Epigenetic, Metabolic and Signal Transduction Pathways Involved in Radiation Response of Normal and Transformed Thyrocytes

ei involved in radiation response and transformed of normal Epigenetic, metabolic, pathways and signal transduction

Epigenetic, metabolic and signal transduction pathways involved in radiation response of normal and transformed thyrocytes

Khalil Abouelaradat Khalil AbouelaradatKhalil

ISBN 978-90-5989-484-6 2011

9 789059 894846 "Once something actually happens somewhere as wildly complicated as the Universe, Kevin knows where it will all end up – where 'Kevin' is any random entity that doesn't know nothin' about nothin'"

Douglas Adams, 'Mostly Harmless' Book 5 of the Trilogy of Five

Promoters: Prof. Dr. Ir. Wim VAN CRIEKINGE

Department of Mathematical Modeling, Statistics and Bioinformatics

Faculty of Bioscience Engineering

Ghent University

Prof. Dr. Sarah BAATOUT

Radiobiology Unit

Belgian Nuclear Research Center (SCK•CEN)

& Department of Molecular Biotechnology

Faculty of Bioscience Engineering

Ghent University

Prof. Dr. Ir. Tim DE MEYER

Department of Mathematical Modeling, Statistics and Bio-informatics /Department of Molecular Biotechnology

Faculty of Bioscience Engineering

Ghent University

Dr. Ir. Sofie BEKAERT

Clinical Research Center

Faculty for Medicine and Health Sciences

Ghent University

Dean: Prof. Dr. Ir. Guido VAN HUYLENBROECK

Rector: Prof. Dr. Paul VAN CAUWENBERGE

iii

Epigenetic, metabolic, and signal transduction pathways involved in radiation response of normal and transformed thyrocytes

Khalil Abouelaradat

Thesis submitted in fulfillment of the requirements for the degree of Doctor (PhD) in Applied Biological Sciences

"Epigenetische, metabole en signaaloverdrachtmechanismen betrokken bij de stralingsrespons van normale en getransformeerde schildkliercellen"

Cover Illustration:

Colored gamma scan (scintigram) of a healthy human thyroid gland, in front view (Centre Jean Perrin/Science Photo Library). A radioactive tracer (Tc-99m) highlights areas of activity in the thyroid where blue represents areas of low activity and green/red areas of high activity. Gamma scans involve injecting a radioactive tracer and then measuring the gamma rays emitted using a gamma camera. Picture was converted to ASCII art using Ascgen2 (http://sourceforge.net/projects/ascgen2). The ASCII picture is made out of repeats of the word RET/PTC.

Cover art design by Khalil Abouelaradat

Abouelaradat Khalil (2011). Epigenetic, metabolic, and signal transduction pathways involved in radiation response of normal and transformed thyrocytes. PhD Thesis. Ghent University.

ISBN: 978-90-5989-484-6

The author and the promoters give the authorization to consult and to copy parts of this work for personal use only. Any other use is limited by the Laws of Copyright. Permission to reproduce any material contained in this work should be obtained from the author.

Members of the Examination Committee

Prof. Dr. Ir. Wim Van Criekinge (Promoter)

Department of Mathematical Modeling, Statistics and Bioinformatics, Faculty of Bioscience Engineering, Ghent University

Prof. Dr. Sarah Baatout (Co-promoter)

Radiobiology Unit, Belgian Nuclear Research Center (SCK•CEN)

Department of Molecular Biotechnology, Faculty of Bioscience Engineering, Ghent University

Prof. Dr. Ir. Sofie Bekaert (Co-promoter)

Clinical Research Center, Faculty of Medicine and Health Sciences, Ghent University

Prof. Dr. Ir. Tim De Meyer (Co-promoter)

Department of Molecular Biotechnology/Department of Mathematical Modeling, Statistics and Bioinformatics, Faculty of Bioscience Engineering, Ghent University

Prof. Dr. Ir. Dirk Reheul (Chairman)

Department of Plant Production, Faculty of Bioscience Engineering, Ghent University

Prof. Dr. Ir. Guy Smagghe (Secretary)

Department of Crop Production, Faculty of Bioscience Engineering , Ghent University

Dr. Anne-Catherine Gérard

Morphology Group, Institute of Fundamental and Clinical Research, Catholic University of Louvain (UCL)

Prof. Dr. Hubert Thierens

Department of Basic Medical Sciences, Faculty of Medicine and Health Sciences, Ghent University

Em. Prof. Dr. Patrick Van Oostveldt

Department of Molecular Biotechnology, Faculty of Bioscience Engineering, Ghent University

Prof. Dr. Jan Philippé

Department of Clinical Chemistry, Microbiology and Immunology, Faculty of Medicine and Health Sciences, Ghent University

vii

Table of Contents List of Figures ...... xiv List of Tables ...... xvi List of Abbreviations ...... xviii

Chapter 1: General Introduction 1. The Nuclear Conundrum ...... 3 1.1 TMI and Chernobyl: Compare and Contrast ...... 3 1.2 Nuclear Weapons Testing ...... 5 1.3 Fukushima: Lessons from the Past ...... 6 1.4 External Irradiation ...... 7 2. The Thyroid ...... 9 2.1 Iodine Metabolism and Hormone Synthesis ...... 9 2.2 Thyroid Homeostasis ...... 13 3. Thyroid Cancer...... 15 3.1 Thyroid Cancer Types ...... 15 3.2 Mutations in PTC ...... 17 3.2.1 RAF ...... 17 3.2.2 RET/PTC ...... 17 3.2.2.1 Prominance of RET/PTC Translocations in Radiation-induced PTCs ...... 19 3.2.3 NTRK Rearrangments ...... 21 3.2.4 Ras ...... 21 4. Low Dose Radiation Introduction ...... 23 4.1 The Linear Non-Threshold Model (LNT) ...... 23 4.2 The LNT: Pros and Cons, Epidemiology ...... 25 4.3 Health Risk Estimates: Molecular ...... 26 4.3.1 γH2AX ...... 26 4.3.2 Cell Survival ...... 27 4.4 Deviations from the LNT ...... 28 4.4.1 HRS/IRR ...... 28 4.4.2 The Bystander Effect ...... 29 4.4.3 Adaptive Response ...... 30

viii

4.4.4 Hormesis ...... 30 4.5 Interindividual Sensitivity ...... 31 4.6 Risk to the Thyroid ...... 31 4.7 Conclusion ...... 32 5. Senescence Introduction: To Divide or Not To Divide...... 33 5.1 The Hayflick Mosaic ...... 34 5.2 Hitting the Breaks: Premature Senescence ...... 36 5.2.1 Hallmarks of Senescence ...... 36 5.2.2 The Role of p53 and pRb ...... 38 5.2.3 Other players on the senescence scene ...... 39 5.3 Final Remarks and Points of Contention ...... 40 6. Epigenetics ...... 43 6.1 DNA Methylation: And the Rest is Silence ...... 43 6.2 The Cancer Methylome ...... 45 6.2.1 Thyroid Cancer: The Epigenetic Factor ...... 46 6.3 DNA Methylation, DNA Repair, and Radiation: a Messy Triangle ...... 47 6.4 Detection and Measurement of DNA Methylation ...... 48 6.5 Sailing Uncharted Waters: Active DNA Demethylation and Gene Body Methylation ...... 50 6.6 MicroRNAs: Seek and Destroy (or not…) ...... 51 6.6.1 MiRandering Around the Rules ...... 53 Thesis Aims ...... 57 Chapter 2: Response to low dose of X-irradiation is p53-dependent in a papillary thyroid carcinoma model system

1. Abstract ...... 61 2. Introduction ...... 62 3. Materials and Methods ...... 64 3.1 Cell Culture ...... 64 3.2 Cell Irradiation ...... 64 3.3 Cell Viability, Counting and Morphology ...... 65 3.4 Cell Cycle and AnnexinV/PI ...... 65 3.5 Caspase-3 levels using Flow Cytometry ...... 67 3.6 Western Blotting ...... 68 3.7 TGF-β1 ELISA ...... 68

3.8 Detection of γ-H2AX foci ...... 69 3.9 Senescence-associated (SA) β-galactosidase quantification by fluoroscence ...... 70 3.10 Cytokine Level Measurement using Multiplex Bead Assay ...... 70 3.11 Statistical Analysis ...... 71 4. Results ...... 72 4.1 Irradiation Causes an Increase in Amount of Double Strand Breaks in TPC-1 Cells ...... 72 4.2 Irradiation Causes a Decrease in Number of TPC-1 Cells ...... 74 4.3 Cell Cycle Analysis Using PI Staining of DNA ...... 75 4.4 Analysis of Cell Death Induced by Irradiation ...... 76 4.5 Levels of Caspase-3 after Irradiation ...... 79 4.6 Changes in Cellular Morphology Following Irradiation ...... 80 4.7 Effect of external X-irradiation on secreted levels of TGF-β1 ...... 83 4.8 Effect of Irradiation on the Secreted Level of Cytokines Using a Multiplex Bead Assay...... 84 4.9 Translational Levels of Molecular Markers by Western blotting ...... 85 4.10 Effect of Irradiation on levels of β-galactosidase ...... 88 5. Discussion ...... 90 Chapter 3: Low Dose Irradiation of Thyroid Cells Reveals a Unique Transcriptomic and Epigenetic Signature in RET/PTC-positive Cells

1. Abstract ...... 101 2. Introduction ...... 102 3. Materials and Methods ...... 104 3.1 Cell Lines ...... 104 3.2 Mice ...... 104 3.3 Mouse Genotyping ...... 105 3.4 Irradiation ...... 105 3.5 RNA Isolation ...... 106 3.6 Protein Extraction and Western blotting ...... 106 3.7 MicroRNA Analysis ...... 107 3.8 Affymetrix Microarray Preparation ...... 108 3.9 Affymetrix Microarray Data Analysis ...... 108 3.10 Real-time RT-PCR (qRT-PCR) ...... 109 3.11 Multiplex Bead Assay Analysis of Protein Phosphorylation ...... 110 3.12 Statistical Analysis ...... 110

4. Results ...... 112 4.1 Thyroid Size and RET/PTC Status in Mice ...... 112 4.2 Differential Expression of Genes Between RET/PTC-positive and –negative thyroids ...... 113 4.3 Microarray Analysis of TPC-1 Cells ...... 113 4.4 Microarray Analysis of Mouse Thyroids ...... 114 4.5 Pathway Analysis using Multiplex Bead Assay ...... 117 4.6 Detection of Protein Phosphorylation in Murine Thyroids by Western Blotting ...... 120 4.7 MicroRNA (miRNA) regulation upon X-irradiation ...... 121 4.8 Western blots on Protein Levels ...... 122 4.9 Pax-8 mRNA Level After X-irradiation ...... 125 5. Discussion ...... 127 Chapter 4: External X-irradiation has an immediate impact on DNA methylation in a cell line of papillary thyroid carcinoma

1. Abstract ...... 159 2. Introduction ...... 160 3. Materials and Methods ...... 162 3.1 Cell Culture ...... 162 3.2 Irradiation of cells ...... 162 3.3 DNA extraction ...... 162 3.4 DNA Fragmentation ...... 163 3.5 Isolation of Methylated DNA using MBD ...... 163 3.6 DNA Sequencing using Illumina Solexa's Genome Analyzer II ...... 163 3.7 qPCR on bisulfite and non-bisulfite treated DNA ...... 164 3.8 Data Processing and Statistical Analysis ...... 164 4. Results ...... 166 4.1 The effect of radiation on the methylation of DNA ...... 166 4.2 The effect of radiation on the demethylation of DNA ...... 166 4.3 The effect of a high dose of radiation on the methylation of DNA ...... 166 4.4 The effect of a high dose of radiation on the demethylation of DNA ...... 166 4.5 Primer optimization for deep sequencing ...... 167 5. Discussion ...... 168 General discussion and perspectives ...... 179 Summary ...... 195

Samenvatting ...... 199 References ...... 205 CURRICULUM VITAE ...... 230

xii

xiii

List of Figures

Figure 1: A. Absorbed dose (in mGy) in the thyroid of children in two regions of the Belarus in the wake of Chernobyl ...... 5 Figure 2: Six of the major nuclear accidents in the 20th century with their ranking according to the international nuclear event scale (INES) ...... 7 Figure 3: Illustration of the human thyroid showing the two lobes connected by an isthmus ...... 9 Figure 4: Illustration of the thyroid hormone synthesis pathway ...... 12 Figure 5: The hallmarks of cancer ...... 16 Figure 6: Formation of the three fusion oncogenes: RET/PTC1, RET/PTC2, and RET/PTC3 ...... 19 Figure 7: Graph showing the Linear Non-threshold model (black) and modifications thereof (green and red) ...... 24 Figure 8: Illustration of the formation of DNA DSB foci ...... 27 Figure 9: States of proliferation and non-proliferation ...... 34 Figure 10: Pathways up-regulated by oncogenic BRAF ...... 39 Figure 11: (A) Cross-talk between DNA methylation and histone methylation. (B) Different types of DNA methylation across species ...... 45 Figure 12: Illumina Solexa's 'bridge PCR' ...... 50 Figure 13: Processing of miRNAs in the cell and their mode of action ...... 52 Figure 14: illustrating the distribution of TPC-1 cells according to their fluorescence in the PI channel (x-axis) and the FITC channel (y-axis) by a flow cytometer………………………………………………………………..67

Figure 15: DNA DSB detection using γ-H2AX immunostaining...... 73 Figure 16: A) Box plot representing median spot number per nucleus vs. irradiation dose in TPC-1 cells 30 minutes post-irradiation...... 74 B) Box plot representing spot occupancy vs. irradiation dose in TPC-1 cells 30 minutes post-irradiation ...... 74 Figure 17: Cell number vs. Irradiation dose in TPC-1 cells at 24 and 48 hours post-irradiation ...... 75 Figure 18: Cell cycle distribution by PI staining ...... 76 Figure 19: Apoptosis measurement using Annexin V/PI double staining ...... 78 Figure 20: Caspase-3 levels in response to irradiation ...... 79 Figure 21: Morphological alterations in TPC-1 cells in response to radiation ...... 81 Figure 22: Morphological alterations in BCPAP cells in response to radiation ...... 82 Figure 23: TGF-β1 levels in response to various doses of X-irradiation in TPC-1 cells ...... 84 Figure 24: Concentration of secreted cytokines by multiplex bead assay ...... 85 Figure 25: Response of various molecular markers to radiation in TPC-1 cells as measured by Western blotting ...... 87 Figure 26: Response of various molecular markers to radiation in BCPAP cells as measured by Western blotting...... 88 Figure 27: β-galactosidase level measurement in TPC-1 cells by fluorescence ...... 89 Figure 28: PCR performed on DNA from RET/PTC-positive and –negative mice ...... 112 Figure 29: Relative size of the thyroids taken from RET/PTC-negative (A) and –positive (B) mice. ... 112

xiv

Figure 30: PCA unsupervised clustering of the gene expression of RET/PTC-positive (A) and –negative (W) thyroids exposed to X-ray...... 116 Figure 31: Hierarchical clustering of genes regulated in RET/PTC-positive (A) and –negative (W) thyroids upon irradiation with X-rays ...... 117 Figure 32: Dot plots representing the mean fluorescence intensity (MFI) for 8 total and phosphorylated proteins in response to X-irradiation ...... 119 Figure 33: Western blot of total and ser15-phosphorylated p53 in RET/PTC-positive and wild-type murine thyroids ...... 120 Figure 34: Western blot of total and phosphorylated SMAD3 in RET/PTC-positive and wild-type murine thyroids...... 121 Figure 35: Venn diagram of miRNAs regulated at three X-ray doses ...... 122 Figure 36: Western blots of four selected proteins in RET/PTC-positive and –negative murine thyroids upon irradiation ...... 123 Figure 37: A)Western blots of four selected genes in TPC-1 cells upon irradiation...... 124 Figure 37: B) Protein levels of Mdm-2 were up-regulated at and above X-ray doses of 0.5 Gy while RFPL-1 (upper band) levels increased at all irradiation doses ...... 125 Figure 38: Transcriptional levels of paired homology box 8 (Pax-8) by qRT-PCR...... 126 Figure 39: Cell Cycle Pathway retrieved from KEGG database and modified to represent genes regulated in RET/PTC-positive thyroids (red), RET/PTC-negative thyroids (blue), and TPC-1 cells (green) ...... 130 Figure 40: qPCR data on bisulfite treated and non-bisulfite treated DNA ...... 172 Figure 41: Summary scheme of thesis results………………………………………………………………………………… 191

List of Tables

Table 1: Percentage of cells in each cell cycle phase in TPC-1 cells……………………………………………….…..96 Table 2: Percentage of cells in each cell cycle phase in BCPAP cells………………………………………………….97

Table 3: Primer sequences used for mouse genotyping………………………………………………………………….135

Table 4: Pathways responsive to radiation……………………………………………………………………………………..136

Table 5: Genes commonly regulated upon irradiation in two RET/PTC systems………………………………141

Table 6: Some genes common to RET/PTC and wild-type thyroids………………………………………………….142

Table 7: MiRNAs regulated upon irradiation in TPC-1 cells……………………………………………………………..144

Table 8: MiRNAs regulated upon irradiation in TPC-1 and their gene targets………………………………….145

Supplement 1: Genes regulated due to radiation in TPC-1………………………………………………………………147

Table 9: Genes demethylated due to radiation in TPC-1…………………………………………………………………172

Table 10: Genes methylated due to radiation in TPC-1………….……………………………………………………….172

Table 11: Genes methylated due to radiation (high dose only)……………………………………………………….173

Table 12: Genes Demethylated due to radiation (high dose only)…………………………………………………174

Table 13: Primer sequences…………………………………………………………………………………………………………..175

xvi

xvii

List of Abbreviations

µM: Micromolar

5meC: Five methyl-cytosine

AC: Adenylyl cyclase

AIT: Apical iodide transporter

ALARA: As low as reasonably achievable

ANOVA: Analysis of variance

AT: Ataxia telangiectasia

ATM: Ataxia telangiectasia mutated

AURKA: Aurora kinase A

Bcl-2: B-cell lymphoma 2

BEIR: Biological effects of ionizing radiations

BER: Base excision repair

Bp: Base pair

Bq: Becquerel

BRCA1: Breast cancer type 1 cAMP: Cyclic adenosine monophosphate

CDH1: Cadherin 1

CDKI: Cyclin dependent kinase inhibitor cDMR: DNA-methylated regions

CFTR: Cystic fibrosis transmembrane conductor regulatory

CGI: CpG islands cGy: Centigray

Ci: Curie

CI: Confidence interval

CITED1: Cbp/p300 interacting transactivators with glutamic acid [E] and aspartic acid [D]-rich C- terminal domain

CLC-5: Cl-/H+ antiporter

CLL: Chronic lymphocytic leukemia

CO2: Carbon dioxide xviii

Cu: Copper

DAC: 5-aza-2‟-deoxycytidine

DDR: DNA damage response

DEHAL: iodotyrosine dehalogenase

Dio1: Deiodinase 1

DMEM: Dulbecco's modified essential medium

DNA: Deoxyribonucleic acid

DNMT1: DNA methyl transferase 1

DSB: Double strand break

DUOX: Dual oxidase

EC: European commission

EDTA: Ethylenediaminetetraacetic acid

EGF: Epidermal growth factor

ELISA: Enzyme linked immunosorbent assay

ER: Estrogen receptor

ERR: Excess of radiation risk

Et al.: et alii

FBS: Fetal bovine serum

FDR: False discovery rate

FMTC: Familial medullary thyroid carcinoma

GDNF: Glial cell line-derived neurotrophic factor

GDF: Growth differentiation factor

GFRα: GDNF-family α receptor

GM-CSF: Granulocyte macrophage colony stimulating factor

GNEF: Guanine nucleotide exchange factor

GPCR: G-protein coupled receptor

GST: Glutathione S-transferase

Gy: Gray

H2O2: Hydrogen peroxide

H2SO4: Sulfuric acid

xix

H3K9Me: Methylated lysine 9 on histone 3

HAT: Histone acetyltransferase

HCl: Hydrochloric acid

HDAC: Histone deacetylase

HDF: Human diploid fibroblast

HDGS: Homology-dependent gene silencing

HPS: Hyperplastic polyposis syndrome

HR: Homologous recombination

HRP: Horseradish peroxidase

HRS: Hyperradiosensitivity

ICRP: International commission on radiological protection

IGFBP: Insulin-like growth factor binding protein

IL: Interleukin

INES: International nuclear event scale

IRR: Increased radioresistance

JNK: c-jun NH2-terminal protein kinase kBq: Kilobecquerel kV: Kilovolt

L: Liter

Let: Lethal

LET: Linear energy transfer

LINE-1: Long interspersed nuclear element 1

LNT: Linear non-threshold

LOH: Loss of heterozygosity

LRP1B: Lipoprotein receptor-related protein 1B

LSS: Life span study mA: Milliampere

MAPK: Mitogen activated protein kinase

MBD: Methyl-cytosine binding domain

MBP: Methyl-cytosine binding protein

MCP-1: Macrophage chemotactic protein-1

MCT8: monocarboxylate transporter 8

MDM2: Mouse double minute 2

MeDIP: Methylated DNA immunoprecipitation

MEN: Multiple endocrine neoplasia

MFI: Mean fluorescence intensity mGy: Milligray miRNA: Micro RNA

MMP: Matrix metalloproteinase mRNA: Messenger RNA

MSP: Methyl-specific polymerase chain reaction

Mt: Megaton mTOR: Target of rapamycin

NADPH2: Nicotinamide adenine dinucleotide phosphate NaOH: Sodium hydroxide

NER: Nucleotide excision repair

NF-κB: Nuclear factor kappa B

NGF: Nerve growth factor

NGS: Next generation sequencing

NIS: Sodium iodide symporter

NPP: Nuclear power plant

PBMC: Peripheral blood mononuclear cell

PBS: Phosphate buffered saline pCi: Picocurie

PCR: Polymerase chain reaction

PDGF: Platelet derived growth factor

PI: Propidium iodide

PI3K: Phosphatidylinositol 3-kinase

PKC: Protein kinase C pM: Picomolar

xxi

PMSF: Phenylmethylsulfonyl fluoride

PP: Proliferation potential

PTC: Papillary thyroid carcinoma

PVDF: Polyvinyl difluoride qPCR: Quantitative PCR

RASSF: Ras association domain family

Rb: Retinoblastoma

RET: Rearranged in Transformation

RFPL1: Ret finger protein like 1

RIN: RNA integrity number

RISC: RNA-induced silencing complex

RIZ-1: retinoblastoma protein-interacting zinc finger 1

ROS: Reactive oxygen species

RPMI: Roswell park memorial institute

RS: Replicative senescence rT3: Reverse T3 RTK: Receptor tyrosine kinase

SA β-gal: Senescence-associated beta galactosidase

SAHF: Senescence-associated heterochromatic foci

SASP: Senescence-associated secretory profile

SBS: Sequence by synthesis

Ser: Serine

SIPS: Stress-induced premature senescence

SMCT-1: Sodium monocarboxylate transporter 1

SOCS1: Suppressor of cytokine signaling 1

SOD: superoxide dismutase

SSB: Single strand break

STAT: Signal transducer and activator of transcription

Sv: Sievert

T3: Triiodothyronine

xxii

T4: Tetraiodothyronine TBq: Terabecquerel

TF: Transcription factor

Tg: Thyroglobulin

TGF-β1: Transforming growth factor beta 1

TIS: Tumor-induced senescence

TMI: Three Mile Island

TPO: Thyroid peroxidase

TSG: Tumor suppressor gene

TSH: Thyroid stimulating hormone

TSHR: Thyroid stimulating hormone receptor

TSLC1: Tumor suppressor in lung cancer 1

TTF-1: Thyroid transcription factor 1

Ub: Ubiquitin

UK: United Kingdom

USA: United States of America

UVB: Ultraviolet type B

VEGF: Vascular endothelial growth factor

XP: Xeroderma pigmentosum

xxiii

Chapter 1: General Introduction

General Introduction

R ~ i.a

R = amount of rumor in circulation. i = importance of the rumor to the person who hears or reads it. a = the level of ambiguity or uncertainty surrounding the rumor.

Allport and Postman's basic law of rumor

Chapter 1

1. The Nuclear Conundrum The transition of nuclear energy from a harmful weapon to a friendly energy source and back to a dangerous foe in the minds of the people was achieved within a span of 40 years only.

The events that have contributed to this shift in people's perception were either intentional

(the atomic bombing of Hiroshima and Nagasaki) or accidental (Three Mile Island and

Chernobyl accidents) and all have been linked to an increase in cases of cancer in local populations.

Almost 25 years after the accident of Chernobyl, a new accident at the Dai'ichi unit of

Fukushima Nuclear Power Plant has revived fears of nuclear power. The latest accident in

Fukushima shares some features with the previous two nuclear power plant accidents (e.g. all involved radioactive iodine release and were caused mainly by lack of proper cooling of the reactor); but is also different in that the first two were put down to human error, while the third was due to a natural disaster [1-3].

The release of radioactive iodine, a product of the fission of uranium which is used as fuel in nuclear power plants, is always of concern because it is quickly taken up by the thyroid via the sodium iodide symporter which does not distinguish between iodine isotopes [4]. The accumulation of radioactive iodine in the thyroid can then cause appearance of thyroid cancer due to the emission of β and γ rays [5].

1.1 TMI and Chernobyl: Compare and Contrast Although the accident at Three Mile Island (TMI) was considered at the time to be the worst nuclear accident ever, the amount of radioactive iodine released was minimal (around 20 Ci) and the elevation in the levels of radioactive iodine was observed in only selected samples of milk from surrounding farms but only averaged 36 pCi/L way below the US environmental protection agency's maximum of 12,000 pCi/L [2]. These numbers are in stark contrast to the

General Introduction

Chernobyl nuclear power plant accident in which the core lay wide open and where the activity due to radioactive iodine was estimated at 4.86x107 Ci which made its way into the soil and food cycle [6]. This difference in radioactive iodine release could be responsible for the difference in the absorbed dose in the thyroids of children of the areas surrounding the site and subsequently for the increase in cases of thyroid cancer in the case of Chernobyl and not

TMI. Average doses to the thyroid in the case of the Chernobyl accident have been reported as ranging from a low dose of 0.09 Gy to as high as 12.5 Gy in the populations surrounding the site [1]. Five years after the Chernobyl accident, the latency time for thyroid cancer in children, an increase in the cases of diagnosed thyroid cancer was observed in children in the surrounding areas of Chernobyl, especially Gomel in the Belarus [7]. Afterwards, an increase in thyroid cancer was observed in adolescents and adults. In total and to date, between 4000 and 6000 cases of thyroid cancer have been attributed to the accident at Chernobyl and are thought to be mostly due to the radioactive iodine release which was quickly incorporated into the children's thyroids due to the state of iodine deficiency in the population [8].

Chapter 1

Figure 1: A. Absorbed dose (in mGy) in the thyroid of children in two regions of the Belarus in the wake of Chernobyl. The Gomel region showed consistently higher doses then other regions (Smith and Beresford, Chernobyl Catastrophe and Consequences, 2005). B. Picture of the destroyed reactor four in the Chernobyl NPP (Mould, Chernobyl Record: The Definitive History of the Chernobyl Catastrophe, 2000). C. The incidence of thyroid cancer as number of cases per 100,000 people in the Belarus in the years 1986 through 2002 (Reiners, Radioactivity and Thyroid Cancer, Thyroid, 2009). D. Two Gryvnya commomorative coin issued by Ukraine to mark the 10th anniversary of the Chernobyl disaster (adapted from Mould, Chernobyl Record: The Definitive History of the Chernobyl Catastrophe, 2000). 1.2 Nuclear Weapons Testing The Chernobyl accident, although regrettable, has offered scientists a unique opportunity to study the effect of severe radioiodine contamination on appearance of thyroid cancer and the mechanism by which that is achieved. Before Chernobyl the greatest man-made source of release of radionuclides was the atmospheric nuclear weapons testing which, because it was controlled and usually got dispersed in non-inhabited areas, did not lead to mass population contamination. An exception was the Marshall Island nuclear weapons testing conducted by the US on Bikini and Enewetak atolls between 1946 and 1958. On the 1st of March 1954, a

15-Mt thermonuclear device was detonated and because of the wind shear, considerable

General Introduction

amount of nuclear fallout occurred on the nearby islands of Rangelop and Utrik [9]. It has been estimated that the adult uptake of I-131 in the population of Rongelap was around

3.7x103 kBq and that the average dose to the thyroid of the population due to the same isotope was 4200 mGy [9, 10]. This was found to lead to an increased risk of thyroid cancer which was projected to have led to 35 excess cases of thyroid cancer in the period from 1948 to 2008

[11]. However, the epidemiological studies were limited by the fact that the Rangelop and

Utrik island population at the time of the accident was small.

1.3 Fukushima: Lessons from the Past The recent accident in Fukushima seems to lie in between the two accidents in TMI and

Chernobyl when it comes to its risk to the thyroid. Hydrogen buildup in the reactor caused an explosion that removed one of the two containment roofs and led to the release of significant amounts of radioactive iodine [12]. The amount of radioactivity released by the Fukushima accident has been estimated to range between 370,000 and 630,000 TBq (9.7x106 to 1.7x107

Ci) but is still considered to be one tenth the activity released from the Chernobyl reactor [3].

Given this fact, the fact that most of the radioactive plume moved on to the Pacific Ocean, the swifter response of the Japanese authorities compared to the Soviet, and the richness of the

Japanese diet with iodine, the accident in Fukushima does not seem to pose as big a risk for the development of thyroid cancer as did the accident in Chernobyl [13]. Nonetheless, the

Fukushima accident has been rated as a 'seven', the worst rating on the INES scale, putting it on level with the Chernobyl accident [14].

Chapter 1

Figure 2: Six of the major nuclear accidents in the 20th century with their ranking according to the international nuclear event scale (INES). The INES scale ranges from zero to seven with seven being the highest (Mould, Chernobyl Record: The Definitive History of the Chernobyl Catastrophe, 2000). 1.4 External Irradiation Internal contamination with radioactive iodine is not the only risk for thyroid cancer, however, as has been demonstrated by such incidents as the atomic bombing of Hiroshima and Nagasaki and children who were irradiated for medical purposes. In the case of

Hiroshima and Nagasaki, an increased risk for the development of benign thyroid nodules and thyroid cancer with dose was found in children who were irradiated when below 15 years of age and in utero [15-17]. In the 1950s, children from North Africa migrating to Israel received

X-rays to the scalp to rid them of the ringworm Tinea capitis. In the course of this treatment, the children received a low to moderate dose of radiation to the thyroid (between 4.5 and 49.5 cGy). In several follow-up studies, an increase in incidence of thyroid cancer and an increased risk for the development of thyroid cancer could be demonstrated. In fact, the excess of radiation risk (ERR) per Gy was estimated by one study to be as high as 20.2 which may point to the fact that the genetic background can also play a role in radiation sensitivity [18-21]. It needs to be pointed out that the ERR/Gy of 20.2 was obtained on a population of males and females and which contains subjects whose ages ranged from one to 54 years so can be assumed to be free of gender and age bias [18]. Nonetheless, a decrease of ERR/Gy was observed with age when the population was broken down into age categories [22]. A review 7

General Introduction

of other studies on radiation and thyroid cancer in children reported high ERR per Gy. These studies included a study on children in New York following scalp irradiation where the average thyroid dose was 60 mGy, the enlarged thymus study where the average thyroid dose was 1.4 Gy, and the French hemangioma study where the dose to the thyroids was 39 mGy.

The ERR per Gy for thyroid cancer development was estimated in one study to be as high as

23, a number comparable to the Tinea capitis study [393]. However, later studies have reached the lower number of 5.25 with a 95% confidence interval (CI) of 1.7 to 27 [22]. One possible reason that the risk to the thyroid after irradiation is consistently higher in children may lie in the fact that the thyroid is more active in children than in adults as it is more involved in growth and development.

Chapter 1

2. The Thyroid The thyroid is a highly vascularized endocrine gland that is involved in the secretion of the hormones that influence aspects of reproduction, growth, differentiation, and metabolism. The thyroid is mainly made up of two types of cells, the follicular cells, that are involved in concentrating iodine and T3 and T4 hormone secretion, and the parafollicular C cells which are mainly involved in the secretion of the calcium-regulating calcitonin [23].

Figure 3: Illustration of the human thyroid showing the two lobes connected by an isthmus. A cross-section of the human thyroid with three follicles formed by thyroid follicular cells around a colloid. An enlargement of one follicular cell showing the formation of colloid droplets on the apical side (Norris, Vertebrate Endocrinology, 4th ed., 2007).

2.1 Iodine Metabolism and Hormone Synthesis The follicular cells are polarized cells with a basal side that is involved in iodine entry from the bloodstream and hormone release into the blood and an apical side that is involved in thyroid hormone synthesis. Iodine which is the main component of the two thyroid hormones,

General Introduction

triiodothyronine (T3) and tetraiodothyronine (T4), enters the follicular cells through the sodium iodide symporter (NIS) as iodide (I-) and is transported to the apical side where it is transferred to the colloid by the transporter slc26A4, otherwise known as pendrin, where iodide is organified by thyroid peroxidase (TPO). Organified iodide is subsequently added onto tyrosines of the thyroid-specific protein thyroglobulin (Tg) to form 3- monoiodinetyrosine (MIT) and diiodotyrosine (DIT). Linking two DITs together results in the hormone T4 while the linking of MIT and DIT results in T3 [23-25]. The movement of iodide across the apical side of the membrane is now thought to be carried out by other anion carriers found on the apical side of the thyrocyte. These carriers are thought to work in tandem or to affect each other in some ways although the true nature of their role in the thyroid is still controversial. These carriers are the sodium monocarboxylate transporter (SMCT)-1 (aka human apical iodide transporter [hAIT] or slc5A8), Cl-/H+ antiporter (CLC-5), and cystic fibrosis transmembrane conductance regulatory (CFTR) [26]. TPO activity depends upon the existence of reactive oxygen species (ROS) such as hydrogen peroxide (H2O2). In the thyroid,

H2O2 is generated by the enzymes dual oxidase (Duox) 1 and 2 from oxygen using nicotinamide adenine dinucleotide phosphate (NADPH2) as a coenzyme [27, 28]. Due to the presence of oxidative challenge, the thyroid is equipped with means to effectively deal with

ROS such as the enzymes superoxide dismutase (SOD), catalase, and glutathione peroxidase

[29-31]. This would mean that the thyroid is perhaps better equipped to handle the challenge of radiation-generated ROS than other organs. However, an increase in the levels of H2O2 due to an increased intake of iodide or a decrease in iodine availability has been intimately linked to goitrogenesis and inflammation [32, 33].

The thyroid mainly synthesizes T4 which is thought to be the precursor for the active hormone form, T3, although the thyroid does synthesize both. The hormones are reabsorbed into the

Chapter 1

follicular cell by a mechanism that is still not entirely clear but probably involves pinocytosis of Tg and its breakdown by cathepsins after which the enzyme deiodinase 1 (Dio1) removes one iodine from T4 to convert it to either T3 or reverse T3 (rT3). The removed iodine is taken back to the apical side where it is again involved in the synthesis of the hormone T4 [34]. The hormones then leave the follicular cell by a process that is also not well elucidated but probably involves a carrier system. It was long thought that the thyroid hormones traverse the lipid bilayer passively. However, it is now thought that the product of the slc16a2 gene, monocarboxylate transporter 8 (MCT8) is involved in the active transport of the hormones out of the thyrocyte [35, 36].

In addition to the activity of the iodotyrosine deiodinases that remove iodines from T4, there are the iodotyrosine dehalogenase (DEHAL) enzymes that remove iodines from MIT and

DIT. There exist two forms of this enzyme, namely DEHAL1 and DEHAL1B which differ by a 127 bp insert [37, 38]. Mutations in the DEHAL enzymes have been linked to various disorders and iodine deficiency [39, 40].

The hormones circulate in the bloodstream and exert their permissive action through binding to the nuclear thyroid receptors (TR) found in target cells. Many cells are also equipped with

Dio1 enzymes as well as two other members of the same family, Dio2 and Dio3 [41]. The enzymes play a role in converting T4 to its active form T3. Finally, the thyroid hormone receptors are nuclear receptors so the thyroid hormones have to first traverse the plasma membrane to reach the receptors. It was long thought that the thyroid hormones cross the membrane by passive diffusion. However, it is now known that only 10% of the hormones enter by diffusion while the majority are actively transported by the solute carrier MCT8 [38,

42, 43].

General Introduction

Figure 4: Illustration of the thyroid hormone synthesis pathway with iodide entry happening on the basal side (number 1) and release of the hormones T3 and T4 at number 7 (Dunn and Dunn, Update on Intrathyroidal Iodine Metabolism, Thyroid, 2001).

Chapter 1

2.2 Thyroid Homeostasis The control of iodine uptake in the thyroid is mediated mainly by the thyroid stimulating hormone (TSH) through its binding to the thyroid stimulating hormone receptor (TSHR) which is a G-protein coupled receptor (GPCR). This binding induces thyroid growth and proliferation and an increase in iodine uptake by the NIS. Actually it was found that the proper functioning of TSHR is important for the proper trafficking of NIS to the basal membrane. On the other hand, overwhelming the thyroid with excess iodine will cause the thyroid to shut down its uptake of iodine in a process known as the Wolff-Chaikoff block.

This is the basis upon which the potassium iodide (KI) tablets were issued to Polish residents after the authorities were informed of the Chernobyl accident [23, 24, 44-46]. Papillary thyroid carcinomas (PTCs) usually lose their TSHR since they have over-expressed a parallel growth pathway making the TSHR unnecessary. It has been found that PTCs show an increased expression of NIS but its uptake of iodine is still lower than that of normal thyrocytes. This could be attributed to the loss of TSHR expression which causes cytoplasmic accumulation of NIS [44]. The downregulation of NIS is also induced by an excess of iodine while a lack of iodine in the diet will elicit an increase in NIS levels in the thyroid. It is now thought that the iodine-related downregulation of NIS is induced at a post-transcriptional manner [47]. The binding of TSH to its receptor mediates the formation of cyclic adenosine monophosphate (cAMP) via the action of the enzyme adenylyl cyclase (AC). This, among other things, activates protein kinase C (PKC) which is found to inhibit iodine metabolism. It also causes the up-regulation of thyroid differentiation markers such as NIS, TPO and even

DEHAL1 [40, 48]. On the other hand, the increase in cAMP and insulin increase iodide uptake and organification [48]. To add a further layer of protection to the thyrocyte, it was reported that TPO and DUOX are inhibited by hydrogen peroxide. The interaction of the two enzymes at the plasma membrane protects DUOX from inhibition by H2O2 [49].

General Introduction

"And the third angel sounded, and there fell a great star from heaven, burning as it were a lamp, and it fell upon the third part of the rivers, and upon the fountains of waters; And the name of the star is called Wormwood; and the third part of the waters became wormwood; and many more died of the waters; because they were made bitter." Book of Revelations, 8:10-11 (King James Bible) "Третий ангел вострубил, и упала с неба большая звезда, горящая подобно светнльнику, и пала на третью часть рек и на источники вод. Имя сей звезде "полынь" и третья часть вод сделалась полынью, и многие из людей умели от вод, мотому что они стали горьки" Book of Revelations, 8:10-11 (Russian Synodal Text)

Chapter 1

3. Thyroid Cancer Although thyroid cancer is the most prominent endocrine malignancy, it is still a rare event, constituting less than 1% of human malignant neoplasms [50]. The incidence and prognosis of thyroid cancer is very much age-dependent with an increasing incidence with age and a more favorable outcome with younger age [51]. Other influencing factors are the state of iodine sufficiency, gender, and genetic background. For example, goitrogenesis, an enlargement of the thyroid due to excessive growth, and thyroid cancer increase with decreasing amounts of dietary iodine while some populations with a certain background (e.g. Jewish people) are at an elevated risk of developing thyroid cancer [50]. Most notable is the finding that females are more at risk to develop thyroid cancer with an average male : female ratio of 1:2.25. This ratio is region dependent and, and for example, ranges from 1:0.9 in Belgium to 1:4.1 in

Denmark [52]. Recent findings indicate that estrogen plays a role in thyroid cancer by signaling through estrogen receptor (ER) α on thyrocytes and thus could explain the disparity in cancer risk between females and males [53-55].

3.1 Thyroid Cancer Types Broadly speaking, thyroid-specific cancers can be classified into two types based on their cell of origin: papillary and follicular carcinomas arising from the follicle cells and medullary carcinomas arising from the parafollicular C cells [56]. The golden standard for identifying tumors, their diagnosis and grading remains histological despite increasing research into genotypic and molecular imaging classification [57-62]. Based on histology, cancers can be further divided into adenomas, carcinomas, and anaplastic [57]. The most common type of thyroid cancer is the papillary carcinoma (40-80%) followed by follicular (10-40%), anaplastic (2-14%), and medullary (1-10%) [63]. Because of its predominance in the cases of thyroid cancer due to irradiation, and as this doctoral thesis focused on this type of cancer, the rest of this section will focus on papillary thyroid carcinoma (PTC).

General Introduction

PTCs are highly differentiated cancers which are identified histologically mainly by their nuclear characteristics called 'ground-glass nuclei'. They are described by Hofstädter (2005) as "enlarged, round-to-oval structures with a pale karyoplasm condensing continuously to the nuclear membrane" [56]. One of the most fundamental characteristics of cancer cells is their ability to maintain chronic proliferation [64, 65] and to this aim PTCs are characterized by specific mutations that are considered initiating steps in cancer. The most important oncogenic mutations in PTCs are the RET rearrangements, NTRK1 rearrangements, RAS and

BRAF mutations [66].

Figure 5: The hallmarks of cancer (Modified from Hanahan and Weinberg, Hallmarks of Cancer: The Next Generation, Cell, 2011). 16

Chapter 1

3.2 Mutations in PTC

3.2.1 RAF RAF is a family of serine/threonine kinases which includes the three members ARAF, BRAF and CRAF and are located downstream the RAS guanine nucleotide exchange factor (GNEF) and upstream the mitogen activated protein kinases (MAPK) [67]. Normally, RAS is activated upon binding of a ligand to a tyrosine receptor kinase (e.g. RET). Subsequently, RAS activates BRAF which in its turn activates the MAPK cascade. However, mutations in BRAF that lead to its constitutive activation cause the oncogenic ligand-independent activation of the

MAPK pathway. Activating point mutations in BRAF have been found on average in 45% of

PTCs [68-73] which made it a lucrative candidate for PTC detection and therapy [74, 75]. The most prevalent BRAF mutation in PTC is the thymine to adenine transversion at position

1799 (T1799A) leading to a substitution of valine with a glutamate at position 600 (V600E)

(previously thought to be a T1796A mutation leading to V599E) [71, 74, 76]. This mutation has been associated with a poor prognosis and a metastatic profile [72].

3.2.2 RET/PTC The RET protooncogene is a receptor tyrosine kinase (RTK) expressed mainly on neural crest-derived and urogenital cells. It forms a docking site for ligands of the glial cell line- derived neurotrophic factor (GDNF) in addition to members of the nerve growth factor (NGF) family [76, 77]. Upon ligand binding, a signaling complex is formed from the ligand-receptor complex and co-receptor (members of the GDNF-family α receptor) which leads to RET receptor dimerization and phosphorylation creating docking sites for binding adaptors that convey the signal downstream. Many phosphorylated tyrosines on the intracellular side of

RET play a role in downstream signaling such as Y752, Y905, Y928, Y981, Y1015, Y1062, and Y1096. However, Y1062 remains among the most important as it is involved in binding

DOK1/4/5, Enigma, FRS2, IRS1/2, Shc, and ShcC. In addition, this tyrosine is involved in

General Introduction

activation of the Ras/ERK, PI3K/AKT, nuclear factor κB (NFκB), and c-Jun NH2-terminal protein kinase (JNK) [78].

Activating point mutations in the RET protooncogene were found to be an initiating step for all autosomal dominant multiple endocrine neoplasia (MEN) 2 cancers such as MEN2A,

MEN2B and familial medullary thyroid carcinoma (FMTC) and thus provided evidence that

RET mutations were important in thyroid cancer initiation [77]. However, it was discovered that the more important form of RET activation was through a translocation which fuses the

RET tyrosine kinase domain to the 5' end of a gene called D10S170 (previously H4, now

CCDC6). This translocation was then found to be the PTC (for papillary thyroid carcinoma) oncogene and was thus dubbed the RET/PTC1 translocation [79, 80]. The other most prevalent translocation in PTC, formed by the fusion of the TK domain of RET with 5' terminal of RFG/ELE1/NCOA4 and termed RET/PTC3, was discovered afterwards by the same group [80, 81]. Since then 10 other fusion partners were found bringing the number of

RET/PTC translocations to 12 [77], however, RET/PTC1 and RET/PTC3 remain the two most important translocations in PTC. Although RET is not expressed in thyroid follicular cells but only in parafollicular C-cells, the RET/PTC oncogene is found in papillary thyroid carcinomas almost exclusively [82] and is targeted to the cytoplasm and not the plasma membrane. These factors could contribute to the oncogenic potential of RET/PTC translocations as well as the overexpression of several pathways related to cellular growth and proliferation such as the MAPK, JNK, and phosphatidylinositol 3-kinase (PI3K) signaling pathways [83-85]. Of the 12 autophosphorylation sites described in RET, 11 are maintained in the RET/PTC fusion protein, includingY1062 [86].

RET/PTC translocations have been found to be sufficient to form an initiating step in PTC as was demonstrated in in vivo studies in mice. When a RET/PTC transcript was introduced into

Chapter 1

a mouse model under the control of the thyroglobulin (Tg) promoter, thyroid tumors with

PTC morphology appeared later in the life of these mice [87]. However, the fact that not all mice develop PTC after the introduction of the RET/PTC translocation means that there are other events necessary for the progression to cancer which supports the multistep model of carcinogenesis in PTC. The RET/PTC translocation is found in around 15-20% of cases of sporadic adult PTC with a range of 5 to 40% which makes it the second most common mutation after the BRAF point mutation described earlier [88].

Figure 6: Formation of the three fusion oncogenes: RET/PTC1, RET/PTC2, and RET/PTC3. Modified from Tallini and Asa, RET Oncogene Activation in Papillary Thyroid Carcinoma, Adv. In Anatomic Pathology, 2001.

3.2.2.1 Prominance of RET/PTC Translocations in Radiation-induced PTCs As mentioned in chapter one, there was an increase in the cases of thyroid cancer in the wake of the Chernobyl disaster in the children of the communities surrounding the Chernobyl nuclear power plant [89, 90]. These cases were mostly of PTC and a striking feature of these cancers was that the RET/PTC translocation was found in around 66-87% of the tumors 19

General Introduction

diagnosed making it the most common mutation in PTCs in these children. This shift in the prominence of RET/PTC translocations was also observed in PTCs from children irradiated for medical reasons and indicated that the increase in DNA double strand breaks (DSB) induced by irradiation probably increased the likelihood that a translocation occurred [91, 92].

This fact has been nicely illustrated by Gandhi et al. (2010) who demonstrated that the RET and H4 (CCDC6) genes were located on fragile sites on chromosome 10 (FRA10C and

FRA10G) and that fragile site-inducing conditions, mimicking irradiation, caused the appearance of the RET/PTC1 translocation and not RET/PTC3 [93]. This however does not explain why the RET/PTC3 was the most dominant translocation after Chernobyl, warranting further research. Perhaps the DNA DSBs in the fragile sites in chromosome 10 that contribute to RET/PTC1 can be more easily repaired in time before they can cause the translocation.

Although RET/PTC3 signals on the same axis as RET/PTC1, it is described as conferring a more aggressive metastatic solid-type papillary carcinoma with distinct morphological features as compared to RET/PTC1-positive PTC [94]. To further illustrate the role of

RET/PTC translocations in radiation-induced PTCs, a study by Caudill et al. (2005) described a dose-dependent increase in RET/PTC translocations in response to γ rays, while follow-up studies on children irradiated for Tinea capitis revealed an increase in said translocations [95,

96]. Furthermore, when the transcriptomic profile of sporadic and Chernobyl irradiation- induced PTCs was compared, it was found that Chernobyl-related PTCs possessed a unique transcriptomic signature that made it possible to distinguish them from sporadic PTCs. This work was carried out by three separate labs on different samples and all came to the same conclusion which means that in addition to the variance in the percentage of RET/PTC-Ras-

BRAF mutations, other events play a role in sporadic and radiation-induced PTCs [97-99].

This unique transcriptomic profile does not only exclusively hold for Chernobyl-related PTCs

Chapter 1

but also extends to children who were irradiated for therapeutic purposes in the head and neck area [100].

3.2.3 NTRK Rearrangments Although RET/PTC translocations are the most prominent in radiation-induced PTCs, other translocations are and continue to be described in the literature. Another RTK, namely

NTRK1, which encodes for a nerve growth factor receptor is also involved in oncogenic transformation. Fusion of the 5' end of various genes to the 3' end of NTRK1 creates several oncogenic recombinant proteins designated as TRK-T1 – T4. TRK-T1 for example is formed by the fusion of the 3' end of NTRK1 to the 5' end of TPR and was found to induce neoplastic transformation of the thyroid epithelium [101, 102].

3.2.4 Ras The fourth type of mutations described in PTCs is the Ras mutation which can involve any of the three members, K-RAS, N-RAS, or H-RAS, although mutations in N-RAS seem to be predominant. Several point mutations in codons 12, 13, and 61 cause a constitutively active form of the RAS protooncogene and an overactivation of the BRAF-ERK-MAPK pathway.

This mutation is usually found in only less than 10% of PTCs [66, 76, 82].

A notable feature of the above-mentioned RAS, BRAF, and RET/PTC mutations in PTC is their mutual exclusivity in that a PTC which is found to harbor a RET/PTC translocation does not contain a BRAF point mutation and vice-versa. This has been described by many groups that looked into these mutations in clinical samples and can be explained by the fact that those proteins signal down the same axis and thus any one mutation is sufficient to overexpress the signaling pathway making another mutation redundant [71, 72, 74, 103].

The above-mentioned mutations cause the over-expression of several pathways that are involved in carcinogenesis but are by no means the only events leading to PTC.

General Introduction

Chapter 1

4. Low Dose Radiation Introduction It is a well-established fact that radiation causes damage to DNA which can be classified into damage to one strand of the DNA molecule, i.e. depurination and depyrimidation, alkylation, oxidative lesions, deaminations, and single strand breaks (SSB), or damage to both strands or double strand breaks (DSB) which is the most harmful. DNA damage can either be repaired, cause cell death or senescence, or the cell can just resume proliferation with the damage. If this confers a proliferation or survival advantage, a cell can become cancerous and this is what is usually meant by health risks associated with radiation. This risk is proportionally linear with absorbed radiation dose at a dose above 100 mSv where effects are mostly deterministic. However debate still rages over the effect of low doses of radiation which are usually identified as any dose below 100 mSv and where stochastic effects prevail [104].

4.1 The Linear Non-Threshold Model (LNT) The most widely accepted model for the prediction of radiation risk is the linear non-threshold

(LNT) model that is adopted by the Biological Effects of Ionizing Radiations (BEIR) VII report of the US National Academy of Sciences and the International Commission on

Radiological Protection (ICRP). This model states that the relationship between health risk and radiation dose is linear and therefore there is no such thing as a „safe‟ dose and is behind the As Low As Reasonably Achievable (ALARA) principle. This model has been widely used except for some cases as in leukemia and exponential cell-sterilization where a linear- quadratic model was employed. This LNT model has been contested by several researchers and agencies as either over- or underestimating the risk to humans. The French Academies report has even claimed that the LNT model is not based on scientific evidence. Opponents of the LNT model claim that if a threshold were to be set, then great amounts of money could be saved on precautionary methods [105].

General Introduction

Figure 7: Graph showing the Linear Non-threshold model (black) and modifications thereof (green and red). Mullenders et al. Assessing Cancer Risks of Low-dose Radiation, Nat. Rev. Cancer, 2009 The LNT model is based on data gathered from the survivors of the atomic bombs of

Hiroshima and Nagasaki (the Life Span Study [LSS]), of gathered epidemiologic data, of workers in the nuclear industry, and of people exposed environmentally [104]. The most important however remains the data from the atomic bomb survivors since those constituted a very large cohort of mixed sexes for whom accurate dosimetric data were found and were followed up for their whole lifetime since the bomb. Also of great value is the fact that around

80% of those survivors were irradiated with what is considered to be low doses (i.e. under 0.1

Sv). Although non-solid tumors (i.e. leukemias) and sarcomas showed a non-linear relationship (curvilinear), other solid cancers displayed a linear pattern down to 0.2 Sv and thus this was employed for all cancers [106, 107]. However, for doses less than 0.1 Sv, many other factors come into play that could interfere with the proper interpretation of results and thus some scientists claim that epidemiologic studies lack statistical power and are less 24

Chapter 1

helpful than molecular ones. Yet for others, distinction can be made between three different regions: doses above 1 Gy (when speaking about X-rays, Grays and Sieverts are interchangeable; however for alpha particles and neutrons, Gy = Sv/radiation quality factor) are defined as high doses; doses down to 10 mGy which are defined as low doses; and then doses from 10 mGy to below 1 mGy, defined as very low doses. For the latter, they claim that the resolution of assays would never be that they can assess risk and therefore there can be no other solution beyond extrapolation from higher doses [108]. The argument goes however, that since the type of DNA damage at very low doses is quite similar to that at low doses with only the number of affected cells being altered, LNT extrapolation from low to very low doses is more feasible than from high to low. However, the LNT model here can predict mechanistic responses and not cancer risks.

4.2 The LNT: Pros and Cons, Epidemiology Although the follow up study on the survivors of the atomic bombs of Hiroshima and

Nagasaki remains the biggest and some claim likely to be unparalleled, other smaller epidemiological studies have either confirmed or contested the LNT model [109]. For example, the risk of breast cancer after low doses of radiation suggests linearity as well as leukemia incidence in children following radiation for medical purposes [110-112]. Whereas follow up studies on Chernobyl liquidators detected an increased risk of leukemias, excluding chronic lymphocytic leukemias (CLL), only at doses above 200 mSv, a small increase in cancer risk was found even at low doses of radiation in nuclear workers from a study involving 15 countries [113, 114]. Again, there was no link between irradiation of the nuclear workers and CLL [115]. Increased risk of leukemia has been established by monitoring workers in the nuclear field in the UK and the USA while an increase in cases of lung cancer due to inhalation of radon gas was found in Europe [116-118]. On the other hand, Tubiana et al. and Averbeck recount several epidemiological examples that counter the above such as the 25

General Introduction

absence of excess risk of any other cancers besides thyroid malignancies in Chernobyl victims, the absence of excess cancer risk in populations exposed to high background irradiation, and the absence of excess cancer risk due to radon inhalation in Japan [105, 119].

However, these studies have been criticized for lack of statistical power which highlights one of the problems with assessing risk at low doses: it requires a very large study with long-term follow-up [110].

4.3 Health Risk Estimates: Molecular

4.3.1 γH2AX One of the most studied markers of DNA damage is the phosphorylated form of the conserved histone H2AX which is known as gamma H2AX (γH2AX). This histone is phosphorylated on

Ser139 upon detection of DNA DSB and serves as a signaling platform for DNA repair molecules and subsequently in the ataxia telangiectasia mutated (ATM) autophosphorylation by the MRN complex [120]. Phosphorylation of H2AX happens seconds after DNA damage and foci are resolved usually by 24 hours post-irradiation. Quantification of γH2AX foci using fluorescence microscopy was found to be sensitive enough to detect DSBs at even low doses and by combining fluorescence microscopy with computer power it was possible to devise an automated method, high content screening, to analyze thousands of cells with relative ease [121, 122]. In addition, it was found that the relationship between the number of

γH2AX foci and irradiation dose is linear down to a dose of 1 mGy and thus would support the LNT model [123, 124]. However, other researchers have found a deviation from linearity in lymphocytes in vivo after irradiation with doses below 6 mGy which indicates an underestimation of the risk of radiation by the LNT model [125]. Furthermore, analysis of

53BP1 and γH2AX foci was found to overlap and thus 53BP1 could be also used as a marker of DSB [126]. However, the number of DNA DSBs per nucleus is not the only indication of increased health risk, and therefore even if linearity could be established between DNA DSBs 26

Chapter 1

and radiation dose, it does not necessarily mean that linearity between irradiation dose and cancer risk holds true.

Figure 8: Illustration of the formation of DNA DSB foci. (a) shows the formation of a DNA DSB. (b) is the early response such as histone H2AX phosphorylation and recruitment of the MRN complex. (c) is the later response such as recruitment of BRCA1 (van Attikum and Gasser, Crosstalk between Histone Modifications during the DNA Damage Response, Trends in Cell Biol., 2009)

4.3.2 Cell Survival For example, the rate and fidelity of DNA repair following irradiation is also a factor that plays a role in cell survival. The golden standard that has been employed in radiation research for assessing cell survival following irradiation is the colony formation assay. This assay

General Introduction

measures the ability of surviving cells to form colonies but is not sensitive enough to be reliable at low doses of radiation. Combining the colony formation assay with cell counting techniques such as flow cytometry or digital microscopy made this technique robust enough to be used for low doses of radiation [127]. This method allowed scientists to determine that there was interindividual variability in radiosensitivity in human fibroblasts and that this was linked to chromosomal aberrations and to induction of premature differentiation [127].

Chromosomal aberrations are directly linked to DNA DSBs and their presence in peripheral blood mononuclear cells (PBMCs) has been long used as a biodosimetric method since the number of chromosomal rearrangements and gene mutations was found to hold a linear relationship with irradiation dose down to a dose of 0.01 Gy [110]. However, some studies have indicated failure to detect chromosomal aberrations in irradiated lymphocytes below 20 and 100 mGy respectively [128, 129]. If this is true, then it might indicate a deviation from the LNT model and lend credence to a threshold model. On the other hand, the report that non-cycling fibroblasts irradiated with a dose below 20 mGy fail to repair DSBs could indicate that low doses are more detrimental to cells and, possibly, that low dose effects are cell cycle dependent [124]. Another factor to consider is the complexity of DSBs; at low doses, the DSBs are sparse and far apart, whereas at higher doses they are closer together forming bigger and more complex foci that are harder to repair and may have more chance to form chromosomal rearrangements [105].

4.4 Deviations from the LNT

4.4.1 HRS/IRR The factor that has been linked to what has been termed hyperradiosensitivity (HRS) is the difference in G2/M checkpoint sensitivity between high and low doses. Research has shown that the G2/M checkpoint is not activated in some cells in response to low doses of radiation and that this contributes to HRS in these cells. In addition, when cells that do activate the 28

Chapter 1

G2/M checkpoint in response to DNA damage were treated with caffeine which allowed them to bypass this checkpoint, they became more radiosensitive [130]. This phenomenon has been uncovered in both normal and tumorigenic cells and often precedes another phenomenon between doses of 0.5 and 1 Gy termed increased radioresistance (IRR) [131-133]. HRS has been linked to the presence of a working copy of p53 and on induction of apoptosis due to a bypass of the G2/M checkpoint with unrepaired DNA DSBs [134, 135]. Apoptosis is thought to play a bigger role in low-dose sensitivity since a tissue is more capable of sacrificing the few cells that are hit at low doses than it is when a large number of cells carries DNA damage at higher doses. The existence of HRS in cells could indicate that the cancer risk at low doses of radiation is exaggerated by the LNT model [131, 136].

4.4.2 The Bystander Effect The bystander effect is a phenomenon that was reported in the scientific literature since the early days of radiation research and is gaining more prominence since its existence could alter the way we perceive the biological effects of low doses of radiation. The bystander effect refers to DNA damage being observed in cells that were not directly exposed to radiation.

This phenomenon was observed when cells were shielded but were in the same culture medium as irradiated cells or when culture medium was transferred from irradiated cells to cells that were not irradiated [137, 138]. These observations were even found in fish [139] and were found to confer a protective effect on non-irradiated fish. The factor responsible for the bystander effect has long been thought to be a small protein and recent research into the role of a panel of cytokines and chemokines point to interleukin (IL)-6 and IL-8 as two of the possible players [140-142]. Bystander cells have been shown to activate stress-related pathways such as JNK and ERK1/2 [143]. A mentioned in section 5.2.1.4, IL-6 and IL-8 are part of the secretome of senescent cells and this secretome was found to alter the response of neighboring cells to radiation. In addition, the finding that cells that were not directly hit by 29

General Introduction

radiation display DNA damage foci is also of significance. At low doses of radiation, this could have a significant effect on the cells' response to radiation and could alter the linear extrapolation below a dose of 100 mGy both at the phenotypic and molecular levels.

4.4.3 Adaptive Response Another low-dose phenomenon that has been linked to the bystander effect is the adaptive response. After exposing cells to a priming low dose, they were found to be more resistant to a subsequent higher dose of radiation. The adaptive response was described by some as a subset of the bystander effect. It was induced by a priming dose of 0.01 to 0.5 Gy and was found to be dependent on intercellular communication by gap junctions or secreted protein factors [144]. Others had reported that the adaptive response was only invoked in normal and not tumorous cell lines [145]. The adaptive response has been linked to the activation of such factors as nuclear factor kappa B (NF-κB), ATM, ERK and cyclin D1 upon priming with a low dose and that the radioprotective adaptive response was dependent on those factors' anti- apoptotic function [146-148]. A recent work by Grudzenski et al. proved that the adaptive response can be induced not only by a priming low dose of irradiation but also by other damaging reagents such as hydrogen peroxide (H2O2). Pretreating cells with 10 µM H2O2 caused more efficient repair of DNA DSBs at low radiation doses, thus establishing that the adaptive response did not only work via inhibition of apoptosis but perhaps by inducing faster and more efficient DNA repair [149].

4.4.4 Hormesis From the above it can be deduced that low doses of radiation could actually have a beneficial effect on health and this is the hormesis concept. This radioprotective effect has been found in a wide array of organisms from the simple prokaryotes to higher mammals as well as in vitro

[150, 151]. Adherents to the hormesis theory claim that it makes sense from an evolutionary point of view; since cells have always been assaulted by a barrage of background radiation

Chapter 1

and reactive oxygen species (ROS) from living in an aerobic environment, they have developed efficient means to dealing with this damage. However, higher doses of radiation are an abnormal occurrence and thus the cell is ill-prepared to deal with them.

4.5 Interindividual Sensitivity A further layer of complexity is added to the above arguments by the presence of interindividual sensitivity. As has been noted above, the use of the colony forming assay on surviving fractions of irradiated fibroblasts revealed differences in their behavior between donors. This probably can be ascribed to the genetic background of each individual. For example individuals with ataxia telangiectasia (AT) suffer a mutation in ATM that interferes with DNA DSB repair and thus makes them more sensitive to radiation-induced damage.

Other individuals with a mutation in the nucleotide excision repair (NER) pathway may suffer from xeroderma pigmentosum (XP) which makes them too sensitive to be exposed to sunlight. These individuals underline the fact that the study of the effect of low doses of radiation on health effects cannot be complete without considering studies on the basis of cell type and genetic background. The use of microarrays and high-throughput gene sequencing could be of immense value in this field of research [152-155].

Interestingly, these insights into the area of low dose irradiation are making their mark on the medical field as more papers are published on the use of lower doses of radiation in the treatment of cancers of the blood, the thyroid and the brain among others [156-160].

4.6 Risk to the Thyroid As mentioned earlier, low doses of radiation have been found to be linked to an increase in thyroid cancer risk [17, 18, 89, 90]. In addition, a pooled analysis of thyroid cancer in people who were irradiated during their childhood for medical purposes and atomic bomb survivors had an ERR per Gy of 7.7 while a Swedish study of patients with skin hemangioma who were

General Introduction

exposed to Ra-226 showed an ERR to the thyroid of 7.5 per Gy for a dose of 120 mGy [161,

162]. However, these estimates are confounded by many factors such as the age of the persons, their gender, their genetic background, the state of iodine deficiency, and latency time [163-165].

4.7 Conclusion From all of the above it is evident how murky the biology of low dose radiation still is.

However, more and more evidence indicate that perhaps an abandonment of the LNT model may be feasible and necessary from a molecular point of view. A replacement model, on the other hand, is still not evident and perhaps without clear and solid epidemiological data the

LNT is perhaps the best that can be achieved for now.

Chapter 1

5. Senescence Introduction: To Divide or Not To Divide A cell is characterized by its ability to reproduce itself and this ability is fundamental to the continuation of life since the days when unicellular organisms were the only life form.

However, whereas proliferation is crucial for the start of life, tissue repair, and survival, continuous and unchecked proliferation could lead to unwanted effects such as cancer.

Therefore, cells have devised several states of non-proliferation and those are quiescence, differentiation, and senescence.

The difference between the three is sometimes subtle; quiescence is basically different from the latter two in that it is reversible. In cell culture for example, cells can enter quiescence by withdrawal of serum and growth factors. This is the G0 phase that is described in the literature and in vivo; this can be found in lymphocytes which remain in this phase until they are needed in cases of infection for example [166].

Adding serum to the cell culture medium is enough for the cell to reenter the cell cycle. In contrast, differentiation and senescence are basically irreversible cellular functions with some exceptions. For example, advanced stages of tumorigenic cells are thought to undergo a process of dedifferentiation whereas some senescent cells are thought to reenter the cell cycle upon the suppression of the cyclin dependent kinase inhibitors (CDKI) [167]. Also, research has indicated that senescence could be reversed in some instances by inhibition of p53 and p16 [168, 169]. Sometimes senescence and differentiation have overlapped with senescent cells displaying signs of differentiation.

Senescence was originally discovered in cultures of human diploid fibroblasts (HDF) that had undergone a certain number of replicative cycles so that their telomeres reached a critical threshold which is now known as the Hayflick limit [170-172]. Senescence was thought to be a protective mechanism to guard against telomere attrition and subsequent chromosomal

General Introduction

instability. This form of senescence was found in all cells except ALT cells and those that had switched on a certain enzyme, telomerase, that elongated telomeres such as stem cells and most cancerous cells. Currently this form of senescence is known as replicative senescence

(RS) to distinguish it from other forms of premature senescence such as oncogene-mediated, stress-induced, or DNA damage-induced senescence.

Figure 9: States of proliferation and non-proliferation (Sang et al., Hijacking HES1: how Tumors co-opt the Anti- differentation Strategies of Quiescent Cells, Trends in Mol. Med., 2010) 5.1 The Hayflick Mosaic As mentioned in the beginning of this section, cellular proliferation is crucial for tissue homeostasis and its regenerative capacity; however, unlimited proliferation also makes cells more prone to the development of mutations and therefore to tumorigenicity. Senescence is 34

Chapter 1

assumed to be a tumor suppressive function of the organism and is thought to be an antagonistic pleiotropic1 process from an evolutionary point of view in that it is beneficial in younger individuals but harmful in older ones. This duality is at the heart of the stress-induced premature senescence (SIPS) conundrum to this day.

Premature senescence is now thought to be a beneficial cellular phenomenon and an acceptable cancer therapy endpoint like apoptosis. Many cytostatic drugs have shown promising results in cancer therapy by inducing senescence and sometimes lowering the concentration of some typical DNA damage-inducing drugs such as doxorubicin was enough to cause tumor-induced senescence (TIS) instead of apoptosis [173, 174]. However, senescent cells were found to possess a unique secretory profile termed a senescence-associated secretory profile (SASP) which includes various cytokines, chemokines and other proteins.

Some of these proteins were described as reinforcing the senescent phenotype via an autocrine manner such as signaling via the CXCR2 receptor and the fact that IL-1α is necessary for the continuous secretion of IL-6 and -8 thus creating a feedback loop [175, 176]. On the other hand, some of the secreted cytokines such as IL-6 are also inflammatory in nature and are thought to play a tumorigenic role in neighboring cells. A paper by Tsai et al. demonstrated how senescent fibroblasts by function of their SASP were able to render nearby breast cancer cells radioresistant while others have indicated that secretion of such factors as IL-6 or matrix metalloproteinases (MMP) induce tumor cell growth and invasion [177-179]. This raises the question of whether senescence is indeed a beneficial or harmful endpoint in cancer therapy.

A vital point here is then how long a senescent cell remains in the tissue. Previous reports indicated that senescent cells may remain in tissues for extended periods of time while a more

1 The antagonistic pleiotropy theory was proposed by George C. Williams as an explanation for senescence. It originally referred to the case where one gene controlled more than one trait and where at least one trait was 35

General Introduction

recent paper by Xue et al. demonstrates that senescent cells are rapidly cleared from the body by the immune system [180-182]. This finding could possibly point to another function of the

SASP, besides reinforcing senescence.

5.2 Hitting the Breaks: Premature Senescence Recent evidence points to the hypothesis that senescence in tumor cells might be a more prevalent response than previously thought. Up-regulating CDKIs in cells with overexpressed cell growth pathways such as the MAPK or Target of Rapamycin (mTOR) pathways will lead to hypertrophy and, subsequently, a senescence-like phenotype. It can therefore be argued that cancerous cells are in a pro-senescent state in that it is enough to up-regulate CDKIs to induce a loss of proliferation potential (PP) [183].

Loss of PP is probably the best way to distinguish senescent cells from other states of non- proliferation. If a cell in culture is provided with growth factors but does not resume proliferation, then it can be thought of as in senescence. However, other hallmarks of senescence exist.

5.2.1 Hallmarks of Senescence

5.2.1.1 Cell Cycle Profile

Senescence is marked by a halt in proliferation which is mediated by activation of the G1/S and the G2/M checkpoints and thus a typical cell cycle profile of senescent cells is an increase in the percentage of cells in the G1 and G2 phases of the cell cycle with a dramatic decrease in the fraction of cells in the S phase [184]. This profile has been extended to also include an increase in 4N cells [185].

5.2.1.2 Senescence-associated β-galactosidase (SA β-gal) The enzyme is thought to be the normal lysosomal β-galactosidase measured at pH 6.0. An increase in SA β-gal histochemical staining was observed in tissue and cells and was the first described sign of senescence [186]. However, this method has its limitations as it is found in 36

Chapter 1

other cases besides senescence such as quiescence and serum starvation [187-189]. Therefore,

SA β-gal staining should be coupled to other signs of senescence when determining senescence onset.

5.2.1.3 Morphology Cells that undergo senescence have a typical morphology that includes an enlarged and flattened phenotype as well as increased vacuolization that can be measured by microscopy or using flow cytometry via the side scatter (SSC) parameter. The SSC parameter is used in flow cytometry as a measure of how the laser bounces off the cell at a 90° angle. An increase in the

SSC parameter is indicative of an increase in cell granularity or complexity and thus could be indicative of increased vacuolization. However, since light scattering is affected by many phenomena, the results have to be interpreted carefully [185].

5.2.1.4 SASP One of the main characteristics of senescent cells is their association with a secretory signature termed SASP. This SASP has been found to include such cytokines as IL-6, IL-8, macrophage chemotactic protein (MCP)-1, vascular endothelial growth factor (VEGF)-A, and eotaxin [190-192]. The appearance of the SASP is thought to be p53-independent as opposed to the growth arrest phenotype. The appearance of the SASP was however dependent on the appearance of irreparable DNA damage foci that were unresolved for days or weeks.

A noteworthy and relevant finding was that cells expressing the BRAF oncogene went into senescence with an increase in secreted IL-6 and when IL-6 or its receptor were suppressed, the cell entered the cell cycle [193] whereas cells with the BRAFV600E mutation evaded senescence when they lacked insulin-like growth factor binding protein (IGFBP)-7 or CXCR2 or when the NF-κB and p38 pathways were blocked [179, 194]. Another important factor in

General Introduction

the development of the senescent profile is the transforming growth factor (TGF)-β1 which was found to be important for G1 arrest and SA β-gal staining in keratinocytes [179].

5.2.2 The Role of p53 and pRb Two parallel and interconnected pathways are implicated in senescence, namely, p53 and the retinoblastoma (Rb) pathway. The tumor suppressor protein p53 is activated by ATM which phosphorylates the former on serine 15. Afterwards, p53 activates the CDKI p21Cip1 which is involved in the G1/S checkpoint activation and has also been linked to resistance to apoptosis

[195, 196]. The up-regulation of p21 is thought to be an activating event of the senescence response in that it causes reversible inhibition of the cell cycle probably for DNA damage repair. On the other hand, p16 is up-regulated afterwards, when p21 levels decline, and is needed for the irreversible cell cycle block [197]. Previous research has indicated that although the p53 and Rb signaling pathways signal independently, there apparently is crosstalk between the two. For example, p53 was required for BRAFV600E-mediated senescence and the induction of p16 in melanocytes. Others have hypothesized that p53 function makes cells permissive for p16-dependent effects [198-200]. A reverse effect was found by others in various cell lines with different statuses of p53 and p16 where up- regulation of p21 was absent in cells with no active p16 [201]. One common factor between the p53 and Rb pathways is the mouse double minute (MDM)-2 protein which was found to promote Rb degradation in a ubiquitin (Ub)-independent manner as well as its Ub-dependent degradation of p53. However, Rb does mediate another hallmark of senescence, that is now thought to be unique to Rb, which is senescence-associated heterochromatic foci (SAHF).

These foci are mediated by the methylation of lysine 9 of histone 3 (H3K9Me) [202].

Yamakoshi et al. reported that the relationship between H3K9Me foci and p16 is mediated by

DNMT1, an inhibitor of p16. Furthermore they and others showed that loss of p53 led to an increase in the levels of p16, thus proving the existence of crosstalk between the two 38

Chapter 1

pathways and highlighting the role of p16 as a tumor suppressor which also explains the increased tumorigenicity in cells with loss of p16 and p53 [203-205]. Interestingly, p16 mutations are a rare event in primary thyroid tumors but more prevalent in thyroid cell lines

[206]. Finally, there have been reports of p53 and p16-independent senescence which underlines the role of other players reported to induce senescence in the literature such as p63, p73, p15, and p27 [207, 208].

Figure 10: Pathways up-regulated by oncogenic BRAF and leading to onset of premature senescence. The pathways converge on Rb and p53 (Cichowski and Hahn, Unexpected Pieces to the Senescence Puzzle, Cell, 2008)

5.2.3 Other players on the senescence scene The advent of high-throughput screening techniques such as DNA microarrays has provided insight into the genetic complexities of many physiological processes. One of the main findings of using microarrays is that SIPS is distinct from RS on the transcriptional level with only a few genes shared in common [209, 210]. As expected, the genes constantly regulated in senescent fibroblasts after high doses of external radiation were involved in cell cycle entry

General Introduction

and exit, mitotic spindle assembly and proper segregation of cells. However, these genes were also found in quiescent cells making it harder to differentiate senescence from other states of non-proliferation [211, 212]. Recent reports have identified aurora kinase A (AURKA) as a player in senescence although there are conflicting reports as to whether it induces or inhibits senescence as over- and under-expression of this protein have both been reported to induce senescence [213, 214]. Cdc-2 has also been shown to be necessary for mediating a senescence response in breast cancer cells [215]. Increased TGF-β1 activity has been linked as well to the appearance of the senescent profile apparently through inhibiting c-Myc. Furthermore, TGF-

β1 was found to induce an anti-proliferative effect in cells of papillary thyroid carcinoma through the up-regulation of p21 [216, 217]. A recent paper has hinted at a crucial role for suppressor of cytokine signaling (SOCS)-1 in the senescence response. It seems SOCS1 acts as a mediator between cytokine signaling and p53-mediated senescence and plays a role in bridging ATM and p53, thus facilitating the latter's phosphorylation on serine 15 which is important for senescence onset [218].

5.3 Final Remarks and Points of Contention So why does a cell undergo senescence in response to DNA damage and not apoptosis for example? Given that p53 is involved in both responses, there should be other factors that tip the balance one way or the other. For example, it was found that overexpression of the anti- apoptotic B-cell Lymphoma (BCL)-2 protein induced the cell to undergo senescence [219].

Furthermore, Bertrand-Vallery et al. reported onset of differentiation in keratinocytes exposed to UVB with up-regulation of p53 and p21 in the absence of p16. This response reverted to senescence when p16 was introduced into the system [220]. In addition, the Rb transcription factor (TF) was found to be involved in quiescence as well as senescence so this signaling pathway is related to proliferation per se and not necessarily to irreversible senescence [221].

Finally, it has recently been shown that senescent cells share some features with 40

Chapter 1

dedifferentiating cells pointing to the possibility that cells pass through a stem cell-like state before committing to senescence [222].

From the above it is clear that much headway has been made in the field of premature senescence in the past decade. However, the field remains not without controversy and some loose ends remain concerning the exact mechanism of senescence onset and whether it is indeed a fail-safe anti-cancer mechanism.

General Introduction

Chapter 1

6. Epigenetics The DNA damage induced by irradiation is not limited to genetic modifications, here meaning any change that occurs to the order of the nucleotide bases as happens in translocations, deletions, point mutations, etc…, but extends to epigenetic changes. Epigenetic changes involve heritable modifications to the DNA structure itself, without altering the actual sequence, e.g. methylation of the nucleotide bases, to the histone proteins, such as acetylation or ubiquitylation. In addition, any changes that affect the stability and translation of the messenger RNA (mRNA), such as microRNA (miRNA)-induced silencing, are also considered as epigenetics sensu lato. Epigenetics has been gaining a lot of interest lately and adds a further layer of complexity to the effects induced by radiation.

Histone modifications usually occur on their N-terminal tails and include a wide array of changes such as acetylation, methylation, phosphorylation, and ubiquitylation. All of these modifications seem to have an impact on chromosome structure, however the most studied changes are histone acetylation and methylation. Histone acetylation and deacetylation are carried out by the two enzymes histone acetyl transferases (HAT) and histone deacetylases

(HDAC) while histone methylation is mediated by histone methyl transferases [223].

However, since the epigenetic effects of radiation that were studies in this thesis include microRNAs and DNA methylation, the bulk of this introduction will deal with those two aspects of epigenetics.

6.1 DNA Methylation: And the Rest is Silence Since all cells typically share the same DNA in one individual of any species, epigenetic marks could be thought of as the determinants of cell fate, implying that the epigenome differs from cell type to cell type. However, not all species possess methylated DNA sequences as has been observed in the yeast Saccharomyces cerevisae and the nematode

Caenorhabditis elegans and the ones that do display different types of methylation. For 43

General Introduction

example, insects' genomes possess a type of patchy methylation known as 'mosaic methylation' as opposed to mammals whose genomes are globally methylated [224]. Most of the mammalian genome is methylated except for clusters of cytosines located in what is known as CpG islands (CGI) which are found in around 50% of the promoters of human genes. Methylating cytosines in these islands located in the promoter region of genes is now agreed to be a silencing mechanism which has been until recently thought to be actively irreversible except through artificial processes or in the early embryo [225]. After establishing the basic methylation pattern by de novo DNA methyltransferases (DNMT) such as

DNMT3A and 3B which act by interacting with previously established methylation markers on histones (e.g. triply methylated lysine 9 on histone 3 [H3K9me3]), methylation patterns are maintained by another methyltransferase, namely DNMT1 [223]. However, during the course of life of any organism, DNA methylation occurs for any one of various reasons such as the normal process of aging, environmental factors such as smoking or radiation, or even viral infections. Methylation of one allele of a tumor suppressor genes (TSG), a process known as loss of heterozygosity (LOH), may lead to cancer when its counterpart allele had been previously mutated as a first hit in the Knudson two-hit hypothesis while the opposite is also true [226]. The fact that methylation was found to increase with age in genes associated with prostate cancer can be one reason of why cancer incidence increases with age [227]. It remains to be stressed that DNA methylation does not lead to silencing per se but it is the binding of proteins, such as the methyl-cytosine binding protein (MBP) protein, that act as transcriptional repressors.

Chapter 1

Figure 11: (A) Cross-talk between DNA methylation and histone methylation. (B) Different types of DNA methylation across species (modified from Suzuki and Bird, DNA Methylation Landscapes: Provocative Insights from Epigenomics, Nat. Rev. Genetics, 2008) 6.2 The Cancer Methylome The disruption of the cancer epigenome has been observed for some time now and it has been held partially responsible for some of the features of cancerous cells. Some genes were found to be hypermethylated (most notably TSGs) while others were hypomethylated

(e.g.oncogenes) whereas a general hypomethylation of the cancer genome is observed in most tumors as compared to matched normal cells [228]. These methylation changes have been found in high CpG island density regions as well as in regions of lower CpG density termed cancer-specific differentially DNA-methylated regions (cDMR) which are thought to account for cancer heterogeneity [229]. Besides the abnormal silencing or activation of genes in cancer, aberrant methylation of the cancer genome may also lead to chromosomal instability.

Some typical TSGs that are silenced by promoter methylation in cancers are the VHL genes, the CDKN2A, and MLH1 gene in colorectal cancer, and RASSF1A which is related to the Ras 45

General Introduction

oncogene. However, the total number of genes that are silenced by promoter methylation is as big or even bigger than the number of genetically mutated genes in cancer. Furthermore, a lot of the genes with aberrant DNA methylation in cancer are DNA repair genes providing a link between epigenetic dysregulation and genetic mutations. Epigenetic changes are now even thought to be initiating events in cancer through such pathways as Wnt and to lead to oncogenic addiction [230].

6.2.1 Thyroid Cancer: The Epigenetic Factor Like any other malignancy, thyroid cancers have displayed wide aberrations in their DNA methylation status. Early reports have identified an aberrant methylation of three genes in thyroid tumors including human GH (hGH), platelet derived growth factor (PDGF)-B and H- ras which were located on different chromosomes [231]. Since then, the number of genes that display aberrant methylation has grown considerably. One of the most prominent genes silenced is that coding for the TSHR in PTCs displaying an up-regulation in the MAPK pathway as has been seen earlier in cells with RET/PTC translocation for example [232]. Up- regulation of MAPK pathway has been linked to epigenetic silencing of TSHR and NIS which is important for cancer cells to maintain ligand-independent growth and proliferation, one of the hallmarks of cancer. In addition, silencing of NIS makes it harder to kill the cancerous cells by radioiodine therapy. Down-regulation of the MAPK pathway was enough to restore the expression of TSHR and NIS [233, 234]. Interestingly, other markers of thyroid differentiation and of iodine metabolism are silenced at different stages of thyroid cancer. For instance, pendrin, the product of the gene slc26A4, is silenced at an early stage of thyroid carcinoma while thyroid transcription factor (TTF)-1 was silenced at later stages [235, 236].

The silencing of several tumor suppressor genes has also been reported in thyroid malignancies and have been linked to the initiation and progression of the tumors. Among the genes silenced by hypermethylation are Cbp/p300 Interacting Transactivators with glutamic 46

Chapter 1

acid [E] and aspartic acid [D]-rich C-terminal domain (CITED1), retinoblastoma protein- interacting zinc finger (RIZ)-1, Ras association domain family (RASSF)-1A, 2, 10,

Rap1GAP, and TMS1 [237-242]. In addition, the sodium monocarboxylate transporter

(SMCT)-1, the product of the gene slc5A8, was also frequently silenced in cases of thyroid carcinoma and this protein is thought to induce apoptosis through pyruvate-dependent mechanisms [243, 244]. Notably, a lot of the aforementioned silenced genes have been found in cases with BRAF mutations, which may provide a link between BRAF activating point mutations and aberrant methylation. A recent study conducted in hyperplastic polyposis syndrome (HPS) revealed promoter hypermethylation of DNMT3L but only in association with KRAS mutations and not BRAF. However, the study was carried on a small number of samples and only silencing by methylation was studied [245].

BRAF mutations, however, have been linked to the silencing of DNA repair genes such as hMLH1, involved in mismatch repair. This finding could link BRAF point mutations to genomic instability observed in PTC [246]. An important gene that was found to be hypermethylated in around 30% of thyroid carcinomas is the product of the CDKN2A gene, p16INK4. This gene is involved in the Rb pathway and its activation is important for the senescence response as mentioned in a previous section [233, 234].

6.3 DNA Methylation, DNA Repair, and Radiation: a Messy Triangle Besides identifying the link between radiation and thyroid cancer, the accident at the

Chernobyl nuclear power plant (NPP) demonstrated the link between radiation and methylation changes. Studies done on plants such as pine trees (Pinus silvestris) growing in the region of Chernobyl revealed genome-wide DNA methylation changes that have been linked to radiation resistance [247, 248]. DNA methylation has been associated with radiation resistance in tumors as well, a phenomenon that was reversible in patients by treatment with

General Introduction

demethylating agents such as the DNMT inhibitor 5-aza-2'-deoxycytidine (DAC) [249]. So what is the link between radiation sensitivity and epigenetic modifications? One possible link is the DNA damage response (DDR) which involves phosphorylation of the histone H2AX. It has been found that DNA repair occurs slower in heterochromatic regions of the genome and that efficient DNA repair involves histone modifications that open up the DNA. For example, recruitment of breast cancer type 1 susceptibility protein (BRCA1) was dependent upon polyubiquitination of the histones H2A, H2B and H2AX [250]. The process of DNA repair itself has been associated with methylation changes that could lead to gene silencing. A case in point is the homologous recombination (HR) mode of DNA DSB repair where this process has been linked to an increase in promoter methylation and gene silencing [251].

Therefore, it is not surprising that Aypar et al. found global DNA hypermethylation in response to low linear energy transfer (LET) irradiation. However, the authors reported no change in the methylation status of the three genes NF-κB, tumor suppressor in lung cancer 1

(TSLC1), and cadherin 1 (CDH1) and hypomethylation of long interspersed nuclear element 1

(LINE-1) [252]. Incidently, many genes involved in DNA repair have been found to be epigenetically silenced in cancers and thus, theoretically, radiation-induced silencing could also involve DNA repair genes such as OGG1, MGMT, and transcription-coupled repair

(TCR) which are involved in SSB repair or BRCA1 and XRCC5 involved in DSB repair [253], thus adding insult to injury.

6.4 Detection and Measurement of DNA Methylation Quantitative measurement of changes in DNA methylation and histone modifications has been described for a long time. However these methods measured only global changes in

DNA methylation or histone modifications [254]. Assessing the methylation status of individual genes had to be done on a gene by gene basis, using methyl-specific polymerase

Chapter 1

chain reaction (MSP) after bisulfite treatment of the DNA. Bisulfite treatment of cytosines converts them to uracils with the exception of methylated cytosines which remain untouched.

Performing polymerase chain reaction (PCR) on the treated DNA using specific primers could detect the methylation status of DNA. With the advent of high throughput methods, it became possible to assess the methylation status of a large number of genes in one shot. The methods used include bisulfite genomic sequencing, methods that utilize the sensitivity of restriction enzymes to CpG methylation within their cleavage site, and the method used in this study which utilizes the affinity of certain proteins to methylated DNA [224, 255]. Using the naturally occurring methyl binding domain (MBD) proteins, it is possible to isolate highly methylated regions of the genome and combined with either quantitative PCR (qPCR) or whole genome sequencing, it's possible to identify changes in DNA methylation [256, 257].

The appearance of next-generation sequencing (NGS) methods on the market has drastically reduced both the price and the time needed to sequence large areas of the genome. The NGS technologies rely on advances made in computers, robotics, and imaging and all require a considerable knowledge of bioinformatics. An example of NGS is the technology developed by Illumina based on the concept 'sequencing by synthesis' (SBS) which produces sequence reads of ~32-40 base pairs (bp) from tens of millions of surface-amplified DNA fragments simultaneously [258]. Those short DNA sequences are then immobilized on glass flow cells and amplified into clusters using a process known as 'bridge PCR'. Afterwards, sequencing occurs after linearization of the DNA using several rounds of extension using modified nucleotides. These nucleotides are bound to four fluorescent labels, one for each of the four nucleotides, and they are 'reversible terminators' so that only one can be added in each elongation cycle [259-261].

General Introduction

Figure 12: Illumina Solexa's 'bridge PCR'. Stretches of DNA are immobilized using adapters and then amplified to form clusters on the flow cell (modified from Shendure and Ji, Next-generation DNA Sequencing, Nat. Biotechnology, 2008) 6.5 Sailing Uncharted Waters: Active DNA Demethylation and Gene Body Methylation As mentioned earlier, DNA demethylation was described in the early embryo; but proving active demethylation in later development stages has proven to be more difficult. Several mechanisms of DNA demethylation in mammalian cells have been described. These include enzymatic removal of the methyl group of 5-methyl cytosine (5meC), oxidative demethylation, and radical S-adenosylmethionine-based demethylation (SEM). The most interesting however is the report that active DNA demethylation could occur by a mechanism involving base excision repair (BER) and nucleotide excision repair (NER) which are also involved in repairing DNA SSB [262].

The finding that a considerable amount of DNA methylation occurs actually in the gene body and outside the promoter region has also led to confusion over the role of methylation in these regions. For example, some studies have shown that methylation of the first exon is linked to gene silencing just like the promoter region [263]. Methylation of other exonic regions, especially at intron-exon boundaries, is thought to play a role in alternative splicing of genes while one study has shown that methylation of an intronic region in testicular cancer regulated the miRNA miR-199a [264-267].

All the above serves to illustrate how complex epigenetic regulation of gene expression is with epigenetic factors serving to control other epigenetic processes.

Chapter 1

6.6 MicroRNAs: Seek and Destroy (or not…) MicroRNA (miRNA) are a class of RNA that are transcribed by RNA polymerase II as pri- miRs and later processed into shorter transcripts, known as pre-miRs, which are further processed into ~22-nucleotide mature double-stranded miRs by the RNase III Drosha. The main strand, the miR, is involved in the silencing of mRNA by either degrading or inhibiting its translation in what is known as the RNA-induced silencing complex (RISC); while the opposite strand, miR*, was thought to be degraded. Now it is thought that the miR* might also be involved in silencing and might even be involved in thyroid cancer [268, 269]. The basic mechanism of miRNA action is what is known as homology-dependent gene silencing

(HDGS) and thus identifying miRNA targets has relied on automated techniques that search for homologous sequences within the genome. This has led to identification of hundreds of potential targets for each miRNA with some targets being shared between different miRNAs as well as some miRNAs being targets of other miRNAs [268, 270].

Since miRNAs have been reported to be involved in the regulation of various physiological processes such as apoptosis, cell cycle regulation, and senescence, it's perhaps not surprising that miRNAs are dysregulated in cancers [271]. An overall increase in the activity of miRNAs was reported in human PTCs and breast cancer and an involvement of miRNAs in the initiation and progression of cancer as well as angiogenesis is now established. Reports that miRNAs can target known oncogenes and tumor suppressor genes highlights a role of miRNAs in cancer; add to that the fact that they themselves can act as oncogenes and tumor suppressor genes [272, 273]. miRNAs have also been reported to alter the response of cancer cells to cytotoxic therapy [274].

General Introduction

Figure 13: Processing of miRNAs in the cell and their mode of action (from Breving and Esquela-Kerscher, The Complexities of microRNA Regulation: mirandering around the rules, Int. J. Biochem. Cell Biol., 2010) A growing interest in miRNAs and their role in normal cellular processes has led to identification of clusters of miRNAs that are involved in different responses to the same stressors among cells. However, there has been little overlap in the miRNAs that are reported in the literature to respond to irradiation in fibroblasts, endothelial cells, and prostate cancer cells [275-277]. Perhaps this can be explained by the finding that miRNA regulation is dose- dependent where some miRNAs are regulated at high doses and not at low ones and vice versa [278]. A study by Simone et al. identified a cluster of miRNAs that were consistently regulated upon oxidative stress which included the Lethal (Let)-7 family [279]. The let-7 family is the first to be identified in C. elegans and is now thought to be one of the most important members of the miRNA family. The targets identified for different members of the let-7 family include the anti-apoptotic member of the B-cell Lymphoma (Bcl)-2 family, Bcl-

XL, c-Myc, and p16 [280, 281]. A link between let-7f and PTC was established when it was

Chapter 1

found that the RET/PTC oncogene down-regulated let-7f and that overexpression of this miRNA inhibited the MAPK pathway [282]. Other notable miRNAs have been reported in the literature whereby their targets had been confirmed such as miR-15 and -16 which target Bcl-

2, and the miR-17-92 cluster which targets the E2F transcription factor [283-286]. However, the most prominent miRNA cluster perhaps remains the family miR-34 whose three different members, miR34a-c, are involved in the p53 pathway and control cell proliferation [287,

288].

6.6.1 MiRandering Around the Rules The picture emerging from the miRNA realm was complex enough before it was discovered that miRNAs can act in an activating as well as a suppressive manner. This paradoxical effect of miRNAs was in part explained by differing cell cycle states of cells. Vasudevan et al. reported that miRNAs in non-cycling cells had an activating effect on their target genes as opposed to actively proliferating cells [289-291].

In addition, several other factors that control miRNA function including other epigenetic processes have now emerged. On the other hand, miRNAs have also been reported to affect the methylation status of some genes [289, 292]. The recently published paper by Prazeres et al. that stipulated convergence of chromosomal, epigenetic, and miRNA factors to silence the gene low-density lipoprotein receptor-related protein (LRP1B) in thyroid cancer would thus make sense in light of the previous findings [293].

General Introduction

Thesis Aims

"Few people have the imagination for reality"

Johann Wolfgang von Goethe

Thesis Aims

Thesis Aims It is clear, from the previous chapter, that a link exists between radiation and the appearance of thyroid cancer and that this is tied to chromosome rearrangements brought on by DNA

DSBs. Evidence exists pointing to an effect of low doses to the thyroid in childhood and the appearance of thyroid cancer later in life. However, these observations are based on clinical evidence from the long-term follow up of children who received moderate to low doses of X- rays to their thyroids and not on mechanistic studies. In addition, RET/PTC, an oncogenic translocation, was found in a large percentage of Chernobyl PTCs. This translocation is an initiating event in PTC but is not alone sufficient to induce thyroid cancer. It would definitely be interesting to know how the presence of the RET/PTC translocation would alter the thyrocytes' response to radiation.

We were interested to investigate the effect of a range of external X-ray doses covering the low to high on PTC cells with a RET/PTC translocation, on thyroid cells with a RET/PTC translocation but no discernable PTC morphology, and on normal thyroids. We chose a range of doses from 62.5 mGy to 4 Gy as it falls within the range of doses to the thyroid received by residents of the areas surrounding the Chernobyl NPP site (90 mGy to 12 Gy) [1]. By using an array of high-throughput techniques, we hoped to identify the difference in the response of these three thyroid systems. Furthermore, since p53 is such a central player in the radiation response, we will use two cell lines of PTC origin with different p53 statuses to investigate the difference in the response of these cells to low dose radiation.

Microarray use has become more and more common in the past decade. The standardization of microarray preparation and the advent of such companies as Affymetrix has facilitated inter-laboratory comparison of results. We will use Affymetrix microarrays to compare the response of a PTC cell line with RET/PTC translocation, murine thyroids with RET/PTC

Thesis Aims

translocation but no PTC, and wild-type thyroids to low, moderate and high doses of X-rays.

This will allow us to identify the role of the RET/PTC translocation in the cellular radiation response. In addition, this will hopefully provide some mechanistic insight into the tumorigenic potential of low doses of radiation in the normal thyroid.

The link between radiation and the epigenetic profile of cells is gaining more interest in the scientific literature. We therefore are interested in the effect of low doses of radiation on microRNA and DNA methylation changes. Next generation sequencing (NGS) techniques are now gaining more interest and are predicted to replace microarrays. We will use NGS technology and low density arrays to investigate the epigenetic response of low, moderate and high X-ray doses on a cell line of PTC with a RET/PTC translocation and tie that with the transcriptomic and translational response of the same cell line.

Chapter 2: Response to Low Dose of X- irradiation is p53-dependent in a Papillary Thyroid Carcinoma Model System Chapter 2

Modified from 'Abou-El-Ardat K, Derradji H, De Vos W, De Meyer T, Bekaert S, Van Criekinge W, Baatout S. Response to low-dose X-irradiation is p53-dependent in a papillary thyroid carcinoma model system. Int. J. Oncol. 39 (2011): 1429-1441'

Response to low radiation dose is p53-dependent

Response to low dose of X-irradiation is p53-dependent in a papillary thyroid carcinoma model system

1. Abstract The link between high doses of radiation and thyroid cancer has been well established in the literature, as opposed to the effect of low doses. We investigated the effects of a low dose of

X-ray irradiation in a papillary thyroid carcinoma model with wild-type and mutated p53. A low dose of 62.5 mGy was enough to cause an up-regulation of p16 and a decrease in number of cells in S phase in TPC-1 cells, but not in the p53 mutant BCPAP. At 0.5 Gy and higher signs of senescence appeared only in TPC-1. We conclude that low doses of X-rays are enough to cause a change in cell cycle distribution, probably p53-dependent p16 activation, but no significant apoptosis. Senescence required higher doses of X-irradiation via a mechanism involving both p16 and p21.

Chapter 2

2. Introduction The causal link between ionizing radiation, DNA damage, and adverse health effects is well established at high doses [294]. However, there is still debate on the biological consequences of effects induced at lower doses, and consequently there are uncertainties in what constitutes a 'low dose' and how safe low doses are. This uncertainty is compounded by the choice of method used for determining the effect of low doses. Currently a linear no-threshold (LNT) model is used to extrapolate effects of low dose exposure from those at high doses. This model has been challenged as the LNT model is suggested to break down below 10 mGy

[295]. A recent paper by Averbeck has listed scientific evidence against the LNT model that have been published to date and came to the conclusion that the LNT model may have to be abandoned as it does not describe what happens at low doses [119]. Indeed, Rothkamm and

Löbrich reported that there was a lack of DNA double-strand break repair in fibroblasts irradiated with X-ray doses as low as 1 mGy [124]. However, the authors did show that detection of γ-H2AX foci is possible even at such low doses making it a suitable method for the study of the biological effects of low doses of radiation. Furthermore, a recent study on the effect of low and very low doses of radiation on a human mesenchymal cell line at different time points supported a non-linear model at low doses [296]. Currently, low doses of radiation are commonly defined as anything between background radiation (~0.01 mSv/day) and high doses of radiation (150 mSv/day and above) [142].

Studies on the effect of low-dose irradiation are non-trivial. After the Chernobyl nuclear disaster, large amounts of radioactive iodine isotopes were released into the atmosphere [89].

The effect of the radioiodine became apparent in the early nineties, when increases in the frequency of thyroid cancer were reported in children from the countries surrounding the site

Response to low radiation dose is p53-dependent

[89, 90]. In addition, a high prevalence of RET/PTC3 rearrangement was reported in these patients [90]. A link between external X-irradiation and thyroid tumor formation was already reported by Christov in rats [297]. Over a period of 18 months, the author noticed an increase in the incidence of thyroid tumors in rats irradiated once with 300 rads of X-rays. The potential for adverse health effects after low doses becomes crucial when taking into account that we are now more than 20 years away from the Chernobyl disaster. The danger of the high doses of irradiation emitted is thus largely gone but the issue of the risks of exposure to low doses of radiation remains to be dealt with.

For the purpose of this study, a model system for papillary thyroid carcinoma was exposed to a range of X-ray doses, ranging from what is considered a 'low dose' (0.0625 Gy) to high doses (0.25, 0.5, 1, and 4 Gy) and the changes in viability, necrosis, apoptosis, cellular morphology, and the cell cycle were monitored. The model system consists of two different cell-lines, TPC-1 and BCPAP. TPC-1 is a papillary thyroid carcinoma (PTC) cell line of human origin which harbors a constitutive RET/PTC1 rearrangement while BCPAP is a cell line of the same origin but with a V600E BRAF mutation and a mutated copy of p53. The

RET/PTC and BRAF mutations were found to work on the same signaling axis and had a common gene signature [103].The main difference is therefore the p53 mutational status, which allows us to infer conclusions on the role of the p53 pathway in low and high dose X- irradiation. Furthermore, cell lines provide a more homogenous model to work with, something which is important when working with low doses of radiation, where the effects could be subtle and easy to miss in a heterogeneous system.

Chapter 2

3. Materials and Methods

3.1 Cell Culture TPC-1 cells (kindly provided by Dr. Horst Zitzelsberger, Helmholtz Zentrum München,

GmbH, Munich, Germany), which are thyroid papillary carcinoma cells of human origin, were cultured in 25 cm2-tissue culture flasks in Dulbecco's Modified Essential Medium

(DMEM) (Gibco, Invitrogen, Paisley, UK) supplemented with 4 mM L-glutamine and 10% v/v fetal bovine serum (FBS) (Gibco, Invitrogen, Paisley, UK) in a humidified incubator

(37°C; 5% CO2). BCPAP (provided by Dr. Jacques Dumont, Université Libre du Bruxelles

(ULB), Brussels, Belgium) were cultured in Roswell Park Memorial Institute (RPMI)-1640 supplemented with 10% FBS in a humidified incubator (37°C; 5% CO2). No antibiotics were added to the culture medium. For all experiments the earliest possible passage of cells was used (p22 for TPC-1 and p7 for BCPAP). For experiments on supernatants in TPC-1 cells, three different passages were used one of which was p22. The cells were tested for mycoplasma contamination and were found to be contamination-free.

3.2 Cell Irradiation Cells were plated at an initial cell density of 1x106 cells per flask 24 hours prior to irradiation.

The medium was replenished prior to irradiation. Cells were irradiated using 250 kV, 15 mA,

1 mm Cu-filtered X-rays (delivered at 5 mGy/sec from a Pentak HF420 RX machine). Doses used were 0.0625, 0.25, 0.5, 1, or 4 Gy, while non-exposed cultures were sham irradiated using a procedure identical to the 4 Gy irradiation. Doses in Gy were calculated using the same criteria used in medical practice to calculate doses given to tissues (i.e. 95% of exposure in Roentgen). The cells were then returned to the humidified incubator until harvest.

Response to low radiation dose is p53-dependent

3.3 Cell Viability, Counting and Morphology Cell viability was assessed using the trypan blue dye exclusion method, cell numbers by a

Beckman Coulter EPICS XL-MCL and fluorescent microfluorospheres at 24 and 48 hours post-irradiation. For the trypan blue exclusion method, cells were mixed 1:1 with the trypan blue dye and then counted in a hemocytometer under an inverted light compound microscope.

Cells that excluded the dye (i.e. white in color) were counted as alive while the ones that incorporated the dye (i.e. blue in color) were counted as dead. Viability was calculated as the fraction of live cells in the whole counted population multiplied by 100%. For the microfluorosphere counting, a volume with a known amount of microfluorospheres was mixed with a volume of PBS containing an unknown number of cells. The mixture was passed in a flow cytometer which counted 10,000 cells and a certain number of microfluorospheres. Using the known concentration of microflurospheres, the number of microfluorospheres counted and the number of cells, it was possible to calculate the initial concentration of cells. Cellular morphology was determined at 24 hours post-irradiation by examining May-Grünwald Giemsa stained cells after centrifuging them onto glass microscope slides.

3.4 Cell Cycle and AnnexinV/PI Cell cycle stages were quantified using an EPICS XL-MCL flow cytometric analysis of propidium iodide stained cells and analyzed using the software System II from Coulter. In summary, cells were collected from their flasks 24 hours after irradiation by trypsinization and then permeabilized using 80% ethanol and stained with a solution of propidium iodide

(PI) and RNase A at 37°C for 1 hour. The cells were then analyzed by flow cytometry. The

Chapter 2

flow cytometer sorts cells according to their DNA content, as measured by the amount of fluorescence emitted by the DNA-binding PI. Cells are then sorted according to their DNA content into cells in G1, S, or G2 using manually chosen gates. Complications due to clumps or doublet cells were avoided by excluding the compromised region by gating. The results are reported as the percentage of cells in each phase (x.xx%).

Apoptotic and necrotic cell numbers were ascertained with FITC-conjugated Annexin V antibody and propidium iodide staining (Bender MedSystems Diagnostics, GmbH, Vienna,

Austria) according to manufacturer‟s instructions. The method relies on the binding of an

FITC-labelled Annexin-V antibody to phosphatidylserine (PS), which in normally functioning cells is kept on the inner leaflet of the bilipid plasma membrane by the enzyme flippase. Upon apoptosis, the PS can flip to the outer leaflet of the membrane and is thus detected by the antibody. PI cannot normally penetrate the plasma membrane and therefore its ability to stain

DNA is based upon loss of membrane integrity which happens during necrosis or late apoptosis. Thus cells that are Annexin-V-positive/PI negative are considered in early apoptosis, those that are Annexin-V-postive/PI-positive are in late apoptosis while Annexin-

V-negative/PI-positive are considered necrotic. Analysis of apoptosis/necrosis by flow cytometry was done using System II software from Coulter using manually set gates.

Response to low radiation dose is p53-dependent

Figure 14: Figure illustrating the distribution of TPC-1 cells according to their fluorescence in the PI channel (x-axis) and the FITC channel (y-axis) by a flow cytometer at 0 (A) and 4 Gy (B). Cells were incubated with propidium iodide and FITC-conjugated antibody against annexin V and passed through a Beckman Coulter EPICS XL flow cytometer as mentioned in the materials and methods section. An increase in annexin V- positive/ PI-positive and FITC-positive/PI-negative cells can be seen in regions N and R at 4 Gy of X- irradiation.

3.5 Caspase-3 levels using Flow Cytometry Irradiated TPC-1 cells were trypsinized and pelleted (1500 rpm for 5 minutes) and the manufacturers' instructions were followed:

For Caspase-3: CaspGLOW Fluorescein Active Caspase-3 Staining Kit (Medical &

Biological Laboratories Co., Ltd., Woburn, MA) was used according to manufacturer‟s instructions. Briefly, a substrate, FITC-linked DEVD-FMK, that is cleaved by active caspase-

3 to release fluorescence was added to 3.3x105 TPC-1 cells and incubated for an hour in a humidified incubator at 37°C. The cells were pelleted and washed twice with PBS and then

Chapter 2

suspended in 1 ml of PBS and read by flow cytometry which sorted the cells according to their fluorescence level.

3.6 Western Blotting Irradiated cells were pelleted by centrifugation and washed with PBS before being pelleted and frozen at -80°C. On the day of experiment, the pellet was thawed and proteins extracted using BioRad's Ready Prep II and a cocktail of protease inhibitors and phenylmethylsulfonyl fluoride (PMSF) following the manufacturer‟s instructions. The proteins were quantified using BioRad's Bradford solution on a spectrophotometer at 595 nm. 30 µg of proteins were loaded onto a 10% polyacrylamide followed by electrophoresis. The bands were transferred electrically onto a polyvinyl difluoride (PVDF) membrane, the loading controlled using

Ponceau S staining, and the membrane probed with primary monoclonal antibodies against p53 (53 kDa), p21 (21 kDa) (Sigma-Aldrich), p73 (73 kDa), Akt (56-60 kDa) , Bcl-2 (26 kDa), and p16 (16 kDa) (Santa Cruz). For p16, an antibody that recognizes both the mutated and normal form of the protein was used as TPC-1 is reported to harbor a mutated copy of the protein [298]. Horseradish peroxidase (HRP)-conjugated secnodary antibodies were used and the bands were visualized on an X-ray film using chemiluminescence (Amersham ECL

Western Analysis System, GE Healthcare, UK) and equal loading was verified using β- tubulin as a housekeeping protein.

3.7 TGF-β1 ELISA Twenty four hours after irradiation, TPC-1 and BCPAP cell supernatants were collected and frozen at -80°C until the day of the experiment. TGF-β1 was quantified using the

Human/Mouse TGF-β1 ELISA Ready-SET-Go! kit (eBioscience, Vienna, Austria) according

Response to low radiation dose is p53-dependent

to the manufacturer's instructions. In brief, the supernatants were thawed and activated using

1N HCl for 10 minutes, after which the HCl was neutralized with 1N NaOH. A 96-well

ELISA plate was coated with an antibody against TGF-β1, the supernatants were loaded onto the plate in triplicates and incubated overnight in the plate at 4°C. The plate was then washed, probed with primary antibody against TGF-β1, washed again and probed with a secondary antibody against the primary. A substrate buffer supplied with the kit was used to visualize the color and 1M H2SO4 was used as a stop solution. The resultant yellow color, corresponding to levels of TGF-β1, was measured using a spectrophotometer at 450 nm. A standard curve for purified TGF-β1 was used to correlate absorbance with concentration of

TGF-β1.

3.8 Detection of γ-H2AX foci Immunohistochemistry: TPC-1 cells were plated at 500,000 cells per coverslip and irradiated with various doses of X-rays. The cells were fixed with 4% paraformaldehyde 30 minutes after irradiation and probed overnight with a primary antibody against the phosphorylated form of the histone H2AX (γ-H2AX) (Abcam, Cambridge, UK) and subsequently a FITC- linked secondary antibody against the primary. The coverslips were mounted onto glass slides and the images were acquired using a widefield Nikon TE2000E epifluorescence microscope

(25 different frames/slide, 5 plains of depth of 1 µm thickness). Images were analyzed using

ImageJ software (Rasband, NIH, Bethesda, MA, USA) and nuclei and γ-H2AX spots along with their respective sizes were calculated as described in [122] (algorithm kindly supplied by

Dr. W. de Vos, Ghent University, Ghent Belgium). In total, around 1300 nuclei from four different coverslips were scored and the number of foci for each nucleus was reported by the algorithm. The algorithm also determined the average spot occupancy, the area of the nucleus

Chapter 2

occupied by one focus, as a means to determine the size of the foci independent of the cell cycle. The median number of foci per nucleus and spot occupancy and their interquartile range was calculated for each irradiation dose using Microsoft Excel and SPSS.

3.9 Senescence-associated (SA) β-galactosidase quantification by fluoroscence Cells were plated and irradiated with various doses of X-rays. The amount of β-galactosidase was measured 24 hours later by monitoring the conversion of a substrate to 4- methylumbelliferone (4-MU) at pH 6.0 (BetaFluor β-Galactosidase Assay Kit (Novagen, WI,

USA) as mentioned in the paper by Gary and Kindell [299]. β-galactosidase was quantified using a fluorometer (Fluoroskan Ascent CF, Thermo Labsystems, Waltham, MA, USA) with excitation at 360 nm and emission at 440. A flask of cells was kept at confluence for comparison. This method allowed for a more quantitative approach to β-galactosidase level measurement.

3.10 Cytokine Level Measurement using Multiplex Bead Assay Supernatants collected from both irradiation and sham-irradiated TPC-1 cells were incubated in a 96-well plate with dye-injected synthetic beads conjugated with antibodies against 90 different cytokines (Millipore Co., MA, USA). Each bead has a certain signature which helps to identify the associated cytokine. The cytokines were targeted with fluorescence-conjugated primary antibodies against that cytokine. The beads were passed through two intersecting lasers in a Luminex 100 instrument, one to excite the beads and the other to excite the probe.

The beads were sorted automatically and the average fluorescence was reported for each well.

Response to low radiation dose is p53-dependent

3.11 Statistical Analysis Unless otherwise stated, 1x106 cells were seeded into 25-cm2 tissue culture flasks 24 hours prior to irradiation. All experiments were done in biological triplicates except for the immunostaining experiments. The Dunnet test, a post-hoc test for multiple comparisons between irradiated and control conditions, was performed to compare experimental samples to the control using the SPSS software package version 11.5. For the ELISA experiment, a univariate analysis of variance (ANOVA) was performed with passage and radiation dose effect as factors. For γ-H2AX foci scoring, the Mann-Witney and Kruskal-Wallis mean rank test was used with a post-hoc modification for multiple comparisions. Results were considered significant at p<0.05. Boxplots were created using SPSS while all other graphs were created using GraphPad Prism version 5.00 (GraphPad Software Inc., La Jolla, CA,

USA).

Chapter 2

4. Results All experiments below have been performed on TPC-1 cells. Since results indicated a role for the p53 pathway, several experiments were repeated on BCPAP-cells to investigate the latter hypothesis.

4.1 Irradiation Causes an Increase in Amount of Double Strand Breaks in TPC-1 Cells To investigate the level of DNA damage upon radiation of TPC-1 cells with various doses of

X-rays, we quantified the amount of DNA double strand breaks (DSB) using a microscopic analysis 30 minutes after irradiation. This method not only provides a sensitive method to detect DNA damage foci but also allows an analysis of their size. Calculation of the foci size was done using a parameter called spot occupancy which measures the percentage of the nucleus occupied by the foci and adjusts for cell cycle variations and spot segmentation problems [141]. Irradiation at a dose of 62.5 mGy caused a slight increase in the percentage of nuclei with observable DNA damage from around 1% in control cells to 8% (see figure 15; p<0.05). At a dose of 4 Gy, the percentage of nuclei with foci greatly increased to around

99% of the nuclei. On the other hand, there was no increase in the median number of foci per nucleus at the lowest irradiation dose compared to the control: 0; [0–0] vs. 0; [0-0] foci/nucleus (median; [interquartile range]). Only at the higher doses was there an increase to

2; [0-9] for 0.5 Gy, and 26; [19-32] foci per nucleus for 4 Gy (figure 16A), both significant at p<0.05. Interestingly, the size of the foci as measured by spot occupancy also increased in a dose-dependent manner in cells irradiated with 0.5 Gy and 4 Gy compared to the control

(p<0.05) . However, for the low dose of 62.5 mGy, the size of the foci did not increase as compared to the control (figure 16B).

Response to low radiation dose is p53-dependent

Since an increase in DNA damage usually leads to a halt in proliferation for DNA repair, we evaluated whether these dose-dependent effects also affected cell.

Figure 15: DNA DSB detection using γ-H2AX immunostaining: Images of TPC-1 cells taken using a widefield Nikon TE2000E epifluorescence microscope fitted with a 40x objective lens. DNA damage foci were visualized by staining for γ-H2AX (FITC) while nuclei were stained with DAPI. Rows 1 to 4 are in order: control, low (62.5 mGy), medium (0.5 Gy), and high (4 Gy) of X-rays. The merge column represents the DAPI (blue) and FITC (green) channels.

Chapter 2

Figure 16: A) Box plot representing median spot number per nucleus vs. irradiation dose in TPC-1 cells 30 minutes post- irradiation. The box boundaries represent the upper and lower quartiles and the thick black line within represents the median. The whiskers represent the 95% confidence interval (CI) while the circles represent outliers and the asterisks extreme cases. An algorithm on ImageJ was used to count number of nuclei and foci in each nucleus. In total around 1300 nuclei were scored from four different coverslips and median of foci per nucleus was calculated. Low = 62.5 mGy, Medium = 0.5 Gy and High = 4 Gy. Only medium and high doses are significant (p<0.05) compared to the control.

B) Box plot representing spot occupancy vs. irradiation dose in TPC-1 cells 30 minutes post-irradiation. The box boundaries represent the upper and lower quartiles and the thick black line within represents the median. The whiskers represent the 95% confidence interval (CI) while the circles represent outliers and the asterisks extreme cases. Spot occupancy is used here as a more robust measure of spot size and was calculated using an algorithm on ImageJ. Average occupancy was calculated for each nucleus and the median calculated. Asterisks represent outliers. Low = 62.5 mGy, Medium = 0.5 Gy and High = 4 Gy. Only medium and high doses are significant compared to the control.

4.2 Irradiation Causes a Decrease in Number of TPC-1 Cells The two cell counting methods used (trypan blue counting and fluorescent microfluorospheres) yielded consistent results (Spearman correlation R=0.861). Low doses of

X-irradiation of TPC-1 cells did not cause any significant change in the number of cells, while higher doses caused a significant decline (Figure 17), both at 24 hours and 48 hours. After irradiation at 4 Gy, half the number of cells was present in cultures 24 hours after irradiation and around 40% at 48 hours compared to the relevant sham-irradiated control cultures. This indicates a possible increase in cell death or a decrease in proliferation (cell cycle arrest) in these cells upon irradiation, the scope of the next two paragraphs.

Response to low radiation dose is p53-dependent

Figure 17: Cell number vs. Irradiation dose in TPC-1 cells at 24 and 48 hours post-irradiation: histogram showing the number of cells (in millions) 24 and 48 hours after irradiation with various doses of X-rays as measured by counting fluorescent beads alongside TPC-1 cells in a flow cytometer. Error bars represent SD. Asterisks indicate significant results compared to control at p<0.05.

4.3 Cell Cycle Analysis Using PI Staining of DNA To check the effect of x-rays on the cell cycle, we analyzed propidium iodide (PI)-treated cells with flow cytometry. Cells were sorted according to their DNA content as measured by the amount of fluorescence emitted by each cell. The results are reported as the percentage of cells in each cell cycle phase to two decimal places. The data were analyzed using Coulter's

System II software for EPICS XL flow cytometers. At 24 and 48 hours post-irradiation there was a dose-dependent decline in number of S-phase cells in TPC-1. This decline was significant at all doses even for the lowest (62.5 mGy) at 24 hours, and reached a maximum of

75% at the highest irradiation dose (4 Gy) at 24 hours and 83% at 48 hours. These results indicate the possible activation of the G1/S and/or G2/M checkpoints in TPC-1 cells upon

Chapter 2

irradiation which is directly proportional to the irradiation dose. The cell distributions at 24 hours post-irradiation are depicted in figure 17. The percentage of cells in each cell cycle phase is detailed in table 1.

To compare the effect that p53 plays in the cell cycle response, we performed the same experiment on BCPAP cells which are known to harbor a mutated copy of p53 [298]. As in

TPC-1 cells, a decline in the number of cells in the S phase of the cell cycle was observed, but only at 1 and 4 Gy. However, there was a decline in G1 phase cells and an increase G2 phase cells. This indicates a possible activation of the G2/M checkpoint without G1/S checkpoint activation, which is not unexpected in cells with a mutant copy of p53 [130] (figure 18). The percentage of cells in each cell cycle phase is detailed in table 2.

Figure 18: Cell cycle distribution by PI staining: bar graph representing the percent of cells in pre-G1, G1, S, and G2. A) TPC-1 cells and B) BCPAP cells. Bars represent an average of 3 measurements. Error bars represent SD. Asterisks represent significant results at p<0.05 compared to the control.

4.4 Analysis of Cell Death Induced by Irradiation To distinguish between apoptosis and necrosis, annexin-V-PI staining was employed. The method depends on the detection of phosphatidylserine on the outer layer of the phospholipid bilayer, which is a sign of apoptosis, using an antibody. At the same time, propidium iodide is

Response to low radiation dose is p53-dependent

used as an indicator of membrane integrity. Therefore, annexin-V-positive and PI-negative cells are considered in early apoptosis, annexin-V-positive and PI-positive cells are considered to be in late apoptosis and PI-positive/annexin-V-negative cells are considered necrotic. A slight decline in the number of viable cells and an increase in apoptosis was observed. However, it was observed that, at 24 hours post-irradiation, the predominant mode of death was necrosis at all irradiation doses. This was also the case at 48 hours, except for the highest dose (4 Gy) where apoptosis became the most predominant.

To check whether apoptosis was further activated at later time points in response to irradiation, annexin V/PI double staining was also carried out on cells irradiated with X-rays after 72 hours. Compared to the control, TPC-1 cells 72 hours post-irradiation displayed an increase in basal levels of Annexin-V-positive cells. This could be attributed to the fact that the cells had been in culture for more than 72 hours without medium replenishment. However, apoptosis didn't appear to be significant as compared to the control except, again, for the highest dose of irradiation (4 Gy). At the same dose, there was an increase in necrotic cells too. At all time points, irradiation with low doses did not increase neither apoptosis nor necrosis significantly (Figure 19 A,B,C). On the other hand, the levels of apoptosis as measured by Annexin V/PI double staining did not increase in BCPAP cells 24 hours post irradiation (data not shown).

Chapter 2

Response to low radiation dose is p53-dependent

Figure 19: Apoptosis measurement using Annexin V/PI double staining: bar graphs representing the percentage of apoptotic and necrotic cells A) 24 hours, B) 48 hours, C) 72 hours post irradiation in TPC-1 cells. Apoptotic cells were taken as annexin V-positive/PI-negative and –positive cells whereas necrotic cells were taken as annexin V-negative/PI- positive cells. Results reported as percentage of cells ± SD. Asterisks represent significant results compared to the control at p<0.05.

4.5 Levels of Caspase-3 after Irradiation The increase in the levels of apoptosis at the later time points for 4 Gy led us to examine whether this was also reflected in the terminal marker caspase-3. Cleaved caspase-3 levels did not alter significantly at 24 hours post-irradiation but at 48 hours post-irradiation exhibited a dose-dependent increase in their levels with a significant up-regulation at 1 and 4 Gy (figure

20A). On the other hand, levels of cleaved caspase-3 did not increase significantly in the

BCPAP cell line neither at 24 nor 48 hours post-irradiation (figure 20B).

Subsequently we wanted to assess whether the significant effects of high-dose irradiation on

TPC-1 cells also induced morphologic alterations, particularly in comparison with the

BCPAP-cells.

Figure 20: Caspase-3 levels in response to irradiation: bar graphs displaying the levels of active caspase-3 in A) TPC-1 and B) BCPAP cells in response to various doses of X-radiation. Results reported as percentage of fluorescence ± SD. Asterisks represent significant results at p<0.05 compared to the control.

Chapter 2

4.6 Changes in Cellular Morphology Following Irradiation Senescence can be brought on by several factors, including stress such as radiation and DNA damage [194]. This state is defined by several morphological and molecular markers. We therefore performed microscopic observations of control and irradiated TPC-1 cells (figure

21A). The morphology of TPC-1 cells following irradiation shows alterations typical of stress induction. There is a marked increase in intracellular vesicle trafficking, as shown by the increased number of vesicles in the cytoplasm of irradiated cells (Figure 21B). Similarly, there is evidence of nuclear fragmentation in some cells at the higher doses of irradiation (4

Gy). In addition, cells appeared to be more flattened and larger in size especially at higher doses of irradiation (0.5 Gy and above) (Figure 21B). This morphology has been associated with an increase in cellular senescence. The appearance of these phenomena increased in a manner proportional to the dose of irradiation received. On the other hand, when we observed

BCPAP (Figures 22B,) cells under the microscope we didn't observe an increase in cell size and flattening in these cells as observed in TPC-1 cells (figures 22 A and B).

Next, we wanted to further elaborate on TPC-1 senescence, with a particular focus on the cytokines and chemokines involved. In the last part, we will further focus on the molecular differences possibly explaining the discrepancy in TPC1 and BCPAP response.

Response to low radiation dose is p53-dependent

Figure 21: Morphological alterations in TPC-1 cells in response to radiation: A) TPC-1 cells observed under an inverted light compound microscope (60x). B) TPC-1 cells stained with the May-Grünwald Giemsa stain (60x).

Chapter 2

Figure 22: Morphological alterations in BCPAP cells in response to radiation: A) BCPAP cells observed under an inverted light compound microscope (60x). B) BCPAP cells stained with the May-Grünwald Giemsa stain (40x).

Response to low radiation dose is p53-dependent

4.7 Effect of external X-irradiation on secreted levels of TGF-β1 Senescent cells are now found to exhibit a distinct secretory profile, termed a senescence- associated secretory phenotype (SASP) [190]. We therefore attempted to check the milieu of

TPC-1 cells for changes in secreted cytokines. A sandwich ELISA on the levels of TGF-β1 was performed on the supernatant of TPC-1 cells collected 24 hours after irradiation with various doses of X-rays. The supernatants were collected from cells of different passages and all supernatants were analyzed in one go in a single 96-well plate. An analysis of TGF-β1 levels and adjustment for passage number revealed that 0.5 Gy was the lowest significant dose to cause an up-regulation in the levels of TGF-β1 (Dunnet test; p<0.05) (Figure 23A). To confirm whether the lack of the morphological changes observed in senescent cells would translate itself to the excreted levels of TGF-β1, we perfored a sandwich ELISA on the supernatent of BCPAP cells similar to what was done on TPC-1 supernatents. Interestingly, there was no increase in the levels of TGF-β1 in the case of BCPAP at all irradiation doses.

Indeed, there was a decrease in TGF-β1 levels upon irradiation but when all three passages of cells were taken into account turned out not to be significant (p<0.05) (figure 23B).

Chapter 2

Figure 23: TGF-β1 levels in response to various doses of X-irradiation in TPC-1 cells: TGF-β1 concentrations (pg/ml) in the milieu of TPC-1 (A) and BCPAP (B) cells 24 hours post-irradiation as measured by sandwich ELISA. Asterisks represent significant results at p<0.05 compared to the control. An increase in the levels of TGF-β1 upon irradiation is evident at 0.5 Gy in TPC-1 cells but not in BCPAP. Error bars represent SD.

4.8 Effect of Irradiation on the Secreted Level of Cytokines Using a Multiplex Bead Assay We further inspected the effect of radiation on a panel of cytokines and chemokines using a multiplex bead assay. We analyzed around 90 different cytokines and chemokines and came up with 6 significantly regulated factors whose regulation was affected by radiation, these factors being eotaxin, granulocyte macrophage colony stimulating factor (GM-CSF), interleukins (IL)-6 and -8, vascular endothelial growth factor (VEGF), and monocyte chemotactic protein -1 (MCP-1). For eotaxin, GM-CSF, IL-8, and VEGF showed only significant changes in their levels at 4 Gy of X-rays where GM-CSF and IL-8 were up- regulated while VEGF and eotaxin were down-regulated. For IL-6 and MCP-1 their levels were significantly up-regulated at 1 and 4 Gy (p<0.05) (figure 24).

Response to low radiation dose is p53-dependent

Figure 24: Concentration of secreted cytokines by multiplex bead assay: measured concentrations (pg/ml) of various cytokines in the milieu of TPC-1 cells 24 hours post-irradiation. A = eotaxin and GM-CSF, B = IL-6, C = IL-8 and VEGF, D = MCP-1. Asterisks represent significant results at p<0.05 compared to the control. Error bars represent SD.

4.9 Translational Levels of Molecular Markers by Western blotting To get an insight into the molecular mechanisms at work in the two cell lines that could underlie the differences in response to low and high doses of radiation, we checked the levels of some of the central players in DNA damage response. A central player in the cellular DNA damage response is the tumor suppressor, p53 [300]. As can be seen in figure 25, p53 levels increase dose-dependently in irradiated TPC-1 cells starting at a dose of 0.5 Gy. To further confirm that p53 was indeed active, the levels of one of its downstream targets, namely p21, were checked. A Western blot of p21 showed a dose-dependent up-regulation upon irradiation

Chapter 2

at doses of 0.5 Gy and above. This up-regulation is similar to that observed with p53. We also analyzed the response of another p53 family member, p73, to irradiation and found it to be absent in this cell line. To assay this cell line's response to apoptosis, the levels of two anti- apoptotic proteins, Bcl-2 and Akt, were assessed and were found to be up-regulated at all doses of radiation except 0.5 Gy (Figure 25 A&B). The lack of up-regulation at 0.5 Gy may point to the fact that Bcl-2 and Akt are not the only two anti-apoptotic factors at play here and there may be other players such as Bcl-XL.

As mentioned above, upon irradiation, cells presented signs of senescence such as distinct morphological features, inhibition of proliferation, resistance to apoptosis and an altered secretory profile, and therefore, we checked the effect of radiation on p16INK4a translational levels. As shown in figure 24C radiation caused an increase in the levels of p16 in TPC-1 cells. The levels of p16 appear to be up-regulated even at the low dose of irradiation (62.5 mGy) which warrants further research into the role of low doses of radiation in senescence.

To check the effect of p53 on induction of p16 and senescence-like profile, we checked the effect of a range of external X-rays on levels of p16 and p73 in BCPAP cells. The tumor suppressor p53 was detected in BCPAP cells but upon irradiation the protein levels of p53 did not increase and neither did the levels of p73 and p16 (Figure 26).

Response to low radiation dose is p53-dependent

Figure 25: Response of various molecular markers to radiation in TPC-1 cells as measured by Western blotting: translational levels of A) the p53 family members p73 and p53 and their downstream effector p21 B) anti-apoptotic factors Akt and Bcl-2 and C) p16. β-tubulin was used as a housekeeping protein.

Chapter 2

Figure 26: Response of various molecular markers to radiation in BCPAP cells as measured by Western blotting: translational levels of p73, p53, and p16 in BCPAP cells. β-tubulin was used as a housekeeping protin.

4.10 Effect of Irradiation on levels of β-galactosidase The increase in the levels of β-galactosidase measured at pH 6.0, now named senescence- associated β-galactosidase (SA β-gal), is one of the hallmarks of senescence but not limited to it [188]. To measure the levels of this enzyme in TPC-1 cells, we adapted a fluorescence assay and used various passages of TPC-1 cells as mentioned above. We noticed that fluorescence levels increased upon irradiation of TPC-1 cells starting at 0.25 Gy (p<0.05).

When cells were harvested until they formed a confluent monolayer, an increase in fluorescence was observed but was not significant when compared to control (p<0.05) (Figure

27).

Response to low radiation dose is p53-dependent

Figure 27: β-galactosidase level measurement in TPC-1 cells by fluorescence: measured levels of SA β-galactosidase levels in TPC-1 cells 24 hours post-irradiation. Confluent cells represent cells that were grown until fully covering the tissue culture flask bottoms while blank refers to case with no cells. Asterisks represent significance at p<0.05 compared to the control. Error bars represent SD.

Chapter 2

5. Discussion Epidemiological evidence indicates a relationship between exposure to ionizing radiation and the induction of thyroid cancer, most notably following the study of children exposed to external irradiation and the victims of the Chernobyl disaster [163, 164, 301]. In addition,

Cardis et al. [114] showed that there was a significant increase in the risk of certain cancers in workers of nuclear facilities [113]. Radiation in the form of radioiodine is also used as a treatment for thyroid cancer albeit at higher doses [302] although recent papers have explored the possibility of using lower doses of radiation in cancer therapy for certain cancers [156-

158]. The risk of thyroid cancer from low dose radiation exposures, such as those received from the environment, or from medical and workplace exposures, is uncertain.

We used a model system of two cell lines of papillary thyroid carcinoma origin, one containing a wild-type copy of p53 and one with a mutated copy and irradiated them with low

(62.5 mGy) and moderate to high doses of X-rays (0.25, 0.5, 1, and 4 Gy) and monitored several parameters that were changed at various irradiation doses. We have shown in this study that a low dose of radiation (62.5 mGy) elicits a response in cells with wt p53 (TPC-1) which is basically different from the response at higher doses and that could be mainly due to the difference in the molecular players involved. At the former dose, we observed a significant increase in percentage of cells with double stranded breaks, associated with a decrease in the fraction of cells in the S phase of the cell cycle and an upregulation of p16 in

TPC-1 cells, but not in BCPAP indicating p53-dependency.

Response to low radiation dose is p53-dependent

The drop in the number of cells in S phase and increase in cells in the G2/M in TPC-1 cells at doses higher than 0.5 Gy was similar in our study at 24 and 48 hours. This might be due to the fact that TPC-1 cell line maintains a functional copy of p53. Although Frasca et al. (2003)

[303] sequenced the TP53 gene in several papillary thyroid carcinoma cell lines and found that TPC-1 harbored an inactivating K286E point mutation in p53, the work of Meireles et al. has shown that TPC-1 carries wild-type p53 [298]. Our results confirm the latter and point in the direction of TPC-1 having at least one functional copy of p53. Translational levels of p53 were up-regulated upon irradiation with X-ray doses of 0.5 Gy and above. p53 is known to activate the cyclin dependent kinase inhibitor (CDKN1A), p21, which is able to cause a G1/S checkpoint activation [304]. Analyzing the levels of p21 revealed a similar increase in p21 levels upon X-ray doses of and above 0.5 Gy. A similar decrease in cells in the S phase was noted by Namba et al. when they irradiated the NPA papillary thyroid carcinoma cell line, at doses above 0.5 Gy and this was also found to be p53-dependent [305]. Another member of the p53 family, p73, has recently been found to be a tumor suppressor and to share the pathways that are associated with p53 [306, 307]. Therefore, p21 could feasibly be up- regulated by p73 and not by p53, and thus it warranted further investigation. By probing for p73, we found that there were no translational levels of this protein in TPC-1 cells. Frasca et al. came to a similar conclusion when checking for transcriptional changes of p73 in TPC-1 cells by RT-PCR [303].

The up-regulation of p16 on the other hand did not follow the same trend as p53 activation and might be responsible for the decrease of the fraction of cells in S phase at the low dose of radiation. On the other hand, BCPAP, a cell line which harbors a mutant copy of p53, did not show any changes in the parameters measured at the low dose of irradiation. Evidence of

Chapter 2

cross-talk between the p53-p21 and p16-pRb pathways has existed for some time [203, 204,

308]. However, the exact nature of this cross-talk is not entirely clear. It has now emerged that p53 may inhibit p16 levels and that in the absence of wild-type p53, p16 levels increase as a back-up mechanism [197, 200]. Our results suggest that the p16 response to radiation is dependent on the p53 status of the cells where the former was up-regulated in response to radiation in the only TPC-1 cell line with wt-p53. However, whether the one activates the other directly is not certain. Furthermore, the levels of p16 increased in response to various doses of X-rays despite the presence of a p53-mediated response. There was a decrease in the fraction of cells in the S phase only at 1 and 4 Gy and no increase in apoptosis as measured by the levels of active caspase 3 or the observed changes in morphology associated with senescence. This was associated with no changes in the translational levels of p53, p73 and p16 in these cell upon external X-irradiation.

The up-regulation of p16 and p21 has been linked to the appearance of a senescence-like phenotype in cells. Senescence was found to proceed via two interconnected pathways in which p53 and p16INK4a lie at the center (refer to the concise review of Cichowski and Hahn

[194]) and the appearance of a senescent profile was mainly linked to the different dynamics of p21 and p16 where p21 is important for the appearance of the senescent profile while p16 maintains the cells in that state [201]. The appearance of a senescent-like profile in the TPC-1 cells at 0.5 Gy and above and its absence at 62.5 mGy may point to the fact that the concomitant activation of both p21 and p16 and not only the latter is necessary for the appearance of this profile. These different dynamics in TPC-1 cells resemble those that were reported for melanocytes upon irradiation with different doses of UV [200]. The difference in the appearance of a senescent-like profile between low and high doses of radiation is reflected

Response to low radiation dose is p53-dependent

in the increase in DNA DSBs in response to radiation; TPC-1 irradiated with a low dose of radiation (62.5 mGy) showed an increase in the percent of nuclei displaying DNA damage with no change in median number of foci or focus size consistent with earlier reports that indicated absence of these phenomena below 0.1 Gy [133]. At higher doses (0.5 and 4 Gy), the percent of nuclei showing DNA damage, the median number and spot occupancy increased significantly compared to the control. More significantly image analysis revealed a great increase in the size of the foci at 4 Gy. The size of the foci is indicative of the amount of

DNA damage and the failure to repair the DNA damage caused the size of the foci to increase. Rodier et al. [190] found that failure to repair DNA damage after 24 hours at a dose of 10 Gy led to appearance of large foci and was associated with a senescent profile.

The appearance of this senescent state at moderate to high doses of radiation would also explain the lack of apoptosis. The up-regulation of such markers as Bcl-2 and Akt could also explain the lack of apoptosis due to the reported role of those two molecules in the anti- apoptotic response [309, 310]. In fact, the up-regulation of Bcl-2 was found to push the cellular response to DNA damage from the apoptotic to the senescent [219]. However, this does not explain the lack of apoptosis at lower doses of radiation especially since previous research has indicated that transformed cell lines of adenocarcinoma and glioma with a functional copy of p53 display low-dose hyper-radiosensitivity which is cell cycle- and apoptosis-dependent [134, 135]. Perhaps the low number of nuclei that exhibit DNA damage and the slight increase in the levels of Bcl-2 and Akt at low doses contribute to this effect.

Another feature of senescent cells is an alteration in their secretory profile which has been reported at low doses of radiation (0.2 Gy) with noticeable effects on the microenvironment

[178]. This secretory associated senescent profile (SASP) has been deemed an important

Chapter 2

player in the tissue microenvironment and in the maintenance of the senescent state in cells

[175, 176]. Our results, point to an alteration of the secretory profile of TPC-1 cells only at doses of 0.5 Gy and higher. We identified six factors that were altered upon X-irradiation:

TGF-β1, VEGF, Eotaxin, IL-6, IL-8, and MCP-1. TGF-β1 was found to contribute to inhibition of proliferation of TPC-1 cells through SMAD2/3 and up-regulation of p21 [216].

IL-6 and -8 are of particular importance since they are involved in the inflammatory response and were linked to the induction and maintenance of senescence and were found to be the two most robustly expressed of the SASP panel [176, 190]. These factors among others have been already detected in the supernatant of stress-induced senescent cells [191].

In conclusion, our results indicate that cell lines of papillary thyroid carcinoma respond to external X-irradiation with a change in their cell cycle distribution irrespective of their p53 status. However, only when a functional copy of p53 is present do the cells alter their cell cycle distribution in response to low doses of radiation, concomitant with p16 up-regulation and up-regulation of some anti-apoptotic markers. On the other hand, a senescent-like profile only appears in these cells at higher irradiation doses. Finally, it might be said that the response of the TPC-1 cells is dependent on the level of DNA damage brought on by irradiation. At a low dose of radiation, there is no increase in the number of DNA damage foci per nucleus and thus p16 up-regulation may cause a reversible cell cycle arrest until the DNA damage is fixed. At higher doses of radiation, DNA damage is more severe and p53, p21, and p16 are up-regulated and this leads to a non-reversible senescent-like profile.

Acknowledgements

Response to low radiation dose is p53-dependent

The authors are grateful to Prof. Michael Atkinson (Helmholtz Zentrum, Munich, Germany) for his critical reviewing of the manuscript and for his continuous support. Khalil Abou-El-

Ardat is supported by a doctoral SCK•CEN/Ghent University grant. This work was funded by the EU Euratom Program (GENRISK-T contract FIP6-2006-036495 on “defining the genetic component of thyroid cancer risk at low doses”) and the DoReMi NoE agreement 249689 on

'low dose research towards multidisciplinary integration').

Chapter 2

Table 1: Percentage of cells in each cell cycle phase in TPC-1 cells. Numbers reported are: percent of cells (%) ± SD.

Dose (Gy) G1 S G2 0 67.27 ± 0.99 9.52 ± 0.51 15.39 ± 0.75 0.0625 67.45 ± 2.01 8.99 ± 0.19 15.69 ± 0.70 0.25 66.58 ± 1.44 8.49 ± 0.38 16.33 ± 1.72 0.5 65.38 ± 0.85 7.97 ± 0.32 18.81 ± 1.16 1 80.01 ± 1.17 3.75 ± 0.15 13.31 ± 0.38 4 69.70 ± 1.42 2.28 ± 0.11 22.76 ± 0.83

Response to low radiation dose is p53-dependent

Table 2: Percentage of cells in each cell cycle phase in BCPAP cells. Numbers are reported as: percent of cells (%) ± SD.

Dose (Gy) G1 S G2 0 82.34 ± 1.89 5.77 ± 0.49 11.09 ± 1.30 0.0625 82.54 ± 2.00 6.52 ± 1.28 10.41 ± 0.63 0.25 79.46 ± 3.33 6.58 ± 1.16 13.42 ± 3.79 0.5 79.56 ± 1.82 5.16 ± 0.81 14.76 ± 1.24 1 82.71 ± 1.13 3.73 ± 0.74 13.21 ± 1.31 4 69.74 ± 1.84 2.53 ± 0.60 27.41 ± 2.09

Chapter 3: Low Dose Irradiation of Thyroid Cells Reveals a Unique Transcriptomic and Epigenetic Signature in RET/PTC-positive Cells

Low dose irradiation reveals distinct signature

Modified from: Abou-El-Ardat K, Monsieurs P, Anastasov N, Atkinson M, Derradji H, De Meyer T, Bekaert S, Van Criekinge W, Baatout S. Low dose irradiation of thyroid cells reveals a unique transcriptomic and epigenetic signature in RET/PTC-positive cells. Mut. Res. (2011), doi: 10.1016/j.mrfmmm.2011.10.006

100

Chapter 3

Low dose irradiation of thyroid cells reveals a unique transcriptomic and epigenetic signature in RET/PTC-positive cells

1. Abstract The high doses of radiation received in the wake of the Chernobyl incident and the atomic bombing of Hiroshima and Nagasaki have been linked to the increased appearance of thyroid cancer in the children living in the vicinity of the site. However, the data gathered on the effect of low doses of radiation on the thyroid remain limited. We have examined the genomewide transcriptional response of a culture of TPC-1 human cell line of papillary thyroid carcinoma origin with a RET/PTC1 translocation to various doses (0.0625, 0.5, and 4

Gy) of X-rays and compared it to response of thyroids with a RET/PTC3 translocation and against wild-type mouse thyroids irradiated with the same doses using Affymetrix microarrays. We have found considerable overlap at a high dose of 4 Gy in both RET/PTC- positive systems but no common genes at 62.5 mGy. In addition, the response of RET/PTC- positive system at all doses was distinct from the response of wild-type thyroids with both systems signaling down different pathways. Analysis of the response of microRNAs in TPC-1 cells revealed a radiation-responsive signature of microRNAs in addition to dose-responsive microRNAs. Our results point to the fact that a low dose of X-rays seems to have a significant proliferative effect on normal thyroids. This observation should be studied further as opposed to its effect on RET/PTC-positive thyroids which was subtle, anti-proliferative and system- dependent.

101

Low dose irradiation reveals distinct signature

2. Introduction Radiation exposure in childhood has been linked to the increased appearance of thyroid cancer [311]. The link between radiation and thyroid cancer was enforced when in 1986 the release of radioactive iodine in the environment after a meltdown in reactor 4 of the

Chernobyl nuclear power plant led to an increase in cases of papillary thyroid carcinoma

(PTC) in the areas surrounding the site in Ukraine, Belarus, and Russia. This effect was most pronounced in children [89, 90]. A large percentage of these cases were found to carry a specific translocation, called rearranged in transformation/papillary thyroid carcinoma

(RET/PTC) translocation, which was found to be enough to initiate transformation to PTC.

Indeed, when the RET/PTC recombinant gene was inserted in animal models under the control of the thyroglobulin (Tg) promoter, it caused the appearance of thyroid tumors with

PTC morphology [87]. RET/PTC translocations are caused by the fusion of the RET kinase domain with the N-terminal part of various genes. The most common variants, RET/PTC1 and RET/PTC3 are caused by the fusion of the RET kinase domain to H4/D10S170 and

RFG/ELE1, respectively [79, 81]. This translocation was found to overexpress the mitogen activated protein kinase (MAPK) pathway, the c-Jun NH2-terminal protein kinase (JNK), and phosphatidylinositol 3-kinase (PI3K) signaling pathways [83-85]. When comparing radiation- induced PTC with sporadic tumors, the frequency of RET/PTC translocation was much higher in the radiation-induced PTC which could be attributed to the higher frequency of DNA double-strand breaks after irradiation [91, 92]. When compared to sporadic PTC tumors, radiation-induced PTC from Chernobyl were found to possess a unique transcriptomic signature [97, 98].

The effects of radiation on the thyroid were found to hold for high doses (i.e. above 150 mSv

[142]) but limited evidence exists in the literature for lower doses. One of the few examples of

102

Chapter 3

the latter is the follow-up study in Israel of children who had received low doses of 9 cGy to the thyroid and which revealed an increase in thyroid tumors with a higher frequency of

RET/PTC translocations [96]. The most commonly used method for assessing effects of low doses of radiation is the linear-no-threshold (LNT) model. However, this model has been contested recently as not being fully accurate [119, 296]. Using microarrays to search for radiation-induced signatures has been done before on blood and on moderate to high doses

[312, 313]. However, no research has been performed on the effect of low doses of irradiation on thyroids. Our previous work showed that low doses of X-rays (62.5 mGy) had a distinct molecular and phenotypic response from moderate to high doses (> 0.5 Gy) on TPC-1, a cell line harboring a RET/PTC1 translocation and a wild-type copy of p53 [314]. In this paper, we investigated the genome-wide transcriptional response of TPC-1 cells to low and high doses of X-rays and compared it to the genome-wide response of murine thyroids both with and without the RET/PTC translocation. Our aim in using both TPC-1 cells and RET/PTC- positive mouse thyroids was the identification of radiation-responsive genes in two independent RET/PTC-positive human and murine systems in vivo and in vitro and contrast that with genes in wild type thyroids. In addition, this will allow us to compare and contrast genes involved in the radiation response in a system of well-defined PTC characteristics

(TPC-1) with one that does not possess all characteristics of thyroid carcinoma and thus identify a RET/PTC-dependent response. Our results point to a unique transcriptional and microRNA response in RET/PTC-bearing thyroids exposed to a low-dose of radiation.

103

Low dose irradiation reveals distinct signature

3. Materials and Methods

3.1 Cell Lines TPC-1 cells, which are human cells of PTC origin and possessing an inherent RET/PTC1 translocation, were obtained from Dr. Horst Zitzelsberger (HelmholtzZentrum München,

GmbH, Munich, Germany) and were grown in Dulbecco's Modified Essential Medium

(DMEM) supplemented with 10% v/v of fetal bovine serum (FBS) (Gibco, UK). The cells were kept in a humidified incubator (37°C, 5% CO2) and were split every two days. On the day of the experiment, 1x106 cells were seeded into 25-cm2 tissue culture flasks and kept overnight in a humidified incubator. On the next day, the cells were irradiated with various doses of X-rays and returned to incubator until harvest. Cells of three different passages (p22, p29, and p37) were used in the experiments.

3.2 Mice C57BL/6J mice of the species Mus musculus were genetically modified to express the

RET/PTC3 translocation under the control of bovine thyroglobulin (Tg) promoter and were obtained from Dr. Jacques Dumont (Université Libre du Bruxelles (ULB), Brussels,

Belgium). These mice were developed by Santoro's group and described in the literature [94,

315]. The mice were bred in our animal facility and mated with wild-type C57BL/6J mice.

The progeny were then genotyped to identify the presence or absence of the RET/PTC translocation as described in section Mouse Genotyping. Treatment of the mice was done in accordance with the ethical guidelines for treatment of laboratory animals as specified in the national legislation and under the continuous supervision of the veterinarian. The local ethical committee of SCK•CEN approved the use of animals in this study (Request no. 08-003).

104

Chapter 3

3.3 Mouse Genotyping Pieces of the tails of mice were cut and the DNA extracted using High Pure PCR Template

Preparation Kit (Roche, Germany). The resultant DNA's concentration and purity was checked using the Nanodrop spectrophotometer (Thermo Scientific, USA). To check for the presence or absence of the RET/PTC translocation, 200 ng of DNA was mixed with 5 µL of each of the forward and reverse primer solutions (see table 3) and 10 µL of 5xC Taq&LOAD

Mastermix (MP Biomedicals, USA) and nuclease-free water was added to complete the volume to 50 µL. The DNA was amplified using a PCR program consisting of a single-step

94°C for three minutes followed by 33 cycles of 94°C for one minute, 60°C for one minute and 72°C for one minute and a final step of 72°C for seven minutes. The PCR products were run on a 1.5%-agarose gel at 150 V for 2 hours and viewed under UV light.

3.4 Irradiation Cells were plated in 25-cm2 tissue culture flasks one day before irradiation at 1x106 cells/flask. Mice (both normal and RET/PTC-transgenic) were assigned to one of four groups

(control, low dose, moderate dose, high dose) each group containing four mice. The mice had an average age of three months and were of mixed sexes (two males and two females in each group). The mice were restrained in a plastic box separated into several chambers; each chamber is enough to restrict the movement of one mouse. Both cells and mice were irradiated with 250 kV, 15 mA, 1 mm Cu-filtered X-rays (delivered at 5 mGy/sec from a

Pentak HF420 RX machine) while non-exposed cultures and mice were sham irradiated using a procedure identical to the 4 Gy irradiation. Doses delivered were calculated using the same criteria used in medical practice to calculate doses given to tissues (i.e. 95% of exposure in

Roentgen). Cells were then returned to humidified incubator until harvest. Mice were returned to their respective cages until sacrifice.

105

Low dose irradiation reveals distinct signature

3.5 RNA Isolation Six hours after irradiation, cells were washed twice with 1x PBS and collected in 1 mL of

TRIzol® (Invitrogen, CA, USA) by scraping and frozen at -80°C. For mice, six hours after irradiation, they were killed by asphyxiation using CO2 after which the thyroids were removed aseptically and stored in 1 mL of RNAlater (Ambion, USA) solution at -80°C. On the day of experiment, 500 µL of TRIzol was added to the thyroids and cells were dissociated and lysed using a hand-held mechanical homogenizer. For both cells and thyroid tissues, the

RNA was extracted using an improved Guanidinium thiocyanate-phenol-chloroform extraction method as described by Chomczynski and Sacchi [316]. Concentration and purity of RNA were determined using the Nanodrop spectrophotometer (Thermo Scientific, USA) while RNA integrity was determined using the RIN number (Agilent's lab-on-chip

Bioanalyzer 2100, Agilent Technologies, USA). Only RNA with a RIN higher than 9.0 was used.

3.6 Protein Extraction and Western blotting Twenty four hours after irradiation, cells were collected by trypsinization, washed twice with

1x PBS and then centrifuged at high speed and the PBS removed carefully before the pellet was frozen at -80°C. Concerning the mouse experiment, 24 hours post-irradiation, mice were sacrificed by asphyxiation as mentioned above and the thyroids were removed aseptically and snap-frozen using liquid nitrogen. The thyroids were then stored at -80°C until the day of the experiment. Proteins were isolated from TPC-1 cells using Bio-Rad's ReadyPrep II (Bio-Rad,

Germany) supplemented with protease inhibitor cocktail (Roche, Germany) and PMSF

(Sigma, MO, USA). For thyroids, proteins were extracted using Tissue Extraction Buffer I

(Invitrogen, CA, USA) supplemented with cOmplete ULTRA Mini EDTA-free protease 106

Chapter 3

inhibitor cocktail tablets (Roche, Germany) and homogenization using a hand-held mechanical homogenizer. The proteins were quantified using the Bradford solution (Bio-Rad,

Germany) for TPC-1 protein extracts and the bicinchoninic acid (BCA) method (Sigma-

Aldrich, MO, USA) for thyroid protein extracts. Thirty µg of TPC-1 proteins and 60 µg of thyroid proteins were loaded onto 10% polyacrylamide gels and run at 100V for two hours.

The protein bands were then transferred electrically onto a PVDF membrane using Bio-Rad's semi-dry transfer unit according to manufacturer's instructions. The membrane was probed overnight with antibodies against MDM2 (90 kDa), RFPL1 (33 kDa) (Santa Cruz

Biotechnology, CA, USA), GDF15 (35 kDa), GDF5 (55 kDa), Aurora A Kinase (AURKA, formerly STK6, 43 kDa), cyclin G (CCNG, 30 kDa), p53 (53 kDa), phospho p53 (Ser15) (53 kDa), SMAD3 (48 kDa), and phospho SMAD3 (Ser423/Ser425) (48 kDa) (Abcam, UK). The membranes were then probed with horseradish peroxidase-conjugated secondary antibodies and treated with ECL reagent (GE Healthcare, UK). Bands were visualized on Fuji X-ray film

(Tokyo, Japan) and scanned to obtain a digital image of the bands.

3.7 MicroRNA Analysis The RNA extracted from irradiated and sham-irradiated TPC-1 cells mentioned above was reverse transcribed into cDNA and the quality and presence of microRNA (miRNA) was then controlled by qPCR for a control miRNA (RNU44). Afterwards, the remaining RNA was reverse transcribed into cDNA and amplified and the result was loaded onto TaqMan Low

Density Array Human MicroRNA Panel v1.0 (Applied Biosystems, USA). A qPCR was performed using a 7900HT Fast Real-Time PCR system fitted with a 384-well block (Applied

Biosystems, USA). Results were normalized to a housekeeping miRNA (RNU44) and fold change of miRNA between irradiation dose and control condition was calculated using the

-ΔΔCT formula 2 where the CT value is the cutoff value of the miRNA, ΔCT is the difference 107

Low dose irradiation reveals distinct signature

between each CT value and its housekeeping counterpart and the ΔΔCT is the difference between each ΔCT of experimental condition and control condition. The average fold-change between the three biological replicates was calculated and only miRNAs which were significant at p<0.05 as explained in the Statistical Analysis section were reported. The predicted targets of significant miRNAs were retrieved from the Sanger database using

TargetScanHuman (www.targetscan.org) [317].

3.8 Affymetrix Microarray Preparation The RNA extracted from TPC-1 cells earlier was quantified and treated using GeneChip®

WT cDNA Synthesis and Amplification Kit according to the manufacturer's instructions. The

RNA was then hybridized to Affymetrix Human Gene GeneChip® Gene 1.0 ST Arrays in the case of TPC-1 cells and Affymetrix Mouse Gene GeneChip® Gene 1.0 ST Arrays

(Affymetrix, Santa Clara, USA) in the case of mouse thyroids. The chips were all scanned at once. In total, 12 chips were prepared for TPC-1 cells (Four doses [0, 0.0625, 0.5, and 4 Gy] and three different passages mentioned in the section Cell Lines). For the mouse thyroids, 24 chips were prepared, 12 for the wild-type thyroids, and 12 for the RET/PTC-positive thyroids.

The chips were subdivided into four doses (0, 0.0625, 0.5, and 4 Gy) each done in triplicates.

3.9 Affymetrix Microarray Data Analysis Raw Affymetrix data were preprocessed using the "Affy" package (ver. 1.22.0) in

BioConductor (ver. 2.4/R version 2.9.0). Background correction was done using the Robust

Multichip Average (RMA) convolution model [318] while quantile normalization was used to homogenize expression values from multiple array chips [319]. The intensities obtained from multiple probes belonging to one probeset were summarized into one expression value per gene using the median polish approach [318]. To test for the differential expression between 108

Chapter 3

the different irradiated conditions and the reference conditions (no irradiation), we used the

Bayesian adjusted t-statistics as implemented in the "LIMMA" package (ver. 2.18.0) [320]. p- values were corrected for multiple testing using the Benjamini and Hochberg method to control for the false discovery rate (FDR) [321]. The statistical power of the test was calculated using Lee and Whitmore's method [322]. For pathway analysis (ArrayTrack ver.

3.3.0; National Center for Toxicological Research, USA), genes that were significant at p<0.005 in the case of TPC-1 cells and p<0.05 for the mouse thyroids were used with no cut- off value on the fold-change. Since our main interest was the effect of low doses of irradiation on gene level, we were interested in gene expression changes no matter how slight. Since we heavily rely on the pathway analysis in the entire study, a Benjamini and Hochberg smart

FDR correction method was employed for the pathways. Pathways with a p-value less than

0.05 after FDR correction were considered highly significant.

3.10 Real-time RT-PCR (qRT-PCR) Extracted RNA was reverse transcribed into cDNA using TaqMan Reverse Transcription kit

(Applied Biosystems, USA). The cDNA was then quantified and 6 µg of that was pipetted into a 96-well PCR plate with 17 µL of SYBR Green mix (MESA Green qPCR MasterMix

Plus for SYBR Assay I Low ROX, Eurogentec, Belgium) and primers for Pax-8 and α- tubulin (see table 1 for primer sequences). Primer sequences were blasted using Blastn

(http://blast.ncbi.nlm.nih.gov) and the target gene verified to be the most significant targets.

Amplification of target sequences was performed on the ABI Prism 7500 (Applied

Biosystems, USA) using the following program: 5 minutes at 95°C, followed by 40 cycles of

15 seconds at 95°C and 1 minute at 60°C. A melting curve was used to guard against the appearance of primer-primer dimers. The fold-change was calculated using the same method described above for miRNAs after ensuring the same efficiency for both primers. 109

Low dose irradiation reveals distinct signature

3.11 Multiplex Bead Assay Analysis of Protein Phosphorylation TPC-1 cells were lysed using lysis buffer (Millipore) supplemented with cOmplete ULTRA

Mini EDTA-free protease inhibitor cocktail tablets (Roche, Germany). The lysate was purified on filter columns (Millipore) by centrifugation at 12,000 g. The lysates were quantified using the BCA assay and 10 µg of the lysate was loaded onto a filter-bottomed 96- well plate. For thyroids, 25 µg of the lysates from section 2.6 were loaded onto the filter- bottomed 96-well plate. The manufacturer's instructions were followed for the rest of the experiment. In summary, five bead sets injected with specific dye signatures against ERK1/2, signal transducer and activator of transcription (STAT)3, JNK, p53, and Rb (Millipore) were added to each well and incubated overnight. On the next day, the wells were washed twice and secondary antibody against total ERK1/2, phosphorylated ERK1/2 (Thr185/Tyr187), total p53, phosphorylated p53 (Ser15), total JNK, phosphorylated JNK (Thr183/Tyr185), total

STAT3, phosphorylated STAT3 (Tyr705), total Rb, and phosphorylated Rb (Thr252)

(Millipore) were added to each well. A streptavidin/phycoerythrin substrate was added to each well followed by an amplification buffer (Millipore). The samples were analyzed using a

Luminex 100 sorter (Luminex, USA) which utilizes two lasers to excite the different beads

(for analyte detection) and the antibodies (for analyte concentration). The results are reported as mean fluorescence intensity (MFI) for each well with the assumption that the higher the

MFI, the more analyte there is. For each analyzed protein, at least one positive and one negative control were analyzed.

3.12 Statistical Analysis For the microRNA and qRT-PCR analysis, the Dunnet test, a post-hoc one-way ANOVA for multiple comparisons between a test and a control condition was used. Results were 110

Chapter 3

considered significant at p<0.05. For the multiplex bead assay, a Dunnet test was used for significantly regulated protein analysis while a two-factor univariate ANOVA was used to analyze the effect of cell type. A Spearman's Rho test was used to correlate presence of

RET/PTC translocation with response to radiation. Analyses were performed using SPSS

Statistics ver. 17 (IBM Corporation, NY, USA).

111

Low dose irradiation reveals distinct signature

4. Results

4.1 Thyroid Size and RET/PTC Status in Mice The RET/PTC status of the experimental mice was checked by PCR using specific primers

(see table I). RET/PTC-positive mice displayed a band at 1 kbp on a 1.5% agarose gel whereas wild-type thyroids lacked this specific band (see fig. 28). Based on visual examination, the size of the thyroid in RET/PTC-positive mice was found to be consistently larger than their counterparts in wild-type C57BL/6J mice (fig. 29).

Figure 28: PCR performed on DNA from RET/PTC-positive and –negative mice were run on a 1.5% agarose gel. Lanes A, B, C, H, I, L, and M possess the 1-kb amplicon and were considered RET/PTC-positive. The other lanes did not possess the amplicon and were considered negative.

A B

Figure 29: Relative size of the thyroids taken from RET/PTC-negative (A) and –positive (B) mice. RET/PTC-positive thyroids were consistently larger than their –negative counterparts.

112

Chapter 3

4.2 Differential Expression of Genes Between RET/PTC-positive and – negative thyroids To check for markers of PTC and the effect of the RET/PTC translocation on gene expression, we compared the expression levels of three RET/PTC-positive thyroids to the gene expression of three RET/PTC-negative thyroids. We took a subset of genes that had a p- value below 0.001 and a fold-change above 2. When we took this list of 57 genes and checked for the most enriched pathways using ArrayTrack, the Ras signaling pathway appeared as the most significant pathway (p-value 0.03) up-regulated in RET/PTC-positive thyroids. When we relaxed the p-value to 0.05 while retaining the fold-change cut-off of 2, the cell cycle appeared among the most significant pathways (p=0.0005) . When we searched for markers of

PTC mentioned in the literature we came up with several genes described as possible biomarkers such as keratin 19 (KRN19) (FC = 10.8) [61], tissue inhibitor of metalloproteinase-1 (TIMP1) (FC = 1.7) [323], and cyclin D2 (CCND2) (FC = 4.4) [62]. We also found considerable but not total overlap with the results obtained by Burniat et al. (2008) in the same mice [324]. It is worth noting that in our original list of 57 genes, only two genes were down-regulated and all the rest up-regulated possibly pointing to the activation of a transcription factor or inactivation of an inhibitor. In addition, several markers of thyroid differentiation and function were up-regulated such as thyroglobulin (Tg) and thyroid peroxidase (TPO) which does not correspond with previous data.

4.3 Microarray Analysis of TPC-1 Cells Analysis of the most significant genes (p<0.005) differentially expressed upon external X- irradiation using KEGG pathway and PathArt in ArrayTrack revealed an activation of the p53 pathway upon irradiation of TPC-1 cells with 0.5 and 4 Gy. Two well-known genes linked to the p53 pathway, CDKN1A and MDM2, were found to be up-regulated at 0.5 and 4

113

Low dose irradiation reveals distinct signature

Gy. In addition, genes involved in G1/S and G2/M checkpoint activation were also up- regulated at abovementioned doses. Furthermore, genes involved in mitotic spindle formation such as aurora A kinase (AURKA) were down-regulated at a dose of 4 Gy. Cyclin G1

(CCNG1) and GDF15 which were found to be involved in the p53-mediated pathway were both up-regulated at 0.5 and 4 Gy. Signs of DNA repair were detected at those doses with base excision repair (BER) and ataxia telangiectasia mutated (ATM) pathways up-regulated.

All these pathways were not found to be differentially expressed at the low irradiation dose of

62.5 mGy. Indeed, the number of genes significantly regulated at this dose was very low (18 genes) compared to 0.5 and 4 Gy and pathways reported as most significant were SOCS and vitamin D3 signaling. Notably, the apoptotic pathway was not one of the regulated pathways in all irradiation doses tested. It is worth mentioning that the gene SENP8 was up-regulated at all doses of radiation as well as the gene coding for the ret finger protein-like 1 (RFPL1)

(Tables provided Annex 1).

4.4 Microarray Analysis of Mouse Thyroids To confirm results obtained in vitro, we repeated the irradiation experiment in vivo on murine thyroids bearing the RET/PTC translocation and compared the results to those obtained with wild-type (wt) thyroids. We revealed, using PCA unsupervised clustering, that irradiation of

RET/PTC-positive and –negative thyroids results in distinct genomic profiles (fig. 30). In addition, irradiation at low doses (62.5 mGy) elicited a distinct genomic profile as compared to higher doses (0.5 and 4 Gy).

The main pathway up-regulated at 4 Gy in RET/PTC-positive thyroids was the p53 pathway and cell cycle checkpoint with the two genes CDKN1A and MDM2 highly significantly up- regulated. Again cyclin G1 and GDF15, both connected to the p53 pathway, were up- regulated at the high radiation doses. In addition, Aurora Kinase A interacting protein 1 was 114

Chapter 3

down-regulated at a dose of 4 Gy. At a low dose of 62.5 mGy, the most significant pathway was the granzyme-mediated pathway and the pathway responsible for Ck1/CDK5 production via type 1 glutamate receptors (table 4). On the other hand, the TGF-β1 pathway was found to be consistently differentially expressed over all irradiation doses tested in wild-type thyroids.

Both RET/PTC-positive and –negative thyroids signalled through CDKN1A with up- regulation of its levels at 4 Gy. In wild-type thyroids, a low dose of 62.5 mGy was shown to up-regulate pathways involved in proliferation such as the cyclin dependent kinases (CDK)

11, 9, 5, 10, 7, and 8, EGFR signaling, and FGF pathways. For a list of the pathways involved in radiation response refer to table 4. In addition, thyroid differentiation markers and the thyroid hormone synthesis pathway were up-regulated at all irradiation doses such as thyroglobulin (Tg), thyroid peroxidase (TPO), and thyroid stimulating hormone receptor

(TSHR). Interestingly, the sodium monocarbon transporter (SMCT)-1, otherwise known as slc5A8 was up-regulated at all irradiation doses in wild-type thyroids and not in RET/PTC- positive thyroids. This protein has been linked to tumor suppressive functions [244, 325, 326].

The apoptotic pathway only appeared at a dose of 4 Gy in RET/PTC-positive thyroids but at all irradiation doses in wild-type thyroids. For a list of genes common at 4 Gy of X-rays between TPC-1 cells and RET/PTC-positive thyroids refer to table III. Figure 31 summarizes genes differentially regulated upon irradiation of RET/PTC-positive and –negative thyroids and table 6 some genes that are commonly regulated between RET/PTC-positive and – negative thyroids.

115

Low dose irradiation reveals distinct signature

Figure 30: PCA unsupervised clustering of the gene expression of RET/PTC-positive (A) and –negative (W) thyroids exposed to X-ray doses of 0.0625, 0.5, and 4 Gy. Clustering according to dose received (PC2 = 20%) and to genetic background (PC1 = 66%) can be observed. Number following letters A and W refers to the dose received: 00625 = 0.0625 Gy, 05 = 0.5 Gy, and 4 = 4 Gy.

116

Chapter 3

Figure 31: Hierarchical clustering of genes regulated in RET/PTC-positive (A) and –negative (W) thyroids upon irradiation with X-rays of 0.0625 Gy, 0.5 Gy, and 4 Gy (p<0.001). Green refers to down-regulated genes, red to up-regulated genes and black to no regulation. 4.5 Pathway Analysis using Multiplex Bead Assay To confirm some of the pathways obtained in our microarray data, we performed a multiplex bead assay on protein lysates from TPC-1 and RET/PTC-positive and wild-type murine thyroids on five proteins representative of pathways important in cell growth and proliferation. The advantage of using this technology is that it is more quantitative than a

Western blot and that it allows analysis of a fairly large number of samples for many analytes in one shot. First of all, and as expected, the levels of all total and phosphorylated proteins in the controls were always higher in TPC-1 cells than in RET/PTC-positive thyroids and in both were higher than in wild-type thyroids. Our results indicated no change in the total levels of

ERK1/2 in both TPC-1 and RET/PTC-positive thyroids (fig. 32A). However, there was a 117

Low dose irradiation reveals distinct signature

decrease in the levels of phosphorylated ERK1/2 at all doses in TPC-1 cells but not in

RET/PTC-positive thyroids (fig. 32B). On the other hand, there was no change in the levels of total JNK in TPC-1 cells but it was down-regulated at all doses in RET/PTC-positive thyroids while the STAT3 levels remained constant in response to radiation in both RET/PTC-positive systems. Phosphorylation of JNK on Thr183/Tyr185 also decreased in response to radiation in

TPC-1 cells but not in RET/PTC-positive thyroids (fig. 32E-H). For wild-type thyroids there was a significant increase in the levels of ERK1/2 at a low dose of 62.5 mGy as well as an increase in the levels of total JNK at the same dose. An increase in the phosphorylation of

STAT3 on Tyr705 was observed at this dose (fig. 32A-H). Interestingly, an increase in the levels of total p53 was observed at 1 and 4 Gy in TPC-1 cells. The phosphorylation level of p53 on Ser15 was low in this cell line in general but increased slightly at 1 and 4 Gy (fig.

32B-C). To determine whether there was a correlation between the presence of the RET/PTC translocation and the cellular response to radiation-induced damage, we performed a non- parametric Spearman's Rho test. A high correlation (R=1.00; p < 0.001) was found between the presence of RET/PTC and the response to radiation. Since the antibodies used for the detection of total and phosphorylated p53 in this assay were only specific to human cells, we had to use Western blotting to detect changes in p53 signaling in the murine samples.

118

Chapter 3

119

Low dose irradiation reveals distinct signature

Figure 32: Dot plots representing the mean fluorescence intensity (MFI) for 8 total and phosphorylated proteins in response to X-irradiation in TPC-1, RET/PTC-positive murine thyroids and wild-type murine thyroids. Each point on the graph represents the average of two MFI measurements while the error bars represent the standard error of the means (SEM). HeLa unstimulated: positive control for total ERK1/2 and STAT3. EGF-treated A431: positive control for total and phosphorylated ERK1/2, total p53, total and phosphorylated JNK, and phosphorylated STAT3. A549 Camptothecin: camptothecin-treated A549 cells, positive control for phosphorylated p53 and negative control for total STAT3. Jurkat: negative control for total p53. HeLa Lambda: lambda phosphatase-treated HeLa cells, negative control for phosphorylated proteins. Table shows significantly regulated proteins at various irradiation doses in the three systems. NS = not significant.

4.6 Detection of Protein Phosphorylation in Murine Thyroids by Western Blotting Analysis of the changes in p53 and SMAD3 levels upon irradiation of RET/PTC-positive and wild-type murine thyroids revealed an increase in the levels of p53 in RET/PTC-positive but not wild-type thyroids. On the other hand, SMAD3 was up-regulated in wild-type but not

RET/PTC-positive thyroids. Phosphorylation of p53 on serine 15 was observed at a dose of 4

Gy in RET/PTC-positive and not in wild-type thyroids (fig. 33&34).

Figure 33: Western blot of total and ser15-phosphorylated p53 in RET/PTC-positive and wild-type murine thyroids. Up- regulation of p53 is observed at 4 Gy in RET/PTC-positive but not wild-type mice. An increase in p53 phosphorylation on serine 15, a sign of activation by ATM, is also observed at 4 Gy in RET/PTC-positive but not wild-type murine thyroids.

120

Chapter 3

Figure 34: Western blot of total and phosphorylated SMAD3 in RET/PTC-positive and wild-type murine thyroids. Up- regulation of SMAD3 is observed in wild-type but not in RET/PTC-positive murine thyroids. An increase in SMAD3 phosphorylation is observed at 4 Gy in wild-type thyroids. 4.7 MicroRNA (miRNA) regulation upon X-irradiation The effect of radiation could potentially have an effect not only at the genetic but also at the epigenetic level. Therefore, we attempted to check the response of various miRNA to radiation as a first step towards investigating the epigenetic effect of radiation. We used low density arrays of 384-well plates with qRT-PCR primers for 380 miRNAs to check the levels of those miRNAs. The results revealed both a radiation-specific and a dose-specific response of miRNAs (fig. 35; table 7). The miRNAs displayed in the venn diagram are those that were statistically significantly regulated (p<0.05). Those at the intersection could be considered radiation responsive in that their regulation was apparent at all radiation doses.The number of miRNAs regulated showed a positive correlation with dose received (R=1.00 using

Spearman's Rho test; p<0.01). We then searched for predicted targets of those miRNAs using

TargetScan (http://www.targetscan.org/) [317] and took the most significant ones (top 50 targets) and performed a search of these targets in our microarray data for the TPC-1 cells.

The miRNA and regulated gene list along with correlation between miRNA and gene regulation is summarized in table 8. It is perhaps worth mentioning that many of the p53 targets and genes involved in cell cycle were among the targets of the miRNAs found to be

121

Low dose irradiation reveals distinct signature

regulated upon irradiation of TPC-1 cells. Most notably, CCNG1 and CDKN1A were predicted targets of let-7g and miR-27b respectively.

Figure 35: Venn diagram of miRNAs regulated at three X-ray doses (0.0625 Gy [low], 0.5 Gy [medium], and 4 Gy [high]) in TPC-1 cells. All miRNAs are significant at p<0.05. MiRNAs at the intersection of the three circles can be considered irradiation-responsive while miRNAs that are located in one circle only can be considered dose-responsive. 4.8 Western blots on Protein Levels To confirm the results of the microarray analysis and to investigate whether the observed changes at the transcription level are translated to the translational level, we picked a group of genes (see table 5) that were common between RET/PTC-positive thyroids and TPC-1 cells and performed a Western blot on both systems in addition to wild-type thyroids. Results reveal an up-regulation of growth differentiation factor (GDF)-5, GDF15, and cyclin G and a down-regulation of aurora kinase A (AURKA) in RET/PTC-positive thyroids upon irradiation

– all of these observations corresponding with the microarray analysis. There was no change in the levels of all but GDF15 which was up-regulated at 0.5 and 4 Gy of X-rays in wild-type thyroids (fig. 36). On the other hand, in TPC-1 cells, up-regulation of GDF5, GDF15, and 122

Chapter 3

cyclin G1 was observed upon irradiation while AURKA down-regulation was observed at doses above 0.25 Gy. In addition, up-regulation of RFPL1 was observed at all irradiation doses while Mdm-2 was up-regulated at doses of 0.5 Gy and above (fig. 37).

Figure 36: Western blots of four selected proteins in RET/PTC-positive and –negative murine thyroids upon irradiation. Up-regulation of cyclin G1 and GDF5 can be seen in RET/PTC-positive thyroids and not RET/PTC-negative thyroids upon irradiation. GDF15 was up-regulated in both systems upon irradiation whereas Aurora Kinase A was down-regulated in RET/PTC-positive thyroids only. Although there is no apparent band at the control condition for cyclin G in RET/PTC- positive thyroids, the band could be seen on the film, although it was very faint and did not appear when digitally scanned.

123

Low dose irradiation reveals distinct signature

Figure 37: A)Western blots of four selected genes in TPC-1 cells upon irradiation. Up-regulation of cyclin G1, GDF5 and GDF15 is observed upon irradiation with X-rays while Aurora Kinase A is greatly down-regulated upon irradiation with doses of and above 0.5 Gy.

124

Chapter 3

Figure 37: B) Protein levels of Mdm-2 were up-regulated at and above X-ray doses of 0.5 Gy while RFPL-1 (upper band) levels increased at all irradiation doses. Lower band probably represents the other RFPL isoform, RFPL-3 which was also detected by the antibody. 4.9 Pax-8 mRNA Level After X-irradiation Radiation doses of 0.5 Gy and higher were found to cause the onset of senescence in TPC-1 cells [327] and increase in differentiation markers in wild-type thyroids. In addition, the gene ret finger protein-like (RFPL)-1 appeared to be up-regulated at all irradiation doses in TPC-1 cells upon irradiation in microarray data. This gene is not described extensively in the literature. However, a paper hypothesized that this protein may be involved in differentiation

[328]. Another paper claimed that RFPL-1 promoter possesses Pax-6 binding sites in its promoter [329]. This led us to examine the levels of Pax-8 in response to X-irradiation as it is the only differentiation marker in TPC-1 cells [298]. Levels of Pax-8 mRNA as checked by qRT-PCR did not change upon irradiation (fig. 38).

125

Low dose irradiation reveals distinct signature

Figure 38: Transcriptional levels of paired homology box 8 (Pax-8) by qRT-PCR. Levels of Pax-8 were normalized to the levels of α-tubulin. There was no change in mRNA levels of Pax-8 upon irradiation of TPC-1 cells with X-ray doses of 0.0625, 0.5, and 4 Gy.

126

Chapter 3

5. Discussion We have shown in this paper that the RET/PTC translocation alters the response of thyroid cells to low and high doses of X-irradiation at the transcriptional and epigenetic levels. In addition, the response of normal and RET/PTC-bearing thyroid cells to low doses of radiation is distinct from that of high doses.

TPC-1 is a cell line derived from cells of PTC taken from a female patient. TPC-1 cells all possess the RET/PTC1 translocation and a wild-type copy of p53 [298]. In addition, cytogenic studies done on this cell line revealed that it had several genetic imbalances and the deletion of chromosome 21 [330]. Genomic instability is a feature of PTC and cancers in general in addition to being a feature of cell lines [331-336]. The RET/PTC-positive mice used in this study were transgenic C57BL/6J mice with a RET/PTC3 translocation under the control of a bovine Tg promoter. According to a study done on the same strain of mice, they do not show any sign of PTC at three months of age, the age used in this study [324]. Therefore, we can assume that those thyroids represent the initiation step in PTC. The use of cells from two different species was forced upon us since we lacked any clinical samples. However, this means that our conclusions are not limited to one species and apply in vitro and in vivo to mouse and human samples.

The Lee and Whitmore test results indicated that a gene fold-change larger than 1.5 and smaller than 0.66 can be detected with 88% statistical power. Any change between 1.5 and

0.66 severely reduces the power of the test. However, we decided to include any fold change in statistically significant genes (FDR corrected p-value < 0.005 for TPC-1 and 0.05 for murine thyroids) as we didn't expect to see large fold-changes at the low dose of radiation and thus we were interested in regulated genes regardless of the fold change. Indeed, in

RET/PTC-positive cells (TPC-1 and RET/PTC-positive thyroids) only a limited number of genes would survive these criteria. Among those genes in TPC-1 is RFPL-1, a gene was also

127

Low dose irradiation reveals distinct signature

found in a proteomic analysis of human gastric carcinoma cells treated with hexamethylene bisacetamide and was postulated to be involved in differentiation of said cells and which was confirmed in our results by Western blotting [328]. In addition, after applying an FDR p-value correction to the pathway analysis results, no pathways appeared significantly regulated in response to a low dose of X-rays in RET/PTC-positive cells. This is in stark contrast to the effect of a low dose of radiation on normal murine thyroids where pathways involved in cell growth and proliferation were significantly regulated. These results were confirmed using a multiplex bead assay where the up-regulation of ERK1/2, JNK and phosphorylated JNK on

Thr183/Tyr185 confirmed the up-regulation of the EGF signaling pathway at 62.5 mGy.

Incredibly, these results were only found at a dose of 62.5 mGy and not higher in wild-type mice. The link between low dose radiation and appearance of thyroid cancer has been indicated before in the literature and our results may point to the fact that the process of hyperplasia may start as early as 6 hours after irradiation and may be specific to low doses of

X-rays. Naturally, further studies with other time points and long-term follow up of the mice need to be performed to confirm these findings [18, 96, 337].

At the higher doses of 0.5 and 4 Gy, the p53 pathway was one of the significantly regulated pathways in TPC-1 cells. This confirms our early results which revealed an increase in p53 by

Western blotting 24 hours after irradiation and was confirmed by multiplex bead assay and

Western blots of the p53 inhibitor, mdm-2 [314]. The p53 pathway was regulated at a high dose of 4 Gy in RET/PTC-positive mice. However, when FDR correction was applied to the pathway analysis, the p53 pathway was no long significant. Nonetheless, we detect an increase in the levels of p53 in RET/PTC-positive thyroids 24 hours after irradiation with 4

Gy. A possible reason why the pathway analysis in RET/PTC-positive thyroids was not as robust as that in TPC-1 cells may lie in the fact that TPC-1 cells are a more homogenous system where the RET/PTC translocation is expressed in all cells whereas the RET/PTC- 128

Chapter 3

positive thyroids are a more heterogeneous system where the RET/PTC translocation is not expressed in every follicular cell [338].

We have also uncovered a subset of genes that are all regulated in both RET/PTC-positive cells and that all are connected to p53 activation. For example, aurora kinase A is a protein involved in mitotic spindle formation and organization and is reported as an oncogene commonly dysregulated in cancers [339]. Aurora kinase A's inhibition is p53-dependent and to our knowledge, this is the first report of aurora kinase A as a radiation-responsive gene.

Aurora Kinase A and its inhibitory partner Aurora Kinase A interacting protein (AURKAIP1) were down-regulated in TPC-1 cells and RET/PTC-positive thyroids respectively. The down- regulation of AURKAIP1 should have increased the levels of aurora A in this system but a

Western blot done on the protein levels of RET/PTC-positive thyroids revealed a down- regulation of aurora A similar to that seen in TPC-1 cells [340, 341]. Perhaps this could indicate a feedback loop in which down-regulation of aurora A causes the down-regulation of its inhibitory partner. GDF15 and GDF5 are extended members of the TGF-β superfamily

[342, 343] . GDF15 was found to be down-regulated in RET/PTC-positive thyroids. Its effect has been disputed in the literature as some studies have linked it to a proliferative effect while others have attributed its effects to a decrease in proliferation and onset of differentiation

[344-347]. Secretion of GDF15 has also been proposed as a p53 pathway biomarker, similar to p21 [348]. The down-regulation of GDF15 in RET/PTC-positive thyroids and up- regulation upon irradiation indicates that the effect of GDF5 and GDF15 is anti-proliferative.

Cyclin G1 is another gene whose up-regulation was confirmed in both of our RET/PTC- positive systems using microarrays and Western blots. Cyclin G1 is the only cyclin not to have a CDK partner and is directly involved in p53 regulation through inhibition of mouse double minute 2 (Mdm-2) [349, 350]. This gene was proposed as a biomarker of radiation as it was found to be up-regulated in response to radiation in peripheral blood mononuclear cells 129

Low dose irradiation reveals distinct signature

(PBMCs) [312]. However, our data reveal that cyclin G1 is not a radiation-responsive gene in normal murine thyroids neither in the transcriptional nor the translational sense.

This could be due to the fact that the p53-responsive pathway does not seem to be activated in thyroids from wild-type mice in response to irradiation as opposed to the TGF-β1 pathway which was activated at all irradiation doses. Our data would indicate that RET/PTC-positive and –negative thyroids respond to irradiation in two distinct manners: RET/PTC-positive thyrocytes hit with doses of 0.5 Gy and higher activate the p53 pathway while wild-type thyrocytes activate the TGF-β1 pathway and both converge on p21Cip1 (see fig. 39).

Figure 39: Cell Cycle Pathway retrieved from KEGG database and modified to represent genes regulated in RET/PTC- positive thyroids (red), RET/PTC-negative thyroids (blue), and TPC-1 cells (green). Involvement of the TGF-β pathway in RET/PTC-negative thyroids can be observed upon irradiation while p53 pathway is activated in RET/PTC-positive thyroids and TPC-1 cells. p21 was involved in the response of all three systems. Arrows represent regulation of gene. Color of arrow indicates to which system it belongs. miRNAs have been added in red to scheme alongside their predicted targets. Merger of colors represents genes that were involved in multiple system response to radiation. 130

Chapter 3

Incidentally, our previous data show that TPC-1 cells irradiated with doses of 0.5 Gy and higher have significantly higher amounts of TGF-β1 in their milieu as compared to non- irradiated cells [314]. This indicates that TGF-β1 pathway is indeed activated in RET/PTC- positive thyrocytes by token of the increased secretion of this cytokine. However, the lack of signaling downstream of the TGF receptor might be explained by the recent evidence that showed a reduced sensitivity to TGF-β1 in RET/PTC-positive cells due to a down-regulation of Smads 3/4 [351]. Another noteworthy difference between the RET/PTC-positive and – negative thyroids was that apoptosis-related pathways were elicited upon irradiation at all doses of radiation in wild-type thyroids but only at 4 Gy in RET/PTC-positive thyroids. This could point to the fact that the RET/PTC translocation confers a protective advantage against cell death probably via induction of a senescent-like response.

Our previous results showed that TPC-1 responds to radiation-induced DNA damage at doses of 0.5 and higher with a senescent-like response. Incidentally, the inhibition of aurora A kinase was associated with senescence induction in tumors in vitro and in vivo [213, 214,

314]. Therefore, we wondered whether the transcriptome signature obtained by us was shared by other senescent cells. We discovered an overlap of 20% common genes between our list of genes at 4 Gy and the genes obtained by the group of Safrany in diploid fibroblasts upon

DNA damage [211].

MicroRNAs (miRNAs) are non-coding 22-nucleotide stretches of RNA that bind with incomplete complementarity to mRNAs and either degrade the mRNA or inhibits its translation [352]. MicroRNAs have been described as up-regulated in cancers as well as responding to cellular stress and radiation-induced reactive oxygen species (ROS) [273, 279,

292]. Interestingly, Wang et al. (2011) described a list of miRNAs that could be responsible for regulating both replicative and radiation-induced senescence and we have significant overlap between our list of miRNAs and Wang et al [353] which could indicate that miRNAs 131

Low dose irradiation reveals distinct signature

are involved in the senescent-like response observed in TPC-1 cells upon irradiation. The let-

7 family of miRNAs are the first to be discovered in cells and the miRNA let-7g has been described as induced by a dose of 2 Gy in human endothelial cells [354]. However, in TPC-1 cells, let-7g was found to be down-regulated in response to 4 Gy of X-rays. Another study has demonstrated that let-7g inhibits the proliferation of hepatocellular carcinoma by up- regulation of p16 [355]. Our studies however have shown that irradiation of TPC-1 cells causes an up-regulation of p16 [327]. Indeed, we have found that there is often a positive correlation between the miRNA and gene regulation in response to irradiation in TPC-1 cells.

This probably could be explained by recent research that miRNAs can also up-regulate translation of mRNA and that this phenomenon is cell cycle-dependent [290, 291, 356]. When searching for predicted targets of the differentially regulated miRNAs in TPC-1 cells upon irradiation in our list of differentially regulated genes, we found significant overlap but only at a dose of 4 Gy. Around 16% of the genes differentially regulated at 4 Gy were predicted targets of the miRNAs regulated at the same dose which indicates that regulation of these genes upon irradiation is partially due to miRNA regulation. It is also significant that most of the genes involved in the miRNA response are involved in the p53 pathway and cell cycle regulation which points to the fact that the main pathway regulated by irradiation at 4 Gy in

TPC-1 cells is both genetically and epigenetically modulated. The fact that there was a common signature of miRNAs that responded to radiation indicates that these miRNAs are probably involved in DNA repair or in the cellular stress response. However, we found only little overlap with the results of Simone et al. who tested the changes in miRNA levels after

10 Gy, H2O2, and etoposide in human fibroblasts. This could be related to the fact that the authors use a higher dose of radiation and use a time-point of one hour [279]. The change in a single parameter such as time-point or hydrogen peroxide concentration, for example, was found to have a significant impact on the miRNA profile as evidenced from the study of Li et 132

Chapter 3

al. on H2O2-treated fibroblasts. The authors use a different concentration of hydrogen peroxide and different time-points and come up with an almost different miRNA profile than the study of Simone et al. on the same cells [279, 357]. Most importantly is the fact that the authors' study, and in fact most of the studies, are performed on normal cells while our study was conducted on transformed cells. It is also possible that the radiation response is not universal across cell types. Unfortunately, only a few of the miRNAs that came up in our data are described in the literature. One of the important links between miRNA and gene regulation was confirmed by Li et al. on miR-106a and p21, a link that is hypothesized in our data [357]. The fact that we have observed not one but two miRNAs associated with p21 regulation illustrates the importance of miRNAs in the regulation of cell cycle arrest and possibly the induction of senescence in response to radiation. Although some of the predicted targets of the miRNAs in our data are important in thyroid cancer, more studies are needed to confirm the link between miRNA levels and gene expression. In addition, miR-17 is a member of the same family of miRNAs as miR-106a so could be theoretically involved in p21 regulation. However, the fact that we don't observe p21 regulation at the protein level at this low dose highlights how complex the relationship between miRNAs and transcription is.

Finally, despite the presence of similarities between the two RET/PTC-positive systems‟ response to moderate to high doses of radiation, the systems did not behave exactly the same.

For example, TPC-1 cells responded to radiation by down-regulating phosphorylated ERK1/2 but not JNK while RET/PTC thryoids responded in the opposite manner. This could be due to the fact that the response radiation is system-specific. Also, the effective dose delivered to the

TPC-1 cells and to the RET/PTC-positive thyroids was different. Whole-body irradiation of the mice means that the effective dose delivered to the thyroids was less than the dose delivered to the TPC-1 cells in tissue culture flasks. In addition, although the RET/PTC translocations both signal along the same axes, the morphology of RET/PTC3-bearing 133

Low dose irradiation reveals distinct signature

thyroids has been described as distinct in the literature and perhaps this also contributed to the difference in response [94]. Furthermore, the difference in response of TPC-1 cells and

RET/PTC-bearing thyroids could also be put down to the fact that at the age of the mice used in this study, a small percentage of the thyroids maintained proliferative changes and tumor characteristics [338] as opposed to TPC-1 cells which have PTC characteristics.

We have shown that the response to low and high doses of X-rays is different in two systems of RET/PTC-positive thyrocytes as compared to wild-type thyroids. This has also been found by Ding et al. (2005) where the p53 pathway was up-regulated in fibroblasts at a dose of 2 Gy and not at a dose of 2 cGy [358]. A low dose of radiation caused the up-regulation of cellular growth and proliferation pathways in normal thyroids but not in RET/PTC-positive thyrocytes. In these thyrocytes, radiation decreased signaling down cellular growth and proliferation pathway with p53 activation at moderate to high doses.

Acknowledgements

The authors are thankful to Dr. M.A. Benotmane and Ms. A. Janssen for their help with the microarray experiments, Ms. A. Michaux for her work on the mouse genotyping, and Mr. K.

Tabury for his help with the multiplex bead assay. K. Abou-El-Ardat is supported by a doctoral SCK•CEN/Ghent University grant. This work was funded by the EU Euratom

Program (GENRISK-T contract FIP6-2006-036495 on "defining the genetic component of thyroid cancer risk at low doses" as well as the DoReMi Network of Excellence on "Low

Dose Research towards Multidisciplinary Integration", agreement n° 249689).

134

Chapter 3

Gene Name Forward (f) Reverse (r)

RET/PTC3 Tg: 5' GGC CAG AGC CCT AAG GTG GGC ELE1: 5' AAG GGA TTC AAT TGC CAT

3' CCA 3'

Pax-8 5' CAA CAG CAC CCT GGA CGA C 3' 5' AGG GTG AGT GAG GAT CTG CC 3'

α-tubulin 5' GCC TAC CAT GAA CAG CTT TC 3' 5' ACG TCG TAC AGT ACG AGG GTC T

Table 3: Primer sequences used for mouse genotyping

135

Low dose irradiation reveals distinct signature

Pathway Fisher P-value Corrected p-value

SOCS mediated pathway 0.012 0.625

Vitamin D3 Signaling Pathway 0.05 1

Low Dose (62.5 mGy) 18 genes

Pathway Fisher P-value Corrected p-value

Notch Signaling Pathway 0.0004 0.0150 p53 Mediated Pathway 0.0006 0.0150

PTEN Signaling Pathway 0.0032 0.0533

G2-M Checkpoint Pathway 0.0050 0.0533

BER Pathway 0.0061 0.0533

ATR-ATM Signaling Pathway 0.0093 0.0775

G1-S Checkpoint Pathway 0.0138 0.98

Medium Dose (0.5 Gy) 21 genes

Pathway Fisher P-value Corrected p-value

Spindle Checkpoint Pathway 0.000001 0.000089 p21 Mediated Pathway 0.000001 0.000089 p53 Mediated Pathway 0.000001 0.000089

Cyclins Mediated Pathway 0.000006 0.00032

G1-S Checkpoint Pathway 0.000006 0.00032

TRAIL Signaling Pathway 0.000386 0.017

ATR-ATM Signaling Pathway 0.000789 0.030

High Dose (4 Gy) 123 genes

136

Chapter 3

A) TPC-1

137

Low dose irradiation reveals distinct signature

Pathway Fisher P-value Corrected p-value

Granzyme Mediated Pathway 0.017 0.625

Regulation of Ck1/CDK5 by Type 1 Glutamate 0.025 0.625

Receptors

Low Dose (62.5 mGy) 59 genes

Pathway Fisher P-value Corrected p-value

CCR2 Mediated Pathway 0.0074 0.0185

Regulation of Ck1/CDK5 by Type 1 Gluatamate Receptors 0.0074 0.0185

CCL2 Signaling Pathway 0.0155 0.258

Rb Signaling Pathway 0.0452 0.500

G1-S Checkpoint Pathway 0.050 0.500

Medium Dose (0.5 Gy) 432 genes

Pathway Fisher P-value Corrected p-value

CDK5 Mediated Pathway 0.0010 0.05 p53 Mediated Pathway 0.0047 0.08

G1-S Checkpoint Pathway 0.0048 0.08

ATR-ATM Signaling Pathway 0.0075 0.09375

TRAIL Mediated Apoptosis 0.0120 0.12

High Dose (4 Gy) 455 genes

B) RET/PTC-positive thyroids

138

Chapter 3

Pathway Fisher P-value Corrected p-value

AGEs Signaling Pathway 0.00037 0.0185

PRL Signaling Pathway 0.00141 0.0352

CDK5 Mediated Pathway 0.00239 0.0398

EGF Signaling Pathway 0.00454 0.0481

IFNg Signaling Pathway 0.00799 0.0678

FGF Signling Pathway 0.00814 0.0678

TGF Beta Induced Apoptosis 0.01971 0.141

Low Dose (62.5 mGy) 2311 genes

Pathway Fisher P-value Corrected p-value

FGF Signaling Pathway 0.01667 0.736

PDGF Signaling Pathway 0.03533 0.736

TGF Beta Induced Apoptosis 0.04420 0.736

Medium Dose (0.5 Gy) 556 genes

Pathway Fisher P-value Corrected p-value

TGF Beta Induced Apoptosis 0.00088 0.0440

Bcr-Abl Signaling Pathway 0.01045 0.261

ATR-ATM Signaling Pathway 0.02283 0.380

G1-S Checkpoint Pathway 0.05191 0.638

High Dose (4 Gy) 629 genes

C) RET/PTC-negative thyroids

139

Low dose irradiation reveals distinct signature

Pathway Fisher P-value Corrected p-value

CDK5 Mediated Pathway (Cell 0.000510 0.0255 cycle)

FGF Signaling Pathway 0.001867 0.0366

Ras Signaling Pathway 0.002212 0.0376

IL2 Signaling Pathway 0.002930 0.0410

EGF Signaling Pathway 0.005812 0.049

D) RET/PTC-positive thyroids

Table 4: Pathways responsive to radiation at low (0.0625 Gy), medium (0.5 Gy), and high (4 Gy) doses of X-rays in A) TPC-1 cells B) RET/PTC-positive thyroids and C) RET/PTC-negative thyroids. Pathways responsive in RET/PTC-positive thyroids when compared to wild-type thyroids is shown in D.

140

Chapter 3

RET/PTC-positive Thyroids TPC-1 p21 (1.72) p21 (2.81)

ABCA16 (1.26) ABCA12 (2.62)

GDF15 (1.67) GDF15 (2.60)

AURKAIP1 (0.78) AURKA (0.40)

Mdm-2 (1.68) Mdm-2 (1.92)

DAGKβ (0.65)* DAGKα (1.73)*

PSRC1 (1.31)* PSRC1 (0.52)*

REEP3 (0.73) REEP4 (0.63)

Actin α2 (0.64)* Actin α2 (1.77)*

Cyclin G1 (1.88) Cyclin G1 (1.57)

UBE2E (0.75) UBE2C (0.59)

SYCP1 (1.37) SYCP1 (1.30)

MRPL22 (0.78) MRPL22 (0.71)

Plexin C1 (1.25) Plexin B2 (1.59)

PLC δ4 (1.54)* PLC δ4 (0.79)*

Table 5: Genes commonly regulated upon irradiation in RET/PTC-positive thyroids and TPC-1 cells. Paralogs and interacting partners were also considered. Numbers in parenthesis represent fold- change where numbers above 1 denote up-regulation while those below 1 denote down-regulation. Asterisks refer to genes that are differentially regulated in both systems. This subset of genes constitutes 12.2% of genes in TPC-1 cells at 4 Gy.

141

Low dose irradiation reveals distinct signature

RET/PTC-negative RET/PTC-positive

GDF15 (2.35) GDF15 (1.67)

Bcl-X (1.97) Bcl-X (1.40)

Bax (1.94) Bax (1.56)

Profilin 2 (0.73) Profilin 2 (0.56) p21 (2.00) p21 (1.72)

Synaptotagmin XII (1.78) Synaptotagmin XII (1.80)

Dio1 (2.13)* Dio1 (0.64)*

PDGFα (1.39) PDGFα (1.30)

Cytochrome c oxidase Vib (1.41) Cytochrome c oxidase Vib (1.68)

ApoE (1.51)* ApoE (0.75)*

Tektin 2 (1.37) Tektin 2 (1.44)

Synaptotagmin-like 1 (1.31) Synaptotagmin-like 1 (1.37)

WDR55 (0.75) WDR55 (0.78)

GAS6 (1.46) GAS6 (1.37)

PLC δ4 (1.30) PLC δ4 (1.54)

CD53 (0.65) CD53 (0.60)

AEN (1.18) AEN (1.20)

142

Chapter 3

PHD finger Protein 20-like 1 (1.30)* PHD finger Protein 20-like 1 (0.67)*

Table 6: Some genes commonly regulated in RET/PTC-positive and –negative thyroids. Numbers in parenthesis refer to fold-change and asterisks were differentially regulated in both systems.

143

Low dose irradiation reveals distinct signature

Micro RNA Predicted Targets Fold Change Dose

Let-7g HMGA2 0.56 High miR-10a CDK6, PPARα 0.31; 0.37; 0.38 All miR-224 SMAD4 3.21; 2.40; 2.37 All miR-296-5p HMGA1 1.07; 1.16; 0.51 All miR-625 IGF-1 0.75; 1.17; 0.49 All miR-30b CCNE2, IL-2Rα, WDR44 0.58 High miR-106a p21 0.93 High miR-152 PTEN, GADD45a, WNT1 0.63; 0.66; 0.74 All miR-183 KIF-2α 2.50 High miR-423-5p Caspase 2 0.35 High miR-454 IGF-1, TNF, p63 0.70 High miR-491 TP53INP1, CDK6 0.32; 0.44 Medium/High miR-376c WDR44 0.24; 0.27 Medium/High miR-27b PLK2, CDK6, XIAP 2.45; 2.12 Medium/High miR-185 VEGFα, CDC42 0.95 Medium miR-17 PPARα, E2F5, HIF-1α 1.39 Low

Table 7: MiRNAs regulated upon irradiation in TPC-1 cells with their predicted targets and fold- change. Numbers are ranked from low dose to high.

144

Chapter 3

Gene Gene Name miRNA Correlation Dose

WDR63 WD Repeat Domain 63 miR-454 Positive High

CDKN1A p21 Let-7g Positive High

miR-106a

SESN1 Sestrin 1 miR-183 Negative High

GADD45a Growth Arrest and DNA miR-152 Positive High

Damage Inducible

TP53INP1 Tumor Protein p53 Inducible miR-30b Positive High

Nuclear Protein 1 miR-106a

miR-454

BTG2 BTG Family, member 2 Let-7g Positive High

miR-27b

ACTA2 Actin alpha 2 miR-27b Negative High

CDCA8 Cell Division Cycle Associated Let-7g Negative High

RDH10 Retinol Dehydrogenase 10 Let-7g Positive High

CCNF Cyclin F Let-7g Negative High

KPNA2 Karyopherin alpha 2 miR-106a Negative High

145

Low dose irradiation reveals distinct signature

miR-376c

C1orf96 Chromosome 1 open reading Let-7g Negative High

frame 96

LBR Lamin B Receptor Let-7g Positive High

miR-152 Positive

miR-224 Negative

TMPO Thymopoietin miR-183 Positive High

DEPDC1 DEP Domain Containing 1 miR-27b Positive High

RACGAP1 Rac GTPase Activating Protein miR-106a Negative High

1 miR-454

LBA1 Lupus Brain Antigen 1 Let-7g Positive High

CCNG1 Cyclin G1 miR-27b Negative High

KIF23 Kinesin Family miR-224 Positive High

Table 8: miRNAs regulated upon irradiation of TPC-1 cells and their predicted target genes that were found in our microarray data. Positive correlation describes situation when miRNA and gene are regulated in the same way. Low refers to 0.0625 Gy, Medium to 0.5 Gy, and High to 4 Gy.

146

Chapter 3

Symbol Description Chromosome GenBank p-value FC

RFPL1 ret finger protein-like 1 22 NM_021026 0.00015961 1.572 7

VNN2 vanin 2 6 NM_004665 0.00026004 1.390 7

C14orf19 chromosome 14 open reading frame 14 NR_002937 0.00146945 1.575 19

OR1D2 olfactory receptor, family 1, subfamily 17 NM_002548 0.00167875 0.671 D, member 2

OR5K1 olfactory receptor, family 5, subfamily 3 NM_00100473 0.00198784 1.366 K, member 1 6

OR56A3 olfactory receptor, family 56, 11 NM_00100344 0.00211155 1.386 subfamily A, member 3 3

RASSF9 Ras association (RalGDS/AF-6) domain 12 NM_005447 0.00246766 0.772 family (N-terminal) member 9

DOC2B double C2-like domains, beta 17 NM_003585 0.00289874 0.739

OR6M1 olfactory receptor, family 6, subfamily 11 NM_00100532 0.00291205 1.326 M, member 1 5

SENP8 SUMO/sentrin specific peptidase 15 NM_145204 0.00298823 1.286 family member 8

MGC3482 Similar to hypothetical gene supported 17 XM_208993 0.00299924 2.124 9 by AL050367; AK022946

PLCD4 phospholipase C, delta 4 2 NM_032726 0.00310876 0.785

KCNS3 potassium voltage-gated channel, 2 NM_002252 0.00362768 1.276 delayed-rectifier, subfamily S, member 3

SPZ1 spermatogenic leucine zipper 1 5 NM_032567 0.0039138 1.280

C15orf26 chromosome 15 open reading frame 15 AK095934 0.00414851 0.773 26

IFLTD1 intermediate filament tail domain 12 NM_152590 0.00428453 0.783 containing 1

SRD5A2 steroid-5-alpha-reductase, alpha 2 NM_000348 0.00446916 1.298 polypeptide 2 (3-oxo-5 alpha-steroid delta 4-dehydrogenase alpha 2)

147

Low dose irradiation reveals distinct signature

CCDC54 coiled-coil domain containing 54 3 NM_032600 0.00462921 0.780

18 Genes

Symbol Description Chromosome GenBank p-value FC

WDR63 WD repeat domain 63 1 NM_145172 0.00046329 1.433 2

CDKN1A cyclin-dependent kinase inhibitor 6 NM_078467 0.00065356 1.538 1A (p21, Cip1) 4

CYFIP2 cytoplasmic FMR1 interacting 5 NM_001037332 0.00070868 1.551 protein 2 5

LOC653665 similar to mCG4465 10 XM_928758 0.0014228 1.511

LOC121952 hypothetical protein LOC121952 13 XR_015298 0.00163517 0.741

PTCHD3 patched domain containing 3 10 NM_001034842 0.0020296 0.769

FAM48B1 family with sequence similarity 48, X XM_001131038 0.00243098 0.655 member B1

SENP8 SUMO/sentrin specific peptidase 15 NM_145204 0.00271743 1.291 family member 8

C6orf138 chromosome 6 open reading 6 NM_001013732 0.00295919 1.534 frame 138

OR1D2 olfactory receptor, family 1, 17 NM_002548 0.00312674 0.695 subfamily D, member 2

RFPL1 ret finger protein-like 1 22 NM_021026 0.00320118 1.357

EDA2R ectodysplasin A2 receptor X NM_021783 0.0033861 1.452

PLEKHM1L pleckstrin homology domain 2 BX648983 0.00358652 1.265 containing, family M, member 1- like

TMC5 transmembrane channel-like 5 16 NM_024780 0.00361403 0.808

APOC4 apolipoprotein C-IV 19 NM_001646 0.00417588 0.789

PRSS8 protease, serine, 8 16 NM_002773 0.00426004 0.787

CCDC129 coiled-coil domain containing 129 7 NM_194300 0.00437047 1.295

FAM128B family with sequence similarity 2 AK024408 0.00441242 0.774 128, member B

MDM2 Mdm2, transformed 3T3 cell 12 NM_002392 0.004616 1.290

148

Chapter 3

double minute 2, p53 binding protein (mouse)

LOC440570 LOC440570 1 AK125737 0.00479313 1.401

C3orf43 chromosome 3 open reading 3 NM_001077657 0.00491844 1.297 frame 43

CCDC54 coiled-coil domain containing 54 3 NM_032600 0.00495621 0.782

22 Genes

149

Low dose irradiation reveals distinct signature

Symbol Description Chromosome GenBank p-value FC

WDR63 WD repeat domain 63 1 NM_145172 2.27e-07 2.459

CDKN1A cyclin-dependent kinase inhibitor 6 NM_078467 5.60e-07 2.807 1A (p21, Cip1)

SESN1 sestrin 1 6 NM_014454 3.11e-06 2.281

PLK1 polo-like kinase 1 (Drosophila) 16 NM_005030 9.58e-06 0.331

ABCA12 ATP-binding cassette, sub-family A 2 NM_173076 1.24e-05 2.620 (ABC1), member 12

CCNB1 cyclin B1 5 NM_031966 1.38e-05 0.413

KIF20A kinesin family member 20A 5 NM_005733 1.72e-05 0.357

KLRC1 killer cell lectin-like receptor 12 NM_002259 2.08e-05 2.106 subfamily C, member 1

GDF15 growth differentiation factor 15 19 NM_004864 2.68e-05 2.602

CYFIP2 cytoplasmic FMR1 interacting 5 NM_001037332 2.84e-05 2.044 protein 2

AURKA aurora kinase A 20 NM_198433 3.09e-05 0.396

CDC20 cell division cycle 20 homolog (S. 1 NM_001255 3.64e-05 0.512 cerevisiae)

FAM83D family with sequence similarity 83, 20 NM_030919 4.60e-05 0.508 member D

GADD45A growth arrest and DNA-damage- 1 NM_001924 5.63e-05 1.559 inducible, alpha

TP53INP1 tumor protein p53 inducible 8 NM_033285 7.53e-05 3.219 nuclear protein 1

MDM2 Mdm2, transformed 3T3 cell 12 NM_002392 7.90e-05 1.921 double minute 2, p53 binding protein (mouse)

C6orf138 chromosome 6 open reading 6 NM_001013732 8.19e-05 2.132 frame 138

OR5K1 olfactory receptor, family 5, 3 NM_001004736 8.73e-05 1.677 subfamily K, member 1

DGKA diacylglycerol kinase, alpha 80kDa 12 NM_201444 0.00010226 1.733 2

150

Chapter 3

TRIM22 tripartite motif-containing 22 11 NM_006074 0.00011957 2.265 6

TNFSF4 tumor necrosis factor (ligand) 1 NM_003326 0.00012030 2.504 superfamily, member 4 (tax- 8 transcriptionally activated glycoprotein 1, 34kDa)

CCNB2 cyclin B2 15 NM_004701 0.00012274 0.536 4

EDA2R ectodysplasin A2 receptor X NM_021783 0.00015636 1.867 8

TRIM55 tripartite motif-containing 55 8 NM_033058 0.00016849 2.446 4

BTG2 BTG family, member 2 1 NM_006763 0.00018206 2.258 5

APOC4 apolipoprotein C-IV 19 NM_001646 0.00019212 0.671 8

C15orf23 chromosome 15 open reading 15 NM_033286 0.00020124 0.525 frame 23 1

FLJ11827 hypothetical protein FLJ11827 3 AK021889 0.00021477 0.605 3

PSRC1 proline/serine-rich coiled-coil 1 1 NM_001032290 0.00021598 0.522 9

CCNA2 cyclin A2 4 NM_001237 0.00022700 0.462 9

REEP4 receptor accessory protein 4 8 NM_025232 0.00023148 0.625 7

XPC xeroderma pigmentosum, 3 NM_004628 0.00023398 1.704 complementation group C 1

GTSE1 G-2 and S-phase expressed 1 22 NM_016426 0.00023450 0.516 9

RFPL1 ret finger protein-like 1 22 NM_021026 0.00024493 1.617 9

MYL4 myosin, light chain 4, alkali; atrial, 17 NM_001002841 0.00026923 0.584 embryonic 3

PRC1 protein regulator of cytokinesis 1 15 NM_003981 0.00029243 0.623 8

151

Low dose irradiation reveals distinct signature

CENPA centromere protein A 2 NM_001809 0.00029909 0.478 4

ACTA2 actin, alpha 2, smooth muscle, 10 NM_001613 0.00033794 1.769 aorta 1

PSTPIP2 proline-serine-threonine 18 NM_024430 0.00036002 1.576 phosphatase interacting protein 2 4

PLK3 polo-like kinase 3 (Drosophila) 1 NM_004073 0.00036154 1.483 6

IHPK2 inositol hexaphosphate kinase 2 3 NM_001005909 0.00041398 1.656 1

C12orf5 chromosome 12 open reading 12 NM_020375 0.00042076 2.096 frame 5 6

CES2 carboxylesterase 2 (intestine, liver) 16 NM_003869 0.00044730 1.446 3

CDCA8 cell division cycle associated 8 1 NM_018101 0.00049102 0.468

PLEKHM1L pleckstrin homology domain 2 BX648983 0.00056283 1.405 containing, family M, member 1- 7 like

SENP8 SUMO/sentrin specific peptidase 15 NM_145204 0.00056357 1.425 family member 8 9

NDE1 nudE nuclear distribution gene E 16 NM_017668 0.00061620 0.677 homolog 1 (A. nidulans) 7

CES7 carboxylesterase 7 16 NM_145024 0.00064389 0.690 6

TPX2 TPX2, microtubule-associated, 20 NM_012112 0.00065575 0.537 homolog (Xenopus laevis) 3

STOM stomatin 9 NM_004099 0.00068198 1.624 5

OR1D2 olfactory receptor, family 1, 17 NM_002548 0.00069658 0.606 subfamily D, member 2 2

ZMAT3 zinc finger, matrin type 3 3 NM_022470 0.00075726 1.718

RDH10 retinol dehydrogenase 10 (all- 8 NM_172037 0.00076820 1.955 trans) 5

RACGAP1P Rac GTPase activating protein 1 12 AF334184 0.0007841 0.585 pseudogene

152

Chapter 3

CDKN3 cyclin-dependent kinase inhibitor 3 14 NM_005192 0.00086468 0.624 (CDK2-associated dual specificity 1 phosphatase)

HJURP Holliday junction recognition 2 NM_018410 0.00089706 0.524 protein 2

KIF2C kinesin family member 2C 1 NM_006845 0.00094699 0.563 1

FHL2 four and a half LIM domains 2 2 NM_201555 0.00096367 1.440 4

TARSL2 threonyl-tRNA synthetase-like 2 15 NM_152334 0.00096479 1.432 7

OR4A16 olfactory receptor, family 4, 11 NM_001005274 0.00096645 1.346 subfamily A, member 16 7

PROKR1 prokineticin receptor 1 2 NM_138964 0.00109858 0.745

CCNF cyclin F 16 NM_001761 0.00109923 0.504

DDB2 damage-specific DNA binding 11 NM_000107 0.00110097 1.905 protein 2, 48kDa

DAND5 DAN domain family, member 5 19 NM_152654 0.00110272 0.689

BUB1 BUB1 budding uninhibited by 2 NM_004336 0.00110961 0.504 benzimidazoles 1 homolog (yeast)

KPNA2 karyopherin alpha 2 (RAG cohort 1, 17 NM_002266 0.00116966 0.639 importin alpha 1)

DCP1B DCP1 decapping enzyme homolog 12 NM_152640 0.00126902 1.617 B (S. cerevisiae)

C1orf96 chromosome 1 open reading 1 BC039241 0.00132859 0.731 frame 96

LBR lamin B receptor 1 NM_002296 0.00136266 0.721

RFWD3 ring finger and WD repeat domain 16 NM_018124 0.00146132 0.668 3

GABRA3 gamma-aminobutyric acid (GABA) X NM_000808 0.00151597 0.764 A receptor, alpha 3

BLOC1S2 biogenesis of lysosome-related 10 NM_001001342 0.00153747 1.581 organelles complex-1, subunit 2

PIF1 PIF1 5'-to-3' DNA helicase homolog 15 NM_025049 0.00157311 0.711

153

Low dose irradiation reveals distinct signature

(S. cerevisiae)

RPS27L ribosomal protein S27-like 15 NM_015920 0.00163226 1.720

CSMD3 CUB and Sushi multiple domains 3 8 NM_198124 0.00163859 1.323

TMPO thymopoietin 12 NM_001032283 0.00166575 0.701

DEPDC1 DEP domain containing 1 1 NM_017779 0.00170147 0.438

BTBD8 BTB (POZ) domain containing 8 1 NM_183242 0.00172655 1.502

ZNF79 zinc finger protein 79 9 NM_007135 0.00181288 1.386

DIPAS DIPLA1-antisense expressed 9 AY623011 0.00191165 1.974

BUB1B BUB1 budding uninhibited by 15 NM_001211 0.00192402 0.503 benzimidazoles 1 homolog beta (yeast)

RACGAP1 Rac GTPase activating protein 1 12 NM_013277 0.00206076 0.575

IGF2 insulin-like growth factor 2 11 AK074614 0.00206316 0.771 (somatomedin A)

CCDC144C coiled-coil domain containing 144C 17 BC036241 0.00213011 0.690

LBA1 lupus brain antigen 1 3 NM_014831 0.00220708 1.603

BIRC5 baculoviral IAP repeat-containing 5 17 NM_001012271 0.00240896 0.661 (survivin)

KCNJ2 potassium inwardly-rectifying 17 NM_000891 0.00251286 1.442 channel, subfamily J, member 2

ZNF479 zinc finger protein 479 7 NM_033273 0.00255741 1.443

LOC388965 FUN14 domain containing 2 2 BC067852 0.00260329 0.730 pseudogene

OR8K3 olfactory receptor, family 8, 11 NM_001005202 0.0027281 0.645 subfamily K, member 3

CDC42BPA CDC42 binding protein kinase 1 NM_003607 0.00273631 1.483 alpha (DMPK-like)

RASD2 RASD family, member 2 22 NM_014310 0.00275632 0.748

TRIAP1 TP53 regulated inhibitor of 12 NM_016399 0.00284354 1.504 apoptosis 1

CCNG1 cyclin G1 5 NM_004060 0.00286532 1.571

154

Chapter 3

KIF23 kinesin family member 23 15 NM_138555 0.00291764 0.491

TNFRSF10B tumor necrosis factor receptor 8 NM_147187 0.00301736 1.435 superfamily, member 10b

FAM110A family with sequence similarity 20 NM_031424 0.00321613 0.744 110, member A

RETSAT retinol saturase (all-trans-retinol 2 NM_017750 0.00322187 1.394 13,14-reductase)

ARHGAP11A Rho GTPase activating protein 11A 15 NM_014783 0.00323685 0.478

DLG7 discs, large homolog 7 (Drosophila) 14 NM_014750 0.00343724 0.476

ISCU iron-sulfur cluster scaffold 12 NM_014301 0.00345027 1.362 homolog (E. coli)

ASAH3L N-acylsphingosine amidohydrolase 9 NM_001010887 0.00345292 1.651 3-like

FBXO22 F-box protein 22 15 NM_147188 0.00365871 1.430

ZNF222 zinc finger protein 222 19 NM_013360 0.00370529 1.531

UBE2C ubiquitin-conjugating enzyme E2C 20 NM_181802 0.00373307 0.592

TP53I3 tumor protein p53 inducible 2 NM_004881 0.00381974 1.464 protein 3

SYCP1 synaptonemal complex protein 1 1 NM_003176 0.00391452 1.303

LMNB1 lamin B1 5 NM_005573 0.00395391 0.664

FMN1 formin 1 15 AL833157 0.00398765 1.374

CR1L complement component (3b/4b) 1 NM_175710 0.00401051 1.359 receptor 1-like

MRPL22 mitochondrial ribosomal protein 5 AK097880 0.00406038 0.706 L22

NCAPH non-SMC condensin I complex, 2 NM_015341 0.00412711 0.582 subunit H

OR52K3P olfactory receptor, family 52, 11 AF143328 0.00415044 0.711 subfamily K, member 3 pseudogene

CDCA2 cell division cycle associated 2 8 NM_152562 0.00434206 0.544

RNF26 ring finger protein 26 11 NM_032015 0.00458945 0.682

155

Low dose irradiation reveals distinct signature

RRM2B ribonucleotide reductase M2 B 8 NM_015713 0.00463759 1.720 (TP53 inducible)

GRM1 glutamate receptor, metabotropic 6 NM_000838 0.00464107 0.726 1

UBE2S ubiquitin-conjugating enzyme E2S 19 NM_014501 0.00467546 0.616

NEK2 NIMA (never in mitosis gene a)- 1 NM_002497 0.00471467 0.448 related kinase 2

PLXNB2 plexin B2 22 NM_012401 0.00472966 1.589

CENPL centromere protein L 1 NM_033319 0.0047436 0.627

TRGV9 T cell receptor gamma variable 9 7 M16768 0.0048785 1.460

VNN2 vanin 2 6 NM_004665 0.00488848 1.279

123 Genes

Supplement 1: List of significantly regulated genes (p<0.005) in response to a low (62.5 mGy), moderate (0.5 Gy), and high (4 Gy) dose of X-irradiation in TPC-1. Table displays gene symbol in first column, full gene name in second column, chromosome number in third column, the NCBI accession number in the fourth column, the p-value in the fifth column and the log-2 fold-change in the final column. The p-value denotes the statistical significance of the gene (i.e. the probability that the gene is differentially regulated upon irradiation) whereas the fold-change refers to the ratio of gene expression in the experimental condition as compared to the control condition. We chose our threshold based on the statistical significance instead of the fold-change.

156

Chapter 4: External X-irradiation has an Immediate Impact on DNA Methylation in a Cell Line of Papillary Thyroid Carcinoma

Modified from: 'Abou-El-Ardat K, De Meyer T, Bekaert S, Van Criekinge W, Baatout S. External X-irradiation has an immediate impact on DNA methylation in a cell line of papillary thyroid carcinoma' [in preparation]

158

Chapter 4

External X-irradiation has an immediate impact on DNA methylation in a cell line of papillary thyroid carcinoma

1. Abstract The role of epigenetic changes in the cellular response to irradiation has been gaining more attention in the scientific world. The connection between radiation, DNA repair and epigenetic changes promises to be an interesting one. After the accident in Chernobyl, methylation changes were detected in the pine trees in the vicinity of the reactor. Changes in the methylation status of genes were also reported in several genes in cases of papillary thyroid carcinoma. We have used the power of next-generation sequencing to sequence isolated methylated stretches of DNA in TPC-1, a cell line of papillary thyroid carcinoma, using the MBD domain of human MeCP2. We have identified changes in the methylation status of several genes in response to a low (62.5 mGy) and high (4 Gy) dose of X-irradiation six hours after irradiation.

159

2. Introduction The link between high energy radiation and DNA damage has been established for quite some time. High energy radiation causes DNA modifications such as single-strand breaks (SSB), double strand breaks (DSB), and other modifications such as depurination and depyrimidination [104]. If not repaired, these modifications could lead to mutations and eventually could lead to cancer. A case in point is the increase in the incidence of childhood thyroid cancer in the residents of areas surrounding the Chernobyl nuclear power plant (NPP) nearly five years after the meltdown of reactor 4 in April of 1986 [89, 90]. A large number of oncogenic translocations, that have been ascribed to radiation-induced DNA DSBs, have been described in the literature since then. One the most important is the rearranged in transformation/papillary thyroid carcinoma (RET/PTC) translocation which has been found to be an initiator of PTCs [79, 80]. By the same token, radiation-induced DNA damage can be used to treat cancer by causing the death or senescence of tumorigenic cells.

The link between epigenetic changes and high energy radiation is not as well described. In the case of the Chernobyl NPP accident, DNA methylation changes have been described in plants growing in the vicinity of the reactor [247, 248]. Since then, more reports on the effect of radiation on the epigenetic profile of biological systems have come out. The methylation of several genes has been linked to the appearance of thyroid cancer or at least was found to be increased in cases of thyroid malignancies [233, 234, 359]. However, no work has tackled the effect of low doses of radiation on the epigenetic profile of normal and transformed thyroid cells.

DNA methylation refers to the methylation of cytosines at the fifth carbon and in humans usually occurs in a CpG context [225]. CpG islands are areas of CpG repeats found in roughly

40% of gene promoters and methylation of these regions in the promoter and first exon has

160

Chapter 4

been associated with transcriptional silencing [263, 360]. The advent of next generation sequencing (NGS) technology has opened up new vistas in biology by making it gradually cheaper and more convenient to sequence whole genomes [258, 259]. Combining isolation of highly methylated regions of the genome with NGS and deep sequencing made possible the identification of genes that are methylated under normal conditions and the changes that occur in response to various stressors down to a single base level [361]. The use of the naturally occurring methyl-CpG binding domain (MBD) protein has proven to be a successful manner in which to easily isolate methylated regions of the genome [257].

The effect of low doses of radiation, defined as any dose below 100 mSv, has been the subject of debate for a long time. The linear non-threshold (LNT) model, the most commonly used model to estimate health risks due to irradiation dose, have come under fire from some scientists as not being fully accurate [105, 110, 119]. Our previous work has shown that a low dose of 62.5 mGy elicits a distinct response on the genetic and epigenetic level in a system of normal and RET/PTC-positive thyroids (manuscript submitted). However, our work on the epigenetic level was limited to the microRNA (miRNA) response to radiation. We have extended our work here to studying early methylation changes in response to a low (62.5 mGy) and high (4 Gy) dose of external X-irradiation in TPC-1, a cell line of PTC, using MBD isolation of regions of methylation in TPC-1 and Illumina Solexa's Genome Analyzer IIx for sequencing.

161

3. Materials and Methods

3.1 Cell Culture TPC-1, a human cell line of PTC origin with a RET/PTC1 translocation, was obtained from

Dr. Horst Zitzelsberger (Helmholtz Zentrum München, Munich, Germany). The cells were grown in Dulbecco's Modified Eagle Medium (DMEM) supplemented with 10% (v/v) of fetal bovine serum (FBS) in a humidified incubator (37°C, 5% CO2) and split every other day.

Cells were tested for the presence of mycoplasma by PCR and were found to be mycoplasma- free. Three different passages of cells were used in the experiment, passages 22, 29, and 37 to guard against passage-dependent changes in methylation due to irradiation.

3.2 Irradiation of cells Cells were seeded in 25 cm2 tissue culture flasks 24 hours prior to irradiation and the medium replenished just before irradiation. Cells were irradiated with X-rays (1 mm Cu-filtered, 250 kV, 15 mA) at a dose rate of 5 mGy/sec from a Pentak HF420 RX machine. A low dose of

62.5 mGy and a high dose of 4 Gy were used while the control cells were sham irradiated in a manner similar to that of the high dose. The absorbed dose in Gray was calculated using the same formula used in medical procedures (i.e. 95% of the exposure in Roentgen). The cells were returned to the humidified incubator until harvest.

3.3 DNA extraction Cells were collected six hours after irradiation by trypsinization. The cells were pelleted and then frozen at -80°C until DNA extraction. DNA extraction was performed using the Qiagen

QIAamp DNA Mini and Blood Mini Kit (Qiagen, Germany) according to the manufacturer's instructions. Extracted DNA was resuspended in nuclease free water and frozen at -20°C. The concentration of DNA was quantified using the Nanodrop spectrophotometer at 260 nm with the 260/280 ratio used as a determinant of DNA purity.

162

Chapter 4

3.4 DNA Fragmentation The DNA concentration was measured again on the day of the experiment using Quant-IT

Picogreen kit (Invitrogen P7589, Merelbeke, Belgium) on a FluoStar Optima plate reader

(BMG Labtech, Offenburg, Germany) and 1 µg of DNA was moved to a microtube with an

AFA intensifier (Covaris, Woburn, MA, USA) containing Tris-EDTA (TE) buffer. The DNA was fragmented into 200 bp stretches using ultrasonic waves generated by a Covaris S2

(Covaris, Woburn, MA, USA) (duty cycle: 10%, intensity: 5, cycles/burst: 200, duration: 200 seconds; power mode: sweeping; temperature: 6-8°C). Fragmentation efficiency was controlled using Agilent's DNA High Sensitivity chip on the 2100 Bioanalyzer (Agilent

Technologies, Santa Clara, CA, USA).

3.5 Isolation of Methylated DNA using MBD Methylated DNA was isolated from the fragmented DNA mentioned earlier using the

MethylCap kit™ (Diagenode AF-100-0048, Liège, Belgium) according to the manufacturer's instructions. Briefly, the methyl binding domain (MBD) of human MeCP2 C-terminally fused with glutathione-S-transferase (GST) containing an N-terminal His6-tag was used to bind methylated segments of the genome. Afterwards, magnetic beads coated with GSH were used to capture those segments while a series of washes were utilized to discard all unbound DNA.

A final elution step resulted in the isolation of pure methylated DNA.

3.6 DNA Sequencing using Illumina Solexa's Genome Analyzer II Twenty-two µl of the isolated methylated DNA was subjected to PCR following the Illumina

Library Amplification Index Protocol (Illumina) with 21 cycles of PCR amplification. PCR products were purified on Qiaquick PCR Purification columns (Qiagen) and eluted in 50 µl elution buffer (1:5). The samples were then concentrated in a rotary evaporator to 10 µl and assessed using an Agilent 2100 High Sensitive DNA chip (Agilent Technologies). The concentration was afterwards determined by qPCR with a PhiX index3 standard solution

163

(Illumina). Pools were then diluted to 10 pM and used for sequencing on an Illumina Genome

Analyzer IIx following the Illumina protocol: 'performing a multiplexed paired-end run' (2 x

45 cycles).

3.7 qPCR on bisulfite and non-bisulfite treated DNA The primer design for selected genes was performed using an algorithm developed by our group. The primers were synthesized by Eurogentec (Belgium). For optimization of primers and to check that they amplify bisulfite-treated and not non-bisulfite-treated DNA, a qPCR on control DNA was performed (EpiTect PCR Control DNA Set (100), Qiagen, Germany). Ten ng of methylated and unmethylated bisulfite converted and unmethylated non-converted DNA was pipetted into a 96-well PCR plate with 0.5 µM of each of the forward and reverse primers and a SYBR Green PCR mixture (MESA Green Low qPCR MasterMix Plus for SYBR Assay

I Low ROX Mix, Eurogentec, Belgium). A qPCR was performed on a 7500 Prism PCR machine (Applied Biosystems) with the following program: 5' activation at 95°C followed by

40 cycles of 15'' denaturation at 95°C, 20'' annealing at 60°, and 40'' extension at 72°C. The efficiency of the primers was calculated using a 1:10 dilution standard curve on methylated bisulfite-treated DNA. To check the amplicon size of the PCR products, they were run on a

2% agarose gel with 0.01% Gel Red (Eurogentec, Belgium) at 100V for 30'. The bands were visualized using UV light and photographed digitally.

3.8 Data Processing and Statistical Analysis Coverage values were summarized using the map of the human methylome, which consists of putatively independently methylated regions (methylated cores) throughout the genome

(manuscript in preparation). For each sample, and each methylation core, the maximum read coverage was used for further analysis. A Poisson background model was used to identify significantly methylated regions (p<0.01) with lambda estimated as the sum of the coverage values over all methylation cores divided by the number of methylation cores (1,660,950). 164

Chapter 4

This approach takes into account coverage differences between samples, although generally low coverage will result in low sensitivity. For putative absence of methylation, a p-value of

>0.05 was used as a cut-off.

165

4. Results

4.1 The effect of radiation on the methylation of DNA Our results on the immediate effect of external X-irradiation on the methylation of DNA irrespective of dose received revealed only three significantly methylated genes: DNAJC22,

NCRO2, and ARID3A. None of these genes exhibited changes in their promoter methylation, which is associated with gene silencing (table 10).

4.2 The effect of radiation on the demethylation of DNA The number of genes demethylated due to external X-irradiation irrespective of dose was also low with only three genes significantly demethylated: PPP2R13B, CCDC85C, and PID1.

Only PID1 was significantly demethylated at the promoter region (table 9).

4.3 The effect of a high dose of radiation on the methylation of DNA Due to the limited number of genes that came out when we extended our results to include both low and high doses of radiation, we focused on the effect of a high dose of radiation (i.e.

4 Gy) on the methylation status of DNA six hours after irradiation of TPC-1. Forty genes now emerged as significantly methylated due to radiation among which only one gene, KCNMB3, was methylated in the first exon and one was methylated in the promoter region. Seventy- eight percent of the genes were methylated in their intronic region (table 11).

4.4 The effect of a high dose of radiation on the demethylation of DNA The number of genes that emerged as signficantly demethylated due to a high dose of irradiation in TPC-1 cells was around half of the genes that were methylated due to a high dose of radiation. Twenty-three genes were demethylated due to a dose of 4 Gy with two genes demethylated in their promoter region and one in the first exon. Seventy-four percent of the genes were demethylated in their intronic region (table 12).

166

Chapter 4

4.5 Primer optimization for deep sequencing We selected six genes that were interesting for us for confirmation. The six genes were:

NCOR2, ARID3A, APCDD1, PID1, NLRC5, and ADAMTS12. We tested nine primers against those six primers (for sequences, check table 13) and we found four primers against the three genes ADAMTS12, NLRC5, and ARID3A that amplified bisulfite treated but not non-bisulifte treated DNA (see figure 40). The amplicon size for those primers was around 300 bp which matched the predicted amplicon size (data not shown).

167

5. Discussion We have attempted in this paper to investigate the effect of external X-irradiation, and particularly in the low dose region, on the epigenetic profile of TPC-1, a cell line of papillary thyroid carcinoma with a RET/PTC1 translocation. We have found that X-rays had an immediate effect on the methylation status of genes in that methylation changes were detected six hours after irradiation in response to both low (62.5 mGy) and high (4 Gy) doses of X- rays.

Sequencing of areas of the genome enriched for methylation is a cheaper alternative to the bisulfite sequencing of the whole genome. Bisulfite sequencing uses bisulfite-treated DNA, which serves to convert cytosine but not 5-methyl-cytosines (5-meC) to uracils [362].

Alternative methods to bisulfite sequencing include DNA-methylation specific bead arrays for bisulfite treated DNA but those are limited by the probe design [363]. Isolating methylated areas of the genome using either antibodies or methyl-binding proteins serves as a cheap genome-wide approach. Research has shown that using methyl-CpG binding domain (MBD) proteins (known as MethylCap-seq or MiGS) is more sensitive than methylated DNA immune-precipitation (MEDIP) given enough coverage [361, 362]. We have used the

MethylCap™ kit from Diagenode to isolate areas of methylated DNA in the TPC-1 cell line, as this kit was shown to have the best performance among five kits tested and demonstrated the highest sensitivity and specificity (De Meyer et al., submitted manuscript).

Aberrant methylation in thyroid cancer has been described for quite some time and silencing of genes by hypermethylation continues to be reported in the literature [233, 234, 241, 359].

The methylated regions of DNA obtained here correspond to peaks of methylation in a library of ~80 normal and cancerous tissues and cell lines (Trooskens et al. manuscript in preparation). As a proof of principle, we checked whether genes reported to be hypermethylated in thyroid cancer in the literature were found in our sequenced DNA. We 168

Chapter 4

found all but two of the genes in our data; RIZ-1 and hMLH1 being the only two genes reported to be silenced in PTC but not found in our list of genes [238, 246]. RASSF10,

RASSF2, CITED1 and SLC26A4 were all among the genes whose promoter was methylated

[235, 237, 239, 240]. In addition, TP73, whose product p73 was found to be absent on the transcriptional and translational level in TPC-1 cells, was among the genes with a methylated promoter region [303, 314]. It remains to be said that the appearance of these genes in our data does not mean that they are necessarily silenced in this cell line, especially since we have not compared our results against normal thyroid follicular cells. This only proves that those genes contain CpG islands in their promoters, as reported in the literature, and that those

CpGs are methylated enough to be detected by the MBD protein.

The DNA damage response (DDR) is a very crucial aspect in the survival of a cell to DNA damage. This response involves first and foremost the DNA damage repair response which is a very complex procedure, especially when DNA double strand breaks (DSB) are involved.

We have found around two-thirds (62%) of the genes involved in DNA repair in our sequenced DNA. All those genes were methylated in their promoter region and involved genes that are implicated in DSB as well as single strand break (SSB) repair such as XRCC5,

XPC, and MLH3.

We have obtained a list of genes that are both methylated and demethylated due to irradiation.

Only three genes were significantly methylated due to irradiation at a low and high dose of radiation and the same number of genes were significantly demethylated at both doses. When we searched for genes that were significantly methylated and demethylated in response to a high dose of radiation, the list grew to 40 genes that were methylated and 23 demethylated genes. The changes in methylation occurred all over the gene body; changes in the methylation of gene promoters, which is associated with gene silencing, were not so common

169

in our case. There were two cases of changes in the methylation of the first exon which has also been linked to gene silencing [263]. Most of the changes in methylation occurred in intronic and exonic sequences with intronic sequences being the most predominant (see tables

11 and 12). Methylation of intronic sequences has been found to be linked in some cases to the transcription of microRNAs (miRNA) while methylation of exonic sequences was found to control alternative splicing [264-267]. However, work on the function of non-promoter sequences is still an ongoing process and other functions of epigenetic changes in these regions will surely be uncovered. The nature of active demethylation of DNA is still a point of contention in the scientific community. Active demethylation has been described in some biological systems and one of the reported mechanisms of active demethylation is a mechanism similar to that used for DNA SSB repair [364].

The gene that was methylated in its promoter due to a high dose of radiation was CYP20A1 while KCNMB3 was methylated in its first exon. GATAD2A and PID1 were demethylated in their promoter region while NEFM was demethylated in its first exon. NYGGF4 (PID1) function has been studied in adipocytes mostly; however, it was found to participate in signaling via Akt through its phosphorylation [365]. Akt is one of the central molecules in the

PI3K signaling pathway, a pathway that is overexpressed in RET/PTC-bearing cells [78].

Among the genes that are methylated in other regions of the gene body besides the promoter and first exon are genes involved in inflammation such as NCOR2 and NLRC5 while some genes are reported in the literature to be possible oncogenes or to be associated with cancer such as IGF2BP2 and PRDM16 [366-369]. There were more genes involved with cancer among the demethylated genes such as CBFA2T3, APCDD1 and ADAMTS12 especially the latter which is reported to be a cancer marker [370-372]. SAPS2 is reported to be involved in

DNA repair which is important in the context of radiation-induced DNA damage [373].

170

Chapter 4

However, since the methylation and demethylation occur in areas outside the promoter and first exon, their significance is not totally clear and should be investigated further. What we have shown is that DNA methylation is altered in cells as early as six hours after irradiation and that these changes could have an impact on aspects of the cellular response to radiation and stress such as inflammation and cell cycle. The fact that the number of genes that appear to be differentially methylated in response to irradiation is dose-dependent might indicate that the change in methylation is dependent on the amount of DNA damage in the cell. The fact that the majority of methylation changes occur in intronic and exonic sequences might indicate that the immediate impact of external X-irradiation on the cell is a shift in the miRNA profile or alternative splicing in genes involved in the immediate stress response rather than the silencing or expression of genes. Perhaps the latter response is more apparent at later time points, which warrants further investigation. In addition, how those methylation changes influence the response of cells to radiation is also very important for future research.

We have shown in this paper that irradiation has an effect on both the methylation and demethylation of genes six hours after irradiation. Irradiation caused both methylation and demethylation of genes involved in various cellular processes such as DNA repair and inflammation which have a direct bearing on cellular response to radiation. Further confirmation of these results by deep sequencing will be needed as well as the elucidation of the impact of these changes on the cellular response to radiation. Confirmation of the results obtained in this study is crucial, as it is in any high-throughput technique, to avoid the inclusion of false-positives.

171

Figure 40: qPCR data on bisulfite treated and non-bisulfite treated DNA. CT value refers to the cycle number when the amplification curve crosses the assigned threshold. MC = methylated bisulfite converted, UMC = unmethylated bisulfite converted, UMNC = unmethylatede non-converted Table 9: Genes demethylated due to radiation (low and high)

Gene Name Chromosome Location

PPP2R13B 14 Intronic CCDC85C 14 Intronic PID1 2 Promoter

Table 10: Genes methylated due to radiation (low and high)

Gene Name Chromosome Location

DNAJC22 12 Exonic NCOR2 12 Intronic ARID3A 19 Intronic

Table 11: Genes methylated due to radiation (high dose only) 172

Chapter 4

Gene Name Chromosome Location

SGMS1 10 Intronic PPP3CB 10 Intronic SORCS1 10 Intronic DNAJC22 12 Exon 3 FBXO21 12 Intronic NCOR2 12 Intronic GTF2F2 13 Intorinc PSTPIP1 15 Intornic NLRC5 16 Exon 20 ABR 17 Intronic TANC2 17 Intronic ARID3A 19 Intronic AC092295.3 19 Intronic NOSIP 19 Intronic PRDM16 1 Intronic SAMD13 1 Intronic LAMC2 1 Exon 16 PLEKHA6 1 Intronic GALNT2 1 Intronic AHCY 20 Intronic APCDD1L 20 Intronic ZBTB46 20 Intronic PLA2G6 22 Exon 3 CYP20A1 2 Promotion KCNMB3 3 Exon 1 IGF2BP2 3 Intronic AC091947.1 5 Intronic

173

RUFY1 5 Intronic BACH2 6 Intronic EXTL3 8 Intronic RGS20 8 Intronic RNF19A 8 Intronic FAM84B 8 Exon 2 ZFAT 8 Intronic PAX5 9 Intronic FAM125B 9 Intronic CAMSAP1 9 Intronic EHMT1 9 Exon 16 AF196972.5 X Intronic PDZD4 X Exon 4

Table 12: Demethylated genes due to irradiation (high dose only)

Gene Name Chromosome Location

C10orf92 10 Intronic SYT12 11 Exon 7 ELK3 12 Intonic GPR81 12 Intronic CCDC85C 14 Intronic PPP1R13B 14 Intronic GABRA5 15 Intronic CBFA2T3 16 Intronic MYO15B 17 Exon 25 APCDD1 18 Exon 3 FAM59A 18 Intronic ME2 18 Intronic

174

Chapter 4

GATAD2A 19 Promoter SAPS2 22 Intronic PID1 2 Promoter AGAP1 2 Intronic ADAMTS12 5 Intronic ZNF354A 5 Intronic C6orf132 6 Intronic EPDR1 7 Intronic KIAA1147 7 Intronic NEFM 8 Exon 1 FBXW2 9 Intronic

Table 13: Primer sequences

Primer Forward Reverse

PID1 5'-TTAATTTATATAGGGGTGAGGAGGG-3' 5'-

AACAAATAACCTAAACCTTATAAACCTCAT-

NCOR2_1 5'-TTGTTGTGTTGAGTTAAAGGGTTAT-3' 5'-

TTAAAATACCAAAACACCCTAAAATCAAC-3'

NCOR2_2 5'-TTGGTTATTTGTGGGTTTTAGTGA-3' 5'-

TTTTAAAACTTCCCATTAACTAAATCCAAC-3'

NLRC5 5'-TTATAAAGAGGATGTGTTTGGGTTTT-3' 5'-ACATACTCTACAAATCTAATACTACCCAA-

ARID3A 5'-TAAGGTTAGGTTTGATTAGGGTTTATTG- 5'-AATCTTAAACTCCTAACCTCAAATAATCC-

3' 3'

APCDD1_1 5'-TTTTTTGGATTATTCGAGGGGG-3' 5'-TAAAAATACTACTTCTCCACCCGA-3'

APCDD1_2 5'- 5'-CCTACACCCAAAAACCCC-3'

175

AAGGTTAGGTTTAGAGTTTATTATAAGGTT-

ADAMTS12_1 5'-TTAGGTTGGAATGTAATGGTGTAATTT-3' 5'-ACTCTTAATCCAATAAAACCACAATTCT-3'

ADAMTS12_2 5'-GATGTTGGAAGTATAAGATTAAGGTGTT- 5'-

3' TAATCCCAACTACTCGAAAAACTAAAATAA-

176

Chapter 5: General Discussion and Perspectives

General discussion and perspectives

"This is not to say that facts or data are nonexistent, but that facts get their importance from what is made of them in interpretation…for interpretations depend very much on who the interpreter is, who he or she is addressing, what his or her purpose is, at what historical moment the interpretation takes place."

Edward Said

178

Chapter 5

General discussion and perspectives It is hard to believe, seeing the panic and pandemonium spread by the mere mention of the word 'nuclear', that only 90 years ago atomic or nuclear energy was synonymous to world peace, technological advancement, and even magic. That was the age of Radithor, water laced with radium that was supposedly good for you, until it claimed its first victim who died of radiation poisoning. It was the age when the shoe section in department stores in the US had to include an X-ray machine so that you and your friends could see how your foot fits inside a new shoe. It didn't stop there; a 1929 European pharmacopoeia listed 80 patent medicines whose active ingredients were radioactive and many doctors were using X-rays to treat warts and excess facial hair2.

Since then, the attitude of people towards anything nuclear has taken an about face. As the link between radiation and mutations and cancer consolidated itself in the minds of the masses and as the images of Hiroshima and Nagasaki and Chernobyl poured in, a general mistrust of nuclear energy and radiation permeated the general discourse. Now the word „nuclear‟ has to be removed from „magnetic resonance imaging‟ so as not to alarm the patients and Germany vows to dismantle all of its nuclear power plants within ten years. With each ICRP

(international committee for radiation protection) report, the maximum permissible annual radiation dose for the public is lowered. The European Commission (EC) is pouring millions of euros into research on the effect of low doses of radiation on various organs and systems in the body. And the debate over the biological effect of low doses of radiation still carries on.

Peoples' perception of nuclear power had changed long before the accident at the Chernobyl

NPP. However, that accident served to show how far human error can go and how one

General discussion and perspectives

mistake could have consequences that reverberate for decades. The accident at Chernobyl was a result of the accumulation of many errors over time from the design of the RBMK type reactors which had no containment vessel to the enormous deviations from the original plans during construction to the security gaps left by the operating team on the night of the fateful accident [1, 6, 8, 89]. Since then, numerous studies have investigated the health effects of the

Chernobyl disaster, thus quelling many of the circulating rumors which greatly exaggerated the numbers of dead and injured [374]. However, most of the studies up until recently focused on the overall health effects of the Chernobyl disaster. Recently, focus has centered on the effect of low doses of radiation on the thyroid since the range of doses received by the population surrounding the site ranged from the low to the very high. The strongest link between the Chernobyl disaster and adverse health effects was that of thyroid cancer in children. The link between the Chernobyl disaster and other cancers such as leukemia and breast cancer was also suggested by the data but was harder to prove. Among the non-cancer effects of the Chernobyl disaster was the link with cardiac problems and cataracts which is gaining more attention in the scientific community [22, 391, 392]. An increased risk for the development of cataracts was found for children and liquidators with a suggestion that the threshold should be lower than the <1 Gy employed for adults.

The aim of this thesis was the identification of the effects of low doses of external radiation on normal and transformed thyrocytes. We have established by several means that low doses of radiation have a distinct effect on RET/PTC-bearing thyrocytes that differ from these thyroid cells‟ response to high doses of external radiation and to the response of normal thyrocytes to said doses.

We chose a range of X-ray doses that ranged from the low to the high and that fell within the range of effective doses to the thyroids of the residents of the area surrounding the Chernobyl

180

Chapter 5

NPP. The range of doses that is reported in the literature starts at 90 mGy and goes up to a high dose of 12 Gy [1]. However, around a million residents received a very low dose of around 1 mGy; our results probably cannot predict the risk of such a low dose.

Like everything, using cell lines for low-dose research has its advantages and disadvantages; cell lines always have the disadvantage of losing some of the properties of the original cells they are derived from with each subsequent passage. Cell lines are also victims to the many stresses of cell culture, from the unnatural matrix they are grown on in 2D cultures to trypsin:EDTA solutions that are used to detach cells and serum starvation as a result of leaving the cells in the same culture medium for extended periods of time such as over a weekend. Furthermore, there is the mix-up that could happen when culturing cells for a long period and handing them down from research group to research group as has happened with some thyroid cell lines [375]. However, despite all these disadvantages, cell lines possess the advantages of being easy to handle and grow. In addition, cell lines are a homogenous system where the subtle effects produced by feeble doses of radiation are not lost or are not muddled by the existence of other types of cells such as might happen in tissues.

We used TPC-1, a cell line from a female patient with papillary thyroid carcinoma (PTC) with a RET/PTC1 translocation, wild-type p53 and genetic imbalances including the deletion of chromosome 21 and several deletions and translocations [298, 376]. Genomic instability is a feature of thyroid cancer in general and PTC in particular while some mutations such as the

BRAF V600E mutation were linked to genomic instability in PTC [331, 332, 377]. Therefore, we can assume that TPC-1 is representative of PTC in that aspect.

Irradiation of TPC-1 with a range of doses of X-rays caused a delay in the cell cycle with an increase in cells in the G1 and G2 phase and a decrease in cells in the S phase. This cell cycle profile has been long linked to the onset of senescence [184]. Senescence is a normal 181

General discussion and perspectives

physiological phenomenon associated with the shortening of the telomeres beyond a certain limit known as the Hayflick limit. This state can also be brought on by several stressors including DNA damage and oncogenes. Stress induced senescence profile (SISP) includes, besides the unique cell cycle profile, such signs as an increase in senescence associated beta- galactosidase (SA β-gal) staining, senescence associated secretory profile (SASP), and an increase in cell size and granularity [179, 188, 378, 379]. Our results showed that all the above-mentioned phenomena appeared at an irradiation dose of 0.5 Gy and higher and never at lower doses which indicates that the senescent phenotype in TPC-1 cells needs a certain

DNA damage threshold to manifest. We've observed that the number of DNA damage foci per nucleus increased at a dose of 0.5 Gy and above but not at a dose of 62.5 mGy. In addition, the size of the foci, an indication of the complexity of the DNA damage, increased at a dose of 0.5 Gy and above and not at 62.5 mGy. Therefore, we can assume that the appearance of a senescent phenotype is dependent on both the number of foci per nucleus and the complexity of the foci. Previous research indicates that the cellular response to DNA damage is dependent on the degree of damage [380]. When the DNA damage is low, pro- survival mechanisms such as DNA repair and cell cycle arrest are activated but when the

DNA damage is more severe, pro-elimination pathways are activated instead. These pathways include apoptosis, differentiation, and senescence and a central player in all of them is the tumor suppressor TP53 [380].

The importance of p53 has been a recurrent theme in this thesis; BCPAP, a cell line of papillary thyroid carcinoma origin with a V600E point mutation and a mutated copy of p53, did not respond to radiation with an up-regulation of neither p53 nor the CDKI p16INK4A nor did it display any signs of senescence in response to radiation. In chapter 2, we explore the differences in the response of TPC-1 and BCPAP cells to a range of low to high doses of X-

182

Chapter 5

rays and conclude that the difference between the response of the two cell lines to radiation lies in their different p53 status. The role of p53 in cells has been much elaborated on in the literature and its importance to the proper function of cells and is a huge obstacle in the face of tumorigenicity. Therefore, it‟s no wonder that p53 is mutated in around 50% of cancers and has earned it the title of „guardian of the genome‟ [381]. The tumor suppressor p53 lies in the center of so many pathways and disrupting it causes myriad changes in cellular physiology from unchecked proliferation to genomic instability.

However, the p53 response does not seem to be relevant to low doses of radiation as we did not detect p53 activation (neither an increase in the total p53 levels nor increase in p53 phosphorylation on serine 15) below a dose of 1 Gy. The lack of p53 activation at low doses could be due to the DNA damage not being too complex, as research has shown that p53 activation follows two patterns: an on/off pattern and a gradient pattern [382, 383]. Our results indicate that p53 follows the on/off pattern of activation where it is activated at doses above 0.5 Gy but not below that.

Naturally, the conclusions presented here must be read within the context of the models used and the experimental setup. The importance of p53 in the low dose response was observed in in vitro culture whose limitations were presented earlier. In addition, BCPAP, which possesses a mutated copy of p53, carries other mutations such as BRAF V600E which could also be involved in the low dose response and thus would need to be factored in. Therefore, drawing absolute and direct cause/effect relationships between the molecular players involved in this study and radiation is impossible without additional experiments such as the ones outlined in the end of this discussion.

RET/PTC is a translocation that was first described by Greco et al. in the early nineties as the long sought after PTC oncogene [79]. The fact that the ret protooncogene was important in 183

General discussion and perspectives

thyroid neoplasia was known before discovery of the RET/PTC translocation. Point mutations of ret were behind such thyroid malignancies as MEN2A and FMTC [78]. The translocation, however, proved to be prevalent in PTCs and especially in PTCs caused by the Chernobyl disaster a few years earlier [90-92, 301]. Would RET/PTC translocations, once induced in some follicular cells in the thyroid, alter those cells‟ response to radiation? In chapter 3 we compared the response of transgenic mice with a RET/PTC3 translocation to X-irradiation to that of TPC-1 cells. We then attempted to compare and contrast both those system‟s response to that of wild-type murine thyroids. The transgenic mice were normal C57BL/6J inbred mice with RET/PTC3 translocation under the control of a bovine thyroglobulin promoter. We chose mice that were three months of age because those mice were described in the literature as not bearing any tumors but only signs of hyperplasia [324, 338]. This would allow us to identify the role of RET/PTC in the response of these cells to radiation as TPC-1 cells possess in addition to the RET/PTC1 translocations several other mutations and deletions and are tumorigenic. By comparing TPC-1, RET/PTC-bearing murine thyroids and wild-type thyroids, we could better pinpoint the genes that are radiation responsive in RET/PTC- positive systems only and thus identify a RET/PTC-specific response to radiation. This method is naturally not ideal since the two RET/PTC systems are from two different species and from cancerous and hyperplastic cells. In addition, the fact that not all the follicular cells in the RET/PTC-positive murine thyroids possess the RET/PTC translocation means that the power of the microarray test is lowered due to the presence of normal follicular cells within the analyzed tissue. This might explain why, when a false discovery rate (FDR) correction was applied to the pathway analysis, most pathways did not pass the significance threshold

(p=0.05) in RET/PTC-positive thyroids but not in TPC-1 cells which all possess the RET/PTC translocation.

184

Chapter 5

Incidentally, one of the major differences between the response of normal and RET/PTC- positive thyroids to low and high doses of radiation was the lack of p53 response in normal thyroids. We have shown by Western blots that p53 levels indeed are not up-regulated in response to radiation in normal thyroids. Instead, the pathway that was always regulated in response to radiation in these thyroids was the TGF-β apoptotic pathway. This fact could be explained by the finding that PTC cells have down-regulated SMAD3/4 and so usually have a dysregulated TGF-β signaling axis [384]. In addition, some of the pathways enriched for at a low dose of radiation in normal thyroids were pathways involved in cellular growth and proliferation. This was confirmed using a multiplex bead assay on proteins extracted from normal thyrocytes which revealed an increase in the levels of total ERK1/2, Total JNK, and phosphorylated JNK at this dose. These pathways are among those up-regulated in RET/PTC- bearing thyrocytes [77, 80, 84, 103, 385]. The activation of these cellular growth and proliferation pathways could possibly imply an increased risk of developing cancer at low doses in normal thyroids. This would also suggest a mechanism by which low doses of irradiation received in the thyroid of children caused an increased risk of PTC later on in life

[96].

When the whole human genome was finally sequenced, large portions appeared to be without any discernable function and thus this DNA was called „junk DNA‟. Later it transpired that some of that DNA actually encoded stretches of RNA that acted by degrading other mRNA thus adding a further layer of complexity to transcriptional control. MicroRNAs, as they were later called, are now found to inhibit mRNA by either degrading them or inhibiting their translation. Furthermore, they were found to be dysregulated in cases of cancer and in response to stress such as radiation [252, 273, 292]. In chapter 3 we describe a cohort of miRNAs that respond to irradiation in TPC-1. In addition to this radiation-responsive miRNA

185

General discussion and perspectives

signature, we uncovered miRNAs that responded in a dose-dependent manner. For example, miR-17 was only regulated at a low dose of 62.5 mGy and not higher. What is the significance of the dose dependency of miRNA regulation? Unfortunately, miRNAs have hundreds of predicted mRNA targets and trying to figure out which of the targets is regulated by miRNA in response to radiation needs extensive testing. Luckily, we had the microarray data on the same RNA we used to analyze miRNA response to radiation. By searching for predicted miRNA targets in our microarray data we could speculate as to the role played by various miRNA in the TPC-1 cells‟ response to radiation. Unfortunately, due to the limited number of genes that were significant at a low dose of radiation, we couldn't find any common genes between the predicted targets of miR-17 and our list of genes. However, one of the predicted targets of miR-17 is the peroxisome proliferator antigen receptor (PPAR)-α which is involved in tumor proliferation and miR-17's up-regulation may be one mechanism by which the inhibition of TPC-1 proliferation occurs [386].

As mentioned earlier, the translation of results from an in vitro system to an in vivo one must be done with care; in vitro cultures are homogenous systems where the clonal expansion of cells ensures little variability. However, changes in gene expression and function could be seen between passages in our data which made it crucial to include cells from different passages in our studies so as to minimize the passage influence. On the other hand, in vivo systems are more complex with various types of cells and factors influencing each other. In our case, the thyroids contained two types of cells as well as the microvasculature in the thyroid. In addition, the effective dose that is delivered to cells in a tissue culture flask and a thyroid in a whole-body-irradiated mouse will be different. Despite all this, we observed the activation of the p53 pathway at 4 Gy in both in vitro and in vivo systems. At lower doses, the response was somewhat different, although signs of cell cycle checkpoint activation was

186

Chapter 5

evident at 0.5 Gy in the two systems. It is possible that at lower doses other factors come into play in the tissue such as the bystander effect especially since we observe a change in the thyrocyte secretome that contribute to the difference in the radiation response in the two systems.

Our in vivo results are probably more relevant to the situation in humans since both are biological systems. The studies of Burniat et al. showed that the transcriptomic data from

RET/PTC3-positive mice at 2 months of age are probably the best model of PTC in humans.

We chose mice of the age of three months because it would equal the age of 8 to 10 years in human years (based on a life expectancy of two years for a normal lab mouse and 70 years for humans). This age has been associated with an excess risk for the development of radiation- associated thyroid cancer and the average age of children who developed thyroid cancer after

Chernobyl. However, there are still other factors to consider when translating data obtained in vivo in mice to humans; although three months in mice is mathematically equivalent to around nine years in human years, the biological reality provides a more complex situation. For example, mice are sexually active between 30 and 50 days of age, their mass-specific metabolic rate is higher than humans, and they have a higher production of ROS with a lower ability to maintain homeostasis. Furthermore, lab mice are inbred and thus have a more homogenous genetic background than human subjects and their biological processes proceed at a faster rate [387-389]. Despite this, mouse models remain a valuable source of information and the data obtained in mice in clinical studies are usually comparable to those obtained in humans. We thus find that the transcriptomic profile obtained in wild-type thyroids after a low dose of irradiation predicts an increased risk of cancer development, the same that was found in children after receiving low doses to their thyroids during medical procedures [162]. In summary, despite the clear limitations of the different model system described above, they

187

General discussion and perspectives

can provide fundamental information on the molecular mechanisms involved that are also relevant for humans such as in pediatric medicine.

The epigenetic changes induced by radiation were not limited to microRNAs in our study but extended to changes in DNA methylation. In chapter 4, we analyzed the changes in the TPC-

1 methylome in response to two doses of X-rays six hours after irradiation. We utilized next generation sequencing (NGS) of regions enriched for methylation. Isolation of highly methylated sequences of the genome using antibodies or naturally occurring proteins has been used as a cheaper alternative to whole genome bisulfite sequencing. The method does have its drawbacks such as the lack of detection of non-CG methylation [257]. However, for our purposes, it was a suitable method for detection of methylation changes in response to low and high doses of X-rays. We detected methylation changes in the genome of TPC-1 cells at an early time-point of six hours, a time that is too early for these cells to divide [390]. This would mean that we are measuring the direct impact of radiation on the genome. Since we observed a small percentage of the methylation changes in the promoter or the first exonic region, this may mean that the effect of radiation has more to do with alternative splicing of genes and other epigenetic processes than with gene silencing. However, there was no correlation between genes methylated in their promoter and first exon regions and genes down-regulated in our microarray data. This is not surprising since we used the same time- point for both and would possibly warrant inspecting the expression changes in these genes at later time-points in this cell line. Finally, active demethylation of DNA is a not so well described process and our results on the active demethylation of DNA in response to irradiation would need further confirmation.

In conclusion, we have attempted in this thesis to uncover the effect of low doses of radiation on normal and transformed thyrocytes. We have taken a rather holistic approach to the matter

188

Chapter 5

of low dose effects by combining several high-throughput techniques such as DNA microarrays, next generation sequencing and low-density qPCR arrays with studies of cellular morphology, and protein levels and modifications to inspect aspects of the transcriptional, translational and epigenetic modifications induced by low doses of radiation. We conclude that the effects of low doses of radiation are system dependent. For example, we observed differences in the thyrocytes' response based on whether they possessed the RET/PTC translocation or whether they carried a working copy of p53. Normal thyroids appeared the most responsive to low doses of radiation, with many cellular growth and proliferation pathways up-regulated at the transcriptional and translational level. RET/PTC-positive cells on the other hand responded to a low dose of radiation by decreased signaling down the

MAPK pathway and activation of cell cycle checkpoints. However, when p53 was mutated, this phenomenon was absent at low doses of radiation. In all cases, there was a distinct difference between the cellular response to low and higher doses of radiation. This was manifest at the transcriptional, translational and epigenetic levels which may also indicate an interplay between all those levels in the response to low doses of radiation.

In addition, we have here focused on the early response of thyrocytes to radiation. This would mean that although we detect the direct effect of radiation on cells, we can mostly infer the phenotypic response of these cells on the short and long term. Perhaps the long-term effect of radiation on RET/PTC transformed and wild-type thyroid can be the focus of further studies.

This could be interesting for thyroid cancer therapy, especially for the high doses, and for dosimetry and biomonitoring, especially for the low doses. In addition, the way that the

RET/PTC translocation alters the response of the thyroid to radiation could have an impact on our understanding of radiation-induced thyroid cancer progression and should be further expounded in future studies. The effect of low doses of radiation on the transcriptomic profile

189

General discussion and perspectives

of wild-type thyroids should be of special concern since it shows a possible increased risk of cancer development on the long run. Therefore, a long-term study on murine thyroids irradiated with a low dose of X-rays where the mice are monitored for thyroid cancer is advisable. This could have an impact on radioimaging in children or on the radiotherapy of head and neck cancers in children where the thyroid could receive low doses of radiation. The

LNT model is based on the linear extrapolation of health effects to low doses of radiation based on data obtained from such studies as the LSS study and epidemiological data as mentioned in the introduction (section 4.1). The model has been widely used as a basis of risk estimation and radioprotection measures and assumes that the cellular response to radiation is uniform at all doses. Deviations from linearity have been consistently reported in the literature

(see section 4.4). Our transcriptomic data on normal murine thyroids at low doses of radiation as well as the altered secretome at higher doses of radiation in TPC-1 cells could be of special interest to legislators for risk assessment at low doses and might suggest deviations from linearity.

Perhaps the effect of chronic low dose radiation would also be interesting. The results of the effect of irradiation on the miRNA profile in thyroid cells should also be expounded in further studies. For example, the subset of miRNAs that responded to irradiation at all doses deserves further research. The targets of these miRNAs would be of special interest in the field of radiation biology. Furthermore, this subset of miRNAs could be invaluable for biomonitoring of radiation exposure and their potential as thus should be inspected further. The next steps would also be to dissect exactly how this interplay causes the difference in response to low doses of radiation and what is the role of such factors as RET/PTC and p53 in mediating this response. Additional studies using siRNA and gene transfection would also be helpful in elucidating the role of some of the important genes in our study to the phenotypic response to

190

Chapter 5

radiation observed. Also, a long-term study on the effect of those mechanisms on the thyroid will help to determine whether an acute low-dose of X-rays will have a detrimental effect on human health in the long run. Below is a schematic summary of the main results and conclusions of this thesis (figure 41).

Figure 41:Summary scheme of thesis results.

191

Summary (English and Dutch)

193

Samenvatting

Calvin Coolidge, the famously taciturn 30th president of the USA, was sitting at dinner when a fellow guest attempted to lure him into conversation. 'I have made a bet, Mr. Coolidge,' she began, 'that I can get more than two words out of you.' 'You lose,' replied Coolidge.

194

Summary

Summary The Chernobyl disaster in April of 1986 opened the world's eyes to the dangers of nuclear power plants. In subsequent years, 'Chernobyl' became synonymous to an off-scale disaster and many rumors started circulating about the extent of the damage and the number of casualties. Extensive research on the effects of the Chernobyl disaster dispelled most of the rumors and the strongest link that emerged between the Chernobyl accident and health risks is the increase in cases of childhood thyroid. Indeed, the release of huge amounts of radioactive iodine into the atmosphere following the blast at reactor 4 was linked to the appearance of cases of childhood papillary thyroid carcinoma (PTC) in the regions surrounding the reactor.

A large percentage of these cases carried an oncogenic translocation, RET/PTC, that is linked to the initiation of PTC. Since the population surrounding the disaster site was exposed to a range of radiation doses to their thyroids ranging from the low to high and since exposure to low doses of radiation during childhood has been linked to the appearance of PTC later in life, we were interested in inspecting the effects of low doses of radiation on normal and

RET/PTC-bearing thyroid cells.

The effects of low doses of radiation in mammalian cells remain controversial. The 'linear non-threshold' (LNT) model has recently come under fire as not being accurate. In chapter 2 we investigated the effect of a range of X-rays on two cell lines of PTC with two different oncogenic mutations and different p53 status. We described a bi-phasic response to radiation centering around a moderate dose of 0.5 Gy in TPC-1 cells with RET/PTC1 translocation and wild-type p53. Below that dose, there were signs of cell cycle checkpoint activation and a decrease in the fraction of actively dividing cells (cells in S phase). At a dose of 0.5 Gy and above, cell cycle checkpoint activation was associated with signs of stress-induced senescence such as an increase in senescence-associated β-galactosidase, increased TGF-β1 secretion and the appearance of an altered secretory profile. This was associated with an increase in the

195

Samenvatting

levels of p53, p21, and p16. We hypothesized that this difference in response to radiation may be due to the nature of DNA damage induced in these cells in response to various doses of X- rays. At a dose of 62.5 mGy, the DNA double-strand breaks were sparse and less complex than those at 0.5 and 4 Gy. We also proposed that p53 is crucial to the phenomena we described as BCPAP, which carries a mutated copy of p53, did not display any signs of senescence and showed no response to radiation-induced DNA damage at doses below 1 Gy.

The role of p53 in the TPC-1 cells' response to irradiation was further elaborated on in chapter 3. Microarray analysis of the response of TPC-1 cells to X-rays revealed an enrichment for genes involved in the p53 response at doses of 0.5 and 4 Gy but not at 62.5 mGy. We also uncovered the p53 pathway activation in the thyroids of transgenic C57BL/6J mice with a RET/PTC3 translocation in response to a dose of 4 Gy. TPC-1 cells differ from

RET/PTC3-bearing thyroids in that the former are an established cell line of PTC while the latter display signs of hyperplasia. On the other hand, the p53 pathway activation was absent in normal murine thyroids in response to X-irradiation while the TGF-β apoptotic pathway was enriched at all irradiation doses in these thyroids. Interestingly, many cellular growth and proliferation pathways were highly enriched in these thyroids at a low dose of 62.5 mGy, a fact that we confirmed at the protein level. These pathways include the mitogen activated protein kinase (MAPK)/ERK1 pathway and the c-jun N-terminal kinase (JNK) pathway and would thus propose a mechanism by which tumorigenesis is induced in normal thyroids in response to low doses of radiation only. However, more experiments need to be done on those thyroids such as long-term follow up accompanied with histological and microarray data analysis.

In chapters 3 and 4, we investigated the epigenetic effect of low and high doses of radiation on TPC-1 cells. We identified a cohort of microRNAs (miRNA) that are radiation-responsive irrespective of the dose received. On the other hand, miRNAs linked to the regulation of p21 196

Summary

and cyclin G1 were only responsive to 4 Gy of X-rays. We also identified miRNAs that are regulated at 62.5 mGy and 0.5 Gy only. Furthermore, in chapter 4, we identified DNA methylation changes in a subset of genes in TPC-1 six hours after irradiation. These changes in methylation occurred at both 62.5 mGy and 4 Gy and happened in the promoter as well as in intronic and exonic regions of the genes. The genes uncovered in our study include those involved in inflammation, senescence and even include biomarkers of cancer.

Our results indicate that a low dose of radiation (62.5 mGy) elicits a distinct transcriptional and epigenetic response in RET/PTC-positive and normal thyrocytes. In normal thyroids, low dose X-rays up-regulated cellular growth and proliferation pathways which could have consequences on thyroid cancer initiation later on. The presence of the RET/PTC translocation in cells seemed to suppress this effect, with a possible central role for p53, and even decreased signaling down those pathways in response to irradiation. At the same time, the RET/PTC-translocation protected cells from apoptosis possibly through induction of senescence at moderate to high doses of radiation. This could have an impact on radioiodine ablation after thyroidectomies in cases of PTC.

Finally we propose that our findings on the difference in behavior of RET/PTC-positive and normal thyroids to low doses of radiation may provide an insight into the long-term risk of low doses of radiation in that organ.

197

Samenvatting

198

Summary

Samenvatting De ramp in Tsjernobyl in april 1986 opende de ogen van de wereld voor de gevaren van kerncentrales. In de daaropvolgende jaren werd 'Tsjernobyl' het synoniem voor een buitengewone ramp en gingen vele geruchten de ronde over de omvang van de schade en het aantal slachtoffers. Uitgebreid onderzoek naar de gevolgen van de ramp in Tsjernobyl hebben de meeste geruchten ontkracht en het sterkste verband tussen het ongeval in Tsjernobyl en een potentieel risico voor de gezondheid blijkt de verhoging in het aantal gevallen van schildklierkanker bij kinderen te zijn. De vrijgave van grote hoeveelheden radioactief jodium in de atmosfeer na de ontploffing in reactor 4 is inderdaad in verband gebracht met het verschijnen van gevallen van schildklierkanker, en in het bijzonder papillair carcinoom

(papillary thyroid carcinoma; PTC), bij kinderen in de gebieden gelegen rond de reactor. Een groot percentage van deze kinderen zijn drager van een oncogene translocatie RET/PTC, welke in verband wordt gebracht met de initiatie van PTC. De bevolking in de omgeving van het ongeval werd onder andere ter hoogte van de schildklier blootgesteld aan een reeks verschillende dosissen straling, gaande van laag tot hoog. Verder wordt de blootstelling aan lage dosissen straling tijdens de kindertijd in verband gelegd met het voorkomen van PTC op latere leeftijd. De combinatie van deze bevindingen brachten ons tot een studie naar de effecten van lage dosis straling op enerzijds normale schildkliercellen en anderzijds schildkliercellen met een RET/PTC translocatie.

De effecten van lage dosis straling in zoogdiercellen blijft een controversieel onderwerp.

Recent werd het 'lineair non-threshold' (LNT) model onder vuur genomen omdat het niet accuraat zou zijn. In hoofdstuk 2 onderzoeken we de effecten van verschillende dosissen X- straling op twee PTC cellijnen die twee verschillende oncogene mutaties dragen en bovendien een verschillende p53 status hebben. Een bifasische respons op straling bij een dosis gelegen rond 0.5 Gy werd waargenomen in TPC-1 cellen met een RET/PTC1 translocatie en wild-type

199

Samenvatting

p53. Beneden deze dosis zijn er indicaties van celcyclus activatie en werd een daling in de fractie van actief delende cellen (cellen in S fase) waargenomen. Bij dosissen hoger dan 0.5

Gy was er een verband tussen de activatie van controlepunten in de celcyclus en stress- geïnduceerde celveroudering. Onder andere celveroudering geassocieerde β-galactosidase, een toename in de secretie van TGF-β1 en een gewijzigd secretieprofiel werden hier geobserveerd. Dit werd in verband gebracht met een toename in p53, p21 en p16 eiwitten.

Onze hypothese is dat deze verschillende respons op straling in deze cellen het gevolg is van de aard van DNA schade geïnduceerd afhankelijk van de dosissen X-straling. Bij een dosis van 62.5 mGy zijn er minder DNA dubbelstreng breuken waar te nemen en zijn deze bovendien minder complex dan bij 0.5 en 4 Gy. We stellen eveneens dat p53 cruciaal is voor dit fenomeen sinds BCPAP cellen, drager van een gemuteerde kopie van p53, geen tekenen van celveroudering vertonen en bovendien niet reageren op DNA schade geïnduceerd door straling met dosissen onder 0.1 Gy. De rol van p53 in het antwoord van TPC-1 cellen op straling werd verder onderzocht in hoofdstuk 3. Microarray analyse van TPC-1 cellen blootgesteld aan X-straling toont een toegenomen differentiële expressie aan van genen die betrokken zijn bij de p53 geregelde respons op dosissen van 0.5 en 4 Gy, maar niet van 62.5 mGy. We hebben eveneens ontdekt dat de p53 pathway wordt geactiveerd in de schildklier van transgene C57BL/6J muizen, die drager zijn van een RET/PTC3 translocatie, na blootstelling aan 4 Gy X-straling. TPC-1 cellen verschillen van schildkliercellen die de

RET/PTC3 translocatie dragen. TPC-1 is een gevestigde cellijn voor de studie van PTC, terwijl de schildkliercellen die de RET/PTC3 translocatie dragen tekenen van hyperplasie vertonen. Toch was de activatie van de p53 pathway afwezig in de schildklier van normale muizen na bestraling terwijl de TGF-β apoptosis pathway was toegenomen bij alle dosissen.

Interessant was dat vele pathways betrokken bij celgroei en proliferatie enorm waren verhoogd bij de lage dosis van 62.5 mGy. Dit werd eveneens bevestigd op eiwitniveau. Onder 200

Summary

deze pathways bevinden zich onder andere de mitogen activated protein kinase

(MAPK)/ERK1 pathway en de c-jun N-terminal kinase (JNK) pathway. Deze kunnen wijzen op een mechanisme waarmee tumoren enkel worden geïnduceerd in normale schildklieren ten gevolge van lage dosis straling. Er dienen echter meer experimenten uitgevoerd te worden met deze schildkliercellen, zoals een opvolging op lange termijn gepaard met een analyse van de histologie en microarrays.

In hoofdstuk 3 en 4 hebben we het epigenetisch effect van lage en hoge dosissen straling op

TPC-1 cellen onderzocht. We hebben een cohort van microRNA's (miRNA)geïdentificeerd die onafhankelijk van de dosis reageren op X-straling. Anderzijds reageerden miRNA's die betrokken zijn bij de regulatie van p21 en cyclin G1 enkel op een dosis van 4 Gy. Verder hebben we hebben miRNA's geïdentificeerd die enkel gereguleerd waren bij 62.5 mGy en 0.5

Gy. Bovendien hebben we in hoofdstuk 4, zes uur na bestraling, veranderingen in DNA methylatie in een subset van genen in TPC-1 geïdentificeerd. Deze veranderingen in methylatie waren te zien bij 62.5 mGy en 4 Gy en vond plaats in zowel de intron- en exongebieden van de genen. De genen ontdekt in onze studie zijn betrokken in onder andere inflammatie en celveroudering en enkele worden zelfs beschouwd als biomerker van kanker.

Onze resultaten tonen aan dat een lage dosis straling (62.5 mGy) een verschillende reactie kan opwekken op transcriptioneel en epigenetisch vlak in zowel RET/PTC-positieve en normale schildkliercellen. In normale schildklieren nam de expressie van pathways betrokken in celgroei en proliferatie toe na blootstelling aan lage dosissen X-stralen, wat een gevolg zou hebben op de latere ontwikkeling van schildklierkanker. De aanwezigheid van de RET/PTC translocatie in cellen schijnt dit effect te onderdrukken, met een mogelijk centrale rol voor p53. Er werd zelfs een afname van de signalisatie in deze pathways waargenomen na bestraling. Tegelijkertijd beschermt de RET/PTC translocatie de cellen tegen apoptosis waarschijnlijk door de inductie van celveroudering bij gemiddeld tot lage dosis straling. Dit 201

Samenvatting

zou een impact kunnen hebben op de verwijdering van radioactief jodium na thyroidectomies bij patiënten met PTC. Ten slotte, kunnen we stellen dat onze bevindingen, betrekking hebbende op de verschillen in gedrag tussen RET/PTC positieve en normale schildkliercellen na een lage dosis bestraling, voorziet een inzicht in de lang termijn risico van de lage dosis straling in de schildklier.

202

References

204

References

1. Smith, J.a.B., N.A., Chernobyl - Catastrophe and Consequences. 2005, Chichester: Praxis Publishing Ltd. 310. 2. Osif, B.A., Baratta, A.J., Conkling, T.W., TMI 25 Years Later: the Three Mile Island Nuclear Power Plant Accident and Its Impact. 1st ed. 2004, University Park: Pennsylvania State University Press. 158. 3. Butler, D., Radioactivity spreads in Japan. Nature, 2011. 471(7340): p. 555-6. 4. Castro, M.R., et al., Immunohistochemical analysis of sodium iodide symporter expression in metastatic differentiated thyroid cancer: correlation with radioiodine uptake. J Clin Endocrinol Metab, 2001. 86(11): p. 5627-32. 5. Christodouleas, J.P., et al., Short-Term and Long-Term Health Risks of Nuclear-Power-Plant Accidents. N Engl J Med, 2011. 6. Mould, R.F., Chernobyl Record: The Definitive History of the Chernobyl Catastrophe. 2000, Bristol: Institute of Physics Publishing. 402. 7. Goulko, G.M., et al., Thyroid dose and thyroid cancer incidence after the Chernobyl accident: assessments for the Zhytomyr region (Ukraine). Radiat Environ Biophys, 1998. 36(4): p. 261- 73. 8. U.S. Government, E.P.A.E., Nuclear Regulatory Commission (NRC), Soviet Union (USSR), 20th Century Nuclear Power Plant Accidents: 1986 Chernobyl Accident and Radioactive Release (Chornobyl Atomic Power Station) USSR, Health Consequences, Cesium, Iodine, Thyroid Cancer, Lessons: Progressive Management Publications. 9. Simon, S.L., et al., Radiation doses and cancer risks in the Marshall Islands associated with exposure to radioactive fallout from Bikini and Enewetak nuclear weapons tests: summary. Health Phys, 2010. 99(2): p. 105-23. 10. Takahashi, T., et al., The relationship of thyroid cancer with radiation exposure from nuclear weapon testing in the Marshall Islands. J Epidemiol, 2003. 13(2): p. 99-107. 11. Land, C.E., et al., Projected lifetime cancer risks from exposure to regional radioactive fallout in the Marshall Islands. Health Phys, 2010. 99(2): p. 201-15. 12. Brumfiel, G., The meltdown that wasn't. Nature, 2011. 471(7339): p. 417-8. 13. Butler, D., Fukushima health risks scrutinized. Nature, 2011. 472(7341): p. 13-4. 14. Tanimoto, T., et al., Safety of workers at the Fukushima Daiichi nuclear power plant. Lancet, 2011. 377(9776): p. 1489-90. 15. Imaizumi, M., et al., Radiation dose-response relationships for thyroid nodules and autoimmune thyroid diseases in Hiroshima and Nagasaki atomic bomb survivors 55-58 years after radiation exposure. JAMA, 2006. 295(9): p. 1011-22. 16. Imaizumi, M., et al., Thyroid diseases in atomic bomb survivors exposed in utero. J Clin Endocrinol Metab, 2008. 93(5): p. 1641-8. 17. Shimizu, Y., H. Kato, and W.J. Schull, Risk of cancer among atomic bomb survivors. J Radiat Res (Tokyo), 1991. 32 Suppl 2: p. 54-63. 18. Sadetzki, S., et al., Risk of thyroid cancer after childhood exposure to ionizing radiation for tinea capitis. J Clin Endocrinol Metab, 2006. 91(12): p. 4798-804. 19. Sadetzki, S. and L. Mandelzweig, Childhood exposure to external ionising radiation and solid cancer risk. Br J Cancer, 2009. 100(7): p. 1021-5. 20. Schafer, D.W., et al., Thyroid cancer following scalp irradiation: a reanalysis accounting for uncertainty in dosimetry. Biometrics, 2001. 57(3): p. 689-97. 21. Lubin, J.H., et al., A reanalysis of thyroid neoplasms in the Israeli tinea capitis study accounting for dose uncertainties. Radiat Res, 2004. 161(3): p. 359-68. 22. Ron, E. and A. Brenner, Non-Malignant Thyroid Diseases after a Wide Range of Radiation Exposures. Radiat Res, 2010. 205

References

23. Norris, D.O., The Hypothalamus-Pituitary-Thyroid (HPT) Axis of Mammals, in Vertebrate Endocrinology, D.O. Norris, Editor. 2007, Elsevier Academic Press. p. 221-242. 24. Miot, F., Dupuy, C, Dumont, J.E., Rousset, B.A., Thyroid Hormone Synthesis and Secretion, in The Thyroid and its Diseases, L.J. De Groot, Editor. 1984, John Wiley and Sons Inc. p. 907. 25. Bizhanova, A. and P. Kopp, Minireview: The sodium-iodide symporter NIS and pendrin in iodide homeostasis of the thyroid. Endocrinology, 2009. 150(3): p. 1084-90. 26. Fong, P., Thyroid iodide efflux: a team effort? J Physiol, 2011. 27. De Deken, X., et al., Cloning of two human thyroid cDNAs encoding new members of the NADPH oxidase family. J Biol Chem, 2000. 275(30): p. 23227-33. 28. Dupuy, C., et al., Thyroid oxidase (THOX2) gene expression in the rat thyroid cell line FRTL-5. Biochem Biophys Res Commun, 2000. 277(2): p. 287-92. 29. Mano, T., et al., Changes in free radical scavengers and lipid peroxide in thyroid glands of various thyroid disorders. Horm Metab Res, 1997. 29(7): p. 351-4. 30. Gerard, A.C., et al., Peroxiredoxin 5 expression in the human thyroid gland. Thyroid, 2005. 15(3): p. 205-9. 31. Kim, H., et al., Role of peroxiredoxins in regulating intracellular hydrogen peroxide and hydrogen peroxide-induced apoptosis in thyroid cells. J Biol Chem, 2000. 275(24): p. 18266- 70. 32. Poncin, S., et al., Oxidative stress in the thyroid gland: from harmlessness to hazard depending on the iodine content. Endocrinology, 2008. 149(1): p. 424-33. 33. Poncin, S., et al., Oxidative stress: a required condition for thyroid cell proliferation. Am J Pathol, 2010. 176(3): p. 1355-63. 34. Dunn, J.T. and A.D. Dunn, Update on intrathyroidal iodine metabolism. Thyroid, 2001. 11(5): p. 407-14. 35. Di Cosmo, C., et al., Mice deficient in MCT8 reveal a mechanism regulating thyroid hormone secretion. J Clin Invest, 2011. 120(9): p. 3377-88. 36. Friesema, E.C., et al., Thyroid hormone transport by the human monocarboxylate transporter 8 and its rate-limiting role in intracellular metabolism. Mol Endocrinol, 2006. 20(11): p. 2761- 72. 37. Gnidehou, S., et al., Iodotyrosine dehalogenase 1 (DEHAL1) is a transmembrane protein involved in the recycling of iodide close to the thyroglobulin iodination site. FASEB J, 2004. 18(13): p. 1574-6. 38. Brix, K., D. Fuhrer, and H. Biebermann, Molecules important for thyroid hormone synthesis and action - known facts and future perspectives. Thyroid Res, 2011. 4 Suppl 1: p. S9. 39. Krause, K., et al., Characterisation of DEHAL1 expression in thyroid pathologies. Eur J Endocrinol, 2007. 156(3): p. 295-301. 40. Moreno, J.C. and T.J. Visser, Genetics and phenomics of hypothyroidism and goiter due to iodotyrosine deiodinase (DEHAL1) gene mutations. Mol Cell Endocrinol, 2010. 322(1-2): p. 91- 8. 41. Zavacki, A.M., et al., Type 1 iodothyronine deiodinase is a sensitive marker of peripheral thyroid status in the mouse. Endocrinology, 2005. 146(3): p. 1568-75. 42. Arjona, F.J., et al., Identification and Functional Characterization of Zebrafish Solute Carrier Slc16a2 (Mct8) as a Thyroid Hormone Membrane Transporter. Endocrinology, 2011. 43. Kleinau, G., et al., Insights into molecular properties of the human monocarboxylate transporter 8 by combining functional with structural information. Thyroid Res, 2011. 4 Suppl 1: p. S4. 44. Saito, T., et al., Increased expression of the sodium/iodide symporter in papillary thyroid carcinomas. J Clin Invest, 1998. 101(7): p. 1296-300. 45. Chung, J.K., Sodium iodide symporter: its role in nuclear medicine. J Nucl Med, 2002. 43(9): p. 1188-200.

206

References

46. Riedel, C., O. Levy, and N. Carrasco, Post-transcriptional regulation of the sodium/iodide symporter by thyrotropin. J Biol Chem, 2001. 276(24): p. 21458-63. 47. Serrano-Nascimento, C., J. Calil-Silveira, and M.T. Nunes, Posttranscriptional regulation of sodium-iodide symporter mRNA expression in the rat thyroid gland by acute iodide administration. Am J Physiol Cell Physiol, 2010. 298(4): p. C893-9. 48. Eggo, M.C., Molecular regulation of thyroid gland function. Curr Opin Endocrinol Diabetes Obes, 2010. 17(5): p. 396-401. 49. Fortunato, R.S., et al., Functional consequences of dual oxidase-thyroperoxidase interaction at the plasma membrane. J Clin Endocrinol Metab, 2010. 95(12): p. 5403-11. 50. Görges, R., The Changing Epidemiology of Thryoid Cancer, in Thyroid Cancer, H.-J. Biersack, Grünwald, F, Editor. 2005, Springer. p. 363. 51. Hundahl, S.A., et al., A National Cancer Data Base report on 53,856 cases of thyroid carcinoma treated in the U.S., 1985-1995 [see commetns]. Cancer, 1998. 83(12): p. 2638-48. 52. Franceschi, S. and C. La Vecchia, Thyroid cancer. Cancer Surv, 1994. 19-20: p. 393-422. 53. Zeng, Q., et al., The contributions of oestrogen receptor isoforms to the development of papillary and anaplastic thyroid carcinomas. J Pathol, 2008. 214(4): p. 425-33. 54. Antico-Arciuch, V.G., et al., Cross-talk between PI3K and estrogen in the mouse thyroid predisposes to the development of follicular carcinomas with a higher incidence in females. Oncogene, 2011. 29(42): p. 5678-86. 55. Kavanagh, D.O., et al., The role of oestrogen receptor {alpha} in human thyroid cancer: contributions from coregulatory proteins and the tyrosine kinase receptor HER2. Endocr Relat Cancer, 2010. 17(1): p. 255-64. 56. Hofstädter, F., Histopathology, Immunohistochemistry, and Molecular Biology, in Thyroid Cancer, H.-J. Biersack, Grünwald, F, Editor. 2005, Springer. p. 363. 57. Asa, S.L., The Pathology of Thyroid Cancer, in Molecular Basis of Thyroid Cancer, S.T. Rosen, Editor. 2004, Kluwer Academic Publishers. p. 444. 58. Weber, F., et al., Genetic classification of benign and malignant thyroid follicular neoplasia based on a three-gene combination. J Clin Endocrinol Metab, 2005. 90(5): p. 2512-21. 59. Shibru, D., et al., Does the 3-gene diagnostic assay accurately distinguish benign from malignant thyroid neoplasms? Cancer, 2008. 113(5): p. 930-5. 60. Fluge, O., et al., Gene expression in poorly differentiated papillary thyroid carcinomas. Thyroid, 2006. 16(2): p. 161-75. 61. Huang, Y., et al., Gene expression in papillary thyroid carcinoma reveals highly consistent profiles. Proc Natl Acad Sci U S A, 2001. 98(26): p. 15044-9. 62. Jarzab, B., et al., Gene expression profile of papillary thyroid cancer: sources of variability and diagnostic implications. Cancer Res, 2005. 65(4): p. 1587-97. 63. Negri, E., et al., A pooled analysis of case-control studies of thyroid cancer. I. Methods. Cancer Causes Control, 1999. 10(2): p. 131-42. 64. Hanahan, D. and R.A. Weinberg, The hallmarks of cancer. Cell, 2000. 100(1): p. 57-70. 65. Hanahan, D. and R.A. Weinberg, Hallmarks of cancer: the next generation. Cell, 2011. 144(5): p. 646-74. 66. Kroll, T.G., Molecular Events in Follicular Thyroid Tumors, in Molecular Basis of Thyroid Cancer, S.T. Rosen, Editor. 2004, Kluwer Academic Publishers. p. 444. 67. Davies, H., et al., Mutations of the BRAF gene in human cancer. Nature, 2002. 417(6892): p. 949-54. 68. Soares, P., et al., BRAF mutations and RET/PTC rearrangements are alternative events in the etiopathogenesis of PTC. Oncogene, 2003. 22(29): p. 4578-80. 69. Cohen, Y., et al., BRAF mutation in papillary thyroid carcinoma. J Natl Cancer Inst, 2003. 95(8): p. 625-7.

207

References

70. Fukushima, T., et al., BRAF mutations in papillary carcinomas of the thyroid. Oncogene, 2003. 22(41): p. 6455-7. 71. Kimura, E.T., et al., High prevalence of BRAF mutations in thyroid cancer: genetic evidence for constitutive activation of the RET/PTC-RAS-BRAF signaling pathway in papillary thyroid carcinoma. Cancer Res, 2003. 63(7): p. 1454-7. 72. Namba, H., et al., Clinical implication of hot spot BRAF mutation, V599E, in papillary thyroid cancers. J Clin Endocrinol Metab, 2003. 88(9): p. 4393-7. 73. Xu, X., et al., High prevalence of BRAF gene mutation in papillary thyroid carcinomas and thyroid tumor cell lines. Cancer Res, 2003. 63(15): p. 4561-7. 74. Salvatore, G., et al., Analysis of BRAF point mutation and RET/PTC rearrangement refines the fine-needle aspiration diagnosis of papillary thyroid carcinoma. J Clin Endocrinol Metab, 2004. 89(10): p. 5175-80. 75. Ouyang, B., et al., Inhibitors of Raf kinase activity block growth of thyroid cancer cells with RET/PTC or BRAF mutations in vitro and in vivo. Clin Cancer Res, 2006. 12(6): p. 1785-93. 76. DeLellis, R.A., Pathology and genetics of thyroid carcinoma. J Surg Oncol, 2006. 94(8): p. 662- 9. 77. Arighi, E., M.G. Borrello, and H. Sariola, RET tyrosine kinase signaling in development and cancer. Cytokine Growth Factor Rev, 2005. 16(4-5): p. 441-67. 78. Wells, S.A., Jr. and M. Santoro, Targeting the RET pathway in thyroid cancer. Clin Cancer Res, 2009. 15(23): p. 7119-23. 79. Grieco, M., et al., PTC is a novel rearranged form of the ret proto-oncogene and is frequently detected in vivo in human thyroid papillary carcinomas. Cell, 1990. 60(4): p. 557-63. 80. Santoro, M., R.M. Melillo, and A. Fusco, RET/PTC activation in papillary thyroid carcinoma: European Journal of Endocrinology Prize Lecture. Eur J Endocrinol, 2006. 155(5): p. 645-53. 81. Santoro, M., et al., Molecular characterization of RET/PTC3; a novel rearranged version of the RETproto-oncogene in a human thyroid papillary carcinoma. Oncogene, 1994. 9(2): p. 509-16. 82. Kim, D.S., et al., Oncogenes in thyroid cancer. Clin Otolaryngol Allied Sci, 2003. 28(5): p. 386- 95. 83. Besset, V., R.P. Scott, and C.F. Ibanez, Signaling complexes and protein-protein interactions involved in the activation of the Ras and phosphatidylinositol 3-kinase pathways by the c-Ret receptor tyrosine kinase. J Biol Chem, 2000. 275(50): p. 39159-66. 84. Chiariello, M., et al., Signalling of the Ret receptor tyrosine kinase through the c-Jun NH2- terminal protein kinases (JNKS): evidence for a divergence of the ERKs and JNKs pathways induced by Ret. Oncogene, 1998. 16(19): p. 2435-45. 85. Segouffin-Cariou, C. and M. Billaud, Transforming ability of MEN2A-RET requires activation of the phosphatidylinositol 3-kinase/AKT signaling pathway. J Biol Chem, 2000. 275(5): p. 3568- 76. 86. Kawamoto, Y., et al., Identification of RET autophosphorylation sites by mass spectrometry. J Biol Chem, 2004. 279(14): p. 14213-24. 87. Knostman, K.A., S.M. Jhiang, and C.C. Capen, Genetic alterations in thyroid cancer: the role of mouse models. Vet Pathol, 2007. 44(1): p. 1-14. 88. Fagin, J.A., Genetics of papillary thyroid cancer initiation: implications for therapy. Trans Am Clin Climatol Assoc, 2005. 116: p. 259-69; discussion 269-71. 89. Cardis, E., et al., Cancer consequences of the Chernobyl accident: 20 years on. J Radiol Prot, 2006. 26(2): p. 127-40. 90. Hatch, M., et al., The Chernobyl disaster: cancer following the accident at the Chernobyl nuclear power plant. Epidemiol Rev, 2005. 27: p. 56-66. 91. Mizuno, T., et al., Preferential induction of RET/PTC1 rearrangement by X-ray irradiation. Oncogene, 2000. 19(3): p. 438-43.

208

References

92. Santoro, M., et al., Gene rearrangement and Chernobyl related thyroid cancers. Br J Cancer, 2000. 82(2): p. 315-22. 93. Gandhi, M., et al., DNA breaks at fragile sites generate oncogenic RET/PTC rearrangements in human thyroid cells. Oncogene, 2010. 29(15): p. 2272-80. 94. Powell, D.J., Jr., et al., The RET/PTC3 oncogene: metastatic solid-type papillary carcinomas in murine thyroids. Cancer Res, 1998. 58(23): p. 5523-8. 95. Caudill, C.M., et al., Dose-dependent generation of RET/PTC in human thyroid cells after in vitro exposure to gamma-radiation: a model of carcinogenic chromosomal rearrangement induced by ionizing radiation. J Clin Endocrinol Metab, 2005. 90(4): p. 2364-9. 96. Sadetzki, S., et al., Ret/PTC activation in benign and malignant thyroid tumors arising in a population exposed to low-dose external-beam irradiation in childhood. J Clin Endocrinol Metab, 2004. 89(5): p. 2281-9. 97. Port, M., et al., A radiation-induced gene signature distinguishes post-Chernobyl from sporadic papillary thyroid cancers. Radiat Res, 2007. 168(6): p. 639-49. 98. Detours, V., et al., Genome-wide gene expression profiling suggests distinct radiation susceptibilities in sporadic and post-Chernobyl papillary thyroid cancers. Br J Cancer, 2007. 97(6): p. 818-25. 99. Detours, V., et al., Gene expression profiles of post-Chernobyl thyroid cancers. Curr Opin Endocrinol Diabetes Obes, 2008. 15(5): p. 440-5. 100. Assaad, A., L. Voeghtly, and J.L. Hunt, Thyroidectomies from patients with history of therapeutic radiation during childhood and adolescence have a unique mutational profile. Mod Pathol, 2008. 21(9): p. 1176-82. 101. Caria, P. and R. Vanni, Cytogenetic and molecular events in adenoma and well-differentiated thyroid follicular-cell neoplasia. Cancer Genet Cytogenet, 2011. 203(1): p. 21-9. 102. Russell, J.P., et al., The TRK-T1 fusion protein induces neoplastic transformation of thyroid epithelium. Oncogene, 2000. 19(50): p. 5729-35. 103. Melillo, R.M., et al., The RET/PTC-RAS-BRAF linear signaling cascade mediates the motile and mitogenic phenotype of thyroid cancer cells. J Clin Invest, 2005. 115(4): p. 1068-81. 104. Mullenders, L., et al., Assessing cancer risks of low-dose radiation. Nat Rev Cancer, 2009. 9(8): p. 596-604. 105. Tubiana, M., et al., The linear no-threshold relationship is inconsistent with radiation biologic and experimental data. Radiology, 2009. 251(1): p. 13-22. 106. Pierce, D.A. and D.L. Preston, Radiation-related cancer risks at low doses among atomic bomb survivors. Radiat Res, 2000. 154(2): p. 178-86. 107. Hall, E.J., Is there a place for quantitative risk assessment? J Radiol Prot, 2009. 29(2A): p. A171-84. 108. Brenner, D.J., Extrapolating radiation-induced cancer risks from low doses to very low doses. Health Phys, 2009. 97(5): p. 505-9. 109. Preston, D.L., et al., Studies of mortality of atomic bomb survivors. Report 13: Solid cancer and noncancer disease mortality: 1950-1997. Radiat Res, 2003. 160(4): p. 381-407. 110. Little, M.P., et al., Risks associated with low doses and low dose rates of ionizing radiation: why linearity may be (almost) the best we can do. Radiology, 2009. 251(1): p. 6-12. 111. Martin, C.J., et al., The radiobiology/radiation protection interface in healthcare. J Radiol Prot, 2009. 29(2A): p. A1-A20. 112. Bartley, K., et al., Diagnostic X-rays and risk of childhood leukaemia. Int J Epidemiol, 2011. 39(6): p. 1628-37. 113. Kesminiene, A., et al., Risk of hematological malignancies among Chernobyl liquidators. Radiat Res, 2008. 170(6): p. 721-35. 114. Cardis, E., et al., Risk of cancer after low doses of ionising radiation: retrospective cohort study in 15 countries. BMJ, 2005. 331(7508): p. 77.

209

References

115. Vrijheid, M., et al., Ionizing radiation and risk of chronic lymphocytic leukemia in the 15- country study of nuclear industry workers. Radiat Res, 2008. 170(5): p. 661-5. 116. Berrington, A., et al., 100 years of observation on British radiologists: mortality from cancer and other causes 1897-1997. Br J Radiol, 2001. 74(882): p. 507-19. 117. Darby, S., et al., Residential radon and lung cancer--detailed results of a collaborative analysis of individual data on 7148 persons with lung cancer and 14,208 persons without lung cancer from 13 epidemiologic studies in Europe. Scand J Work Environ Health, 2006. 32 Suppl 1: p. 1- 83. 118. Mohan, A.K., et al., Cancer and other causes of mortality among radiologic technologists in the United States. Int J Cancer, 2003. 103(2): p. 259-67. 119. Averbeck, D., Does scientific evidence support a change from the LNT model for low-dose radiation risk extrapolation? Health Phys, 2009. 97(5): p. 493-504. 120. Kinner, A., et al., Gamma-H2AX in recognition and signaling of DNA double-strand breaks in the context of chromatin. Nucleic Acids Res, 2008. 36(17): p. 5678-94. 121. Mistrik, M., et al., Low-dose DNA damage and replication stress responses quantified by optimized automated single-cell image analysis. Cell Cycle, 2009. 8(16): p. 2592-9. 122. De Vos, W.H., et al., High content image cytometry in the context of subnuclear organization. Cytometry A, 2010. 77(1): p. 64-75. 123. Lobrich, M., et al., In vivo formation and repair of DNA double-strand breaks after computed tomography examinations. Proc Natl Acad Sci U S A, 2005. 102(25): p. 8984-9. 124. Rothkamm, K. and M. Lobrich, Evidence for a lack of DNA double-strand break repair in human cells exposed to very low x-ray doses. Proc Natl Acad Sci U S A, 2003. 100(9): p. 5057- 62. 125. Beels, L., et al., gamma-H2AX foci as a biomarker for patient X-ray exposure in pediatric cardiac catheterization: are we underestimating radiation risks? Circulation, 2009. 120(19): p. 1903-9. 126. Lassmann, M., et al., In vivo formation of gamma-H2AX and 53BP1 DNA repair foci in blood cells after radioiodine therapy of differentiated thyroid cancer. J Nucl Med, 2010. 51(8): p. 1318-25. 127. Rave-Frank, M., et al., In vitro response of human dermal fibroblasts to X-irradiation: relationship between radiation-induced clonogenic cell death, chromosome aberrations and markers of proliferative senescence or differentiation. Int J Radiat Biol, 2001. 77(12): p. 1163- 74. 128. Huber, R., et al., Dose-response relationships of micronuclei in human lymphocytes induced by fission neutrons and by low LET radiations. Mutat Res, 1994. 306(2): p. 135-41. 129. Lloyd, D.C., et al., Chromosomal aberrations in human lymphocytes induced in vitro by very low doses of X-rays. Int J Radiat Biol, 1992. 61(3): p. 335-43. 130. Bache, M., et al., Loss of G2/M arrest correlates with radiosensitization in two human sarcoma cell lines with mutant p53. Int J Cancer, 2001. 96(2): p. 110-7. 131. Joiner, M.C., et al., Low-dose hypersensitivity: current status and possible mechanisms. Int J Radiat Oncol Biol Phys, 2001. 49(2): p. 379-89. 132. Nasonova, E.A., et al., Cytogenetic effects of low-dose radiation with different LET in human peripheral blood lymphocytes. Radiat Environ Biophys, 2006. 45(4): p. 307-12. 133. Short, S.C., et al., DNA damage responses at low radiation doses. Radiat Res, 2005. 164(3): p. 292-302. 134. Enns, L., et al., Low-dose radiation hypersensitivity is associated with p53-dependent apoptosis. Mol Cancer Res, 2004. 2(10): p. 557-66. 135. Krueger, S.A., et al., Role of apoptosis in low-dose hyper-radiosensitivity. Radiat Res, 2007. 167(3): p. 260-7.

210

References

136. Skov, K.A., Radioresponsiveness at low doses: hyper-radiosensitivity and increased radioresistance in mammalian cells. Mutat Res, 1999. 430(2): p. 241-53. 137. Mothersill, C. and C. Seymour, Radiation-induced bystander effects: past history and future directions. Radiat Res, 2001. 155(6): p. 759-67. 138. Mothersill, C. and C. Seymour, Radiation-induced bystander effects and adaptive responses-- the Yin and Yang of low dose radiobiology? Mutat Res, 2004. 568(1): p. 121-8. 139. Smith, R.W., et al., Evidence for a protective response by the gill proteome of rainbow trout exposed to X-ray induced bystander signals. Proteomics, 2007. 7(22): p. 4171-80. 140. Dieriks, B., et al., High content analysis of human fibroblast cell cultures after exposure to space radiation. Radiat Res, 2009. 172(4): p. 423-36. 141. Dieriks, B., et al., Medium-mediated DNA repair response after ionizing radiation is correlated with the increase of specific cytokines in human fibroblasts. Mutat Res, 2011. 687(1-2): p. 40- 8. 142. Bonner, W.M., Low-dose radiation: thresholds, bystander effects, and adaptive responses. Proc Natl Acad Sci U S A, 2003. 100(9): p. 4973-5. 143. Chaudhry, M.A., Bystander effect: biological endpoints and microarray analysis. Mutat Res, 2006. 597(1-2): p. 98-112. 144. Tapio, S. and V. Jacob, Radioadaptive response revisited. Radiat Environ Biophys, 2007. 46(1): p. 1-12. 145. Jiang, H., et al., Low-dose radiation induces adaptive response in normal cells, but not in tumor cells: in vitro and in vivo studies. J Radiat Res (Tokyo), 2008. 49(3): p. 219-30. 146. Ahmed, K.M. and J.J. Li, NF-kappa B-mediated adaptive resistance to ionizing radiation. Free Radic Biol Med, 2008. 44(1): p. 1-13. 147. Ahmed, K.M., et al., Coactivation of ATM/ERK/NF-kappaB in the low-dose radiation-induced radioadaptive response in human skin keratinocytes. Free Radic Biol Med, 2009. 46(11): p. 1543-50. 148. Ahmed, K.M., et al., Cyclin D1 in low-dose radiation-induced adaptive resistance. Oncogene, 2008. 27(53): p. 6738-48. 149. Grudzenski, S., et al., Inducible response required for repair of low-dose radiation damage in human fibroblasts. Proc Natl Acad Sci U S A, 2010. 107(32): p. 14205-10. 150. Mitchel, R.E., Low doses of radiation are protective in vitro and in vivo: evolutionary origins. Dose Response, 2006. 4(2): p. 75-90. 151. Mitchel, R.E., Low doses of radiation reduce risk in vivo. Dose Response, 2007. 5(1): p. 1-10. 152. Sankaranarayanan, K. and R. Chakraborty, Cancer predisposition, radiosensitivity and the risk of radiation-induced cancers. I. Background. Radiat Res, 1995. 143(2): p. 121-43. 153. Bourguignon, M.H., et al., Genetic and epigenetic features in radiation sensitivity. Part II: implications for clinical practice and radiation protection. Eur J Nucl Med Mol Imaging, 2005. 32(3): p. 351-68. 154. Bourguignon, M.H., et al., Genetic and epigenetic features in radiation sensitivity Part I: cell signalling in radiation response. Eur J Nucl Med Mol Imaging, 2005. 32(2): p. 229-46. 155. Trosko, J.E., et al., Low-dose ionizing radiation: induction of differential intracellular signalling possibly affecting intercellular communication. Radiat Environ Biophys, 2005. 44(1): p. 3-9. 156. Cuttler, J., Can Cancer Be Treated with Low Doses of Radiation? Journal of American Physicians and Surgeons, 2003. 8(4): p. 4. 157. Haas, R.L., Low dose radiotherapy in indolent lymphomas, enough is enough. Hematol Oncol, 2009. 27(2): p. 71-81. 158. Mallick, U., C. Harmer, and A. Hackshaw, The HiLo trial: a multicentre randomised trial of high- versus low-dose radioiodine, with or without recombinant human thyroid stimulating hormone, for remnant ablation after surgery for differentiated thyroid cancer. Clin Oncol (R Coll Radiol), 2008. 20(5): p. 325-6.

211

References

159. Martin, N.E. and A.K. Ng, Good things come in small packages: low-dose radiation as palliation for indolent non-Hodgkin lymphomas. Leuk Lymphoma, 2009. 50(11): p. 1765-72. 160. Puri, D.R., W. Tereffe, and J. Yahalom, Low-dose and limited-volume radiotherapy alone for primary dural marginal zone lymphoma: treatment approach and review of published data. Int J Radiat Oncol Biol Phys, 2008. 71(5): p. 1425-35. 161. Lundell, M., T. Hakulinen, and L.E. Holm, Thyroid cancer after radiotherapy for skin hemangioma in infancy. Radiat Res, 1994. 140(3): p. 334-9. 162. Ron, E., et al., Thyroid cancer after exposure to external radiation: a pooled analysis of seven studies. Radiat Res, 1995. 141(3): p. 259-77. 163. Tucker, M.A., et al., Therapeutic radiation at a young age is linked to secondary thyroid cancer. The Late Effects Study Group. Cancer Res, 1991. 51(11): p. 2885-8. 164. Schneider, A.B., et al., Dose-response relationships for radiation-induced thyroid cancer and thyroid nodules: evidence for the prolonged effects of radiation on the thyroid. J Clin Endocrinol Metab, 1993. 77(2): p. 362-9. 165. Boltze, C., et al., Radiation-induced thyroid carcinogenesis as a function of time and dietary iodine supply: an in vivo model of tumorigenesis in the rat. Endocrinology, 2002. 143(7): p. 2584-92. 166. Yusuf, I. and D.A. Fruman, Regulation of quiescence in lymphocytes. Trends Immunol, 2003. 24(7): p. 380-6. 167. Tapia, N. and H.R. Schöler, p53 connects tumorigenesis and reprogramming to pluripotency. J Exp Med, 2010. 207(10): p. 2045-8. 168. Beausejour, C.M., et al., Reversal of human cellular senescence: roles of the p53 and p16 pathways. EMBO J, 2003. 22(16): p. 4212-22. 169. Dirac, A.M. and R. Bernards, Reversal of senescence in mouse fibroblasts through lentiviral suppression of p53. J Biol Chem, 2003. 278(14): p. 11731-4. 170. Hayflick, L., The Limited in Vitro Lifetime of Human Diploid Cell Strains. Exp Cell Res, 1965. 37: p. 614-36. 171. Hayflick, L. and P.S. Moorhead, The serial cultivation of human diploid cell strains. Exp Cell Res, 1961. 25: p. 585-621. 172. Shay, J.W. and W.E. Wright, Hayflick, his limit, and cellular ageing. Nat Rev Mol Cell Biol, 2000. 1(1): p. 72-6. 173. Ewald, J.A., et al., Therapy-induced senescence in cancer. J Natl Cancer Inst, 2010. 102(20): p. 1536-46. 174. Villalonga-Planells, R., et al., Activation of p53 by nutlin-3a induces apoptosis and cellular senescence in human glioblastoma multiforme. PLoS One, 2011. 6(4): p. e18588. 175. Orjalo, A.V., et al., Cell surface-bound IL-1alpha is an upstream regulator of the senescence- associated IL-6/IL-8 cytokine network. Proc Natl Acad Sci U S A, 2009. 106(40): p. 17031-6. 176. Acosta, J.C., et al., Chemokine signaling via the CXCR2 receptor reinforces senescence. Cell, 2008. 133(6): p. 1006-18. 177. Liu, D. and P.J. Hornsby, Senescent human fibroblasts increase the early growth of xenograft tumors via matrix metalloproteinase secretion. Cancer Res, 2007. 67(7): p. 3117-26. 178. Tsai, K.K., et al., Low-dose radiation-induced senescent stromal fibroblasts render nearby breast cancer cells radioresistant. Radiat Res, 2009. 172(3): p. 306-13. 179. Kuilman, T. and D.S. Peeper, Senescence-messaging secretome: SMS-ing cellular stress. Nat Rev Cancer, 2009. 9(2): p. 81-94. 180. Jeyapalan, J.C., et al., Accumulation of senescent cells in mitotic tissue of aging primates. Mech Ageing Dev, 2007. 128(1): p. 36-44. 181. Xue, W., et al., Senescence and tumour clearance is triggered by p53 restoration in murine liver carcinomas. Nature, 2007. 445(7128): p. 656-60.

212

References

182. Collado, M., et al., Tumour biology: senescence in premalignant tumours. Nature, 2005. 436(7051): p. 642. 183. Blagosklonny, M.V., Cell cycle arrest is not senescence. Aging (Albany NY), 2011. 3(2): p. 94- 101. 184. Sherwood, S.W., et al., Defining cellular senescence in IMR-90 cells: a flow cytometric analysis. Proc Natl Acad Sci U S A, 1988. 85(23): p. 9086-90. 185. Zahnreich, S., et al., Radiation-induced premature senescence is associated with specific cytogenetic changes. Mutat Res, 2010. 701(1): p. 60-6. 186. Dimri, G.P., et al., A biomarker that identifies senescent human cells in culture and in aging skin in vivo. Proc Natl Acad Sci U S A, 1995. 92(20): p. 9363-7. 187. Cristofalo, V.J., SA beta Gal staining: biomarker or delusion. Exp Gerontol, 2005. 40(10): p. 836-8. 188. Severino, J., et al., Is beta-galactosidase staining a marker of senescence in vitro and in vivo? Exp Cell Res, 2000. 257(1): p. 162-71. 189. Yang, N.C. and M.L. Hu, The limitations and validities of senescence associated-beta- galactosidase activity as an aging marker for human foreskin fibroblast Hs68 cells. Exp Gerontol, 2005. 40(10): p. 813-9. 190. Rodier, F., et al., Persistent DNA damage signalling triggers senescence-associated inflammatory cytokine secretion. Nat Cell Biol, 2009. 11(8): p. 973-9. 191. Coppe, J.P., et al., Senescence-associated secretory phenotypes reveal cell-nonautonomous functions of oncogenic RAS and the p53 tumor suppressor. PLoS Biol, 2008. 6(12): p. 2853-68. 192. Novakova, Z., et al., Cytokine expression and signaling in drug-induced cellular senescence. Oncogene, 2010. 29(2): p. 273-84. 193. Tsirpanlis, G., Cellular senescence and inflammation: a noteworthy link. Blood Purif, 2009. 28(1): p. 12-4. 194. Cichowski, K. and W.C. Hahn, Unexpected pieces to the senescence puzzle. Cell, 2008. 133(6): p. 958-61. 195. Gartel, A.L., p21(WAF1/CIP1) and cancer: a shifting paradigm? Biofactors, 2009. 35(2): p. 161-4. 196. Abbas, T. and A. Dutta, p21 in cancer: intricate networks and multiple activities. Nat Rev Cancer, 2009. 9(6): p. 400-14. 197. Stein, G.H., et al., Differential roles for cyclin-dependent kinase inhibitors p21 and p16 in the mechanisms of senescence and differentiation in human fibroblasts. Mol Cell Biol, 1999. 19(3): p. 2109-17. 198. Haferkamp, S., et al., The relative contributions of the p53 and pRb pathways in oncogene- induced melanocyte senescence. Aging (Albany NY), 2009. 1(6): p. 542-56. 199. Schmitt, C.A., et al., A senescence program controlled by p53 and p16INK4a contributes to the outcome of cancer therapy. Cell, 2002. 109(3): p. 335-46. 200. Abd Elmageed, Z.Y., et al., Characterization of coordinated immediate responses by p16INK4A and p53 pathways in UVB-irradiated human skin cells. J Invest Dermatol, 2009. 129(1): p. 175-83. 201. Al-Mohanna, M.A., et al., The p16INK4a tumor suppressor controls p21WAF1 induction in response to ultraviolet light. Nucleic Acids Res, 2007. 35(1): p. 223-33. 202. Udayakumar, T., et al., The E2F1/Rb and p53/MDM2 pathways in DNA repair and apoptosis: understanding the crosstalk to develop novel strategies for prostate cancer radiotherapy. Semin Radiat Oncol, 2010. 20(4): p. 258-66. 203. Leong, W.F., J.F. Chau, and B. Li, p53 Deficiency leads to compensatory up-regulation of p16INK4a. Mol Cancer Res, 2009. 7(3): p. 354-60. 204. Yamakoshi, K., et al., Real-time in vivo imaging of p16Ink4a reveals cross talk with p53. J Cell Biol, 2009. 186(3): p. 393-407.

213

References

205. Mirzayans, R., et al., Single-cell analysis of p16(INK4a) and p21(WAF1) expression suggests distinct mechanisms of senescence in normal human and Li-Fraumeni Syndrome fibroblasts. J Cell Physiol, 2010. 223(1): p. 57-67. 206. Calabro, V., et al., Status and expression of the p16INK4 gene in human thyroid tumors and thyroid-tumor cell lines. Int J Cancer, 1996. 67(1): p. 29-34. 207. Shay, J.W. and I.B. Roninson, Hallmarks of senescence in carcinogenesis and cancer therapy. Oncogene, 2004. 23(16): p. 2919-33. 208. Roninson, I.B., Tumor cell senescence in cancer treatment. Cancer Res, 2003. 63(11): p. 2705- 15. 209. Pascal, T., et al., Comparison of replicative senescence and stress-induced premature senescence combining differential display and low-density DNA arrays. FEBS Lett, 2005. 579(17): p. 3651-9. 210. Shelton, D.N., et al., Microarray analysis of replicative senescence. Curr Biol, 1999. 9(17): p. 939-45. 211. Kis, E., et al., Microarray analysis of radiation response genes in primary human fibroblasts. Int J Radiat Oncol Biol Phys, 2006. 66(5): p. 1506-14. 212. Zhou, T., et al., Profiles of global gene expression in ionizing-radiation-damaged human diploid fibroblasts reveal synchronization behind the G1 checkpoint in a G0-like state of quiescence. Environ Health Perspect, 2006. 114(4): p. 553-9. 213. Huck, J.J., et al., MLN8054, an inhibitor of Aurora A kinase, induces senescence in human tumor cells both in vitro and in vivo. Mol Cancer Res, 2010. 8(3): p. 373-84. 214. Zhang, D., et al., Aurora A overexpression induces cellular senescence in mammary gland hyperplastic tumors developed in p53-deficient mice. Oncogene, 2008. 27(31): p. 4305-14. 215. Elmore, L.W., et al., Evasion of a single-step, chemotherapy-induced senescence in breast cancer cells: implications for treatment response. Clin Cancer Res, 2005. 11(7): p. 2637-43. 216. Matsuo, S.E., et al., Transforming growth factor-beta1 and activin A generate antiproliferative signaling in thyroid cancer cells. J Endocrinol, 2006. 190(1): p. 141-50. 217. Wu, S., et al., TGF-beta enforces senescence in Myc-transformed hematopoietic tumor cells through induction of Mad1 and repression of Myc activity. Exp Cell Res, 2009. 315(18): p. 3099-111. 218. Calabrese, V., et al., SOCS1 links cytokine signaling to p53 and senescence. Mol Cell, 2009. 36(5): p. 754-67. 219. Wang, Y., et al., High prevalence and mutual exclusivity of genetic alterations in the phosphatidylinositol-3-kinase/akt pathway in thyroid tumors. J Clin Endocrinol Metab, 2007. 92(6): p. 2387-90. 220. Bertrand-Vallery, V., et al., Repeated exposures to UVB induce differentiation rather than senescence of human keratinocytes lacking p16(INK-4A). Biogerontology, 2011. 11(2): p. 167- 81. 221. Coller, H.A., L. Sang, and J.M. Roberts, A new description of cellular quiescence. PLoS Biol, 2006. 4(3): p. e83. 222. Damri, M., et al., Senescing cells share common features with dedifferentiating cells. Rejuvenation Res, 2009. 12(6): p. 435-43. 223. Cedar, H. and Y. Bergman, Linking DNA methylation and histone modification: patterns and paradigms. Nat Rev Genet, 2009. 10(5): p. 295-304. 224. Suzuki, M.M. and A. Bird, DNA methylation landscapes: provocative insights from epigenomics. Nat Rev Genet, 2008. 9(6): p. 465-76. 225. Herman, J.G. and S.B. Baylin, Gene silencing in cancer in association with promoter hypermethylation. N Engl J Med, 2003. 349(21): p. 2042-54. 226. Ting, A.H., K.M. McGarvey, and S.B. Baylin, The cancer epigenome--components and functional correlates. Genes Dev, 2006. 20(23): p. 3215-31.

214

References

227. Kwabi-Addo, B., et al., Age-related DNA methylation changes in normal human prostate tissues. Clin Cancer Res, 2007. 13(13): p. 3796-802. 228. Esteller, M., Epigenetic gene silencing in cancer: the DNA hypermethylome. Hum Mol Genet, 2007. 16 Spec No 1: p. R50-9. 229. Hansen, K.D., et al., Increased methylation variation in epigenetic domains across cancer types. Nat Genet, 2011. 230. Baylin, S.B. and J.E. Ohm, Epigenetic gene silencing in cancer - a mechanism for early oncogenic pathway addiction? Nat Rev Cancer, 2006. 6(2): p. 107-16. 231. Matsuo, K., et al., Aberrant deoxyribonucleic acid methylation in human thyroid tumors. J Clin Endocrinol Metab, 1993. 77(4): p. 991-5. 232. Xing, M., et al., Methylation of the thyroid-stimulating hormone receptor gene in epithelial thyroid tumors: a marker of malignancy and a cause of gene silencing. Cancer Res, 2003. 63(9): p. 2316-21. 233. Russo, D., et al., Epigenetics of thyroid cancer and novel therapeutic targets. J Mol Endocrinol, 2011. 46(3): p. R73-81. 234. Kondo, T., S.L. Asa, and S. Ezzat, Epigenetic dysregulation in thyroid neoplasia. Endocrinol Metab Clin North Am, 2008. 37(2): p. 389-400, ix. 235. Xing, M., et al., Hypermethylation of the Pendred syndrome gene SLC26A4 is an early event in thyroid tumorigenesis. Cancer Res, 2003. 63(9): p. 2312-5. 236. Kondo, T., et al., Epigenetic silencing of TTF-1/NKX2-1 through DNA hypermethylation and histone H3 modulation in thyroid carcinomas. Lab Invest, 2009. 89(7): p. 791-9. 237. Sassa, M., et al., Aberrant Promoter Methylation in Overexpression of CITED1 in Papillary Thyroid Cancer. Thyroid, 2011. 21(5): p. 511-7. 238. Lal, G., et al., RIZ1 is epigenetically inactivated by promoter hypermethylation in thyroid carcinoma. Cancer, 2006. 107(12): p. 2752-9. 239. Schagdarsurengin, U., et al., Frequent epigenetic inactivation of RASSF2 in thyroid cancer and functional consequences. Mol Cancer, 2010. 9: p. 264. 240. Schagdarsurengin, U., et al., Frequent epigenetic inactivation of RASSF10 in thyroid cancer. Epigenetics, 2009. 4(8): p. 571-6. 241. Zuo, H., et al., Downregulation of Rap1GAP through epigenetic silencing and loss of heterozygosity promotes invasion and progression of thyroid tumors. Cancer Res, 2010. 70(4): p. 1389-97. 242. Siraj, A.K., et al., Demethylation of TMS1 gene sensitizes thyroid cancer cells to TRAIL-induced apoptosis. J Clin Endocrinol Metab, 2011. 96(1): p. E215-24. 243. Park, J.Y., et al., Candidate tumor suppressor gene SLC5A8 is frequently down-regulated by promoter hypermethylation in prostate tumor. Cancer Detect Prev, 2007. 31(5): p. 359-65. 244. Thangaraju, M., et al., SLC5A8 triggers tumor cell apoptosis through pyruvate-dependent inhibition of histone deacetylases. Cancer Res, 2006. 66(24): p. 11560-4. 245. Drini, M., et al., Investigating the potential role of genetic and epigenetic variation of DNA methyltransferase genes in hyperplastic polyposis syndrome. PLoS One, 2011. 6(2): p. e16831. 246. Guan, H., et al., Hypermethylation of the DNA mismatch repair gene hMLH1 and its association with lymph node metastasis and T1799A BRAF mutation in patients with papillary thyroid cancer. Cancer, 2008. 113(2): p. 247-55. 247. Kovalchuk, O., et al., Genome hypermethylation in Pinus silvestris of Chernobyl--a mechanism for radiation adaptation? Mutat Res, 2003. 529(1-2): p. 13-20. 248. Kovalchuk, I., et al., Molecular aspects of plant adaptation to life in the Chernobyl zone. Plant Physiol, 2004. 135(1): p. 357-63. 249. Hofstetter, B., et al., Impact of genomic methylation on radiation sensitivity of colorectal carcinoma. Int J Radiat Oncol Biol Phys, 2010. 76(5): p. 1512-9.

215

References

250. van Attikum, H. and S.M. Gasser, Crosstalk between histone modifications during the DNA damage response. Trends Cell Biol, 2009. 19(5): p. 207-17. 251. Cuozzo, C., et al., DNA damage, homology-directed repair, and DNA methylation. PLoS Genet, 2007. 3(7): p. e110. 252. Aypar, U., W.F. Morgan, and J.E. Baulch, Radiation-induced epigenetic alterations after low and high LET irradiations. Mutat Res, 2011. 707(1-2): p. 24-33. 253. Lahtz, C. and G.P. Pfeifer, Epigenetic changes of DNA repair genes in cancer. J Mol Cell Biol, 2011. 3(1): p. 51-8. 254. Pantazis, P. and W.M. Bonner, Quantitative determination of histone modification. H2A acetylation and phosphorylation. J Biol Chem, 1981. 256(9): p. 4669-75. 255. Xie, H., et al., Genome-wide quantitative assessment of variation in DNA methylation patterns. Nucleic Acids Res, 2011. 39(10): p. 4099-108. 256. Kimura, N., et al., Rapid and quantitative detection of CpG-methylation status using TaqMan PCR combined with methyl-binding-domain polypeptide. Clin Biochem, 2009. 42(10-11): p. 1113-22. 257. Serre, D., B.H. Lee, and A.H. Ting, MBD-isolated Genome Sequencing provides a high- throughput and comprehensive survey of DNA methylation in the human genome. Nucleic Acids Res, 2010. 38(2): p. 391-9. 258. Mardis, E.R., The impact of next-generation sequencing technology on genetics. Trends Genet, 2008. 24(3): p. 133-41. 259. Shendure, J. and H. Ji, Next-generation DNA sequencing. Nat Biotechnol, 2008. 26(10): p. 1135-45. 260. Schuster, S.C., Next-generation sequencing transforms today's biology. Nat Methods, 2008. 5(1): p. 16-8. 261. Zhang, Y. and A. Jeltsch, The Application of Next Generation Sequencing in DNA Methylation Analysis. Genes, 2010. 1(1): p. 85-101. 262. Wu, S.C. and Y. Zhang, Active DNA demethylation: many roads lead to Rome. Nat Rev Mol Cell Biol, 2010. 11(9): p. 607-20. 263. Brenet, F., et al., DNA methylation of the first exon is tightly linked to transcriptional silencing. PLoS One, 2011. 6(1): p. e14524. 264. Sharov, A.A., D.B. Dudekula, and M.S. Ko, Genome-wide assembly and analysis of alternative transcripts in mouse. Genome Res, 2005. 15(5): p. 748-54. 265. Schwartz, S. and G. Ast, Chromatin density and splicing destiny: on the cross-talk between chromatin structure and splicing. EMBO J, 2010. 29(10): p. 1629-36. 266. Edwards, J.R., et al., Chromatin and sequence features that define the fine and gross structure of genomic methylation patterns. Genome Res, 2010. 20(7): p. 972-80. 267. Cheung, H.H., et al., Methylation of an intronic region regulates miR-199a in testicular tumor malignancy. Oncogene, 2011. 268. Wang, Y. and C.G. Lee, MicroRNA and cancer--focus on apoptosis. J Cell Mol Med, 2009. 13(1): p. 12-23. 269. Jazdzewski, K., et al., Polymorphic mature microRNAs from passenger strand of pre-miR-146a contribute to thyroid cancer. Proc Natl Acad Sci U S A, 2009. 106(5): p. 1502-5. 270. Bernstein, E., A.M. Denli, and G.J. Hannon, The rest is silence. RNA, 2001. 7(11): p. 1509-21. 271. Wang, Y., et al., MicroRNA Regulation of Ionizing Radiation-Induced Premature Senescence. Int J Radiat Oncol Biol Phys, 2011. 272. Hummel, R., D.J. Hussey, and J. Haier, MicroRNAs: predictors and modifiers of chemo- and radiotherapy in different tumour types. Eur J Cancer, 2010. 46(2): p. 298-311. 273. Israel, A., et al., Increased microRNA activity in human cancers. PLoS One, 2009. 4(6): p. e6045.

216

References

274. Weidhaas, J.B., et al., MicroRNAs as potential agents to alter resistance to cytotoxic anticancer therapy. Cancer Res, 2007. 67(23): p. 11111-6. 275. Maes, O.C., et al., Changes in MicroRNA expression patterns in human fibroblasts after low- LET radiation. J Cell Biochem, 2008. 105(3): p. 824-34. 276. Wagner-Ecker, M., et al., MicroRNA expression after ionizing radiation in human endothelial cells. Radiat Oncol, 2010. 5: p. 25. 277. Josson, S., et al., Radiation modulation of microRNA in prostate cancer cell lines. Prostate, 2008. 68(15): p. 1599-606. 278. Cha, H.J., et al., Identification of specific microRNAs responding to low and high dose gamma- irradiation in the human lymphoblast line IM9. Oncol Rep, 2009. 22(4): p. 863-8. 279. Simone, N.L., et al., Ionizing radiation-induced oxidative stress alters miRNA expression. PLoS One, 2009. 4(7): p. e6377. 280. Lan, F.F., et al., Hsa-let-7g inhibits proliferation of hepatocellular carcinoma cells by downregulation of c-Myc and upregulation of p16(INK4A). Int J Cancer, 2010. 128(2): p. 319- 31. 281. Shimizu, S., et al., The let-7 family of microRNAs inhibits Bcl-xL expression and potentiates sorafenib-induced apoptosis in human hepatocellular carcinoma. J Hepatol, 2010. 52(5): p. 698-704. 282. Ricarte-Filho, J.C., et al., Effects of let-7 microRNA on Cell Growth and Differentiation of Papillary Thyroid Cancer. Transl Oncol, 2009. 2(4): p. 236-41. 283. He, S., et al., Aurora kinase A induces miR-17-92 cluster through regulation of E2F1 transcription factor. Cell Mol Life Sci, 2010. 67(12): p. 2069-76. 284. Pickering, M.T., B.M. Stadler, and T.F. Kowalik, miR-17 and miR-20a temper an E2F1-induced G1 checkpoint to regulate cell cycle progression. Oncogene, 2009. 28(1): p. 140-5. 285. Cimmino, A., et al., miR-15 and miR-16 induce apoptosis by targeting BCL2. Proc Natl Acad Sci U S A, 2005. 102(39): p. 13944-9. 286. Liu, Q., et al., miR-16 family induces cell cycle arrest by regulating multiple cell cycle genes. Nucleic Acids Res, 2008. 36(16): p. 5391-404. 287. Corney, D.C., et al., MicroRNA-34b and MicroRNA-34c are targets of p53 and cooperate in control of cell proliferation and adhesion-independent growth. Cancer Res, 2007. 67(18): p. 8433-8. 288. He, L., et al., A microRNA component of the p53 tumour suppressor network. Nature, 2007. 447(7148): p. 1130-4. 289. Breving, K. and A. Esquela-Kerscher, The complexities of microRNA regulation: mirandering around the rules. Int J Biochem Cell Biol, 2011. 42(8): p. 1316-29. 290. Vasudevan, S., Y. Tong, and J.A. Steitz, Switching from repression to activation: microRNAs can up-regulate translation. Science, 2007. 318(5858): p. 1931-4. 291. Vasudevan, S., Y. Tong, and J.A. Steitz, Cell-cycle control of microRNA-mediated translation regulation. Cell Cycle, 2008. 7(11): p. 1545-9. 292. Marsit, C.J., K. Eddy, and K.T. Kelsey, MicroRNA responses to cellular stress. Cancer Res, 2006. 66(22): p. 10843-8. 293. Prazeres, H., et al., Chromosomal, epigenetic and microRNA-mediated inactivation of LRP1B, a modulator of the extracellular environment of thyroid cancer cells. Oncogene, 2011. 30(11): p. 1302-17. 294. Little, J.B., Cellular, molecular, and carcinogenic effects of radiation. Hematol Oncol Clin North Am, 1993. 7(2): p. 337-52. 295. Hooker, A.M., et al., The linear no-threshold model does not hold for low-dose ionizing radiation. Radiat Res, 2004. 162(4): p. 447-52.

217

References

296. Jin, Y.W., et al., Comprehensive analysis of time- and dose-dependent patterns of gene expression in a human mesenchymal stem cell line exposed to low-dose ionizing radiation. Oncol Rep, 2008. 19(1): p. 135-44. 297. Christov, K., Thyroid cell proliferation in rats and induction of tumors by X-rays. Cancer Res, 1975. 35(5): p. 1256-62. 298. Meireles, A.M., et al., Molecular and genotypic characterization of human thyroid follicular cell carcinoma-derived cell lines. Thyroid, 2007. 17(8): p. 707-15. 299. Gary, R.K. and S.M. Kindell, Quantitative assay of senescence-associated beta-galactosidase activity in mammalian cell extracts. Anal Biochem, 2005. 343(2): p. 329-34. 300. Sengupta, S. and C.C. Harris, p53: traffic cop at the crossroads of DNA repair and recombination. Nat Rev Mol Cell Biol, 2005. 6(1): p. 44-55. 301. Boice, J.D., Jr., Radiation-induced thyroid cancer--what's new? J Natl Cancer Inst, 2005. 97(10): p. 703-5. 302. Luster, M., et al., Thyroid cancer in childhood: management strategy, including dosimetry and long-term results. Hormones (Athens), 2007. 6(4): p. 269-78. 303. Frasca, F., et al., p73 tumor-suppressor activity is impaired in human thyroid cancer. Cancer Res, 2003. 63(18): p. 5829-37. 304. Garner, E. and K. Raj, Protective mechanisms of p53-p21-pRb proteins against DNA damage- induced cell death. Cell Cycle, 2008. 7(3): p. 277-82. 305. Namba, H., et al., Radiation-induced G1 arrest is selectively mediated by the p53-WAF1/Cip1 pathway in human thyroid cells. Cancer Res, 1995. 55(10): p. 2075-80. 306. Rosenbluth, J.M. and J.A. Pietenpol, The jury is in: p73 is a tumor suppressor after all. Genes Dev, 2008. 22(19): p. 2591-5. 307. Levrero, M., et al., The p53/p63/p73 family of transcription factors: overlapping and distinct functions. J Cell Sci, 2000. 113 ( Pt 10): p. 1661-70. 308. Soussi, T. and L. Hjortsberg, When mutant p53 plays hide and seek: a new challenge for diagnosis and therapy? Trends Mol Med, 2009. 15(1): p. 1-4. 309. McCubrey, J.A., et al., Roles of the RAF/MEK/ERK and PI3K/PTEN/AKT pathways in malignant transformation and drug resistance. Adv Enzyme Regul, 2006. 46: p. 249-79. 310. Antonsson, B. and J.C. Martinou, The Bcl-2 protein family. Exp Cell Res, 2000. 256(1): p. 50-7. 311. Blackburn, D.J., et al., Occurrence of thyroid papillary carcinoma in young patients. A Chernobyl connection? J Pediatr Endocrinol Metab, 2001. 14(5): p. 503-6. 312. Amundson, S.A., et al., Identification of potential mRNA biomarkers in peripheral blood lymphocytes for human exposure to ionizing radiation. Radiat Res, 2000. 154(3): p. 342-6. 313. Kang, C.M., et al., Possible biomarkers for ionizing radiation exposure in human peripheral blood lymphocytes. Radiat Res, 2003. 159(3): p. 312-9. 314. Abou-El-Ardat, K., et al., Response to low-dose X-irradiation is p53-dependent in a papillary thyroid carcinoma model system. Int J Oncol, 2011. 315. Santoro, M., et al., Development of thyroid papillary carcinomas secondary to tissue-specific expression of the RET/PTC1 oncogene in transgenic mice. Oncogene, 1996. 12(8): p. 1821-6. 316. Chomczynski, P. and N. Sacchi, Single-step method of RNA isolation by acid guanidinium thiocyanate-phenol-chloroform extraction. Anal Biochem, 1987. 162(1): p. 156-9. 317. Lewis, B.P., C.B. Burge, and D.P. Bartel, Conserved seed pairing, often flanked by adenosines, indicates that thousands of human genes are microRNA targets. Cell, 2005. 120(1): p. 15-20. 318. Irizarry, R.A., et al., Exploration, normalization, and summaries of high density oligonucleotide array probe level data. Biostatistics, 2003. 4(2): p. 249-264. 319. Bolstad, B.M., et al., A comparison of normalization methods for high density oligonucleotide array data based on variance and bias. Bioinformatics, 2003. 19(2): p. 185-193.

218

References

320. Smyth, G.K., Linear Models and Empirical Bayes Methods for Assessing Differential Expression in Microarray Experiments. Statistical Applications in Genetics and Molecular Biology, 2004. 3(1): p. 25. 321. Benjamini, Y. and Y. Hochberg, CONTROLLING THE FALSE DISCOVERY RATE - A PRACTICAL AND POWERFUL APPROACH TO MULTIPLE TESTING. Journal of the Royal Statistical Society Series B-Methodological, 1995. 57(1): p. 289-300. 322. Lee, M.L. and G.A. Whitmore, Power and sample size for DNA microarray studies. Stat Med, 2002. 21(23): p. 3543-70. 323. Wasenius, V.M., et al., Hepatocyte growth factor receptor, matrix metalloproteinase-11, tissue inhibitor of metalloproteinase-1, and fibronectin are up-regulated in papillary thyroid carcinoma: a cDNA and tissue microarray study. Clin Cancer Res, 2003. 9(1): p. 68-75. 324. Burniat, A., et al., Gene expression in RET/PTC3 and E7 transgenic mouse thyroids: RET/PTC3 but not E7 tumors are partial and transient models of human papillary thyroid cancers. Endocrinology, 2008. 149(10): p. 5107-17. 325. Gupta, N., et al., SLC5A8 (SMCT1)-mediated transport of butyrate forms the basis for the tumor suppressive function of the transporter. Life Sci, 2006. 78(21): p. 2419-25. 326. Ganapathy, V., et al., Biological functions of SLC5A8, a candidate tumour suppressor. Biochem Soc Trans, 2005. 33(Pt 1): p. 237-40. 327. Abou-El-Ardat, K., Derradji, H, De Meyer, T, Bekaert, S, Van Criekinge, W, Baatout, S, Response to low-dose X-irradiation is p53-dependent in a papillary thyroid carcinoma model system. Int J Oncol, 2011. 328. Zhang, G., et al., Applying proteomic methodologies to analyze the effect of hexamethylene bisacetamide (HMBA) on proliferation and differentiation of human gastric carcinoma BGC- 823 cells. Int J Biochem Cell Biol, 2004. 36(8): p. 1613-23. 329. Bonnefont, J., et al., Evolutionary forces shape the human RFPL1,2,3 genes toward a role in neocortex development. Am J Hum Genet, 2008. 83(2): p. 208-18. 330. Hieber, L., et al., Chromosomal rearrangements in post-Chernobyl papillary thyroid carcinomas: evaluation by spectral karyotyping and automated interphase FISH. J Biomed Biotechnol, 2011. 2011: p. 693691. 331. Unger, K., et al., Array CGH demonstrates characteristic aberration signatures in human papillary thyroid carcinomas governed by RET/PTC. Oncogene, 2008. 27(33): p. 4592-602. 332. Finn, S., et al., Low-level genomic instability is a feature of papillary thyroid carcinoma: an array comparative genomic hybridization study of laser capture microdissected papillary thyroid carcinoma tumors and clonal cell lines. Arch Pathol Lab Med, 2007. 131(1): p. 65-73. 333. Hu, S., et al., Association of aberrant methylation of tumor suppressor genes with tumor aggressiveness and BRAF mutation in papillary thyroid cancer. Int J Cancer, 2006. 119(10): p. 2322-9. 334. Martin, C.L., et al., Chromosomal imbalances in oral squamous cell carcinoma: examination of 31 cell lines and review of the literature. Oral Oncol, 2008. 44(4): p. 369-82. 335. Kryh, H., et al., Comprehensive SNP array study of frequently used neuroblastoma cell lines; Copy neutral loss of heterozygosity is common in the cell lines but uncommon in primary tumors. BMC Genomics, 2011. 12(1): p. 443. 336. Feys, T., et al., A detailed inventory of DNA copy number alterations in four commonly used Hodgkin's lymphoma cell lines. Haematologica, 2007. 92(7): p. 913-20. 337. Hess, J., et al., Gain of chromosome band 7q11 in papillary thyroid carcinomas of young patients is associated with exposure to low-dose irradiation. Proc Natl Acad Sci U S A, 2011. 108(23): p. 9595-600. 338. Jin, L., et al., Human thyroid tumours, the puzzling lessons from E7 and RET/PTC3 transgenic mice. Br J Cancer, 2008. 99(11): p. 1874-83.

219

References

339. Karthigeyan, D., et al., Biology of Aurora A kinase: Implications in cancer manifestation and therapy. Med Res Rev. 340. Carmena, M., S. Ruchaud, and W.C. Earnshaw, Making the Auroras glow: regulation of Aurora A and B kinase function by interacting proteins. Curr Opin Cell Biol, 2009. 21(6): p. 796-805. 341. Lim, S.K. and G. Gopalan, Aurora-A kinase interacting protein 1 (AURKAIP1) promotes Aurora- A degradation through an alternative ubiquitin-independent pathway. Biochem J, 2007. 403(1): p. 119-27. 342. Yokoyama-Kobayashi, M., et al., Human cDNA encoding a novel TGF-beta superfamily protein highly expressed in placenta. J Biochem, 1997. 122(3): p. 622-6. 343. Hotten, G., et al., Cloning and expression of recombinant human growth/differentiation factor 5. Biochem Biophys Res Commun, 1994. 204(2): p. 646-52. 344. Yamaguchi, K., et al., Identification of nonsteroidal anti-inflammatory drug-activated gene (NAG-1) as a novel downstream target of phosphatidylinositol 3-kinase/AKT/GSK-3beta pathway. J Biol Chem, 2004. 279(48): p. 49617-23. 345. Li, P.X., et al., Placental transforming growth factor-beta is a downstream mediator of the growth arrest and apoptotic response of tumor cells to DNA damage and p53 overexpression. J Biol Chem, 2000. 275(26): p. 20127-35. 346. Bauskin, A.R., et al., Role of macrophage inhibitory cytokine-1 in tumorigenesis and diagnosis of cancer. Cancer Res, 2006. 66(10): p. 4983-6. 347. Mimeault, M. and S.K. Batra, Divergent molecular mechanisms underlying the pleiotropic functions of macrophage inhibitory cytokine-1 in cancer. J Cell Physiol. 224(3): p. 626-35. 348. Yang, H., et al., Macrophage inhibitory cytokine-1: a novel biomarker for p53 pathway activation. Mol Cancer Ther, 2003. 2(10): p. 1023-9. 349. Chen, X., Cyclin G: a regulator of the p53-Mdm2 network. Dev Cell, 2002. 2(5): p. 518-9. 350. Zhao, L., et al., Cyclin G1 has growth inhibitory activity linked to the ARF-Mdm2-p53 and pRb tumor suppressor pathways. Mol Cancer Res, 2003. 1(3): p. 195-206. 351. D'Inzeo, S., et al., Role of reduced expression of SMAD4 in papillary thyroid carcinoma. J Mol Endocrinol. 45(4): p. 229-44. 352. Huntzinger, E. and E. Izaurralde, Gene silencing by microRNAs: contributions of translational repression and mRNA decay. Nat Rev Genet. 12(2): p. 99-110. 353. Wang, Y., et al., MicroRNA Regulation of Ionizing Radiation-Induced Premature Senescence. Int J Radiat Oncol Biol Phys. 354. Wagner-Ecker, M., et al., MicroRNA expression after ionizing radiation in human endothelial cells. Radiat Oncol. 5: p. 25. 355. Lan, F.F., et al., Hsa-let-7g inhibits proliferation of hepatocellular carcinoma cells by downregulation of c-Myc and upregulation of p16(INK4A). Int J Cancer. 128(2): p. 319-31. 356. Breving, K. and A. Esquela-Kerscher, The complexities of microRNA regulation: mirandering around the rules. Int J Biochem Cell Biol. 42(8): p. 1316-29. 357. Li, G., et al., Alterations in microRNA expression in stress-induced cellular senescence. Mech Ageing Dev, 2009. 130(11-12): p. 731-41. 358. Ding, L.H., et al., Gene expression profiles of normal human fibroblasts after exposure to ionizing radiation: a comparative study of low and high doses. Radiat Res, 2005. 164(1): p. 17-26. 359. Smith, J.A., et al., Methylation status of genes in papillary thyroid carcinoma. Arch Otolaryngol Head Neck Surg, 2007. 133(10): p. 1006-11. 360. Cedar, H., et al., Effect of DNA methylation on gene expression. Cold Spring Harb Symp Quant Biol, 1983. 47 Pt 2: p. 605-9. 361. Li, N., et al., Whole genome DNA methylation analysis based on high throughput sequencing technology. Methods, 2010. 52(3): p. 203-12.

220

References

362. Bock, C., et al., Quantitative comparison of genome-wide DNA methylation mapping technologies. Nat Biotechnol, 2011. 28(10): p. 1106-14. 363. Sandoval, J., et al., Validation of a DNA methylation microarray for 450,000 CpG sites in the human genome. Epigenetics, 2011. 6(6): p. 692-702. 364. Wu, S.C. and Y. Zhang, Active DNA demethylation: many roads lead to Rome. Nat Rev Mol Cell Biol, 2010. 11(9): p. 607-20. 365. Zhang, C.M., et al., Over-expression of NYGGF4 inhibits glucose transport in 3T3-L1 adipocytes via attenuated phosphorylation of IRS-1 and Akt. Acta Pharmacol Sin, 2009. 30(1): p. 120-4. 366. Magalhaes, J.G., et al., What is new with Nods? Curr Opin Immunol, 2011. 23(1): p. 29-34. 367. van Agthoven, T., et al., CITED2 and NCOR2 in anti-oestrogen resistance and progression of breast cancer. Br J Cancer, 2009. 101(11): p. 1824-32. 368. Christiansen, J., et al., IGF2 mRNA-binding protein 2: biological function and putative role in type 2 diabetes. J Mol Endocrinol, 2009. 43(5): p. 187-95. 369. Ghisletti, S., et al., Cooperative NCoR/SMRT interactions establish a corepressor-based strategy for integration of inflammatory and anti-inflammatory signaling pathways. Genes Dev, 2009. 23(6): p. 681-93. 370. Shimomura, Y., et al., APCDD1 is a novel Wnt inhibitor mutated in hereditary hypotrichosis simplex. Nature, 2010. 464(7291): p. 1043-7. 371. Wang, D., et al., Expression of ADAMTS12 in Colorectal Cancer-Associated Stroma Prevents Cancer Development and Is a Good Prognostic Indicator of Colorectal Cancer. Dig Dis Sci, 2011. 372. Moncada-Pazos, A., et al., The ADAMTS12 metalloprotease gene is epigenetically silenced in tumor cells and transcriptionally activated in the stroma during progression of colon cancer. J Cell Sci, 2009. 122(Pt 16): p. 2906-13. 373. Svendsen, J.M., et al., Mammalian BTBD12/SLX4 assembles a Holliday junction resolvase and is required for DNA repair. Cell, 2009. 138(1): p. 63-77. 374. Rahu, M., Health effects of the Chernobyl accident: fears, rumours and the truth. Eur J Cancer, 2003. 39(3): p. 295-9. 375. Schweppe, R.E., et al., Deoxyribonucleic acid profiling analysis of 40 human thyroid cancer cell lines reveals cross-contamination resulting in cell line redundancy and misidentification. J Clin Endocrinol Metab, 2008. 93(11): p. 4331-41. 376. Hieber, L., et al., Chromosomal rearrangements in post-Chernobyl papillary thyroid carcinomas: evaluation by spectral karyotyping and automated interphase FISH. J Biomed Biotechnol. 2011: p. 693691. 377. Rhoden, K.J., et al., RET/papillary thyroid cancer rearrangement in nonneoplastic thyrocytes: follicular cells of Hashimoto's thyroiditis share low-level recombination events with a subset of papillary carcinoma. J Clin Endocrinol Metab, 2006. 91(6): p. 2414-23. 378. Campisi, J. and F. d'Adda di Fagagna, Cellular senescence: when bad things happen to good cells. Nat Rev Mol Cell Biol, 2007. 8(9): p. 729-40. 379. Rodier, F., et al., DNA-SCARS: distinct nuclear structures that sustain damage-induced senescence growth arrest and inflammatory cytokine secretion. J Cell Sci, 2011. 124(Pt 1): p. 68-81. 380. Helton, E.S. and X. Chen, p53 modulation of the DNA damage response. J Cell Biochem, 2007. 100(4): p. 883-96. 381. Farnebo, M., V.J. Bykov, and K.G. Wiman, The p53 tumor suppressor: a master regulator of diverse cellular processes and therapeutic target in cancer. Biochem Biophys Res Commun, 2010. 396(1): p. 85-9. 382. Hill, R., et al., p53 Binding to the p21 promoter is dependent on the nature of DNA damage. Cell Cycle, 2008. 7(16): p. 2535-43.

221

References

383. Jõers, A., et al., p53-dependent transcription can exhibit both on/off and graded response after genotoxic stress. Oncogene, 2004. 23(37): p. 6175-85. 384. D'Inzeo, S., et al., Role of reduced expression of SMAD4 in papillary thyroid carcinoma. J Mol Endocrinol, 2010. 45(4): p. 229-44. 385. Mitsutake, N., et al., BRAF mediates RET/PTC-induced mitogen-activated protein kinase activation in thyroid cells: functional support for requirement of the RET/PTC-RAS-BRAF pathway in papillary thyroid carcinogenesis. Endocrinology, 2006. 147(2): p. 1014-9. 386. Pyper, S.R., et al., PPARalpha: energy combustion, hypolipidemia, inflammation and cancer. Nucl Recept Signal, 2010. 8: p. e002. 387. Demetrius, L., Of mice and men. When it comes to studying ageing and the means to slow it down, mice are not just small humans. EMBO Rep, 2005. 6 Spec No: p. S39-44. 388. Gutmann, D.H., K. Hunter-Schaedle, and K.M. Shannon, Harnessing preclinical mouse models to inform human clinical cancer trials. J Clin Invest, 2006. 116(4): p. 847-52. 389. Behr, T.M., D.M. Goldenberg, and W.S. Becker, Radioimmunotherapy of solid tumors: a review "of mice and men". Hybridoma, 1997. 16(1): p. 101-7. 390. Ouyang, B., et al., Mechanisms of aneuploidy in thyroid cancer cell lines and tissues: evidence for mitotic checkpoint dysfunction without mutations in BUB1 and BUBR1. Clin Endocrinol (Oxf), 2002. 56(3): p. 341-50. 391. Ainsbury, E.A., et al., Radiation cataractogenesis: a review of recent studies. Rad Res, 2009. 172(1): p. 1-9. 392. European Commission. Radiation Protection No. 170: Recent scientific findings and publications on the health effects of Chernobyl. Summary Report. Directorate-General for Energy, Directorate D – Nuclear Energy, Unit D.4 – Radiation Protection, 2011. 393. Jacob, P., et al., Childhood exposure due to the Chernobyl accident and thyroid cancer risk in contaminated areas of Belarus and Russia. Br. J. Cancer, 1999. 80(9): p. 1461-9.

222

Epilogue Acknowledgements

Long is the way, and hard, that out of Hell leads up to light

John Milton, "Paradise Lost"

224

Acknowledgements

person is the sum of his experiences and the events that occur in our lives shape them

into what they are at any given moment. Therefore, it would make sense to thank every A person that crossed my path since I was born as an integral player in where I am now. However, for reasons of practicality I will refrain from doing that and will instead apologize in advance to anyone I might have forgotten here.

A scientific mind always strives to form order from chaos and since this is a scientific document (at least I hope it is), I'll order the people below in chronological order of appearance in my life.

Naturally, the first two persons to whom I'm indebted are my parents, Ahmad and Esther, who among other things taught me that we are first and foremost human beings with more similarities than differences despite our creed or nationality. They also taught me the value of education and perseverance and this thesis is testament to the great job they both did. Next is my sister, Adelle, who is two years my junior and a rising star in the world of nutrition sciences.

It's been more than four years ago that I first set foot in Belgium. I was a stranger in a foreign land that had three different districts, the same number of languages, and the same number of governments. You'd assume that a stranger would be lost in such a labyrinth and that's where you'd be wrong. I've found the people to be warm and kind and that greatly helped alleviate any sense of homesickness I had. The first face I saw in Belgium was that of Sarah Baatout, my co-promoter and mentor at SCK•CEN. I appreciate both her guidance throughout my PhD and the moments of freedom that she gave me to pursue whatever line of thought or to prove a hypothesis. I knew that I could always drop in to ask a question or to talk about my research and that I could always count on her for support. Wim Van Criekinge, my promoter, was always ready with a new idea and always prepared to nudge me in the right direction. I greatly enjoyed our meetings and his mantra

‘everything will be fine’ was to me like the ‘Don’t Panic’ button on the Hitchhiker’s Guide to the

Galaxy. Sofie Bekaert, my co-promoter, was always there with an encouraging remark despite the move from one faculty to another. Whereas everybody has one to two promoters, I was blessed with

225

Acknowledgements

four. Last but not least of my co-promoters is Tim De Meyer whose help was crucial for the success of this thesis. Always full of ideas and energy, his comments on my manuscripts were forever illuminating.

Since my arrival to the radiobiology group in SCK-CEN, I’ve watched it grow gradually with the addition of new PhD students and post-doctoral researchers. Michaël Beck became a friend, colleague and office mate for the the duration of my stay here. Our time in the office was filled with discussions of a scientific and personal nature. And naturally with lots of laughs that probably disturbed the whole corridor. Also sharing that ‘great room on the roof’ were two other PhD students and friends, Myriam Ghardi and Nada Samari, who were always there to share an anecdote or a cup of Senseo coffee. The world is truly a small village; I realized that when I met Hussein El

Saghire whose university was a few kilometers from mine in Beirut but whom I met only in Mol! The arrival of Giuseppe Pani made this a truly international group. The list of PhD students grew with the addition of Matthias D’huyvetter, Kristel Mijnendonckx, Marlies Gijs, Annelies Suetens, and

Charlotte Rombouts. Although I didn’t have as much time to get to know you all, your company and friendship was greatly appreciated. Charlotte and Annelies deserve special mention for their contribution of the Dutch translation of the summary to this thesis. Thanks! From across the biological divide, namely the group of microbiology, Hanène Badri and Salem Ben Hammouda, hope you have a great life together. The newcomers Tom Verbiest and Tine Verreet who arrived just as I was preparing to leave. Hope you have a successful time in our group. And I can’t mention the new without remembering the old: Benny Pycke, Nicolas Morin, and Aurélie Crabbé who have long left but are still remembered.

Of course I can’t forget the experience of the PIs: Rafi Benotmane, Louis De Saint-Georges, Paul

Jacquet, Nathalie Leys, Hugo Moors and Paul Janssen. Talking to you was always a pleasure.

Furthermore, I can’t forget the great support from Hanane Derradji, An Aerts, Marjan Moreels, Roel

Quintens, and Pieter Monsieurs. The latter deserves special mention for his great help with the

226

Acknowledgements

microarrays all through the thesis. He was always ready to explain a complicated statistical concept.

Working with him was a pleasure.

The silent soldiers of our lab: the technicians Ann Janssen, Arlette Michaux, Mieke Neefs, Jasmine

Buset, and Liselotte Leysen: you were always eager to help and I’m forever grateful. A special thanks is extended to Kevin Tabury who at the moment is enjoying the Florida sun. His smile was enough to brighten any day and his great energy in the lab kept it running perfectly.

During the course of my PhD, I had the opportunity to travel about Europe and I spent some time at the Helmholtz Zentrum in Munich, Germany where I met Nataša Anastasov and Mike Atkinson with whom the collaboration on microRNAs was a fruitful one.

Finally, I’d like to mention some people from out of the lab whose friendship meant a lot to me during this PhD: Praveen Pandey, Irina Nikolova, Naresh Rawat, Deepak Pant, Vinayaraj OK, Boris

Minov, and Wendy Vandendries. A special thank you to Maya Serhan who was always there for me in more ways than one and still is despite the distance.

227

Acknowledgements

228

Curriculum Vitae Curriculum Vitae

If I knew all there was to learn, every twist and turn, why do I still try?

Vangelis, "Losing Sleep (Still My Heart)"

230

Curriculum Vitae

CURRICULUM VITAE KHALIL ABOUELARADAT

Gender Male

Current Address Boeretang 204/15 2400 Mol Belgium 0032-486775484 (Mobile) [email protected]

Education - Universiteit Gent, Ph.D., Doctor of Applied Biological Sciences, expected defense: 21st November 2011 - American University of Beirut, M.Sc., Biology, February 2006 - American University of Beirut, B.Sc., Biology, February 2004

Honors and Awards - American University of Beirut's Dean's Honor List (average above 85% in any given semester): Spring 2000, Fall 2000, Spring 2001, and Fall 2002 - Oral presentation selection award from the Belgian Society of Molecular Biology and Cytometry (200 euros cash prize) November 2008 - AWM PhD grant from SCK-CEN and Universiteit Gent

Research Experience - Epigenetic, Metabolic and Signal Transduction Changes in Normal and Transformed Thyrocytes upon Exposure to Low Doses of X-rays, laboratory of Sarah Baatout and Wim Van Criekinge, Belgian Nuclear Research Center (SCK•CEN) and Universiteit Gent, 2007-2011 - Effect of Low Doses of Radiation on a Cell Line of Papillary Thyroid Carcinoma, TPC-1, laboratory of Sarah Baatout, Belgian Nuclear Research Center (SCK•CEN), April – July 2007 - Effect of a Solution of PEG-coated Nanoparticles on the Anti-proliferative Effects of Epican Forte on Cell Lines of Leukemic Origin, laboratory of Elias Baydoun, American University of Beirut, September 2006 – March 2007 - Effect of a Solution of Indapamide on Morphological and Molecular Changes in the Left Ventricle by Hypertension, laboratory of Michael Nasser, American University of Beirut, October 2005 – June 2006 - Effect of Epigallocatechin Gallate, a Green Tea Extract, on Cell Lines of Leukemic Origin with and without HTLV-1, American University of Beirut, February 2004 – February 2006 - Effect of Milk Formula on Iron Intake of Children, laboratory of Dr. Muwaket, Children's Cancer Center, Beirut, Lebanon, June – August 2003

231

Curriculum Vitae

Teaching Experience - Gave three lectures on apoptosis and system's biology to students following the European Master's in Radiation Biology, February 2010 and February 2011, Belgian Nuclear Research Center (SCK•CEN), Mol, Belgium - Laboratory instructor, Introductory Biology I, American University of Beirut, Fall 2004-05 - Laboratory instructor, Introductory Biology II, American University of Beirut, Spring 2004, Spring 2005, Fall 2005-06 - Tutor, Commoner's Public Service, May – July 2006 (Lebanese System 8th grade biology)

Publications with Peer Review

- Gérard AC, Humblet K, Wilvers C, Poncin S, Derradji H, de Ville de Goyet C, Abou- El-Ardat K, Baatout S, Sonveaux P, Denef JF, Colin I. Iodine deficiency-induced long lasting angiogenic reaction in thyroid cancers occurs via a VEGF-HIF-1, but not a ROS dependent pathway. Thyroid. In review. - Abou-El-Ardat K, Monsieurs P, Anastasov N, Atkinson M, Derradji H, De Meyer T, Beakert S, Van Criekinge W, Baatout S. Low dose irradiation of thyroid cells reveals a unique transcriptomic and epigenetic signature in RET/PTC-positive cells. Mut. Res. 2011; doi: 10.1016/j.mrfmmm.2011.10.006. - Abou-El-Ardat K, Derradji H, De Meyer T, Bekaert S, Van Criekinge W, Baatout S. Response to low-dose X-irradiation is p53-dependent in a papillary thyroid carcinoma model system. Int. J. Oncol. 2011; 39: 1429-41 - Derradji H, Abou-El-Ardat K, Aerts A, Faraj Akram K, Baatout S, Harakeh S. 2010. Antioxidants: a new approach to tackle radiation-induced cancer, In: Herbal Medicine, A Cancer Chemopreventive and Therapeutic Perspective. 1st ed. Jaypee Brothers Medical Publishers (P) Ltd. pp. 441-472 - Derradji H, Bekaert S, De Meyer T, Jacquet P, Abou-El-Ardat K, Ghardi M, Michaux A, Baatout S. Ionizing radiation-induced gene modulations, cytokine content changes and telomere shortening in mouse fetuses exhibiting forelimb defects. Dev. Biol. 2008; 322: 302-313 - Abou-El-Ardat K, Derradji H, Bekaert S, De Meyer T, Van Criekinge W, Baatout S. GENRISK-T Project: response of the TPC-1 papillary thyroid carcinoma cell line to a range of low to high doses of irradiation. Commun. Agric. Appl. Biol. Sci. 2008; 73: 79-83 - Harakeh S, Abu-El-Ardat K, Diab-Assaf M, Niedzwiecki A, El-Sabban M, Rath M. Epigallocatechin-3-gallate induces apoptosis and cell cycle arrest in HTLV-1 positive and –negative leukemia cells. Med. Oncol. 2008; 25: 30-39 - Harakeh S, Diab-Assaf M, Khalife J, Abu-El-Ardat K, Baydoun E, Niedzwiecki A, El-Sabban M, Rath M. Ascorbic acid induces apoptosis in adult T-cell leukemia. Anti- cancer Res. 2007; 27: 289-298 - Harakeh S, Diab-Assaf M, Abu-El-Ardat K, Niedzwiecki A, Rath M. Mechanistic aspects of apoptosis induction by L-lysine in both HTLV-1-positive and –negative cell lines. Chemo-biological Interactions 2006; 164: 102-114 - Harakeh S, Diab-Assaf M, Niedzwiecki A, Khalife J, Abu-El-Ardat K, Rath M. Apoptosis induction by Epican Forte in HTLV-1 positive and negative malignant T- cells. Leukemia Res. 2006; 30: 869-881

232

Curriculum Vitae

- Harakeh S, Afifi Soweid R, Cortbawi H, Abu-El-Ardat K, Ossama A, Accaoui R, Bendaly E, Hakim W, Kadri A, Masroujeh R, Obeid M, Shatila K. Knowledge, attitudes and behaviors concerning cow disease among physicians in Lebanon. Food Protection Trends 2005; 25: 250-255

Selected Oral and Poster Presentations - Abou-El-Ardat K, Monsieurs P, Janssen A, Beck M, Michaux A, Anastasov N, Atkinson M, Benotmane R, Derradji H, Bekaert S, Van Criekinge W, Baatout S. Distinctive Genomic Profiles of Normal and Transformed Thyrocytes Irradiated with Low vs. High Doses of X-irradiation both in vitro and in vivo. 14th International Thyroid Congress. September 11-16, 2010, Paris, France (Poster Presentation) - Abou-El-Ardat K. Radiation and Thyroid Cancer: What Chernobyl Told Us. Luminex Planet xMAP 2009. October 7-9, 2009, Amsterdam, The Netherlands (invited speaker) - Abou-El-Ardat K, De Vos W. The Boundaries of Proliferation: Two Means to an End. Belgian Society for Analytical Cytology's Molecular Biology and Cytometry Course. May 7-8, 2009, Mol, Belgium (Oral Presentation) - Abou-El-Ardat K. External X-irradiation Causes a Decrease in Proliferation of Two Papillary Thyroid Carcinoma Cell Lines With No Significant Increase in Cell Death. Annual Meeting of the Belgian Society for Analytical Cytology. November 14, 2008, Liège, Belgium (Award winning oral presentation) - Abou-El-Ardat K, Derradji H, Bekaert S, De Meyer T, Van Criekinge W, Baatout S. GENRISK-T: Response of TPC-1 Papillary Thyroid Carcinoma Cell Line to a Range of Low to High Doses of Irradiation. 33rd Annual Meeting of the European Thyroid Association. September 20-24, 2008, Thessaloniki, Greece (Poster Presentation) - Harakeh S, Abu-El-Ardat K, Diab-Assaf M, Niedzwiecki A, Khalife J, Roomi W, Rath M. Epigallocatechin-3-gallate Induces Apoptosis and Cell Cycle Arrest in HTLV-1-positive and -Negative Cell Lines. XI International Congress of Bacteriology and Applied Microbiology. (International Union of Microbiological Societies) July 23-28, 2005, San Francisco, USA (Poster presentation). - Harakeh S, Diab-Assaf M, Niedzwiecki A, Khalife J, Abu-El-Ardat K, Roomi W, Rath M. Apoptosis Induction by Ascorbic Acid in HTLV-1 Positive and Negative MalignanT-cells. 1st International Conference on Environmental, Industrial and Applied Microbiology (BioMicroWorld-2005) March 15-18th 2005, Badajoz, Spain (Poster presentation). - Harakeh S, Diab-Assaf M, Niedzwiecki A, Khalife J, Abu-El-Ardat K, Roomi W, Rath M. Induction of apoptosis by Epican Forte in HTLV-1 positive and negative malignant T-cellsWorld Conference on Dosing of Anti-infectives, Dosing the Magic Bullets, September 9-11, 2004, Nürnberg, Germany (Invited oral presentation). - Harakeh S, Diab-Assaf M, Niedziecki A, Khalife J, Abu-El-Ardat K, Roomi W, Rath M. Anti-proliferative Effects of Antioxidants using HTLV-1 Positive and Negative Malignant T-cells. American Society for Microbiology. New Orleans May, 2004 (Poster presentation).

Reports - Participated in writing the report: FI6R-CT-2006-36495: GENRISK-T on the effect of radiation on the thyroid and which was presented to the European Commission. - Participated in the preparation of the final reports for the EU FP6 GENRISK-T contract FI6R-CT-2006-36495: Project Deliverable A 'Scientific Summary', Project 233

Curriculum Vitae

Deliverable B 'GENRISK-T Defining the Genetic Component of Thyroid Cancer Risk at Low Doses', Project Deliverable C 'Detailed Scientific Report' - S. Baatout, M.A. Benotmane, K. Abou-El-Ardat, A. Aerts, M. Beck, L. de Saint- Georges, H. Derradji, P. Jacquet, P. Monsieurs, R. Quintens, C. Rombouts, N. Samari. FP7 EU DoReMi annual activity report 2010. January 29, 2011 - S. Baatout, K. Abou-El-Ardat, A. Aerts, M.A. Benotmane, H. Derradji, H. el- Saghire, P. Jacquet, R. Quintens, C. Rombouts, M. Beck, L. de Saint-Georges, N. Impens, P. Monsieurs, N. Samari. FP7 EU DoReMi annual activity report 2011. July 24, 2011

234