chemistry 278, 27593-27604. Life Technologies (2011). T-REx System: A Tetracycline-Regulated Expression System for Mammalian Cells. In, (Invitrogen). Love, C., Sun, Z., Jima, D., Li, G., Zhang, J., Miles, R., Richards, K. L., Dunphy, C. H., Choi, W. W., Srivastava, G., et al. (2012). The genetic landscape of mutations in Burkitt lymphoma. Nature genetics 44, 1321-1325. Luo, K. J., Wen, J., Xie, X., Fu, J. H., Luo, R. Z., Wu, Q. L., and Hu, Y. (2012). Prognostic relevance of Id-1 expression in patients with resectable esophageal squamous cell carcinoma. The Annals of thoracic surgery 93, 1682-1688. Lyden, D., Young, A. Z., Zagzag, D., Yan, W., Gerald, W., O'Reilly, R., Bader, B. L., Hynes, R. O., Zhuang, Y., Manova, K., and Benezra, R. (1999). Id1 and Id3 are required for neurogenesis, angiogenesis and vascularization of tumour xenografts. Nature 401, 670-677. Martinez, J. C., and Otley, C. C. (2001). The management of melanoma and nonmelanoma skin cancer: a review for the primary care physician. Mayo Clinic proceedings 76, 1253-1265. Martinsen, B. J., and Bronner-Fraser, M. (1998). Neural crest specification regulated by the helix- loop-helix repressor Id2. Science 281, 988-991. Massari, M. E., and Murre, C. (2000). Helix-loop-helix proteins: regulators of transcription in eucaryotic organisms. Molecular and cellular biology 20, 429-440. Mathew, S., Chen, W., Murty, V. V., Benezra, R., and Chaganti, R. S. (1995). Chromosomal assignment of human ID1 and ID2 genes. Genomics 30, 385-387. Matsumura, M. E., Lobe, D. R., and McNamara, C. A. (2002). Contribution of the helix-loop-helix factor Id2 to regulation of vascular smooth muscle cell proliferation. The Journal of biological chemistry 277, 7293-7297.

113

Mern, D. S., Hasskarl, J., and Burwinkel, B. (2010a). Inhibition of Id proteins by a peptide aptamer induces cell-cycle arrest and apoptosis in ovarian cancer cells. British journal of cancer 103, 1237-1244. Mern, D. S., Hoppe-Seyler, K., Hoppe-Seyler, F., Hasskarl, J., and Burwinkel, B. (2010b). Targeting Id1 and Id3 by a specific peptide aptamer induces E-box promoter activity, cell cycle arrest, and apoptosis in breast cancer cells. Breast Cancer Res Treat 124, 623-633. Min, B. M., Baek, J. H., Shin, K. H., Gujuluva, C. N., Cherrick, H. M., and Park, N. H. (1994). Inactivation of the p53 gene by either mutation or HPV infection is extremely frequent in human oral squamous cell carcinoma cell lines. European journal of cancer Part B, Oral oncology 30B, 338-345. Neville, J. A., Welch, E., and Leffell, D. J. (2007). Management of nonmelanoma skin cancer in 2007. Nat Clin Pract Oncol 4, 462-469. Nickoloff, B. J., Chaturvedi, V., Bacon, P., Qin, J. Z., Denning, M. F., and Diaz, M. O. (2000). Id-1 delays senescence but does not immortalize keratinocytes. The Journal of biological chemistry 275, 27501-27504. Nindl, I., Dang, C., Forschner, T., Kuban, R. J., Meyer, T., Sterry, W., and Stockfleth, E. (2006). Identification of differentially expressed genes in cutaneous squamous cell carcinoma by microarray expression profiling. Molecular cancer 5, 30. Nishimine, M., Nakamura, M., Mishima, K., Kishi, M., Kirita, T., Sugimura, M., and Konishi, N. (2003). Id proteins are overexpressed in human oral squamous cell carcinomas. Journal of oral pathology & medicine : official publication of the International Association of Oral Pathologists and the American Academy of Oral Pathology 32, 350-357. Norton, J. D. (2000). ID helix-loop-helix proteins in cell growth, differentiation and tumorigenesis. Journal of cell science 113 ( Pt 22), 3897-3905. Norton, J. D., and Atherton, G. T. (1998). Coupling of cell growth control and apoptosis functions of Id proteins. Molecular and cellular biology 18, 2371-2381. O'Brien, C. A., Kreso, A., Ryan, P., Hermans, K. G., Gibson, L., Wang, Y., Tsatsanis, A., Gallinger, S., and Dick, J. E. (2012). ID1 and ID3 regulate the self-renewal capacity of human colon cancer-initiating cells through p21. Cancer cell 21, 777-792. O'Bryan, K. W., and Ratner, D. (2011). The role of targeted molecular inhibitors in the management of advanced nonmelanoma skin cancer. Seminars in cutaneous medicine and surgery 30, 57-61. Oh, Y. T., Liu, X., Yue, P., Kang, S., Chen, J., Taunton, J., Khuri, F. R., and Sun, S. Y. (2010). ERK/ribosomal S6 kinase (RSK) signaling positively regulates death receptor 5 expression through co- activation of CHOP and Elk1. The Journal of biological chemistry 285, 41310-41319. Ohtani, N., Zebedee, Z., Huot, T. J., Stinson, J. A., Sugimoto, M., Ohashi, Y., Sharrocks, A. D., Peters, G., and Hara, E. (2001). Opposing effects of Ets and Id proteins on p16INK4a expression during cellular senescence. Nature 409, 1067-1070. OncomineTM (2014). Disease summary for Id3. In. Pagliuca, A., Bartoli, P. C., Saccone, S., Della Valle, G., and Lania, L. (1995). Molecular cloning of ID4, a novel dominant negative helix-loop-helix human gene on chromosome 6p21.3-p22. Genomics 27, 200-203. Perk, J., Iavarone, A., and Benezra, R. (2005). Id family of helix-loop-helix proteins in cancer. Nature reviews Cancer 5, 603-614. Pomerantz, J. L., and Baltimore, D. (1999). NF-kappaB activation by a signaling complex containing TRAF2, TANK and TBK1, a novel IKK-related kinase. The EMBO journal 18, 6694-6704. Ratushny, V., Gober, M. D., Hick, R., Ridky, T. W., and Seykora, J. T. (2012). From keratinocyte to cancer: the pathogenesis and modeling of cutaneous squamous cell carcinoma. The Journal of clinical investigation 122, 464-472.

114

Reiss, M., Brash, D. E., Munoz-Antonia, T., Simon, J. A., Ziegler, A., Vellucci, V. F., and Zhou, Z. L. (1992). Status of the p53 tumor suppressor gene in human squamous carcinoma cell lines. Oncol Res 4, 349-357. Rescan, P. Y. (1997). Identification in a fish species of two Id (inhibitor of DNA binding/differentiation)-related helix-loop-helix factors expressed in the slow oxidative muscle fibers. European journal of biochemistry / FEBS 247, 870-876. Richter, J., Schlesner, M., Hoffmann, S., Kreuz, M., Leich, E., Burkhardt, B., Rosolowski, M., Ammerpohl, O., Wagener, R., Bernhart, S. H., et al. (2012). Recurrent mutation of the ID3 gene in Burkitt lymphoma identified by integrated genome, exome and transcriptome sequencing. Nature genetics 44, 1316-1320. Riechmann, V., van Cruchten, I., and Sablitzky, F. (1994). The expression pattern of Id4, a novel dominant negative helix-loop-helix protein, is distinct from Id1, Id2 and Id3. Nucleic acids research 22, 749-755. Riker, A. I., Enkemann, S. A., Fodstad, O., Liu, S., Ren, S., Morris, C., Xi, Y., Howell, P., Metge, B., Samant, R. S., et al. (2008). The gene expression profiles of primary and metastatic melanoma yields a transition point of tumor progression and metastasis. BMC Med Genomics 1, 13. Roberts, E. C., Deed, R. W., Inoue, T., Norton, J. D., and Sharrocks, A. D. (2001). Id helix-loop- helix proteins antagonize pax transcription factor activity by inhibiting DNA binding. Molecular and cellular biology 21, 524-533. Rogers, H. W., Weinstock, M. A., Harris, A. R., Hinckley, M. R., Feldman, S. R., Fleischer, A. B., and Coldiron, B. M. (2010). Incidence estimate of nonmelanoma skin cancer in the United States, 2006. Archives of dermatology 146, 283-287. Sano, D., Xie, T. X., Ow, T. J., Zhao, M., Pickering, C. R., Zhou, G., Sandulache, V. C., Wheeler, D. A., Gibbs, R. A., Caulin, C., and Myers, J. N. (2011). Disruptive TP53 mutation is associated with aggressive disease characteristics in an orthotopic murine model of oral tongue cancer. Clin Cancer Res 17, 6658-6670. Sawada, S., Mese, H., Sasaki, A., Yoshioka, N., and Matsumura, T. (2003). Combination chemotherapy of paclitaxel and cisplatin induces apoptosis with Bcl-2 phosphorylation in a cisplatin- resistant human epidermoid carcinoma cell line. Cancer chemotherapy and pharmacology 51, 505-511. Schlingemann, J., Habtemichael, N., Ittrich, C., Toedt, G., Kramer, H., Hambek, M., Knecht, R., Lichter, P., Stauber, R., and Hahn, M. (2005). Patient-based cross-platform comparison of oligonucleotide microarray expression profiles. Lab Invest 85, 1024-1039. Schmitz, R., Young, R. M., Ceribelli, M., Jhavar, S., Xiao, W., Zhang, M., Wright, G., Shaffer, A. L., Hodson, D. J., Buras, E., et al. (2012). Burkitt lymphoma pathogenesis and therapeutic targets from structural and functional genomics. Nature 490, 116-120. Schmitz, S., Ang, K. K., Vermorken, J., Haddad, R., Suarez, C., Wolf, G. T., Hamoir, M., and Machiels, J. P. (2014). Targeted therapies for squamous cell carcinoma of the head and neck: Current knowledge and future directions. Cancer Treat Rev 40, 390-404. Seitz, V., Butzhammer, P., Hirsch, B., Hecht, J., Gutgemann, I., Ehlers, A., Lenze, D., Oker, E., Sommerfeld, A., von der Wall, E., et al. (2011). Deep sequencing of MYC DNA-binding sites in Burkitt lymphoma. PloS one 6, e26837. Shao, N., Chai, Y., Cui, J. Q., Wang, N., Aysola, K., Reddy, E. S., and Rao, V. N. (1998). Induction of apoptosis by Elk-1 and deltaElk-1 proteins. Oncogene 17, 527-532. Sharrocks, A. D. (2002). Complexities in ETS-domain transcription factor function and regulation: lessons from the TCF (ternary complex factor) subfamily. The Colworth Medal Lecture. Biochemical Society transactions 30, 1-9. Shepherd, T. G., Theriault, B. L., and Nachtigal, M. W. (2008). Autocrine BMP4 signalling regulates ID3 proto-oncogene expression in human ovarian cancer cells. Gene 414, 95-105.

115

Shiio, Y., Suh, K. S., Lee, H., Yuspa, S. H., Eisenman, R. N., and Aebersold, R. (2006). Quantitative proteomic analysis of myc-induced apoptosis: a direct role for Myc induction of the mitochondrial chloride ion channel, mtCLIC/CLIC4. The Journal of biological chemistry 281, 2750-2756. Shin, S. Y., Kim, C. G., Lim, Y., and Lee, Y. H. (2011). The ETS family transcription factor ELK-1 regulates induction of the cell cycle-regulatory gene p21(Waf1/Cip1) and the BAX gene in sodium arsenite-exposed human keratinocyte HaCaT cells. The Journal of biological chemistry 286, 26860- 26872. Sikder, H. A., Devlin, M. K., Dunlap, S., Ryu, B., and Alani, R. M. (2003). Id proteins in cell growth and tumorigenesis. Cancer cell 3, 525-530. Simbulan-Rosenthal, C. M., Daher, A., Trabosh, V., Chen, W. C., Gerstel, D., Soeda, E., and Rosenthal, D. S. (2006). Id3 induces a caspase-3- and -9-dependent apoptosis and mediates UVB sensitization of HPV16 E6/7 immortalized human keratinocytes. Oncogene 25, 3649-3660. Simbulan-Rosenthal, C. M., Trabosh, V., Velarde, A., Chou, F. P., Daher, A., Tenzin, F., Tokino, T., and Rosenthal, D. S. (2005). Id2 protein is selectively upregulated by UVB in primary, but not in immortalized human keratinocytes and inhibits differentiation. Oncogene 24, 5443-5458. Simpson, C. L., Patel, D. M., and Green, K. J. (2011). Deconstructing the skin: cytoarchitectural determinants of epidermal morphogenesis. Nature reviews Molecular cell biology 12, 565-580. Skin Cancer Foundation (2014). Skin Cancer Information: Sqamous Cell Carcinoma. In. Skvortsov, S., Skvortsova, I., Stasyk, T., Schiefermeier, N., Neher, A., Gunkel, A. R., Bonn, G. K., Huber, L. A., Lukas, P., Pleiman, C. M., and Zwierzina, H. (2007). Antitumor activity of CTFB, a novel anticancer agent, is associated with the down-regulation of nuclear factor-kappaB expression and proteasome activation in head and neck squamous carcinoma cell lines. Molecular cancer therapeutics 6, 1898-1908. Snuderl, M., Kolman, O. K., Chen, Y. B., Hsu, J. J., Ackerman, A. M., Dal Cin, P., Ferry, J. A., Harris, N. L., Hasserjian, R. P., Zukerberg, L. R., et al. (2010). B-cell lymphomas with concurrent IGH-BCL2 and MYC rearrangements are aggressive neoplasms with clinical and pathologic features distinct from Burkitt lymphoma and diffuse large B-cell lymphoma. The American journal of surgical pathology 34, 327-340. St John, L. S., Sauter, E. R., Herlyn, M., Litwin, S., and Adler-Storthz, K. (2000). Endogenous p53 gene status predicts the response of human squamous cell carcinomas to wild-type p53. Cancer gene therapy 7, 749-756. Stinson, J., Inoue, T., Yates, P., Clancy, A., Norton, J. D., and Sharrocks, A. D. (2003). Regulation of TCF ETS-domain transcription factors by helix-loop-helix motifs. Nucleic acids research 31, 4717-4728. Suh, K. S., Malik, M., Shukla, A., Ryscavage, A., Wright, L., Jividen, K., Crutchley, J. M., Dumont, R. A., Fernandez-Salas, E., Webster, J. D., et al. (2012). CLIC4 is a tumor suppressor for cutaneous squamous cell cancer. Carcinogenesis 33, 986-995. Sun, X. H., Copeland, N. G., Jenkins, N. A., and Baltimore, D. (1991). Id proteins Id1 and Id2 selectively inhibit DNA binding by one class of helix-loop-helix proteins. Molecular and cellular biology 11, 5603-5611. Talantov, D., Mazumder, A., Yu, J. X., Briggs, T., Jiang, Y., Backus, J., Atkins, D., and Wang, Y. (2005). Novel genes associated with malignant melanoma but not benign melanocytic lesions. Clin Cancer Res 11, 7234-7242. Tang, M. J., Tai, T. C., and Tai, I. T. (2009). SPARC-Mediated Binding of Sp1 and Elk1 to the Caspase 8 Promoter Increases Apoptosis in Chemotherapy Resistant Colorectal Cancer Cells. Paper presented at: American Gastroenterological Association (Elsevier Inc.). Tepper, C. G., Gregg, J. P., Shi, X. B., Vinall, R. L., Baron, C. A., Ryan, P. E., Desprez, P. Y., Kung, H. J., and deVere White, R. W. (2005). Profiling of gene expression changes caused by p53 gain-of-function mutant alleles in prostate cancer cells. The Prostate 65, 375-389.

116

Tournay, O., and Benezra, R. (1996). Transcription of the dominant-negative helix-loop-helix protein Id1 is regulated by a protein complex containing the immediate-early response gene Egr-1. Molecular and cellular biology 16, 2418-2430. Trabosh, V. A., Daher, A., Divito, K. A., Amin, K., Simbulan-Rosenthal, C. M., and Rosenthal, D. S. (2009a). UVB upregulates the bax promoter in immortalized human keratinocytes via ROS induction of Id3. Experimental dermatology 18, 387-395. Trabosh, V. A., Divito, K. A., B, D. A., Simbulan-Rosenthal, C. M., and Rosenthal, D. S. (2009b). Sequestration of E12/E47 and suppression of p27KIP1 play a role in Id2-induced proliferation and tumorigenesis. Carcinogenesis 30, 1252-1259. Troester, M. A., Hoadley, K. A., Parker, J. S., and Perou, C. M. (2004). Prediction of toxicant- specific gene expression signatures after chemotherapeutic treatment of breast cell lines. Environ Health Perspect 112, 1607-1613. Tsuchiya, T., Okaji, Y., Tsuno, N. H., Sakurai, D., Tsuchiya, N., Kawai, K., Yazawa, K., Asakage, M., Yamada, J., Yoneyama, S., et al. (2005). Targeting Id1 and Id3 inhibits peritoneal metastasis of gastric cancer. Cancer science 96, 784-790. Weith, A., Martinsson, T., Cziepluch, C., Bruderlein, S., Amler, L. C., Berthold, F., and Schwab, M. (1989). Neuroblastoma consensus deletion maps to 1p36.1-2. Genes Chromosomes Cancer 1, 159-166. Wong, R. P., Tsang, W. P., Chau, P. Y., Co, N. N., Tsang, T. Y., and Kwok, T. T. (2007). p53-R273H gains new function in induction of drug resistance through down-regulation of procaspase-3. Molecular cancer therapeutics 6, 1054-1061. Yan, W., Liu, G., Scoumanne, A., and Chen, X. (2008). Suppression of inhibitor of differentiation 2, a target of mutant p53, is required for gain-of-function mutations. Cancer research 68, 6789-6796. Yates, P. R., Atherton, G. T., Deed, R. W., Norton, J. D., and Sharrocks, A. D. (1999). Id helix-loop- helix proteins inhibit nucleoprotein complex formation by the TCF ETS-domain transcription factors. The EMBO journal 18, 968-976. Yuan, J., Shaham, S., Ledoux, S., Ellis, H. M., and Horvitz, H. R. (1993). The C. elegans cell death gene ced-3 encodes a protein similar to mammalian interleukin-1 beta-converting enzyme. Cell 75, 641- 652. Zebedee, Z., and Hara, E. (2001). Id proteins in cell cycle control and cellular senescence. Oncogene 20, 8317-8325.

117