CHARACTERIZATION OF URACIL DNA GLYCOSYLASE AS A THERAPEUTIC TARGET FOR SENSITIZATION OF FLOXURIDINE IN CANCER WITH P53 MUTATION OR DEFICIENCY

by

YAN YAN

Submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy

Thesis Advisor: Dr. Stanton L. Gerson, M.D.

Department of Pharmacology

CASE WESTERN RESERVE UNIVERSITY

August, 2017

CASE WESTERN RESERVE UNIVERSITY SCHOOL OF GRADUATE STUDIES

We hereby approve the thesis/dissertation of

Yan Yan

candidate for the degree of Doctor of Philosophy *.

Committee Chair Youwei Zhang

Committee Member Stanton Gerson

Committee Member Maria Hatzoglou

Committee Member Derek Taylor

Committee Member Alexandru Almasan

Date of defense April 24th, 2017

*We also certify that written approval has been obtained for any proprietary material contained therein.

Page | ii TABLE OF CONTENTS

Table of Contents…………………………….…………………………………………...iii

List of Tables……………………………………….………………….…………………vii

List of Figures…………………….……………………………….…….………………viii

Acknowledgements………………………………………………………….……………xi

List of Abbreviations…………………………………………………….……….………xii

Abstract……………………………………………………………………………………1

Chapter 1. Introduction……………………………………………………………….……4

1.1 TS inhibitors and the directed effects………………………………………………4

1.1.1 TS inhibition by fluoropyrimidines………………………...…………….….6

1.1.2 TS inhibition by antifolates…………………………………………...……...8

1.1.3 Thymineless death…………………………………………………………...9

1.2 Impact of BER repair pathway on TS inhibitors……………...…….…………….10

1.2.1 Uracil and 5-FU excision by DNA glycosylases……………………………13

1.2.2 Downstream components of BER………………………………….……….16

1.3 Molecular determinants of response to TS inhibitors…………………………….19

1.3.1 Thymidylate synthase………………………………………………………19

1.3.2 p53………………………………………………………………………….20

1.3.3 dUTPase……………………………………………………………………22

1.3.4 Other DNA repair pathways…….……………………….…….…….….….24

Page | iii 1.4 Tumor suppressor p53 and its regulation…………………………………………27

1.5 The role of p53 in apoptosis…………………………………………….………...29

1.6 p53 germline and somatic mutations…………….………………….…………...30

1.7 Mutant p53 gain of functions and their link to chemoresistance………....……….31

1.8 Statement of objectives…………………………………………………………...35

Chapter 2. Inhibition of uracil DNA glycosylase (UDG) sensitizes cancer cells to 5- fluorodeoxyuridine through replication fork collapse-induced DNA damage……...... …38

2.1 Abstract…………………………………….……………...………………….….39

2.2 Introduction……………………………………………….……………………...40

2.3 Materials and methods………………….………………………………………...42

2.4 Results……………………………………………………………………………48

2.4.1 UDG removes uracil and 5-FU incorporated into DNA following 5-FdU

treatment…………………………………………………………………………48

2.4.2 Loss of UDG enhances cytotoxicity of 5-FdU in cancer cells...... 50

2.4.3 Thymidine treatment after 5-FdU exposure cannot fully rescue the enhanced

cytotoxicity in UDG depleted cells due to the retention of genomic uracil and 5-

FU.……………………………………………………………………………….51

2.4.4 UDG depletion leads to cell cycle arrest at late G1 and early S phase by 5-

FdU…...... 52

Page | iv 2.4.5 Loss of UDG inhibits DNA replication progression in response to 5-FdU

treatment…………………………………………………………………………53

2.4.6 DNA damage persists in UDG depleted cells and is not due to apoptosis by

5-FdU treatment………………………………………………………………….54

2.5 Discussion………………………………….…………………………………….56

Chapter 3. Knockdown of uracil DNA glycosylase selectively re-sensitizes p53 mutant and deficient human cancer cells to 5-fluorodeoxyuridine………………………….…....72

3.1 Abstract…………………………………….……………...………………….….73

3.2 Introduction………………………………………………………………………74

3.3 Materials and methods………………….………………………………………...77

3.4 Results……………………………………………………………………………80

3.4.1 p53 mutation or deficiency affords 5-FdU resistance among different types

of cancer cells……………………………………………………………….……80

3.4.2 UDG depletion sensitizes cancer cells with p53 mutation or deficiency to 5-

FdU exposure…………………………………………………………………….81

3.4.3 5-FdU resistance in p53 knockout (KO) or knockdown (KD) cells is reversed

by UDG depletion………………………………………………………….…….82

3.4.4 UDG depletion selectively sensitizes p53 KO cancer cells to pemetrexed and

5-FU...... 83

Page | v 3.4.5 5-FdU activates cell death in p53 KO cancer cells with depleted

UDG………….………….……………………………………………………….84

3.5 Discussion……………………………….……………………………………….85

Chapter 4. Conclusions and future directions…………………………………………...101

4.1 Conclusions…………………….……………………………………………….101

4.2 Future directions………………………………………………………………...103

4.2.1 Evaluation of the interplay between p53 and thymidylate synthase (TS) on

cytotoxicity of TS inhibitors…………………………………………………….103

4.2.2 Characterization of the effect of other glycosylases (SMUG1, TDG, MBD4)

in processing genomic uracil and 5-FU induced by 5-fluorodeoxyuridine...…....106

4.2.3 Exploitation of DNA damage response signaling as a target to achieve

‘synthetic sickness or lethality’ with UDG depletion in p53 mutant cancer cells

after DNA damage……….……………………………………………………...114

4.2.4 Exploration of the potential gain of functions for p53 mutations that are

responsible for 5-FdU resistance……………….……………………………….118

4.2.5 Identification of small molecule inhibitors of UDG via high-throughput

screening.……………………………………………………………………….120

Appendix…………………………………………………………………………….….122

Reference……………………………………………………………………………….128

Page | vi LIST OF TABLES

Table 1-1. Summary of uracil and 5-FU lesions repaired by mammalian DNA glycosylases…………………………….……………………………………………….14

Table 1-2. Role of downstream BER proteins in response to TS inhibitors………………18

Table 1-3. Select list of mutant p53 upregulated genes and chemoresistance…………….34

Table 3-1. Cell lines and strains used in this work…………………………….….………88

Page | vii LIST OF FIGURES

Figure 1-1. Thymidylate synthesis and two classes of TS inhibitors……….…...………….5

Figure 1-2. Metabolism of fluoropyrimidines…………………….……………………….7

Figure 1-3. Base excision repair (BER) of mis-incorporated uracil (U) and 5-FU……...12

Figure 1-4. dUTPase prevents dUTP/FdUTP from incorporation into DNA during TS inhibition…………………………………………………………………………………23

Figure 1-5. Mismatch repair (MMR) of mis-incorporated 5-FU………………………….26

Figure 2-1. UDG depletion causes incorporation of uracil and 5-FU into genomic DNA by

5-FdU…………………………………………………………………………………….60

Figure 2-2. UDG depletion enhances 5-FdU sensitivity in cancer cells...... 62

Figure 2.3. Thymidine treatment after 5-FdU exposure cannot fully rescue increased cytotoxicity in UDG depleted cells……………………………………………………….64

Figure 2-4. Loss of UDG induces cell cycle arrest at late G1 and early S phase by 5-FdU exposure………………………………………………………………………………….66

Figure 2-5. UDG depletion inhibits replication fork progression following 5-FdU treatment...... 68

Figure 2-6. DNA damage accumulates in UDG depleted cells in a caspase independent manner. …………………………………………………………….…………………….70

Page | viii Figure 3-1. 5-FdU resistance in different types of cancer cells with p53 mutation or deficiency………………………………………………………………………………...89

Figure 3-2. UDG depletion selectively sensitizes cells with p53 mutation or deficiency to

5-FdU…………………………………………………………………………………….91

Figure 3-3. 5-FdU resistance due to loss of p53 is reversed by UDG depletion…………...93

Figure 3-4. p53 knockdown re-sensitizes cancer cells with UDG depletion to 5-FdU……95

Figure 3-5. UDG depletion selectively sensitizes p53 KO cells to pemetrexed and 5-

FU...... 97

Figure 3-6. UDG depletion induces cell death caused by 5-FdU in p53 KO cancer cells…99

Figure 4-1. Evaluation of uracil and 5-FU levels in TDG treated genomic DNA extracted following 5-FdU exposure via AP site detection assay………………………………….109

Figure 4-2: Evaluation of UDG and TDG activities via glycosylase activity assay...…...111

Figure 4-3. Synthetic sickness or lethality between DNA damage response signaling and p53………………………………………………………………………………….….117

Page | ix Appendix 1. Retention of uracil and 5-FU in HEC1A UDG depleted cells during thymidine recovery following 5-FdU exposure…………………………………………………….123

Appendix 2. Loss of UDG induces HEC1A cell cycle arrest at late G1 and early S phase following 5-FdU exposure………………………………………………………………124

Appendix 3. DNA damage accumulates in HEC1A UDG depleted cells in a caspase independent manner…………………………………………………………………….125

Appendix 4. Effect of irradiation on p21 induction in various cancer cells with different p53 status………………………………….…………………………………………….127

Page | x ACKNOWLEDGEMENTS

First and foremost, it has been a privilege to have Dr. Stanton Gerson as my mentor.

I would like to express my deepest gratitude for his patient guidance, immense knowledge,

and gracious support throughout the course of my project and my career as a scientist.

I would also like to express my sincere appreciation to my committee members, Dr.

Youwei Zhang, Dr. Maria Hatzoglou, Dr. Derek Taylor, and Dr. Alexandru Almasan, for their insightful comments, hard questions, and tremendous help alongside my Ph.D. pursuit.

For their encouragement, kindness, and collaboration, I would like to thank Xiangzi

Han, Shuming Yang, Shashank Gorityala, Yan Xu, and the past and present members of

Dr. Stanton Gerson’s group: Yulan Qing, Hua Fung, Johnathan Kenyon, Amar Desai,

Lachelle Weeks, Allison Condie, and Mya Nguyen.

My sincere thanks also go to the faculty and staff of Pharmacology department and

Case Comprehensive Cancer Center. I very much appreciated Dr. John Pink for his persistent help and valuable comments.

Finally, I want to thank my parents, parents-in-law, and all my family members for their unconditional love and support. Most of all I want to thank my dear husband, Dr.

Kezhen Yin, for his unwavering understanding, companion, and support during those toughest times.

Page | xi LIST OF ABBREVIATIONS

5-FU: 5-fluorouracil

5-FdU: floxuridine or 5'-fluoro-2'-deoxyuridine

5,10-CH2THF: 5,10-methylenetetrahydrofolate

AP site: apurinic/apyrimidinic/abasic site

APE: AP endonuclease

AICARFT: aminoimidazole carboxamide ribonucleotide transformylase

BER: base excision repair

BrdU: 5-bromo-2-deoxyuridine

CldU: 5-chloro-2-deoxyuridine

DHF: dihydrofolate

DHFR: dihydrofolate reductase dNTP: deoxyribonucleotide triphosphates

DSB: DNA double strand breaks dTMP: deoxythymidine monophosphate dTTP: deoxythymidine triphosphate dUMP: deoxyuridine monophosphate dUTP: deoxyuridine triphosphate

Page | xii dUTPase: deoxyuridine triphosphate nucleotidohydrolase

Exo1: endonuclease 1

FEN1: flap endonuclease 1

FdUMP: fluorodeoxyuridine monophosphate

FdUTP: fluorodeoxyuridine triphosphate

GARFT: glycinamide ribonucleotide formyltransferase

HNPCC: hereditary non-polyposis colorectal cancer

HR: homologous recombination

IC50: concentration inhibiting 50% of growth, survival or activity

IdU: 5-iodo-2-deoxyuridine

KD: knockdown

KO: knockout

LFS: Li-Fraumeni Syndrome

MBD4: methyl CpG binding domain protein 4

MEF: mouse embryonic fibroblast

MMR: mismatch repair mRNA: messenger RNA

MSI: microsatellite instability

Page | xiii NHEJ: non-homologous end joining

NIR: near infrared

OPRT: orotate phosphoribosyltransferase

PARP: poly(ADP-ribose) polymerase phosphate

PCNA: proliferating cell nuclear antigen

PCR: polymerase chain reaction

PI: propidium iodide

Polβ: DNA polymerase beta

RFC: replication factor C

RPA: replication protein A

RT-PCR: reverse transcriptase polymerase chain reaction

SHMT: serine hydroxymethyltransferase shRNA: short hairpin RNA

SMUG: single strand monofunctional uracil DNA glycosylase

SSB: DNA single strand breaks

TDG: thymine DNA glycosylase

THF: tetrahydrofolate

Thy: thymidine

Page | xiv Thymineless death: TLD

TS: thymidylate synthase protein

UDG: uracil DNA glycosylases (protein encoded by gene on chr12)

Ugi: bacteriophage uracil DNA glycosylase inhibitor protein

UP: uridine phosphorylase

Unt: Untreated

WT: wild-type

γ-H2AX: phosphorylated form of histone H2A.X

Page | xv Characterization of Uracil DNA Glycosylase as a Therapeutic Target for Sensitization of

Floxuridine in Cancer with P53 Mutation or Deficiency

Abstract

By

YAN YAN

Cancer is a collection of genetic diseases which occur via a step-by-step mutagenic process whereby cancer cells gain a selective growth advantage over normal epithelial cells.

Mutation in the TP53 tumor suppressor gene is one of the most frequent events in tumorigenesis, as p53 acts as a master regulator of numerous signaling pathways that drive tumor suppressive activities. In particular, mutant p53 proteins not only lose their tumor suppressor function but often endow cells with novel abilities that promote tumor progression as well as chemoresistance.

Chemotherapeutic agents such as thymidylate synthase (TS) inhibitors stand out as some of the most successful drugs in the management of cancer. Floxuridine (5- fluorodeoxyuridine, 5-FdU) is a classic TS inhibitor that can impede DNA metabolism and ultimately introduce uracil and 5-FU incorporation into the genome. According to in vitro kinetic studies, uracil DNA glycosylase (UDG) is the predominant enzyme that removes genomic uracil and 5-FU lesions. However, little is known about how uracil and 5-FU are

Page | 1 processed in cancer cells, and whether UDG is a determinant of floxuridine sensitivity that

can be exploited as a therapeutic target for potentiation of its cytotoxicity.

In the present study, we first demonstrate that UDG plays a key role in limiting

uracil and 5-FU incorporation into genomic DNA following 5-FdU treatment. The

persistent uracil and 5-FU incorporation due to loss of UDG highly enhance the

sensitization of cancer cells to 5-FdU. UDG is required for recovery from cell cycle arrest after 5-FdU, while UDG depleted cells arrest at late G1 and early S phase. Mechanistically,

5-FdU significantly delays replication fork progression in UDG depleted cells. Importantly, the sustained DNA damage, likely due to replication fork collapse, contributes to the genotoxicity of 5-FdU in the absence of UDG, which cannot be rescued by a caspase inhibitor. Together, these results indicate a critical role for UDG in regulating the cell killing effect of 5-FdU.

Since p53 has been shown previously to be a critical determinant of sensitivity to

TS inhibitors, we further interrogate the cytotoxicity of 5-FdU after UDG depletion with

regard to p53 status. By analyzing a panel of cancer cells with differing p53 status, we find

that cells with p53 mutations or deficiencies are highly resistant to 5-FdU. UDG depletion

re-sensitizes p53 mutant or deficient cancer cells to 5-FdU, whereas p53 wild-type cells

are not affected by loss of UDG. Utilizing paired HCT116 p53 wild-type (WT) and p53

knockout (KO) cell lines, we report that loss of p53 improves cell survival after 5-FdU,

and UDG depletion only sensitizes p53 KO cells. Additionally, re-sensitization is also

detected after treatment with the TS inhibitors pemetrexed and 5-FU in p53 KO cells. In

p53 WT cells the apoptosis pathway induced by 5-FdU is activated similarly regardless of

whether UDG is present or not. However, in p53 KO cells, apoptosis is compromised in

Page | 2 UDG expressing cells, but dramatically elevated in UDG depleted cells. Collectively, these findings provide evidence that UDG greatly impacts the cytotoxicity of TS inhibitors in p53 mutant and deficient cancer cells, which are commonly refractory to cell death in response to TS-targeted chemotherapy. Therefore, targeting UDG can be manipulated for the therapeutic purposes to augment the cell killing effect of TS inhibitors selectively in p53 mutant and deficient cancers.

Page | 3 CHAPTER 1. INTRODUCTION

1.1 TS inhibitors and the directed effects

Duplication of pathways is a pivotal feature of nucleotide metabolism. One notable exception to this is thymidylate synthase (TS), a “bottleneck” enzyme that provides the only pathway for de novo synthesis of deoxythymidine monophosphate (dTMP). Using

5,10-methylenetetrahydrofolate (5,10-CH2THF) as a cofactor, it catalyzes the reductive methylation of deoxyuridine monophosphate (dUMP) to deoxythymidine monophosphate

(dTMP), which can be subsequently phosphorylated to deoxythymidine triphosphate

(dTTP) (Figure 1-1) [1, 2]. TS maintains the dTTP pool essential for DNA synthesis and repair, highlighting the crucial role of this enzyme in cell proliferation and survival, especially in cancers. In particular, TS is highly expressed in a variety of tumors and demonstrates oncogene-like phenotypes when overexpressed in mammalian cells [3].

Therefore, chemotherapeutic agents that target TS have remained among the most successful drugs in the treatment of cancer.

TS is an enzyme which forms a ternary complex with its substrate dUMP and the methyl donating cofactor 5,10-CH2THF. Two structurally distinct classes of inhibitors, nucleotides or folate analogs have shown the capacity to block TS activity [2]. The class of fluoropyrimidines, such as 5-fluorouracil (5-FU) and floxuridine (5-fluorodeoxyuridine,

5-FdU), target the nucleotide binding site of TS, whereas the antifolates, such as pemetrexed, target its folate binding site (Figure 1-1).

Page | 4

Figure 1-1. Thymidylate synthesis and two classes of TS inhibitors.

A carbon group is transferred from serine to tetrahydrofolate (THF) by serine hydroxymethyltransferase (SHMT) to form 5,10-methylenetetrahydrofolate (5,10-

CH2THF). Using 5,10-CH2THF as the methyl-group donor, TS catalyzes the conversion of dUMP to dTMP. To sustain the cycle, the resultant dihydrofolate (DHF) will be reduced to THF by dihydrofolate reductase (DHFR). Fluoropyrimidines (e.g. 5- fluorouracil (5-FU) and 5-fluorodeoxyuridine (5-FdU)) and antifolates (e.g. pemetrexed) exert their actions through the inhibition of TS.

Page | 5 1.1.1 TS inhibition by fluoropyrimidines

5-FU and 5-FdU penetrate into cells rapidly by a facilitated transmembrane carrier system. Once inside the cells, they can be metabolized to two forms of active metabolites that interfere with DNA metabolism: fluorodeoxyuridine monophosphate (FdUMP) and fluorodeoxyuridine triphosphate (FdUTP) (Figure 1-2) [4, 5]. FdUMP forms an inhibitory ternary structure with TS and 5,10-CH2THF due to the inability of TS to break the fluorine- carbon bond of FdUMP [6-8]. TS inhibition by FdUMP ultimately leads to depletion of dTTP and accumulation of dUTP pools in cells. Since DNA polymerases are unable to discriminate between dUTP/FdUTP and dTTP, both dUTP and FdUTP can be utilized as nucleotides during DNA synthesis in place of thymidine, causing uracil and 5-FU mis- incorporation into DNA.

Fluoropyrimidines exposure can also result in RNA-directed cytotoxicity through the incorporation of fluorouridine triphosphate (FUTP) into different types of RNAs, disrupting RNA processing such as mRNA splicing, post-transcriptional modification of tRNA, and rRNA maturation [9-14]. The enzymes uridine phosphorylase (UP) and orotate phosphoribosyltransferase (OPRT) can transform 5-FU into its ribonucleoside or ribonucleotide versions respectively, with further phosphorylation to fluorouridine triphosphate (FUTP), a substrate for RNA polymerase (Figure 1-2). Recently, studies indicate that the cytotoxicity of 5-FU is largely RNA-based, being rescued by exogenous uridine rather than by thymidine [15-17]. However, unlike 5-FU, 5-FdU kills cells mainly by causing DNA damages (i.e., by inhibiting TS and causing the incorporation of uracil and 5-FU into genomic DNA).

Page | 6

Figure 1-2. Metabolism of fluoropyrimidines.

Once enter into cells, 5-FU is rapidly metabolized to both 5-fluorouridine (5-FUrd) and

5-fluoro-2’deoxyuridine (5-FdU) by phosphorylases (UP and TP) that add ribose or deoxyribose ring. Both 5-FUrd and 5-FdU can be further converted to mono-, di-, and tri-phosphate metabolites that are incorporated into RNA (1) and DNA (2), respectively.

5-FdUMP blocks TS and subsequently leads to increased dUMP levels with further phosphorylation to be incorporated into DNA (3).

Page | 7 1.1.2 TS inhibition by antifolates

Antifolates are a class of drugs exert their killing effects by inhibiting folate- dependent enzymes required for the biosynthesis of DNA nucleotides. Several antifolates are currently utilized in the clinic, including methotrexate, raltitrexed, and pemetrexed. Of note, pemetrexed has become the subject of substantial clinical studies and was approved by FDA as the first-line anticancer therapy in combination with cisplatin against advanced or recurrent non-small cell lung cancer [18], as well as malignant pleural mesothelioma

[19]. Currently, evaluation of the efficacy of pemetrexed-based combinatory therapies with other DNA damaging agents are under extensive clinical trials for different types of solid tumors including breast, colorectal, and pancreatic cancers.

Pemetrexed gains entry into cells via the reduced folate carrier [20]. As a folate analog, intracellular pemetrexed undergoes polyglutamation by folylpolyglutamate synthase. The polyglutamated derivatives exhibit a long half-life and potent inhibitory activity towards TS and several other folate-dependent enzymes, including dihydrofolate reductase (DHFR), aminoimidazole carboxamide ribonucleotide transformylase

(AICARFT), and glycinamide ribonucleotide transformylase (GARFT) [21, 22]. Inhibition of DHFR reduces TS activity indirectly via blocking folate recycling, while inhibition of

AICARFT and GARFT prevent purine synthesis. Specifically, since the pemetrexed- mediated cytotoxicity can be largely rescued by treatment with exogenous thymidine, inhibition of TS is considered as its predominant mechanism of action [21-24]. Similar to fluoropyrimidines, blocking TS by antifolates leads to nucleotide pool imbalance with decreased dTTP pools and increased dUTP levels, which can be eventually utilized by

DNA polymerase to produce uracil-containing DNA.

Page | 8 1.1.3 Thymineless death

Thymineless death (TLD) was first discovered in mutant E. coli with defective TS which suffers severe cell death when grown in a medium devoid of thymine. This phenomenon has also been reported in mammalian cells treated with TS-targeted chemotherapies that elicit thymine deficiency [25]. The model generated initially to explain

TLD was called “unbalanced growth” observed as ongoing protein synthesis and RNA transcription in the absence of DNA replication during thymine starvation [26]. However, this notion was too vague and the molecular mechanisms that cause the cell death still remained muddled.

DNA fragmentations in terms of the occurrence of single-strand breaks (SSBs) or double-strand breaks (DSBs) were later detected during TLD [27]. The underlying molecular event was proposed as a result of futile cycles of uracil DNA glycosylase

(UDG)-initiated base excision repair (BER) pathway which will be discussed in detail in the following section. The basis of this model is that blocking TS increases dUTP/dTTP pool ratios, leading to uracil incorporation into genome that triggers its removal by UDG.

Since dUTP/dTTP ratios still remain high, dUTP will be reinserted during DNA repair process. Thus, the repetitive cycles of uracil excision and reinsertion are proposed to cause

DNA strand breaks and ultimate cell death. However, little evidence supports this hypothesis since no effect on the cytotoxicity of TS inhibitors was observed with either overexpression or inhibition of UDG. Furthermore, DNA DSBs were observed in thymine- starved cells with the treatment of rifamipicin or chloramphenicol which can prevent TLD, indicating DSBs are not sufficient to cause TLD [27].

Page | 9 More recently, the killing mechanisms of thymine starvation is correlated with the initiation of replication. Studies in bacteria have shown that the lethality of thymine starvation is associated with new rounds of DNA replication initiation, while this death can be suppressed by inactivating DnaA protein to inhibit new initiations [27, 28]. Additionally, replication origin destruction was also reported in bacteria during thymine starvation that linked with cell death [29], but the nature of the degradation occurs preferentially at the replication origin region is unclear. Thus, more studies are required to delineate the contribution of DNA repair pathways or other proteins might lead to the formation or resolution of the DNA strand breaks during thymine starvation.

1.2 Impact of BER repair pathway on TS inhibitors

Base excision repair (BER) is the primary pathway that copes with various non- bulky small nucleobase lesions such as deamination, depurinatiation/depyrimidination, oxidation, and methylation, that can be caused by environmental chemicals, radiation, or treatment with cytotoxic drugs [30, 31]. The major steps of BER involve [32, 33]: (1) recognition and removal of a flipped-out damaged base, such as uracil or 5-FU, by a damage-specific DNA glycosylase, leaving an abasic site (AP site), (2) generation a single strand break on the sugar-phosphate backbone at the resulting AP site by an endonuclease, such as APE1, yielding a 3’hydroxyl group and a 5’dRP moiety, (3) clean-up 5’dRP moiety by dRP lyase domain of a DNA polymerase which allows further repair by either short patch BER (the predominant pathway) or long patch BER, depending on the relative concentrations of ATP or the nature of 5’dRP intermediates, (4) in short patch BER, polymerase β (Polβ) completes the gap filling by inserting a single nucleotide to DNA

Page | 10 strand. However, if the ATP concentrations is low, or 5′dRP moiety is reduced or oxidized, long patch BER will be initiated so that the gap filling is catalyzed by polymerase

δ/polymerase ε (Polδ/Polε) followed with the removal of at least two flapped damaged nucleotides by flap endonuclease 1 (FEN1), (5) sealing of the DNA nick by a DNA ligase,

LIG3 or LIG1 for short or long patch repair respectively, completing the repair pathway.

In the process, the binding of poly(ADP-ribose) polymerase 1 (PARP1) to DNA strand breaks will stimulate its activation, which is necessary for the recruitment of XRCC1 to provide the scaffold for both short patch and long patch BER. The fundamental molecular events and the predominant mammalian proteins participated in BER are depicted in Figure

1-3. As many of the clinically used chemotherapeutic agents are competent to induce DNA damages which are often repaired by BER pathway, it has been shown that inhibition of a particular BER protein can augment the efficacy and minimize the resistance of current chemotherapeutic regimes [34].

Page | 11

Figure 1-3. Base excision repair (BER) of mis-incorporated uracil (U) and 5-FU.

In brief, mis-incorporated uracil (U) and 5-FU following exposure to TS inhibitors are removed by uracil DNA glycosylase (UDG), creating an abasic site (AP site). AP sites are then cleaved by AP endonuclease I (APE1) to generate single strand breaks. The downstream pathways can be divided into either short patch (predominant repair mode) or long patch repair pathway. PARP-1 is localized to the site of DNA damages so as to recruit XRCC1 and other BER repair proteins. DNA polymerase (Polβ or Polδ/Polε) initiates the nucleotides re-synthesis to the single strand gap. DNA ligase (LIG3 or

LIG1) finally seals the gap and completes the repair.

Page | 12 1.2.1 Uracil and 5-FU excision by DNA glycosylases

TS inhibition by fluoropyrimidines and antifolates results in a high intracellular concentration of dUTP or 5-FdUTP, which can be mis-incorporated into DNA. Repair of uracil- and 5-FU-containing DNA is predominantly conducted by glycosylases initiated

BER pathway. In mammalian glycosylases family, there are four members listed in Table

1-1 that exhibit substrate specificity to both uracil and 5-FU [35]: uracil DNA glycosylase

(UDG) (alternative splicing leads to two different isoforms: the mitochondrial UDG1 and the nuclear UDG2), single-strand-selective monofunctional uracil-DNA glycosylase

(SMUG1), G/T mismatch-specific thymine DNA glycosylase (TDG), and methyl-CpG- binding domain protein 4 (MBD4 or MED1). Most notably, according to in vitro kinetic analysis among different glycosylases, UDG has the highest turnover rate for removing both uracil and 5-FU from DNA irrespective of whether they are paired with adenine or guanine [36]. In addition, studies reported that, by using its N-terminus to interact with replication protein A (RPA) and proliferating cell nuclear antigen (PCNA), UDG comes along with the rapid moving replication foci and excises incorporated uracil and 5-FU in a post-replicative manner [37, 38]. Although glycosylases initiated BER account for the dominant activity that excises genomic uracil and 5-FU from DNA, whether BER events are the downstream determinants of the cytotoxicity of TS inhibitors still remain controversial.

Page | 13 Table 1-1. Summary of uracil and 5-FU lesions repaired by mammalian DNA glycosylases

Glycosylases Substrates UDG uracil DNA glycosylase U, 5-FU in ss and dsDNA, alloxan, 5- hydroxyuracil, isodialuric acid

SMUG1 single-strand-selective monofunctional 5-hmU, U:G > U:A > uracil-DNA glycosylase ssU, 5-FU, εC in ss and dsDNA

TDG G/T mismatch-specific thymine DNA U:G > T:G, 5-hmU in glycosylase dsDNA, 5-FU

MBD4 methyl-CpG-binding domain protein 4 U:G and T:G, 5-hmU in (MED1) CpG context, εC, 5-FU in dsDNA

Abbreviations in this table: 5-hmU (5-hydroxymethyluracil), A (adenine), C (cytosine), dsDNA (double strand DNA), G (guanine), ssDNA (single strand DNA), T (thymidine), U (uracil)

Page | 14 Previously, it was hypothesized that thymine-less induced futile cycles of mis- incorporation, excision, and repair ultimately lead to DNA strand breaks and cell death [39,

40]. If the futile cycles of cell death were dependent on UDG mediated excision of uracil and 5-FU, UDG expression levels would determine the cytotoxicity of TS inhibitors.

However, several studies using various approaches do not appear to support this correlation.

For instance, a study reported that the increased expression and activity of UDG does not affect the cytotoxicity of antifolate TS inhibitors, ZD9331 and raltitrexed [41]. Several other studies using UDG+/+ and UDG-/- mouse embryonic fibroblasts (MEF) found that there are no significant differences in terms of cell death observed after treatment with fluoropyrimidine TS inhibitors, 5-FU and 5-FdU [42, 43]. Additionally, blocking UDG by expression a protein inhibitor (Ugi) in HEK293 cells did not precipitate the cytotoxicity of raltitrexed, 5-FU, and 5-FdU [44]. Although the increased dUTP and 5-FdUTP pools were detected following treatment with raltitrexed or 5-FdU, the differential responses were also not perceived in chicken DT40 B cells with or without UDG expression [36]. In particular, it is important to point out that 5-FU has been shown to cause cell damage mainly in RNAs; thus UDG which works on DNA may not necessarily determine the cytotoxicity of 5-FU

[45]. In contrast to these studies, both our and Karnitz group using cancer cell lines reported that UDG depletion by shRNA or siRNA greatly augments the cytotoxicity of pemetrexed or 5-FdU [46, 47]. This enhanced killing effect via UDG depletion was associated with increased incorporation of both uracil and 5-FU into genomic DNA [47]. Moreover, our group also analyzed a panel of cancer cell lines with inherently different p53 status, and found the potentiated cytotoxicity of TS inhibitors in cells lacking UDG is highly reliant on p53 status, which will be discussed in detail in Chapter 3.

Page | 15 In addition to UDG, the discrepant findings were also observed with other DNA glycosylases. For example, in the MEF cell models, the expression of SMUG1 was reported to protect cells against 5-FU induced cytotoxicity [48], whereas the expression of

TDG or MBD4 precipitated the cytotoxicity of 5-FU [49, 50]. Mechanistically, the enhanced cytotoxicity caused by TDG was explained due to its slow dissociation from AP sites after uracil and 5-FU removal, thus blocking downstream BER pathway [49]. The

MBD4 triggered cytotoxicity was proposed as a result of interaction with MLH1, a key component of mismatch repair (MMR) pathway which can signal to cell death in response to DNA damage [4]. Indeed, future studies aimed at elucidating the detailed mechanisms of how expression of TDG or MBD4 contributes to the cytotoxicity of TS inhibitors are necessary.

1.2.2 Downstream components of BER

Since base lesions generated by TS inhibition are chiefly repaired by BER pathway, the impact of key BER proteins downstream of glycosylases on cytotoxicity of TS inhibitors was also evaluated by several studies, which are summarized in Table 1-2.

The first BER component following glycosylases is AP endonuclease I (APE1), which accounts for the major AP endonuclease activity in mammalian cells [51]. A study reported that expression of a dominant negative APE1 in Chinese hamster ovary (CHO) cells significantly enhances the cell killing effect of 5-FU and 5-FdU [52]. Furthermore, knockdown of APE1 expression by siRNA also augmented the cytotoxicity of 5-FdU in colon cancer cells, indicating BER pathway protects cells from TS inhibition induced DNA damages [53].

Page | 16 Following the single strand breaks generated by APE1, more downstream repair proteins such as Pol β, XRCC1, PARP, and FEN1 were also assessed for their link with the response to TS inhibitors. DNA polymerase β (Pol β) is a non-replicative polymerase that exhibits two biochemical functions in both short and long patch BER: 5’dRP excision and nucleotide gap filling. Surprisingly, studies in Pol β-/- MEFs found that loss of Pol β renders cells more resistant to raltitrexed and 5-FdU in comparison with wild-type MEFs

[54, 55]. Similar as glycosylases, inconsistent results were also described in cells lacking

XRCC1, a scaffold protein interacts with Pol β, DNA ligase III, and PARP during base excision repair. Hamster cells bearing XRCC1 deficiency displayed no significant changes in response to raltitrexed [54] and 5-FU [56] when compared to the wild-type counterparts.

However, depletion of XRCC1 by siRNA in OVCAR8 cells increased the cell killing effect of 5-FdU [53]. Moreover, studies on PARP revealed that PARP inhibition by siRNA or

PARP inhibitor(s) AZD2281 or ABT-888 profoundly synergized with 5-FdU but not 5-FU in multiple cancer cells [53, 57], suggesting 5-FU exerts its cytotoxicity largely by introducing RNA damages. Recently, studies also revealed that knockdown of FEN1 by siRNA or expression a mutant FEN1 sensitizes cancer cells to 5-FU treatment [58, 59].

Collectively, the vast majority of studies supported an important role of BER pathways to combat with the DNA damages caused by TS inhibition. Some of the differential responses might be due to the altered choice for BER sub-pathways on the cell type examined [60], and the detailed contribution of each pathway to the repair of TS inhibition induced DNA damages needs to be further investigated.

Page | 17 Table 1-2. Role of downstream BER proteins in response to TS inhibitors.

BER BER sub- Cell manipulation TS inhibitors Response Ref protein pathway

APE1 upstream of expression of 5-FU, 5-FdU sensitive 52, 53 sub- dominant negative pathways APE1 protein (ED) in CHO cells, KD of APE1 by siRNA in HT29 cells

Pol β short- & Pol β -/- MEF Ratiltrexed, resistant 54, 55 long-patch 5-FdU

XRCC1 short- & Mutant hamster cells Ratiltrexed, no 54, 56 long-patch with XRCC1 5-FU difference deficiency, KD of XRCC1 in OVCAR8 and HT29 cells by siRNA

XRCC1 short- & KD of XRCC1 in 5-FdU sensitive 53 long-patch OVCAR8 and HT29 cells by siRNA

PARP short- & KD of PARP by 5-FU no 53, 57 long-patch siRNA, or inhibition difference of PARP by PARP inhibitors AZD2281 or ABT-888 in multiple cancer cells

PARP short- & KD of PARP by 5-FdU sensitive 53, 57 long-patch siRNA, or inhibition of PARP by PARP inhibitors AZD2281 or ABT-888 in multiple cancer cells

FEN1 long-patch KD of FEN1 by 5-FU sensitive 58, 59 siRNA in MCF7 cells, expression of mutant FEN1 in SW480 cells

Abbreviations in this table: Chinese hamster ovary cells (CHO), knockdown (KD)

Page | 18 1.3 Molecular determinants of response to TS inhibitors

An extensive area of the basic and clinical investigation is to characterize the biological factors that can impact the cellular response to TS inhibitors. Insights into these molecular mechanisms not only build up a pivotal step to predict those patients who are most likely to benefit from TS-targeted chemotherapy but also provide the strategies to overcome chemoresistance. This section describes the changes in several central elements that are capable of rendering cell resistance to TS inhibitors. These encompass alterations of thymidylate synthase, loss of wild-type p53, dUTPase overexpression, deficiencies in mismatch repair pathway, and involvement of double strand breaks repair pathway.

1.3.1 Thymidylate synthase

Thymidylate synthase (TS) is a key determinant of the sensitivity to TS inhibitors.

Numerous studies have demonstrated that increased TS protein expression due to gene amplification is an important mechanism of resistance to TS inhibitors [61-63].

Interestingly, since TS gene promoter is polymorphic which can appear either as a double or triple tandem repeat of a 28-bp sequence, studies reported that transcription of TS under the promoter that contains a triple repeat generates higher levels of mRNA, which explains the clinical studies that patients utilizing the promoter that is homozygous for the triple repeat are less likely to respond to 5-FU treatment [64-67]. Moreover, an acute increase in

TS enzyme after exposure to both fluoropyrimidines and antifolates has been observed [68,

69]. Indeed, this induction was attributed to the auto-regulatory mechanism for TS, as the ligand-free TS protein can bind to TS mRNA, limiting its own translation [70, 71].

According to this paradigm, when TS is bound with its inhibitors, the restrained complex

Page | 19 formed between TS and its mRNA will be disrupted, releasing TS mRNA for translation

[72]. However, studies with TS mutation at residue 303 (P303L) located at the protein- message-binding site failed to increase TS translation; rather, it significantly reduced the half-life of TS that lead to the resistance to TS inhibitors [73, 74]. Besides Pro303-to-Leu substitution, other TS mutations in cell culture models have also been described that responsible for resistance to TS inhibitors. For instance, a Tyr-to-His substitution at residue

33 of TS was defined that confers fluoropyrimidines resistance due to its reduced affinity to FdUMP [75-77]. However, to date, no mutations in TS have been identified in patient tumors undergoing TS-targeted chemotherapy.

1.3.2 p53

p53 is one of the most extensively studied tumor suppressors. It maintains DNA integrity and halts cell proliferation by inducing cell-cycle arrest and apoptosis in response to DNA damage. Loss of wild-type p53 protein has been shown to be associated with resistance to TS inhibitors [1, 78]. For example, a study reported that disruption of both alleles of p53 in HCT116 cancer cells endows the cells remarkably refractory to apoptosis induced by 5-FU [79]. Furthermore, the resistance to 5-FU was also observed in various types of cancer cells bearing p53 mutations [80-83]. In support of these pre-clinical studies, a number of clinical findings with tumor specimens derived from patients have correlated the p53 overexpression, a surrogate marker for p53 mutations, to 5-FU resistance [84-86].

However, there are conflicting results against this correlation, and this inconsistency might be, in part, due to the variations in immunohistochemical protocol, especially the utilization

Page | 20 of different p53 antibodies, making it difficult to draw conclusions of its clinical value as a predictive marker.

Using cell models, mechanisms of resistance to TS inhibitors in cells lacking wild- type p53 have been reported. As a number of studies described that p53 directly activates the transcription of a series of apoptotic genes including FAS/APO1, and PUMA [87, 88], a link has been established between the limited efficacy of TS inhibitors with the impaired induction of apoptotic protein as a result of the p53 loss. For example, studies revealed that expression of Fas, a cell surface protein capable of triggering apoptosis upon ligand binding, is greatly down-regulated in p53 mutant cancer cells following 5-FU and raltitrexed treatment when compared with p53 wild-type cells, and expression of a p53 wild-type vector in p53 mutant cells significantly elevates Fas expression in response to 5-FU [89,

90]. Inhibition of Fas by siRNA in p53 wild-type cells markedly reduced 5-FU-mediated

DNA fragmentation and caspase activity [91]. Furthermore, mRNA levels of PUMA, one of the most potent killers among Bcl-2 family members, was also induced in p53 wild-type cell after treatment with 5-FU, while this induction was dramatically abolished when p53 is inactive [92]. Interestingly, several studies also characterized the upstream effectors that are responsible for the activation of p53 following treatment with 5-FU. One study found that 5-FU treatment stimulates extracellular Ca2+ influx into the cells, activating p53 by post-translational phosphorylation through calmodulin activity [93]. Another study which emphasizes the RNA stress caused by 5-FU described that 5-FU treatment triggers the induction of ribosome-free L5, L11, and L23 ribosomal proteins, which interact with

MDM2, and subsequently lead to stabilization of p53 [94].

Page | 21 1.3.3 dUTPase

As discussed previously, aberrant nucleotides dUTP and FdUTP can be incorporated into genomic DNA as a consequence of treatment with fluoropyrimidines or antifolates, causing DNA damages. Studies with the enzyme deoxyuridine 5’-triphosphate nucleotidohydrolase (dUTPase) reported that it is capable of catalyzing the conversion of the triphosphate nucleotides dUTP and FdUTP to their monophosphate versions, protecting cells from abnormal mis-incorporation of dUTP and FdUTP during TS inhibition (Figure

1-4) [95, 96]. Importantly, dUTPase overexpression has been observed in a number of cancers, including colorectal, lung, and breast cancer [97-100], and thus blocking this enzyme provides the chance to improve the efficacy of TS-targeted therapies. In consistent with this hypothesis, inhibition of dUTPase by siRNA or small molecules in different types of cancer cells has been shown to augment the cytotoxicity of TS inhibitors both in fluoropyrimidines and antifolates, causing potentiated uracil and 5-FU mis-incorporation and subsequent synthetic lethality [97, 98, 101-104]. In 2014, one of a potent uracil- derivatives blocking dUTPase termed TAS-114 has been first investigated in human, which demonstrated encouraging pharmacokinetic characteristics. TAS-114 is now entering the clinical trial in combination with fluoropyrimidines to treat patients with advanced solid tumors [105-107].

Page | 22

Figure 1-4. dUTPase prevents dUTP/FdUTP from incorporation into DNA during

TS inhibition.

Inhibition of TS by fluoropyrimidines and antifolates blocks the formation of dTMP from dUMP, resulting in a corresponding accumulation of dUMP. Fluoropyrimidines can also be directly metabolized to FdUMP. Both dUMP and FdUMP can be further phosphorylated to dUTP and FdUTP by uridine monophosphate-cytidine monophosphate kinase/nucleoside diphosphate kinase (UMP-CMPK/NDK). In the context, dUTPase catalyzes the reversion of dUTP and FdUTP to their monophosphate structures, preventing the appearance of uracil and 5-FU in DNA, and thus protects the cells from TS inhibition induced DNA damages. TAS-114, a novel dUTPase inhibitor is currently under clinical trials in combination with fluoropyrimidines to treat patient tumors.

Page | 23 1.3.4 Other DNA repair pathways

Mismatch repair (MMR) pathway corrects base:base mismatches and insertion/deletion mispairs caused during DNA replication or recombination. It has been documented that deficient MMR is linked with genomic instability, defects in meiosis, predisposition to certain types of cancer such as hereditary non-polyposis colorectal cancer

(HNPCC), and resistance to chemotherapeutic agents including temozolomide, platinum, and some nucleoside analogs. Biochemically, a fraudulent mismatch or deletion and insertion is recognized by either MSH2:MSH6 or MSH2:MSH3 heterodimers.

Subsequently, downstream processing complex MLH1:PMS2 is recruited, which not only signals to endonuclease 1 (Exo1) to excise the mispaired base on the daughter strand, but also recruits other repair components such as replication protein A (RPA), proliferating cell nuclear antigen (PCNA), replication factor C (RFC), and DNA polymerases to resynthesize and complete the repair (Figure 1-5) [108, 109]. In particular, MMR can also signal to apoptotic pathway upon recognition of DNA damage [4, 110]. Therefore, defects of MMR in genetic or epigenetic levels will be tolerable to DNA damage, causing resistance to DNA damaging agents.

By using a biochemical assay, it has been described that human MSH2:MSH6 complex can recognize 5-FU in the DNA when mispaired with guanine [111], and cell extracts from MMR proficient cells can efficiently repair 5-FU:G mismatches [112].

Interestingly, multiple studies have identified that deficiencies in MMR component of

MSH2 or MLH1 are resistant to 5-FU and 5-FdU in both human and mouse cell lines [111,

113, 114]. These results are consistent with the findings that MMR can sense the DNA lesions and trigger apoptosis. In addition, following treatment with 5-FdU, studies also

Page | 24 revealed that higher levels of 5-FU:G pairs assessed by incubation with MBD4 enzyme are

observed in the DNA of MMR deficient cells when compared with their MMR competent

counterparts, indicating the presence of 5-FU:G mispairs are being tolerated as a result of

loss of MMR [111]. However, there is an ambiguous aspect of these studies regarding 5-

FU incorporation following fluoropyrimidines. Since polymerases generally insert FdUTP

opposite adenine, and cytosine deamination generates U:G mispairs, it is unclear how

genomic 5-FU:G mismatches appear after treatment with fluoropyrimidines.

As noted above, MMR is involved in determining the sensitivity of fluoropyrimidines in human cell lines. Some studies have reported that patients with tumors proficient in MMR are associated with improved survival from 5-FU based adjuvant chemotherapy [115, 116]. Nevertheless, some other studies demonstrated that there is no obvious difference in terms of 5-FU response among patients with either low or high microsatellite instability (MSI), a hypermutable phenotype results from MMR deficiency.

[117, 118]. It is important to point out that other determinants of 5-FU treatment including

BER components, p53 status, and dUTPase activity were not evaluated simultaneously in above models, which makes it difficult to characterize the impact of MMR on the response to 5-FU adjuvant therapy.

Page | 25

Figure 1-5. Mismatch repair (MMR) of mis-incorporated 5-FU

The introduction of a mispaired base (e.g. 5-FU:G) is recognized by the MSH2-MSH6

heterodimer, whereas insertion or deletion is recognized by the MSH2-MSH3 heterodimer. The bindings of heterodimers with damaged sites result in the recruitment

of PMS2 and MLH1-MLH3 heterodimer, which will further recruit nuclease Exo1 to excise the mis-incorporated base. Other repair components including replication protein

A (RPA), proliferating cell nuclear antigen (PCNA), replication factor C (RFC), and

DNA polymerase δ or ε (Pol δ or Pol ε) are involved in nucleotide resynthesis, and DNA ligase finally seals the DNA, completing the repair.

Page | 26 DNA double strand breaks (DSBs) are considered as the most toxic lesions because a single unrepaired DSB is able to induce cell death. In eukaryotic cells, there are two main

DSBs repair pathways, namely non-homologous end-joining (NHEJ), and homologous recombination (HR). Previously, it has been demonstrated that HR is activated to resolve stalled replication forks [119, 120], and sister chromatid exchange mediated by HR has been observed in cells treated with antifolate TS inhibitor raltitrexed [121]. However, implications of DSBs repair pathway following TS inhibition are not as widely studied as

BER and MMR pathways. A couple of studies found that using siRNA to deplete RAD51, an essential component of HR, augments the sensitivity to raltitrexed [122]. Conversely, overexpression of RAD51 conferred raltitrexed resistance [123]. Moreover, a study described that expression of a mutant RAD51 (T309A), acting as a dominant negative protein, sensitizes cells to raltitrexed [123]. Given that the unresolved strand break intermediates generated during the repair after exposing to TS inhibitors are likely to be processed by HR, it is reasonable to examine the components of HR more carefully in the context of TS inhibition.

1.4 Tumor suppressor p53 and its regulation

p53 is a tumor suppressor which acts as a central node downstream of a myriad of different stress signals to suppress cancer by triggering antiproliferative programs such as cell cycle arrest and apoptosis. Currently, the main cellular stress identified to activate p53 includes DNA damage, oncogene expression, hypoxia, nutrient scarcity, ribosome dysfunction, and telomere shortening [124]. Under conditions of relatively low levels of stress, p53 drives a series of protective events such as cell cycle arrest, DNA repair, and

Page | 27 production of antioxidant proteins to maintain genome integrity and cell survival. However, in response to the prolonged or extensive stress signals with irreparable damages, p53 elicits cell death through apoptosis or senescence programs, for the ultimate benefit of the whole organism. As a transcription factor, p53 can either enhance or repress the expression of numerous genes and microRNAs [125, 126]. Beyond its key role in transcription, p53 can also direct its interaction with cytoplasmic proteins, such as effectors in apoptotic pathway [127].

There are two key models, protein stabilization and post-translational modification briefly summarized here that can regulate p53 activation in response to stress signals [128-

130]. p53 stabilization is mainly achieved under circumstances that interrupt its interaction with the ubiquitin ligase MDM2, a negative regulator that not only targets p53 for degradation but also inhibits p53 activity through occluding its transcriptional activation domain. In particular, p53 itself also positively regulates MDM2 by increasing its transcription, allowing for an autoregulatory negative feedback mechanism to attenuate p53 activation. The critical role of MDM2 in regulating p53 has been shown that MDM2 deficient mice die during very early embryonic development. This phenotype is attributed to the unrestrained p53 activity, as the lethality can be fully rescued by codeletion of p53

[131]. Importantly, p53 is also subjected to multiple facets of post-translational modifications, including phosphorylation, acetylation, methylation, ubiquitylation, glycosylation, sumoylation, hydroxylation, and ADP ribosylation [132]. For example, phosphorylation is generally considered as the first crucial step of p53 stabilization or activation, which are undertaken by a number of kinases, such as ATM, ATR, Chk1/Chk2, p38, JNK, and others. Acetylation and ubiquitination have been shown to be mutually

Page | 28 exclusive, as acetylation is considered as the blockage of p53 ubiquitination by MDM2.

Currently, there are approximately 50 individual modifications identified throughout the p53 peptide, which function in cooperation with the interaction of p53 protein-binding partners to control the transcriptional dependent and independent functions of p53. And the ability of p53 to integrate the intricate constellation of post-translational modifications allows it to respond to various types of stress signals adequately.

1.5 The role of p53 in apoptosis

One of the most widely studied areas in p53 surrounds its ability to control apoptosis. In many cell types, apoptosis occurs through two major pathways, the intrinsic mitochondrial or extrinsic death receptor pathway, both of which can be modulated by p53

[133]. It has been demonstrated p53 drives irreversible apoptosis by triggering the transcriptional induction of proapoptotic genes involved in both intrinsic and extrinsic death pathways, such as BAX, FAS, PUMA, NOXA and others [125, 129]. Alternatively, p53 can also repress the transcription of antiapoptotic genes, such as survivin, thus promoting cell death [134].

Despite its well-known transcriptional activities in apoptosis activation, a novel transcriptionally independent proapoptotic pathway mediated by cytoplasmic p53 has been identified [127, 135]. Mechanistic insights into the function of p53 in cytoplasm and/or mitochondria revealed that p53 can directly interact with members of the Bcl-2 family, which are essential in governing mitochondrial outer membrane permeabilization. For instance, studies reported that p53 induces apoptosis via either interaction and sequestration of Bcl2 and Bcl-XL to activate Bax and Bak, or directly binding to Bak to set

Page | 29 it free from its inhibitory complex mediated by Mcl-1 [136, 137]. According to protein modeling and mutagenesis function analysis, these protein-protein interactions were reliant upon the DNA binding domain of p53 [135]. Hence, the missense mutant p53 proteins which lack the ability to bind to Bcl-2 family proteins are unable to promote mitochondrial apoptosis. In addition to its actions on mitochondrial, an alternative model depicted that the nuclear and cytoplasmic activities of p53 actually cooperate in concert to activate the apoptosis pathway [124, 135]. This model reveals that under unstressed conditions, basal levels of cytosolic p53 are sequestered through interaction with Bcl-XL. In response to stress signals, the nuclear p53 first activates the transcription of PUMA, and PUMA will subsequently bind to Bcl-XL, releasing p53 to activate Bax in the cytoplasm.

1.6 p53 germline and somatic mutations

Mutation or loss of p53 is highly associated with an increased susceptibility to human cancer. Germline TP53 mutations predispose a rare type of cancer termed as Li-

Fraumeni Syndrome (LFS), a genetically heterogenous autosomal dominant disorder [138].

This syndrome is described as the early-onset of a wide spectrum of neoplasms. Somatic

TP53 mutations lead to sporadic cancer that are found in 50% of all human cancers. Of note, the majority of cancer related p53 mutations are missense mutations that localized in p53 DNA-binding domain, illuminating the importance of DNA binding for p53’s tumor suppressive activities. There are six amino acid residues termed “hot spots” (R175, G245,

R248, R249, R273, and R282) are known to be most commonly mutated [139]. These mutations either affect the direct contact of p53 to its responsive elements (contact mutations) or lead to improper folding of the protein (structure mutants). Normally p53

Page | 30 mutation results in the formation of a stable full-length protein that is unable to activate p53 targeted genes since it functions as dominant negative protein to inactivate wild-type p53 from the remaining allele. Mutant p53 proteins accumulate in the nuclear at extremely high levels due to the disruption of the negative feedback of MDM2 by mutant p53 proteins that abrogate the transcription of MDM2 for p53 degradation.

1.7 Mutant p53 gain of functions and their link to chemoresistance

TP53 is the most frequently mutated gene in human cancer [140]. Mutations in

TP53 can result in the generation of mutant p53 proteins, which display gain of functions to promote tumor survival [139]. It has been demonstrated in animal models that loss of p53 and expression of mutant p53 are not analogous, as mice with one p53 mutant allele exhibit a different and broader tumor spectrum [141, 142]. A plethora of oncogenic functions of mutant p53 have been identified in cell culture models, including the ability to drive enhanced proliferation, invasion, metastasis, angiogenesis, survival, expansion of stem cell, and tissue remodeling [143]. Because the majority of p53 mutations lie in its

DNA binding domain, indicating DNA-binding is the critical activity for functional p53, alterations in its transcriptional target genes could be important events of mutant p53 to promote tumor survival and progression [144, 145].

Since p53 mutations with gain of functions show a series of tumor promoting phenotypes, their association with chemoresistance and the underlying molecular mechanisms have also been demonstrated. Generally, the majority of studies documented that mutant p53 proteins protect cells from chemotherapy induced DNA damage through upregulation of numerous genes involved in the following (as summarized in Table 1-3):

Page | 31 (1) Drug efflux: multidrug resistance genes (MDR1 and MRP1) which belong to ATP- binding cassette subfamily B member 1 (ABCB1) have been shown to be upregulated by mutant p53 which requires an Ets-binding site [146-148]. (2) Drug metabolism: studies reported that ectopic expression of mutant p53 upregulates CYP3A4, a drug-metabolizing enzyme, and is associated with resistance to a number of chemotherapeutic agents such as etoposide, rapamycin, and elesclomol [149]. (3) Cell proliferation and survival mediated by transcriptional factors: mutant p53 enhanced the expression of NF-κB capable of promoting cell proliferation by triggering histone acetylation through recruitment of CBP and STAT2 to the promoter, and depletion of NF-κB by RNAi reduced the resistance to chemotherapeutic drugs etoposide and paclitaxel [150, 151]. In addition, another study revealed that mutant p53 can bind to the promoter of early growth response 1 (EGR1) and up-regulates its expression, which is linked to cisplatin resistance [152]. (4) Apoptosis: a study with Galectin-3 (Gal-3) found that wild-type p53 represses Gal-3 expression required for p53 mediated apoptosis, while overexpression of mutant p53 augments Gal-3 expression which correlates with cisplatin resistance [153]. (5) DNA repair: there is evidence indicated that mutant p53 enhances cisplatin resistance through upregulation of

BRCA1 [154]. Moreover, a 5′-tyrosyl DNA phosphodiesterase (TDP2) involved in the repair of DNA lesions induced by etoposide has been identified as a transcriptional target of mutant p53 [155]. (6) Others: studies demonstrated that mutant p53 is able to inactivate p73, a p53 family member, that can promote resistance to a number of drugs, including camptothecin, etoposide, cisplatin, doxorubicin, and taxol [156]. Furthermore, cells harboring p53 mutants upregulated Tim50 expression, a protein functions to transport proteins into mitochondria, and knockdown of this protein expression decreased the

Page | 32 resistance to paclitaxel [157]. Particularly, studies described that conditional expression of p53 mutants result in induced transcription of dUTPase, an enzyme involved in resistance to TS inhibitors [158]. Taken together, studies on different p53 mutations can vary widely in terms of the extent of chemoresistance in response to a defined chemo-agent, and the mechanisms of chemoresistance mediated by mutant p53 may also depend on the exact site of the mutation as well as the properties of the drug.

Page | 33 Table 1-3. Select list of mutant p53 upregulated genes and chemoresistance

Category Gene Name p53 mutants Resistant to Ref symbol drugs

Drug efflux ABCB1 ATP-binding R175H, R248Q, cisplatin 146, cassette subfamily B D281G 147, member 1; MDR1 148

Drug CYP3A4 Cytochrome P450 R282W etoposide, 149 metabolism 3A4 rapamycin, elesclomol, MK- 2206, and NVP- BEZ235

Cell NFKB2 NF-kB (p52) R175H, R273H, etoposide 150, proliferation R273C, D281G, 151 and survival R, G266E EGR1 Early growth R175H, H179E, cisplatin 152 response 1 R248W, R273H, D281G

Apoptosis LGALS3 Lectin, galactoside- R273H cisplatin 153 binding, soluble; Galectin-3

DNA repair BRAC1 BRCA1 R273H, R273C cisplatin 154

TDP2 Tyrosyl-DNA R248W etoposide 155 phosphodiesterase 2

Others TP73 p73 ND campothecin, 156 etoposide, cisplatin, doxorubicin, and taxol TIMM50 Mitochondrial R175H, R273H paclitaxel 157 import inner membrane translocase subunit

TIM50

DUT dUTPase R175H, R248W 5-FU 158

Page | 34 1.8 Statement of objectives

Chemotherapeutic agents generally cause cell death or prevent cell growth by inhibiting DNA synthesis or inducing various types of DNA damages. A finely coordinated

DNA repair pathway has evolved to cope with the corresponding DNA lesions and protect against DNA damages after exposure to a certain type of chemotherapy. Activation of

DNA repair pathways can highly contribute to resistance to DNA damaging agents.

Therefore, blocking a DNA repair pathway harbors the potential to augment the killing effect of chemotherapeutic agents, which has been the fundamental rationale for the development of inhibitors targeting DNA repair components.

Chemo-agents targeting thymidylate synthase (TS) such as fluoropyrimidines (e.g.

5-FU and floxuridine (5-FdU)) or antifolates (e.g. pemetrexed) are extensively used in the treatment of a variety of tumors. For instance, 5-FU is given as a single or combinatory agent therapy for colon cancer patients. However, the response rates for 5-FU-based adjuvant therapy are only 40-50%, highlighting the novel strategies for resistance reversal.

Mechanistically, TS inhibitors have been shown to induce DNA damages through thymidine deficiency and incorporation of aberrant bases, uracil and 5-FU, into genomic

DNA. It has been reported in vitro that base excision repair pathway (BER) initiated by uracil DNA glycosylase (UDG) is responsible for the primary activity to remove base lesions induced by TS inhibition. Thus, the studies herein hypothesized that inhibition of

UDG, which retains uracil and 5-FU incorporation into DNA, may alter the thymine-less cell death caused by 5-FdU treatment and potentiate 5-FdU cytotoxicity in cancer cells.

And this hypothesis was addressed as below:

Page | 35 First, we identified that UDG is the dominant enzyme that limits the aberrant bases

(uracil and 5-FU) from incorporation into DNA following exposure to 5-FdU. BER can be initiated by a number of different glycosylases and four glycosylases UDG, SMUG1, TDG, and MBD4 display the substrates specificity to both uracil and 5-FU. By incubation of nuclear extracts from UDG competent or depleted cells with uracil-containing oligomers, we confirmed that UDG is the major enzyme in cells that responsible for uracil removal.

5-FdU treatment was able to produce a significant induction of genomic uracil and 5-FU in UDG depleted cells, which was assessed ex vivo by using a fluorescent AP-site probe synthesized from our lab that can covalently bind to AP sites, a surrogate method for detection of a combined levels of uracil and 5-FU in DNA. In addition, we also examined the individual levels of uracil and 5-FU incorporation by using mass spectrometry, which is informative to distinguish which incorporated base is paramount in response to 5-FdU exposure.

Second, we characterized the sensitivity of 5-FdU following UDG depletion and the underlying mechanisms account for the altered cytotoxicity. UDG has been shown capable of removal base lesions uracil and 5-FU as a result of treatment with TS inhibitor

5-FdU. Thus we further addressed that uracil and 5-FU incorporation are persistently remained in cells after 5-FdU, and the retained uracil and 5-FU incorporation lead to increased sensitization of cancer cells to 5-FdU in the absence of UDG. The mechanisms of enhanced cytotoxicity of 5-FdU in UDG depleted cells were demonstrated as a result of

DNA damages likely due to DNA replication stall-induced collapse.

Finally, we also defined the role of p53 in determining the sensitivity of UDG depleted cells in response to 5-FdU. p53 is a key determinant of the sensitivity to TS

Page | 36 inhibitors. It has been shown previously that loss of UDG in MEFs does not affect the cytotoxicity of TS inhibitors, which might not be the case in the cancer cells bearing p53 mutations. By evaluating the sensitivity of 5-FdU in cancer cells with different p53 status

(p53 wild-type, p53 null, and p53 mutations), we observed that, in general, cells with p53 mutation or deficiency are strongly resistant to 5-FdU. Interesting, UDG depletion re- sensitized 5-FdU in p53 mutant or deficient cancer cells, but did not affect the cytotoxicity of 5-FdU in p53 wild-type cells, indicating the enhanced cytotoxicity caused by UDG depletion is dependent on p53 status. We also identified that the differential responses after

UDG depletion in cells containing distinct p53 status are due to p53 mediated cell death.

In total, these studies have revealed that UDG plays a central role in removing genomic uracil and 5-FU incorporation following 5-FdU treatment, and the retention of uracil and 5-FU incorporation highly increased sensitivity to 5-FdU in cancer cells, suggesting a critical role of UDG in regulating the cell killing effect of 5-FdU. Importantly, this 5-FdU resistance has been observed in p53 deficient and mutant cancer cells, and targeting UDG selectively re-sensitized p53 mutated or deficient cancer cells to 5-FdU.

Based on these data, future studies can be exploited to understand the clinic values of UDG and p53 expression as predictive marker for TS inhibitors, and development of UDG inhibitors for evaluation their activities to enhance the efficacy of TS-targeted therapies.

Page | 37 CHAPTER 2.

Inhibition of uracil DNA glycosylase sensitizes cancer cells to 5-fluorodeoxyuridine

through replication fork collapse-induced DNA damage

Yan Yan1, Xiangzi Han1, Yulan Qing2, Allison G. Condie3, Shashank Gorityala4, Shuming

Yang2, Yan Xu2,4, Youwei Zhang1, and Stanton L. Gerson2*

Authors and affiliations:

1 Department of Pharmacology, Case Western Reserve University, Cleveland, OH, United

States

2 Department of Hematology and Oncology, Case Comprehensive Cancer Center, Case

Western Reserve University, Cleveland, OH, United States

3 Division of Radiopharmaceutical Science, Case Center for Imaging Research,

Department of Radiology, Chemistry, and Biomedical Engineering, Case Western Reserve

University, Cleveland, OH, United States

4 Department of Chemistry, Cleveland State University, Cleveland, OH, United States

* Corresponding Author:

Note: This chapter has been adapted from the publication by Yan, Y., et al., Inhibition of uracil DNA glycosylase sensitizes cancer cells to 5-fluorodeoxyuridine through replication fork collapse-induced DNA damage, to the journal Oncotarget.

Page | 38 2.1 Abstract

5-fluorodeoxyuridine (5-FdU, floxuridine) is active against multiple cancers through the inhibition of thymidylate synthase, which consequently introduces uracil and

5-FU incorporation into the genome. Uracil DNA glycosylase (UDG) is one of the main enzymes responsible for the removal of uracil and 5-FU. However, how exactly UDG mediates cellular sensitivity to 5-FdU, and if so whether it is through its ability to remove uracil and 5-FU have not been well characterized. In this study, we report that UDG depletion led to incorporation of uracil and 5-FU in DNA following 5-FdU treatment and significantly enhanced 5-FdU’s cytotoxicity in cancer cell lines. Co-treatment, but not post- treatment with thymidine prevented cell death of UDG depleted cells by 5-FdU, indicating that the enhanced cytotoxicity is due to the retention of uracil and 5-FU in genomic DNA in the absence of UDG. Furthermore, UDG depleted cells were arrested at late G1 and early

S phase by 5-FdU, followed by accumulation of sub-G1 population indicating cell death.

Mechanistically, 5-FdU dramatically reduced DNA replication speed in UDG depleted cells. UDG depletion also greatly enhanced DNA damage as shown by γH2AX foci formation. Notably, the increased γH2AX foci formation was not suppressed by caspase inhibitor treatment, suggesting that DNA damage precedes cell death induced by 5-FdU.

Together, these data provide novel mechanistic insights into the roles of UDG in DNA replication, damage repair, and cell death in response to 5-FdU and suggest that UDG is a target for improving the anticancer effect of this agent.

Page | 39 2.2 Introduction

Fluoropyrimidines including 5-fluorouracil (5-FU) and its deoxyribonucleoside metabolite 5-fluorodeoxyuridine (5-FdU, floxuridine) have been widely used in the treatment of various solid tumors, most notably for colorectal cancer [1, 159, 160]. Both

5-FU and 5-FdU can be converted into two forms of active metabolites in cells that disrupt

DNA metabolism: fluorodeoxyuridine monophosphate (FdUMP) and fluorodeoxyuridine triphosphate (FdUTP) [4, 5]. FdUMP inhibits thymidylate synthase (TS), which consequently causes intracellular nucleotide pool imbalance with decreased dTTP and increased dUTP levels. As a result, cells will incorporate dUTP and FdUTP instead of dTTP into their DNA as the modified bases uracil and 5-FU. In addition, 5-FU can also be converted into ribonucleotide fluorouridine triphosphate (FUTP) which can then be incorporated into RNA [4, 5]. A large body of studies suggests that TS inhibition is the widely accepted mechanism by which fluropyrimidines exert their anticancer effects [1, 4,

5]. Therefore, 5-FU combined with leucovorin, which specifically prolongs the duration of inhibition on TS by FdUMP, is currently considered as the standard systematic chemotherapy for advanced colorectal tumors in the clinic [161-163].

Unlike the metabolism of 5-FU into RNA, 5-FdU is primarily phosphorylated into

FdUMP as a potent TS inhibitor and putatively introduces uracil and 5-FU incorporation into DNA, which therefore mainly disrupts DNA metabolism with little RNA-directed action [4-5, 164]. Additionally, 5-FdU appears to be more cytotoxic than 5-FU in a wide range of cancer cell lines and animal tumor systems [165, 166]. Although the metabolism of 5-FdU into nucleotide and DNA has been described [4, 5], it remains unclear how the

DNA damage and the downstream repair pathways would impact the effectiveness of this

Page | 40 drug. According to in vitro kinetic studies, base excision repair (BER) initiated by uracil

DNA glycosylase (UDG) accounts for the dominant cellular activity that removes uracil and 5-FU from DNA compared with other DNA glycosylases [36]. However, whether

UDG-directed BER is an effector that determines the sensitivity of TS inhibitors remains controversial. Based on studies in the yeast system [167], two models were established to explain the role of UDG in determining the cytotoxicity of TS inhibitors [5, 168]. In the first model, futile cycles of uracil and/or 5-FU incorporation and their removal by UDG lead to DNA fragmentation. One piece of evidence supporting this model showed that

UDG-targeted knockdown increased the resistance to 5-FdU [169]. In the second model, accumulation of uracil and/or 5-FU in, rather than their excision from, DNA contributes to the cytotoxicity. For example, recent studies revealed that loss of UDG enhanced the cytotoxicity of cancer cells to pemetrexed and 5-FdU [46, 170]. On the other hand, several studies demonstrated that overexpression or inhibition of UDG did not affect the sensitivity of TS inhibitors in human, mouse, or chicken DT40 cells [15, 36, 41-45]. In addition, the discrepant findings have also been observed with other DNA glycosylases: SMUG1, TDG and MBD4. Enhanced sensitivity to 5-FU was reported in SMUG1 knockout murine cells due to elevated uracil and 5-FU retention [48], whereas increased resistance to 5-FU and

5-FdU was found in genetically depleted TDG or MBD4 mouse embryonic cells [49, 50].

Since UDG activity is significantly higher in colorectal tumors than in normal tissues [171], the question remains as to the role of UDG in cancer cells in response to fluoropyrimidines. In this study we investigated the impact of UDG on the sensitivity of cancer cells to 5-FdU and explored the underlying molecular mechanisms. We found that depletion of UDG induced significant accumulation of both uracil and 5-FU in genomic

Page | 41 DNA, which indicates a prevailing role of UDG in preventing the persistence of these DNA lesions by 5-FdU treatment. Loss of UDG highly enhanced the cytotoxicity of 5-FdU.

Interestingly, this increased cytotoxicity and retention of uracil and 5-FU could not be reversed by thymidine treatment after 5-FdU exposure, suggesting that the cell killing effect of 5-FdU is a result of uracil and 5-FU incorporation into DNA. UDG depleted cells were arrested at late G1 and early S phase during 5-FdU exposure; accordingly, the DNA replication speed detected by the DNA fiber assay was significantly reduced by loss of

UDG, suggesting replication fork stalling or falling. Consistently, UDG depleted cells displayed sustained DNA damage following 5-FdU treatment. Collectively, these findings suggest that UDG plays an important role in the removal of uracil and 5-FU and therefore determines at least partially the therapeutic outcome of fluoropyrimidines in the clinic.

2.3 Materials and Methods

Cell lines and drugs.

DLD1 colon cancer cells were purchased from American Type Culture Collection, and

HEC1A cells were a gift from Dr. Sanford Markowitz at Case Western Reserve University.

Cells were maintained in growth medium DMEM supplemented with 10% dialyzed fetal bovine serum containing penicillin and streptomycin. Cells were incubated at 37°C in a humidified atmosphere of 5% CO2. Drugs and chemicals used in this study are: 5- fluorodeoxyuridine (Sigma Aldrich), thymidine (Sigma Aldrich), pemetrexed (LC laboratories), temozolomide (Ochem Inc), cisplatin and doxorubicin (kindly provided by

Dr. John Pink at Case Western Reserve University).

Page | 42 Lentiviral shRNA knockdown.

UDG knockdown was performed via shRNA transduction with validated clone from

Sigma-Aldrich. The ID of UDG shRNA clone is NM_003362.2-656s21c1. A non-targeted scramble control shRNA clone (Sigma-Aldrich) was also used. Transfection of shRNA clones was performed according to manufacturer’s specifications from Lipofectamine

2000 (Invitrogen). Lentiviral particles were produced via HEK293 cells, and targeted cells were infected and selected with puromycin. The stable UDG knockdown levels were verified for q-PCR and western blot analysis.

Glycosylase activity assay.

UDG activity was determined by using a green emitting Alexa 532 labeled 40-mer duplex

DNA containing a U:A base pair that was synthesized by IDT with the sequence:

5’-TCCTGGGTGACAAAGCUAAACACTGTCTCCAAAAAAAATT [Alexa]-3’

3’-AGGACCCACTGTTTCGATTTGTGACAGAGGTTTTTTTTAA-5’

For the reaction, 5 pmol (10 μL) diluted DNA aliquots were incubated with either purified enzymes UDG and APE (New England Biolabs) sequentially or 10 µg nuclear extracts isolated from cells at 37°C for 20 minutes. Nuclear extracts were prepared by using the

NucBuster isolation procedure (EMD Bioscience Calbiochem). Reaction products were resolved in the dark by electrophoresis on 20% denaturing polyacrylamide gels (5.3 g urea,

5.0 mL 40% acrylamide, 2.3 mL 5X TBE buffer, 200 μL 10% APS, and 20 μL TEMED).

Gels were visualized by a Typhoon Tri + Variable Mode Imager (Amersham Biosciences).

Page | 43 Apyrimidinic (AP) site detection.

The amount of cellular AP sites was assessed as we previously described by using a NIR cyanine-based AP site probe [172]. Briefly, following 5-FdU exposure, genomic DNA was obtained from phenol-chloroform extraction, dissolved in 1X UDG reaction buffer (20 mM

Tris-HCl, 1 mM EDTA and 1 mM dithiothreitol, pH 8.0), and incubated with either the

UDG enzyme (1 µL, 5 units) or 1 µL UDG storage buffer (10 mM Tris-HCl, 50 mM KCl,

1 mM DTT, 0.1 mM EDTA, 0.1 mg/ml BSA, 50% Glycerol, pH 7.4) as a vehicle control at 37 °C for 1 h. After the reaction, AP site probe with a final concentration of 25 µM was added and incubated at 37 °C for 1 h. Following incubation, extracted DNA was precipitated, and the supernatant was discarded. DNA pellets were resuspended in H2O, and DNA concentrations were measured and adjusted. The fluorescence intensities of each sample were analyzed with 760 nm excitation and emission scan of 790-847 nm.

Quantitative determination of uracil and 5-FU incorporated in cellular DNA by LC-

MS/MS.

Genomic DNA was extracted from cells treated with 5-FdU via phenol-chloroform mixture.

80 µg of DNA sample was dissolved in 1X UDG reaction buffer (20 mM Tris-HCl, 1 mM

EDTA and 1 mM dithiothreitol, pH 8.0) and incubated with UDG enzyme (1 µL, 5 units) for 1 h at 37°C. For LC-MS/MS analysis of DNA-incorporated uracil and 5-FU, 75 µL of the enzyme reaction mixture was obtained, and uracil-1,3-15N2 was used as the internal standard (Sigma-Aldrich). All uracil and 5-FU standards, internal standard, and QC samples were prepared in 1X UDG reaction buffer. The separation of analytes were achieved by a Shimadzu LC-20AD HPLC system with a Shimadzu SIL-20AC autosampler

Page | 44 (Shimadzu) on a Waters Xbridge HILIC pre-column (2.1 x 10 mm, 3.5 µm) and a Xbridge

HILIC column (2.1 x 100 mm, 3.5 µm) (Waters Corporation) using a mobile phase consisting of 87.5% acetonitrile and 12.5% 10 mM ammonium formate at a flow rate of

0.200 mL/min. Quantitation of the analytes was accomplished by a AB Sciex API 3200 triple quadrupole tandem mass spectrometer (AB Sciex), which was operated in the negative multiple-reaction-monitoring (MRM) mode with mass transitions of m/z 110.8 >

42.0 for uracil, m/z 112.9 > 43.0 for uracil-1,3-15N2 and m/z 129.0 > 42.0 for 5-FU. This method has lower limits of quantitation of 2.50 ng/mL and linear calibration ranges up to

500 ng/mL for both uracil and 5-FU with a sample injection volume of 15 µL, as well as a total analysis time of 6 min.

Colony survival assay.

DLD1 (200 cells/well) or HEC1A (300 cells/well) cells were plated in 6-well culture dishes and allowed to adhere for 16 h. Cells were treated with drug for 24 h, gently washed twice with 1X PBS, and incubated with fresh media for at least 10 days to allow individual colonies to form. Colonies were stained with methylene blue, and only colonies containing

≥50 cells were counted. The percentage of survival was determined relative to untreated control averaged over 3 independent experiments.

Cell cycle and bromo-deoxyuridine (BrdU)/PI labeling analysis.

For cell cycle analysis, DLD1 cells were synchronize by serum starvation for 48 h and released in fresh media for 16 h. The cells were then treated with 100 nM 5-FdU for 4, 8,

12, 20, 24, 28, 32, 36, 48, 72, and 96 h. At each time point, cells were harvested and fixed with methanol. Fixed cells were incubated with DNase-free RNaseA (Roche) and stained

Page | 45 with 50 μg/mL PI solution (Sigma-Aldrich). For BrdU/PI labeling analysis, cells were treated with 100 nM 5-FdU for 24 h and pulsed with 10 μM BrdU (BD Biosciences

Pharmingen, BrdU Flow Kit) for 45 minutes before collecting cells. According to manufacturer’s instructions from BD Biosciences Pharmingen, cells were fixed, treated with DNAse for 1 h at 37°C, stained with FITC anti-BrdU for 20 minutes, and incubated with PI staining solution (50 μg/mL PI, 10 mM Tris-HCl pH 7.5, 5 mM MgCl2, 10 µg/mL

DNase-free RNaseA) for 30 minutes at 37°C. For both assays, cells were analyzed on a

BD LSRII instrument.

DNA fiber assays.

DNA fiber analysis was performed as described [173]. Cells treated with 100 nM 5-FdU for 24 h were pulse-labeled with 100 µM chlorodeoxyuridine (CIdU) for 20 minutes, washed with PBS, and 25 µM Iododeoxyuridine (IdU) for 20 minutes. Cells were collected in PBS, and 2.5 µL of cell suspension was dropped on glass slide. 7.5 µL of lysis buffer

(0.5% SDS, 200 mM Tris-HCl pH 7.4, 50 mM EDTA) was dropped on the cell suspension and lysis for 10 minutes. Slides were then tilted at 15° to spread the suspension and placed horizontally to allow air-dry. After drying, slides were fixed in 3:1 methanol:acetic acid for 15 minutes, washed with water, and placed at -20 °C overnight. Slides were then treated with 2.5 M HCl for 1 h, washed with PBS containing 0.1% Tween-20, washed twice with

PBS, blocked in PBS containing 5% BSA and 0.1% Tween 20 for 20 minutes, and rinsed with PBS three times. After washing, 100 µL primary antibodies: mouse anti-BrdU/IdU

(Becton Dickinson, 1:100) and rat anti-BrdU/CIdU (AbD Serotec, 1:400) diluted in PBS containing 5% BSA and 0.1% were added to incubate in a humid chamber for 4-6 h. After incubation, slides were washed with PBS three times, incubated with secondary fluorescent

Page | 46 antibodies: sheep anti-mouse Alexa Fluor 488 (Life technologies) and donkey anti-rat

Alexa Fluor 594 (Life technologies) diluted in PBS containing 5% BSA for 1 h. Slides were washed with PBS three times and mounted with Vectashield mounting medium.

Image acquisition was performed on a Leica laser microscope. DNA fiber length was measured by using ImageJ software (NCI/NIH).

Immunofluorescence staining.

Cells cultured on glass coverslips were treated with 5-FdU in the presence or absence of

10 μM caspase inhibitor Q-VD-OPH (BioVision Inc). Cells were fixed in 3.7% formaldehyde for 10 minutes, blocked with PBS containing 10% FBS and 0.1% Triton X-

100 for 20 minutes, washed with PBS three times, and incubated with primary anti-γH2AX antibody (Millipore, dilution: 1:150) in PBS containing 0.1% Triton X-100 at 4°C overnight. The cells were then washed with PBS three times, incubated with secondary antibodies (Alexa Fluor 594, Life Technologies; dilution: 1:400) in PBS containing 0.1%

Triton X-100 for 1 h, and washed with PBS three times. The slides were mounted with antifade solution with DAPI (Cell Signaling) and visualized on a Leica laser microscope.

Western blots and qPCR.

Western blots were performed as described [174]. Antibodies used were as follows: Anti-

UDG (FL-313) (Santa Cruz Biotechnology), anti-Cleaved PARP (Asp214)(19F4) (Cell

Signaling), and anti-α-Tubulin (Calbiochem). For quantitative RT-PCR, total RNA from cells was extracted using RNeasy Plus Mini Kit (Qiagen), and cDNA synthesis was carried out by using SuperScript III First Strand Kit (Life Technologies). Q-PCR was performed with validated TaqMAN MGB FAMTM dye labeled probes (Applied Biosystems) for

Page | 47 UDG on an ABI 7500 Fast Real-time PCR System (Applied Biosystems). β-Actin was used as an endogenous control, and relative gene expression was calculated as 2−ΔΔCt.

Statistics.

Statistical significance between two treatment groups was determined by unpaired 2-tailed student’s t test. Significance was assigned for a P-value < 0.05. Standard software

GraphPad Prism (San Diego, CA, USA) and Excel 2013 (Microsoft Corp., Redmond, WA) were used for all statistical analysis.

2.4 Results

2.4.1 UDG removes uracil and 5-FU incorporated into DNA following 5-FdU treatment

Studies have demonstrated that the nuclear form of UDG is responsible for the removal of uracil and 5-FU from DNA in vitro in comparison with other glycosylases [36].

To confirm this activity of UDG in vivo, we generated DLD1 colon cancer cells whose expression of UDG was depleted by shRNA (Figure 2-1A, B). We then determined if the enzymatic activity of UDG is reduced in UDG depleted cells by the glycosylase activity assay. In brief, we incubated isolated nuclear extracts with a fluorescently tagged 40-mer

DNA duplex that contains a U:A base pair. If the activity of UDG is intact, the uracil base will be removed, creating an abasic/apyrimidinic (AP) site. AP sites will be subsequently cleaved by the downstream BER protein AP endonuclease (APE) to generate a 23-mer band that can be visualized by gel electrophoresis (Figure 2-1C). As expected, purified

UDG and APE enzymes efficiently removed uracil in the DNA duplex (Figure 2-1D, lane

Page | 48 1-3), serving as a positive control. Nuclear extracts from non-targeted scramble (shSCR)- transfected cells almost completed removed uracil bases (oligo cutting) (Figure 2-1D, lane

4). However, extracts from shUDG-transfected cells exhibited markedly reduced activity of removing uracil (minimal cutting) (Figure 2-1D, lane 5). These results confirm that UDG is the major contributor to the uracil removal from DNA in cells.

To further study the role of UDG in removing genomic uracil and/or 5-FU, we assessed the levels of uracil and 5-FU in cellular DNA after 5-FdU treatment by the AP site detection assay. Since dUTP and 5-FdUTP pools are not elevated in cancer cells cultured with standard serum in response to 5-FdU [36], we used medium containing 10% dialyzed serum in this study. We first extracted DNA from cells treated with 5-FdU, exposed the DNA to exogenous UDG to remove residual uracil and 5-FU bases, and then the newly generated AP sites were detected by a novel near infrared (NIR) cyanine-based probe that we previously synthesized and reported [172]. The results showed that the levels of AP sites in shSCR-transfected cells remained low after 5-FdU treatment even at high concentrations (Figure 2-1E). In contrast, DNA from shUDG-transfected cells displayed a dramatic increase in the levels of detected AP sites in a 5-FdU dose dependent manner

(Figure 2-1E), suggesting accumulation of genomic uracil and 5-FU in UDG depleted cells.

AP sites are the common product of removal of uracil and/or 5-FU from DNA.

Therefore, the AP site detection assay provides an assessment of the combined cellular levels of uracil and 5-FU but cannot distinguish which one is dominant. Since the pathways of uracil and 5-FU incorporation differ (TS inhibition leads to uracil incorporation, whereas phosphorylation of 5-FdU leads to 5-FU incorporation), the individual levels of uracil and

5-FU may determine which pathway predominantly contribute to UDG removable lesions.

Page | 49 To address this issue, we isolated genomic DNA from cells treated with 5-FdU, incubated the DNA with purified UDG, and measured the levels of released uracil and 5-FU by LC-

MS/MS. Very low levels of uracil and 5-FU were detected from shSCR-transfected cells even after treatment with high concentrations of 5-FdU (Figure 2-1F), indicating efficient removal of these bases from DNA by UDG. On the other hand, a significant increase of both uracil and 5-FU was detected from shUDG-transfected cells after 5-FdU treatment

(Figure 2-1F). These data demonstrate that 5-FdU treatment leads to roughly equivalent incorporation of both uracil and 5-FU into DNA, indicating that both lesions can contribute to the genotoxicity. These results further suggest that UDG plays a major role in removing these bases and limiting such toxicity.

2.4.2 Loss of UDG enhances cytotoxicity of 5-FdU in cancer cells

To address the role of UDG in determining the cytotoxicity of 5-FdU, we measured the cell survival of DLD1 colon cancer cells and HEC1A endometrial cancer cells in response to 5-FdU by colony survival assays. The results showed that 5-FdU caused a moderate loss of cell viability in shSCR-transfected cells at high concentrations (Figure 2-

2A, B). Notably, loss of UDG highly sensitized cancer cells to 5-FdU treatment (Figure 2-

2A, B). This sensitization was also observed in UDG depleted DLD1 and HEC1A cancer cells treated with pemetrexed (Figure 2-2C, D), an antifolate that can also block TS and introduce uracil incoporation into DNA. In contrast, UDG depleted DLD1 and HEC1A cells displayed no further sensitivitiy to cisplatin (Figure 2-2E, F), a crosslinking agent, doxorubicin (2-2G, H), a DNA intercalating agent, or temozolomide (Figure 2-2I, J), an alkylating agent, indicating that UDG is not involved in removing crosslinked, intercalated,

Page | 50 or methylated nucleotides from DNA. Collectively, these data demonstrate that loss of

UDG increases the sensivity of cancer cells to agents that induce uracil or 5-FU incorporation into DNA, suggesting that UDG plays an important role in determining the cell killing effect of these drugs.

2.4.3 Thymidine treatment after 5-FdU exposure cannot fully rescue the enhanced cytotoxicity in UDG depleted cells due to the retention of genomic uracil and 5-FU

Thymidine deficiency has been generally considered as the main cytotoxic mechanism of TS inhibitors [1, 4-5]. However, our data suggest that it is the incorporation and the lack of removal of genomic uracil and 5-FU lesions that caused the enhanced cytotoxicity of UDG depleted cells to 5-FdU. If so, the replenishment of thymidine should bypass the thymidine deficiency induced by 5-FdU and also reduce the incorporation of either uracil or 5-FU into DNA, a downstream effect of a shortage of thymidine pool. To test this hypothesis, we first examined the effect of simultaneous treatment of thymidine and 5-FdU (shSCR+Thy, shUDG+Thy), which was intended to completely block the thymidineless effect from the beginning. Under these conditions, there was almost no killing in either shSCR-transfected or shUDG-transfected cells (Figure 2-3A). However, when thymidine was replenished 24 h after 5-FdU treatment (shSCR+Thy (24h post), shUDG+Thy (24h post)), it barely inhibited cell death of UDG depleted cells caused by 5-

FdU (Figure 2-3B), indicating that the enhanced killing effect by UDG depletion is due to the incorporation of uracil and 5-FU into DNA instead of the lack of thymidine. To further prove that uracil and 5-FU lesions are indeed retained in DLD1 UDG depleted cells even during recovery in the presence of thymidine, we performed the AP site detection assay in

Page | 51 cells treated with thymidine after 24 h of 5-FdU exposure. The results showed that UDG depleted cells accumulated about three times higher the level of uracil and 5-FU in DNA than shSCR-transfected cells following 24 h of 5-FdU treatment (Figure 2-3C). After 24 h of 5-FdU exposure, cells were washed and placed in drug-free medium supplemented with thymidine. Notably, we observed that the uracil and 5-FU levels in UDG depleted cells remained persistent during 6, 12, and 24 h of thymidine recovery (Figure 2-3C).

Furthermore, the retention of uracil and 5-FU during thymidine recovery following 5-FdU treatment was also detected in HEC1A UDG depleted cells (Appendix 1). Taken together, these data suggest that the enhanced cytotoxicity in UDG depleted cells is attributed to the retention of uracil and 5-FU in DNA.

2.4.4 UDG depletion leads to cell cycle arrest at late G1 and early S phase by 5-FdU

Studies have shown that TS inhibition leads to S phase arrest by blocking DNA replication as a result of dTTP deficiency [175-177]. To elucidate the molecular mechanisms by which UDG regulates cellular sensitivity to 5-FdU, we monitored cell cycle progression by propidium iodide (PI) staining. DLD1 cells were synchronized at G0/G1 phase through serum starvation, resumed growth by placing in medium containing 10% dialyzed FBS for 16 h which did not result in progression through cell cycle, and then exposed to 5-FdU for an additional 0 to 96 h. In the absence of 5-FdU, both shSCR- transfected and shUDG-transfected cells progressed similarly through S and G2/M phases by 8 and 12 h, respectively (Figure 2-4A, B), indicating that UDG depletion did not affect normal cell cycle progression. As expected, 5-FdU slowed the progression of shSCR- transfected cells through S phase by 36 h, and cells entered the next cell cycle by 48 h with

Page | 52 a relatively small portion of cells at sub-G1 phase (Figure 2-4A, B). However, 5-FdU treatment triggered a strong cell cycle arrest of UDG depleted cells at late G1 and early S phase which lasted for 48 h and later displayed a chaotic cell cycle distribution pattern at

72 h and 96 h with substantially increased sub-G1 population (Figure 2-4A, B).

To confirm the cell cycle arrest results, we monitored the S phase population of unsynchronized cells by BrdU and PI co-staining in DLD1 cancer cells. Consistently, we observed S phase arrest especially at middle and late S phase in shSCR-transfected cells as a result of TS inhibition after 24 h of 5-FdU exposure (Figure 2-4C, D). In contrast, DLD1 shUDG-transfected cells were arrested at late G1 and early S phase following 24 h of 5-

FdU exposure (Figure 2-4C, E). In addition, the G1/S phase arrest was also confirmed in

HEC1A UDG depleted cells (Appendix 2). Together, these findings implicate that loss of

UDG affects cell cycle progression at early S phase in response to continuous 5-FdU exposure, likely due to the accumulation of uracil and 5-FU in genomic DNA that blocks

DNA replication.

2.4.5 Loss of UDG inhibits DNA replication progression in response to 5-FdU treatment

To directly investigate the mechanism by which 5-FdU arrests UDG depleted cells at G1/S phase, we monitored replication fork progression by DNA fiber analysis [173].

Following 24 h 5-FdU treatment, DLD1 cells were sequentially pulsed with halogenated nucleotides chlorodeoxyuridine (CldU) and iododeoxyuridine (IdU) for 20 minutes (Figure

2-5A). DNA fibers stained with both CldU (red, not shown) and IdU (green) were included

Page | 53 in the following analysis. To assess the impact on DNA replication progression, we measured the track length of IdU as it represents the ongoing replication fork. In the absence of 5-FdU, the mean fiber length for both shSCR- and shUDG-transfected cells was around 7.5 μm (Figure 2-5B). Following 24 h 5-FdU exposure, the mean fiber length of nascent DNA strands reduced by 23% to 5.7 μm in shSCR-transfected cells, consistent with the temporal S phase arrest results (Figure 2-4). Strikingly, UDG depleted cells displayed significantly shorter fiber track with the mean value at 2.8 μm, representing a

63% reduction (Figure 2-5B), consistent with the prolonged G1/S arrest. These results illustrate that loss of UDG inhibits DNA replication in response to 5-FdU by severely reducing the elongation of nascent DNA strands.

2.4.6 DNA damage persists in UDG depleted cells and is not due to apoptosis by 5-

FdU treatment

The dramatic increase in sub-G1 population in UDG depleted cells by 5-FdU indicates that these cells are undergoing apoptotic cell death. However, what caused the cell death remains unclear. Prolonged replication fork stalling due to dNTP imbalance can lead to fork collapse and the generation of DNA double strand breaks (DSBs) [178, 179], a highly mutagenic and toxic form of DNA damage. To understand if UDG depleted cells accumulate DNA damage by 5-FdU treatment, we performed immunostaining to assess the generation of DSBs using specific antibodies to detect foci formation of the phosphorylated histone variant H2AX (γH2AX), a marker of DSBs (Figure 2-6A). In DLD1 shSCR- transfected cells, 5-FdU caused the maximal increase in the level of DSBs and the percentage of cells with over 10 foci by 12 h of treatment, which then gradually declined

Page | 54 despite the presence of 5-FdU (Figure 2-6B-D), indicating cells expressing UDG are able to repair DNA damage even in the presence of 5-FdU. On the other hand, both the foci number and the percentage of cells with over 10 foci remained persistent during 5-FdU exposure in DLD1 UDG depleted cells (Figure 2-6B-D), suggesting sustained DNA damage in the absence of UDG. Consistently, in HEC1A shSCR-transfected cells, the maximal level of DSBs and the percentage of cells with over 10 foci were detected at 48 h of 5-FdU treatment, which then reduced at 72 h and 96 h of treatment (Appendix 3A-D).

However, in HEC1A UDG depleted cells, the foci number and the precentage of cells with over 10 foci remained high during 5-FdU exposure (Appendix 3A-D).

Caspase activation during apoptosis also leads to DNA fragmentation and damage

[180, 181]. Therefore, to prove that the formation of DNA damage is the cause, but not the consequence of cell death induced by 5-FdU, we monitored γH2AX foci in both DLD1 and

HEC1A cells in the presence or absence of a broad-spectrum caspase inhibitor Q-VD-OPh

[182-184]. If DNA damage were the consequence of caspase activation, then we would expect that the caspase inhibitor should abolish γH2AX foci formation. However, we observed that the number of γH2AX foci and the percentage of γH2AX positive cells were essentially the same between Q-VD-OPh treated and non-treated shSCR-transfected or shUDG-transfected cells (Figure 2-6B-D, and Appendix 3B-D). These data strongly suggest that the increased DNA damage induced by 5-FdU is not the result of caspase activation. To prove that the caspase inhibitor indeed blocked the apoptotic signaling, we examined the expression of cleaved PARP, a marker of apoptosis, in parallel samples. We found that cleaved PARP by 5-FdU treatment was almost completely blocked by the Q-

VD-OPh treatment in both shSCR-transfected and shUDG-transfected cells (Figure 2-6E,

Page | 55 and Appendix 3E). In addition, the appearance of cleaved PARP in DLD1 and HEC1A cells was not evident until after 24 h and 96h of 5-FdU treatment, respectively (Figure 2-

6E, and Appendix 3E), whereas DSBs formation was readily detected at 12 h and 48 h of treatment (Figure 2-6B-D, and Appendix 3B-D). Collectively, these results demonstrate that the formation of DSBs precedes the apoptosis signaling caused by 5-FdU in UDG depleted cells, suggesting that DNA damage is the cause of cell death.

2.5 Discussion

5-FdU metabolite blocks TS, causing nucleotide pool imbalance, which favors uracil and 5-FU incorporation into DNA. Previously, studies have demonstrated elevated genomic uracil levels in UDG deficient non-cancer cells [42, 44], and increased levels of genomic 5-FU in UDG depleted cancer cell following 5-FdU exposure [15]. To the best of our knowledge, the present study is the first to illustrate the individual levels of incorporated uracil and 5-FU simultaneously in response to 5-FdU. Using both the AP site detection assay and mass spectrometry analysis, we found similar levels of uracil and 5-

FU incorporated into cellular DNA following 5-FdU exposure in UDG depleted cancer cells. Collectively, these studies indicate that the absence of uracil and 5-FU in UDG competent cells reflects a predominant role of UDG in preventing abnormal base accumulation in genomic DNA.

Previous studies reported inconsistent roles of UDG in determining the sensitivity of TS inhibitors and, of note, most of these studies were conducted in non-cancer cells [36,

41-45]. In human colon tumors, the UDG activity has been reported to be significantly higher than in normal bowel tissues [171], suggesting UDG as a potential biomarker to

Page | 56 predict 5-FdU resistance in colon cancer and also a potential target for inhibition. The study of Huehl et al. (2016) firstly showed that loss of UDG sensitized cancer cells to 5-FdU [46], which is consistent with our findings. Here we further demonstrate that this sensitization can only be rescued when thymidine was added simultaneously, but not after 5-FdU treatment, indicating that the cytotoxicity was mainly caused by the accumulation of uracil and 5-FU bases in DNA in the absence of UDG. We previously observed sensitization of

UDG depleted cancer cells to another TS inhibitor, pemetrexed, suggesting that the genomic uracil incorporation alone is toxic to cells [170]. Although Huehl et al. (2016) suggested that incorporation of 5-FU into DNA played a more important role than uracil in contributing to 5-FdU-induced cell death [46], they did not measure incorporated uracil levels and therefore may have underestimated the contribution of uracil incorporation to 5-

FdU’s cytotoxicity and DNA replication fork disruption. Nevertheless, these findings together demonstrate that loss of UDG in cancer cells enhances the killing effect of 5-FdU, a TS inhibitor through the incorporation of the abnormal bases uracil and 5-FU into DNA.

In addition to UDG that acts on uracil and 5-FU in DNA, TDG was also reported to preferentially excise uracil and 5-FU that is mispaired with guanine [185-187]. However, loss of TDG that confers 5-FU resistance has been observed in MEF cells in a manner different from that of UDG in our study [49]. The excision of uracil and 5-FU from DNA by TDG was thought to precipitate the cytotoxicity of 5-FU due to the slow dissociation of

TDG from AP sites, therefore blocking the downstream repair pathway [36, 49].

Importantly, it has been reported previously that TDG is absent from S phase cells, while

UDG expression, on the contrary, is highly induced in S phase [188, 189]. Our findings revealed that UDG depletion leads to accumulation of DNA lesions including both uracil

Page | 57 and 5-FU incorporation during S phase in response to 5-FdU treatment. Once the cells exit

S phase where TDG is expressed, these lesions will be recognized by TDG that slows down the repair process and contributes to additional cytotoxicity. Under these assumptions,

TDG would synergize with the inhibition of UDG following 5-FdU exposure.

Knowing that the presence of UDG significantly compromised the cytotoxic effect of 5-FdU by limiting the existence of DNA lesions of uracil and 5-FU, we sought to understand how these lesions led to cytotoxicity in cancer cells. We found that genomic uracil and 5-FU incorporation, which is a downstream effect of TS inhibition by 5-FdU, induced cell cycle arrest at late G1 and early S phase, indicating replication fork stalling at early phases of DNA synthesis. This severely stalled or collapsed DNA replication was confirmed and quantified via the DNA fiber analysis. We previously reported that uracilated DNA induced by pemetrexed treatment arrested UDG depleted DLD1 cells at S phase, which is in agreement with the current findings [170]. However, 5-FdU induced more profound DNA replication arrest than pemetrexed after UDG depletion. We propose two possibilities to explain these differences. First, low doses of pemetrexed were used in the previous study, which likely led to less production of dUTP than 5-FdU at the doses used herein. Second, while pemetrexed primarily induces dUTP production, 5-FdU leads to the generation of both dUTP and 5-FdUTP.

Activation of homologous recombination-induced DNA damage repair in response to TS inhibition promotes cell survival [122, 190], which could explain the disappearance of DNA damage of shSCR-transfected cells even in the presence of 5-FdU. However, loss of UDG induces accumulation of significant amounts of abnormal bases in genomic DNA followed by DNA replication arrest, which consequently leads to the generation of

Page | 58 numerous DSBs that are likely beyond the cell’s repair capability. As a result, UDG depleted cells displayed continuous DNA damage. Although numerous studies have indicated activation of apoptosis following exposure to TS inhibitors [191-193], our results suggested that it is DNA damage that induces cell death, but not the other way around. The time course studies of γH2AX foci formation in the presence or absence of caspase inhibitor confirmed this idea, in which DNA damage precedes the activation of cell death signaling pathway in cells depleted of UDG. These results strongly support the idea that loss of UDG significantly enhances the cell killing effect of 5-FdU through the generation of excessive DNA damage.

While uracil and/or 5-FU incorporation into DNA has been recognized for decades, this is one of the few studies to define its mechanism of toxicity in the absence of removal by UDG. Further, from a clinical point of view, our studies clarify the utility of targeting

UDG to improve the anti-cancer efficacy of commonly used chemotherapeutic agents.

Page | 59 Figure 2-1. UDG depletion causes incorporation of uracil and 5-FU into genomic DNA by 5-FdU.

Lentiviral non-targeted scramble control shRNA (shSCR) or UDG-directed shRNA

(shUDG) were transfected into DLD1 colon cancer cells, and stable cell lines were established. (A) UDG mRNA and (B) protein expression levels were determined by qPCR and western blot, respectively. The shRNA that we used targets both mitochondrial and nuclear UDG, which are collectively termed UDG in this study. (C) Schematic diagram of glycosylase activity assay by using 3’-Alexa tagged 40-mer DNA duplex with a uracil incorporation paired with adenine. (D) 10 µg nuclear extracts from DLD1 shSCR or shUDG cells were incubated with 3’-Alexa labeled oligonucleotide containing U:A base pair for 20 minutes at 37°C. Reactions with purified enzymes were used as controls.

Cellular UDG activity was visualized by denaturing gel electrophoresis to separate intact

40-mer from 23-mer. (E) DLD1 shSCR and shUDG cells were treated with 0, 50, 100, and

200 nM 5-FdU for 48 h. Genomic DNA was extracted and treated in vitro with purified

UDG (+ UDG) or vehicle control (- UDG). AP sites detection was performed by incubation of DNA with a cyanine-based AP site probe. Data represent mean and SD of relative fluorescence intensity normalized to 5-FdU untreated shSCR -UDG sample from three independent experiments. (*, P < 0.05) (F) DLD1 shSCR and shUDG cells were untreated

(Unt) or treated with 5-FdU 100 and 200 nM for 48 h. Genomic DNA was extracted and incubated in vitro with purified UDG enzyme. Uracil and 5-FU were quantified by LC-

MS/MS as described in the Materials and Methods. Data represent mean and SD from three independent experiments. (*, P < 0.05)

Page | 60 Figure 2-1. UDG depletion causes incorporation of uracil and 5-FU into genomic DNA by 5-FdU.

Page | 61 Figure 2-2. UDG depletion enhances 5-FdU sensitivity in cancer cells.

Colony survival assays in (A) DLD1 and (B) HEC1A shSCR and shUDG cancer cells treated with increasing doses of 5-FdU, and cell survival was measured as described in

Materials and Methods. UDG expression level in HEC1A cells was determined by western blot (inset). Colony survival assays in (C) DLD1 and (D) HEC1A shSCR and shUDG cells treated with increasing doses of pemetrexed. Colony survival assays in DLD1 and HEC1A shSCR and shUDG cells treated with increasing doses of (E, F) cisplatin, (G, H) doxorubicin, or (I, J) temozolomide (TMZ). Viable colonies (>50 cells) stained with methylene blue after 10 d of culture were counted. All survival data represent mean and

SEM from at least 3 independent experiments. (*, P < 0.05)

Page | 62 Figure 2-2. UDG depletion enhances 5-FdU sensitivity in cancer cells.

Page | 63 Figure 2.3. Thymidine treatment after 5-FdU exposure cannot fully rescue increased cytotoxicity in UDG depleted cells.

(A) Colony survival assay in DLD1 shSCR and shUDG cells treated with 0 to 200 nM 5-

FdU alone, or supplemented with 20 µM thymidine simultaneously during 5-FdU treatment

(+Thy). (B) Colony survival assay in DLD1 shSCR and shUDG cells treated with 0 to 200 nM 5-FdU alone, or supplemented with 20 µM thymidine 24 h after 5-FdU treatment

((+Thy (24h post)). Data represent mean and error from at least 3 independent experiments.

(*, P < 0.05) (C) DLD1 shSCR and shUDG cells were treated with 100 nM 5-FdU for 24 h, then washed twice with PBS, and incubated in drug-free media supplemented with 20

µM thymidine (Thy) for 6, 12, or 24 h. Genomic DNA was extracted and treated in vitro with purified UDG. AP sites detection was performed by incubation of DNA with a cyanine-based AP site probe. Data represent mean and SD of relative fluorescence intensity normalized to the shSCR DNA without 5-FdU treatment from three independent experiments.

Page | 64 Figure 2.3. Thymidine treatment after 5-FdU exposure cannot fully rescue increased cytotoxicity in UDG depleted cells.

Page | 65 Figure 2-4. Loss of UDG induces cell cycle arrest at late G1 and early S phase by 5-

FdU exposure.

(A) DLD1 shSCR and shUDG cells were synchronized at G0/G1 phase by serum starvation for two days indicated as Starve. Cell cycle and growth were resumed by releasing cells into medium containing 10% dialyzed FBS for 16 h. Cells were then exposed to 100 nM

5-FdU for indicated times (0-96 h). Cell cycle of untreated and treated cells was analyzed by PI mediated flow cytometry. (B) Quantification of each phases of the cell cycle for shSCR and shUDG cells from A. (C) Unsynchronized DLD1 shSCR and shUDG cells were untreated (Unt) or treated with 100 nM 5-FdU for 24 h and pulsed with BrdU for 45 minutes. Cells were collected, fixed and stained with anti-BrdU antibody and PI dye. Cell cycle profiles were analyzed by flow cytometry. eS = early S-phase; mS = mid-S-phase; lS

= late S/G2-phase. Quantification of each phases of the cell cycle for DLD1 (D) shSCR and (E) shUDG cells from C. Data for a representative experiment that has been performed three times is shown.

Page | 66 Figure 2-4. Loss of UDG induces cell cycle arrest at late G1 and early S phase by 5-

FdU exposure.

Page | 67 Figure 2-5. UDG depletion inhibits replication fork progression following 5-FdU treatment.

(A) DLD1 shSCR and shUDG cells were untreated (Unt) or treated with 100 nM 5-FdU for 24 h, washed, pulsed with CIdU and IdU sequentially for 20 minutes. Cells were lysed and DNA fragments were spread on the slide. The fixed samples were stained with anti-

CIdU and anti-IdU antibodies. DNA fibers were visualized on fluorescence microscope

(100X oil lens). (Scale bar: 5 μm) (B) Quantification of the DNA fiber length. The statistical analysis of DNA fiber length across the populations analyzed (n > 200 fibers per population) is shown as a scatter plot with medians and the interquartile ranges. To monitor the replication progression speed, we only counted the IdU track as it represents ongoing replication length.

Page | 68 Figure 2-5. UDG depletion inhibits replication fork progression following 5-FdU treatment.

Page | 69 Figure 2-6. DNA damage accumulates in UDG depleted cells in a caspase independent manner.

(A) Schematic diagram of the treatment of DLD1 cells with 5-FdU in the presence or absence (+/-) of 10 µM caspase inhibitor Q-VD-OPh at indicated time points. (B) DLD1 shSCR and shUDG cells were treated with 50 nM 5-FdU for 12, 24, and 48 h with (+) and without (-) 10 µM Q-VD-OPh. Cells were fixed and stained with anti-γH2AX antibodies.

γH2AX foci was visualized on a fluorescence microscope. (C) Quantification of the number of γH2AX foci per cell for 0, 12, 24, and 48 h of 5-FdU treatment in the presence

(+) or absence (-) of Q-VD-OPh. The statistical analysis of γH2AX foci per cell across the populations analyzed (n > 100 cells per population) is shown as a scatter plot with medians and the interquartile ranges. (D) Quantification of the percentage of cells with >10 γH2AX foci per cell for 0, 12, 24, and 48 h of 5-FdU treatment. Statistical analysis was performed as in C. (E) In parallel samples from B, the expression level of cleaved PARP was analyzed for cells untreated (Unt) or treated with 50 nM 5-FdU for 12, 24, and 48 h in the presence

(+) or absence (-) of 10 µM caspase inhibitor Q-VD-OPh.

Page | 70 Figure 2-6. DNA damage accumulates in UDG depleted cells in a caspase independent manner.

Page | 71 CHAPTER 3.

Knockdown of uracil DNA glycosylase selectively re-sensitizes p53 mutant and deficient human cancer cells to 5-fluorodeoxyuridine

Yan Yan1, Yulan Qing2, John Pink2, and Stanton L. Gerson2*

Authors and affiliations:

1 Department of Pharmacology, Case Western Reserve University, Cleveland, OH, USA

2 Case Comprehensive Cancer Center, Seidman Cancer Center, University Hospitals

Cleveland Medical Center, and Case Western Reserve University, Cleveland, OH, USA

* Corresponding author

Note: An adaptation of this chapter has been submitted to the journal Cancer Research for peer review.

Page | 72 3.1 Abstract

Thymidylate synthase inhibitors including fluoropyrimidines (5-fluorouracil (5-

FU), and floxuridine (5-FdU)) and antifolates (pemetrexed) are widely used in the treatment of solid tumors. Previously we reported that shRNA knockdown (KD) of uracil

DNA glycosylase (UDG) highly sensitized cancer cells to 5-FdU or pemetrexed. However, some studies have not confirmed sensitization of TS inhibitors via UDG depletion. To further interrogate this discrepancy with regard to p53 status, we investigated whether the cytotoxicity of TS inhibitors following UDG depletion is reliant upon p53 activity. In a panel of cancer cells with differing p53 status, we show that cancer cells with p53 mutations or deficiencies are highly resistant to 5-FdU compared to cells with wild-type p53.

Depletion of UDG by shRNA re-sensitizes p53 mutant or deficient cancer cells to 5-FdU, whereas p53 wild-type cells are not affected under similar conditions. Utilizing paired

HCT116 p53 wild-type (WT) and p53 knockout (KO) cell lines, we report that loss of p53 improves cell survival after 5-FdU, and UDG depletion only significantly sensitizes p53

KO cells. This sensitization can also be recapitulated by UDG depletion in cells with p53

KD by shRNAs. Additionally re-sensitization is also detected with TS inhibitor pemetrexed in p53 KO cells, but not with 5-FU, most likely due to a lesser degree of DNA incorporation.

Importantly, in p53 WT cells, the apoptosis pathway induced by 5-FdU is activated at a similar level regardless of whether UDG is present or not. However, in p53 KO cells, apoptosis is compromised in UDG expressing cells, but dramatically elevated in UDG depleted cells. Collectively, these results provide evidence that UDG removal of uracil or

5-FU lesions in DNA protects cells from cytotoxicity of TS inhibitors and suggest that loss

Page | 73 of UDG catalyzes significant cell death signals only in cancer cells mutant or deficient in p53.

3.2 Introduction

Thymidylate synthase (TS) is a key enzyme that catalyzes the only means for de novo synthesis of deoxythymidine monophosphate (dTMP) [194]. TS utilizes 5,10- methylenetetrahydrofolate (5,10-CH2THF) as the methyl-group donor and catalyzes the reductive methylation of deoxyuridine monophosphate (dUMP) to dTMP [194]. dTMP is subsequently phosphorylated to deoxythymidine triphosphate (dTTP), a critical precursor for DNA replication and repair. As TS contains binding sites for the substrate nucleotide

(dUMP) and the cofactor folate (5,10-CH2THF), two structurally different classes of inhibitors, nucleotide or folate analogues block the activity of TS [2]. The class of fluoropyrimidines including 5-fluorouracil (5-FU) and floxuridine (5-FdU) target the nucleotide binding site, whereas the antifolates such as pemetrexed target the folate binding site of TS.

Fluoropyrimidines are widely used in the treatment of various types of malignancies for their broad antitumor activity [1, 5]. Once taken into cells, fluoropyrimidines can be metabolized into fluorodeoxyuridine monophosphate (FdUMP) and fluorodeoxyuridine triphosphate (FdUTP) [1, 2]. The metabolite FdUMP inhibits TS by forming a stable ternary complex with TS and CH2THF [6-8], which ultimately leads to the depletion of dTTP and accumulation of deoxyuridine triphosphate (dUTP). The resulting imbalance of deoxynucleotide pools favors the utilization of dUTP and FdUTP during DNA replication and results in the accumulation of both uracil and 5-FU in DNA

Page | 74 [1, 4, 5]. Multi-targeted antifolates such as pemetrexed have been approved as components of first-line therapy in combination with cisplatin for the treatment of advanced non-small cell lung cancer [18]. Pemetrexed inhibits several folate-dependent enzymes, however TS is its predominant target [21-24]. Administration of pemetrexed leads to a global reduction in nucleotide synthesis as well as accumulation of dUTP [195]. As a result, dUTP is used in DNA synthesis in place of dTTP, generating uracil mis-incorporation into DNA [196].

Mis-incorporated uracil and 5-FU are both primarily recognized and repaired by the uracil DNA glycosylase (UDG) initiated base excision repair (BER) pathway [36].

Although incorporation of uracil and 5-FU into DNA is well documented as a consequence of exposure to TS inhibitors [196], the impact of the downstream repair pathway directed by UDG on cell survival is not consistent. It has been hypothesized that thymine-less futile cycles of uracil mis-incorporation, excision by UDG, and further dUTP re-insertion result in DNA strand breaks and cell death [40]. If thymine-less cell death was dependent on

UDG mediated removal of uracil and 5-FU, one would expect a correlation between the cytotoxicity of TS inhibitors and UDG expression. However, the majority of studies reported that neither overexpression, nor inhibition of UDG affected the sensitivity to TS inhibitors in human, mouse, or chicken DT40 cells [15，36, 41-45]. In contrast, recently both our and the Karnitz group observed that loss of UDG highly potentiated the cytotoxicity of 5-FdU in several cancer cell lines, indicating that uracil and 5-FU incorporation played a key role in cell killing [46, 47].

As the mediators of cell killing due to persistent uracil and 5-FU lesions in DNA are not clear, we assessed the likely pathways and noted that one of the major differences in these disparate findings is that cancer cells bearing p53 mutations were used in our and

Page | 75 Karnitz’s experimental system [46, 47], whereas non-transformed or p53 WT cancer cells were used in the majorities of others [41-44]. Mutation of TP53 is the most frequently observed gene alteration in cancers [140]. Mutations in p53 have been shown to influence cellular response to chemotherapeutic agents such as cisplatin, etoposide, and 5-FU [144,

197]. Notably, substantial evidence reveals that loss of p53, or p53 mutations, are linked to resistance to 5-FU due to inability to activate apoptosis pathway. For example, a study using isogenic cell systems demonstrated that deletion of p53 from a p53 WT colon cancer cell line (HCT116) rendered cells remarkably resistant to apoptosis induced by 5-FU [79].

In addition, 5-FU resistance was also described in a variety of p53 mutated cancer cells, including colon, bladder, pancreatic, and gastric cancer [80-83]. However, few studies have reported on the link of p53 status with the response to other TS inhibitors such as 5-FdU.

Given the divergent cell models with different p53 status used in our and other studies, the following questions remain unanswered: 1) does loss or mutation of p53 render cells resistant to 5-FdU, and 2) is the potentiated cytotoxicity of 5-FdU after UDG depletion reliant upon p53 status? To gain insight into these questions, we tested the impact of UDG depletion on 5-FdU cytotoxicity in a number of cancer cell lines with differing p53 status.

We found that, in general, loss or mutation of p53 remarkably reduced the sensitivity to 5-

FdU, and depletion of UDG selectively re-sensitized p53 deficient or mutated cancer cells to 5-FdU. In order to understand the underlying mechanism contributes to the distinct response after UDG depletion, we utilized paired HCT116 cell lines with, or without, deletion of the TP53 gene, and observed that loss of UDG selectively re-sensitized HCT116 cells with p53 deletion. This re-sensitization was also observed with pemetrexed, but to a lesser extent with 5-FU, which mainly causes damage in RNA [15-17, 57]. In the presence

Page | 76 of wild type p53, 5-FdU treatment induced activation of the apoptosis pathway in both

UDG competent, or UDG depleted cells at comparable levels. However, in the absence of wild type p53, apoptosis activation was compromised in UDG expressing cells and dramatically elevated in UDG depleted cells. Collectively, these findings suggest that loss, or mutation, of p53 is associated with 5-FdU resistance, and UDG depletion can significantly restore sensitivity, indicating that UDG may serve as a therapeutic target to improve the clinical effectiveness of 5-FdU.

3.3 Materials and methods

Cell lines and drugs.

HCT116 p53 KO cells were a gift from Dr. Guangbin Luo (Department of Genetics, Case

Western Reserve University, Cleveland, OH). The other cancer cells were purchased from

American Type Culture Collection. Details of the cell lines used in this study are listed in

Table 3-1. All cells were maintained in DMEM supplemented with 10% dialyzed fetal bovine serum, 2mM L-glutamine, 100 U/mL penicillin and 100 μg/mL streptomycin. Cells were incubated at 37°C in a humidified atmosphere of 95% air and 5% CO2. 5-FdU and 5-

FU were purchased from Sigma-Aldrich, dissolved respectively in Milli-Q water and

DMSO, and stored as a 10 mM stock at -80°C. Pemetrexed was purchased from LC laboratories, and prepared fresh for each experiment by dissolving in Milli-Q water.

Lentiviral shRNA knockdown.

Page | 77 p53 or UDG knockdown was achieved via shRNA transduction. Lentiviral vectors LV-

THM-shp53 (which also expresses a GFP reporter) or LV-Bleo-shp53 to perform p53 KD in WT HCT116 cells were obtained from Dr. Mark Jackson’s laboratory Case Western

Reserve University, Cleveland, OH [198]. Lentiviral vector targeting GFP (sh-GFP) was used as control. UDG shRNA vector (NM_003362.2-656s21c1) was purchased from

Sigma, and a scramble targeting shRNA vector (Sigma) was used as paired control. The lentiviral production and infection were performed as previously described [47]. Cells stably transfected with LV-THM-p53 vector were isolated by cell sorting on the basis of their GFP expression. Cells stably transfected with LV-Bleo-p53 vector were assessed with zeocin (Sigma). Selection of positive UDG KD cells was assessed with puromycin (Sigma).

Clonogenic survival assay.

As described previously [47], cancer cells (200-300 cells/well) were seeded in 6-well culture dishes and allowed to adhere overnight. For 5-FdU, cells were treated for 24 h, then gently washed with PBS, and incubated with fresh media for at least 10 days to allow individual colonies to form. For 5-FU or pemetrexed, cells were treated continuously for at least 10 days to form colonies. After 10-18 days, the plates were stained with methylene blue. Colonies containing ≥50 cells were counted. The percentage of survival was determined relative to untreated control averaged over 3 independent experiments.

Western blots and qPCR.

Page | 78 Western blots were performed as previously described [174]. Twenty microgram of protein was loaded on SDS-polyacrylamide gel. The following antibodies were used to detect proteins on the membrane: α-Tubulin (Calbiochem); GAPDH (Santa Cruz Biotechnology);

UDG (FL-313) (Santa Cruz Biotechnology); cleaved PARP (Asp214)(19F4) (Cell

Signaling); cleaved caspase 3 (Cell Signaling); p53 (FL-393) (Santa Cruz Biotechnology); and p21 (Santa Cruz Biotechnoogy). For quantitative RT-PCR, total RNA from cells was extracted by using RNeasy Plus Mini Kit (Qiagen). cDNA synthesis was performed by using SuperScript III First Strand Kit (Life Technologies). Q-PCR was achieved with validated TaqMAN MGB FAMTM dye labeled probes (Applied Biosystems) for nuclear

UDG on an ABI 7500 Fast Real-time PCR System (Applied Biosystems). β-Actin was used as an endogenous control, and relative gene expression was calculated as 2−ΔΔCt.

Flow cytometric assay of apoptosis.

Cells were seeded in 6-well tissue culture plates (1.5X105 cells/well) and allowed to attach overnight. Cells were then treated with 25 nM 5-FdU for 24 h, washed twice with PBS, replenished with drug-free medium at 48, 72, and 96 h. After recovery, the cells floating in the medium were collected. The adherent cells were trypsinized, pelleted, washed in ice- cold PBS, and resuspended in 1X Binding Buffer according to the manufacturer’s instructions (FITC Annexin V Apoptosis Detection Kit, BD Pharmingen). Cells were then stained with FITC Annexin V and PI for 15 minutes at room temperature in the dark.

Annexin V-FITC detects translocation of phosphatidylinositol from the inner to the out cell membrane during early apoptosis, and PI can enter the cells in late apoptosis or necrosis.

Untreated cells were used as control for the double staining. The cells were analyzed

Page | 79 immediately after staining using a Attune NXT instrument and FlowJo software. For each measurement, at least 20,000 cells were counted.

Statistical analysis.

Statistical significance between two treatment groups was analyzed using unpaired 2-tailed student’s t-test. Significance was assigned for a P-value < 0.05. Standard software

GraphPad Prism (San Diego, CA, USA) and Excel 2013 (Microsoft Corp., Redmond, WA) were used for all statistical analysis.

3.4 Results

3.4.1 p53 mutation or deficiency affords 5-FdU resistance among different types of cancer cells

Given that p53 mutations or deficiencies are frequently observed in cancers, and studies have demonstrated that mutations of p53 reduce 5-FU cytotoxicity [79-83]. To understand whether these mutations also alter the response to 5-FdU, a panel of human cancer cell lines from colon, lung, ovarian, pancreas, skin, and endometrium with intrinsically differing p53 status were utilized in this study. The p53 status of each cell line is listed in Table 3-1. To determine p53 protein functionality in p53 WT and p53 mutant

(Mut) or deficient cancer cell lines, we assessed p53 levels and expression of p21, a widely accepted initiator of p53 activated signaling [199], 24 hours after administration of 8Gy gamma-irradiation. All the p53 WT cancer cell lines used in this work induced p21 expression after irradiation, indicating functional p53 in these cell lines (Appendix 4). In

Page | 80 order to establish the relationship between p53 status and 5-FdU sensitivity, we evaluated the cytotoxicity of 5-FdU in these cell lines. As shown in Figure 3-1A, the cell lines tested displayed a spectrum of 5-FdU sensitivities with IC50 values ranging from 1.32 ± 0.33 to

269.55 ± 0.73 nM for A2780 and H1299 lines, respectively. Importantly, we observed that, in general, cell lines with p53 mutation or deficiency (Figure 3-1A, solid lines) were significantly more resistant to 5-FdU than p53 WT cells (Figure 3-1A, dashed lines), with the exception of A375 which has wild-type p53 but a IC50 of 110.81 ± 1.80 nM. In addition, except for A375, the IC50 values for the p53 WT cancer lines clustered together at a lower dose range, whereas p53 mutant or deficient lines clustered at a higher range (Figure 3-1B).

These observations are consistent with the hypothesis that p53 mutation or deficiency is associated with resistance to 5-FdU.

3.4.2 UDG depletion sensitizes cancer cells with p53 mutation or deficiency to 5-FdU exposure

Previously, the discordant findings on sensitization to 5-FdU following UDG depletion were reported using cell models with differing p53 status [15, 36, 41-47]. To understand whether the divergent responses could be attributed to p53 status, we explored whether UDG depletion could sensitize p53 mutant or deficient cancer cells to 5-FdU differentially. For these experiments, we used shRNA to deplete UDG in various cancer cells lines with differing p53 status, as listed in Table 3-1. UDG stable knockdown was evaluated by western blot (Figure 3-2A and B, insert). Based on a clonogenic cell survival assay, we observed that UDG depletion selectively sensitized cells with p53 mutation or deficiency to 5-FdU exposure (Figure 3-2A). However, in p53 WT cell lines, UDG

Page | 81 depletion did not alter the cytotoxicity of 5-FdU (Figure 3-2B). Collectively, these results demonstrate that UDG depletion re-sensitizes p53 mutant or deficient cancer cells, providing a novel therapeutic target for patients with p53 mutant tumors.

3.4.3 5-FdU resistance in p53 knockout (KO) or knockdown (KD) cells is reversed by

UDG depletion

Since many studies have identified gain of various functions for specific p53 mutated proteins [143, 144], we next asked whether loss of wild-type p53 protein expression can alter the response to 5-FdU. To address this, we utilized paired HCT116 colon cancer cell lines with or without genetic TP53 deletion and tested their sensitivity to

5-FdU, and the loss of p53 expression was evaluated by western blot (Figure 3-3A). Using a clonogenic survival assay, we demonstrated that p53 KO cells were more resistant to 5-

FdU than p53 WT cells (Figure 3-3B). Knockdown of p53 by shRNA recapitulates the resistance observed in p53 KO cells (Figure 3-3B), indicating that p53 status is a key mediator of the response of HCT116 cells to 5-FdU.

To understand whether loss of p53 protein will affect the response to 5-FdU after

UDG depletion, we knocked down UDG by shRNA in both HCT116 p53 WT and p53 KO cells. UDG knockdown levels were shown to be greater than 90% as evaluated by Q-PCR and western blot (Figure 3-3C, D). In agreement with our data using p53 mutant cells, UDG depletion greatly enhanced cytotoxicity of 5-FdU in p53 KO cells but did not significantly affect p53 WT cells (Figure 3-3E, F), indicating that p53 is involved in regulating the response to 5-FdU following UDG depletion. In addition, depletion of UDG also

Page | 82 potentiated 5-FdU cytotoxicity in two HCT116 cancer cells with different shRNAs targeted to p53 (Figure 3-4A-E). Collectively, these results confirm that loss of p53 protein renders cells resistant to 5-FdU, and UDG depletion selectively re-sensitizes p53 KO and KD cells to 5-FdU.

3.4.4 UDG depletion selectively sensitizes p53 KO cancer cells to pemetrexed and 5-

FU

Although all TS inhibitors have the ability to block TS, disrupting DNA replication and leading to uracil incorporation into DNA, differences among distinct TS inhibitors have been reported in terms of their other metabolism mediated cytotoxic pathways [2].

For example, pemetrexed polyglutamate derivatives also demonstrate inhibitory activity for other folate-dependent enzymes such as glycinamide ribonucleotide, but to a lesser extent [21-24]. Moreover, unlike 5-FdU, which mainly exerts its cytotoxicity due to effects at the DNA level [47], studies have revealed that the cytotoxicity of 5-FU is primarily

RNA-mediated, as 5-FU is metabolized to fluorouridine triphosphate (FUTP) which affects multiple RNA processes following its incorporation into RNAs [15-17, 57]. In order to address the question of whether p53 status is responsible for differences in sensitivity to other TS inhibitors, including pemetrexed and 5-FU, in UDG depleted cells, we evaluated cell viability following drug exposure in UDG depleted p53 WT and p53 KO cancer cells.

Similar to our observations with 5-FdU, no significant survival differences were found between UDG expressing and UDG depleted cells in the presence of p53 (Figure 3-5A, B).

However, in the absence of p53, UDG depletion sensitized cells to pemetrexed (Figure 3-

5C), while loss of UDG only moderately sensitized cells to 5-FU at high concentrations

Page | 83 (Figure 3-5D), reaffirming that the primary cytotoxic effect of 5-FU depends on RNA incoporation. Together, these results indicate that UDG depletion also sensitizes cells without p53 to other TS inhibitors, mainly through generation of DNA damage.

3.4.5 5-FdU activates cell death in p53 KO cancer cells with depleted UDG

To understand whether 5-FdU resistance observed in p53 KO cells is due to a failure to activate cell death pathways, we monitored cell death progression by Annexin V and propidium iodide (PI) staining. Cells were exposed to 5-FdU for 24 h, washed with

PBS, and then allowed to recover in drug free medium for a total of 48, 72, and 96 h (Figure

3-6A). In cells with wild-type p53, 5-FdU caused significant cell death (Annexin V and PI positive) at 48 h which was retained at 72 h and 96 h in both UDG expressing and UDG depleted cells (Figure 3-6B, C). However, in the absence of p53, cell death caused by 5-

FdU was significantly lower in UDG expressing cells, while in UDG depleted cells, cell death was detected at 24 h and significantly elevated at 48 to 96 h (Figure 3-6B, C). These data suggest that 5-FdU induced cell death is dependent upon p53, supporting the observation that drug resistance can be observed as a result of abrogation of the p53 mediated cell death pathway. Importantly, UDG depletion significantly potentiates death of cells lacking wild type p53 activity through a p53 independent pathway.

To further elucidate whether the cell death caused by 5-FdU is due to apoptosis, we examined expression of proteins involved in the activation of the apoptotic pathway. In wild-type p53 cells, we observed that p53 expression was induced at 24 h and the induction remained for 96 hours in both shSCR and shUDG cells following 5-FdU exposure (Figure

Page | 84 3-6D). The expression of cleaved PARP, a hallmark of apoptotic cell death, was induced at 48 h and persisted through 72 h and 96 h in p53 WT cells regardless of whether UDG was present or not (Figure 3-6D). In addition, cleaved caspase 3 was also detected in both

UDG expressing or depleted p53 WT cells (Figure 3-6D). In the absence of p53, induction of cleaved PARP or activated caspase 3 was not detected in cells expressing UDG after 5-

FdU exposure (Figure 3-6E), while both were robustly induced from 48 h to 96 h in cells depleted of UDG (Figure 3-6E). Taken together, our results suggest that 5-FdU induced apoptosis is mediated through p53 and the lack of apoptosis activation due to loss of p53 is responsible for the enhanced cell survival observed in p53 KO cells. However, in p53

KO cell with coincident UDG depletion, 5-FdU selectively activates a p53-independent apoptotic pathway through a mechanism which needs further investigation.

3.5 Discussion

In the present study, we utilized multiple cancer cells bearing differing p53 status with or without UDG expression. We observed that loss of UDG selectively re-sensitized cancer cells with p53 mutation or deficiency to 5-FdU, but did not alter the response of p53 wild-type cells. These results demonstrate that UDG, through its function of removing uracil or 5-FU, plays a major role in the effect of 5-FdU on the response of cells lacking wild type p53 activity. Our findings resolve the unexplained discrepancy observed in a number of prior studies regarding the role of UDG in sensitivity to TS inhibitors. Prior studies revealed that either loss of UDG enhanced the cytotoxicity of 5-FdU or pemetrexed in cancer cells [46, 47], or overexpression or inhibition of UDG had no effect on the

Page | 85 sensitivity of human or mouse cells to TS inhibition [36, 41-47]. The difference, we propose, is dependent on p53 status.

p53 plays a key role in determining the sensitivity of cells to 5-FU. A number of studies reported that enhanced 5-FU resistance has been observed in cells bearing TP53 deletion or mutations [79-83]. However, unlike other TS inhibitors, 5-FU exposure caused only slightly potentiated cytotoxicity at higher doses in UDG depleted, p53 KO cell lines.

Recently, several studies have observed that the cytotoxicity of 5-FU is more dependent on its incorporation into RNA than its inhibition of TS, diminishing its effect on DNA [15-17,

57]. In addition, activation of p53 following 5-FU exposure has been identified as working through RNA mechanisms [94, 200]. Since UDG recognizes only DNA lesions, it is not surprising that depletion of UDG does not significantly alter cellular responses to agents that primarily affect RNA function. Together, this suggests that the increased cytotoxicity of 5-FdU and pemetrexed observed in UDG depleted cells is primarily due to uracil and 5-

FU incorporation into DNA.

The present results illustrate that cells with p53 mutation or deficiency are significantly resistant to 5-FdU in comparison with p53 WT cells. It is clear that many different mutant p53s also acquire oncogenic functions that are distinct from the activities of wild-type p53 [143, 144]. Some p53 mutants provide enhanced resistance to apoptosis induced by a variety of treatments, including certain chemotherapeutic drugs [143, 144].

In particular, one study identified that p53 mutants activate expression of dUTPase [158], which has been related to the resistance to TS inhibitors [97, 98, 101-104]. Our results on a selected group of p53 mutants as well as p53 KO cell lines revealed resistance to 5-FdU treatment. However, the p53 KO cell line is much less resistant to 5-FdU than other p53

Page | 86 mutant cell lines, suggesting the potential for enhanced resistance due to gained functions for certain p53 mutants.

Our results demonstrated that inhibition of UDG selectively sensitized p53 mutant and deficient cancer cells to 5-FdU, but did not alter the response in p53 WT cells.

Importantly, we have observed that apoptosis following 5-FdU is efficiently induced in the presence of p53 but highly compromised in cells lacking p53, indicating that the activation of the 5-FdU induced cell death pathway is dependent on p53. Further studies with different p53 WT cell lines also revealed cells highly sensitive to 5-FdU with IC50 values lower than 10 nM. One exception we observed was in the A375 melanoma cells line, which has a wild type TP53 gene. A375 was relatively insensitive to 5-FdU and had an IC50 of 110.81

± 1.80 nM. Clearly, more knowledge is needed regarding the p53 mediated cell death pathway and how 5-FdU, with or without UDG, causes damage and triggers cell death. In response to 5-FdU, cells lacking wild-type p53, combined with UDG depletion, activate cell death in a p53 independent manner, which reverses chemoresistance and selectively re-sensitizes these cancer cells to 5-FdU.

Taken together, these results provide an explanation for the discordant findings in previous published data regarding the role of UDG in mediating the cytotoxicity of TS inhibitors and suggest that UDG is an attractive therapeutic target in cancer cells with p53 mutation or deficiency, to enhance their response to TS inhibitors.

Page | 87 Table 3-1. Cell lines and strains used in this work.

Page | 88 Figure 3-1. 5-FdU resistance in different types of cancer cells with p53 mutation or deficiency.

(A) Colony survival assay in cancer cells shown in Table 3-1 in response to increasing doses of 5-FdU. Cells with WT p53, dashed lines; cells with deficient p53, solid lines; The results represent three independent experiments that were done in duplicate each time. (B)

IC50 values of 5-FdU for cancer cells with WT p53 or deficient p53, respectively.

Page | 89 Figure 3-1. 5-FdU resistance in different types of cancer cells with p53 mutation or deficiency.

Page | 90 Figure 3-2. UDG depletion selectively sensitizes cells with p53 mutation or deficiency to 5-FdU.

Cancer cells stably transfected with shSCR or shUDG were analyzed by the Western blots to examine UDG levels (insert). Colony survival assays in (A) p53 mutated or deficient, and (B) p53 WT cancer cells treated with increasing doses of 5-FdU. The results represent three independent experiments that were done in duplicate each time. (*, P < 0.01).

Page | 91 Figure 3-2. UDG depletion selectively sensitizes cells with p53 mutation or deficiency to 5-FdU.

Page | 92 Figure 3-3. 5-FdU resistance due to loss of p53 is reversed by UDG depletion.

(A) p53 expression levels were analyzed by Western blot in HCT116 cells with wild-type p53 (p53WT), knockout of p53 (p53KO), shGFP expressing vector (shGFP), and shp53 expressing vector (shp53). (*, non-specific bands) (B) Colony survival assay for increasing doses of 5-FdU in HCT116 p53WT, p53KO, shGFP, and shp53 cells. HCT116 p53WT and p53KO cells stably transfected with non-targeted scramble control shRNA (shSCR) or

UDG-directed shRNA (shUDG) were analyzed by Western blot (C) and qPCR (D) to examine UDG levels. Colony survival assays for increasing doses of 5-FdU in (E) p53WT cells alone, or with shSCR or shUDG, and (F) p53KO cells alone, or with shSCR or shUDG.

Viable colonies (>50 cells) stained with methylene blue after 10 d of culture were counted.

The results represent three independent experiments that were done in duplicate. (*, P <

0.01).

Page | 93 Figure 3-3. 5-FdU resistance due to loss of p53 is reversed by UDG depletion.

Page | 94 Figure 3-4. p53 knockdown re-sensitizes cancer cells with UDG depletion to 5-FdU.

(A) HCT116 cells stably transfected with shGFP or shp53 (shp53-THM or shp53-Bleo) shRNAs were analyzed by the Western blots to examine p53 knockdown levels. (*, non- specific bands) (B) HCT116 cells expressing shGFP or shp53 (shp53-THM or shp53-Bleo) vectors were further transfected with non-targeted scramble control shRNA (shSCR) or

UDG-directed shRNA (shUDG). UDG mRNA levels were determined by qPCR. Colony survival assays for increasing doses of 5-FdU in (C) shGFP, (D) shp53-THM, and (E) shp53-Bleo transfected cells alone, or with shSCR or shUDG. Viable colonies (>50 cells) stained with methylene blue after 10 d of culture were counted. The results represent three independent experiments that were done in duplicate. (*, P < 0.01).

Page | 95 Figure 3-4. p53 knockdown re-sensitizes cancer cells with UDG depletion to 5-FdU.

Page | 96 Figure 3-5. UDG depletion selectively sensitizes p53 KO cells to pemetrexed and 5-

FU.

Colony survival assays in shSCR and shUDG p53 WT cells treated with increasing doses of (A) pemetrexed and (B) 5-FU. Colony survival assays in shSCR and shUDG p53 KO cells treated with increasing doses of (C) pemetrexed and (D) 5-FU. Viable colonies (>50 cells) stained with methylene blue after 10 d of culture were counted. The results represent three independent experiments that were done in duplicate. (*, P < 0.01).

Page | 97 Figure 3-5. UDG depletion selectively sensitizes p53 KO cells to pemetrexed and 5-

FU.

Page | 98 Figure 3-6. UDG depletion induces cell death caused by 5-FdU in p53 KO cancer cells.

(A) Schematic diagram of the treatment for HCT116 p53WT (shSCR and shUDG) and p53KO (shSCR and shUDG) cells with 25 nM 5-FdU for 24 h, washed, replenished with drug-free medium at indicated time points. (B) Untreated (Unt) or treated cells were subjected to FITC Annexin V and propodium iodide (PI) staining and analyzed by flow cytometry. Representative flow plots of three independent experiments are shown. (C) Cell death is expressed as 100%-viable cells (Annexin V negative and PI negative). Values indicate mean values ± SD. All experiments were performed independently for three times.

(*P < 0.01). Protein expression involved in regulation of apoptotic cell death in response to 5-FdU were detected in HCT116 (D) p53 WT (shSCR and shUDG) and (E) p53 KO

(shSCR and shUDG) cells. (*, non-specific bands)

Page | 99 Figure 3-6. UDG depletion induces cell death caused by 5-FdU in p53 KO cancer cells.

Page | 100 CHAPTER 4. CONCLUSIONS AND FUTURE DIRECTIONS

4.1 Conclusions

The present work provides novel mechanistic insights into the role of uracil DNA glycosylase as a DNA repair protein in the effectiveness of chemotherapeutic agents like

5-FdU in cancer cells with regard to different p53 status. Key research findings include:

1. UDG plays a key role in elimination of uracil and 5-FU incorporation from genomic

DNA following 5-FdU treatment. Loss of UDG results in persistent uracil and 5-

FU incorporation into DNA, which lead to increased sensitization of cancer cells to

5-FdU. These data suggest a critical role for UDG in regulating the cell killing

effect of 5-FdU.

2. UDG is required for recovery from cell cycle arrest induced by 5-FdU treatment.

Accordingly, UDG depleted cells are arrested at late G1 and early S phase and

display stalled replication progression following 5-FdU exposure.

3. The sustained DNA damage, likely due to replication fork collapse, contributes to

the genotoxicity of 5-FdU in the absence of UDG, which cannot be rescued by a

caspase inhibitor. These results indicate that the persistent DNA damage is the

cause but not the consequence of cell death.

Page | 101 4. Cells with p53 mutation or deficiency exhibit 5-FdU resistance which can be

significantly reversed by UDG depletion. These data provide a potential therapeutic

approach to restore the sensitivity of 5-FdU in p53 mutant or deficient cancer cells

by targeting UDG.

5. The resistance of 5-FdU in p53 mutant or deficient cancer cells is due to the

inability of cells to activate the apoptosis pathway, while loss of UDG which

induces genomic uracil and 5-FU incorporation catalyzes significant cell death

signals in cancer cells mutant or deficient in p53.

Currently, chemotherapies are still the mainstay of cancer treatment and are likely to remain so in the foreseeable future. However, severe side effects are often incompatible with such therapies. Consequently, much effort now is concentrated on the exploitation of the difference between normal and cancer cells in order to selectively potentiate the cytotoxicity of chemotherapies in cancer. Significantly, our findings demonstrated that the

UDG initiated BER pathway is a key determinant of the cytotoxicity of TS inhibitor 5-FdU in p53 mutant and deficient cancer cells, which are commonly refractory to cell death as a result of chemotherapy. Thus, UDG can be manipulated for the therapeutic purposes so as to selectively augment the cell killing effect of TS-targeted chemotherapy in p53 mutant and deficient cancers with a broader therapeutic index.

Page | 102 4.2 Future directions

In order to extend our current understanding of the enhanced sensitivity to 5-FdU via UDG depletion in p53 mutant and deficient cancer cells, our continued investigative efforts will focus on the following major areas:

4.2.1 Evaluation of the interplay between p53 and thymidylate synthase (TS) on cytotoxicity of TS inhibitors

Background and significance:

In addition to acting as an important chemotherapeutic target, thymidylate synthase

(TS) itself can function as a RNA binding protein and regulate protein translation [72].

This function was first identified by Allegra’s group who found that the ligand-free TS mediates an autoregulatory mechanism via interaction with its own mRNA that suppresses translation [70]. After that, several studies have revealed that TS can also form a ribonucleoprotein complex with and repress the translation of many other cellular mRNAs, including the mRNA encoded by TP53 gene [72, 201]. Based on these studies, it has been reported that inducible TS expression in p53 wild-type cells blocks p53 induction following treatment with TS-targeted antifolates, but not 5-FU, and inactivation of p53 significantly increases the resistance to TS inhibitors, indicating that the modulation of p53 by TS affects the efficacy of TS inhibitors [202]. Subsequently, another study, from a different angle, found that expression of mutant p53 in cells elevates TS expression, which causes increased resistance to TS inhibitors [203]. Since both TS and p53 status are identified as

Page | 103 critical determinants of the sensitivity to TS inhibitors, the interplay between TS and p53 may govern the response to TS-targeted chemotherapy. Therefore, we hypothesize that in p53 wild-type cancer cells, resistance to 5-FdU is due to the repression of p53 through enhanced TS expression, while in p53 mutant cancer cells, resistance is caused by p53 mutations mediated TS induction. More importantly, understanding the relationship between TS and p53 will provide us the mechanistic insights into the selective sensitivity to 5-FdU observed in p53 mutant or deficient cancer cells through UDG depletion.

Experimental design and expected results:

To delineate the interplay between TS and p53 that may potentially affect the cytotoxicity of 5-FdU in p53 wild-type and mutant cancer cells, we will (1) systematically evaluate the status of TS and p53 in p53 wild-type and mutant cancer cells with and without

5-FdU exposure at genetic, transcriptional and translational levels, and correlate these knowledge to the corresponding IC50 and the levels of genomic uracil and 5-FU incorporation as a result of 5-FdU treatment; (2) establish a model cell line with wild-type p53 but are resistant to 5-FdU through exposing cells to sub-lethal doses of 5-FdU at step- wise incremental levels, and subsequently assess the TS expression level, p53 status, and incorporated uracil and 5-FU levels in the presence and absence of UDG. Of note, one of our p53 wild-type cell line, A375 is distinct from the wild-type p53 group in that it displays significantly higher resistance to 5-FdU, thus the TS status and the link with 5-FdU resistance can also be evaluated within this cell line; (3) understand the mechanisms of TS induction driven by different p53 mutations in a transcriptionally dependent or independent

Page | 104 manner, and investigate whether the induction of TS activity is a consequence of gain of functions for p53 mutants.

We propose that in 5-FdU sensitive cell lines with wild-type p53, TS expression or induction level will be low, and 5-FdU treatment will highly activate p53 mediated cell death pathway. In 5-FdU resistant cell model with wild-type p53, TS expression will be significantly elevated which will consequently suppress p53 expression as well as p53 mediated cell death pathway. While, in 5-FdU resistant cell lines with mutant p53, TS expression will be highly induced as a result of gain of functions for mutant p53. However, since the molecular determinants of the sensitivity to 5-FdU may or may not coordinate with each other in an intricate network, and there might be other components such as dUTPase that can also contribute to this complexity, the interplay between p53 and TS may not necessarily correlate with the cytotoxicity of 5-FdU, or even one of them may override the pathway to determine the sensitivity of 5-FdU.

Page | 105 4.2.2 Characterization of the effect of other glycosylases (SMUG1, TDG, MBD4) in processing genomic uracil and 5-FU induced by 5-fluorodeoxyuridine

Background and significance:

The base excision repair (BER) machinery is vital to protect cells from endogenous and exogenous DNA damages in the form of small chemical modifications to the DNA base, such as oxidation, alkylation, and deamination, which can be mutagenic or cytotoxic to the cells. DNA glycosylases initiate BER pathway and play an essential role in the recognition and elimination of the damaged bases as well as maintenance of the DNA architecture with high accuracy. As described in previous sections, TS inhibition by fluoropyrimidines and antifolates generates abnormal bases uracil and 5-FU incorporation into DNA. There are four glycosylases identified in mammalian cells which constitute the removal activities towards uracil and 5-FU [35]. Uracil DNA glycosylase (UDG) and single-strand-selective monofunctional uracil-DNA glycosylase (SMUG1) prefer single stranded DNA as substrate, but also excise uracil and 5-FU from double stranded DNA, whereas G/T mismatch-specific thymine DNA glycosylase (TDG) and methyl-CpG- binding domain protein 4 (MBD4 or MED1) are strictly specific to double stranded DNA and have very slow turnover rate [196, 204].

Among these four glycosylases, UDG is the most efficient enzyme in processing uracil and 5-FU in DNA. Moreover, binding of nuclear UDG (UDG2) with RPA and

PCNA localizes it to the sites of DNA replication, where it can rapidly eliminate uracil and

5-FU incorporation in a post-replicative manner [37, 38]. SMUG1 was initially found capable of removal uracil in extracts derived from Xenopus and appears to act as a UDG

Page | 106 backup enzyme in limiting uracil accumulation [205]. In the context of 5-FU exposure, studies reported that excision of uracil and 5-FU from DNA by SMUG1 reduces drug cytotoxicity, indicating accumulation of these abnormal bases in the genome is correlated with sensitivity to 5-FU [48]. TDG and MBD4 specifically remove mismatched uracil or thymine in double-stranded DNA and exhibit highest processing efficiency within CpG context throughout the genome [206]. Of particular note are the findings that TDG prefers

U:G over T:G pairs due to the steric hindrance caused by the C-5 methyl group, and is able to excise 5-FU from DNA with high efficiency, regardless of whether paired with adenine or guanine [207, 208]. However, unlike other uracil glycosylases, TDG has an extremely low enzymatic turnover number. After removal of DNA base, it is tightly bound to the resultant abasic site (AP site). Thus the release of TDG from its abasic product is considered as the rate-limiting step for its glycosylase reaction [205, 206]. Interestingly, in contrast with SMUG1, a study has reported that expression of TDG contributes to the cytotoxicity of 5-FU, possibly explained by its slow dissociation after lesion removal which accumulates the toxic AP-site intermediates and blocks the assembly of downstream repair machinery [49]. This effect ultimately results in delayed S-phase progression, formation of double strand breaks, and activation of DNA damage response in the cells.

Provided that other glycosylase SMUG1, TDG, or MBD4 has been shown in vitro assays to act on uracil and 5-FU pairs and also to affect the cellular cytotoxic activity of

TS inhibitor 5-FU, we raised the question that if any of the cellular uracil or 5-FU presents in DNA after treatment with 5-FdU in UDG depleted cells would be recognized and removed by other glycosylases. To address this question, the AP site detection assay developed from our lab was utilized which provide us a feasible method to quantify the

Page | 107 mis-incorporated bases that might be potentially excised by different glycosylases.

Importantly, understanding the levels of uracil or 5-FU removal by these glycosylases in

cells will provide the direct evidence for the molecular mechanisms responsible for 5-FdU

cytotoxicity.

Preliminary experiments:

Evaluation of cellular levels of U:G and 5-FU:G pairs recognized by TDG through

AP site detection assay.

Given the fact that in addition to UDG, TDG also recognizes uracil and 5-FU which

are preferentially mispaired with guanine, we isolated genomic DNA from shSCR-

transfected and shUDG-transfected cells treated with 100 or 200 nM of 5-FdU for 24 hours,

and exposed the extracted DNA to exogenous glycosylase UDG, TDG, or UDG+TDG to

evaluate the level of AP sites by a near infrared (NIR) cyanine-based probe. We expected

that the fluorescent intensity of the probe (reflecting levels of AP sites) would correlate

with the amount of U:G and 5-FU:G pairs after incubation with TDG. In shSCR-transfected

cells, levels of AP sites remained low and consistent irrespective of 5-FdU after the

incubation with the equal amount of UDG, TDG, or combined the equal amount of

UDG+TDG (Figure 4-1A). In shUDG-transfected cells, levels of AP sites were elevated

following 24 h of 5-FdU exposure in UDG treated DNA but remained unchanged in TDG

treated DNA, and there was no addictive effect following UDG+TDG treatment (Figure 4-

1B). These data indicate that the main mispairs are U:A and/or 5-FU:A following 5-FdU exposure, and UDG, but not TDG is the predominant glycosylase in the cells to prevent uracil and 5-FU incorporation into DNA.

Page | 108 Figure 4-1. Evaluation of uracil and 5-FU levels in TDG treated genomic DNA

extracted following 5-FdU exposure via AP site detection assay.

DLD1 shSCR (A) and shUDG (B) cells were treated with 0, 100 and 200 nM 5-FdU for

24 h. Genomic DNA was isolated from phenol-chloroform extraction following treatment

with 5-FdU and subsequently dissolved in reaction buffer (20 mM Tris-HCl, 1 mM EDTA and 1 mM dithiothreitol, pH 8.0). The DNA extracts were treated in vitro with purified

UDG (50ng), TDG (50ng), or UDG(50ng)+TDG(50ng) at 37 °C for 1 h. AP sites detection was performed by incubation of DNA with a cyanine-based AP site probe with a final concentration of 25 µM. Following incubation, extracted DNA was precipitated, and the supernatant was discarded. DNA pellets were resuspended in H2O, and DNA concentrations were measured and adjusted. The fluorescence intensities of each sample were analyzed with 760 nm excitation and emission scan of 790-847 nm. Data represent mean and SD of relative fluorescence intensity normalized to the 5-FdU untreated shSCR

+UDG from three replicates.

Page | 109 Evaluation of UDG and TDG activities in vitro through glycosylase activity assay.

Since we did not observe the uracil or 5-FU removal activity of TDG in cells by an ex vivo AP site detection assay following 5-FdU exposure, we further did a glycosylase cutting assay with DNA duplexes containing uracil or 5-FU base to prove that the TDG protein (purchased from Abcam) is not a dead enzyme. In this assay, we incubated UDG enzyme (50ng) which serves as a control or increasing amount of TDG (0, 50, 250, 500ng) with a 40-mer DNA oligomer containing U:A, U:G, 5-FU:A, or 5-FU:G mispairs for 1 h at 37 degree. Subsequently, APE was utilized to generate the single strand breaks at resultant AP sites, which can be separated by gel electrophoresis as a way to evaluate glycosylases activity. According to the cutting assay, the activity of TDG compared with

UDG was very weak in terms of the ability to cut U:G pairs when we add the same amount of protein (50ng) (Figure 4-2). Furthermore, we did not observe any activity of TDG in removing 5-FU from 5-FU:G or 5-FU:A pairs (Figure 4-2). Therefore, this results may suggest that UDG is significantly superior in its activity of removal uracil and 5-FU lesions, while TDG activity in terms of these two substrates is not very efficient. In addition, this data may explain the lack of AP site induction in TDG treated genomic DNA after 5-FdU as a result of its poor activity.

Page | 110 Figure 4-2: Evaluation of UDG and TDG activities via glycosylase activity assay

Incubation of 3’-Tamra labeled 40-mer oligonucleotide containing U:G, 5-FU:G, 5-FU:A, or U:A base pair with UDG (50ng) or increasing amount of TDG (0, 50, 250, 500ng) for

1h at 37 degree. Subsequent single stand breaks were generated by incubation with APE and resolved on 20% denaturing polyacrylamide gels (5.3 g urea, 5.0 mL 40% acrylamide,

2.3 mL 5X TBE buffer, 200 μL 10% APS, and 20 μL TEMED) by electrophoresis in the dark. Gels were visualized by a Typhoon Tri + Variable Mode Imager (Amersham

Biosciences).

Page | 111 So far, we have conducted a careful analysis of the possible involvement of TDG in the elimination of uracil and 5-FU incorporation in the context of mispairing with guanine. Based on the results that genomic uracil and 5-FU incorporation were not perceived by AP site detection assay in TDG treated DNA extracts, we might be able to conclude that U:G or 5-FU:G are not the prevailing mispairs formed as a result of treatment with 5-FdU. However, although at this moment, these data do not support a strong role of

TDG in the removal of abnormal base pairs at these particular experimental settings, we cannot exclude the following possibilities: First, endogenous TDG expressed in cells removed U:G and 5-FU:G pairs before the DNA is extracted for analysis. Second, the AP site detection assay is not sensitive enough to detect the amount of AP sites released from

U:G and 5-FU:G pairs. In addition, the glycosylase activity towards 5-FU base paired with either adenine or guanine is not obviously observed in our cutting assays, which further contributes to the uncertainty about the presence of U:G or 5-FU:G pairs as a consequence of 5-FdU exposure.

Experimental design and expected results:

To further clarify the role of other glycosylases in processing uracil and 5-FU bases after 5-FdU exposure, we will (1) perform the AP site detection assay to quantify the mis- incorporated uracil and 5-FU bases that can be recognized by other glycosylases; (2) deplete each of the glycosylase alone or together with UDG and evaluate the cytotoxicity of 5-FdU subsequently; (3) overexpress TDG or MBD4 to see if the cytotoxicity of 5-FdU will be exacerbated in UDG depleted cells.

Page | 112 Previous studies revealed that expression of TDG contributes to the cytotoxicity of

5-FU in a manner distinct from UDG [49]. As the turnover rate for TDG is extremely slow,

TDG binds to AP site with a high affinity, which blocks the accessibility of downstream

BER repair proteins. Interestingly, it has been shown that expression of both UDG and

TDG is under cell cycle regulation [188, 189]. UDG expression and activity are low in early G1 phase but highly induced at late G1 and early S phase, while TDG levels were degraded through the ubiquitin-protease system just when UDG starts to come up, suggesting a finely coordinated uracil repair system by distinct pathways throughout the cell cycle. Our study so far indicates that UDG depletion results in a high degree of uracil and 5-FU incorporation into DNA induced by 5-FdU exposure during S phase, which precipitates the cytotoxicity of 5-FdU. Thus it is possible to envision that once these large amounts of unresolved uracil and 5-FU lesions formed during S phase enter into G2/M phase, the slow acting enzyme TDG will not only be overwhelmed by the lesions, but also binds to the lesions and obstructs the downstream repair pathway in a manner different from UDG. Under these assumptions, expressing of TDG would synergize with the loss of

UDG to potentiate the cytotoxicity of TS inhibitors.

Page | 113 4.2.3 Exploitation of DNA damage response signaling as a target to achieve ‘synthetic sickness or lethality’ with UDG depletion in p53 mutant cancer cells after DNA damage

Background and Significance:

Targeting DNA damage response and cell cycle checkpoint signaling pathways is now an emerging field to manage the efficacy of chemotherapy [209-211]. As TS inhibition blocks DNA replication and incorporates uracil and 5-FU into DNA during S phase, this strategy is highly relevant to TS-targeted therapy. The ATM-Chk2 and ATR-Chk1 pathways are two central signal transduction processes to facilitate DNA repair and promote cell survival in response to DNA damage. ATM-Chk2 is potently activated by double strand breaks caused by radiation or genotoxic chemotherapies, while ATR-Chk1 is strongly triggered when DNA replication is impeded [212]. Generally, TS inhibition causes cell cycle arrest at S phase due to the lack of dTTP [175-177]. It has been shown that ATR-Chk1 signaling pathway can be activated through TS inhibition. For instance,

Chk1 phosphorylation was activated following 5-FdU, raltitrexed and methotrexate treatment [53, 213-215]. Loss of functional ATR due to mutation or siRNA sensitized cancer cells to antifolates and 5-FdU [53, 214], and genetically ablation or pharmacological inhibition of Chk1 by UCN-01 also improved 5-FU killing effect [216-218].

Because G1 checkpoint is primarily dependent on the function of p53 and its downstream target p21, and cancer cells often suffer from the loss of wild-type p53, it is conceivable that cancer cells deficient in functional p53 would be more reliant on the Chk1 mediated cell cycle arrest than would p53 wild-type cells. Thus the synthetic sickness or

Page | 114 lethality might be achieved in tumors deficient in p53 through disruption of the complementary checkpoint component following DNA damage (Figure 4-3) [219-220].

This concept has been revealed in several studies that inhibition of Chk1 selectively sensitized p53 mutant or deficient cancer cells to chemotherapeutic agents including topoisomerase inhibitors, camptothecin, doxorubicin, and cisplatin [221-223]. According to our cell cycle and DNA replication analysis, we have observed that the p53 mutant UDG expressing cells undergo a transient slowing of S phase and ultimately resume the normal cycle progression, while UDG depleted cells experience a prolonged cell cycle arrest at late G1 and early S phase. This sustained cell cycle arrest, we propose, is due to activation of Chk1. Therefore, based on this model, we hypothesize that synthetic sickness or lethality can be achieved through UDG depletion (which induces persistent S phase arrest through uracil and 5-FU incorporation) and Chk1 inhibition (which disrupts S phase arrest) in p53 mutant or deficient cancer cells after treatment with 5-FdU.

Experimental design and expected results:

To define whether inhibition of Chk1 would trigger the synthetic sickness or lethality, we will (1) document whether ATR-Chk1 pathway is differentially activated in

UDG wild-type and UDG depleted cells, and compare the levels in p53 wild-type and p53 mutant cells after 5-FdU; (2) block the activity of Chk1 pharmacologically or genetically, and subsequently evaluate the pattern of cell cycle progression and cellular cytotoxic response following 5-FdU exposure.

We speculate that inhibition of Chk1 will synergize with the loss of wild-type p53 to abolish cell cycle arrest induced by 5-FdU in UDG depleted cells, which will ultimately

Page | 115 lead to severe cell death like the mitotic catastrophe. However, there are some studies identified additional compensatory pathways involving p38MAPK/MK2 to arrest cell cycle and contribute to cell survival after DNA damage [222]. The activation of this pathway may limit the effectiveness of 5-FdU in our proposed models.

Page | 116

Figure 4-3. Synthetic sickness or lethality between DNA damage response

signaling and p53

p53 is required for G1 checkpoint, while Chk1 is required for both S-phase checkpoint

and G2 checkpoint. Genotoxic agents that induce replicative stress trigger checkpoints

to arrest cell cycle for DNA repair and survival. Cancer cells with p53 mutation or

deficiency become dependent on G2/M arrest. When p53 mutant or deficient cells are

subjected to genotoxic agents in combination with inhibition of Chk1, loss of all

checkpoints will lead to mitotic catastrophe without repairing damaged DNA. This

strategy provides a selective killing effect in p53 mutant and deficient cancer cells.

Page | 117 4.2.4 Exploration of the potential gain of functions for p53 mutations that are responsible for 5-FdU resistance

Background and Significance:

p53 mutations promote resistance to chemotherapy through multiple gain of functions including increased drug efflux, suppression of apoptotic pathways, potentiation of DNA repair, which are systematically reviewed in Chapter 1.7. However, key molecular effectors driven by p53 mutation that may contribute to 5-FdU resistance still remain elusive. Since higher IC50 values of 5-FdU for p53 mutant cells are observed when compared to p53 KO cells, it is highly likely that p53 mutations acquire gain-of-function activities that dampen 5-FdU sensitivity. Hence, we hypothesize that p53 gain-of-function mutation endows the p53 protein with novel abilities that render cell resistance to 5-FdU.

In particular, among the six “hotspot” missense mutations, the most frequently mutated residues, R248 and R273, will be the focus of this study.

Experimental design and expected results:

To dissect whether p53 mutation alters target genes that may result in 5-FdU resistance, we will (1) express mutant p53 (R248Q or R273H) in a p53-null background and address the changes in transcriptional profile by chromatin immunoprecipitation

(ChIP)-sequencing; (2) map sequence reads in genome, identify enrichment in gene promoters that associated with mutant p53 in our data set, and compare with the genome- wide analysis of wild-type p53 binding promoters; (3) focus on genes involved in pathways including metabolism of TS inhibitors, DNA damage response, and DNA repair, and select

Page | 118 a set of genes to validate our data set. (4) characterize whether the identified genes targeted by mutant p53 alter biological effects that are responsible for 5-FdU resistance. We propose that the identified mutant p53 DNA-binding sites will be different from the wild-type p53 consensus binding sites. As mutant p53 lacks sequence-specific DNA binding activity, some studies have shown the recruitment of mutant p53 to promoters through protein- protein interactions with other transcription factors [139]. Thus the DNA binding sites that come up from our data set may not necessarily indicate a direct interaction with mutant p53. However, by analyzing the consensus sequence for other transcriptional factors that will be over-represented in the data set, the protein-protein interaction partner for mutant p53 can be identified. In addition, this assay only considers possible gain-of-function at transcriptional level and will exclude other possible gain-of-functions happen in cytosolic compartment, which may also be able to cause 5-FdU resistance.

Page | 119 4.2.5 Identification of small molecule inhibitors of UDG via high-throughput screening

Background and Significance:

Because the enhanced cytotoxicity was observed in a panel of cancer cells with

UDG depletion by shRNA, we propose that pharmacological compounds capable of blocking either catalytic activity or DNA binding ability of UDG will result in potentiation of the cytotoxicity of TS-targeted chemotherapy. In vitro high-throughput screening of small molecules combined with a series of confirmatory experiments will be conducted to identify cellular compatible compounds that block UDG. And the chemo-potentiation effect of putative compounds will be evaluated in cellular and animal models.

Experimental design:

(1) In vitro screening of the candidate compounds via molecular beacon assay: We

have developed and optimized a molecular beacon assay with Drew Adams,

director of the Small Molecule Screening SR of the Case CCC. In this assay, we

utilized a DNA oligomer with hairpin configuration that contains multiple uracil

bases with a 5’-FAM end quenched by 3’-dabsyl moiety. When multiple uracils are

removed from the oligomer after exposing to UDG, two paired strands will separate

spontaneously, releasing fluorescence on the FAM group. Commercially

synthesized purified human UDG with full-length (34kDa) expressed in E. coli will

be incubated with small molecules library, and compounds which give rise to

reduced fluorescence intensity over time will be characterized. The most active

Page | 120 compounds will be selected and confirmed in a secondary glycosylase activity

assay that has been well characterized in the lab. Since we are evaluating the

reduction of fluorescence, possible components that would interfere with the

fluorescent activity will be considered. In addition, because the screening is based

on an in vitro assay, potential cellular permeability, specificity and activity of the

compounds might be a rate-limiting step in the future.

(2) In vivo characterization of the active compounds for the enhanced cytotoxicity: we

will utilize the AP site detection assay developed from our lab to identify the

inhibitory activity of the compounds toward UDG in cells. And the synergistic

effect of the combinatory treatment with UDG and TS inhibitors will be evaluated

in cells and animal models.

Page | 121

Appendix

Page | 122 Appendix 1. Retention of uracil and 5-FU in HEC1A UDG depleted cells during thymidine recovery following 5-FdU exposure.

HEC1A shSCR and shUDG cells were treated with 100 nM 5-FdU for 24 h, then washed twice with PBS, and incubated in drug-free media supplemented with 20 µM thymidine

(Thy) for 6, 12, or 24 h. Genomic DNA was extracted and treated in vitro with purified

UDG. AP sites detection was performed by incubation of DNA with a cyanine-based AP site probe. Data represent mean and SD of relative fluorescence intensity normalized to the shSCR DNA without 5-FdU treatment from three replicates.

Page | 123 Appendix 2. Loss of UDG induces HEC1A cell cycle arrest at late G1 and early S phase following 5-FdU exposure.

Quantification of each phases of the cell cycle for unsynchronized HEC1A (A) shSCR and

(B) shUDG cells untreated (Unt) or treated with 100 nM 5-FdU for 24 h. After treatment, cells were pulsed with BrdU for 45 minutes, fixed, and stained with anti-BrdU antibody and PI dye. Cell cycle profiles were analyzed by flow cytometry. eS = early S-phase; mS

= mid-S-phase; lS = late S/G2-phase. Data represent mean and SD from 3 independent experiments.

Page | 124 Appendix 3. DNA damage accumulates in HEC1A UDG depleted cells in a caspase independent manner.

(A) Schematic diagram of the treatment of HEC1A cells with 5-FdU in the presence or absence (+/-) of 10 µM caspase inhibitor Q-VD-OPh at indicated time points. (B) HEC1A shSCR and shUDG cells were treated with 50 nM 5-FdU for 48, 72, and 96 h with (+) and without (-) 10 µM Q-VD-OPh. Cells were fixed and stained with anti-γH2AX antibodies.

γH2AX foci was visualized on a fluorescence microscope. (C) Quantification of the number of γH2AX foci per cell for 0, 48, 72, and 96 h of 5-FdU treatment in the presence

(+) or absence (-) of Q-VD-OPh. The statistical analysis of γH2AX foci per cell across the populations analyzed (n > 100 cells per population) is shown as a scatter plot with medians and the interquartile ranges. (D) Quantification of the percentage of cells with >10 γH2AX foci per cell for 0, 48, 72, and 96 h of 5-FdU treatment. Statistical analysis was performed as in C. (E) In parallel samples from B, the expression level of cleaved PARP was analyzed for cells untreated (Unt) or treated with 50 nM 5-FdU for 48, 72, and 96 h in the presence

(+) or absence (-) of 10 µM caspase inhibitor Q-VD-OPh.

Page | 125 Appendix 3. DNA damage accumulates in HEC1A UDG depleted cells in a caspase independent manner.

Page | 126 Appendix 4. Effect of irradiation on p21 induction in various cancer cells with different p53 status.

Western blots analysis of p53 and p21 expression in cells treated with (+IR) or without (-

IR) gamma-irradiation for 8Gy. Cells were collected 24 hours after irradiation. Tubulin was used as loading control.

Page | 127 REFERENCES

1. Longley DB, Harkin DP, Johnston PG. 5-fluorouracil: mechanisms of action and

clinical strategies. Nat Rev Cancer. 2003; 3(5):330-8.

2. Wilson PM, Danenberg PV, Johnston PG, Lenz HJ, Ladner RD. Standing the test of

time: targeting thymidylate biosynthesis in cancer therapy. Nature reviews Clinical

oncology. 2014; 11(5):282-98.

3. Rahman L, Voeller D, Rahman M, Lipkowitz S, Allegra C, Barrett JC, Kaye FJ,

Zajac-Kaye M. Thymidylate synthase as an oncogene: a novel role for an essential

DNA synthesis enzyme. Cancer cell. 2004; 5(4):341-51.

4. Li LS, Morales JC, Veigl M, Sedwick D, Greer S, Meyers M, Wagner M, Fishel R,

Boothman DA. DNA mismatch repair (MMR)‐dependent 5‐fluorouracil

cytotoxicity and the potential for new therapeutic targets. British journal of

pharmacology. 2009; 158(3):679-92.

5. Wyatt MD, Wilson DM. Participation of DNA repair in the response to 5-

fluorouracil. Cellular and molecular life sciences. 2009 Mar 1;66(5):788-99.

6. Santi DV, McHenry CS. 5-Fluoro-2′-deoxyuridylate: covalent complex with

thymidylate synthetase. Proceedings of the National Academy of Sciences. 1972;

69(7):1855-7.

7. Matthews DA, Appelt K, Oatley SJ, Xuong NH. Crystal structure of Escherichia coli

thymidylate synthase containing bound 5-fluoro-2′-deoxyuridylate and 10-

propargyl-5, 8-dideazafolate. Journal of molecular biology. 1990 Aug

20;214(4):923-36.

Page | 128 8. Kaiyawet N, Rungrotmongkol T, Hannongbua S. Effect of halogen substitutions on

dUMP to stability of thymidylate synthase/dUMP/mTHF ternary complex using

molecular dynamics simulation. Journal of chemical information and modeling.

2013; 53(6):1315-23.

9. Samuelsson T. Interactions of transfer RNA pseudouridine synthases with RNAs

substituted with fluorouracil. Nucleic acids research. 1991; 19(22):6139-44.

10. Patton JR. Ribonucleoprotein particle assembly and modification of U2 small

nuclear RNA containing 5-fluorouridine. Biochemistry. 1993; 32(34):8939-44.

11. Gustavsson M, Ronne H. Evidence that tRNA modifying enzymes are important in

vivo targets for 5-fluorouracil in yeast. RNA. 2008; 14(4):666-74.

12. Ghoshal K, Jacob ST. Specific inhibition of pre-ribosomal RNA processing in

extracts from the lymphosarcoma cells treated with 5-fluorouracil. Cancer research.

1994; 54(3):632-6.

13. Zhao X, Yu YT. Incorporation of 5-fluorouracil into U2 snRNA blocks

pseudouridylation and pre-mRNA splicing in vivo. Nucleic acids research. 2007;

35(2):550-8.

14. Silverstein RA, de Valdivia EG, Visa N. The incorporation of 5-fluorouracil into

RNA affects the ribonucleolytic activity of the exosome subunit Rrp6. Molecular

Cancer Research. 2011; 9(3):332-40.

15. Pettersen HS, Visnes T, Vågbø CB, Svaasand EK, Doseth B, Slupphaug G, Kavli

B, Krokan HE. UNG-initiated base excision repair is the major repair route for 5-

fluorouracil in DNA, but 5-fluorouracil cytotoxicity depends mainly on RNA

incorporation. Nucleic acids research. 2011; 39(19):8430-44.

Page | 129 16. Pritchard DM, Watson AJ, Potten CS, Jackman AL, Hickman JA. Inhibition by

uridine but not thymidine of p53-dependent intestinal apoptosis initiated by 5-

fluorouracil: evidence for the involvement of RNA perturbation. Proceedings of the

National Academy of Sciences. 1997; 94(5):1795-9.

17. Brody JR, Hucl T, Costantino CL, Eshleman JR, Gallmeier E, Zhu H, van der

Heijden MS, Winter JM, Wikiewicz AK, Yeo CJ, Kern SE. Limits to thymidylate

synthase and TP53 genes as predictive determinants for fluoropyrimidine

sensitivity and further evidence for RNA-based toxicity as a major influence.

Cancer research. 2009; 69(3):984-91.

18. Scagliotti GV, Parikh P, Von Pawel J, Biesma B, Vansteenkiste J, Manegold C,

Serwatowski P, Gatzemeier U, Digumarti R, Zukin M, Lee JS. Phase III study

comparing cisplatin plus gemcitabine with cisplatin plus pemetrexed in

chemotherapy-naive patients with advanced-stage non–small-cell lung cancer.

Journal of clinical oncology. 2008; 26(21):3543-51.

19. Goudar RK. Review of pemetrexed in combination with cisplatin for the treatment

of malignant pleural mesothelioma. Therapeutics and clinical risk management.

2008; 4(1):205.

20. Chattopadhyay S, Zhao R, Krupenko SA, Krupenko N, Goldman ID. The inverse

relationship between reduced folate carrier function and pemetrexed activity in a

human colon cancer cell line. Molecular cancer therapeutics. 2006; 5(2):438-49.

21. Hanauske AR, Chen V, Paoletti P, Niyikiza C. Pemetrexed disodium: a novel

antifolate clinically active against multiple solid tumors. The Oncologist. 2001;

6(4):363-73.

Page | 130 22. Goldman ID, Zhao R. Molecular, biochemical, and cellular pharmacology of

pemetrexed. InSeminars in oncology 2002 (Vol. 29, No. 6, pp. 3-17). WB Saunders.

23. Grindey GB, Shih C, Barnett CJ, Pearce HL, Engelhardt JA, Todd GC, Rinzel SM,

Worzalla JF, Gossett LS, Everson TP, Wilson TM. LY231514, a novel

pyrrolopyrimidine antifolate that inhibits thymidylate synthase (TS). InProc Am

Assoc Cancer Res 1992 (Vol. 33, p. 411)

24. Shih C, Chen VJ, Gossett LS, Gates SB, MacKellar WC, Habeck LL, Shackelford

KA, Mendelsohn LG, Soose DJ, Patel VF, Andis SL. LY231514, a pyrrolo [2, 3-d]

pyrimidine-based antifolate that inhibits multiple folate-requiring enzymes. Cancer

research. 1997; 57(6):1116-23.

25. Ahmad SI, Kirk SH, Eisenstark A. Thymine metabolism and thymineless death in

prokaryotes and eukaryotes. Annual Reviews in Microbiology. 1998; 52(1):591-

625.

26. Hanawalt PC. A balanced perspective on unbalanced growth and thymineless death.

Frontiers in microbiology. 2015; 6:504.

27. Khodursky A, Guzmán EC, Hanawalt PC. Thymineless death lives on: new insights

into a classic phenomenon. Annual review of microbiology. 2015; 69:247-63.

28. Guzmán EC, Martín CM. Thymineless death, at the origin. Frontiers in

microbiology. 2015; 6:499.

29. Sangurdekar DP, Hamann BL, Smirnov D, Srienc F, Hanawalt PC, Khodursky AB.

Thymineless death is associated with loss of essential genetic information from the

replication origin. Molecular microbiology. 2010; 75(6):1455-67.

Page | 131 30. Curtin NJ. DNA repair dysregulation from cancer driver to therapeutic target.

Nature Reviews Cancer. 2012; 12(12):801-17.

31. Krokan HE, Bjørås M. Base excision repair. Cold Spring Harbor perspectives in

biology. 2013; 5(4):a012583.

32. Robertson AB, Klungland A, Rognes T, Leiros I. DNA repair in mammalian cells.

Cellular and molecular life sciences. 2009; 66(6):981-93.

33. Kim YJ, M Wilson III D. Overview of base excision repair biochemistry. Current

molecular pharmacology. 2012; 5(1):3-13.

34. Sharma RA, Dianov GL. Targeting base excision repair to improve cancer therapies.

Molecular aspects of medicine. 2007; 28(3):345-74.

35. Jacobs AL, Schär P. DNA glycosylases: in DNA repair and beyond. Chromosoma.

2012; 121(1):1-20.

36. Grogan BC, Parker JB, Guminski AF, Stivers JT. Effect of the thymidylate synthase

inhibitors on dUTP and TTP pool levels and the activities of DNA repair

glycosylases on uracil and 5-fluorouracil in DNA. Biochemistry. 2011; 50(5):618-

27.

37. Otterlei M, Warbrick E, Nagelhus TA, Haug T, Slupphaug G, Akbari M, Aas PA,

Steinsbekk K, Bakke O, Krokan HE. Post‐replicative base excision repair in

replication foci. The EMBO journal. 1999; 18(13):3834-44.

38. Hagen L, Kavli B, Sousa MM, Torseth K, Liabakk NB, Sundheim O, Peňa‐Diaz J,

Otterlei M, Hørning O, Jensen ON, Krokan HE. Cell cycle‐specific UNG2

phosphorylations regulate protein turnover, activity and association with RPA. The

EMBO journal. 2008; 27(1):51-61.

Page | 132 39. Ingraham HA, Dickey L, Goulian M. DNA fragmentation and cytotoxicity from

increased cellular deoxyuridylate. Biochemistry. 1986; 25(11):3225-30.

40. Ladner RD. The role of dUTPase and uracil-DNA repair in cancer chemotherapy.

Current Protein and Peptide Science. 2001; 2(4):361-70.

41. Welsh SJ, Hobbs S, Aherne GW. Expression of uracil DNA glycosylase (UDG)

does not affect cellular sensitivity to thymidylate synthase (TS) inhibition.

European Journal of Cancer. 2003; 39(3):378-87.

42. Andersen S, Heine T, Sneve R, König I, Krokan HE, Epe B, Nilsen H.

Incorporation of dUMP into DNA is a major source of spontaneous DNA damage,

while excision of uracil is not required for cytotoxicity of fluoropyrimidines in

mouse embryonic fibroblasts. Carcinogenesis. 2005; 26(3):547-55.

43. Kemmerich K, Dingler FA, Rada C, Neuberger MS. Germline ablation of SMUG1

DNA glycosylase causes loss of 5-hydroxymethyluracil-and UNG-backup uracil-

excision activities and increases cancer predisposition of Ung−/− Msh2−/− mice.

Nucleic acids research. 2012; 40(13):6016-25.

44. Luo Y, Walla M, Wyatt MD. Uracil incorporation into genomic DNA does not

predict toxicity caused by chemotherapeutic inhibition of thymidylate synthase.

DNA repair. 2008; 7(2):162-9.

45. Nagaria P, Svilar D, Brown AR, Wang XH, Sobol RW, Wyatt MD. SMUG1 but

not UNG DNA glycosylase contributes to the cellular response to recovery from 5-

fluorouracil induced replication stress. Mutation Research/Fundamental and

Molecular Mechanisms of Mutagenesis. 2013; 743:26-32.

Page | 133 46. Huehls AM, Huntoon CJ, Joshi PM, Baehr CA, Wagner JM, Wang X, Lee MY,

Karnitz LM. Genomically incorporated 5-fluorouracil that escapes UNG-initiated

base excision repair blocks DNA replication and activates homologous

recombination. Molecular pharmacology. 2016; 89(1):53-62.

47. Yan Y, Han X, Qing Y, Condie AG, Gorityala S, Yang S, Xu Y, Zhang Y, Gerson

SL. Inhibition of uracil DNA glycosylase sensitizes cancer cells to 5-

fluorodeoxyuridine through replication fork collapse-induced DNA damage.

Oncotarget. 2016; 7(37):59299-313.

48. An Q, Robins P, Lindahl T, Barnes DE. 5-Fluorouracil incorporated into DNA is

excised by the Smug1 DNA glycosylase to reduce drug cytotoxicity. Cancer

research. 2007; 67(3):940-5.

49. Kunz C, Focke F, Saito Y, Schuermann D, Lettieri T, Selfridge J, Schär P. Base

excision by thymine DNA glycosylase mediates DNA-directed cytotoxicity of 5-

fluorouracil. PLoS biology. 2009; 7(4):e1000091.

50. Cortellino S, Turner D, Masciullo V, Schepis F, Albino D, Daniel R, Skalka AM,

Meropol NJ, Alberti C, Larue L, Bellacosa A. The base excision repair enzyme

MED1 mediates DNA damage response to antitumor drugs and is associated with

mismatch repair system integrity. Proceedings of the National Academy of

Sciences. 2003; 100(25):15071-6.

51. Demple B, Sung JS. Molecular and biological roles of Ape1 protein in mammalian

base excision repair. DNA repair. 2005; 4(12):1442-9.

Page | 134 52. McNeill DR, Lam W, DeWeese TL, Cheng YC, Wilson DM. Impairment of APE1

function enhances cellular sensitivity to clinically relevant alkylators and

antimetabolites. Molecular Cancer Research. 2009; 7(6):897-906.

53. Geng L, Huehls AM, Wagner JM, Huntoon CJ, Karnitz LM. Checkpoint signaling,

base excision repair, and PARP promote survival of colon cancer cells treated with

5-fluorodeoxyuridine but not 5-fluorouracil. PLoS One. 2011; 6(12):e28862.

54. Li L, Berger SH, Wyatt MD. Involvement of base excision repair in response to

therapy targeted at thymidylate synthase. Molecular cancer therapeutics. 2004;

3(6):747-53.

55. Li L, Connor EE, Berger SH, Wyatt MD. Determination of apoptosis, uracil

incorporation, DNA strand breaks, and sister chromatid exchanges under

conditions of thymidylate deprivation in a model of BER deficiency. Biochemical

pharmacology. 2005; 70(10):1458-68.

56. Haveman J, Kreder NC, Rodermond HM, Van Bree C, Franken NA, Stalpers LJ,

Zdzienicka MZ, Peters GJ. Cellular response of X-ray sensitive hamster mutant cell

lines to gemcitabine, cisplatin and 5-fluorouracil. Oncology reports. 2004; 12:187-

92.

57. Huehls AM, Wagner JM, Huntoon CJ, Geng L, Erlichman C, Patel AG, Kaufmann

SH, Karnitz LM. Poly (ADP-Ribose) polymerase inhibition synergizes with 5-

fluorodeoxyuridine but not 5-fluorouracil in ovarian cancer cells. Cancer research.

2011; 71(14):4944-54.

Page | 135 58. He L, Zhang Y, Sun H, Jiang F, Yang H, Wu H, Zhou T, Hu S, Kathera CS, Wang

X, Chen H. Targeting DNA Flap Endonuclease 1 to Impede Breast Cancer

Progression. EBioMedicine. 2016; 14:32-43.

59. Sun H, He L, Wu H, Pan F, Wu X, Zhao J, Hu Z, Sekhar C, Li H, Zheng L, Chen

H. The FEN1 L209P mutation interferes with long-patch base excision repair and

induces cellular transformation. Oncogene. 2017; 36(2):194-207.

60. Almeida KH, Sobol RW. A unified view of base excision repair: lesion-dependent

protein complexes regulated by post-translational modification. DNA repair. 2007;

6(6):695-711.

61. Copur S, Aiba K, Drake JC, Allegra CJ, Chu E. Thymidylate synthase gene

amplification in human colon cancer cell lines resistant to 5-fluorouracil.

Biochemical pharmacology. 1995; 49(10):1419-26.

62. Peters GJ, Backus HH, Freemantle S, Van Triest B, Codacci-Pisanelli G, Van der

Wilt CL, Smid K, Lunec J, Calvert AH, Marsh S, McLeod HL. Induction of

thymidylate synthase as a 5-fluorouracil resistance mechanism. Biochimica et

Biophysica Acta (BBA)-Molecular Basis of Disease. 2002; 1587(2):194-205.

63. Zhao R, Goldman ID. Resistance to antifolates. Oncogene. 2003;22(47):7431-57.

64. Horie N, Aiba H, Oguro K, Hojo H, Takeishi K. Functional analysis and DNA

polymorphism of the tandemly repeated sequences in the 5'-terminal regulatory

region of the human gene for thymidylate synthase. Cell structure and function.

1995; 20(3):191-7.

Page | 136 65. Marsh S, McKay JA, Cassidy J, McLeod HL. Polymorphism in the thymidylate

synthase promoter enhancer region in colorectal cancer. International journal of

oncology. 2001; 19(2):383-6.

66. Pullarkat ST, Stoehlmacher J, Ghaderi V, Xiong YP, Ingles SA, Sherrod A, Warren

R, Tsao-Wei D, Groshen S, Lenz HJ. Thymidylate synthase gene polymorphism

determines response and toxicity of 5-FU chemotherapy. The pharmacogenomics

journal. 2001; 1(1):65-70.

67. Kawakami K, Omura K, Kanehira E, Watanabe Y. Polymorphic tandem repeats in

the thymidylate synthase gene is associated with its protein expression in human

gastrointestinal cancers. Anticancer research. 1998; 19(4B):3249-52.

68. Chu E, Koeller DM, Johnston PG, Zinn S, Allegra CJ. Regulation of thymidylate

synthase in human colon cancer cells treated with 5-fluorouracil and interferon-

gamma. Molecular pharmacology. 1993; 43(4):527-33.

69. Swain SM, Lippman ME, Egan EF, Drake JC, Steinberg SM, Allegra CJ.

Fluorouracil and high-dose leucovorin in previously treated patients with metastatic

breast cancer. Journal of Clinical Oncology. 1989; 7(7):890-9.

70. Chu E, Koeller DM, Casey JL, Drake JC, Chabner BA, Elwood PC, Zinn S, Allegra

CJ. Autoregulation of human thymidylate synthase messenger RNA translation by

thymidylate synthase. Proceedings of the National Academy of Sciences. 1991;

88(20):8977-81.

71. Chu E, Voeller DM, Jones KL, Takechi T, Maley GF, Maley F, Segal S, Allegra

CJ. Identification of a thymidylate synthase ribonucleoprotein complex in human

colon cancer cells. Molecular and cellular biology. 1994; 14(1):207-13.

Page | 137 72. Liu J, Schmitz JC, Lin X, Tai N, Yan W, Farrell M, Bailly M, Chen TM, Chu E.

Thymidylate synthase as a translational regulator of cellular gene expression.

Biochimica et Biophysica Acta (BBA)-Molecular Basis of Disease. 2002;

1587(2):174-82.

73. Kitchens ME, Forsthoefel AM, Barbour KW, Spencer HT, Berger FG. Mechanisms

of acquired resistance to thymidylate synthase inhibitors: the role of enzyme

stability. Molecular pharmacology. 1999; 56(5):1063-70.

74. Kitchens ME, Forsthoefel AM, Rafique Z, Spencer HT, Berger FG. Ligand-

mediated Induction of Thymidylate Synthase Occurs by Enzyme Stabilization

IMPLICATIONS FOR AUTOREGULATION OF TRANSLATION. Journal of

Biological Chemistry. 1999; 274(18):12544-7.

75. Hughey CT, Barbour KW, Berger FG, Berger SH. Functional effects of a naturally

occurring amino acid substitution in human thymidylate synthase. Molecular

pharmacology. 1993; 44(2):316-23.

76. Reilly RT, Forsthoefel AM, Berger FG. Functional effects of amino acid

substitutions at residue 33 of human thymidylate synthase. Archives of

biochemistry and biophysics. 1997; 342(2):338-43.

77. Gonen N, Assaraf YG. Antifolates in cancer therapy: structure, activity and

mechanisms of drug resistance. Drug Resistance Updates. 2012; 15(4):183-210.

78. Lai D, Visser-Grieve S, Yang X. Tumour suppressor genes in chemotherapeutic

drug response. Bioscience reports. 2012; 32(4):361-74.

79. Bunz F, Hwang PM, Torrance C, Waldman T, Zhang Y, Dillehay L, Williams J,

Lengauer C, Kinzler KW, Vogelstein B. Disruption of p53 in human cancer cells

Page | 138 alters the responses to therapeutic agents. The Journal of clinical investigation.

1999; 104(3):263-9.

80. Petak I, Tillman DM, Houghton JA. p53 dependence of Fas induction and acute

apoptosis in response to 5-fluorouracil-leucovorin in human colon carcinoma cell

lines. Clinical cancer research. 2000; 6(11):4432-41.

81. Stravopodis DJ, Karkoulis PK, Konstantakou EG, Melachroinou S, Thanasopoulou

A, Aravantinos G, Margaritis LH, Anastasiadou E, Voutsinas GE. Thymidylate

synthase inhibition induces p53-dependent and p53-independent apoptotic

responses in human urinary bladder cancer cells. Journal of cancer research and

clinical oncology. 2011; 137(2):359-74.

82. Eisold S, Linnebacher M, Ryschich E, Antolovic D, Hinz U, Klar E, Schmidt J.

The effect of adenovirus expressing wild-type p53 on 5-fluorouracil

chemosensitivity is related to p53 status in pancreatic cancer cell lines. World

journal of gastroenterology. 2004; 10(24):3583-9.

83. Osaki M, Tatebe S, Goto A, Hayashi H, Oshimura M, Ito H. 5-Fluorouracil (5-FU)

induced apoptosis in gastric cancer cell lines: role of the p53 gene. Apoptosis. 1997;

2(2):221-6.

84. Liang JT, Huang KC, Cheng YM, Hsu HC, Cheng AL, Hsu CH, Yeh KH, Wang

SM, Chang KJ. P53 overexpression predicts poor chemosensitivity to high‐dose 5‐

fluorouracil plus leucovorin chemotherapy for stage IV colorectal cancers after

palliative bowel resection. International journal of cancer. 2002; 97(4):451-7.

85. Russo A, Bazan V, Iacopetta B, Kerr D, Soussi T, Gebbia N. The TP53 colorectal

cancer international collaborative study on the prognostic and predictive

Page | 139 significance of p53 mutation: influence of tumor site, type of mutation, and

adjuvant treatment. Journal of clinical oncology. 2005; 23(30):7518-28.

86. Kandioler D, Mittlböck M, Kappel S, Puhalla H, Herbst F, Langner C, Wolf B,

Tschmelitsch J, Schippinger W, Steger G, Hofbauer F. TP53 mutational status and

prediction of benefit from adjuvant 5-fluorouracil in stage III colon cancer patients.

EBioMedicine. 2015 Aug 31;2(8):825-30.

87. Owen-Schaub LB, Zhang W, Cusack JC, Angelo LS, Santee SM, Fujiwara T, Roth

JA, Deisseroth AB, Zhang WW, Kruzel E. Wild-type human p53 and a

temperature-sensitive mutant induce Fas/APO-1 expression. Molecular and

Cellular Biology. 1995; 15(6):3032-40.

88. Nakano K, Vousden KH. PUMA, a novel proapoptotic gene, is induced by p53.

Molecular cell. 2001; 7(3):683-94.

89. Petak I, Tillman DM, Houghton JA. p53 dependence of Fas induction and acute

apoptosis in response to 5-fluorouracil-leucovorin in human colon carcinoma cell

lines. Clinical cancer research. 2000; 6(11):4432-41.

90. Longley DB, Allen WL, McDermott U, Wilson TR, Latif T, Boyer J, Lynch M,

Johnston PG. The roles of thymidylate synthase and p53 in regulating Fas-mediated

apoptosis in response to antimetabolites. Clinical cancer research. 2004;

10(10):3562-71.

91. Borralho PM, da Silva IB, Aranha MM, Albuquerque C, Leitão CN, Steer CJ,

Rodrigues CM. Inhibition of Fas expression by RNAi modulates 5-fluorouracil-

induced apoptosis in HCT116 cells expressing wild-type p53. Biochimica et

Biophysica Acta (BBA)-Molecular Basis of Disease. 2007; 1772(1):40-7.

Page | 140 92. Yu J, Zhang L, Hwang PM, Kinzler KW, Vogelstein B. PUMA induces the rapid

apoptosis of colorectal cancer cells. Molecular cell. 2001; 7(3):673-82.

93. Can G, Akpinar B, Baran Y, Zhivotovsky B, Olsson M. 5-Fluorouracil signaling

through a calcium–calmodulin-dependent pathway is required for p53 activation

and apoptosis in colon carcinoma cells. Oncogene. 2013; 32(38):4529-38.

94. Sun XX, Dai MS, Lu H. 5-fluorouracil activation of p53 involves an MDM2-

ribosomal protein interaction. Journal of Biological Chemistry. 2007;

282(11):8052-9.

95. Parsels LA, Parsels JD, Wagner LM, Loney TL, Radany EH, Maybaum J.

Mechanism and pharmacological specificity of dUTPase-mediated protection from

DNA damage and cytotoxicity in human tumor cells. Cancer chemotherapy and

pharmacology. 1998; 42(5):357-62.

96. Ladner RD. The role of dUTPase and uracil-DNA repair in cancer chemotherapy.

Current Protein and Peptide Science. 2001; 2(4):361-70.

97. Wilson PM, LaBonte MJ, Lenz HJ, Mack PC, Ladner RD. Inhibition of dUTPase

Induces Synthetic Lethality with Thymidylate Synthase–Targeted Therapies in

Non–Small Cell Lung Cancer. Molecular cancer therapeutics. 2012; 11(3):616-28.

98. Wilson PM, Fazzone W, LaBonte MJ, Deng J, Neamati N, Ladner RD. Novel

opportunities for thymidylate metabolism as a therapeutic target. Molecular cancer

therapeutics. 2008; 7(9):3029-37.

99. Ladner RD, Lynch FJ, Groshen S, Xiong YP, Sherrod A, Caradonna SJ,

Stoehlmacher J, Lenz HJ. dUTP nucleotidohydrolase isoform expression in normal

Page | 141 and neoplastic tissues: association with survival and response to 5-fluorouracil in

colorectal cancer. Cancer research. 2000; 60(13):3493-503.

100. Kawahara A, Akagi Y, Hattori S, Mizobe T, Shirouzu K, Ono M, Yanagawa T,

Kuwano M, Kage M. Higher expression of deoxyuridine triphosphatase (dUTPase)

may predict the metastasis potential of colorectal cancer. Journal of clinical

pathology. 2009; 62(4):364-9.

101. Koehler SE, Ladner RD. Small interfering RNA-mediated suppression of

dUTPase sensitizes cancer cell lines to thymidylate synthase inhibition. Molecular

pharmacology. 2004; 66(3):620-6.

102. Webley SD, Hardcastle A, Ladner RD, Jackman AL, Aherne GW. Deoxyuridine

triphosphatase (dUTPase) expression and sensitivity to the thymidylate synthase

(TS) inhibitorD9331. British journal of cancer. 2000; 83(6):792.

103. Canman CE, Lawrence TS, Shewach DS, Tang HY, Maybaum J. Resistance to

fluorodeoxyuridine-induced DNA damage and cytotoxicity correlates with an

elevation of deoxyuridine triphosphatase activity and failure to accumulate

deoxyuridine triphosphate. Cancer research. 1993; 53(21):5219-24.

104. Miyahara S, Miyakoshi H, Yokogawa T, Chong KT, Taguchi J, Muto T, Endoh

K, Yano W, Wakasa T, Ueno H, Takao Y. Discovery of a novel class of potent

human deoxyuridine triphosphatase inhibitors remarkably enhancing the antitumor

activity of thymidylate synthase inhibitors. Journal of medicinal chemistry. 2012;

55(7):2970-80.

105. Saito K, Nagashima H, Noguchi K, Yoshisue K, Yokogawa T, Matsushima E,

Tahara T, Takagi S. First-in-human, phase I dose-escalation study of single and

Page | 142 multiple doses of a first-in-class enhancer of fluoropyrimidines, a dUTPase

inhibitor (TAS-114) in healthy male volunteers. Cancer chemotherapy and

pharmacology. 2014; 73(3):577-83.

106. US National Library of Medicine. ClinicalTrials.gov [online],

http://clinicaltrials.gov/ct2/show/NCT01610479 (2014).

107. US National Library of Medicine. ClinicalTrials.gov [online],

http://clinicaltrials.gov/ct2/show/NCT02025803 (2013).

108. Jiricny J. The multifaceted mismatch-repair system. Nature reviews Molecular

cell biology. 2006; 7(5):335-46.

109. Li GM. Mechanisms and functions of DNA mismatch repair. Cell research. 2008;

18(1):85-98.

110. Stojic L, Brun R, Jiricny J. Mismatch repair and DNA damage signalling. DNA

repair. 2004; 3(8):1091-101.

111. Meyers M, Wagner MW, Mazurek A, Schmutte C, Fishel R, Boothman DA. DNA

mismatch repair-dependent response to fluoropyrimidine-generated damage.

Journal of Biological Chemistry. 2005; 280(7):5516-26.

112. Fischer F, Baerenfaller K, Jiricny J. 5-Fluorouracil is efficiently removed from

DNA by the base excision and mismatch repair systems. Gastroenterology. 2007;

133(6):1858-68.

113. Carethers JM, Chauhan DP, Fink D, Nebel S, Bresalier RS, Howell SB, Boland

CR. Mismatch repair proficiency and in vitro response to 5-fluorouracil.

Gastroenterology. 1999; 117(1):123-31.

Page | 143 114. Meyers M, Wagner MW, Hwang HS, Kinsella TJ, Boothman DA. Role of the

hMLH1 DNA mismatch repair protein in fluoropyrimidine-mediated cell death and

cell cycle responses. Cancer research. 2001; 61(13):5193-201.

115. Jover R, Zapater P, Castells A, Llor X, Andreu M, Cubiella J, Piñol V, Xicola RM,

Bujanda L, Reñé JM, Clofent J. Mismatch repair status in the prediction of benefit

from adjuvant fluorouracil chemotherapy in colorectal cancer. Gut. 2006 Jun

1;55(6):848-55.

116. Sargent DJ, Marsoni S, Monges G, Thibodeau SN, Labianca R, Hamilton SR,

French AJ, Kabat B, Foster NR, Torri V, Ribic C. Defective mismatch repair as a

predictive marker for lack of efficacy of fluorouracil-based adjuvant therapy in

colon cancer. Journal of Clinical Oncology. 2010; 28(20):3219-26.

117. Popat S, Hubner R, Houlston RS. Systematic review of microsatellite instability

and colorectal cancer prognosis. Journal of Clinical Oncology. 2005; 23(3):609-18.

118. Lamberti C, Lundin S, Bogdanow M, Pagenstecher C, Friedrichs N, Büttner R,

Sauerbruch T. Microsatellite instability did not predict individual survival of

unselected patients with colorectal cancer. International journal of colorectal

disease. 2007; 22(2):145-52.

119. Nagaraju G, Scully R. Minding the gap: the underground functions of BRCA1 and

BRCA2 at stalled replication forks. DNA repair. 2007; 6(7):1018-31.

120. Thorslund T, West SC. BRCA2: a universal recombinase regulator. Oncogene.

2007; 26(56):7720-30.

121. Li L, Connor EE, Berger SH, Wyatt MD. Determination of apoptosis, uracil

incorporation, DNA strand breaks, and sister chromatid exchanges under

Page | 144 conditions of thymidylate deprivation in a model of BER deficiency. Biochemical

pharmacology. 2005; 70(10):1458-68.

122. Yang Z, Waldman AS, Wyatt MD. DNA damage and homologous recombination

signaling induced by thymidylate deprivation. Biochemical pharmacology. 2008;

76(8):987-96.

123. Yang Z, Waldman AS, Wyatt MD. Expression and regulation of RAD51 mediate

cellular responses to chemotherapeutics. Biochemical pharmacology. 2012;

83(6):741-6.

124. Brady CA, Attardi LD. p53 at a glance. J Cell Sci. 2010; 123(15):2527-32.

125. Riley T, Sontag E, Chen P, Levine A. Transcriptional control of human p53-

regulated genes. Nature reviews Molecular cell biology. 2008 May 1;9(5):402-12.

126. Hermeking H. MicroRNAs in the p53 network: micromanagement of tumour

suppression. Nature Reviews Cancer. 2012; 12(9):613-26.

127. Green DR, Kroemer G. Cytoplasmic functions of the tumour suppressor p53.

Nature. 2009; 458(7242):1127-30.

128. Kruse JP, Gu W. Modes of p53 regulation. Cell. 2009; 137(4):609-22.

129. Zilfou JT, Lowe SW. Tumor suppressive functions of p53. Cold Spring Harbor

perspectives in biology. 2009; 1(5):a001883.

130. Kruiswijk F, Labuschagne CF, Vousden KH. p53 in survival, death and metabolic

health: a lifeguard with a licence to kill. Nature reviews Molecular cell biology.

2015; 16(7):393-405.

131. Parant J, Chavez-Reyes A, Little NA, Yan W, Reinke V, Jochemsen AG, Lozano

G. Rescue of embryonic lethality in Mdm4-null mice by loss of Trp53 suggests a

Page | 145 nonoverlapping pathway with MDM2 to regulate p53. Nature genetics. 2001;

29(1):92-5.

132. Meek DW, Anderson CW. Posttranslational modification of p53: cooperative

integrators of function. Cold Spring Harbor perspectives in biology. 2009;

1(6):a000950.

133. Elmore S. Apoptosis: a review of programmed cell death. Toxicologic pathology.

2007; 35(4):495-516.

134. Mirza A, McGuirk M, Hockenberry TN, Wu Q, Ashar H, Black S, Wen SF, Wang

L, Kirschmeier P, Bishop WR, Nielsen LL. Human survivin is negatively regulated

by wild-type p53 and participates in p53-dependent apoptotic pathway. Oncogene.

2002; 21(17):2613.

135. Vaseva AV, Moll UM. The mitochondrial p53 pathway. Biochimica et Biophysica

Acta (BBA)-Bioenergetics. 2009; 1787(5):414-20.

136. Mihara M, Erster S, Zaika A, Petrenko O, Chittenden T, Pancoska P, Moll UM.

p53 has a direct apoptogenic role at the mitochondria. Molecular cell. 2003;

11(3):577-90.

137. Leu JJ, Dumont P, Hafey M, Murphy ME, George DL. Mitochondrial p53

activates Bak and causes disruption of a Bak–Mcl1 complex. Nature cell biology.

2004; 6(5):443-50.

138. Upton B, Chu Q, Li BD. Li-Fraumeni syndrome: the genetics and treatment

considerations for the sarcoma and associated neoplasms. Surgical oncology clinics

of North America. 2009; 18(1):145-56.

Page | 146 139. Muller PA, Vousden KH. p53 mutations in cancer. Nature cell biology. 2013;

15(1):2-8.

140. Olivier M, Hollstein M, Hainaut P. TP53 mutations in human cancers: origins,

consequences, and clinical use. Cold Spring Harbor perspectives in biology. 2010;

2(1):a001008.

141. Lang GA, Iwakuma T, Suh YA, Liu G, Rao VA, Parant JM, Valentin-Vega YA,

Terzian T, Caldwell LC, Strong LC, El-Naggar AK. Gain of function of a p53 hot

spot mutation in a mouse model of Li-Fraumeni syndrome. Cell. 2004; 119(6):861-

72.

142. Olive KP, Tuveson DA, Ruhe ZC, Yin B, Willis NA, Bronson RT, Crowley D,

Jacks T. Mutant p53 gain of function in two mouse models of Li-Fraumeni

syndrome. Cell. 2004; 119(6):847-60.

143. Oren M, Rotter V. Mutant p53 gain-of-function in cancer. Cold Spring Harbor

perspectives in biology. 2010; 2(2):a001107.

144. Freed-Pastor WA, Prives C. Mutant p53: one name, many proteins. Genes &

development. 2012; 26(12):1268-86.

145. Brosh R, Rotter V. When mutants gain new powers: news from the mutant p53

field. Nature Reviews Cancer. 2009; 9(10):701-13.

146. Chin KV, Ueda K. Modulation of activity of the promoter of the human MDR1

gene by Ras and p53. Science. 1992; 255(5043):459.

147. Sampath J, Sun D, Kidd VJ, Grenet J, Gandhi A, Shapiro LH, Wang Q, Zambetti

GP, Schuetz JD. Mutant p53 cooperates with ETS and selectively up-regulates

Page | 147 human MDR1 not MRP1. Journal of Biological Chemistry. 2001; 276(42):39359-

67.

148. Tonigold M, Rossmann A, Meinold M, Bette M, Märken M, Henkenius K, Bretz

AC, Giel G, Cai C, Rodepeter FR, Beneš V. A cisplatin-resistant head and neck

cancer cell line with cytoplasmic p53mut exhibits ATP-binding cassette transporter

upregulation and high glutathione levels. Journal of cancer research and clinical

oncology. 2014; 140(10):1689-704.

149. Xu J, Wang J, Hu Y, Qian J, Xu B, Chen H, Zou W, Fang JY. Unequal prognostic

potentials of p53 gain-of-function mutations in human cancers associate with drug-

metabolizing activity. Cell death & disease. 2014; 5(3):e1108.

150. Scian MJ, Stagliano KE, Anderson MA, Hassan S, Bowman M, Miles MF, Deb

SP, Deb S. Tumor-derived p53 mutants induce NF-κB2 gene expression. Molecular

and cellular biology. 2005; 25(22):10097-110.

151. Vaughan CA, Singh S, Windle B, Sankala HM, Graves PR, Yeudall WA, Deb SP,

Deb S. p53 mutants induce transcription of NF-κB2 in H1299 cells through CBP

and STAT binding on the NF-κB2 promoter and gain of function activity. Archives

of biochemistry and biophysics. 2012; 518(1):79-88.

152. Weisz L, Zalcenstein A, Stambolsky P, Cohen Y, Goldfinger N, Oren M, Rotter

V. Transactivation of the EGR1 gene contributes to mutant p53 gain of function.

Cancer Research. 2004; 64(22):8318-27.

153. Lavra L, Ulivieri A, Rinaldo C, Dominici R, Volante M, Luciani E, Bartolazzi A,

Frasca F, Soddu S, Sciacchitano S. Gal‐3 is stimulated by gain‐of‐function p53

Page | 148 mutations and modulates chemoresistance in anaplastic thyroid carcinomas. The

Journal of pathology. 2009; 218(1):66-75.

154. Li J, Yang L, Gaur S, Zhang K, Wu X, Yuan YC, Li H, Hu S, Weng Y, Yen Y.

Mutants TP53 p. R273H and p. R273C but not p. R273G enhance cancer cell

malignancy. Human mutation. 2014; 35(5):575-84.

155. Do PM, Varanasi L, Fan S, Li C, Kubacka I, Newman V, Chauhan K, Daniels SR,

Boccetta M, Garrett MR, Li R. Mutant p53 cooperates with ETS2 to promote

etoposide resistance. Genes & development. 2012; 26(8):830-45.

156. Irwin MS, Kondo K, Marin MC, Cheng LS, Hahn WC, Kaelin WG.

Chemosensitivity linked to p73 function. Cancer cell. 2003; 3(4):403-10.

157. Sankala H, Vaughan C, Wang J, Deb S, Graves PR. Upregulation of the

mitochondrial transport protein, Tim50, by mutant p53 contributes to cell growth

and chemoresistance. Archives of biochemistry and biophysics. 2011; 512(1):52-

60.

158. Pugacheva EN, Ivanov AV, Kravchenko JE, Kopnin BP, Levine AJ, Chumakov

PM. Novel gain of function activity of p53 mutants: activation of the dUTPase gene

expression leading to resistance to 5-fluorouracil. Oncogene. 2002; 21(30):4595.

159. Malet-Martino M, Martino R. Clinical studies of three oral prodrugs of 5-

fluorouracil (capecitabine, UFT, S-1): a review. The oncologist. 2002; 7(4):288-

323.

160. Power DG, Kemeny NE. The role of floxuridine in metastatic liver disease.

Molecular cancer therapeutics. 2009; 8(5):1015-25.

Page | 149 161. Wolmark N, Rockette H, Fisher B, Wickerham DL, Redmond C, Fisher ER, Jones

J, Mamounas EP, Ore L, Petrelli NJ. The benefit of leucovorin-modulated

fluorouracil as postoperative adjuvant therapy for primary colon cancer: results

from National Surgical Adjuvant Breast and Bowel Project protocol C-03. Journal

of Clinical Oncology. 1993; 11(10):1879-87.

162. Zaniboni A. Adjuvant chemotherapy in colorectal cancer with high-dose

leucovorin and fluorouracil: impact on disease-free survival and overall survival.

Journal of clinical oncology. 1997; 15(6):2432-41.

163. André T, Boni C, Mounedji-Boudiaf L, Navarro M, Tabernero J, Hickish T,

Topham C, Zaninelli M, Clingan P, Bridgewater J, Tabah-Fisch I. Oxaliplatin,

fluorouracil, and leucovorin as adjuvant treatment for colon cancer. New England

Journal of Medicine. 2004; 350(23):2343-51.

164. Parker WB, Cheng YC. Metabolism and mechanism of action of 5-fluorouracil.

Pharmacology & therapeutics. 1990; 48(3):381-95.

165. Willmore E, Durkacz BW. Cytotoxic mechanisms of 5-fluoropyrimidines:

relationships with poly (ADP-ribose) polymerase activity, DNA strand breakage

and incorporation into nucleic acids. Biochemical pharmacology. 1993; 46(2):205-

11.

166. Van Laar JA, Rustum YM, Ackland SP, Van Groeningen CJ, Peters GJ.

Comparison of 5-fluoro-2′-deoxyuridine with 5-fluorouracil and their role in the

treatment of colorectal cancer. European Journal of Cancer. 1998; 34(3):296-306.

Page | 150 167. Gadsden MH, McIntosh EM, Game JC, Wilson PJ, Haynes RH. dUTP

pyrophosphatase is an essential enzyme in Saccharomyces cerevisiae. The EMBO

journal. 1993; 12(11):4425.

168. Meyers M, Hwang A, Wagner MW, Bruening AJ, Veigl ML, Sedwick WD,

Boothman DA. A role for DNA mismatch repair in sensing and responding to

fluoropyrimidine damage. Oncogene. 2003; 22(47):7376-88.

169. Fischer JA, Muller-Weeks S, Caradonna SJ. Fluorodeoxyuridine modulates

cellular expression of the DNA base excision repair enzyme uracil-DNA

glycosylase. Cancer research. 2006; 66(17):8829-37.

170. Bulgar AD, Weeks LD, Miao Y, Yang S, Xu Y, Guo C, Markowitz S, Oleinick

N, Gerson SL, Liu L. Removal of uracil by uracil DNA glycosylase limits

pemetrexed cytotoxicity: overriding the limit with methoxyamine to inhibit base

excision repair. Cell death & disease. 2012; 3(1):e252.

171. Dusseau C, Murray GI, Keenan RA, O'Kelly T, Krokan HE, McLeod HL.

Analysis of uracil DNA glycosylase in human colorectal cancer. International

journal of oncology. 2001; 18(2):393-400.

172. Condie AG, Yan Y, Gerson SL, Wang Y. A fluorescent probe to measure DNA

damage and repair. PloS one. 2015; 10(8):e0131330.

173. Han X, Pozo FM, Wisotsky JN, Wang B, Jacobberger JW, Zhang Y.

Phosphorylation of minichromosome maintenance 3 (MCM3) by checkpoint kinase

1 (Chk1) negatively regulates DNA replication and checkpoint activation. Journal

of Biological Chemistry. 2015; 290(19):12370-8.

Page | 151 174. Yan L, Bulgar A, Miao Y, Mahajan V, Donze JR, Gerson SL, Liu L. Combined

treatment with temozolomide and methoxyamine: blocking apurininc/pyrimidinic

site repair coupled with targeting topoisomerase IIα. Clinical Cancer Research.

2007; 13(5):1532-9.

175. Longley DB, Ferguson PR, Boyer J, Latif T, Lynch M, Maxwell P, Harkin DP,

Johnston PG. Characterization of a thymidylate synthase (TS)-inducible cell line.

Clinical Cancer Research. 2001; 7(11):3533-9.

176. Sakoff JA, Ackland SP. Thymidylate synthase inhibition induces S-phase arrest,

biphasic mitochondrial alterations and caspase-dependent apoptosis in leukaemia

cells. Cancer chemotherapy and pharmacology. 2000; 46(6):477-87.

177. Chen KC, Yang TY, Wu CC, Cheng CC, Hsu SL, Hung HW, Chen JW, Chang

GC. Pemetrexed induces S-phase arrest and apoptosis via a deregulated activation

of Akt signaling pathway. PloS one. 2014; 9(5):e97888.

178. Zeman MK, Cimprich KA. Causes and consequences of replication stress. Nature

cell biology. 2014; 16(1):2-9.

179. Cortez D. Preventing replication fork collapse to maintain genome integrity. DNA

repair. 2015; 32:149-57.

180. Enari M, Sakahira H, Yokoyama H, Okawa K, Iwamatsu A, Nagata S. A caspase-

activated DNase that degrades DNA during apoptosis, and its inhibitor ICAD.

Nature. 1998; 391(6662):43-50.

181. Rogakou EP, Nieves-Neira W, Boon C, Pommier Y, Bonner WM. Initiation of

DNA fragmentation during apoptosis induces phosphorylation of H2AX histone at

serine 139. Journal of Biological Chemistry. 2000; 275(13):9390-5.

Page | 152 182. Caserta TM, Smith AN, Gultice AD, Reedy MA, Brown TL. Q-VD-OPh, a broad

spectrum caspase inhibitor with potent antiapoptotic properties. Apoptosis. 2003;

8(4):345-52.

183. Feraud O, Debili N, Penninger JM, Kroemer G. Broad-spectrum caspase

inhibitors: from myth to reality?. Cell death and differentiation. 2007;14:387-91.

184. Kuželová K, Grebeňová D, Brodská B. Dose-dependent effects of the caspase

inhibitor Q-VD-OPh on different apoptosis-related processes. J Cell Biochem.

2011; 112(11):3334-42.

185. Morgan MT, Bennett MT, Drohat AC. Excision of 5-Halogenated Uracils by

Human Thymine DNA Glycosylase ROBUST ACTIVITY FOR DNA

CONTEXTS OTHER THAN CpG. Journal of Biological Chemistry. 2007;

282(38):27578-86.

186. Fitzgerald ME, Drohat AC. Coordinating the initial steps of base excision repair

apurinic/apyrimidinic endonuclease 1 actively stimulates thymine DNA

glycosylase by disrupting the product complex. Journal of Biological Chemistry.

2008; 283(47):32680-90.

187. Bennett MT, Rodgers MT, Hebert AS, Ruslander LE, Eisele L, Drohat AC.

Specificity of human thymine DNA glycosylase depends on N-glycosidic bond

stability. Journal of the American Chemical Society. 2006; 128(38):12510.

188. Hardeland U, Kunz C, Focke F, Szadkowski M, Schär P. Cell cycle regulation as

a mechanism for functional separation of the apparently redundant uracil DNA

glycosylases TDG and UNG2. Nucleic acids research. 2007; 35(11):3859-67.

Page | 153 189. Haug T, Skorpen F, Aas PA, Malm V, Skjelbred C, Krokan HE. Regulation of

expression of nuclear and mitochondrial forms of human uracil-DNA glycosylase.

Nucleic acids research. 1998; 26(6):1449-57.

190. Rytelewski M, Ferguson PJ, Vareki SM, Figueredo R, Vincent M, Koropatnick J.

Inhibition of BRCA2 and thymidylate synthase creates multidrug sensitive tumor

cells via the induction of combined “complementary lethality”. Molecular

Therapy—Nucleic Acids. 2013; 2(3):e78.

191. Houghton JA, Harwood FG, Tillman DM. Thymineless death in colon carcinoma

cells is mediated via fas signaling. Proceedings of the National Academy of

Sciences. 1997; 94(15):8144-9.

192. Nita ME, Nagawa H, Tominaga O, Tsuno N, Fujii S, Sasaki S, Fu CG, Takenoue

T, Tsuruo T, Muto T. 5-Fluorouracil induces apoptosis in human colon cancer cell

lines with modulation of Bcl-2 family proteins. British journal of cancer. 1998;

78(8):986.

193. Backus HH, Wouters D, Ferreira CG, Van Houten VM, Brakenhoff RH, Pinedo

HM, Peters GJ. Thymidylate synthase inhibition triggers apoptosis via caspases-8

and-9 in both wild-type and mutant p53 colon cancer cell lines. European journal

of cancer. 2003; 39(9):1310-7.

194. Carreras CW, Santi DV. The catalytic mechanism and structure of thymidylate

synthase. Annual review of biochemistry. 1995; 64(1):721-62.

195. Van Triest B, Pinedo HM, Giaccone G, Peters GJ. Downstream molecular

determinants of response to 5-fluorouracil and antifolate thymidylate synthase

inhibitors. Annals of Oncology. 2000; 11(4):385-91.

Page | 154 196. Berger SH, Pittman DL, Wyatt MD. Uracil in DNA: consequences for

carcinogenesis and chemotherapy. Biochemical pharmacology. 2008; 76(6):697-

706.

197. Tchelebi L, Ashamalla H, Graves PR. Mutant p53 and the response to

chemotherapy and radiation. InMutant p53 and MDM2 in Cancer 2014 (pp. 133-

159). Springer Netherlands.

198. Cipriano R, Patton JT, Mayo LD, Jackson MW. Inactivation of p53 signaling by

p73 or PTEN ablation results in a transformed phenotype that remains susceptible

to Nutlin-3 mediated apoptosis. Cell Cycle. 2010; 9(7):1373-9.

199. Abbas T, Dutta A. p21 in cancer: intricate networks and multiple activities. Nature

Reviews Cancer. 2009; 9(6):400-14.

200. Akpinar B, Bracht EV, Reijnders D, Safarikova B, Jelinkova I, Grandien A,

Vaculova AH, Zhivotovsky B, Olsson M. 5-Fluorouracil-induced RNA stress

engages a TRAIL-DISC-dependent apoptosis axis facilitated by p53. Oncotarget.

2015; 6(41):43679.

201. Ju J, Pedersen-Lane J, Maley F, Chu E. Regulation of p53 expression by

thymidylate synthase. Proceedings of the National Academy of Sciences. 1999;

96(7):3769-74.

202. Longley DB, Boyer J, Allen WL, Latif T, Ferguson PR, Maxwell PJ, McDermott

U, Lynch M, Harkin DP, Johnston PG. The role of thymidylate synthase induction

in modulating p53-regulated gene expression in response to 5-fluorouracil and

antifolates. Cancer research. 2002; 62(9):2644-9.

Page | 155 203. Giovannetti E, Backus HH, Wouters D, Ferreira CG, Van Houten VM, Brakenhoff

RH, Poupon MF, Azzarello A, Pinedo HM, Peters GJ. Changes in the status of p53

affect drug sensitivity to thymidylate synthase (TS) inhibitors by altering TS levels.

British journal of cancer. 2007; 96(5):769-75.

204. Sjolund AB, Senejani AG, Sweasy JB. MBD4 and TDG: multifaceted DNA

glycosylases with ever expanding biological roles. Mutation

Research/Fundamental and Molecular Mechanisms of Mutagenesis. 2013; 743:12-

25.

205. Haushalter KA, Stukenberg PT, Kirschner MW, Verdine GL. Identification of a

new uracil-DNA glycosylase family by expression cloning using synthetic

inhibitors. Current biology. 1999; 9(4):174-85.

206. Cortázar D, Kunz C, Saito Y, Steinacher R, Schär P. The enigmatic thymine DNA

glycosylase. DNA repair. 2007; 6(4):489-504.

207. Hardeland U, Bentele M, Jiricny J, SchaÈr P. The versatile thymine DNA‐

glycosylase: a comparative characterization of the human, Drosophila and fission

yeast orthologs. Nucleic acids research. 2003; 31(9):2261-71.

208. Morgan MT, Maiti A, Fitzgerald ME, Drohat AC. Stoichiometry and affinity for

thymine DNA glycosylase binding to specific and nonspecific DNA. Nucleic acids

research. 2010: gkq1164.

209. Ashwell S, Zabludoff S. DNA damage detection and repair pathways—recent

advances with inhibitors of checkpoint kinases in cancer therapy. Clinical Cancer

Research. 2008; 14(13):4032-7.

Page | 156 210. Toledo LI, Murga M, Fernandez‐Capetillo O. Targeting ATR and Chk1 kinases

for cancer treatment: a new model for new (and old) drugs. Molecular oncology.

2011; 5(4):368-73.

211. Weber AM, Ryan AJ. ATM and ATR as therapeutic targets in cancer.

Pharmacology & therapeutics. 2015; 149:124-38.

212. Smith J, Mun Tho L, Xu N, Gillespie DA. The ATM-Chk2 and ATR-Chk1

pathways in DNA damage signaling and cancer. Advances in cancer research. 2010;

108(C):73-112.

213. Parsels LA, Parsels JD, Tai DC, Coughlin DJ, Maybaum J. 5-fluoro-2′-

deoxyuridine-induced cdc25A accumulation correlates with premature mitotic

entry and clonogenic death in human colon cancer cells. Cancer research. 2004;

64(18):6588-94.

214. Wilsker D, Bunz F. Loss of ataxia telangiectasia mutated–and rad3-related

function potentiates the effects of chemotherapeutic drugs on cancer cell survival.

Molecular cancer therapeutics. 2007; 6(4):1406-13.

215. Yang Z, Waldman AS, Wyatt MD. DNA damage and homologous recombination

signaling induced by thymidylate deprivation. Biochemical pharmacology. 2008;

76(8):987-96.

216. Xiao Z, Xue J, Sowin TJ, Zhang H. Differential roles of checkpoint kinase 1,

checkpoint kinase 2, and mitogen-activated protein kinase–activated protein kinase

2 in mediating DNA damage–induced cell cycle arrest: implications for cancer

therapy. Molecular cancer therapeutics. 2006; 5(8):1935-43.

Page | 157 217. Robinson HM, Jones R, Walker M, Zachos G, Brown R, Cassidy J, Gillespie DA.

Chk1-dependent slowing of S-phase progression protects DT40 B-lymphoma cells

against killing by the nucleoside analogue 5-fluorouracil. Oncogene. 2006;

25(39):5359-69.

218. Hsueh CT, Kelsen D, Schwartz GK. UCN-01 suppresses thymidylate synthase

gene expression and enhances 5-fluorouracil-induced apoptosis in a sequence-

dependent manner. Clinical cancer research. 1998; 4(9):2201-6.

219. Ma CX, Janetka JW, Piwnica-Worms H. Death by releasing the breaks: CHK1

inhibitors as cancer therapeutics. Trends in molecular medicine. 2011; 17(2):88-96.

220. Gurpinar E, Vousden KH. Hitting cancers’ weak spots: vulnerabilities imposed

by p53 mutation. Trends in cell biology. 2015; 25(8):486-95.

221. Chen Z, Xiao Z, Gu WZ, Xue J, Bui MH, Kovar P, Li G, Wang G, Tao ZF, Tong

Y, Lin NH. Selective Chk1 inhibitors differentially sensitize p53‐deficient cancer

cells to cancer therapeutics. International journal of cancer. 2006; 119(12):2784-94.

222. Reinhardt HC, Aslanian AS, Lees JA, Yaffe MB. p53-deficient cells rely on

ATM-and ATR-mediated checkpoint signaling through the p38MAPK/MK2

pathway for survival after DNA damage. Cancer cell. 2007; 11(2):175-89.

223. Reinhardt HC, Jiang H, Hemann MT, Yaffe MB. Exploiting synthetic lethal

interactions for targeted cancer therapy. Cell cycle. 2009; 8(19):3112-9.

Page | 158