The Pennsylvania State University

The Graduate School

College of Medicine

DIFFERENTIAL P53 SIGNALING IN RESPONSE TO 5-FU AND

ETOPOSIDE IN MODULATING TOXICITY VIA DPYD AND CHK2 IN

CANCER THERAPY

A Dissertation in

Molecular Medicine

by

Prashanth R Gokare

Ó 2017 Prashanth R Gokare

Submitted in Partial Fulfillment of the Requirements for the Degree, of

Doctor of Philosophy

December 2017

The dissertation of Prashanth R Gokare to be reviewed and approved* by the following:

Wafik S. El-Deiry, Professor, Department Hematology/Oncology William Wikoff Smith Chair in Cancer Research Deputy Cancer Center Director for Translational Research Co-Leader Molecular Therapeutics Fox Chase Cancer Center Dissertation Adviser Committee Co-Chair

Charles H. Lang, Distinguished Professor of Cellular and Molecular Physiology Director of the Molecular Medicine Graduate Program Committee Co-Chair

Rosalyn Irby, Associate Professor, Department of Medicine

Jin-Ming Yang, Professor of Pharmacology, Department of Medicine,

Nelson S. Yee, (Special Member), Assistant Professor, Department of Hematology Department of Medicine,

*Signatures are required on file for the Graduate School

ii ABSTRACT

Tp53 is a major transcription factor that controls a multitude of processes involved in the response to cellular stress. It plays a critical role in cell cycle arrest and/or apoptosis after

DNA damage. p53 downstream effects can impact on cellular metabolism and these are increasingly becoming unraveled. The studies here in highlights previously unappreciated signaling by Tp53 that is relevant to toxicity and efficacy of chemotherapeutic agents.

Nucleotide metabolism can influence malignant behavior and intrinsic resistance to cancer therapy. Alterations in pyrimidine metabolism are key to the mechanism of action of the chemotherapeutic antimetabolite 5-fluorouracil (5-FU). The initial study describes a novel role of p53 in controlling the rate-limiting enzyme in the pyrimidine catabolic pathway, dihydropyrimidine dehydrogenase (DPYD) and its effect on pharmacokinetics of and response to 5-FU. P53 binds to a p53 DNA-binding site (p53BS) downstream of the DPYD gene and reduces its expression at both the mRNA and protein level. The reduced expression of DPYD follows the inhibition of thymidylate synthase (TS) and is dependent on DNA-dependent protein kinase (DNA-PK) and Ataxia telangiectasia mutated (ATM) signaling. Overall this study highlights the regulation of DPYD and its implications on toxicity and efficacy of 5-FU.

(Chk2) is a serine/threonine kinase that transduces DNA damage response (DDR) signals from the kinases ATM and to some extent also Ataxia Telangiectasia and Rad3-Related

Protein (ATR). It plays a critical role in inducing cell death following radiation in a p53- dependent manner. However, the role of Chk2 in toxicity of chemotherapeutics is less

iii well studied with regard to the involvement of the ChK2-ATM-p53 pathway. Our experiments addressed the role of Chk2 in Dose Limiting toxicity (DLT) of Topisomearase

II (TOP2) inhibitors. We found Chk2 mediates toxicity from TOP2 inhibitors but not with other classes of chemotherapeutics, both in-vitro and in-vivo. Functional screens identified NSC105171 as a novel Chk2 inhibitor. NSC105171 protected from DLT following Etoposide treatment. This study has implications for a potentially effective strategy to preventing DLT.

Thus, we have identified a role for p53 in controlling nucleotide metabolism through repression of DPYD following DNA damage through the involvement of TS inhibition. The study highlights different responses of chemotherapeutic agents that signal through p53 activation. In the second study, we described targeting the ATM-Chk2 p53 pathway in the context of Topoisomerase 2 inhibition to preferentially limit toxicity. Depending on the nature and type of damage there appears to be a differential response mediated by p53.

Taken together, this research explores ways in which p53 signaling and biology can be used to enhance efficacy and limit toxicity following specific chemotherapy treatments.

iv TABLE OF CONTENTS

LIST OF FIGURES ...... viii LIST OF TABLES ...... x ABBREVIATIONS ...... xi ACKNOWLEDGEMENTS ...... xiii CHAPTER 1: LITERATURE REVIEW ...... 1 P53 and cancer ...... 1 P53 and tumor suppression ...... 2

WT P53 induces cell cycle arrest after cellular stress or DNA damage ...... 3

P53 and apoptosis ...... 4

Is role of p53 in cell cycle arrest and apoptosis important for tumor suppression? ... 6

P53 and DNA damage repair ...... 8 P53 and metabolism ...... 9

P53 regulation of glycolysis and the Pentose Phosphate Pathway (PPP) ...... 10

P53 control of mitochondrial metabolism and ROS ...... 12

P53 and nucleotide metabolism ...... 13 Exploiting p53 status and signaling pathway for therapy ...... 15 Can p53 act as transcriptional repressor? ...... 17

Indirect repression by p53 ...... 17

Recruitment to chromatin modifiers ...... 18

Interference and competition with transcription factors ...... 18

Non-coding RNA and p53 mediated repression ...... 19

v Direct repression by p53 ...... 19 Chemotherapeutics that trigger p53 responses ...... 20

5-Fluorouracil (5-FU) ...... 20 Mechanism of action ...... 21 Modes of action and mechanisms of resistance to 5-FU ...... 22 Thymidylate synthase (TS) inhibition by 5-FU ...... 22 DNA damage, TS, and 5-FU ...... 23 RNA damage by 5-FU ...... 24 Dihydropyrimidine dehydrogenase (DPYD) and 5-FU sensitivity ...... 25 Tumor suppressor p53 status and 5-FU response ...... 25 Topoisomerase poisons, inhibitors and p53 pathway ...... 26 Mechanism and modes of action ...... 26 Mechanisms of resistance and toxicity of TOP drugs ...... 27 Dihydropyrimidine Dehydrogenase (DPYD) gene regulation ...... 29

Characterization of the gene and the protein ...... 29

Regulation of the DPYD gene ...... 29

Clinical Pharmacogenetics of DPYD ...... 31 Check-point kinase 2 (Chk2) regulation of p53 ...... 34 Chk2 and DNA Damage Repair (DDR) ...... 36 Chk2 and p53-dependent cell cycle arrest ...... 37 Chk2 and p53-dependent apoptosis ...... 38 Role of Chk2 in mitosis ...... 39 SCOPE OF THE THESIS ...... 41 CHAPTER 2 ...... 43 P53 represses pyrimidine catabolic gene dihydropyrimidine dehydrogenase (DPYD) expression in response to thymidylate synthase (TS) targeting...... 43

vi Introduction ...... 43 Material and Methods ...... 45 Results ...... 52 Discussion ...... 72 CHAPTER 3 ...... 77 Targeting of Chk2 as a countermeasure to dose-limiting toxicity triggered by topoisomerase-II (TOP2) poisons ...... 77 Introduction ...... 77 Materials and Methods ...... 79 Results ...... 84 Discussion ...... 100 CHAPTER 4 ...... 103 Differential expression of Dihdropyrimidine dehydrogenase (DPYD) in mutant p53 colorectal cancer cells and its modulation to study mutant specific sensitivity to 5-FU...... 103 Introduction...... 103 Materials and Method ...... 103 Results ...... 105 CHAPTER 5 ...... 113 DISCUSSION ...... 113 References ...... 127 Appendix: Letters of Permission ...... 154

vii LIST OF FIGURES

Figure 1- 1: The complex network of p53 signaling: p53 activation and downstream signaling in tumor suppression ...... 3 Figure 1- 2: p53 metabolic regulation in cancer: ...... 10 Figure 1- 3: 5-FU mechanism of action ...... 21 Figure 1- 4: Mechanism of action of Topoisomerase II inhibitors ...... 26 Figure 1- 5: Chk2 structure and signaling ...... 35 Figure 2-1:Combined in-silico and chromatin immunoprecipitation(ChIP) identifies p53 downstream DNA binding sites(p53BS) in DPYD gene ...... 54 Figure 2- 2: p53-dependent repression of DPYD expression in intact liver and impact of human p53 polymorphic variants on liver expression of DPYD...... 56 Figure 2- 3: The tumor suppressor p53 represses dihydropyrimidine dehydrogenase (DPYD) expression ...... 58 Figure 2- 4: DPYD p53 co-localization and Epigenetic marks at DPYD promoter ...... 59 Figure 2- 5: Analysis of DPYD protein expression in CRISPR edited p53 binding site in HCT- 116 cells ...... 60 Figure 2- 6: Impact of bodyweight and survival of mice targeted by Gimeracil following IV administration of 5-FU ...... 62 Figure 2- 7: Assessment of hematological parameters following IV 5-FU administration ...... 62 Figure 2- 8: Tp53 specific liver depletion upregulates the catabolism of 5-FU through DPYD . 65 Figure 2- 9: Assessment of liver toxicity following IV 5-FU treatmen of tumor bearing mice .. 66 Figure 2- 10: p53 represses the expression of DPYD specifically following thymidylates synthase (TS) inhibition due to thymidine deficiency ...... 68 Figure 2- 11: p53 dependent repression of DPYD is dependent on signaling fro ATM and DNA- PK following TS inhibition ...... 70 Figure 2- 12: TCGA analysis of DPYD and p53 expression and correlation with overall survival of CRC patients ...... 72 Figure 3- 1: Chk2 targeting protects from toxicity triggered by DNA damage in-vivo and In-vitro ...... 86 Figure 3- 2: Chk2 is a mediator of toxicity triggered by TOP2-poisons ...... 89

viii Figure 3- 3:Chk2 triggers dose limiting toxicity in mice in-vivo following etoposide exposure ...... 93 Figure 3- 4: Pharmacological inhibitors of Chk2 (Chk2i) protect normal human and mouse cells from etoposide-induced killing...... 95 Figure 3-5: A combined in-silico and genetic screen identifies NSC105171, a novel lead Chk2ki with activity in-vitro and in-vivo ...... 97 Figure 3- 6: Schematic showing the development of a screening strategy to isolate novel pharmacological Chk2i with in-vivo activity ...... 98 Figure 4-1: Expression of DPYD protein in HCT-116 cells differing in p53 status ...... 105 Figure 4-2:Cellular viability of cells following 5-FU administration ...... 106 Figure 4-3: Tumor-sphere formation of HCT-116 cells ...... 108 Figure 4-4: DPYD mRNA expression in colorectal cancer patients from TCGA based on different p53 mutation status...... 109 Figure 5- 1: Benefits outlining S-1 therapy ...... 119 Figure 5-2: Different outcomes of p53 signaling following DNA damage therapy ...... 121

ix LIST OF TABLES

Table 1-1: List of DPYD genotypes and pharmacological dosing recommendation ...... 31

Table 2-1: Patient characteristics of Figure 2-11E ...... 71 Table 5-1: Summary of clinical trials to date investigating the efficacy and safety of S-1 in Gastrointestinal cancer……………………………………………………...117

x ABBREVIATIONS

5-Fluorouracil (5-FU)

Ataxia Telangiectasia and Rad3-Related Protein (ATR)

Ataxia Telangiectasia Mutated (ATM)

Base excision repair (BER)

Check-point kinase 1 (Chk1)

Check-point kinase 2 (Chk2)

Colorectal Cancer (CRC)

Deoxythymidine triphosphate /thymidine(dTTP)

Deoxyuridine triphosphate (dTUP)

Dihydropyrimidine dehydrogenase (DPYD)

DNA-damage-Repair (DDR)

DNA-dependent protein kinase (DNA-PK)

Homologous Recombination (HR)

MicroRNA (miRNA)

Mismatch repair (MMR)

Mitochondrial DNA (mtDNA)

Mitochondrial Membrane permeabilization (MOMP)

Mouse Embryonic Fibroblast (MEF)

Non-homologous end joining (NHEJ)

Non-small-cell lung cancer (NSCLC)

Nucleotide excision repair (NER)

xi Oxidative Phosphorylation (OXPHOS)

Peripheral Blood Mono Nuclear Cells (PBMC)

Small cell lung cancer (SCLC)

Tegafur-gemiracil-oteracil (S-1)

Thymidylate synthase (TS)

Topoisomerase 1/i (Top1)

Topoisomerase 2/ii (TOP2)

Transcription-coupled Nucleotide excision repair (TC-NER)

Tumor suppressor p53 (p53)

xii ACKNOWLEDGEMENTS

At the very outset, I want to thank God for blessing me with strength, courage, determination and perseverance to pursue my passion for scientific research. I want to thank my parents and my sister who have been very motivating, encouraging and supportive all throughout my life. They have provided and enabled me to achieve all my dreams and been a great driving force. I owe my Ph.D. to them.

I am humbled, honored and deeply thankful to my dissertation mentor Dr. Wafik S. El-

Deiry, who has been a cradle for my achievements. His extreme passion, dedication and commitment to scientific research has been a great source of inspiration and motivation.

I am very proud to be part of a journey of such a prolific researcher who has not only made seminal discoveries in cancer research but also dedicated his life to finding cures to cancer patients. I am very thankful to him for realizing my potential, trusting my abilities and investing his time and effort in mentoring me for a successful career.

I would like to deeply thank my program director Dr. Charles H. Lang, a student champion, for continued support, advice, assistance and mentoring throughout my graduate career.

I would like to thank my dissertation committee members Dr. Rosalyn Irby, Dr. Jin-Ming yang, Dr. Nelson Yee for providing insightful advice, comments encouragement and general support during my graduate work. Furthermore, I would like to thank all the faculty, staff, colleagues and peers of my graduate program at Penn State Hershey

Medical Center. Especially, I would like to thank Kathy Simon and Kathy Shuey, who have

xiii played role of “school moms” taking care of school paperwork, reminding me of deadlines and providing “things to do” lists.

I would like to thank all my past and present lab members of the El-Deiry lab, especially

Dr. Niklas Finnberg, a great collaborator and mentor and a true supporter of all my efforts.

I would like thank Liz Hernandez, David Dicker and Denise Andrisani (“Lab Mom”) for being wonderful friends and lab-mates whom I will cherish beyond my stay in the lab. I would also like to thank all my colleagues and staff at the Fox Chase Cancer Center for helping me secure my grant.

My friends outside the lab Turan Aghayaev, Yulya Peshkova, Kishore Punnath and Lovely to name a few, certainly not an exhaustive list, have been very supportive providing best memories and laughter much needed in graduate school. I am thankful to my parents who brought me into this world. I am also very thankful to my in-laws for their great support. Last, and most important, I wish to thank my wonderful loving wife Vasudha

Bharatula, an outstanding companion, and soul-mate for being a core strength of my personal and professional life.

xiv CHAPTER 1: LITERATURE REVIEW

P53 and cancer

P53 is one of the most intensely studied molecules in the history of cancer research. P53 was originally identified in the late 1970’s as a putative oncoprotein co-associated with simian SV40 T-antigen or as a human tumor antigen (1-6). It was not until a decade later, during which the p53 gene was cloned (7-12) and identified to be mutated in tumor versus normal tissue or inactivated by viral antigens (13-16), that it became to be known as a tumor suppressor gene (17-19). Cancer in general has mutations in genes involved in various processes such cell growth, survival, proliferation, metabolism etc. Some of the highly mutated genes include PTEN, KRAS, BRAF, PI3 Kinase, and MYC. Most of these genes such as BRAF are highly mutated in one type of cancer such as melanoma where its frequency is about 50% whereas in other cancers these mutations are present anywhere between 5-7%. Some genes are mutated as parts of pathways such as the p16-CDK-Rb-E2F or Wnt-APC-beta-catenin pathway. p53 is mutated in more than 50% of all human cancers while virtually 100% of tumors have defects in the p53 pathway.

Soon after the discovery of p53 as tumor suppressor, there were several observations of p53 mutations made in many tumor types. Cancer such as high grade serous ovarian cancer have almost 98% mutation rate in p53 (20) whereas some testicular germ cell cancers have only 1-2%. Over 3 decades of research on p53, it has been very clear that p53 plays a central role in cancer initiation, progression and cell death. It plays critical role at various levels in tumor biology. At the cellular level, p53 is involved in various processes such as cell cycle arrest, apoptosis, antioxidant response, migration, invasion,

1 metabolism, autophagy, angiogenesis and DNA damage repair. At the tissue level it controls cellular dynamics, differentiation, immune activation, clonal expansion and selection thus contributing to organismal homeostasis (although p53 knockout mice develop normally). A major question has been how p53 is able to coordinate such complex processes. As a transcription factor, p53 acts as a central hub in coordinating major cellular processes through complex transcriptional networks, protein-protein interactions and signaling cascades thus carefully balancing different cellular outputs to diverse inputs. In addition, it plays a central role in determining outcomes to cancer therapeutics. Hence, p53 is not only a “Guardian of the genome” but also a “Guardian of

Homeostasis” (Figure1-1). Understanding the what, when, how and why of p53 function is a continuing effort and challenge to basic and translational research.

P53 and tumor suppression

One of the major outstanding questions has been how p53 suppresses tumors. Even though much is known about p53 function, and its effects on various hallmarks of cancer, the key determinants of the p53 response and the precise molecular mechanisms which help suppress tumor initiation still remain less obvious. Perhaps by now it is not hard to imagine that p53 potentially does this by various mechanisms using multiple pathways synergistically. We review key functions of p53 that are relevant to this thesis and play a major role in determining chemotherapeutic agent toxicity and efficacy ultimately affecting tumor suppression and regression.

2

Figure 1- 1: The complex network of p53 signaling: p53 activation and downstream signaling in tumor suppression

Schematic diagram of regulation of p53 and various downstream regulated cellular processes

Source adapted from Kathryn et al, Nat Rev Cancer. 2014 May;14(5):359-70.

WT P53 induces cell cycle arrest after cellular stress or DNA damage

The earliest discovery of p53 gene dates back to its detection in sea anemone. Its primary role in sea anemone was to maintain genomic integrity in gametes (21). Over a billion years of evolution of this gene, in almost all higher order animals, its role in maintenance of genomic integrity still remains one of the key effector functions of this gene. One of the well-studied functions of p53 in bringing about G1 cell cycle arrest is through activation of

3 p21 (22,23). P21 was one of the first genes downstream of P53 to be identified and the first universal cell cycle inhibitor (24). Binding of p53 to the p21 promoter increases its transcriptional activity strongly. P21 activation leads to its binding to CKD4/cyclin D and

CDK2/cyclin E thus preventing phosphorylation of the downstream target gene Rb (25).

Hypophosphorylated Rb binds to the E2F family including E2F1 and prevents target genes required for S-phase of the cell cycle to be activated. Apart from acting in G1, p53 can also activate GADD45 and 14-3-3-s which in turn block the formation or prevent translocation of the cyclin B/CKD1 complex to the nucleus thus causing G2 arrest (26,27).

More recently, p53 has been shown to induce microRNA’s such as mir34a which brings about cell cycle arrest by downregulating genes such as Myc (28), CCND1, or CDK4/6

(29,30).

P53 and apoptosis

The role of p53 in apoptosis was one of its earliest known functions and is highly conserved. In drosophila, p53 cannot induce p21 or cell cycle arrest as it is generally thought this organism does not need those functions as most of its adult tissue is post- mitotic in nature (31,32). Consistent with this role for p53 in apoptosis, in mammalian tissue p53 brings about apoptosis in two main ways, namely intrinsic and extrinsic apoptosis. Intrinsic apoptosis by p53 involves activation of BH3-only Bcl-2 family members like Puma, Noxa, Bad, Bax, p53 AIP1(33-37). Intrinsic apoptosis mainly controls mitochondrial outer membrane permeabilization (MOMP) (38). P53 activates Puma and

Noxa which prevent Bcl2 family members such as Mcl-1, BcL-XL and BcL2 from inhibiting

Bax and Bak (36,39). Activated Bax and Bak interact with tBid and Bim to localize to outer

4 mitochondrial membrane releasing cytochrome c that now forms a complex with Apaf1

(40). This leads to the formation of the apoptosome and further activation of executioner caspases such as 3 and 7. The extrinsic apoptosis engaged by p53 brings about cell death in a paracrine fashion. The activation of FAS/CD95 (41) and DR5 (42) sensitizes cells to the ligands FASL and TRAIL, respectively, and activates caspase 8. This in turn activates executioner caspases bringing about cell death (43). It is believed that DR5 is important for chemotherapy induced toxicity and efficacy. DR5 is major p53 target gene upregulated following chemotherapy and radiation (44). Work from our lab has shown that DR5 activation sensitizes intestinal epithelial cells to chemotherapy induced cell death. Specifically, DR5 activation sensitizes intestinal crypt Lgr5+ stem cells to chemotherapy induced cell death (45). Silencing of DR5 in-vivo confers colon tumor xenograft tumor resistance to 5-FU (46). In addition, transcription- independent functions of p53 in apoptosis have been observed through direct promotion of MOMP. Localization of p53 at mitochondria is observed and precedes MOMP and caspase activation (47,48).

The activation of Puma relieves p53 from the p53-BcL-XL complex which then activates

Bax through direct binding. P53 can also form a complex with Bcl2-Bcl-XL releasing tBid, which in turn can activate Bax (40,48,49). Activation of Bak by p53 can also occur through tBid or from direct binding of p53 to Bak, promoting its release from an inhibitory Mcl-1 complex (50).

5 Is the role of p53 in cell cycle arrest and apoptosis important for tumor suppression?

The role of p53-dependent apoptosis and cell cycle arrest in determining tumor suppression is a much-as there are several observations and studies which argue both for and against a role for p53 activation of these two processes with regards to tumor suppression. For example, studies from the a Eµ-Myc mouse model suggest that loss of p53 in pre-B and B lymphocytes greatly accelerates lymphoma growth where decreased apoptosis was observed in these cells (51-53). In this model, the average latency for tumor development in p53-knockout mice was 29 days compared to those which had intact p53 (latency > 110 days) (54). Overexpressing the Bcl2 in these mice eliminated the necessity to lose p53 thus highlighting the importance of apoptosis in this model (55).

Puma-knockout in these mice further accelerated tumor progression emphasizing the fact that apoptosis is important for suppression of lymphomas (56,57). P53-null mice develop spontaneous thymic lymphomas. It is well known that apoptosis in the thymus is important for clonal selection of T-cells as well as in determining the sensitivity of T-cells to ionizing radiation and etoposide treatment (58). Studies in choroid plexus brain tumors carrying defective SV40 T-antigen showed enhanced tumor formation with infrequent apoptosis suggesting again that apoptosis is important for tumor suppression (59).

Furthermore, reactivation of p53 expression in H-RasV12-induced hepatocellular carcinoma caused tumor regression and immune clearance (60). On the contrary, mouse models carrying various point-mutations in p53 have questioned the importance of the role of p53-dependent cell cycle arrest and apoptosis in tumor suppression. A humanized knock-in mouse carrying p53-R172P (equivalent of human R175P) found in human

6 tumors and Li-Fraumeni patients failed to develop thymic lymphomas (61). The p53-

E117R mutant mouse (Human E180R) which retains partial cell cycle arrest but is completely deficient in apoptosis also failed to induce thymic lymphomas (62). This mutant however retained activities related to cellular senescence and metabolic regulation. The p533KR mouse defective in 3 acetylation sites was able to suppress tumor initiation in the absence of both apoptosis and cell cycle arrest. However, these retained energy metabolism and an antioxidant response (63). Studies on a p5325,26 mutant allele defective in activating cell cycle arrest and apoptosis retained full capability of cellular senescence and tumor suppression by activating genes in cytoskeleton remodeling, DNA repair, or lysosomal transport (64). Finally, a Puma/Noxa/p21 triple knockout mouse also efficiently prevented spontaneous tumor formation in the absence of cell cycle arrest and apoptosis (65). However, it should be noted that additional targets of p53 such as DR5,

Bid, Bax, Bad, caspase 10 (66-68) etc. have not been studied and hence the role of cell cycle arrest and apoptosis in tumor suppression cannot be ruled out.

Therefore, it is likely that the full spectrum of p53 function is needed for tumor suppression. However, it is important to acknowledge that emerging roles of p53 outside its role in cell cycle arrest and apoptosis may be involved in tumor suppression. These roles of p53 are likely to play an important role not just in tumor suppression but also will likely decide outcomes to anti-cancer drugs.

7 P53 and DNA damage repair

Cells in our body are constantly attacked by genotoxic insults arising from endogenous and exogenous agents such as free radicals and oxidative stress. It is highly imperative that genomic integrity of cells be maintained in all times to prevent unnecessary death or cancer. P53 being the guardian of the genome orchestrates a multitude of DNA damage repair processes in response to genomic insults. P53 has been known to play roles in different repair processes such as NER, BER, NHEJ and HR. P53 transactivates the genes DDB2 and XPC encoding proteins involved in NER (69-71). These two proteins play a role in the resolution of CPD and 6-4PP by identification and targeting excision repair. P53 can directly interact with XPB and XPD subunits of TFIIH and decreases their helicase activity (72,73). In TC-NER, p53 helps in global recognition of damages sites by cooperating with TFIIH in relaxing the chromatin (74). P53 has both transcription-coupled and transcription-independent roles in BER and the relation seems to be quite reciprocal in nature. While the BER protein APE1/Ref-1 can enhance p53 binding and tetramerization (75,76), p53 can in turn repress its transcription as part of its tumor suppressor function to prevent cancer cells from undergoing repair (77). It is believed that p53 can increase BER activity in the G0/G1 phase but has limited activity in the G2/M phase of the cell cycle (78). Additionally, 3-MeAde DNA glycosylase is repressed by p53 consistent with a tumor suppresser role in maintaining genomic integrity (79). P53 can transactivate the OGG1 (80) and MUTYH (81) genes and can physically interact with

OGG1 and APE-1 to remove 8-oxoG adducts from DNA (82). Finally, studies have also shown that direct binding of p53 to DNA polymerase b through its N-terminal domain increases DNA polymerase b stability and replication (83).

8 P53 and metabolism

The role of p53 in controlling metabolism is a relatively recently understood phenomenon adding to its repertoire of ever growing functions As an emerging guardian of homeostasis p53 carefully manages energy requirement, nutrient availability, and antioxidant response of cells by controlling a multitude of metabolic processes which otherwise unchecked can cause energy depletion, nutrient starvation, altered cellular proliferation or arrest and finally cell death (84). Importantly, the role of p53 in metabolic regulation is very essential for tumor suppression. Figure 1-3 highlights the role of p53 role in metabolic regulation. In the following sections, we have limited our discussion to metabolic activities controlled by p53, which are pertinent to the response to cellular stress caused by some chemotherapeutics such as 5-FU.

9 Metabolites 2017, 7, 21 2 of 18

It is therefore not surprising that 50% of all human tumours carry genetic alterations that lead to the inactivation of the p53 pathway. Mostly, these alterations are missense mutations in the coding region of the TP53 gene, but this varies among different tumour types [17,18]. P53 mutations are mainly found in solid tumours and occur at high frequency in inflammation-associated cancers [19–22]. Many p53 mutations cause conformational changes of the DNA binding domain of the p53 protein, leading to reduced binding of p53 to the promoters of its target genes [23]. Importantly, as p53 functions as a tetramer [24], the presence of mutant p53 in cancer cells has a dominant negative effect on wild type p53 function even in heterozygous cells. Moreover, since mutant p53 cannot activate the expression of its negative regulator MDM2, mutant p53 protein is stabilised [25] and can exert additional tumour promoting functions [26]. In general, loss of p53 function causes resistance to DNA damage and prevents apoptosis or senescence in cancer cells [27–29]. Tumour development is accompanied by changes in cellular metabolic activity, which allows cancer cells to grow and proliferate under adverse conditions. The influence of p53 on cellular metabolism is complex and involves multiples nodes of regulation (summarised in Figure 1). p53 changes the activity of multiple metabolic pathways, including glycolysis, mitochondrial oxidative phosphorylation and fatty acid synthesis via transcriptional and non-transcriptional regulation. In addition, p53 governs the adaptation of cancer cells to nutrient and oxygen deprivation, which is crucial for the survival under the metabolically compromised conditions shaped by the tumour microenvironment. Importantly, it has been shown that the regulation of metabolic activity is essential to the tumour suppressive function of p53 [30].

Figure 1- 2: Figurep53 metabolic 1. Regulation ofregulation glycolysis and in mitochondrial cancer: metabolism by p53.

Source:p53 regulates Adapted glycolysis from Floter and J., mitochondrial et al 2014, Metabolites. metabolism 2017 through May multiple 20;7(2) mechanisms. It reduces glucose uptake by directly repressing the transcription of genes coding for the glucose transporters GLUT1 and GLUT4 and by indirectly repressing GLUT3. p53 also reduces expression of HK2, which controls the production of G6P. In response to acute activation by DNA damage, p53 induces the The above figure depicts the central role of p53 in controlling glycolytic and mitochondrial metabolism pathways which play a central role in the response to cellular stress. The genes that are activated by p53 are shown in green, whereas genes that are inhibited by p53 are shown in red.

P53 regulation of glycolysis and the Pentose Phosphate Pathway (PPP)

Aerobic glycolysis, also known as the “Warburg effect,” is a hallmark of cancer cells (85).

Cancer cells rapidly consume glucose and convert it to lactate for quick ATP production.

For a long time, it was poorly understood how cells could undergo aerobic glycolysis when

OXPHOS could produce more ATP for cells. However, it is clear now that cancer cells utilize glycolysis mainly as a source for channeling metabolites for production of building

10 blocks such as nucleotides, amino acids, lipids, redox balance, and also for epigenetic rewiring by controlling SAM production. Thus, the glycolytic pathway plays a central role in altering metabolic flux under conditions of stress. p53 controls the import of glucose by direct transcriptional repression of GLUT1 and GLUT4 (86) and indirect repression of

GLUT3 (87). Recently, it has been noted that suppression of GLUT1 and GLUT3 is important for protecting normal cells from 5-FU induced toxicity (88). Glucose transporters are well known to be regulated by HIF1-alpha and clearly, they can be co-regulated by p53. P53 directly represses HK2 by binding to the promoter regions of the gene (89). The coupled control of glucose import and kinase activity limits the commitment to glycolysis.

In addition, p53 controls TIGAR which reduces F2,6BP to F6P (90). F2,6BP is required for activity of PFK1 which converts F6P to F1,6BP. Activation of TIGAR leads to inhibition of PFK1 thereby channeling of G6P to PPP to restore nucleotide production during times of DNA repair (91). In fact, TIGAR induction by p53 in response to 5-FU was observed in gastric cancer cell lines, and this was mediated by NF-kB. In the same study, polymorphic variants of p53 such as R72P were shown to affect this process (92). Further, p53 has recently been shown to upregulate PFKFB3 in response to genotoxic stress and coordinate DNA repair by enhancing nucleotide synthesis through the PPP pathway (93).

PPP can also be inhibited by p53 through its binding to G6PD, the first rate limiting enzyme of this pathway, thus limiting its enzymatic activity. Since the PPP pathway is important for supply of nucleotides as well as for NADPH production, the opposing effects of p53 on the PPP pathway may control the supply of nucleotides during acute DNA damage and repair, while eliminating excess supply of them during aberrant proliferation

(94).

11 P53 control of mitochondrial metabolism and ROS

The mitochondria are an essential organelle for ATP production and cell survival. P53 actions are imperative to control such cellular functions. P53 increases OXPHOS through promoting ETC function. For example, it increases the expression of SCO2 which is needed for assembly of complex 4 of ETC as well as MT-CO2 which is subunit of complex

I (95). p53 controls mtDNA copy number and DNA mass by activating p53R2 (95-97).

P53 controls mitochondrial quality and turnover through control of MIEAP (98). In addition, p53 controls the shuttling of pyruvate to acetyl coA by transcriptional repression of PDK2 which otherwise inhibit PDH activity (99,100). P53 further represses MCT1 a key transporter which helps in regeneration of NAD+ equivalents by converting pyruvate to lactate (99). Glutamine is an important nutrient source for several tumors as it is a key metabolite of TCA anaplerotic processes such as production of fatty acids, lipids and pyrimidine metabolism and antioxidant defense. P53 increases the expression of GLS2 which converts glutamine to glutamate which is further converted into alpha-ketoglutarate

(a-KG), a key metabolite in the TCA cycle, thus rescuing stressed normal cells (101,102).

However, several tumors exploit by upregulating GLS1 which is p53 independent to gain access to a multitude of other biosynthetic molecules which now may be synthesized as result of glutamine anaplerosis.

Control of mitochondrial ROS production by p53 is highly context-dependent as p53 can promote or inhibit ROS production. In response to DNA damage or other cellular stress p53 can induce the expression of pro-oxidant genes such as PIG3, Puma, Noxa further exacerbating the oxidative stress perhaps at times of excessive damage contributing to

12 cell death (35,37,39,103). Conversely, p53 also suppresses of pro-oxidant genes such as

COX2 and NOS2 (104,105). During states of transient stress p53 can activate ALDH4,

TP53INP1 and Sestrins 1/2 (106-108). P53 by cooperating with p21 can directly activate

Nfr2, a transcriptional factor that upregulates the antioxidant response (109). p53 also represses the expression of the antioxidant gene SOD2 that removes oxygen radicals from mitochondria (110). NAPDH production is key for generation of GSH to reduce ROS levels. P53 controls ME1 and ME2 contributing to cellular NADPH production (111).

Conversely p53 represses SLC7A11, a cysteine/glutamate antiporter important for cysteine import for GSH production (112). This decrease in SLC7A11 can induce ferroptosis and act as a mechanism of tumor suppression (113).

P53 and nucleotide metabolism

Even though the relation between cellular proliferation, DNA synthesis and p53 dates back several decades, the direct regulation of nucleotide metabolism genes in by p53 in tumor suppression is relatively less studied. Most research in this area has evaluated the status of p53 to the response of DNA damaging agents including antimetabolites that interfere with DNA synthesis or free nucleotide pools. The first major indication that p53 status could be important to certain chemotherapeutic agents came from studies carried out in isogenic MEF’s lacking p53 gene, where loss of p53 greatly increased resistance to these agents due to attenuated apoptosis (114). Later it was observed that, p53 reacts to altered nucleotide levels, specifically to ribonucleotide depletion in the absence of DNA damage in normal human fibroblasts. The normal cells underwent reversible Go/G1 phase arrest in response to UTP and GTP synthesis inhibitors resembling that of quiescent cells

13 (115). However, studies in tumor cells previously had shown more of a heterogeneous response to antimetabolites with no clear indication to the importance of p53 status.

Subsequently, Vogelstein’s group convincingly. characterized that, loss of p53 greatly resulted in increased resistance to the drug 5-FU, a major anti-metabolite used to treat many human cancers (116). However, today it is clear that the response of p53 in cell cycle arrest and apoptosis alone may not explain sensitivity to these agents. Such data indicates that cellular mechanisms exists that could activate p53 downstream of nucleotide stress or pool imbalance, and potentially enzymes in pathways of nucleotide biosynthesis or salvage may play a role in determining response to antimetabolites. The first major target of p53 controlling DNA biosynthesis was the discovery of ribonucleotide reductase subunit 2 (p53R2), that supplies nucleotides by reducing NTP’s to dNTPs at the sites of DNA damage (117). Studies which followed showed an increasingly complex relationship between p53 and nucleotide synthesis such as the reciprocal relation between TS and p53. P53 mRNA has been shown to be directly regulated by the TS enzyme where inhibition of the enzyme releases p53mRNA from its complex with TS and increases it translation (118). Further, p53 repress dUTPase, the enzyme that removes

UTP from DNA strands. GMPS, an enzyme that converts XMP to GMP, an early step in the guanidine synthesis pathway that forms a complex with USP7 (deubiquitinase) and stabilizes p53 in the nucleus (119). P53 has been reported to control regulation at the level of microRNA (miRNAs). MiR29b and miR 192 can suppress MTHFR and DHFR genes, key enzymes responsible for production of 5-methyltetrhydrofluorate (5-THF) and methyltetrahydrofolate (mTHF) required for purine and pyrimidine synthesis (120,121).

Mir34a can also represses IMPDH having direct effect on Ras signaling (122). Thus,

14 together such studies are now beginning to unravel direct mechanisms by which p53 can control nucleotide metabolism in conjunction with coordinating DNA replication, stress, damage and repair, however like mentioned before, it is surprising how relatively little is known in this regard and thus more studies are needed to focus on p53 control of nucleotide metabolism to understand and improve the use of chemotherapeutics.

Understanding deeper the role of p53 in nucleotide metabolism could further identify patient subgroups who would benefit from antimetabolite therapies while minimizing drug toxicity.

Exploiting p53 status and signaling pathway for therapy

Targeting p53 function has been a central focus for many years as its signaling pathway is dysfunctional in most if not all cancers. Not surprisingly, many approaches have been taken to restore the p53 pathway. Among them the bulk of studies that have advanced to the clinic have focused on MDM2, MDM4 and MDMX pathways. These studies have identified molecules which block the WT p53 from interacting with MDM family members.

Other work has attempted to convert mutant p53 into its WT conformation thus trying to restore p53 activity. Some studies have a employed more indirect approach of restoring p53 pathway through p53 family members such as p63 and p73. Others have focused on inhibiting mutant p53 Gain-of-Function (GOF) (123). However, here we discuss only those pertaining to exploiting synthetic lethality with loss of WT p53 function including ways to exploit DNA damage, cell cycle checkpoint and metabolism.

Another early approach involved gene therapy. Gendicine is an adenoviral based vector which restores WT p53 function. This drug has been approved for clinical use in China

15 but remains investigational in the USA. However, some of the current challenges to p53 restoration involve the various cellular outcomes following p53 activation leading to potential toxicity concerns. The exact molecular events, pathway activation and outcomes following p53 restoration are very diverse and highly context and cell type- dependent. In addition, mutant p53 poses a higher barrier to prevent tumor killing due to its dominant negative effect on WT p53 and finally heterogeneity of cancer poses an added problem.

But in recent years exploitation of the p53 pathway has been carried out with increased knowledge of the type of chemo-radiotherapy to be employed, tumor type, along with drug combination strategies to safely enhance efficacy and limit toxicity. As mentioned above, selective employment of strategies to eliminate p53-deficient tumors have made use of cell cycle checkpoints as a vulnerability. This is termed “Cyclotherapy”. The rationale is that p53-proficient cells can induce G1 arrest with the induction of p21 and prevent further damage in G2/M phase of the cell cycle, in contrast, p53-deficient cells will not arrest in

G1 and thus progress through G2/M phase with damage leading to catastrophic mitosis and cell death. The G2-checkpoint is mediated by ATM and ATR kinases, therefore, selective targeting of these enzymes might provide clinical benefit. As expected, in-vitro studies have revealed selective killing of p53-deficient cells when DNA damaging agents when combined with ATR inhibitors(124). Patient-derived xenograft (PDX) models in

TNBC mouse models also show enhanced efficacy when p53 targeting is combined with

CHK1 inhibitors (125). ATR inhibition in RAS oncogene-overexpressing cells has shown an improved cancer killing increasing the therapeutic window (126-128). Similarly, ATM inhibition along with DNA damaging agents have shown promising results in preclinical analysis (129,130). Apart from the ATM-ATR pathway, inhibition of MK2/p38MAPK in

16 KRAS-driven NSCLC autochthonous models displayed an enhanced effect against tumors with combined loss of p53 and MK2 (131). Preclinical drug combination studies whereby the first drug selectively arrest WT p53 cells and the second drug targeting mutant p53 cycling cells have shown some promising results. Recent studies have shown benefit to WT p53 cells when pretreated with Nutlin-3A followed by PLK inhibition. WTp53 cells were protected from cell death while p53-deficient colorectal cancer cells had increased sensitivity to PLK1 inhibition in a xenograft model (132). Altogether this indicates promising avenues for exploiting the p53 pathway for rationally designed clinical trials to provide optimal benefit to patients.

Can p53 act as transcriptional repressor?

It is well-established that p53 is an efficient trans-activator of genes. Hundreds of p53 targets have been studied belonging to all aspect of p53 functions such as cell cycle arrest, apoptosis, metabolism, and DNA repair with functional consequences well understood. However, the mechanistic, functional and net biological consequences of p53-mediated repression remain less-well understood.

Indirect repression by p53

Indirect repression by p53 is carried out in many ways. For example, exclusion of transcription factors (TF) from activating the gene has been observed in case of Sgk.

Glucocorticoid receptor (GCR) binding to the Sgk promoter is inhibited by p53 through direct binding to the former and prevention of p53 binding to the glucocorticoid receptor element (GRE) (133). IGF1R transcription is inhibited through p53 through binding to the

17 p53RE and recruitment of HDAC (134-136). Sequestration of the Sp1 transcription factor by p53 prevents hTERT activation (137).

Recruitment to chromatin modifiers

P53 can recruit chromatin modifiers to the promoter regions of genes through its binding to p53RE. p53 recruits msin3A and HDAC to the promoters of the Mad1gene (138). The same two factors are also involved in repressing Nanog, C-Myc and survivin expression

(139-141). AFP gene expression is inhibited by recruitment of mSin3A and by a H3K9 binding factor HP1 (142,143).

Interference and competition with transcription factors

P53 can compete with transcription factors and prevent accessibility to their target sequence. HNF3 binding to the AFP promoter when occupied by p53 results in eviction of HNF3 and thus impedes transcriptional activation of AFP (143). Competition with the

Sp1 binding site at the promoter of POLD1 prevents its transactivation by Sp1 (144). P53 binds to p53RE on the HBV1 enhancer element to repress its expression in coordination with EP binding site next to it. EP interferes with p53 binding restricting its transactivation ability from this element (145).

Cooperation with p21 to repress genes

Several cell cycle genes which were previously thought to be direct targets of p53 mediated repression have also been found to be in some ways regulated by p21. Genes such as CDK1, Cdc25, Cyclin A and B, MYBL2 and PLK1 are all thought to require p21

18 for their downregulation (146). The proposed mechanism by which p21 does this, is through coordination with the DREAM complex which consists of proteins p130, p107 and

E2F all of which are all known to be recruited to several cell cycle target genes. Loss of p21 has in these systems abrogates their repression by p53 (146,147). Thus, some of the p53-dependent gene repression is mediated by p21, however, the exact molecular mechanisms remain unknown.

Non-coding RNA and p53 mediated repression

Mir34a/b are both well-known targets of p53 that downregulate Cdk4 and Bcl2 (30,148).

Mir34a suppress Sirt1 expression thus having a positive feedback loop in increasing p53 acetylation and stabilization (149). Mir 192, 194 and 196 can be upregulated by p53 and which in-turn degrade MDM2 mRNA thus affecting the p53/MDM2 axis in multiple myeloma (150). Mir145 can affect c-Myc transcription (151). P53 can enhance processing of mir-16-1, mir143 and mir 145 through a transcription-independent way by controlling the DROSHA complex by associating with the DEAD-box RNA helicase p68/DDX5 (152).

Long noncoding RNA such as linc-RNA-21 has been shown to be activated by p53 which mediates p53 global repression of target genes (153).

Direct repression by p53

Since the identification of a global consensus sequence for p53 characterized by

RRRCWWGYYY (0-13bp) RRRCWWGYYY (24), many genes were identified as bona fide targets of p53. These targets were mainly trans-activated by p53 (154). Even though the presence of mismatches within the consensus sequence was known previously, the

19 importance of these binding sites to gene repression has been recently recognized as new technologies have been developed for genomics. It is estimated that approximately

20% of genes are transcriptionally repressed by p53 to possess p53RE. Genes such as

VEGF, Stm1, c-Myc and Cdc25 can directly repressed by p53 (155-158). Variations of the p53 consensus sequence have been suggested to explain p53-dependent transcriptional repression. For example, variation in the length of the spacer between 2 consensus half-sites has been regarded to potentially mediate repression. Use of two different consensus p53 binding sites within the same gene has been noted to result in differential regulation for genes in a context-dependent manner such as Plk1 and Cdc25c

(157,159). Alteration in the arrangements (head-head, head-tail and tail-tail) of consensus binding sequences has also been observed in case of Mdr1 and CD44 genes. The original p53RE has 4 binding sites with head to tail arrangement instead of head to head

(160,161). Further LASP1 gene repression is thought to occur because of mismatches in the core sequences of the decamer (i.e CWWG) in the p53RE. Therefore, overall p53 mediated repression is beginning to be recognized as an integral part of the p53’s transcription factor function in controlling various cellular processes that take part in tumor suppression and the stress response.

Chemotherapeutics that trigger p53 responses

5-Fluorouracil (5-FU)

5-FU was the first rationally made chemotherapeutic. It was first synthesized by

Heidelberger in 1957 (162). It was back then discovered due to the observation that rat hepatomas incorporated uracil faster than normal tissue (163) and many antimetabolites

20 such as 6-Azauracil (6-AZA) had good anti-tumor activity (164). Since its discoveryREVIEWS 5-FU hasSummary been a widely used chemotherapeutic for a variety ofTS tumor inhibition. typesTS catalyses including the reductive colorectal, methylation of deoxyuridine monophosphate (dUMP) to deox- •The fluoropyrimidine 5-fluorouracil (5-FU) is an antimetabolite drug that is widely ythymidine monophosphate (dTMP), with the breast,used for the pancreatic, treatment of cancer,particularly head and for neck, colorectal gastric, cancer. ovarian and esophagreduced FOLATEeal cancer5,10-methylenetetrahydrofolate. The greatest

•5-FU exerts its anticancer effects through inhibition ofthymidylate synthase (TS) and (CH2THF) as the methyl donor (FIG.2).This reaction beneficalincorporation impact of its metabolites of 5 into-FU RNA has and DNA.been seen in colorectal cancerprovides where the sole de usually novo source it of is thymidylate, combined which •Modulation strategies,such as co-treatment with leucovorin and methotrexate,have is necessary for DNA replication and repair. The 36- been developed to increase the anticancer activity of 5-FU. kDa TS protein functions as a dimer, both subunits of with•Molecular oxaliplatin biomarkers thator predictirinotecan tumour sensitivity as first to 5-FU and have second been identified, line therapy.which contain a nucleotide-binding site and a binding including mRNA and protein expression levels of TS. site for CH 2THF.The 5-FU metabolite FdUMP binds to the nucleotide-binding site of TS,forming a stable •DNA microarray analysis of5-FU-responsive genes will greatly facilitate the TERNARY COMPLEX with the enzyme and CH THF,thereby identification of new biomarkers, novel therapeutic targets and the development of 2 rational drug combinations. blocking binding of the normal substrate dUMP and inhibiting dTMP synthesis8,9 (FIG.2). Mechanism of action The exact molecular mechanisms that mediate events downstream of TS inhibition have not been fully 5-FU elucidated. Depletion of dTMP results in subsequent DPD depletion of deoxythymidine triphosphate (dTTP), which induces perturbations in the levels of the other DHFU deoxynucleotides (dATP,dGTP and dCTP) through various feedback mechanisms10.Deoxynucleotide pool imbalances (in particular, the dATP/dTTP ratio) are thought to severely disrupt DNA synthesis and repair, resulting in lethal DNA damage11,12 (FIG.2).In addition, DHFU DPD 5-FU TP FUDR TS inhibition results in accumulation of dUMP,which PRPP might subsequently lead to increased levels of deoxyuri- TK UP OPRT dine triphosphate (dUTP)13,14.Both dUTP and the 5-FU

PREDICTIVE BIOMARKERS metabolite FdUTP can be misincorporated into DNA. FUR UK FUMP FdUMP TS Molecular markers that predict inhibition Repair of uracil and 5-FU-containing DNA by the tumour sensitivity to nucleotide excision repair enzyme uracil-DNA-glycosy- chemotherapy. lase (UDG)15 is futile in the presence of high FUDP RR FdUDP (F)dUTP/dTTP ratios and only results in further false- FOLATES Family of essential vitamins that nucleotide incorporation.These futile cycles of misin- act as cofactors in one-carbon corporation, excision and repair eventually lead to DNA transfer reactions. FUTP FdUTP strand breaks and cell death. DNA damage due to dUTP O misincorporation is highly dependent on the levels of TERNARY COMPLEX DNA F the pyrophosphatase dUTPase,which limits intracellu- A stable complex that is formed HN damage 16,17 between 5-fluorouracil, RNA lar accumulation of dUTP (FIG.2).Thymidylate can damage thymidylate synthase and 5,10- O N H be salvaged from thymidine through the action of methylene tetrahydrofolate,and thymidine kinase,thereby alleviating the effects of TS that blocks synthesis of deficiency (FIG. 2).This salvage pathway represents a thymidylate by the enzyme. Figure 1 | 5-Fluorouracil metabolism. 5-Fluorouracil (5-FU; see structure) is converted to three main active metabolites: potential mechanism of resistance to 5-FU18. rRNA fluorodeoxyuridine monophosphate (FdUMP), (Ribosomal RNA).The RNA fluorodeoxyuridine triphosphate (FdUTP) and fluorouridine RNA misincorporation. The 5-FU metabolite FUTP is component of ribosomes,which triphosphate (FUTP). The main mechanism of 5-FU activation extensively incorporated into RNA, disrupting normal translateFigure mRNA into1- protein. 3: 5-FU ismechanism conversion to fluorouridine of monophosphate action (FUMP), either directly by orotate phosphoribosyltransferase (OPRT) with RNA processing and function. Significant correlations tRNA phosphoribosyl pyrophosphate (PRPP) as the cofactor, or between 5-FU misincorporation into RNA and loss of (Transfer RNA).tRNAs bond indirectly via fluorouridine (FUR) through the sequential action clonogenic potential have been shown in human colon Source: Longley et al., Nat Rev Cancer. 2003 May;3(5):330-8 with amino acids and transfer of uridine phosphorylase (UP) and uridine kinase (UK). FUMP is and breast cancer cell lines19,20.5-FU misincorporation them to the ribosomes,where then phosphorylated to fluorouridine diphosphate (FUDP), can result in toxicity to RNA at several levels. It not 5proteins-FU are is assembled either convertedwhich can into be either fluorodeoxyuridine further phosphorylated to the monophosphate active (FUMP) by the action of orate according to the genetic code only inhibits the processing of pre-rRNA into mature metabolite fluorouridine triphosphate (FUTP), or converted to 21,22 that is carried by mRNA. fluorodeoxyuridine diphosphate (FdUDP) by ribonucleotide rRNA ,but also disrupts post-transcriptional modi- phosphoribosyl transferase (OPRT) or by combined action of Uridine phosphorylase23,24 and uridine kinase reductase (RR). In turn, FdUDP can either be phosphorylated fication of tRNAs and the assembly and activity of snRNA or dephosphorylated to generate the active metabolites FdUTP snRNA/protein complexes,thereby inhibiting splicing of (UK).(Small nuclear FUMP RNA).Small is then incorporated into all RNA species. 5-FU is also converted into fluorodeoxyuridine and FdUMP, respectively. An alternative activation pathway 25,26 nuclear RNAs have key roles in pre-mRNA .In addition,rRNA,tRNA and snRNA all (FUDR)the splicing of which pre-mRNA is into convertinvolvesed into the thymidine FdUMP phosphorylase by the action catalysed of conversion thymidine of Kinasecontain the(TK) modified which base then pseudouridine, goes on to andinhibit 5-FU mature mRNA. 5-FU to fluorodeoxyuridine (FUDR), which is then has been shown to inhibit the post-transcriptional phosphorylated by thymidine kinase (TK) to FdUMP. Thymidylate synthase (TS), the only de-novo enzyme supplying thymidineconversion for of uridineDNA replication.to pseudouridine As in a these result, RNA mRNA Dihydropyrimidine dehydrogenase (DPD)-mediated conversion species27. POLYADENYLATION OF mRNA is inhibited at rela- FdUTP(Messenger RNA).RNA which that is subsequentlyof 5-FU to dihydrofluorouracil generated (DHFU) is isthe rate-limitingmisincorporated step of into DNA causing DNA damage 28 serves as a template for protein 5-FU catabolism in normal and tumour cells. Up to 80% of tively low 5-FU concentrations .These in vitro studies synthesis. administered 5-FU is broken down by DPD in the liver. indicate that 5-FU misincorporation can potentially

NATURE REVIEWS | CANCER VO LUME 3 | MAY2003 |21 331 © 2003 Nature Publishing Group Dihydropyrimidine dehydrogenase (DPYD) is the rate limiting enzyme catabolizing ~80% of the administered 5-FU in the liver (165). It converts 5-FU to Dihydrofluorouracil (DHFU), the first metabolite in a 3-step catabolic process. 5-FU is sometimes delivered as an oral prodrug as capecitabine (166) to overcome its catabolism by DPYD. Capecitabine is converted into 5′-deoxy-5-fluorouridine (5′DFUR) by the action of carboxyesterase and cytidine deaminase and then believed to be converted into FUDR or FUR in cancer cells as they express high TP and UP actively(167,168).

Modes of action and mechanisms of resistance to 5-FU

As 5-FU is a pyrimidine analogue, the mechanisms of action and the primary pathways by which the drug acts involve pathways in DNA metabolism, particularly pathways involved in pyrimidine salvage and DNA repair. As a result, the consequence of its action and molecular mechanism connecting the response to the drug not only affects pathways involved in nucleotide biosynthesis and DNA repair, but also extends to pathways such as cell cycle arrest, apoptosis, transcription and ribosome biosynthesis. Therefore, it is evident that molecular determinants of the response to the drug are very diverse involving many mechanisms. Hence, here we will only review those that are primarily important for the work subsequently described.

Thymidylate synthase (TS) inhibition by 5-FU

One of the major targets for 5-FU is TS. As discussed above, 5-FU blocks TS activity by forming an inactive substrate enzyme complex thus preventing the synthesis of thymidine

(dTTP) required for DNA synthesis. As expected, in-vitro and clinical studies have reported an increase in intra-tumoral TS expression as a mechanism of resistance to 5-

FU (169-173). TS enzymatic activity requires a folate precursor, specifically, 5,10-

Methylenetetrahydrofolate (CH2THF) or one of its polyglutamate forms. In the absence of

22 CH2THF or insufficient amount, TS cannot form a stable complex with FdUMP resulting in its poor inhibition (174,175). This mechanism is further complicated by the regulation of TS mRNA by TS enzyme. The TS enzyme is believed to bind TS mRNA thus inhibiting its translation. However, in a scenario where the TS enzyme is blocked, its mRNA is no longer occupied by the TS enzyme and thus the TS mRNA will be translated into new TS protein molecules (176,177). On the contrary, other studies have indicated that accumulation of inhibited TS enhances protein stability and turnover, in turn acting as a resistance mechanism. Polymorphisms in the promoter of TS are associated with resistance. The TS gene promoter usually has two (TSER*2) or three (TSER*3) 28 bp tandem repeat sequences. Patients who are homozygous for TSER*3 are thought to respond poorly to 5-FU based therapies compared to those who have TSER*2 (178-180).

In vitro biochemical analysis has shown that, TSER*3 leads to a 3-fold increase in TS mRNA compared to when TSER*2 is present (181). Finally, thymidine can be acquired from the extracellular environment through a salvage pathway by the action of TK further exacerbating the problem of resistance (182).

DNA damage, TS, and 5-FU

Downstream of TS targeting, the mechanisms leading to recognition of DNA damage and the responses are diverse. Soon after TS inhibition, there is depletion of the dTTP pools which in turn affects the overall pools of other nucleotides exacerbating the damage inflicted to the DNA such as through incorporation of dUTP in the DNA. This is not surprising as complex networks of nucleotide metabolizing enzymes are organized in pathways involving positive and negative feedback to closely coordinate DNA replication

23 and repair. dUTP incorporation in the DNA is recognized by many DNA glycosylases acting in BER. Uracil DNA glycosylase (UDG) excises uracil and 5-FU based adducts in the DNA. Thymidine Glycosylase (TDG) and SMUG1 are also thought to play a role in the removal of 5-FU adducts from DNA (183,184). MMR components recognizes 5-FU adducts. MutSa (MSH2:MSH6) recognizes 5-FU:G adducts (183). As one might expect loss of these enzymes significantly affects 5-FU sensitivity by altering DNA repair and cell death. It is now known in various in-vitro systems that loss of MMR components such as

MSH2 and MLH1 confer resistance to 5-FU (185-187). Thus, some clinical correlation has been made regarding MSI status and 5-FU sensitivity. However, other studies have found otherwise (188,189)

RNA damage by 5-FU

5-FU and its metabolites are extensively incorporated into RNA. rRNA species are important components of ribosomal function. 5-FU disrupts the pre-rRNA processing and maturation in turn affecting mature rRNA generation (190,191). This further greatly affects ribosomal biogenesis and translation. 5-FU affects post-translational modifications of tRNA such as pseudouridylation which is also a major modification seen in non-coding

RNA species (192-194). Further snoRNA/protein complexes are disrupted due to 5-FU meddling with snoRNA processing, assembly and activity (195). Recent reports suggests that formation of stress granules in response to RNA incorporation of 5-FU, but this effect was general to all other RNA incorporating drugs (196). Thus, it is clear that 5-FU can affect many aspects of RNA biology.

24 Dihydropyrimidine dehydrogenase (DPYD) and 5-FU sensitivity

One of the major mechanisms of 5-FU resistance is due to high expression of DPYD.

Several in-vitro and in-vivo observations have been made in relation to DPYD expression and 5-FU response. Studies in colorectal cancer cell lines have shown that high DPYD expression can decrease cytotoxicity. Higher DPYD expression in tumors from patients treated with 5-FU chemotherapy has shown poor response to the therapy (197-200).

Correlation of DPYD expression, mutation and polymorphism is explained further in detail in a later chapter and other sections.

Tumor suppressor p53 status and 5-FU response

P53 plays a central role in the 5-FU response. Many in-vitro and in-vivo studies have revealed that loss of p53 can confer resistance to the cytotoxicity of 5-FU (201-203).

However, clinical correlations have been less clear. In a recent study in breast cancer patient’s TS and p53 levels independently correlated with failure of 5-FU based therapy

(204). In colorectal cancer, p53 overexpression, a surrogate marker for p53 mutation has been linked to poor outcomes following 5-FU chemotherapy. Others have no found such correlation (205). This is particularly due to the lack of robustness in methods such as qPCR or immunohistochemistry in determing p53 status. In addition, overexpression of p53 determined by immunohistochemistry cannot be considered as mutant form in all instances (206). It has also been reported that the dUTPase gene which is responsible for removing UTP from DNA could further confer resistance to 5-FU (207,208).

25 Topoisomerase poisons, inhibitors and p53 pathway

Eukaryotic topoisomerase family consists of 6 topoisomerases that can be broadly

classified into 2 categories, type I and type 2. Type1 can be further classified into type IA

(consisting of Top3a TOP3b) and type IB (TOP1 and TOP1mt). The type 2 classification

consists of TOP2a and TOP2b. Type 1 cleaves only one strand of DNA whereas type 2

cleaves both strands of DNA. Here we will discuss only aspects of TOP2 poisons and

inhibitors, mechanism of action, and determinants of response and toxicity since they are

important for the work subsequently described. REVIEWS Mechanism and modes of action

T segment

G segment Merbarone

ATP ADP + Pi

Aclarubicin ICRF-187

Pi

Etoposide Figure 1 | Mechanisms of inhibiting topoisomerase II. Topoisomerase II (TOP2) can be inhibited at several different Figurepoints in the1- enzyme4: Mechanism reaction cycle, which of action can have differentof Topoisomerase biochemical and cellular II inhibitorsconsequences.Nature OneRevie simplews | Canc er mode of inhibition is to inhibit a step early in the enzyme reaction cycle. For example, competitive inhibitors of Source:ATP binding Nitiss, prevent JL., strand Nat passageRev Cancer. and do not 2009 generate May;9(5):338 enzyme-mediated-50 DNA damage. Although agents such as novobiocin and coumermycin (not shown) inhibit both prokaryotic and eukaryotic Top2 enzymes, they Topoisomeraseare either less potent activity and nonspecific can be (forinhibited example, at novobiocin) various steps or are poorly. Agents taken such up by asmammalian aclarubicin cells (for prevent TOP2 from example, coumermycin). Similar effects would occur with inhibitors that prevent the binding of TOP2 to DNA such as aclarubicin. Agents that prevent DNA cleavage by TOP2 such as merbarone would also be expected to act as simple bindingcatalytic to inhibitors. DNA in Although the first merbarone place. Other clearly agents prevents can DNA block cleavage ATP by hydrolysis TOP2 (REF. 127) and, merbarone thus prevent affects enzyme other -mediated targets besides TOP2. A second mode of inhibition is blocking the catalytic cycle after DNA is cleaved but before DNA DNAre-ligation. damage. This modeAgents of inhibition such as occurs Merbarone for most currentlycan prevent used TOP2 DNA targeting cleavage agents, without including inhibiting anthracyclines ATP. Overall, these and epipodophyllotoxins (such as etoposide), as well as for agents that target prokaryotic type II topoisomerases. typesThese of agents agents prevent prevent enzyme the turnover catalytic and are activity therefore of strongthe enzyme inhibitors and of catalytic thus are activity; classified however, as the TOP2 most inhibitors. The obvious effect of these inhibitors is the generation of high levels of TOP2–DNA covalent complexes. Therefore, these inhibitors generate DNA damage and interfere with many DNA metabolic events such as transcription and replication. As agents of this class convert TOP2 into an agent that induces cellular damage, they have been termed topoisomerase poisons. TOP2 can be inhibited after strand passage is completed but before ATP hydrolysis and dissociation of amino-terminal dimerization. Bisdioxopiperazines such as dexrazoxane (ICRF-187) inhibit both ATP 26 hydrolysis and maintain the TOP2 structure as a closed clamp74. As is the case with TOP2 poisons, bisdioxopiperazines inhibit TOP2 catalytic activity mainly by blocking enzyme turnover. Although these agents are frequently termed catalytic inhibitors, they leave TOP2 trapped on DNA and might interfere with DNA metabolism in a manner that might be analagous to TOP2 poisons. Nonetheless, as bisdioxopiperazines are specific for TOP2, they are the most 143 commonly used catalytic inhibitors of TOP2 in mammalian cells . Pi, inorganic phosphate. Figure is modified, with permission, from Nature REF. 173 (2002) Macmillan Publishers Ltd.

alleles of TOP2 have been identified, no clear pattern confer drug resistance36. For example, in previous that might lead to the identification of a drug-binding structures, the TOPRIM domain was located far from site has emerged using this approach. Many muta- the active site tyrosine. During cleavage and re-ligation, tions that reduce TOP2 catalytic activity can also lead these two elements must interact and therefore must be to drug resistance, complicating the interpretation of close to each other. Because TOP2 poisons act at the possible drug-binding sites. A large-scale screen point of cleavage and re-ligation, the relevant drug- of drug-resistant mutants of yeast Top2 failed to iden- binding pocket might be formed by residues that form tify any mutants with separable effects on etoposide both the TOPRIM and winged-helix domains (FIG. 2). versus mAMSA sensitivity34. However, many of the The hypothesis that the winged-helix and the TOPRIM TOPRIM domain mutants in TOP1 that led to camptothecin resist- domains come together to form a drug-binding pocket A conserved domain found in ance were not understood until a three-dimensional is supported by studies of fluoroquinolone action against topoisomerases, primases and structure of a drug–DNA–enzyme ternary com- prokaryotic Top2 enzymes. Fluoroquinolone-resistant other DNA metabolic enzymes. 35 The TOPRIM domain adopts a plex was determined . The newly described struc- mutations occur in both the TOPRIM and winged- Rossman fold, and is involved ture of TOP2 bound to DNA might be one step helix domains, and rarely occur in other parts of the in divalent cation binding. forward in helping to rationalize why specific mutants protein (reviewed in REF. 37). Although this localization

340 | MAY 2009 | VOLUME 9 www.nature.com/reviews/cancer other common agents are those that promote TOP2 catalytic complex (TOP2cc) but prevent re-ligation of the DNA strand. Examples of these are etoposide, Teniposide (both prevent, re-ligation of DNA strands) and doxorubicin (DNA intercalator), whereas quinolones CP-115,953, the ellipticines, azatoxins promote and stabilize (TOP2cc). These agents are collectively called TOP2 poisons as they increase DNA damage and DNA downstream processing (209).

Mechanisms of resistance and toxicity of TOP drugs

TOP2 poisons promote TOP2 catalytic complex (TOP2cc). The repair of the lesions involves recognition of the DNA-enzyme complex that subsequently creates double strand breaks. If the recognition of the DNA-enzyme complex fails then the damage can remain undetectable or becomes irrreversible. Since the detection requires any DNA processing enzyme to interact with the complex to create double strand breaks, the cytotoxicity of TOP2 poisons is cell cycle independent because both transcription and replication dependent enzymes can initiate this process. This is exactly what is seen for cytotoxicity caused by etoposide (a common clinically used TOP2 poison). As the end result of the recognition is double-strand breaks, it is clear that DNA repair enzymes involved in double-strand break repair such as components of NHEJ and HR are important determinants of response. Indeed, in a recent study the top three genes that provided resistance to doxorubicin (another clinically used TOP2 poison) were Tp53,

CHK2 and TOP2A (210). 5’-tyrosine phosphodiesterase (TTRAP/Tdp2) has been shown to provide resistance to TOP2 poisons by excision of the adduct (211). Rad27FEN1,

MRE11 and CtIP are other DNA repair enzymes whose expression determines sensitivity to TOP2 poisons (209,212). Helicase SNFL11 that broadly plays a role in DDR has been recently shown to correlate with resistance to TOP2 poisons (213). However, there are

27 no studies that clearly implicate a particular pathway in repairing the double-strand breaks downstream of a TOP2 poison and hence it has been difficult to assess biomarkers of response to TOP2 agents in the clinic. It is likely that combination of NHEJ and HR elements that have a vast number of mediators in these pathways would play a collective role in responding to TOP2 mediated damage (214,215). Despite this complexity experimental inhibitors of DNA-Pk (NU7026) have shown some efficacy in combination with TOP2 poisons (216).

TOP2 protein expression level itself can confer resistance to TOP2 poisons. Cells expressing high TOP2 levels are more sensitive to TOP2 poisons compared to cells with low expression of TOP2 (217,218). In-vitro and in-vivo RNAi based studies have shown that reduced TOP2 levels confer resistance to TOP2 poisons. On the contrary, individuals with TOP2 amplification seen in some breast cancer patients who have ERBB2 amplification, show a better response to anthracyclines. ERBB2 and TOP2 are located closely on chromosome 17 and tumors with co-amplification of these two genes are generally more sensitive to TOP2 poisons (219-221). One of major drawbacks of using

TOP2 poisons is the risk of secondary malignancies. Acute Myeloid Leukemia (AML) has been observed in patients treated with etoposide and teniposide and this is generally considered to be due to inhibition of TOP2b which results in chromosomal translocation at the MLL locus (222-224). A recent report suggests that TOP2-mediated DSB in AML in response to etoposide is brought about by chromosomal loop anchors independent of transcription (225). Cardiotoxicity is another result of use of anthracyclines (226).

Myelosuppression is another frequent toxic side effects of TOP2 poisons (227). Hence

28 understanding the underlying repair pathways along with key players are essential to limit toxicity.

Dihydropyrimidine Dehydrogenase (DPYD) gene regulation

Characterization of the gene and the protein

The DPYD gene is 840 kb in length and has been mapped to chromosome 1p22. The gene spans 23 exons and its coding mRNA is about ~3kb in length (228-230). Several internal transcripts have been observed but none which codes for the protein. They are annotated as antisense and intronic transcripts whose function is not yet known (231).

The protein structure was obtained from recombinant pig DPYD. The protein is a homodimer of 111 kd with each subunit having a NADH, FAD, FMN and 4 Fe-S clusters with a total of 1025 aa (232). The catalytic domain of the DPYD enzyme relates to the family of oxidoreductases. The closest family member sharing a similar domain to DPYD is DHODH, a key enzyme in the mitochondria that is involved in the rate-limiting step in commitment to de-novo pyrimidine synthesis and a recent therapeutic target in AML

(233).

Regulation of the DPYD gene

DPYD gene regulation is poorly understood. Early studies indicated that the expression of DPYD can be induced in response to phorbol-12-myristate-13-acetate (PMA) by binding of the AP-1 transcription factor (234). Subsequently it was shown that constitutive expression of this gene is regulated by the Sp1 and Sp3 transcription factors whose binding sites were identified in the promoter regions of the DPYD gene (235). Earlier

29 studies also focused on the circadian nature of DPYD gene expression. Initial observations were made in rat liver and further in humans (236,237). It was observed that the circadian nature of DPYD expression was disturbed in patients with gastrointestinal tumors when leukocytes from these patients were analyzed for DPYD expression (238).

Such results suggested that earlier 5-FU administration in patients has to be adjusted according the time of the day and thus some studies have also tested metronomic (based on time of day) 5-FU dosing in the clinic. At the gene expression level, post transcriptional regulation of DPYD has been seen through miRNA. Mir27a and mir27b are known to regulate DPYD in the murine liver (239). Mir494 and Mir 302b have been shown in

SW480 colorectal cancer and HepG2 liver cancer cell lines to inhibit DPYD expression

(240,241). Epigenetic regulation of the gene is also thought to occur. Particularly methylation of the CpG sequence near the promoter of the DPYD gene has been observed in RKO colorectal cancer cell lines (242). Aberrant methylation of the promoter has been shown to affect DPYD expression and cellular sensitivity to 5-FU (243,244). In- vivo, PMBC’s from patients who had a known history of toxicity have shown increased methylation of the DPYD promoter, compared to healthy volunteers (244). In contrast some have argued against this regulation by citing no correlation between mRNA expression and DNA methylation and little correlation with sensitivity to 5-FU-based chemotherapy in patients (245). It remains unclear whether methylation affects DPYD expression. It has been suggested that methylation can directly affect the gene silencing through epigenetic changes. For example, an increase in DPYD can be seen expression upon treatment with low dose of methylation inhibitors such as AzaC and trichostatin ,

30 whereas no change is observed following treatment with trichostatin alone (246). Hence further studies are needed to elucidate the role of DNA methylation in DPYD expression.

Clinical Pharmacogenetics of DPYD

The first clinical evidence of alterations in the DPYD gene came from observations in patients who had inborn errors in pyrimidine catabolism. Notably patients with severe neurological problems were found to have excess thymine, uracil and 5- hdroxymethyluracil in the urine. In 1984 Berger et al. made the first direct link of DPYD deficiency to excessive thymine-uraciluria (247). Since then DPYD deficiency has been associated with several neurological symptoms such as seizures, developmental delay, psychomotor retardation, microcephaly, muscular hypotonia and eye abnormalities.

However, the first observation that DPYD could be involved in 5-FU drug metabolism came in 1985 when patients with familial pyrimidinemia and pyrimidinuria experienced severe toxicity to 5-FU (248). Currently, there have been at least ~900 variants associated with 5-FU toxicity, out of which 200 occur in the DPYD coding region. As variants have been identified and continue to be identified only a few have been actionable in the clinic

(Table 1-1 summarizes actionable variants (249,250).

Table 1-1: List of DPYD genotypes and pharmacological dosing recommendation

Genotype Effect on Enzyme Phenotype 5-FU dosing activity recommendation DPYD*2A (IVS14 + Severe dose Heterozyote=~50 1G > A, c.1905 + 1G toxicity and very reduction > A, rs3918290) 50% reduced clearance Homozygote: of 5-FU Avoid 5-FU Effect: Result in chemotherapy Exon 14 skipping

31 (pyrimidine binding part) DPYD*13 (c.1679T Severe dose Heterozyote=~50 > G), rs55886062) toxicity and very reduction Ile560Se 50% reduced clearance Homozygote: of 5-FU Avoid 5-FU Effect: chemotherapy Destabilization of protein structure

c.2846A > T Toxicity observed. Heterozyote=50- (rs67376798) Reduced clearance 75% reduction of 5-FU and Homozygote: 25- 50% Effect: Inhibits co- 25% factor binding and structure

1129-5923C > G Toxicity observed. Heterozyote=75% (rs75017182, Reduced clearance reduction (haplotype B3, in 25% of 5-FU and Homozygote: 50% linkage with reduction c.1236G > A)

Effect: Aberrant mRNA splicing

Three decades of DPYD research have not established a robust methodology to asses for DPYD deficiency. This is partly due to a lack of clear genotype to phenotype link in assessing the risk associated with toxicity. Apart from the above-mentioned variants, whose upfront genotyping would benefit patients from severe grade 3 or 4 toxicity, other variants lack significant clinical validity and thus utility. Moreover, all variants including those mentioned above are seen with low frequencies in the population which further differs by race. Hence, it is important to conduct large genotype based studies to assess

32 more robustly the clinical importance of the genotype. A recent prospective study in breast cancer patients highlights that extensive whole DPYD exome sequencing can in fact provide benefits to patients as combined analysis of these variants increases the predictive power along with determining susceptibility to 5-FU toxicity (251). However, it is becoming apparent that phenotypic assessment for DPYD activity is a critical component to be supplemented with genotypic analysis. PBMC DPYD enzymatic activity is a well-established method for assessing the phenotype. Clinical validation of the PBMC

DPYD activity relationship to 5-FU toxicity has been corroborated. It also correlates with liver DPYD activity (252); however there is lack of prospective studies incorporating the

PBMC based dosing of 5-FU. The method is also expensive and time consuming and not suitable for routine clinical testing because it also involves measuring radioactive metabolites (249). Measuring the plasma pretreatment DHU/U ratio has been utilized as an alternative. Several studies have found a relationship between marked susceptibility to 5-FU toxicity and decreased DHU/U ratio (253-258). However, large prospective studies have given equivocal results. In fact, a French study conducted in 252 patients showed that the uracil concentration as opposed to the DHU/U ratio correlated well with

5-FU clearance (258). Moreover, In 550 patients, the pretreatment uracil concentration was found to be an independent predictor of severe and fatal toxicity to 5-FU (259).

Therefore, the clinical utility has to be established for DHU/U vs uracil measurements and the better of the two methods needs to be considered along with other factors such as circadian rhythm, metabolic variations arising from food source for predicting DPYD deficiency.

33 Finally, two indirect methods have been also used such as measuring a 13C-uracil breath test and uracil test dose. The clinical utility of the former has been studied in a small

13 setting in which 6 out of 13 patients had an increase in exhaled CO2 produced upon degradation of 13C-uracil (260). Whereas the latter has shown correlations with DPYD deficiency but not with 5-FU clearance or toxicity (261). Therefore, in the future, overall clinical diagnosis of DPYD deficiency will likely be a combination of genotypic and phenotypic assessment with proven clinical validity and utility that will likely guide the dosing of 5-FU and prevent severe and sometimes fatal toxicity.

Check-point kinase 2 (Chk2) regulation of p53

Chk2, the human homologue of yeast Rad53 was initially identified as a mediator of DNA damage and replication block (262). Since its initial discovery, Chk2 kinase has been established to play a central role in genomic integrity by coupling DDR sensing to a variety of processes such as DNA repair, cell cycle arrest, apoptosis, mitochondrial genome integrity, telomere maintenance and mitosis. Chk2 protein has 3 domains including a

Serine-glutamine and threonine-glutamine pair, an SQ/ST domain SCD, a Fork-head associated domain (FHA), canonical kinase domain and T loop (263). Chk2 is activated by phosphorylation of its SCD domain by PI3K family members ATM and ATR (264,265).

34 REVIEWS REVIEWS

Double-strand breaks Alkylating agents Replication stress a founder mutations present at varying frequencies in RPA S19/ 33/ 35 T68 S260 T383 T387 S456 S516 normal healthy individuals in different populations4,5 RAD1 (TABLE 1) ATRIP HUS1 . Therefore 1100delC and I157T MRE1cannot1 pre- DNA-damage NBS1 RAD9 sensordisposes to highly penetrant LFS. Furthermore, analysis RAD17 SQ/TQ FHA Kinase domain NLS of CHEK2 in LFS families has revealed onlyRAD5 a few0 rare RFC 11969 115 175 226 486 5155522 43 mutations, and these have not been demonstrated to T loop 8 co-segregate with cancer in families with LFSAT. M I157T 1100delC ATR b pT68 pT68 CHEK2 as a multiorgan cancer susceptibility gene. S317 SQ/TQ DNAFurther-damage studies on the prevalence of T61100delC8 and P S345 N N signal transducers P P Y72 I157T founder mutations in cancer families andCHK2 other CHK1 P cancer cases compared with normal healthyP indi- P FHA 50–70Å T383 viduals indicate with statistical significanceT387 that these S296 P92 S210 P92 mutations are in fact moderately penetrant cancer S367 susceptibility mutations, increasing the risk of devel- P 61–69 MDM2 MDM4 oping breast and prostate cancer . I157T has also S123 D207 S504 been demonstrated to increase the risks of develop- P S178 S216 S988 S117 S364 P P Kinase ing ovarian, colorectal, kidney,P thyroidP and bladderP S20 S366 CDC25A P S292 BRCA1 PML E2F1 P P CDC25C DNA-damage 68,70–73 T387 signalcancers effecto andrs leukaemias . However, other studies p53 P have indicated that 1100delC and I157T probably act in synergy with other genes or factors to cause 61–63,74,75 T-loop exchange cancer . For example, 1100delC is more common in patients with breast cancer who have a first-degree Senescence S-phase arrest Figure 2 | Structure and activation of human CHK2 protein. a |Na Thetur 543e Re amino-acidviews | Canc er relative with cancer orDNA a family repair historyApoptosis of this disease, Apoptosis G1 arrest G2 arrest G2 arrest CHK2 protein is characterized by an amino-terminal domain rich in serine or threonine and the mean age of cancer incidence in 1100delC car- residues followedFigure by glutamine 1- (SQ/TQ 5: Chk2motif – the consensusstructure site for phosphoinositide- and signalingFigure 1 | Schematic overview of the DNA-damage response signalling pathway. Cells are constantly subjected riers is lower than that in non-carriers in case–control Nature Reviews | Cancer kinase-related kinases (PIKKs) such as ataxia telangiectasia mutated (ATM) and ATM- tostudies DNA 62,65damage,74,75. The and carriers DNA breaks of 1100delC that arise have either also endogenously been during normal cellular processes such as genome and Rad3-related (ATR)), a forkhead-associated (FHA) domain, which typically binds replicationfound to have or exogenously an increased by exposure risk for bilateralto genotoxic breast agents, such as UV radiation, radiotherapeutics and phosphothreonine residues, and a carboxy-terminal kinase domain that contains the chemotherapeutics. Damage to the DNA triggers the recruitment of specific damage sensor protein complexes. On cancer62,76–79. Interestingly, the highest percentage of activation T loop followed by a nuclear localisation signal (NLS). On the top of the figure the one hand, the MRN (MRE11–RAD50–NBS1) complex is required for the activation of ataxia telangiectasia the main residues that are phosphorylated to regulate the function of CHK2 are shown mutatedcarriers (ATM)for these in response mutations to double-strand is breast cancer breaks patients (DSBs). On the other hand the ATM- and Rad3-related (ATR)- in bold, along with other residues that are phosphorylated in response to DNA damage. interactingwho develop protein contralateral (ATRIP) complex breast canceris recruited after to having sites of single-strand breaks and activates ATR. Specifically, the Source Antony L et al., 2001, Nat Rev Cancer. 2007 Dec;7(12):925-36 76 The main mutations in CHK2 in human tumours are shown on the bottom. Following CHK2received pathway radiation is activated therapy byfor DSBs their thatfirst occurbreast either cancer directly. by exposure to ionizing radiation (and radiomimetic treatment with ionizing radiation (IR), ATM phosphorylates CHK2 on T68 agents)Additionally, or indirectly 1100delC by topoisomerase carriers have IIalso inhibitors. been shown It can also be activated by replication-mediated DSBs as a result (REFS 126,132,133)Left,. Although(A) Different the related ATRdomains kinase can alsoof phosphorylateChk2 kinase CHK2 enzymeofto base-pair have with an increasedexcision phosphorylation generated risk of breast by alkylating cancer sites after agents expoas or single-strandindicated.- breaks (B) caused 3D Structureby topoisomerase I inhibition. in vitro, little is known about its role in activating CHK2 in the cell126,134. Phosphorylation Dependingsure to non-mammographic on the type of stress, X-rays activated80. These CHK1 data and indi CHK2- can phosphorylate a number of overlapping or distinct on T68 and subsequent full activation of CHK2 was recently shown to require priming downstreamcate that environmental effectors, which DNA-damaging results in the activation factors such of DNA repair, cell-cycle arrest, senescence or apoptosis. phosphorylationof the on adjacent Chk2 residues kinase, by Polo-like domains kinase 3 (PLK3) and and theirthe dual- interaction.Symbolsas IR increaseRight, in green the areChk2 risk CHK2-specific of cancer signaling in substrates,carriers. pathway symbols in activationblue are CHK1-specific following substrates DNA and symbols in red are specificity tyrosine and serine/threoninekinase TTK/hMPS1. Additionally TTK appears sharedAll substrates.these findings Downstream are consistent CHK1-specific with the substrates hypoth- are not represented here. to phosphorylate T68 (REFS 135,136). Phosphorylation of T68 promotes the binding of esis that CHEK2 is a multiorgan cancer susceptibility the N-terminaldamage SQ/TQ-rich stress. cluster of one Chk2 CHK2 molecule is mainly with the activatedFHA domain of downstream of DNA double stranded break recognition by DNA another CHK2 molecule. This dimerization is essential for full CHK2 activation by trans- gene that acts in synergy with other genes or factors to 54,55 autophosphorylation of T383 and T387 in the T loop of the kinasePenetrance domain137–140. Both causeTaken cancer. together, To support these data this clearlyhypothesis, show a that recent a number report familial cancer syndrome . About 75% of classical type 2A proteindamage phosphatase sensors (PP2A) and the like PP2C phosphatase, MRN The (MRÈ11/RAD50/NBS1) PPM1D/WIP1,likelihood a given bind gene to ofshows parallel, that but the complex notrisk necessarily from CHEK2 and exclusive, mutations further pathways in prostate activates form LFS families many have downstream a heterozygous germline mutation and dephosphorylate T68 on CHK2, thus providing a recoverypresent mechanism in the germ from line DNA of an theand DNA-damage colon cancer mayresponse, be restricted which is to essential individuals in cancer with in the gene encoding p53, TP53 (REFS 56,57). These damage141–143. Although T68 phosphorylation is necessary for organisminitial activation will result inof disease. CHK2, prevention,a particular and genotype that CHK2 of CDKN1 is an importantB (the gene player encoding in this individuals have an 85% risk of developing cancer it is not required for its subsequent kinase activity either as a dimer or a monomer once (REF. 81) Loss of heterozygosity response.the CDK Indeed,inhibitor, genetic p27 studies). inThese the germgenes, line however, and in over their lifetime and a 42% risk between birth and effectors in138,144 various cellular processes. the T loop is phosphorylated . Moreover, a number of otherClassical residues tumour-suppressor on CHK2 are tumoursdo not include indicate BRCA1 that CHK2 and BRCA2has a role (R EFSin tumorigenesis. 61–63) possi- the age of 16 years58. Tumour initiation in some of phosphorylated following IR, although their role in regulatinggenes CHK2 are activity considered is presently to be bly because the biological mechanisms underlying the loss of heterozygosity 140,145–147 these families is associated with a unclear . However, phosphorylation of S456 was recentlyrecessive. shown Thus, to regulate cells that the Theincreased role ofrisk CHEK2 of breast in cancer cancer in CHEK2 mutation carri- stability of CHK2 in response to DNA damage148. b | Schematic model of a full-length (LOH), a phenomenon first described by Knudson for contain one normal and one ers are already subverted in BRCA1 or BRCA2 mutation CHK2 dimer114,150. This model shows the atypical dimeric arrangementmutated form for ofCHK2, a tumour- The cloning in 1998 of human CHEK2 and the identifi- inherited classical tumour suppressor genes, which in allowing exchange of the T loops and thus providing a mechanismsuppressor by which gene in the germ cationcarriers, of itsconsistent link to the with DNA-damage proteins participating response, led in tothe a this case results in a total loss of function of TP53 61 dimerization-drivenHowever, activation of its CHK2 catalytic by trans-phosphorylation activationline are occurs functionally114. Part isnormal. b broughtis searchsame pathway for by CHEK2 autophosphorylation (FIG. mutations 1). in both somatic and heredi of- (RtheEF. 59) T. In loop other families, (366 it- 406has been hypothesized modified, with permission, from REF. 114 (2006) Nature PublishingThe condition Group. that results in tarySubsequent human cancers. studies Since have then, identified there has beenthree an other ava- that the germline TP53 mutation is either dominant the loss of the remaining lanchefounder of publicationsmutations: S428F, on this presenttopic from in thewhich Ashkenazi it is now negative or has a gain of function such that there is normal allele is known as loss Jewish population, has been shown to confer an of heterozygosity. clear that CHEK2 is indeed a cancer susceptibility gene, no selective pressure in these tumours for loss of the aa) which resides to the C-terminal side within the kinase82 domain. Finally, it has the The identification of three different CHEK2 mutations, butincreased not a tumour risk of suppressor breast cancer gene in; theIV S2classical + 1G>A sense. has remaining wild-type TP53 allele57,60. I157T, 1100delC and 1422delT,Dominant in one or negative more individ- been reported to increase the risk of breast, prostate However, a subset of LFS families do not harbour 6,66,68,83 uals from each of three separateA dominant-negative LFS families led mutation to the CHEK2and thyroid and cancer LFS. The first; and indication 5395del has of been the roleshown of germline TP53 mutations. Given the potential role of occurs when the mutated gene 5,84 nuclear localizationproposal that CHEK2 issignal one of the alternative(NLS) genes domain pre- CHEK2to increase for in itscancer the risksubcellular came of breast from and a study prostate localization that cancerreported CHK2 (Fig and ureCHK1 in1 -the5) mammalian. G2 checkpoint, disposing to LFS53. However, itproduct was subsequently adversely affects found the (TABLE 1). Full analysis of CHEK2 in families with normal, wild-type gene the presence of germline mutations in CHEK2 in a study was undertaken to determine whether muta- that 1422delT is a polymorphismproduct in within a non-processed the same cell. familiesprostate withcancer, LFS prostate53. LFS cancer is a rare cases highly unselected penetrant for tions were present in these genes in LFS families53. pseudogene3 and that both I157T and 1100delC are a family history and families with breast cancer has

928 | DECEMBER 2007 | VOLUME 7 www.nature.com/reviews/cancer NATURE REVUponIEWS | CANCER DNA damage, cells recruit specific damage VOsensorLUME 7 | DE CproteinEMBER 2007 | 929complexes such as the

MRN (MRE11–RAD50–NBS1) which activate ATM in response to double-strand breaks

(DSBs). Likewise (ATRIP) complex is recruited to sites of single-strand breaks which

inurn activates ATR. Chk2 activation by carried out by its phosphorylation ATM on T68

35 (266). This induces Chk2 dimerization through SCD domains which then sets a cascade of autophosphorylations on S269, T432, T383 and T387 which in turn leads to disassociation into active monomers (267). The disassociated monomers quickly are dephosphorylated by phosphatases. Proteins such as Msh2 which interact with Chk2 and

PLK3, the later of which phosphorylates Chk2 on S62, S73 are known to enhance its activation by ATM (268,269). Activation of Chk2 primarily occurs in response to DSB’s.

Chk2 can also be activated by ATR when single stranded breaks get converted to DBS’s.

Downstream events mediated by Chk2 are very diverse with well-known roles in DNA repair, apoptosis, cell cycle G1arrest and senescence (270). (Fig 1-5) The Inactive Chk2 is known to reside in Promyelocytic Leukemia protein (PML-NB) bodies. A fraction of activated Chk2 in PML-NB phosphorylates PML and promotes p53-dependent apoptosis

(271). On the contrary Chk2 is maintained in an inactivated state by many phosphatases such as protein phosphatase 2A/1 (PP2A and PP1), protein phosphatase 1D (WIP1), and

Polo-like-kinase-1(PLK-1) (272-275). Chk2 degradation has also been proposed.

Ubiquitination by RING-H2 protein (PIRH2) on phosphorylated serine S456 Chk2 is degraded by the proteasome (276).

Chk2 and DNA Damage Repair (DDR)

The earliest known function of Chk2 was in DNA repair. Chk2 phosphorylates BRCA1 and BRCA2 proteins promoting HR repair in response to DNA damage and in turn suppresses NHEJ (277,278). Chk2 phosphorylates FOXM1, which increases BRCA1 stability and XRCC1 promoting HR and BER (279). Phosphorylation of BRCA1 by Chk2 helps in the recruitment of Rad51 that suppresses NHEJ through Mre11 (280). BRCA2

36 phosphorylation by Chk2 helps in the dissociation of Rad51-BRCA2 complex. Rad51 further localizes and helps in strand invasion and exchange (278). Chk2 cooperates with

ATM in phosphorylating a common substrate KRAB-binding protein KAP-1. KAP-1 is first phosphorylated by ATM on S824, and subsequently by Chk2 on S473 (281). This action helps KAP-1 to dissociate with its partners CHD3 and HP-1b, nucleosome remodeler protein and heterochromatin packaging factor respectively (282,283); the combined action of which results in increased access to repair proteins.

Chk2 and p53-dependent cell cycle arrest

The role of Chk2 in cell cycle function is well-known playing a role in G1/S and G2/M checkpoints. Chk2 phosphorylates Cdc25A on residues S123, S78, S292, which is subsequently degraded by the proteasome (284,285). This prevents dephosphorylation of Cdk2, a key regulator for G1/S- and S-phase progression, causing cell cycle arrest.

Chk2 in cooperation with ATM activates p53 by phosphorylating p53 on S20 (286,287), which can in turn transcriptionally activate p21 causing G1/S arrest. Chk2 phosphorylation increases p53 stabilization by preventing its degradation by MDM2 signaling. Further, it can also enhance p53 transcriptional function by promoting its interaction with p300 histone deacetylase (288). It also phosphorylates p53 on other residues such as S366 and T387 which regulate acetylation of p53 and thus its transcriptional function (289).

Other ways Chk2 can bring about G1/S arrest is by phosphorylating pRb at S612 enhancing the stability of the pRB/E2F-1 complex preventing E2F-1 transcription of cell cycle target genes (290). In addition, it can phosphorylate serine/threonine protein kinase

LATS2 which plays a role in G1/S arrest (291).

37 In the G2/M phase of cell cycle, Chk2 phosphorylates Cdc25c on S216 which increases its association with 14-3-3s and promotes its nuclear export (292,293) delaying G2/M progression. Chk2 can phosphorylate CHE-1 that increases p53 transcriptional function and maintains G2/M arrest (294). Serine/threonine kinase receptor-associated protein

(STRAP) phosphorylation by Chk2 in the nucleus increases its stability thus augmenting p53-mediated G2/M arrest (295). Finally, it can also stabilize the dual specificity protein kinase TTK1/hMps1 promoting G2/M arrest (296) .

Chk2 and p53-dependent apoptosis

Chk2 occupies an important role in controlling p53-dependent apoptosis. Phosphorylation of p53 by ATM prevents its degradation, and the combined action of ATM and Chk2 helps to further stabilize p53 further by inhibiting MDMX/MDM4 (295). Persistent DNA damage can lead to increased stabilization of p53 and apoptosis. DNA-PK can also activate Chk2 and bring about p53-dependent apoptosis (297,298). Chk2 can bring about apoptosis by a p53-independent mechanism through phosphorylation and stabilization of E2F-1 resulting in increased transcription of E2F-1 target genes involved in apoptosis (299).

BIRC5, also known as survivin, is an anti-apoptotic gene whose release from mitochondria is mediated by Chk2 promtes mitochondrial apoptosis (300). In addition,

Chk2 phosphorylates PML protein in response to radiation thus promoting pro-apoptotic pathways bought about by PML-NB’s (271,301). Recently Chk2 has been reported to inhibit HuR, an mRNA stabilizing protein by phosphorylating it on S88, S100 and T188.

Chk2 prevents HuR binding to SIRT1 mRNA thus enhancing SIRT1 mRNA decay and cell death (302).

38 Role of Chk2 in mitosis

Apart from role in G1/S, intra S and G2/M, Chk2 can also act in mitosis. Chk2 interacts with PLK1 and PLK3 to maintain centromere stability and thus genomic stability

(303,304). Chk2 and BRCA1 are thought to co-localize to the centrosome in the absence of DNA damage (305). DNA-PK’s are also thought to phosphorylate Chk2 on the centrosome, mid-bodies and kinetochores stabilizing the centrosome and spindle formation in response to DNA damage (306). Overall while the role of Chk2 in mitosis is increasing, more studies will be needed to delineate the molecular mechanismby which

Chk2 can prevent mitotic catastrophe.

Therapeutic targeting of Chk2

Chk2 targeting is a much-debated topic. Clearly, Chk2 acts as a guardian of DNA damage by orchestrating different repair processes and if the damage persists it can promote arrest, apoptosis and even senescence. It is involved in all phases of the cell cycle ultimately responding to different stimuli that result in some sort of a DNA lesion.

Therefore, the question becomes whether to inhibit Chk2 or activate it. It remains unclear under what context targeting Chk2 would provide benefit to therapy. Inhibition of Chk2 in- vitro can add benefit to DNA damaging agents such as IR. In HEK-293 cells, inhibition of

Chk2 prevents activation of a cell cycle checkpoint and promotes apoptosis (307).

Overexpressing dominant negative Chk2-DN prevents survivin release that inhibits apoptosis (300). Further, it was found to enhance doxorubicin mediated killing of HCT-

116 colorectal cancer in a xenograft model. In support of this VRX046667 a selective inhibitor of Chk2 showed anti-proliferative effects in long-term culture of cells with this

39 inhibitor (308,309). Chk2 inhibition has proven great feasibility in radioprotection. Chk2 inhibition in normal cells suppresses apoptosis and promotes cell survival. Some of the selective inhibitors of Chk2 such as PV1019, CCT241533, 2-(4-(4-Chloro- phenoxy)phenyl)-1H-benzo[d]-imidazole-5-carboxamide (Chk2 inhibitor II), provide radioprotection (310). VRX046667 was shown to protect murine T-lymphocytes from radiation induced toxicity. However, Chk2 inhibition in the context of chemotherapy has been disappointing. For example, VRX046667 failed to show any cell killing in combination with cisplatin or doxorubicin (309). Chk2 inhibitor II also failed to sensitize and on the contrary decreased cell killing by oxaliplatin (311). siRNA knockdown of Chk2 also failed to sensitize cells to gemcitabine and 5-FdUR. In response to topoisomerase inhibitors Chk2 did not show any sensitivity to the Top1 poison topotecan or its prodrug

SN38. Just one study has shown no benefit with the TOP2 poison etoposide or doxorubicin in Hela or HT29 cells (312).

Specific Chk2 targeting has been slow because most of the inhibitors developed were based on in-vitro catalytic activity which invariably also targets Chk1. This is reflected in the fact that the majority of inhibitors in the clinic so far are either Chk1 inhibitors or dual inhibitors of Chk1/Chk2. Hence, together it highlights the fact that the context for therapeutic Chk2 inhibition has not been fully elucidated. There seems to be a lack of consensus among various studies partly due to the use of multiple cell lines and different

DNA damaging agents to study Chk2 function. It is becoming quite clear that the nature and context of DNA damage are of particular importance given the fact that Chk2 can activate different repair and cellular processes thus having different outcomes. This may

40 be further complicated by the genomic background and heterogeneity observed in different tumor types. Therefore, careful evaluation of the context, tumor type, nature of

DNA damaging agent and genetic background, such as p53 mutation status, is warranted to fully elucidate and exploit Chk2 targeting for cancer therapy.

Since Chk2 offers radio protection, it is less clear if it does so, following chemotherapy.

As reviewed above, chemotherapy consists of different classes of molecules with different mechanisms of action. Since Chk2 is involved in various processes (discussed above chapter), its signaling outcome is highly context-dependent and hence this thesis will focus on identifying a specific context for Chk2 targeting during cancer therapy for limiting toxicity.

SCOPE OF THE THESIS

The focus of the thesis is to address the role of p53 signaling in response to chemotherapeutics 5- FU and Etoposide in order to modulate the toxicity of the drugs. We have addressed this overall goal in 3 parts.

1) The first part, that is chapter 2 addresses the control of pyrimidine metabolism

through regulation of the gene DPYD by p53. Here we discuss studies conducted

to address transcriptional repression of DPYD expression by p53 in response to 5-

FU highlighting some of the mechanisms discussed in chapter 1. The study also

focuses on functional effects of this repression on 5-FU metabolism by analyzing

plasma levels of 5-FU as a measure of DPYD activity and its consequences onto

41 tumor growth. Further, the mechanistic basis of this interaction is studied in the

context of DNA damage and TS inhibition.

2). Chapter 2 focuses on classical ATM-Chk2-p53 signaling pathway modulation in the

context of TOP2 poisons. Here we describe a role of Chk2 in mediating toxicity

specifically to TOP2 poison following its activation by ATM. Further we explore the

genetic and pharmacological targeting of Chk2 with goal of identifying inhibitors

suitable for Chk2 targeting in the clinic

3) Chapter 3 extends observations in Chapter 1, in exploring the role of mutant p53

in controlling the pyrimidine catabolic gene DPYD. The goal of this chapter is to

identify specific p53 mutations with differential DPYD expression in order to further

modulate its expression for potentially increasing the efficacy to the drug 5-FU.

42 CHAPTER 2

P53 represses pyrimidine catabolic gene dihydropyrimidine dehydrogenase (DPYD) expression in response to thymidylate synthase (TS) targeting.

This chapter has been published (Gokare et al., Sci Rep. 2017 Aug 29;7(1):9711)

Introduction

Dihydropyrimidine dehydrogenase (DPYD) is the initial rate-limiting enzyme of the pyrimidine catabolic pathway that catalyzes the reduction of the nucleotide bases uracil and thymine(313). Although recent findings have implicated DPYD in epithelial-to- mesenchymal transition (EMT) of breast cancer cells(314) most interest in DPYD has stemmed from its role in limiting the bioavailability of the chemotherapeutic anti- metabolite fluorouracil (5-FU) that exerts its therapeutic activity at least partly through anabolic uptake(162,315-318). Hepatic catabolism of 5-FU by DPYD is primarily responsible for the rate-limiting catabolic conversion to 5-fluoro-5,6-dihydrouracil (5-

FUH2) which is a substrate in additional enzymatic steps resulting in the urinary excretion of fluorinated forms of beta-alanine and urea. Variable expression of DPYD and genetic polymorphisms in the DPYD gene have been linked to a high intra- and inter-patient variability in the plasma levels of 5-FU with associated toxicity and cancer drug resistance(319-321). Subsequently, different pharmacological DPYD inhibitors such as gimeracil(322) and eniluracil(323-325) have been added to oral 5-FU formulations that currently are either approved for clinical practice or undergoing clinical trials in order to improve 5-FU bioavailability(326).

43 The tumor suppressor gene TP53 that encodes for the transcription factor p53 is mutated and/or inactivated in the majority of human cancer. Canonical p53 signaling involves induced transcription of genes involved in cell cycle arrest, DNA damage repair and programmed forms of cell death. However, it is becoming increasingly clear that p53 also modulates additional cellular processes such as metabolic pathways that can have a profound impact on cancer cell invasion and treatment refractoriness. Moreover, p53’s role as a transcriptional repressor may contribute to the biological phenotypes of its tumor suppressive action. In the context of 5-FU-based therapies, TP53 mutation status has been correlated with treatment response and survival. Colorectal cancer patients with mutant TP53 have a shorter overall survival as compared to patients with wild-type

TP53(201-203). Indeed, colorectal cancer cells deficient for TP53 and subjected to treatment with 5-FU in pre-clinical experiments are protected from cell death(116).

Interestingly, it is less clear if TP53 status dictates the response to other DNA-damaging chemotherapies used for the treatment of colorectal cancer such as oxaliplatin and irinotecan that could indicate intrinsic differences in the p53 response between these chemotherapeutics(327,328).

To to investigate how p53 selectively modulates the cellular response to 5-FU we performed an in-silico screen for p53 DNA-binding sites (p53BS) in the proximity of or within genes involved in nucleotide metabolism. By combining this analysis with chromatin immunoprecipitation (ChIP), expression analysis in-vitro and in-vivo we show that the expression of DPYD is negatively regulated by p53 in the context of inhibition of thymidylate synthase (TS). We show that this observation is correlated with increased

44 relative levels of the 5-FUH2 catabolite and reduced tumor growth delay in mice lacking

TP53 in their livers following treatment with 5-FU. The data also indicate a role of the

TP53 codon R72P polymorphism in DPYD expression. Together, our current study provides novel insight into the role of p53 as a repressor of the key rate-limiting enzyme

DPYD and indicates that p53 may function as a negative regulator of pyrimidine catabolism. Our results have implications for the toxicity of 5-FU as well as its efficacy in the treatment of cancer.

Material and Methods

Cell culture and treatments

Authenticated cell lines were obtained from ATCC between the year 2011-12 and were within 20 passages when used for the experiments. All cell lines were routinely tested for

Mycoplasma for every 3 months by DAPI staining and PCR. Every 6 months STR profiling was performed for verification of cell line origin. HCT-116-p53WT and HCT-116-p53-/- were cultured in McCoy’s 5A media, H460 were cultured in RPMI1640. A549, U87MG,

HT-1080 and R72P MEF were cultured in DMEM. NHF cells were purchased from Coriell

Institute for Medical Research (Camden, NJ, USA) and were grown in DMEM (15%PBS and non-essential amino acids). For the assessment of protein and mRNA expression, 5 x 105 – 8 x 105 cells were plated in 6-well plates (Corning) and treated for up to 24-hrs.

Mice and treatments

Six to eight-week old male C57BL/6J (wild-type), B6.129P2-Trp53tm1Brn/J (p53 loxP/loxP) and 129-Trp53tm1Tyj/J (p53-/-) and B6.Cg-Tg(Alb-cre)21Mgn/J (AlbCre) mice were

45 purchased from Jackson Laboratory (Jackson Laboratory, ME). All mice were housed in a controlled environment with regard to light, temperature and humidity. Mice were euthanized through cervical dislocation following ketamine/xylazine anesthesia. All animal work was conducted based on the procedure and guidelines approved by Penn

State University Hershey Medical Center Institutional Animal Care and Use Committee.

5-FU (TEVA pharmaceutical) was injected IV at 100 mg/kg bolus for 4 hrs for analyzing p53 binding and for 6 hrs for DPYD western blotting experiments using frozen liver tissues. For plasma analysis of 5-FU and 5-FUH2, mice were given 150 mg/kg IV 5-FU or Vehicle (PBS) for 6 hrs and then a second dose of bolus 5-FU 150 mg/kg IV was administered (since the plasma half-life of 5-FU is short ~20min) to all the mice for 30 min after which about 300-500 µl of blood were immediately collected by cardiac puncture and stored in -80oC for HPLC/MS analysis.

Mouse syngeneic colorectal cancer model

A total of 1x106 p53-Ras-Myc (329) cells were injected onto the left and right flank of

C57Bl/6J Albcre; mT/mg; p53Δ/mice and Albcre; mT/mg; p53Δ/Δ (N = 3-5/group) in 1:1 ratio of BD Matrigel and PBS. Tumors were allowed to grow till an average volume of ~40 mm3 following which a weekly dose of 5-FU (100 mg/kg) was administered IV for 6 weeks.

Tumors were measured and volume calculated by the formula (W2XL)/2. Histology sections from tumors were manually counted for apoptotic and mitotic foci in a blinded manner. Body weights of the mice were also monitored.

46 Chromatin immunoprecipitation analysis (ChIP Analysis)

In-vivo ChIP analysis was conducted by EZ-Magna ChIP A/G (Catalog # 17-10086) as per the manufacturer’s instructions. Briefly, about 50 mg of liver tissue was used and homogenized using a dounce homogenizer. The homogenized tissue was sonicated using a Misonix 3000 sonicator for approximately 9 min (on 30 sec, off 30 sec) using a microtip. The fragments were between (200-800 bp). About 1% of the chromatin was used as the input. For immunoprecipitation anti-p53 (FL393) (Cat# sc-6243 (Santa cruz)) and

IgG (Santa Cruz) were used. PCR primers were p53DBS (Forward primer –

AACACTCCTTCGTTGCTCGT: Reverse primerTGAGGGACATCTGGGTTCTT) p21

(Forward primer -CCTTTCTATCAGCCCCAGAGGATACC: Reverse primer -

GACCCCAAAATGACAAAGTGACAA). Primers for other regions are listed in supplemental Table 1. Fold enrichment of in-vivo ChIP was calculated by using Bio-Rad

CFX96 Touch under qPCR Conditions were [950C for 30 sec; 550 C for 45 sec; 720C for

45 sec] x 39 cycles. In-vitro ChIP analyses in HCT-116 p53+/+ cells were conducted according to Simple ChIP protocol according to the manufacturer’s instructions (Cell signaling Cat # 9003). H3K9Ac (Cell signaling, Rabbit mAb #9649) H3K4me3 (Abcam

Rabbit polyclonal #8580), H3K27me3 (Abcam Mouse mAb#6002) was used for pull- down. qPCR primers for DPYD promoter were Primer 1(Forward primer-

GAAGGGAGGGAGGGAGTAGA; Reverse- AGTGCAAGTTGTTGCTTGGA). Primer

2(Forward primer-CTCGAGTCTGCCAGTGACAA; Reverse primer-

CCCTAGTCTGCCTGTTTTCG) Primer 3 (Forward primer-

CGAGTCGAAAACAGGCAGAC; Reverse primer-TCTACTCCCTCCCTCCCTTC).

47 LC/MS/MS Method for 5-Fluorouracil (5-FU) and 5-fluoro-5,6-dihydrouracil (5-FUH2)

One (1) mg/mL of stock solution was prepared in water for 5-FU, 5-FUH2 and in 60% acetonitrile (ACN) in water for 5-chlorouracil (5-CU). Intermediate stock solutions of 10 or 100 µg/mL of prepared for 5-FU and 5-CU by dilution of the 1 mg/mL stock solutions with water. Intermediate combined working standard solutions (5-FU/5-FUH2) of 0.1/5,

0.2/10, 1/50, 2/100, 6/300, 10/500 in µg/mL were prepared from the stock and intermediate stock solutions in water. 5-CU solution was prepared as 2000 ng/mL in water as internal standard (IS). The mouse serum 5-FU/5-FUH2 working calibration standard was prepared by spiking 10 µL of the corresponding 5-FU/5-FUH2 intermediate working solutions (0.1/5, 0.2/10, 1/50, 2/100, 6/300, 10/500 in µg/mL) to 100 µL of blank mouse serum to generate concentrations of 5-FU/5-FUH2 at 0.01/0.5, 0.02/1.0, 0.1/5.0, 0.2/10.0,

0.6/30.0, 1.0/50.0 µg/mL. Spiking 20 µL of the 10/500 standard (5-FU/5-FUH2) to provide the calibration standard at 2.0/100.0 µg/mL. 20 µL of 5-CU (2000 ng/mL) was added to each calibration standard as internal standard. 20 µL of 5-CU (2000 ng/mL) was added to every 100 µL of mouse serum samples as an internal standard. To all the prepared working calibration standards and mouse serum samples, 10 µL of 4% phosphoric acid in aqueous, 1 mL of 16% 2-propanol in ethyl acetate were added. The glass tubes containing the mixture were vigorously mixed for 2 min on a Vortex mixer, centrifuged for

10 min at 3700 rpm/rcf, and the upper organic layers were transferred into 13 x 100 mm disposable culture tubes (Fisher). Samples were evaporated on a SpeedVac

Concentrator (Savant SPD121P-115, Thermo Electron Corporation), reconstituted in 100

µL of acetonitrile, and injected into the LC/MS/MS system. 10 µL of the standards or the samples were injected on an Agilent 1100 HPLC system coupled with ABSciex 4000

48 Qtrap mass spectrometer. The separation was achieved with a SeQuantTM ZIC®-cHILIC column (3.0 µm, 150 x 2.1 mm, 100Å) at gradient of 0.0-2.0 min, 95% B; 2.0-4.0 min, 95-

90% B; 4.1-6.0, 10% B and 6.1-12.0 min, 95% B with flow rate 0.3 mL/min. Mobile phase

A consisted of 100 mM ammonium formate in water, and mobile phase B was ACN. The detection was conducted at electrospray negative ion mode with MRM of m/z 129.0->42.0 for 5-FU, m/z 131.0->42.0 for 5-FUH2 and m/z 145.2->42.0 for 5-CU.

Cell viability assays

For cell viability assays, cells were seeded into 96-well black-walled plates at a concentration of 1-2×104 (cancer cell lines) per well in fresh media and in a volume of 100

µL per well. At endpoint, CellTiter-Glo™ (Promega) assays were performed according to the manufacturer's protocol. Luminescence values were background subtracted and normalized to No drug/No treatment control.

Immunoblotting

The following antibodies were used: Rabbit mAb DPYD (D35A8, 1:500, Cell Signaling and Rabbit pAb Invitrogen PA5-22302), TS Rabbit mAb (D26G11, 1:1000, Cell Signaling), p53 HRP (D0-1, 1:1000, Santa Cruz Biotechnology), mouse p21 (OP-64, 1:200,

Calbiochem), mouse monoclonal b-actin (A5541, 1:10000, Sigma). Secondary Goat anti-

Mouse IgG (Thermo Scientific 31430, 1:10000) and Goat anti-rabbit IgG (Thermo

Scientific 31460 1:10000).

49 Semi-quantitative and qRT-PCR

Quantitative PCR (qPCR) for DPYD, p21 and GAPDH were conducted in triplicates on

Bio-Rad CFX96 Touch for 35 cycles, PCR cycle conditions: 94oC for 10 min, [94oC for

30 secs; 53-55oC for 30 sec; 72oC for 45 sec] x 35 cycles. Gene expression values were normalized to GAPDH for each respective sample. The primer sequences used were as follows: For GAPDH, the forward primer was 5’-ACAACTTTGGTATCGTGGAAGG-3’, and the reverse primer sequence was 5’-GCCATCACGCCACAGTTTC-3’. The Human

DPYD forward primer sequence was 5’-GGTGGTGATGTCGTTGGTTT-3’ and the reverse primer was 5’-GCAGAAACGGAAGCTCCATA-3’. Mouse DPYD forward primer

5’-GACTTCAGTTTCTTCATAGTGGTGC Reverse 5’-AGCAGGGCTTTGAGTCCAGT-3’

The human p21 forward primer was 5’-CTGAGACTCTCAGGGTCGAA-3’ and the reverse primer was 5’-CGGCGTTTGGAGTGGTAGAA-3’. Mouse p21 Forward primer 5’-

TCTCAGGGCCGAAAACGGAG-3’ Reverse primer 5’-ACACAGAGTGAGGGCTAAGG-

3’ and 18S universal primers (Ambion #AM1716) Analysis of DPYD, p21 and 18S in HCT-

116 and HCT-116-p53-/- was done by semi-quantitative PCR using same conditions as qPCR. The fold induction was quantified based on DCT method i.e. 2-DDCT.

Measurement of free dUTP in cell culture in-vitro

The measurement of dUTP has been previously described by Wilson et al. 2011

(330).

Statistical analysis

All results are presented as the mean ± SEM of data. Statistical analyses were performed by the Student’s-t-test (two tail) and P-value mentioned in the figures were analyzed by

50 Graph Pad Prism V5, unless otherwise mentioned in figure legends. Western blots are representative images of 3 independent experiments. All other experiments were conducted in triplicate unless otherwise mentioned in figure legends.

Knockout of the p53 binding site using CRISPR:

We designed and tested 3 sgRNA sequences for the p53 binding site and selected the best one i.e sg1 ATACAACCTATGGCTTGCCT, which was cloned into pLentiCRISPR-E

(Addgene Plasmid #78852). The lentivirus was generated in HEK-293T cells in 10cm dish. Viral supernatant was added to HCT-116WT cells in 6well plates (Ratio 1:1) and then selected with puromycin at (1µg/ml) for 15 days to get pooled clones of p53BSKO cells.

H&E Staining: Blinded analysis of liver sections from two 5-FU treated and Vehicle treated mice was performed by a pathologist.

Survival Curve Analysis

Two subsets of TCGA colorectal patients were defined, based on levels of RNA expression (RSEM-normalized expression from RNA-seq data for Colon adenocarcimona

(COAD) and Rectal adenocarcinoma (READ) datasets. Groups of interests were those patients with (1) high TP53 and low DPYD expression (relative to the overall medians for each gene), and (2) low TP53 and high DPYD expression. Overall survival was compared for these two groups using Kaplan-Meier analysis, and the calculated P-value is from a log-rank test.

51 Scatterplot of DPYD expression versus TP53WT expression

RSEM-normalized gene expression values for TP53 and DPYD from TCGA colorectal

RNA-seq data (COAD + READ) are shown, for subjects with no reported TP53 mutations.

The Spearman rank correlation is rho=-0.1501, indicating a downward trend, but is not statistically significant (p = 0.0523). The horizontal line segments show the median (log10) expression levels of DPYD in subjects with low (< 1000) and high (> 1000) expression of

TP53.

Results

Fluorouracil (5-FU) has been a mainstay of cancer chemotherapy for over 50 years.

Based on prior work implicating p53 in 5-FU induced apoptosis (116), we hypothesized that p53 may regulate the cellular toxicity of 5-FU potentially through effects on 5-

FU metabolism. To gain insight into how p53 could influence metabolic pathways with particular relevance for the cellular response to 5-FU we employed an in-silico approach

[Genomatix (http://www.Genotmatix.de/index.html)] to identify genes with potential p53 binding sites (p53BS) within and ± 20 kb upstream and downstream of the genes involved in 5-FU metabolism. We screened for p53 binding as defined by the presence of two copies of 5’-PuPuPu-CWWG-PyPyPy-3’ with or without an intervening spacer (24). We found an enrichment of genes encoding rate-limiting enzymes in the pyrimidine catabolic pathway with p53BS. These include enzymes involved in 5-FU metabolism, whose regulation currently is poorly understood. Three genes in particular stood out, having multiple p53BS within either the 5’ promoter region or within gene introns; DPYS, DPYD

52 and UBP1. We focused on DPYD as it’s the key rate-limiting enzyme in the pyrimidine catabolic pathway.

P53 binds to a conserved high-affinity p53BS downstream of the DPYD gene and represses gene expression.

Our in-silico screen identified several putative half- and full-p53BS in the DPYD gene.

However, based on the highest scoring matrix conforming to the consensus binding motif, we selected six (6) putative p53BS for further characterization (Fig. 2-1A). ChIP analysis on genomic DNA isolated from the livers of mice subjected to treatment with 5-FU showed enrichment for the downstream p53BS R-0.840 (Chr1: 119451237-119451256; 5’-ACA-

CATG-CTC-CAC-CATG-TTC-3’) (Fig. 2-1B). R-0.840 was found to be conserved as a p53BS downstream (p53hDBS; chr1: 97299933–97299953; 5’-TGG-CTTG-CCT-GGG-

CATG-CCT-3’) of the DPYD gene that was previously identified (331). These two sequences are located at 18,319 bp (R-0.840) (p53DBS) and 15,955 bp (p53hDBS) downstream of the mouse and human DPYD genes respectively. In our ChIP analysis, we found that R-0.840 was enriched more than 2-fold following treatment of the mice with

5-FU (Fig. 2-1C). Thus, it is clear that p53 increasingly occupies a p53BS downstream of the DPYD gene following 5-FU.

53

Figure 2-1:Combined in-silico and chromatin immunoprecipitation(ChIP) identifies p53 downstream DNA binding sites(p53BS) in DPYD gene

(A) Schematic representation of the approach for in silico screening for putative p53BS within the DPYD gene and 20 kb upstream and downstream of the gene. (B) Fold-enrichment over untreated control of the p53BS (R-0.84) and other binding sites predicted by the in-silico analysis in mouse liver DNA as detected by ChIP analysis following IV 5-FU (150 mg/kg) administration for 6 hr (N=3). (NS=Not significant) (C) Validation of p53DBS enrichment when treated as in (B), binding to p21 promoter is used as positive control. (N=3) P-values determined by multiple t-test)

We have previously established a mouse model for assessing the in-vivo response to 5-

FU-based toxicities and gene expression changes in the gut (45). Using this model, we assessed the p53-dependent expression of DPYD in the liver, a key organ in the biotransformation of 5-FU. Following administration of IV 5-FU, we observed that liver

54 DPYD mRNA and DPYD protein expression was reduced by 2.2- and 1.7-fold respectively in p53+/+ as compared to liver mRNA and protein isolated from p53-/- mice

(Fig. 2-2A & B). Thus, our data indicate that p53 can modulate the expression of DPYD by binding to a p53BS downstream of the DPYD gene.

The R72P polymorphism in TP53 modulates DPYD expression following 5-FU exposure.

We hypothesized that polymorphisms in the TP53 gene that modulate the capacity of p53 to bind DNA could influence its ability to repress DPYD. Codon-72 polymorphism is a frequent polymorphism observed in the TP53 gene and is characterized by an arginine

(R) or proline (P) substitution at codon position 72. The P72 variant of p53 is capable of increased DNA binding and activation of transcription of target genes (332). To address the impact of the R72 and P72 alleles on inhibiting the expression of DPYD, we used

MEF’s isolated from humanized knock-in p53 mice (HUPKI) carrying variant alleles of the codon R72P polymorphism in TP53. Indeed, we found that the P72 allele of p53 suppresses DPYD expression when compared to the R72 p53 allele following 5-FU treatment in HUPKI MEF’s (Fig. 2-2C). Furthermore, human fibroblasts taken from two different patients with carrying homozygous polymorphic alleles (P/P and R/R) confirmed the enhanced ability of the P72 TP53 allele to repress DPYD expression following 5-FU treatment (Fig. 2-2D). Taken together these results suggest that p53 inhibits DPYD expression and indicate that polymorphisms of the TP53 gene could potentially indirectly alter systemic bioavailability of 5-FU that may impact the efficacy of 5-FU in cancer therapy.

55

Figure 2- 2: p53-dependent repression of DPYD expression in intact liver and impact of human p53 polymorphic variants on liver expression of DPYD.

(A and B) Fold-change in expression of DPYD mRNA and protein in livers of p53+/+ (wild-type,) and p53-/- mice. P-values are determined Multiple t tests .and one-way Anova [NT vs 5-FU in p53+/+ mice for DPYD mRNA is p=0.08; DPYD protein p=0.0021and NT vs 5-FU in p53-/ mice for DPYD mRNA is p=0.040; DPYD protein p=0.09] (C) HUPKI Codon R72P MEF-R/R or MEF-P/P were treated with 5-FU (384 µM) up to 24 hr and DPYD protein expression in MEF-P/P or MEF-R/R allele was evaluated by Western blot. (D) DPYD protein expression in Normal human fibroblast cell line harboring Codon R72P polymorphism, i.e., NHF P/P and NHF R/R is evaluated by Western blot after treatment with 5-FU for the indicated times.

p53 represses DPYD expression following 5-FU administration in human cancer cells.

To verify that p53 mediates repression of DPYD expression following exposure of cells to

5-FU, we tested human cells and tissues. We first treated the isogenic colorectal cancer

56 cell lines with functional TP53 (HCT-116 WT) and a deleted TP53 gene (HCT-116 p53-/-) with 5-FU (333). RT-PCR analysis indicated that the mRNA expression of DPYD decreased over time in HCT-116 WT cells as compared to the HCT-116 p53-/-cells (Fig.

2-3A, left panel).. This was observed at the protein level where a gradual increase in protein was noted over time following 5-FU treatment in the TP53-/- cells (Fig. 2-3A, right panel). This was further supported by immunofluorescence staining for DPYD whose expression decreased following 5-FU exposure (Fig. 2-4A right panel). As we observed increased p53 binding in-vivo in mouse liver to the conserved DNA-binding site, we evaluated the importance of the binding site to overall DPYD repression using HCT-116

WT cells. We knocked-out half of the binding site using CRISPR/Cas9 technology and analyzed for DPYD protein expression (Fig. 2-5). We found that repression of DPYD was not rescued suggesting this site alone is not sufficient to account for the observed repression. A possible reason is, since the DPYD gene is 850 kb in length, potentially there are other unknown p53 non-canonical binding regions that may cooperatively bring about transcriptional repression or there could be different mechanism involved inherent to tumor and normal cells. This would not be unexpected for a p53-regulated gene that typically harbors multiple p53 response elements scattered throughout the genomic sequence. This may include sites within promoters, several hundred base pairs or several

57 Kb upstream of the promoters, within introns or several kb downstream of the gene.

However, to confirm the generality of the p53-dependent DPYD repression after

Figure 2- 3: The tumor suppressor p53 represses dihydropyrimidine dehydrogenase (DPYD) expression

(A) mRNA and protein expression of DPYD in HCT-116 p53+/+ and HCT-116 p53-/- cell lines at indicated times after 5-FU (384 µM) treatment. (B) Fold-expression of mRNA in A549 and H460 cell lines at 24 hr

after 5-FU (384 µM) treatment with and without siRNA knockdown of p53 (P=0.0011 N=3). (C) H3K9 Acetylation at DPYD promoter following 5-FU treatment for indicated time points. Values are normalized in the sequence [input>IgG>total H3>No Treatment (NT)] (N=3). (D, E, F) Protein expression of DPYD in A549, U87MG, HT-1080 and H460 is shown after western blotting at the indicated time points with and without siRNA knockdown of p53. (G) H3K4me3 and H3K27me3 at DPYD promoter at 24hrs after 5-FU treatment. Values normalized in the sequence [input>IgG>total H3] (N=3).

58 cellular exposure to 5-FU, we used the A549, H460, U87MG and HT-1080 cancer cell lines that are wild-type for TP53 and trigger the expression of canonical p53 target genes.

Indeed, the mRNA and proteins levels of DPYD decreased at 24 hr following treatment of the lung cancer cell lines A549 and H460 cells with 5-FU as compared to cells subjected to siRNA targeting of TP53 (Fig. 2-3B and Fig 2-4A&B). In a similar manner, the U87MG (glioblastoma) and HT-1080 (fibrosarcoma) cell lines also repressed DPYD expression in a TP53-dependent manner (Fig. 2-3D, E, F). The correlative decrease of

DPYD protein to its decrease in mRNA were not so apparent in the liver as it was in cancer cells something which has previously observed (334-337). To further investigate

Figure 2- 4: DPYD p53 co-localization and Epigenetic marks at DPYD promoter

(A) Representative images of HCT-116 and H460 cell show decreased expression of DPYD and increased p53 nuclear staining following administration of 5-FU (384uM) for 24hrs. (B). Merged images of DPYD and p53 showing cells with increased p53 expression and nuclear localization and reduced DPYD expression (see arrows). (C) H3K9Ac and H3K27m3 marks at DPYD promoter in H460 cells following 5-FU(384uM) for 24hrs (n=3) P-value calculated by multiple t-test.

59 transcriptional mechanism of DPYD repression, we evaluated H3K9 acetylation,

H3K4me3 and H3K27me3 at the DPYD promoter region following 5-FU administration in

HCT-116 and in H460 cells. H3K9 acetylation progressively decreased and was lowest

at 24 hr following 5-FU treatment consistent with lower DPYD promoter activity (Fig 2-

3C). H3K27me3 was also increased at 24 hr but no changes were observed in H3K4m3

(Fig. 2-3G and Fig. 2-4C). Taken together our data indicate, regardless of the genetic or

epigenetic background, that p53 negatively regulates the expression of the DPYD gene

in human cells in-vitro and that loss of TP53 abrogates the repression of DPYD

expression.

Gokare P et al Repression of DPYD by p53

Supplemental Fig 1

Figure 2- 5: Analysis of DPYD protein expression in CRISPR edited p53 binding site in HCTFigure-116 S1: cellsAnalysis of DPYD protein expression in CRISPR edited p53 binding site in HCT-116 cells. (A) TIDE analysis of depicting the frequency of deletion in the p53BS in HCT-116, Over 78.8% of clones have lost half of the p53 binding sites, Total editing efficiency

is 96% (Lower panel) Region of decomposition depicting aberrant nucleotide sequence of edited cells. (green) vs WT cells (Black). (B)

60 (A) TIDE analysis of depicting the frequency of deletion in the p53BS in HCT-116, Over 78.8% of clones have lost half of the p53 binding sites, Total editing efficiency is 96% (Lower panel) Region of decomposition depicting aberrant nucleotide sequence of edited cells. (green) vs WT cells (Black). (B) Repression of DPYD following 5-FU (384uM) for 24hrs in p53BSKO HCT-116 cells. (Lower panel) shows the p53 binding site sequence and sequence in red is lost after the cut by CAS9/sgRNA complex indicated by the downward arrow)

Liver specific deletion of TP53 increases systemic catabolism of 5-FU and accelerates syngeneic tumor growth

Approximately 80% of administered 5-FU is eliminated through catabolism by hepatic

DPYD(165). To establish the significance of liver DPYD in limiting the bioavailability of intravenously (IV) administered 5-FU treatment we subjected mice to treatment with IV 5-

FU in the presence or absence of the specific DPYD inhibitor gimeracil and monitored parameters of the acute toxic response to the drug (Fig. 2-6). Indeed, administration of gimeracil (Santa Cruz) along with IV 5-FU triggered increased loss of body weight of mice over the course of ten (10) days as compared to IV 5-FU alone (Fig. 2-6A). The group of mice subjected to the 5-FU/gimeracil combination treatment were also more moribund

(Fig. 2-6B). In addition, we monitored the levels of leukopenia and thrombocytopenia in mice subjected to the combination of 5-FU/gimeracil (Fig. 2-7). These data indicate that

DPYD has a profound effect on the toxicity of accumulated high doses of 5-FU administered IV in mice in-vivo.

61 Gokare P et al Repression of DPYD by p53

Supplemental Fig 2

A B

Figure 2- 6: Impact of bodyweight and survival of mice targeted by Gimeracil following IV administration of 5-FU

Gokare P et alFigure(A) WT S2: C57BL/6 Impact on mice bodyweight were given and survival vehicle, of 5 mice-FU targeted50mg/kg by BW Gimeracil Repression of DPYD by p53 IV or 5 following-FU and IVGimeracil administration (22.4 mg/kgof 5-FU. BW (A) WT IV) or Gimeracil alone every 3 days over a week. Body weights of these mice were measured every other C57BL/6 mice were given vehicle, 5-FU 50mg/kg BW IV or 5-FU and Gimeracil (22.4 mg/kg BW IV) or Gimeracil alone every 3 days day and represented as percentage of Day 0. (B) The moribundicity of mice following treatment as in (A). over a week. Body weights of these mice were measured every other day and represented as percentage of Day 0. (B) The moribundicity SupplementalStatistical analysis Fig 3 was carried out using student t-test and P-values are indicated in figures. of mice following treatment as in (A). Statistical analysis was carried out using student t-test and P Values are indicated in figures.

FigureFigure S3: Asse2- 7ssment: Assessment of Hematological of hematological parameters following parameters IV 5-FU following administration IV 5-. FUCBC analysis following treatment withadministration 5-FU and Gimeracil as treated in Fig S1.

62

CBC analysis following treatment with 5-FU and Gimeracil as treated in Fig 2-6.

We further explored if the repression of DPYD expression by p53 in-vivo could modulate the pharmacokinetics of 5-FU. We hypothesized that targeting of DPYD-expression through deletion of TP53 in hepatocytes in-vivo may impact the systemic bioavailability of 5-FU. To this end we generated mice with liver-specific deletion of the TP53 gene i.e., through the use of mice expressing Cre recombinase under the control of the Albumin promoter(338). Indeed, Albcre;mT/mg;p53Δ/Δ mice showed expression of Cre recombinase in hepatocytes and no histological changes were detected in other organs following deletion of TP53 in the liver (Fig. 2-8A). The plasma half-life of 5-FU is approximately ~20 min (339,340), Therefore, to measure the impact of DPYD repression on 5-FU bioavailability, we followed a treatment schedule in which mice were either given

Vehicle or 5-FU for the first 6 hrs, i.e. a time point at which DPYD expression is repressed

(Fig 2B), followed by second dose of 5-FU for 30 min. As expected, a lower ratio of 5-

Δ/+ FUH2/5-FU was observed in Albcre;mT/mg;p53 mice as compared to the

Albcre;mT/mG;p53Δ/Δ mice (Fig 2-8B&C). Western blot analysis revealed expression levels of DPYD in these mice consistent with our earlier observation of repression of

DPYD by p53 (Fig 2-8F). We further investigated whether the apparent decrease in 5-FU catabolism indicated by a lower plasma ratio of 5-FUH2/5-FU in the presence of wild-type

TP53 could have a functional consequence on the in-vivo therapeutic response to 5-FU.

We used syngeneic (C57BL6/J) malignantly transformed mouse colonocytes to model colorectal cancer with mutated KRAS, p53 and Myc-overexpression (p53dmc-Ras-

Myc)(341). p53dmc-Ras-Myc cells were injected subcutaneously in the flanks of mice and the mice were subjected to treatment with 5-FU (100 mg/kg/week for 6 weeks) with a

63 follow-up time of up to 40 days. Indeed, Albcre;mT/mg;p53Δ/+ mice showed a significant increase in the tumor doubling-time (vehicle versus 5-FU-treated tumor doubling-times were 6.22 versus 11.06 days, respectively) following treatment with 5-FU (Fig. 2-8D). In comparison Albcre;mT/mg;p53Δ/Δ mice displayed a clearly blunted response to 5-FU when compared to vehicle-treated controls (vehicle compared to 5-FU-treated tumor doubling-times were 8.83 vs. 8.93 days, respectively). Furthermore, tumors from 5-FU- treated Albcre;mT/mg;p53Δ/+ mice exhibited increased levels of apoptosis and decreased levels of mitosis when compared to tumors from 5-FU-treated Albcre;mT/mg;p53Δ/Δ mice

(Fig. 2-8E). To eliminate the possibility of hepatotoxicity affecting the outcomes we analyzed liver sections from both 5-FU-treated and -untreated mice and found no differences in tissue histology, or any change in body weight (Fig. 2-9) Taken together these results strongly implicate p53 in controlling liver catabolism and therapeutic efficacy of 5-FU

64

Figure 2- 8: Tp53 specific liver depletion upregulates the catabolism of 5-FU through DPYD

(A). Liver specific expression of Cre in Albcre;mT/mG;p53Δ/Δ mice as seen by expression of GFP and normal histology of liver, Bone marrow (BM) and colon in these mice (B) 5-FU treatment schedule for mice (p53Δ/+ and p53Δ/Δ) with liver specific deletion of the TP53 gene, [S-1= First dose Vehicle (6hrs)+ second dose 5-FU(30 min)]; S-2=[First dose 5-FU(6hrs)+second dose 5-FU(30min)]. (C) Ratio of the amount of 5- Δ/+ Δ/Δ FUH2/5-FU in plasma of liver specific p53 and p53 genotypes following the treatment plan described in (B) (Values represent median n=5-7; p=0.0313 Wilcoxon-rank-sum test). (D) Tumor growth delay (TGD) of syngeneic p53dmc-Ras-Myc colonocytes injected subcutaneously (s.c.) into liver specific p53Δ/+ and p53Δ/Δ and treated with vehicle or 5-FU IV (100 mg/kg/week, for a total of 6 weeks) (N=3-5, Doubling time calculated by exponential growth equation). (E) Analysis of cell death and proliferation in syngeneic tumors on p53Δ/+ and p53Δ/Δ at the study in (C) as indicated by number of apoptotic and mitotic nuclei (N=3). (F) Representative Western blot showing expression of DPYD in liver of p53Δ/+ and p53Δ/Δ mice with prior treatment of 5-FU for 6hrs as indicated in (B) and (C).

65

Figure 2- 9: Assessment of liver toxicity following IV 5-FU treatmen of tumor bearing mice

(A)&(B) H&E Sections of liver from NT (i.e Vehicle) and 5-FU treated (6 weeks) in Albcre;mT/mG;p53Δ/Δ mice. (C) Average body weights of mice from syngeneic tumor study (N=3-5).

Repression of DPYD by p53 is a consequence of thymidylate synthase inhibition and thymidine deficiency.

To better our understanding of the upstream signaling events behind the p53-dependent repression of DPYD by p53 we asked whether this was specifically related to the activity of 5-FU or just simply part of the cellular DNA damage response (DDR). Treating HCT-

116 cells with different chemotherapeutic agents indicated that the repression of DPYD was not a result of DNA damage per se since treatments with the Topoisomerase-I and -

II poisons irinotecan (CPT-11) and etoposide respectively did not reveal a p53-dependent inhibition of DPYD expression (Fig. 2-10A). Both irinotecan and etoposide caused

66 induction of DPYD expression in a somewhat p53-dependent manner. The therapeutic effect of 5-FU is considered to arise from incorporation of 5-FU in DNA, RNA and mainly by inhibition of thymidylate synthase (TS)(326). We hypothesized that the effect of 5-FU to repress DPYD may stem from TS targeting and the resulting depletion of thymidine pools that may trigger p53 activation. To address this specifically, we treated HCT-116-

WT and HCT116 TP53-/- cells with 5-FU and Tomudex (Raltitrexed), a specific inhibitor of

TS. 5-FU forms an inactive ternary complex with TS, which causes a slight upward shift in the TS band signifying TS inhibition (342), whereas TS inhibitors such as Tomudex,

Pemetrexed (PX; Alimta®) and methotrexate relieve feed-back inhibition of TS mRNA translation causing modest increases in TS protein expression (176,342-345). We found a p53- and dose-dependent repression of DPYD expression following 5-FU and

Tomudex-treatment (Fig 2-10B). As an alternative approach to block TS, we employed the TS-inhibitor methotrexate, not a direct inhibitor per se and direct siRNA targeting TS.

Treatment of H460 cells with methotrexate and with TS siRNA generated similar results of p53-dependent DPYD repression as seen with 5-FU and Tomudex (Fig 2-10C). The results were further expanded to Pemetrexed (PX; Alimta®) and Fluorodeoxyuridine

(FdU; floxuridine) that also selectively act on TS (Fig 2-10D). As data using five (5) different inhibitors yielded similar results with respect to p53-dependent inhibition of

DPYD expression, we sought to determine whether thymidine supplementation could overcome the p53-mediated repression of DPYD. As expected addition of exogenous thymidine restored DPYD protein expression following targeting of TS with siRNA (Fig 2-

10E). As a control experiment to directly verify the impact of siRNA-mediated targeting of

TS we analyzed changes in intracellular nucleotide pools. Inhibition of TS is known to

67 cause reduced conversion of dUMP to dTMP and as a result compensatory increased conversion of dUMP to dUTP (326). We found the levels of free dUTP increase in cells subjected to siRNA-targeting of TS and 5-FU (Fig 2-10F). Interestingly, targeting of TS by siRNA was approximately three (3) times more effective than 5-FU to generate increased levels of cellular dUTP, indicating the functional relevance of the TS-inhibition approach. Taken together the data suggest that following efficient TS-inhibition, p53 may specifically repress the expression of DPYD protein and downregulate pyrimidine catabolism

Figure 2- 10: p53 represses the expression of DPYD specifically following thymidylates synthase (TS) inhibition due to thymidine deficiency

(A) mRNA expression of DPYD following Etoposide and CTP-11 in HCT-116 p53+/+ and p53-/- colorectal cancer cells (N=3). (B) DPYD Protein expression following TS inhibition in HCT-116 p53+/+ and p53-/- cell lines treated with 5-FU and Tomudex (Raltitrexed) for 24 hr. (All lysates were evaluated on same gel). (C) Rescue of DPYD protein repression by inhibition of the DNA damage response, with ATM (KU-55933) and DNA-PK (D-PK) (NU7026) inhibitors for 24 hr of H460 cells treated with methotrexate (MTX). (D) Repression of DPYD protein expression with various TS inhibitors [Fluorouracil (FU), Fluorodeoxyuridine

68 (FdU), methotrexate (MTX), Pemetrexed (PX),] and TS knockdown (TSsi). (E) Rescue of DPYD protein repression by addition of thymidine-H460 cells that were incubated with thymidine following treatment with 5-FU or TS knockdown for 24 hrs. (F) Increase in dUTP levels following treatment with 5-FU and TSsi.

p53-dependent repression of DPYD is dependent on ATM and DNA-PK signaling following TS inhibition.

A decrease in the thymidine pools can result in nucleotide imbalance and cause DNA damage that in turn may lead to p53 activation(342). We hypothesized that blocking upstream DDR kinases that impact on p53 stabilization potentially could relieve inhibition of DPYD repression. We treated HCT-116 TP53 WT or TP53-/- cells with inhibitors of ATM

(KU-55933) or DNA-PK (NU7026) in the context of 5-FU-induced damage. We observed that in wild-type p53-expressing cells the 5-FU induced repression of DPYD was alleviated when cells were co-treated with the kinase inhibitors (Fig. 2-11A). By contrast, in TP53-/- cells no change in the expression of DPYD was observed following co-treatment with KU-55933 or NU7026. We extended our analysis to include siRNA to TS (TSsi) and

Tomudex, (Fig. 2-11B and 2-11C). The results indicate that following more selective targeting of TS, p53-dependent inhibition of DPYD expression requires functional DNA-

PK and ATM. Despite some observed discrepancies between DNA damaging chemotherapeutics such as etoposide, CPT-11 and 5-FU in the triggering p53-dependent repression of DPYD (Fig. 2-10A), our data indicate that key DNA-damage signaling kinases such as ATM and DNA-PK are required to signal p53-dependent repression of

DPYD expression following TS inhibition. To functionally validate how DPYD). influences the cancer cell intrinsic response to TS inhibitory drugs we targeted DPYD expression in p53-null cancer cells with siRNA and subjected them to treatment with Tomudex and 5-

69 FU. Knockdown of DPYD in HCT-116-p53-/- cells significantly sensitized them to the toxic effects of 5-FU and Tomudex (Fig. 2-11D). Thus, it appears that elevated

Figure 2- 11: p53 dependent repression of DPYD is dependent on signaling fro ATM and DNA-PK following TS inhibition

(A) Rescue of DPYD repression by inhibition of DNA damage response, HCT-116-WT and p53-/- cells treated with 5-FU, with ATM (KU-55933) and DNA-PK (NU7026) inhibitors for 24 hrs. (B, C) Co-treatment with ATM and DNA-PK inhibitors relieves inhibited DPYD expression following targeting of TS with siRNA

(TSsi) and Tomudex. The effects of siRNA targeting of DPYD on 72-hr cell-viability of HCT-116-WT and p53-/- cells following treatment with isotoxic doses of 5-FU (D) and Tomudex. (E) High expression of DPYD mRNA predicts poor disease-free survival in colorectal cancer patients mainly with Duke Stage B and C as analyzed from the GSE14333 (Melbourne) data set, P-value calculated by cox regression analysis (Cox P- value=0.0010z

70 levels of DPYD may contribute to a resistance phenotype to TS inhibitors observed in cancers that have lost functional p53. Furthermore, this may indicate that DPYD can confer drug resistance to TS inhibitory drugs that is independent of systemic catabolism and reduced bioavailability.

Table 2-1: Patient characteristics of Figure 2-11E

Number DUKE STAGE, N (%) ADJCTX, N (%) ADJXRT N, (%) COX P- HR [CI95%] VALUE of A B C Patients N=226 41 94 91 87 (38.5) 22(9.7) 0.001050 1.68 [1.23 - 2.29] (18.1) (41.6) (40)

DPYD mRNA expression correlates with poor disease-free survival in colorectal cancer.

We analyzed whether tumor DPYD mRNA expression could predict poor outcome in colorectal cancer patients treated with chemotherapy. Analysis of the GSE14333 data- set (346) in 226 colorectal cancer patients comprised mainly of Duke Stage B and C cancers revealed that high DPYD mRNA expression correlates with poor disease-free survival as compared to those expressing low DPYD mRNA levels (Fig. 2-11E & Table

2-1). Analysis of a TCGA cohort for correlation between TP53 and DPYD expression revealed a strong inverse trend, where higher TP53 expression showed lower DPYD expression along with poorer survival rate in patients having TP53Low/DPYDHigh vs

TP53High/DPYDLow (Fig. 2-12). However, these analyses did not reach statistical significance (P=0.12).

71 Taken together, our data suggest that an imbalance in the cellular nucleotide pool

resulting from reduced levels of thymidine required for the synthesis of DNA triggers p53- Gokare P et al Repression of DPYD by p53 dependent inhibition of the key rate-limiting enzyme of pyrimidine catabolism DPYD which

Supplementalin turn reduces Fig 55 -FU catabolism.

Fig. S5 (A) TCGA analysis of DPYD and p53 expression and correlation with overall survival of CRC patients. RSEM-normalized geneFigure expression 2 -values 12: for TCGA TP53 and analysis DPYD The Spearmanof DPYD rank andcorrelation p53 is rho=expression-0.1501, indicating and a downwardcorrelation trend, (p with= 0.0523). The horizontal line segments show the median (log10) expression levels of DPYD in subjects with low (< 1000) and high (> 1000) exproverallession of TP53 survival. (B) “Kaplan of- MeierCRC plot patients of overall survival in patients with Tp53High/DPYDlow vs Tp53Low/DPYDHigh groups.

(A) RSEM-normalized gene expression values for TP53 and DPYD The Spearman rank correlation is rho=-

0.1501, indicating a downward trend, (p = 0.0523). The horizontal line segments show the median (log10)

expression levels of DPYD in subjects with low (< 1000) and high (> 1000) expression of TP53. (B) “Kaplan-

Meier plot of overall survival in patients with Tp53High/DPYDlow vs Tp53Low/DPYDHigh groups.

Discussion

We show for the first time that the p53 tumor suppressor protein controls 5-FU catabolism

by repressing the expression of the key rate limiting enzyme in pyrimidine degradation,

DPYD. p53 has a well-documented function in the cell death response following 5-FU

treatment in pre-clinical experiments in-vitro and in-vivo (116,347). In clinical settings

72 TP53 mutation status has been correlated with survival following 5-FU-based chemotherapy (116,202,203). However, it has been difficult to identify key genes downstream of p53 that predict the 5-FU response in patients in-vivo and TP53 mutation status may not necessarily be a predictor of the response to other chemotherapy agents.

This may partly be due to the complexity of the mechanism of action of 5-FU that ranges from interference with both DNA and RNA synthesis through direct incorporation in such polynucleotide strands and subsequent repair thereof and inhibition of dTTP production through targeting of TS (348).

A critical determinant of the efficacy of 5-FU treatment directly relates to the expression of key metabolic genes that are responsible for the biotransformation of the drug. DPYD is an important catabolic enzyme that limits the bioavailability of 5-FU to therapeutically relevant anabolic pathways and has been linked to 5-FU efficacy and toxicity (349). We show that high expression levels of DPYD is linked to poor outcomes in colorectal cancer.

Interestingly, the link we have uncovered between TP53 and DPYD suggests that expression of both of these genes may serve as markers in determining the response to

5-FU. Furthermore, observations in humanized TP53 knock-in mice and patient samples suggest that polymorphisms in the TP53 gene can influence the inhibition of DPYD expression. Based on our data, the capability of the P72 TP53 allele, which is a more active transcriptional variant of TP53, in repressing DPYD expression over the R72 TP53 allele suggest that p53 may alter the 5-FU drug sensitivity heterogeneously in the general population through its impact on 5-FU catabolism of normal tissues such as the liver (Fig.

2-2C & D). The tumor response observed in our syngeneic mouse model clearly

73 demonstrates the importance of systemic effects of this interaction (Fig. 2-8D). However, it should be noted that the syngeneic tumors in our study expressed low levels of DPYD in comparison to liver tissue. Evidence suggests that some tumors express high levels of

DPYD compared to surrounding normal tissues as something that may in particular be true for colorectal cancers that metastasize to the liver (350). Overexpression of TS and

DPYD is seen frequently in colorectal tumors (197,351). It is generally known in the clinical setting that tumors with low DPYD expression, low TS expression and that are wild-type for TP53 show a favorable response rate following treatment with 5-FU(352).

Subsequently, loss of functional p53-signaling in colorectal cancer, a typical late-stage event in the disease, may fail to suppress DPYD expression and add another level of complexity to 5-FU treatment by contributing to drug resistance. This idea is supported by the DPYD gene expression profile in p53 WT cell lines (Figure 2-3) as well as in advanced stage colorectal tumor patients where higher expression of DPYD predicts poor disease-free survival (Fig. 2-11E and Fig. 2-12). It remains to be seen to what extent tumor levels of DPYD can make a significant contribution to 5-FU catabolism in-vivo.

Our data in tumor cell lines suggest that inhibition of TS by 5-FU, methotrexate, Tomudex or Pemetrexed causes p53-dependent repression of DPYD expression (Fig. 2-10D). We provide evidence that repression by p53 of DPYD expression is related to changes in the dUTP/dTTP ratio and signaling from ATM and DNA-PK following TS-inhibition unlike other DNA damaging agents. A potential compensatory response to the reduction in dTTP levels, as a result of TS-inhibition, could be to reduce catabolism of pyrimidines to salvage thymine and uracil necessary for dTTP synthesis and DNA replication recovery.

74 Suppression of pyrimidine catabolism by p53 following nucleotide imbalance may be another way for the TP53 tumor suppressor to control the integrity of DNA synthesis by favoring nucleotide salvage of thymidine and prevent errors during replication, However, since blocking pyrimidine catabolism would also affect catabolism of 5-FU, this would in turn cause an increase in 5-FU bioavailability emphasizing the deleterious effects in wild- type p53-expressing cells. In line with this, siRNA targeting of DPYD in TP53-null cells sensitizes to chemotherapeutics that target TS such as Tomudex (Fig. 2-11D).

Our data suggest that inhibiting DPYD would work synergistically with TS inhibitors. In support of our observations, use of a clinical DPYD inhibitor gimeracil as a component of the oral S-1 (tegafur, gimeracil and oteracil) mix has already proved promising in many solid tumors and is approved in more than 50 countries but not in the US. Early indications of synergy between S-1 and TS inhibitors have been reported for 5-FU resistant tumors

(353). S-1 has been combined with HDAC inhibitors, which are known to downregulate

TS expression (354). Moreover, targeting DPYD with the oral irreversible inhibitor eniluracil, has significantly improved PFS and OS of patients with metastatic breast cancer in comparison to patients who were refractory to capecitabine (oral 5-FU) (323).

Eniluracil also limited the frequency of hand-foot syndrome, a toxicity phenotype believed to result from metabolites of 5-FU catabolism. Considering that recent evidence indicates that DPYD may play a role in breast cancer metastasis (314), it would be interesting to determine if a DPYD inhibitor might provide added benefit to patients by limiting toxicity as well as targeting tumors that undergo EMT.

75 In conclusion, our study provides the first evidence for a role of the tumor suppressor p53 protein in downregulating pyrimidine and 5-FU catabolism by repressing DPYD gene expression following TS inhibition. These effects are not observed with other DNA damaging chemotherapeutic drugs like the topoisomerase inhibitors etoposide or irinotecan. Further studies would need to evaluate the interplay between combined use of 5-FU and irinotecan as compared to 5-FU alone with regard to p53 dependent regulation of DPYD. The idea that mutant p53 could up-regulate DPYD as a resistance mechanism to 5-FU treatment is a focus of our future research as 5-FU is used in multiple consecutive regimens in the therapy of evolving colorectal tumor. Overall our data document a role for tumor suppressor p53 in controlling pyrimidine catabolism through

DPYD, particularly following metabolic stress imposed by nucleotide imbalance, and signaling effects through DNA-PK and ATM. The findings have implications for the toxicity and efficacy of the cancer therapeutic 5-FU. Previous findings from our lab have demonstrated that monitoring 5-FU levels can minimize toxicity and improve outcomes

(355) and thus this study can provide some avenues for future design of potentially more effective treatments or treatment monitoring.

76 CHAPTER 3

Targeting of Chk2 as a countermeasure to dose-limiting toxicity triggered by topoisomerase-II (TOP2) poisons

This chapter has been published (Gokare et al., Oncotarget. 2016 May 17;7(20):29520-

30)

Introduction

Myelosuppression is one of the most common acute dose-limiting toxicities (DLT) following systemic chemotherapy that may result in delayed or discontinued treatment with a negative impact on patient quality-of-life and survival. As such, there is a need for clinical translation of targeted toxicity countermeasures that prevent killing of normal cells without interfering with the efficacy of cancer treatment. The cell cycle checkpoint kinase

2 (Chk2) is a serine/threonine kinase that transduces DNA damage response (DDR) signals from the kinases ATM and to some extent also ATR. ATM phosphorylates Chk2, promotes dimerization and subsequent trans-autophosphorylation of Serine 516 required for fulminant activity (356,357). Loss of the Chk2-gene in mice is associated with reduced levels of apoptosis in hematopoietic cells and improved survival following lethal doses of

γ-radiation (IR) (286).

Interestingly, evidence suggests different emphasis on Chk2’s function in normal and cancer cells. Chk2 promotes programmed cell death in normal cells in part through the tumor suppressive p53-pathway. In contrast, cancer cells frequently carry inactivating mutations of the p53 gene and increasingly rely on p53-independent cell cycle

77 checkpoints in G2/M something that may be exploited therapeutically (358,359). Indeed,

Chk2 has been shown to trigger p53-independent but CDC25-dependent cell cycle checkpoints and subsequent DNA repair following damage inflicted by cancer therapy

(360). Consequently, pharmacologic Chk2-inhibitors (Chk2i) have been developed with the intent to sensitize cancer cells to radiochemotherapy. Chk2i’s may also at the same time protect normal human and mouse cells from cell death following ionizing radiation in-vitro (361-363). However, it remains unclear to what extent pharmacologic Chk2i’s are an effective strategy to prevent toxicities from radiochemotherapy in-vivo.

The use of Chk2i may be particularly beneficial for patients subjected to chemotherapy given that such treatments generate a systemic exposure to genotoxicity that typically contrasts that of modern radiotherapy. However, chemotherapeutics constitutes a heterogeneous group of compounds with diverse modes of action that may not rely entirely on ATM-Chk2-p53 signaling (or DDR at all) to trigger DLT’s (364). It is therefore possible that successful molecular targeting of Chk2 for the purpose of inhibiting chemotherapy DLT is highly treatment-dependent. Successful clinical translation of Chk2i may thus depend on identifying a specific context where Chk2 inhibition would be most beneficial as a toxicity countermeasure.

To address this, we performed screening in cells that were proficient and deficient for

Chk2 (Chk2-/-) to identify chemotherapeutics that trigger toxicity through Chk2.

Collectively, our data indicates that chemotherapy belonging to the class of topoisomerase II (TOP2) inhibitors is particularly likely to trigger Chk2-dependent cell

78 death and toxicity in-vitro. Furthermore, we also designed an in-silico screen that would allow for the condensation of small molecule compound libraries to lead compounds with an affinity to bind to the ADP binding pocket of Chk2. By assessing the Chk2 kinase- and cell death inhibitory activities of the compounds in this condensed library we were able to identify the antiviral compound ptu-23/NSC105171 as a Chk2i that reduces etoposide toxicity in-vivo.

Materials and Methods

Mice and treatments

Six to eight-week old male C57BL/6J (wild-type), C57BL/6-Bbc3tm1Ast/J (puma-/-) and

B6.129S6 (Cg)-Cdkn1atm1Led/J (p21-/-) were purchased from Jackson Laboratory

(Jackson Laboratory, ME). Chk2-/- mice have been described previously (365). All mice were housed in a controlled environment with regard to light, temperature and humidity.

Mice were euthanized through cervical dislocation following ketamine/xylazine anesthesia. An Institutional Animal Care and Use Committee approved all animal care and treatment procedures employed. Four- to six-week-old wildtype and Chk2−/− mice were subject to 5 Gy of whole-body radiation (WBR) with a 137Cs gamma source at a rate of 1.4 Gy/min. At 6 hrs following radiation animals were euthanized by cervical dislocation with an approved Institutional Animal Care and Use Committee Protocol, which followed the recommendations of the Panel on Euthanasia of the American Veterinary Medical

Association. The femur and tibia of the animals were isolated and aspirated with RPMI-

1640 medium supplemented with FBS and penicillin and streptomycin. Red blood cells were lysed in ACK buffer (0.15 M NH4Cl, 10.0 mM KHCO3, 0.1 mM Na2EDTA, pH 7.4)

79 and the remaining bone marrow cells were washed and re-suspended in cold PBS. The bone marrow cells were then stained with propidium iodide and subjected to flow cytometric sorting to quantitate the percentage of cells with sub-G1 DNA (see below). For the in-vivo etoposide-toxicity model, mice were injected with etoposide at a dose level and protocol equivalent to a clinically relevant protocol of etoposide treatment.

Subsequently etoposide was injected IV at 32 mg/kg bw three times over 5 days with one day of rest between doses. The body weight of the mice was measured throughout the study. For hematology of etoposide-treated mice, wildtype and Chk2−/− animals were subjected to two (2) consecutive doses of etoposide (32 mg/kg bw IV). Mice were euthanized two days following the last dose of etoposide and blood was collected by cardiac puncture and submitted for analysis.

Cell culture, cell viability assays, and reagents

Cell lines were obtained from ATCC and cultured in ATCC-recommended media in a humidified incubator at 5% CO2 and 37°C. For cell viability assays, cells were seeded into 96-well black-walled plates at a concentration of 2×105 cells (splenocytes) and 50,000 cells (cancer cell lines) per well in fresh media and in a volume of 100 µL per well. Cells were allowed to adhere overnight and were treated the next day as indicated. At endpoint,

CellTiter-Glo™ (Promega) assays were performed according to the manufacturer's protocol, and the bioluminescent readout was recorded on an IVIS imaging system

(Xenogen). Camptothecin, teniposide, etoposide, daunorubicin, adriamycin, topotecan, taxol and CPT-11 were obtained from the Penn State Hershey Cancer Institute

Pharmacy. MG132 was purchased from Sigma-Aldrich. Chk2 inhibitors PV1019 (362),

80 CI2 (361) and CI3 (363) were purchased from EMD Millipore. CellTiter-Glo™ was purchased from Promega.

Splenocyte isolation and compound screening

Splenocytes were isolated from 4 – 6-week old wild-type and Chk2-/- mice and cultured in 96-well plates at 1.0 x 105 cells/well in RPMI 1640 medium supplemented with 10% fetal bovine serum (Hyclone) and antibiotics. The splenocytes were subjected to treatment for twenty-four (24) hours with different chemotherapy after which cell viability was assessed by using CellTiter-Glo® (Promega) and a xenogen imaging system. For the determination of compound protection from Chk2-dependent killing wildtype and

Chk2-/- splenocytes were treated with equitoxic (IC50) doses of daunorubicin for twenty- four hours in the presence and absence of 5 µM of compounds from the in silico condensed compound library and the cell viability was determined using the CellTiter-

Glo® assay.

In-silico screening of compound libraries

The crystal structure of human CHK2 in complex with ADP, debromohymnialdisine- derived inhibitors and NSC109555 was retrieved from the Protein Data Bank (PDB:

2CN5, 2CN08, 2W07 (362,366)). The electrostatic Potential (ESP) methodology was used to calculate charges (with the help of the R.E.D. Server (367)) to help assess inter- molecular dynamic interactions within the ADP-binding pocket of Chk2. Docking protocols were generated using MOE-Dock (http://www.chemcomp.com) GOLD

(www.ccdc.cam.ac.uk/) and Glide (368). To generate an optimally performing docking

81 protocol, ADP and NSC109555 were re-docked to the ADP-binding pocket of the crystal structure of human Chk2 with several combinations of scoring and algorithms for docking function using the Schrödinger Small-molecule Drug Discovery Suite. This docking protocol was subsequently applied to the Diversity Set II compound library (DTP/NIH).

Flow cytometry

Single cell suspensions were prepared from the tibia and femur and analyzed for sub-G1

FACS analysis. Bone marrow cells were collected and fixed with 70% ethanol at 4°C. The samples were stained with propidium iodide (Sigma) in the presence of RNase and subjected to flow cytometric analysis using Epics Elite flow cytometer (Beckman Coulter,

Fullerton, CA).

Retroviral transfection

The generation of stable p53 knock-down human colorectal cancer cell lines was performed in a similar manner as described previously for mammary epithelial cells (369).

Human colorectal cancer cells HCT116 were grown as described previously (370). Briefly, amphotrophic retroviruses were made by transfecting Phoenix-Ampho cells with the

Lipofectamine2000 reagent (Invitrogen) by following the manufacturer’s instructions. After

2 days of transfection, the filtered viral supernatant was used to infect target HCT116 cells using spin centrifugation. Retroviruses were infected serially and cells having puromycin resistance (together with scrambled shRNA or shRNA targeting p53) were isolated and subsequently expanded in-vitro. The generation of E1A-immortalized mouse embryo fibroblasts (MEF) was performed in a similar manner as described previously

82 (343). Ecotropic Phoenix cells were transfected with LPC-E1A (kindly provided by Dr.

Scott Lowe, Memorial Sloan-Kettering Cancer Center). The supernatant containing recombinant retrovirus was harvested 48 h following transfection. Wildtype and Chk2-/-

MEF’s were seeded at a density of 105 and incubated 1 h with the retroviral supernatant.

Cells stably expressing E1A were selected by incubation with either 2 μg of puromycin per ml for 3 days.

Histology and Immunohistochemistry

Following necropsy, the bone marrow, colon, small intestine, spleen, testis and thymus were collected and fixed in 4% paraformaldehyde overnight at 4°C, embedded in paraffin and cut into 4-µm sections for blinded evaluation by histology and immunohistochemistry.

Cut sections were stained with hematoxylin and eosin (H&E) and analyzed by microscopy. Immunohistochemistry was performed by rehydrating slides and subjecting them to antigen-retrieval through boiling in 1 mM citric acid buffer (pH 6.0). Endogenous peroxidases were blocked by submerging slides in 3% H2O2. Briefly, primary antibodies were left on sections overnight at 4°C. Sections were washed and primary antibodies were conjugated with either peroxidase conjugates using the ImmPRESS™-system

(Vector laboratories) counterstaining was performed using hematoxylin. Representative depiction of histology and immunohistochemistry was made using IP lab software (BD

Biosciences).

83 Western blotting

Cells were homogenized and sonicated in RIPA-buffer (1xPBS, 1% NP-40, 0.5% sodium deoxycholate, 0.1% SDS, 1mM PMSF, Complete protease inhibitor cocktail (Roche).

Protein concentrations were determined by the Bradford method (Bio-Rad) and proteins separated on sodium dodecylsulfate 12.5% polyacrylamide gels and transferred to polyvinylidene difluoride membranes. Labeling of the transferred proteins were performed using the following primary antibodies: phospho-ATM (Rockland Immunochemicals),

Thr1989-ATR (Genetex #GT222), S296-Chk1 (Cell Signal#2349S), S516-Chk2 (Cell

Signal #2669S), cleaved caspase-9 (Asp353) (Cell Signal #95095S), PARP (Cell Signal

#9542S) and Ran (BD #610340). Membranes were incubated with horseradish peroxidase conjugated secondary antibodies (1:4,000) and detected by the ECL procedure (Amersham).

Statistical Analysis

The statistical significance of differences between data sets was analyzed by either Two-

Way ANOVA with Bonferroni correction or the log-rank (Mantel-Cox) test using the

GraphPad Prism software. P<0.05 was considered as statistically significant.

Results

Chk2 triggers cell death in normal tissues following DNA damage inflicted by whole-body gamma-radiation (WBR)

Previous findings have shown that Chk2 is a mediator of cell death in normal hematopoietic tissues following WBR (365). Indeed, mice proficient (wild type [WT]) and

84 lacking Chk2 (Chk2-/-) were subjected to WBR (5 Gy) and assessed for cell death in hematopoietic organs (Fig. 1A and B). Consistent with previous data, mice lacking Chk2 showed a reduced frequency of TUNEL-positive (+) cells in their white pulps following radiation exposure compared to littermates indicating ameliorated apoptosis following

DNA damage. Furthermore, Chk2-deficient mice were also able to tolerate increased expression of p53 while at the same time expressing reduced levels of the pro-apoptotic protein Bax in the white pulp as determined by IHC (Fig. 1A). This is consistent with prior investigations where Chk2 does not appear to be required for the stabilization of p53 protein following DNA damage. However, the induction of downstream target genes required for p53-dependent cell cycle arrest and cell death is hampered by loss of Chk2

(365). One p53-dependent gene that is regulated positively by Chk2 is MDM2. MDM2 is an E3 ubiquitin-protein ligase that promotes p53 degradation suggesting that p53 protein half-life may potentially increase Chk2 ablation. On the molecular level our IHC data are consistent with published data showing phosphorylation by Chk2 at sites in the N- and C- terminal domains of p53 is required for fulminant stimulation of transcription from p53- target genes (289).

Considering that myelosuppression is one of the most frequently occurring DLTs seen in patients subjected to chemotherapy, we aimed to further investigate if targeting of Chk2 could prevent this type of toxicity in mice. Bone marrow (BM) aspirates from wild-type and

Chk2 (Chk2-/-) mice subjected to WBR and were assessed for the presence of apoptotic cells with sub-G1 DNA content by flow cytometry. Indeed, BM aspirates from irradiated wild-type mice showed a higher percentage of nucleated cells with sub-G1 content

85 more readily underwent apoptosis compared to Chk2-/- consistent with previous studies where Chk2 was found to E1A MEF’s following treatment with the TOP2-inhibitor be a facilitator of chemotherapy- and IR-induced apoptosis etoposide (Figure 1D). Western blot assessment of PARP in MEF’s and normal mouse hematopoietic tissues [11, cleavage and cleavage of caspase-9 (CC9) showed that 15]. However, our data indicates that not all DNA Chk2-defcient MEF’s were increasingly protected from damaging chemotherapy triggers apoptosis and toxicity in PARP and caspase-9 cleavage following treatment with a Chk2-dependent manner. etoposide compared to MEF’s with intact Chk2 (Figure We also assessed Chk2-dependent killing of primary 1E). The ratio of cleaved PARP (p89) to full-length PARP splenocytes isolated from wild type (WT) and Chk2 “null” (p116) ratio (p89:p116) and the normalized band density (Chk2-/-) mice following treatment with etoposide (Figure of CC9 for the highest dose of etoposide was 1.25 and 2A). The dose-response analysis indicated that Chk2-/-

2.27 respectively for WT MEF’s compared to 0.37 and splenocytes displayed an approximately 3-fold higher IC50 0.32 respectively for Chk2-/- MEFs. This indicates that in vitro compared to WT splenocytes following etoposide- induction of etoposide-induced apoptosis is defcient treatment (10.18 [95%CI: 8.651-11.97] vs. 3.274 µg/ml following loss of Chk2. In comparison following treatment [95% CI: 2.522 - 4.250]) suggesting protection from Chk2- with the TOP1-poison CPT-11, only limited expression defciency over a broad dose-range of etoposide (Figure 2A, of PARP p89 and CC9 was observed indicating modest 2B and Table 1). In conclusion our data indicates that Chk2 onset of apoptosis downstream and canonical ATM- may trigger toxicity in normal cells following some DNA Chk2-p53 signaling following CPT-11. Moreover, little damaging chemotherapy but not others. relative protection was observed from Chk2-defciency withcompared respect to to the mice expression lacking of Chk2 cleaved (Fig. PARP 1B). (the Our interpretatChk2 triggerion ofcell this death finding following is that treatment Chk2 is with p89:p116 ratio was 0.09 and 0.07 respectively for wild TOP2-inhibitors but not following inhibitors of typea and mediator Chk2 “null” of cells apoptosis respectively also following in the 1.6 BM µM and thisTOP1 may contribute to Chk2-dependent of CPT-11) and CC9 (the normalized CC9 band density ofmyelosuppression 0.83 and 0.61 was observed following for WTE1A DNA anddamage Chk2- inflicted byTOP1-inhibitors ionizing radiation. trigger predominantly DNA /-E1A MEF’s respectively following 1.6 µM of CPT- single strand breaks (SSB) throughout the cell cycle that 11) (Figure 1E). To some extent our observations are are converted to DSBs during S-phase when escaping

Figure Figure 1: Chk2-targeting 3- 1: Chk2 protects targeting from protectstoxicity triggered from by toxicity DNA damage triggered in vivo andby inDNA vitro .damage A. Immunohistochemistry in-vivo forand TUNEL In- vitro(apoptosis), p53 and bax in spleens isolated from wild type and Chk2-/- mice that were subjected to 5 Gy of whole-body ionizing radiation (IR). Representative images are shown. Size bar is 40 µM. B. Flow cytometric analysis of the frequency of apoptotic cells with sub-G DNA present in BM aspirates from the femur and tibia of wild type and Chk2-/- mice subjected to either sham (control) A. Immunohistochemistry1 for TUNEL (apoptosis), p53 and bax expression in spleens isolated from wild- or 5 Gy of whole-body radiation (WBR). C. The sub-G1 EC50 was determined by fow cytometric analysis of E1A-immortalized wild typetype (WTE1A) and Chk2 and Chk2-/--/- mice (Chk2-/-E1A) that were MEF’ssubjected following to 5 twenty-four Gy of whole (24)- bodyhours ofioniz treatmenting radiation with different (IR). chemotherapy. Representative Error bars represent the standard deviation from the mean and N=3/treatment. *Indicates P<0.05, Student’s t-test. D. Immunocytochemistry analysis to imagesdetect the are presence shown. of p53Size (Cy3; bar red)is 40 and μM. cleaved B. Flow caspase-8 cytometric (Cy2; green) analysis in E1A-immortalized of the frequency wild of type apoptotic (WT) and cells Chk2-defcient with (Chk2-/-)sub-G1 MEF’s DNA subjected present to in treatment BM aspirates with two differentfrom the doses femur of etoposide and tibia (100 of µM wild and-type 1 mM). and DAPI Chk2 (blue)-/- mice was employed subjected to visualize to cellular nuclei. Representative images are shown. E. Western blot analysis for the detection of full-length PARP (p116), cleaved PARP (p89)either and shamcleaved (control) caspase-9 or(CC9; 5 Gy p37) of inwhole WTE1A-body and radiationChk2-/-E1A (WBR). MEF’s subjectedC. The subto treatment-G1 EC50 with was CPT-11 determined and etoposide. by owRan was used as a loading control. Densiometry was performed using the NIH Image J 1.45S software. The p89/p116 ratio indicates the densiometric cytometric analysis of E1A-immortalized wild type (WTE1A) and Chk2-/- (Chk2-/-E1A) MEF’s following ratio of cleaved PARP p89 to full-length PARP p116 in each lane respectively. Densiometric quantitation of CC9 p37 normalized to the loadingtwenty control-four (Ran) (24) arehours shown. of treatment with different chemotherapy. Error bars represent the standard deviation from the mean and N=3/treatment. *Indicates P<0.05, Student’s t-test. D. Immunocytochemistry analysis www.impactjournals.com/oncotarget 29522 Oncotarget to detect the presence of p53 (Cy3; red) and cleaved caspase-8 (Cy2; green) in E1A-immortalized wild type (WT) and Chk2-deficient (Chk2-/-) MEF’s subjected to treatment with two different doses of etoposide (100 μM and 1 mM). DAPI (blue) was employed to visualize cellular nuclei. Representative images are shown. E. Western blot analysis for the detection of full-length PARP (p116), cleaved PARP (p89) and cleaved caspase-9 (CC9; p37) in WTE1A and Chk2-/-E1A MEF’s subjected to treatment with CPT-11 and etoposide. Ran was used as a loading control. Densiometry was performed using the NIH Image J 1.45S software. The p89/p116 ratio indicates the densiometric ratio of cleaved PARP p89 to full-length PARP p116 in each lane respectively. Densiometric quantitation of CC9 p37 normalized to the loading control (Ran) is shown.

86 Chk2 triggers apoptosis in normal cells following exposure to select chemotherapeutics

To facilitate in-vitro screenings of new drugs where inhibition of Chk2 may be most beneficial to prevent DLT’s, we generated non-malignant E1A-immortalized MEF's from wild type (WTE1A) and Chk2-/- (Chk2-/-E1A) mice. In contrast to normal MEF’s, which undergo senescence following DNA damage, E1A-transfected MEF’s readily undergo p53-dependent apoptosis following such cellular stress (114,371). We hypothesized that

Chk2 may preferentially trigger cell death following DNA-damaging chemotherapeutics with certain genotoxic modes of action. Previous data have not addressed this facet of

Chk2-targeting in detail. Subsequently we decided to undertake a small screen to identify chemotherapy that triggered cell death predominantly in a Chk2-dependent manner.

Indeed, data from this screen indicated that the TOP2-inhibitors etoposide and doxorubicin triggered apoptosis in a Chk2-dependent manner (Fig.3- 1C). In contrast, the

TOP1-inhibitor CPT-11, the antimicrotubule agent taxol and the antimetabolite fluorouracil (5-FU) did not trigger cell death in E1A-immortalized MEF’s in a Chk2- depedent manner (Fig. 3-1C) Interestingly, the proteasome inhibitor MG132 triggered apoptosis in the MEF’s in a Chk2-dependent manner. Previous data have shown that

MG132 can force accumulation of nuclear p53 potentially indicating that cell death was p53- and Chk2-dependent following inhibition of proteasomal degradation. Consistent with data from our screen, immunocytochemistry indicated that WTE1A MEF’s expressed higher levels of p53, cleaved caspase-8 and more readily underwent apoptosis compared to Chk2-/-E1A MEF’s following treatment with the TOP2-inhibitor etoposide (Fig. 3-1D).

Western blot assessment of PARP cleavage and cleavage of caspase-9 (CC9) showed

87 that Chk2-deficient MEF’s were increasingly protected from PARP and caspase-9 cleavage following treatment with etoposide compared to MEF’s with intact Chk2(Fig. 3-

1E). The ratio of cleaved PARP (p89) to full-length PARP (p116) ratio (p89:p116) and the normalized band density of CC9 for the highest dose of etoposide was 1.25 and 2.27 respectively for WT MEF’s compared to 0.37 and 0.32 respectively for Chk2-/- MEFs.

These findings indicates that induction of etoposide-induced apoptosis is deficient following loss of Chk2. In comparison following treatment with the TOP1-poison CPT-11, only limited expression of PARP p89 and CC9 was observed indicating modest onset of apoptosis downstream and canonical ATM-Chk2-p53 signaling following CPT-11.

Moreover, little relative protection was observed from Chk2-deficiency with respect to the expression of cleaved PARP (the p89:p116 ratio was 0.09 and 0.07 respectively for wild type and Chk2 “null” cells respectively following 1.6 µM of CPT-11) and CC9 (the normalized CC9 band density of 0.83 and 0.61 was observed for WTE1A and Chk2-/-

E1A MEF’s respectively following 1.6 µM of CPT-11) (Fig. 3-1E). To some extent our observations are consistent with previous studies where Chk2 was found to be a facilitator of chemotherapy- and IR-induced apoptosis in MEF’s and normal mouse hematopoietic tissues (365,372). However, our data indicates that not all DNA damaging chemotherapy triggers apoptosis and toxicity in a Chk2-dependent manner.

We also assessed Chk2-dependent killing of primary splenocytes isolated from wild-type

(WT) and Chk2 “null” (Chk2-/-) mice following treatment with etoposide (Fig. 3-2A). The dose-response analysis indicated that Chk2-/- splenocytes displayed an approximately 3- fold higher IC50 in-vitro compared to WT splenocytes following etoposide-treatment (10.18

88 cell cycle checkpoints and subsequent DNA repair. In DNA damage [16, 17]. Indeed, this analysis showed contrast, TOP2-inhibitors have the capacity to trigger increased ATM-activation as evident by phosphorylation DSB’s in all phases of the cell cycle that are highly lethal at the ATM autophosphorylation site S1981 following in G2/M phase. In order, to address the generality of our treatment with TOP2-poisons (Figure 2D). However, fnding we assessed the dose-response parameters of a Chk2-activation, as evident by expression of S516 panel of TOP1- and TOP2-inhibitors in a primary mouse phosphorylated Chk2, occurred following treatment splenocytes isolated from wild type and Chk2-defcient with both TOP1 and TOP2-poisons and did not correlate

mice. Indeed, an IC50-shift was observed in Chk2-defcient well with ATM-activation (Figure 2D). Furthermore, splenocytes predominantly following treatment with pharmacologic inhibition of Chk2 using Chk2 inhibitor

TOP2-inhibitors (Figure 2C). In contrast a limited IC50- II (CI2) suggested that in the case of CPT-11, Chk2 shift (or even sensitization in the case of CPT-11) was targeting might result in increased ATM activation observed following treatment with TOP1-inhibitors. This perhaps as a result of increased DNA damage. Thus gave further support to the hypothesis that Chk2 controls in rapidly dividing cancer cells Chk2 may become toxicity following certain types of DNA damage such as activated in an ATM-independent manner. that[95%CI: inficted by 8.651 TOP2-poisons.-11.97] vs. 3.274 µg/ml [95% CI: 2.522We - 4.250]) also addressed suggesting activation protectionof the ATM-Chk2 from and It has previously been shown that DSB is the ATR-Chk1 signaling following treatment with equimolar predominant activator the canonical ATM-Chk2-p53 and equitoxic doses of CPT-11 and etoposide in normal pathwayChk2 -butdeficiency Chk2 may overalso be a activatedbroad dose by ATR-range and of etoposideprimary human (Fig. bone3-2A, marrow B). mono In conclusion, nucleated cells (hBM-our SSB’s as a result of e.g. collapsed replications forks MNC). Human BM-MNC’s show limited proliferation in duringdata S-phase. indicate In supportthat Chk2 of this, may western trigger blot analysis toxicity in normalculture cells where following more than ninety-fvesome DNA (95) damaging percent of the

of the p53-profcient and exponentially growing H460 cells remain in G0-G1 phase of the cell cycle (data not lung cancer cell line treated with the panel of TOP1- shown). We frst determined the BM-MNC’s IC to CPT- chemotherapy but not others. 50 and TOP2-inhibitors was performed. It has previously 11 and etoposide and then subjected them to short-term been shown that this cell line elicit a robust DDR (6 hrs) treatment to equitoxic doses of the TOP1- and including phosphorylation of Chk-proteins following TOP2-poison respectively that did not trigger cell death

Figure 2: Chk2 is a mediator of toxicity triggered by TOP2-poisons. A. The viability of primary mouse splenocytes isolated fromFigure wild type 3 (WT)- 2: andChk2 Chk2-/- is micea mediat followingo treatmentr of toxicity with etoposide triggered in vitro was by assessed TOP2 by -thepoisons CellTiter-Glo ®assay. B. The dose-

response IC50 for primary WT and Chk2-/- mouse splenocytes following long-term (72-hrs) treatment with etoposide was determined by the ® CellTiter-GloA. The viability assay. Error of primarybars represent mouse the standard splenocytes error from isolated the mean. from N=3/treatment wild-type and (WT) genotype. and C. Chk2 The IC-/50- -shiftmice was following determined for TOP1- and TOP2-inhibitors in primary splenocytes isolated from littermate Chk2-/- and WT mice. Error bars represent the standard errortreatment from the mean. with N=3/treatmentetoposide in and-vitro genotype. was assessedD. Protein expression by the CellTiter as detected-Glo®assay. by western blotting B. The of phosphorylated dose- response ATM and IC50 Chk2 at theirfor primary autophosphorylation WT and Chk2 sites -S1981/- mouse and S516 splenocytes respectively following following long6 hours-term of treatment (72-hrs) of treatment the human lungwith cancer etoposide cell line was H460 with TOP1 and TOP2-inhibitors. E. The IC50’s (µg/mL) for primary human bone marrow mono nucleated cells (hBM-MNC) treated with thedetermined TOP1- and TOP2-inhibitors by the CellTiter CPT-11 -(CPT-11)Glo® assay. and etoposide Error for bars 24-hrs representrespectively were the determined standard using error the fromCellTiter-Glo® the mean. system. The error bars represent the standard error from the shown mean. *P<0.05, Student’s t-test. F. Western blot of lysates from hBM-MNC’s N=3/treatment and genotype. C. The IC50-shift was determined for TOP1- and TOP2-inhibitors in primary treated with isomolar (0.13 mM) and isotoxic (IC50) doses (0.13 mM and 0.53 mM) of etoposide and CPT-11 for 6-hrs respectively. The expressionssplenocytes of ATM, isolated Chk2, ATR from and littermate Chk1 phosphorylated Chk2-/- and at the WT indicated mice. autophosphorylationError bars represent sites werethe standardassessed. Gamma-(γ) error from H2AX the is a downstream substrate for the ATM kinase and marker for the presence of DSB. Ran was used as a loading control. mean. N=3/treatment and genotype. D. Protein expression as detected by Western blotting of www.impactjournals.com/oncotargetphosphorylated ATM and Chk2 at their autophosphorylation29523 sites S1981 and S516 respectively followingOncotarget 6 hours of treatment of the human lung cancer cell line H460 with TOP1 and TOP2-inhibitors. E. The IC50’s (μg/mL) for primary human bone marrow mono nucleated cells (hBM-MNC) treated with the TOP1- and TOP2-inhibitors CPT-11 (CPT-11) and etoposide for 24-hrs respectively were determined using the CellTiter-Glo® system. The error bars represent the standard error from the shown mean. *P<0.05, Student’s t-test. F. Western blot of lysates from hBM-MNC’s treated with isomolar (0.13 mM) and isotoxic (IC50) doses (0.13 mM and 0.53 mM) of etoposide and CPT-11 for 6-hrs respectively. The expressions of ATM, Chk2, ATR and Chk1 phosphorylated at the indicated autophosphorylation sites were assessed.

89 Gamma-(γ) H2AX is a downstream substrate for the ATM kinase and marker for the presence of DSB. Ran was used as a loading control.

Chk2 triggers cell death following treatment with TOP2-inhibitors but not following inhibitors of TOP1

TOP1-inhibitors trigger predominantly DNA single strand breaks (SSB) throughout the cell cycle that are converted to DSBs during S-phase when escaping cell cycle checkpoints and subsequent DNA repair. In contrast, TOP2-inhibitors have the capacity to trigger DSB’s in all phases of the cell cycle that are highly lethal in G2/M phase. To address the generality of our finding we assessed the dose-response parameters of a panel of TOP1- and TOP2-inhibitors in primary mouse splenocytes isolated from wild- type and Chk2-deficient mice. Indeed, an IC50-shift was observed in Chk2-deficient splenocytes predominantly following treatment with TOP2-inhibitors (Fig. 3-2C). In contrast, a limited IC50-shift (or even sensitization in the case of CPT-11) was observed following treatment with TOP1-inhibitors. This provided further support to the hypothesis that Chk2 controls toxicity following certain types of DNA damage such as that inflicted by TOP2-poisons.

It has previously been shown that DSB is the predominant activator the canonical ATM-

Chk2-p53 pathway but Chk2 may also be activated by ATR and SSB’s as a result of e.g. collapsed replications forks during S-phase. In support of this, Western blot analysis of the p53-proficient and exponentially growing H460 lung cancer cell line treated with the panel of TOP1- and TOP2-inhibitors was performed. It has previously been shown that this cell line elicits a robust DDR including phosphorylation of Chk-proteins following DNA

90 damage (373,374). Indeed, this analysis showed increased ATM-activation as evident by phosphorylation at the ATM autophosphorylation site S1981 following treatment with

TOP2-poisons (Fig. 3-2D). However, Chk2-activation, as evident by expression of S516 phosphorylated Chk2, occurred following treatment with both TOP1 and TOP2-poisons and did not correlate with ATM-activation (Fig. 3-2D). Furthermore, pharmacologic inhibition of Chk2 using Chk2 inhibitor II (CI2) suggested that in the case of CPT-11, Chk2 targeting might result in increased ATM activation perhaps as a result of increased DNA damage. Thus, in rapidly dividing cancer cells Chk2 may become activated in an ATM- independent manner.

We also addressed activation of the ATM-Chk2 and ATR-Chk1 signaling following treatment with equimolar and equitoxic doses of CPT-11 and etoposide in normal primary human bone marrow mono nucleated cells (hBM-MNC). Human BM-MNC’s show limited proliferation in culture where more than ninety-five (95) percent of the cells remain in G0-

G1 phase of the cell cycle. We first determined the BM-MNC’s IC50 to CPT-11 and etoposide and then subjected them to short-term (6 hrs) treatment to equitoxic doses of the TOP1- and TOP2-poison respectively that did not trigger cell death (Fig. 3-2E and F).

Indeed, Western blot data indicated that etoposide more robustly activated ATM-Chk2 signaling compared to CPT-11 (Fig. 2F). Moreover, at equitoxic doses (0.13 mM) etoposide was more potent in the triggering of DSB as evident of an increased expression of gamma-H2AX following such treatment (Fig. 3-2F). However, CPT-11 was in comparison to its relative potency to trigger expressions of phosphorylated species of

ATM and Chk2 increasingly potent in activating ATR at both isomolar (0.13 mM) and

91 equitoxic doses (0.53 mM) and Chk1 S296 autophosphorylation at equitoxic doses. Thus, this finding suggests that etoposide increasingly triggers a DDR that activates ATM and

Chk2 in the bone marrow whereas the DDR following CPT-11 involves a more isolated activation of ATR and Chk1 in-vitro.

Chk2 controls toxicity in mice in-vivo following the TOP2-poison etoposide

To assess the importance of Chk2 in mediating DLT’s following chemotherapy, we employed an in-vivo mouse model for treatment with the TOP2 poison etoposide. This model was based on a clinical repeat-dose etoposide-treatment protocol, where etoposide is administered three (3) to five (5) times during a period of five days. Surface area doses of etoposide employed in patients were converted to doses for retro-orbital

IV injection in mice according to Freireich et al. (375). Interestingly, when etoposide was injected five times during the time course of 5 days (5x32 mg/kg bw IV), no significant contribution of Chk2 to toxicity in-vivo was observed because both wild-type and mice lacking the Chk2 gene succumbed following this treatment schedule (Fig. 3-3A).

However, when the accumulated dose of etoposide was reduced to 96 mg/kg (3x32 mg/kg bw IV) and given over 5 days with one day in between, a significant protection was observed in mice lacking Chk2 as well as in mice lacking the p53-responsive genes p21

(p21-/-; a cell cycle inhibitory protein) and Puma (puma-/-; a pro-apoptotic BH3-only protein) (Fig. 3-3B). Furthermore, wild-type and mice heterozygous for Chk2 showed reduced levels of circulating white blood cells (WBC) following treatment with two doses of etoposide (2x32 mg/kg bw) compared to mice that were completely null for Chk2 (Fig.

3-3C). As expected, toxicity in this model was associated with BM and gastrointestinal

92 treatment in vivo (Figure 5I). Unfortunately, delivering through Chk2 since splenocytes lacking the Chk2 gene

the Chk2i’s by surgically implanted osmotic pumps had an approximately 30-fold higher IC50 compared to in mice did not improve activity in this model and that of wild type splenocytes (Figure 2C and 2D). The it also became apparent that some of these Chk2i’s Z-scores for the functional screens of the wild type and triggered increased weight loss in the etoposide-treated Chk2-/- splenocytes were 0.89 and 0.87 respectively mice (data not shown). Therefore, in order to identify indicating ‘excellent’ sensitivity in both assays [20]. lead Chk2i with in vivo countermeasure activity we Based on the results from the combined screens we chose designed a combined computational, functional and a cut-off for protection at 1.3-fold change (f.c.) (Figure cell-based screen for such compounds (Figure S1). 5B). Furthermore, we also assessed the compounds of The chemical structures present in the Diversity Set II the condensed library for inhibition of Chk2 activity in a (NCI/NTP) were compared to the crystal structure of cell-free kinase assay (Figure 5D and Figure S1). ADP bound to Chk2 from the Protein Data Base (PDB: A total of seven (7) compounds were found 2CN5). From this approach we were able to condense to improve the survival of wild type but not Chk2- aplasia (Fig. 3-3D). Consistent with the previous in-vitro findings, when a toxic (EC50) the number of candidate compound structures from /- splenocytes following treatment with daunorubicin 1,364 to 299 (22% of the initial library size) (Figure (Figure 5C). Three (3) of these compounds were removed 5A).dose Subsequently, of the we TOP1 used -thispoison condensed CPT compound-11 was administeredfrom further to assessment Chk2-/ -based mice, on information no apparent retrieved library to screen wild type (WT) and Chk2-/- mouse from e.g. the TOXLINE database (http://toxnet.nlm.nih. splenocytesprotection for protectionwas observed (improved (Fig. cell 3survival)-3E). In to contrast,gov/) such indicating mice unfavorable appeared toxicities. to be Forsensitized example, the

isotoxic doses (defned from genotype-specifc IC50’s) caffeine derivative NSC524385 was somewhat less potent of thewhen TOP2-poison compared daunorubicin to wild-type (Figure mice. 5B andThus, 5C). our datathan are the consistent other lead compounds. with the notion Furthermore, that Chk2caffeine is The inclusion of cells devoid of Chk2 was performed a potent inhibitor of ATM, a kinase upstream of Chk2 in ordertrigger to somit DLT compoundss following that offeredexposure protection to TOP2 in -poisons.required In for contrast, fulminant Chk2 activity may notfollowing trigger DSB’s a manner independent of Chk2. Furthermore, in order and triggers toxicity in laboratory animals. NSC106570 to trigger potent Chk2-activation in the splenocytes we is a muscle relaxant and a psychotropic drug which usedtoxicity the TOP2-poison following daunorubicin. TOP1-poisons Our data in- indicatevivo. makes this lead compound less suitable for further in vivo that daunorubicin was most potent at killing splenocytes assessment.

Figure 3: Chk2 triggers dose-limiting toxicity in mice in vivo following etoposide. A. Wild type and Chk2-/- mice were treated with Figureetoposide once3- 3 daily:Chk2 for fve triggers (5) days at thedose indicated limiting doses. B. toxicity Wild type, Chk2-/-in mice and micein-vivo lacking following the p53 responsive apoptosis- and cell cycleetoposide regulating genesexposure puma (puma-/-) and p21 (p21-/-) were treated every other day for 5 days with etoposide. Statistical signifcant differences (P<0.05) in survival relative to treated wild type mice were analyzed by log rank (Mantel Cox) tests. C. Hematology assessment in wild type (WT), Chk2+/- and Chk2-/- mice suggest a Chk2-dependent reduction in the white blood cell count (WBC) following etoposide.

Each data point represents data from one mouse. Black lines indicate the median of each group. “*” Indicates a P<0.05 and “**”indicate a P<0.01 by 1-way ANOVA analysis with Bonferonni correction. D. Histology (H/E staining) of the bone marrow (BM), colon (C) and small intestineA. Wild (SI) -oftype mice and succumbing Chk2-/ -to mice etoposide-treatment. were treated Sizewith bars etoposide represent once100 µm daily (25x) for and (5)25 µm days (100x). at the E. Survivalindicated of wild doses. type and

Chk2-/-B. Wild mice- treatedtype, Chk2 with EC-/-50 and-doses mice of CPT-11. lacking the p53 responsive apoptosis- and cell cycle regulating genes puma www.impactjournals.com/oncotarget(puma-/-) and p21 (p21-/-) were treated every other29525 day for 5 days with etoposide. Statistically signiOncotargetficant differences (P<0.05) in survival relative to treated wild-type mice were analyzed by log rank (Mantel Cox) tests. C. Hematology assessment in wild-type (WT), Chk2+/- and Chk2-/- mice suggest a Chk2-dependent reduction in the white blood cell count (WBC) following etoposide. Each data point represents data from one mouse. Black lines indicate the median of each group. “*” Indicates a P<0.05 and “**”indicate a P<0.01 by 1-way ANOVA analysis with Bonferonni correction. D. Histology (H/E staining) of the bone marrow (BM),

93 colon (C) and small intestine (SI) of mice succumbing to etoposide-treatment. Size bars represent 100 μm (25x) and 25 μm (100x). E. Survival of wild type and Chk2-/- mice treated with EC50-doses of CPT-11.

Pharmacologic inhibition of Chk2 prevents killing of normal cells in-vitro by the

TOP2-poison etoposide

To verify if our findings with regard to genetic ablation of Chk2 translates to pharmacologic

(Chk2i) targeting of Chk2’s kinase function we treated human normal fibroblasts with etoposide and a panel of Chk2i’s. Our data indicate that CI3 (363) protected MCR5 cells from the toxic effect of etoposide following treatment with 165 µg/ml of the chemotherapeutic (Fig. 4A and B). To address if the pharmacologic inhibitors would phenocopy splenocytes lacking the Chk2 gene we subjected wild-type splenocytes to etoposide-treatment in the absence and presence of PV1019 (362), CI2 (361) and CI3

(Fig. 3-4A and C). Our finding indicates that CI3 (IC50: 63.47 µg/mL, 95%CI: 51.63-78.03) offered mouse splenocytes a 6-fold protection from etoposide-induced toxicity compared to vehicle (IC50: 10.29 µg/mL, 95%CI: 7.484-14.15) (Table 2). We also subjected mouse splenocytes to ionizing radiation (IR) in the absence and presence of CI3 in-vitro (Fig. 3-

4D). Here we observed a modest but statistically significant protection by CI3 from IR- induced killing of the splenocytes compared to vehicle (control) (36.7 ± 0.53 vs 29.8 ±

0.62 % respectively). In contrast, CI3 completely protected the splenocytes from etoposide-induced killing. Collectively we conclude that pharmacologic targeting of Chk2 by Chk2i such as CI3 may be particularly potent in protecting from cell death from TOP2- poisons such as etoposide. Thus, it may be a feasible approach to pharmacologically inhibit Chk2 to countermeasure DLT’s following etoposide treatment.

94 NSC105171 is a pharmacologic countermeasure As previously noted, neither of the three Chk2i’s that show to etoposide-induced toxicity in vivo activity in cells in vitro (CI2, CI3 and PV1019) protected mice from etoposide-induced toxicity compared to mice The four (4) remaining compounds (red box) subjected to vehicle (Figure 5I). In contrast, NSC105171 were re-evaluated for inhibition Chk2-activity in the did provide signifcant protection (P<0.05; log rank test, kinase assay (Figure 5C and 5D). Only NSC105171 N=5/treatment group) compared to vehicle-treated mice (E) were found to suffciently inhibit CHK2-dependent and this fnding was correlated improved bone marrow phosphorylation of a synthetic oligopeptide encoding for histology (Figure 5I and ‘data not shown’). This indicates a Chk1/2 -phospho site of CDC25A, in the presence of that the combined in silico and functional screening recombinant Chk2 and ATP (Figure 5E-5G). NSC105171 approach may identify lead Chk2i for in vivo application

showed an IC50 that was approximately eight (8) times to countermeasure DLT’s from TOP2-poisons. higher than the hymenialdesine derivative CI3 [9] (Figure 5F and 5G). This indicates that in comparison, NSC105171 DISCUSSION is a less potent Chk2i in vitro. In A549 cells, NSC105171 inhibited the expression of Chk2 phosphorylated at the Inhibition of Chk2 may offer a novel strategy autophosphorylation site Serine 516 (S516) following of clinical intervention to prevent DLT’s from DNA treatment with daunorubicin indicating cell in vivo activity damaging chemotherapy. Inhibition of Chk2 may curtail (Figure 5H). We decided to test the effcacy of NSC105171 propagation of DDR signals that trigger hematopoietic in preventing etoposide-induced toxicity in mice in vivo. stem cell death and BM aplasia. This molecularly targeted

Figure 3- 4: Pharmacological inhibitors of Chk2 (Chk2i) protect normal human Figure 4: Pharmacologic inhibitors of Chk2 (Chk2i) protects normal human and mouse cells from etoposide-induced killing.and A. mouse The chemical cells structures from of etoposide PV1019, CI2 -andinduced CI3 B. Quantitation killing. of MRC5 cell survival following etoposide-treatment in the presence and absence (vehicle) of CI3. The statistical signifcance was determined by the two-way ANOVA test with Bonferroni multiple comparison correction. The relevant ‘P’ values are indicated for N=6. C. Twenty-four (24) hour dose-response CellTiter-Glo® A. The chemical structures of PV1019, CI2 and CI3 B. Quantitation of MRC5 cell survival following assay assessing the survival of primary mouse splenocytes isolated from C57BL6J mice in the presence and absence of Chk2i following

treatmentetoposide with- etoposide.treatment The in curve the ft presence was analyzed and using absence the GraphPad (vehicle) Prism 5.0 of software. CI3. The IC50’s statisticalwith non-overlapping significance 95%-confdence was intervals (CI) were considered statistically signifcantly different. An N=3/treatment was employed for this particular experiment. D. The celldetermined viability of mouseby the splenocytes two-way relativeANOVA untreated test with control Bonferroni was determined multiple by thecomparison CellTiter-Glo correction.® assay. The Theassay relevantwas performed ‘P’ at twenty-twovalues are (22) indicated hours following for N=6. treatment C. Twenty with etoposide-four (24)(18 µg/mL) hour doseand 10-response Gy of ionizing CellTi radiationter-Glo® (IR) inassay the presence assessing and absence the of the Chk2i CI3. The splenocytes were pre-treated with CI3 two (2) hours prior to initiation of treatments. Error bars represent the standard errorsurvival from the of mean. primary Statistical mouse assessment splenocytes was performed isolated by from a two-way C57BL6J ANOVA mice test inwith the Bonferroni presence multiple and absencecomparison of correction. Chk2i www.impactjournals.com/oncotargetfollowing treatment with etoposide. The curves were29526 analyzed using the GraphPad Prism 5.0 software.Oncotarget IC50’s with non-overlapping 95%-confidence intervals (CI) were considered statistically significantly different. An N=3/treatment was employed for this particular experiment. D. The cell viability of mouse splenocytes relative untreated control was determined by the CellTiter-Glo® assay. The assay was performed at twenty-two (22) hours following treatment with etoposide (18 μg/mL) and 10 Gy of ionizing radiation (IR) in the presence and absence of the Chk2i CI3. The splenocytes were pre-treated with CI3 two (2) hours prior to initiation of treatments. Error bars represent the standard error from the mean. Statistical assessment was performed by a two-way ANOVA test with Bonferroni multiple comparison correction.

95 A combined in silico and functional compound screen identifies NSC105171 as a

Chk2i with toxicity countermeasure activity

Although the Chk2i’s showed promising activity following treatment with etoposide in- vitro, they failed to prevent toxicity following etoposide-treatment in-vivo (Fig. 5I).

Unfortunately, delivering the Chk2i’s by surgically implanted osmotic pumps in mice did not improve activity in this model and it also became apparent that some of these Chk2i’s triggered increased weight loss in the etoposide-treated mice. Therefore, in order to identify a lead Chk2i with in-vivo countermeasure activity we designed a combined computational, functional and cell-based screen for such compounds (Fig.3-6). The chemical structures present in the Diversity Set II (NCI/NTP) were compared to the crystal structure of ADP bound to Chk2 from the Protein Data Base (PDB: 2CN5). From this approach, we were able to condense the number of candidate compound structures from

1,364 to 299 (22% of the initial library size) (Fig. 3-5A). Subsequently, we used this condensed compound library to screen wild-type (WT) and Chk2-/- mouse splenocytes for protection (improved cell survival) to isotoxic doses (defined from genotype-specific

IC50’s) of the TOP2-poison daunorubicin (Fig. 3-5B and C). The inclusion of cells devoid of Chk2 was performed to omit compounds that offered protection in a manner independent of Chk2. Furthermore, to trigger potent Chk2-activation in the splenocytes we used the TOP2-poison daunorubicin. Our data indicate that daunorubicin was most potent at killing splenocytes through Chk2 since splenocytes lacking the Chk2 gene had an approximately 30-fold higher IC50 compared to that of wild-type splenocytes (Fig. 3-

2C and D). The Z-scores for the functional screens of the wild-type and Chk2-/- splenocytes were 0.89 and 0.87, respectively, indicating ‘excellent’ sensitivity in both

96 assays(376). Based on the results from the combined screens we chose a cut-off for

Table 2: Splenocyte IC50 to etoposide without and with Chk2i protection at 1.3-fold change (f.c.) (Fig. 3-5B). Furthermore, we also assessed the Splenocytes IC50 (µg/mL) 95%-CI IC50 ratio 95%-CI Vehicle 10.29 7.484 – 14.15 - - compounds of the condensed library for inhibition of Chk2 activity in a cell-free kinase PV-1019 (10 µM) 11.52 8.996 – 14.76 1.275 0.8781 – 1.673 CI2assay (10 µM) (Fig. 3-5D and Fig.5.952 S1). 5.149 – 6.881 0.6690 0.4605 – 0.8776 CI3 (10 µM) 63.47 51.63 – 78.03 6.949 4.773 – 9.126

FigureFigure 5: A 3combined-5: A combined in silico and in genetic-silico screen and identifes genetic NCS105171 screen identifies- a novel lead NSC105171, Chk2i with countermeasure a novel activitylead inChk2ki vitro and with in vivo activity. A. Construction in-vitro of in silico and docking in-vivo protocols using the high-resolution crystal structure of ADP bound to the Chk2 kinase. B. Identifed lead compounds from the condensed compound library that protects mouse primary splenocytes from cell death following treatment with the TOP2-poison daunorubicin. The cutoff was set to 1.3-fold (indicated in red) for protection. C. Four lead compounds A. Construction that show Chk2-dependent of in silico docking protection protocols from daunorubicin-dependent using the high-resolution cell death. D. crystal The kinase structure inhibitory of ADPactivity bound of the leadto Chk2i’sthe wasChk2 assessed kinase. in a B. cell-free Identified Chk2 kinaselead compounds assay. E. The chemicalfrom the structure condensed of the leadcompound compound library NSC105171 that protects(ptu-23) is mouse depicted. Graphs showing dose-dependent inhibition of Chk2 kinase activity by NSC105171 F. and CI3 G. in the Chk2 kinase assay. H. NSC105171 inhibitsprimary the expression splenocytes of Chk2 from phosphorylated cell death followingat the autophosphorylation treatment with site the Serine TOP2 516-poison in human daunorubicin. A549 lung cancer The cells cutoff following was treatment with daunorubicin. I. Survival of mice subjected to repeat-dose exposure of etoposide and co-treated with commercially available Chk2i’sset and to NSC105171. 1.3-fold (indicated in red) for protection. C. Four lead compounds that show Chk2-dependent protection from daunorubicin-dependent cell death. D. The kinase inhibitory activity of the lead Chk2i’s was strategy takes advantage of a common differences between Chk2 may stimulate DNA repair and exercise increased assessed in a cell-free Chk2 kinase assay. E. The chemical structure of the lead compound NSC105171 normal and cancer cells in their capacity to trigger p53- control of cell cycle progression through G2/M transition dependent(ptu-23) cell is death depi cted.following Graphs DNA-damaging showing dose therapies-dependent inhibitionin the absence of Chk2 of p53, kinase something activity that by NSC105171may diminish F.the since cancer cells frequently carry inactivating p53- potency of DNA damaging therapy in some cancer cells. and CI3 G. in the Chk2 kinase assay. H. NSC105171 inhibits the expression of Chk2 phosphorylated at the mutations [21]. Moreover, some cancers frequently show Subsequently, Chk2 inhibition may both sensitize cancer highautophosphorylation expression of T68 phosphorylated site Serine 516 Chk2 in humanindicating A549 lung cellscancer to DNA cells damage following and treatment at the same with time daunorubicin. prevent DLT’s. high activity of Chk2. A genomic and proteomic study of Our data indicate that Chk2 induces cell death I. Survival of mice subjected to repeat-dose exposure of etoposide and co-treated with commercially the NCI-60 cell line panel indicated that 12% of the panel following DNA damage in a highly chemotherapeutic- had available high endogenous Chk2i’s activationand NSC105171. of Chk2 and this was specifc manner. For example, targeting of Chk2 may always associated with loss of p53 [22]. It has previously have limited value as a toxicity countermeasure following been shown that cancer cells lacking functional p53 may treatment with TOP1-inhibitors since loss of Chk2 did be increasingly sensitive to damage that trigger G2/M not protect E1A-immortalized MEF’s from cell death checkpoints such as that by anti-mitotics [23, 24]. Indeed, following treatment with CPT-11 nor did somatic loss97 of www.impactjournals.com/oncotarget 29527 Oncotarget www.impactjournals.com/oncotarget/ Oncotarget, Supplementary Materials 2016

FigureSupplementary 3- 6: Schematic Figure S1: Schematic showing showing thethe development development of a screening of a strategy screening to isolate strategy novel pharmacologic to isolateChk2 inhibitors novel withpharmacological in vivo activity. The approach Chk2i shows with an in silicoin- vivodocking activityprotocol employed on the Diversity Set II (A) based on the high resolution crystal structure of ADP bound to Chk2. This was performed to help condense this compound library and allow for quicker identifcation of lead compounds in subsequent a cell-based screen. This functional screen utilized employed a cell-free kinase assay in parallel to assess the kinase inhibitory potency of the lead compounds (B.1). Furthermore, normal primary mouse splenocytes that with an intact (CHK2+/+) and mutated and functionally inactivated chek2 gene (CHK2-/-) (B.2) in order to help identify small molecule Theinhibitors approach of Chk2 shows (Chk2i) an with in minimal-silico off-target docking effect. protocol The candidates employed were assessed on the for Diversityfavorable pharmacokinetics Set II (A) based and toxicology on the data high - resolutionbased on acrystal priori knowledge structure through of theADP use boundof the TOXNET to Chk2. data base This (C) was followed performed by further experimental to help condense assessment of this activity compound (D). library and allow for quicker identification of lead compounds in subsequent a cell-based screen. This functional screen utilized employed a cell-free kinase assay in parallel to assess the kinase inhibitory potency of the lead compounds (B.1). Furthermore, normal primary mouse splenocytes that with an intact (CHK2+/+) and mutated and functionally inactivated chek2 gene (CHK2-/-) (B.2) To help identify small molecule inhibitors of Chk2 (Chk2i) with minimal off-target effect. The candidates were assessed for favorable pharmacokinetics and toxicology data based on a priori knowledge through the use of the TOXNET data base (C) followed by further experimental assessment of activity (D).

A total of seven (7) compounds were found to improve the survival of wild-type but not

Chk2-/- splenocytes following treatment with daunorubicin (Fig. 3-5C). Three (3) of these compounds were removed from further assessment based on information retrieved from e.g. the TOXLINE database (http://toxnet.nlm.nih.gov/) indicating unfavorable toxicities.

98 For example, the caffeine derivative NSC524385 was somewhat less potent than the other lead compounds. Furthermore, caffeine is a potent inhibitor of ATM, a kinase upstream of Chk2 required for fulminant Chk2 activity following DSB’s and triggers toxicity in laboratory animals. NSC106570 is a muscle relaxant and a psychotropic drug which makes this lead compound less suitable for further in-vivo assessment.

NSC105171 is a pharmacologic countermeasure to etoposide-induced toxicity in- vivo

The four (4) remaining compounds (red box) were re-evaluated for inhibition Chk2-activity in the kinase assay (Fig. 3-5C and 3-5D). Only NSC105171 (E) was found to sufficiently inhibit CHK2-dependent phosphorylation of a synthetic oligopeptide encoding for a

Chk1/2 -phosphorylation site of CDC25A, in the presence of recombinant Chk2 and ATP

(Fig. 3-5E-G). NSC105171 showed an IC50 that was approximately eight (8) times higher than the hymenialdesine derivative CI3 (363) (Fig. 3-5F and G). This indicates that in comparison, NSC105171 is a less potent Chk2i in-vitro. In A549 cells, NSC105171 inhibited the expression of Chk2 phosphorylated at the auto-phosphorylation site Serine

516 (S516) following treatment with daunorubicin indicating cell in-vivo activity (Fig. 5H).

We tested the efficacy of NSC105171 in preventing etoposide-induced toxicity in mice in- vivo. As previously noted, neither of the three Chk2i’s that show activity in cells in-vitro

(CI2, CI3 and PV1019) protected mice from etoposide-induced toxicity compared to mice subjected to vehicle (Fig. 3-5I). In contrast, NSC105171 provided significant protection

(P<0.05; log rank test, N=5/treatment group) compared to vehicle-treated mice and this finding was correlated to improved bone marrow histology (Fig. 3-5I). This indicates that

99 the combined in silico and functional screening approach may identify lead Chk2i for in- vivo application to counteract DLT’s from TOP2-poisons.

Discussion

Inhibition of Chk2 may offer a novel strategy of clinical intervention to prevent DLT’s from

DNA damaging chemotherapy. Inhibition of Chk2 may curtail propagation of DDR signals that trigger hematopoietic stem cell death and BM aplasia. This molecularly targeted strategy takes advantage of common differences between normal and cancer cells in their capacity to trigger p53-dependent cell death following DNA-damaging therapies as cancer cells frequently carry inactivating p53-mutations(377). Moreover, some cancers frequently show high expression of T68 phosphorylated Chk2 indicating high activity of

Chk2. A genomic and proteomic study of the NCI-60 cell line panel indicated that 12% of the panel had high endogenous activation of Chk2 and this was always associated with loss of p53 (378). It has previously been shown that cancer cells lacking functional p53 may be increasingly sensitive to damage that triggers G2/M checkpoints such as that by anti-mitotic agents (379,380). Indeed, Chk2 may stimulate DNA repair and exercise increased control of cell cycle progression through G2/M transition in the absence of p53, something that may diminish the potency of DNA damaging therapy in some cancer cells.

Subsequently, Chk2 inhibition may both sensitize cancer cells to DNA damage and at the same time prevent DLT’s.

Our data indicate that Chk2 induces cell death following DNA damage in a highly chemotherapeutic-specific manner. For example, targeting of Chk2 may have limited

100 value as a toxicity countermeasure following treatment with TOP1-inhibitors because loss of Chk2 did not protect E1A-immortalized MEF’s from cell death following treatment with

CPT-11 nor did germline loss of the Chk2 gene protect mice from lethal toxicity of CPT-

11. Furthermore, loss of Chk2 did not significantly change the in-vitro dose-response relationship of splenocytes to a panel of TOP1-inhibitors. These observations stand in contrast to data on a panel of TOP2-inhibitors where loss of Chk2 in mouse splenocytes caused a profound protective increase of the IC50’s. Moreover, in-vivo data indicated that mice devoid of Chk2 are protected from lethal toxicities of the TOP2-inhibitor etoposide.

Based on this we conclude that pharmacologic inhibition of Chk2 to prevent DLT’s may be most efficient following chemotherapy such as TOP2-inhibitors.

In contrast to TOP1-inhibitors, TOP2-inhibitors trigger DSB directly through inhibition of

TOP2alpha and may rely on canonical ATM-Chk2-p53 signaling pathway throughout the cell cycle (381). Inhibition of TOP1 may trigger toxicity primarily through the generation of

SSB and the generation of pathological levels of single stranded DNA (ssDNA) following replication fork stalling. As a consequence of this, this type of cellular DDR may rely increasingly on the activation of ATR and Chk1 and less on activation of ATM and Chk2

(299,343,382). This would be particularly apparent in organs such as the BM, that under unstressed conditions display a low frequency of cycling cells. In such organs, SSB’s inflicted by TOP1-inhibitors may be slowly and infrequently converted to DSB’s (383).

This is consistent with our data on primary human BM cells that indicate limited activation of ATM and Chk2 following treatment with CPT-11 compared to treatment with equitoxic doses of etoposide. In the absence of Chk2, cells might be less prone to instigate p53-

101 dependent G1-S and intra-S checkpoints (384). Subsequently, loss of Chk2 may on the contrary render normal tissues with frequently cycling cells such as the GI tract susceptible to TOP1-inhibitors that rely on S-phase progression to generate lethal DSB’s.

To isolate novel Chk2i suitable as countermeasures to DLT’s of TOP2-poison, we undertook a combined in silico and functional compound screen. Our collective efforts suggested that the carbanilide-derivative NSC105171 inhibits the kinase activity of Chk2 and protects Chk2-proficient mouse splenocytes from daunorubicin-dependent killing.

Interestingly, NSC105171 is also known as the carbanilide derivative ptu-23, an antiviral compound that protects mice from lethal infections of several Coxsackie virus strains

(385,386). Furthermore, NSC105171 appears to have a permissive safety profile with a reported mouse LD50 of 1,000 mg/kg bw. The combination of a permissive toxicity profile and in-vivo activity of NSC105171 has spurred us to investigate NSC105171-analogues as a novel class of small molecule in-vivo competent Chk2i.

In summary, our data indicate that Chk2 is the predominant trigger of toxicity following

TOP2-inhibitors and incorporating Chk2i’s into chemotherapy protocols that employ

TOP2-inhibitors may fully take advantage of the benefits such strategies have to offer for clinical translation.

102 CHAPTER 4

Differential expression of Dihdropyrimidine dehydrogenase (DPYD) in mutant p53 colorectal cancer cells and its modulation to study mutant specific sensitivity to 5-FU.

Introduction.

Earlier in chapter 2, we identified a novel role for p53 in regulating DPYD expression.

Specifically, WT p53 represses the expression of DPYD in response to TS inhibition. We also noted that the p53-null tumor cell lines were resistant to cytotoxicity caused by 5-FU.

With this knowledge, we hypothesized that the specific mutations of p53 in colorectal cancer cells would equip p53 to upregulate DPYD expression potentially altering the response to 5-FU therapy.

To address this possibility, we utilized isogenic HCT-116 cells carrying different p53 gene status to study the expression and loss thereof of DPYD in these cells in 2-D and 3-D cultures. We also analyzed DPYD expression in correlation to mutant p53 status in TCGA cohort of colorectal cancer patients.

Materials and Method

Cell viability assays

For cell viability assays, cells were seeded in 96-well black-walled plates at a concentration of 1-2×104 (cancer cell lines) per well in fresh media and in a volume of 100

103 µL per well. At the endpoint, CellTiter-Glo™ (Promega) assays were performed according to the manufacturer's protocol. Luminescence values were background subtracted and normalized to No drug/No treatment control.

Western blotting

The following antibodies were used: Rabbit polyclonal DPYD (1:500, Thermoscientific and Rabbit pAb Invitrogen PA5-22302), Mouse mAb p53 HRP (D0-1, 1:1000, Santa Cruz

Biotechnology), Mouse mAb Actin (A5541, 1:10000, Sigma). Secondary Goat anti-Mouse

IgG (Thermo Scientific 31430, 1:10000) and Goat anti-rabbit IgG (Thermo Scientific

31460 1:10000).

Tumor-spheres assays

HCT-116 WT, p53-null, R175H and R273H were seeded in 6 well ultra low attachment plates at 1X103 cells per well in MamoCult medium for 3-5 days to allow the formation of spheres. For DPYD knockdown experiments, cells were transfected with siRNA in 2-D cultures. The knockdown was carried out for 24 hrs following which cells were seeded as as above.

DPYD expression analysis from TCGA Nature 2012 cohort.

RSEM (RNA-Seq by Expectation Maximization) normalized expression values from patients were segregated based on Tp53 mutation status using DAVID algorithm

(387,388).

104 Results

Tp53 mutant R175H carrying HCT-116 cells express high DPYD.

Figure 4-1: Expression of DPYD protein in HCT-116 cells differing in p53 status

Western blot analysis shows the expression of DPYD protein following treatment with 5-FU (384µM) for 24

hrs.

We analyzed the variation of the DPYD expression in different mutant p53 expressing cells on an isogenic background. In our analysis, we focused on R175H and R273H mutant forms of p53 which are the most commonly mutated residues in cancer.

Analysis for DPYD protein expression (Figure 4-1) revealed that the p53R175H mutant cells expresses a very high amount of DPYD protein compared to the cells carrying p53WT (WT), p53-null and p53 R273H forms of p53. Unlike WT, p53 mutant p53R175H, p53R273H and NULL carrying cells did not represses DPYD expression following 5-FU treatment. These results were expected and consistent with our previous observation in chapter 2. It is noteworthy that DPYD levels are very high in R175H mutant p53 carrying cells.

105

Mutant p53R175H cells are desensitized to 5-FU following knockdown of DPYD.

Based on the expression pattern of DPYD in p53 mutant cells we predicted that the upregulation of DPYD observed in mutants such as p53R175H would confer increased resistance to 5-FU. To test our hypothesis, we transiently knockdown DPYD in these cells

Figure 4-2:Cellular viability of cells following 5-FU administration

Viability of HCT-116, WT, NULL, p53F175H and p53R273H cells treated with 5-FU (0-384uM) for 72hrs

N=3 for each concentration

and assessed for the viability of cells to various doses of 5-FU. Surprisingly, contrary to our expectation, loss of DPYD in p53R175H mutant increased the resistance of these cells to 5-FU. The p53-null cells however showed enhanced sensitivity to 5-FU consistent with our previous observation

106 Mutant p53R175H tumor-spheres are resistant to 5-FU following transient knockdown of DPYD.

Since DPYD was implicated in EMT, we further tested the effect of knockdown of DPYD on tumor sphere formation which are thought to enrich for mesenchymal features. Again, much to our surprise the effect of knockdown of DPYD in p53R175H cells was to increase the formation of tumor-spheres compared to WT and p53R273H (Figure 4-3). However, p53-null cells did not form tumor-spheres in our assay consistent with recent observations which show that anchorage independent growth is a characteristic of mutant p53 unlike p53 deficient cells (389) .

107

Figure 4-3: Tumor-sphere formation of HCT-116 cells

Cell were allowed to form spheres in low anchorage conditions for 1 week with/without knockdown of DPYD. (A) WT, (B) NULL, (C) R273H and (D) R175H.

108 Analysis of tumor DPYD expression in the TCGA colorectal cancer cohort reveals significant differential expression of tumor DPYD based on p53 status

In order to understand and see the potential relevance of our finding to human tumors, we analyzed tumor DPYD mRNA expression in the TCGA colorectal cancer cohort based on different p53 status. Analysis revealed that the TCGA DPYD mRNA data did not correlate with the DPYD protein levels we observed in colorectal cancer cells for the corresponding mutation. However, a significant variation of DPYD expression was

Figure 4-4: DPYD mRNA expression in colorectal cancer patients from TCGA based on different p53 mutation status.

Gene expression of DPYD on agilent microarray from TCGA Nature 2012 Cohort with respect to mtp53 status. (R175H, n=13; R273H, n=8; R248W, n=9; R213*, n=4; Y226*, n=4; WT, n=102;) statistics: Kruskal

Wallis Test p=0.002; Man Whitney test *** p<0.001, ** P<0.01, * P<0.0

observed between different groups of patients carrying different p53 mutations (Figure 4-

4). Further analysis of these subgroups of patients for DPYD expression and their

109 correlation with disease free survival, overall survival and/or predictive value to 5-FU therapy, might help in better understanding the relationship between these two genes.

Discussion

Extending our previous observations that WT p53 suppresses the expression of DPYD, we analyzed the expression of DPYD in colorectal cells carrying mutant p53 status and the biological consequences of altering its expression. Our assessment indeed revealed that some the p53 mutants such as revealed a binding site for p53 on DPYD downstream of the gene (p53DBS). the p53R175H indeed express significantly higher levels of DPYD protein compared to others. Since nearly 50% of colorectal tumors carry p53 mutation and based on previous observations that overexpression of DPYD is seen frequently in colorectal primary and metastatic colorectal tumors (197,351,390,391), we predicted that knockdown of DPYD in mutant p53 expressing cells would sensitize them to 5-FU.

However, to our surprise p53R175H expressing cells showed increased resistance nearly

~2-fold change in IC50. To further explore the effect of DPYD in long-term assay conditions we grew the HCT-116 cells carrying different p53 mutations as tumor-spheres.

It is thought that tumor-spheres enrich for stem and mesenchymal features and in light of recent observations that DPYD expression and activity was enriched as a mesenchymal feature we hypothesized that potentially altering DPYD in this setting could bring out the differences between different p53 mutants following knockdown of DPYD. However, much to our surprise p53R175H mutant expressing cells formed robust spheres and in fact showed slightly higher capacity to form spheres in the presence of 5-FU and knockdown of DPYD. Presently we do not fully understand the reason for these unexpected results. However, possible explanations which could reconcile our

110 observations are that mutant p53 acquires gain-of-function capabilities and some mutants such as p53R175H can rewire cellular metabolism to sustain growth and proliferation

(392). Recent evidence suggests that mutant forms of p53 could rewire nucleotide metabolism (393). Indeed, DPYD activity, specifically the metabolic product Dihdyrouracil was required for EMT of high grade breast cancer cell lines harboring mutant p53. Thus, it possible that an immediate product of DPYD enzymatic activity could be used for channeling into other metabolic activities required to sustain mutant p53 activities. As such any interference with this process could affect cellular growth, such as treatment with 5-FU. One can imagine, as an antimetabolite, 5-FU could also inhibit, enzymes or signaling molecules which use pyrimidine base analogues such as P2X and P2Y receptors (394). Hence, our observations of increased viability and tumor-spheres formation of p53 R175H cells when DPYD is knocked down could be due to reduced 5-

FU metabolites interfering with rewired metabolism. Further, observations in combined analysis of gastric cancer, leukemia, liver cacer, lung adenocarcinoma, ovarian cancer, and pancreatic cancer, for oncometabolites (altered metabolites found in cancer cells) has identified the DPYD gene along with other well-known genes such as IDH1 and IDH2, among the top 20 genes contributing to formation of oncometabolites. Particularly in liver cancer Dihydrouracil and Dihdrothymine were identified as oncometabolites (395).

Therefore, together these observations provide a strong explanation for why potentially we saw an unexpected increase in viability and tumor-sphere formation. If true, it suggests that DPYD could be an Achilles heel for certain p53 mutant tumors. However, we recognize that our data is preliminary and further characterization of DPYD expression

111 in different p53 mutant cells and the resultant effects of modulating DPYD expression on these cells need to be explored.

112 CHAPTER 5

DISCUSSION

Identification of a novel target of p53 in pyrimidine metabolism

Earlier work had implicated the role of p53-dependent apoptosis in sensitization to 5-FU in colorectal cancer cells. Previous studies from our lab indicated varying plasma levels of 5-FU in patients to whom the drug was administered. Coupled with the observation that p53 status is important for clinical efficacy, we hypothesized that p53 may control 5-FU metabolism with implications not only towards efficacy, but also with regard to the toxicity caused by the drug. To this end, efforts were focused on identifying the direct targets of p53 in 5-FU metabolism. Therefore, we set out to identify p53 binding (p53RE) elements within and ±20kb upstream or downstream of the genes in the 5-FU metabolism pathway.

We used an in-silico approach to identify several putative p53-binding sequences in the pyrimidine catabolic gene DPYD. In order to assess the role of this binding site in controlling DPYD expression, we initially carried out ChIP experiments using mouse liver, an organ where DPYD is highly expressed and contributes to physiological 5-FU levels.

Our analysis revealed a binding site for p53 downstream of the DPYD gene (p53DBS).

P53 binding to the p53BDS increased following 5-FU administration. We next wanted to study the functional consequence of p53 binding to this element. Analysis for DPYD mRNA and protein expression revealed repression of the gene expression. This was surprising as the binding of p53 to its consensus binding motif is generally thought to trans-activate genes. The mechanistic basis of this repression has not been further explored in this thesis but remains an area for future investigation. The potential ways to

113 address this aspect are described below in the future directions. Further, looking at the bigger picture, the results were also counterintuitive for another reason. As the primary role of p53 is to overcome cellular stress, we expected that DPYD would be upregulated to metabolize 5-FU resulting in inactivation of the drug. However, this was not the case and a probable explanation is discussed later in the context of TS inhibition. Thus, the above finding implicates a novel role for p53 in controlling pyrimidine degradation.

P53 controls systemic 5-FU metabolism with subsequent impact on tumor growth and a p53 R72P polymorphism could have an effect on this process.

Expanding on our novel finding of control of DPYD expression by p53, we were interested to learn if the plasma variations of 5-FU levels could be explained by alterations in p53.

To this end, we created a mouse model carrying liver-specific deletion of the Tp53 gene, i.e. Albcre;mT/mG;p53Δ/Δ. Analysis of plasma 5-FU levels in these mice revealed that

Albcre;mT/mg;p53Δ/+ have reduced 5-FU clearance, evident by a decreased ratio of 5-

FUH2/5-FU compared to Albcre;mT/mG;p53Δ/Δ mice. Furthermore Albcre;mT/mg;p53Δ/+ mice displayed delayed tumor progression compared to Albcre;mT/mg;p53Δ/Δ in response to 5-FU therapy. This implies that possibly alterations in the p53 gene in human populations could affect 5-FU levels, and could potentially explain our previous observation of varying 5-FU levels in patients. With this knowledge, we next hypothesized that polymorphisms in the p53 gene could alter DPYD expression. To address this, we collaborated with Dr. Maureen Murphy at the Wistar Institute to obtain humanized p53 knock-in mouse embryonic fibroblasts (HUPKI-MEF’s). HUPKI-MEF’s carrying either

114 proline (P/P) or Arginine (R/R) at codon 72 of p53 gene were analyzed for DPYD expression following 5-FU treatment. The codon R72P polymorphism has been extensively studied for altering p53 biological function. Analysis for DPYD expression revealed that the P/P allele was a better transcriptional repressor of DPYD expression than the R/R allele. This observation has implications in the clinic as, it suggests that germline variants of the p53 gene observed in the population may potentially impact on the toxicity of 5-FU in specific patients. Thus, in the future, patients may need to be screened for the polymorphism before the administration of 5-FU as it might impact 5-FU clearance. Additionally, our study calls for potential biochemical analysis for DPYD activity based on tumor p53 status before and after 5-FU treatment. This aspect is further explained in future directions.

The role p53 in 5-FU metabolism extends to tumor cells and is context-dependent

Based on our findings in liver and MEF’s regarding DPYD repression by p53, we were interested to learn the extent of this interaction in cancer cells which often have mutations in p53 that could impact on the efficacy of the drug. In our analysis, we observed that p53

WT tumor cell lines repressed DPYD in response to 5-FU. This repression of DPYD was abrogated when p53 was knocked down using RNAi. Since 5-FU causes DNA damage, we wanted to explore the role of ATM, ATR and DNA-PK in mediating repression of DPYD expression by p53. Our result indicated that ATM and DNA-PK signaling are key to this process. This prompted us to analyze the general role of DNA damaging chemotherapeutics in p53-mediated repression of DPYD. We used etoposide and irinotecan to address this question. Very surprisingly, the repression of DPYD by p53 was

115 specific to 5-FU treatment suggesting a possible link intrinsic to the mechanism of action.

Since the major mechanism of action of 5-FU involves TS inhibition, we hypothesized that lack of dTTP supply could trigger this repression, in order to salvage nucleotides for repair as DPYD is a rate limiting enzyme controlling catabolism of free bases. Indeed, this was the case as addition of thymidine to the cells rescued DPYD expression. Thus, our results indicate that not just the nature of DNA damage but the context of DNA damage determines the outcome of different therapeutic agents. This could potentially explain why p53 status matters to the outcome of 5-FU therapy. Our results are supported by few clinical studies where p53 status was shown to be important for overall survival of patients who have undergone 5-FU therapy (201-203). However, it has been difficult to firmly establish whether tumor p53 status matters for 5-FU therapy. Recent efforts to identify 5-

FU metabolic pathway genes have not identified any consistent features (396). This is partly because, as evident from our work, in addition to p53 status, expression of DPYD and potentially other metabolism genes in the 5-FU pathway together determine the outcome of 5-FU.

Targeting DPYD could sensitize p53-deficient and certain p53 mutant tumors to 5-

FU based therapy.

Our studies highlight that loss of p53 results in failure to repress DPYD expression.

Hence, intact (de-repressed) DPYD protein is now able to catabolize 5-FU and decrease the sensitivity to the drug. Further, our data suggests that DPYD inhibition particularly sensitizes p53-deficient cells to 5-FU therapy. Therefore, therapeutic inhibition of DPYD makes an attractive strategy with 5-FU therapy. As discussed in chapter 2 in detail, DPYD

116 inhibitor such as Gimeracil and Eniluracil have shown clinical benefit in many solid tumors

including breast and gastric cancer. In fact, therapeutic targeting of DPYD has been

approved in many countries in Europe and Australia but remains investigational in the

GastrointestinalUnited States. Oncology

TableTable 5 -1: 1: Summary Summary of ofClinical clinical Trials trials to Dateto date Investigating investigating the theEfficacy efficacy and Safetyand safety of S-1 in ofGastrointestinal S-1 in Gastrointestinal Cancers Cancers

Trial Name Patient Cohort Trial Design Therapy Results Safety Reference GastrointestinalSOFT mCRC, n=512 OncologyNon-inferiority trial, mFOLFOX + Bev mPFS 11.5 m FOLFOX + Bev Similar rate of AEs 42 median follow-up: vs SOX + Bev vs 11.7 m SOX + Bev 11.4 m TableACTS-CC 1: SummaryResected colon of ClinicalPhase III, Trials median to DateS-1 vsInvestigating UFT/LV DFS 75.5 the % S-1 Efficacy vs 72/5 % and SafetyLower incidence of S-1 of haemotologicalin 57 Gastrointestinal cancer, n=1,504 Cancers follow-up: 41.3 m UFT/LV toxicities, neurotoxicities with S-1 GEST Locally advanced or Phase III, open-label GEM vs S-1 vs Median OS 8.8 m (GEM) vs All treatments were generally well 26 Trial Name Patientmetastatic Cohort PC, n=834,Trial Design TherapyGEM + S-1 Results 9.7 m (S-1), and 10.1 m (GEM)Safety tolerated, although hematologicalReference SOFT mCRC,Japan n=512 and Taiwan Non-inferiority trial, mFOLFOX + Bev mPFS 11.5+ S-1.m FOLFOX Non-inferiority + Bev of S-1Similar to rate ofand AEs GI toxicities were more42 severe median follow-up: vs SOX + Bev vs 11.7 m SOX + Bev GEM was demonstrated in the GEM plus S-1 group 11.4 m ACTS-CC Resected colon Phase III, median S-1 vs UFT/LV DFS 75.5 % S-1 vs 72/5 % Lower incidencethan ofin haemotologicalthe GEM group 57 JASPAC-01 cancer,Resected n=1,504 PC, follow-up:Phase 41.3 III open-label m GEM vs S-1UFT/LV 2 yr OS 70 % (S-1) vs 53 %toxicities, (GEM) neurotoxicitiesFewer haematological with S-1 and hepatic 22 GEST Locallyn=378, advanced Japan or Phasemulticentre III, open-label GEM vs S-1 vs Median OSp<0.0001 8.8 m (GEM) for non-inferiority vs All treatments and toxicities were generally associated well with26 S-1 metastatic PC, n=834, GEM + S-1 9.7 m (S-1),superiority. and 10.1 m2-yr (GEM) RFS 49 %tolerated, (S-1) vs although treatment. hematological Stomatitis and Japan and Taiwan + S-1. Non-inferiority of S-1 to and GI toxicities were more severe 29 % (GEM). Median RFS 23.2 m diarrhoea (<5%) with S-1 GEM was demonstrated in the GEM plus S-1 group (S-1) vs 11.2 m (GEM) than in the GEM group JASPAC-01ACTS-GC ResectedResected PC, stage II PhasePhase III open-label III, open-label GEM vsSurgery S-1 + S-12 yr OS 703-year % (S-1) OS vs was 53 % 80.1 (GEM) % (S-1)Fewer vs haematologicalGrade 3 or and 4 AEs hepatic with S-122 group: 32 n=378,and III Japan GC, multicentre vs surgery p<0.0001alone 70.1 for non-inferiority% (surgery only) and toxicities associatedanorexia with (6.0 S-1 %), nausea (3.7 %) n=1,059, Japan superiority. 2-yr RFS 49 % (S-1) vs treatment. Stomatitisand diarrhoea and (3.1 %). 29 % (GEM). Median RFS 23.2 m diarrhoea (<5%) with S-1 JCOG9912 mGC, n=704, Japan Phase III, open label, Irinotecan + OS 10.8 m (5-FU) vs 11.4 m (S-1), Three treatment-related deaths 38 (S-1) vs 11.2 m (GEM) ACTS-GC Resected stage II Phasemedian III, open-label follow-up Surgerycisplatin + S-1 vs3-year OS12.3 was months 80.1 % (S-1)(irinocetan vs +Grade 3 or 4occurred AEs with S-1in the group: irinotecan 32 plus and III GC, 1 year vs surgery5-FU alone vs S-1 70.1 % (surgerycisplatin). only) Non-inferiority anorexia of S-1 to (6.0 cisplatin %), nausea group (3.7 and%) one was n=1,059, Japan 5-FU demonstrated. Irinocetanand diarrhoea + recorded (3.1 %). in the S-1 group JCOG9912 mGC, n=704, Japan Phase III, open label, Irinotecan + OS 10.8 mcisplatin (5-FU) vs not 11.4 superior m (S-1), to 5-FUThree treatment-related deaths 38 median follow-up cisplatin vs 12.3 months (irinocetan + occurred in the irinotecan plus FLAGS Advanced gastric Phase III, open-label 5-FU + cisplatin OS 8.6 m (CS) vs 7.9 m (5-FU + S-1 causes fewer haematological 40 1 year 5-FU vs S-1 cisplatin). Non-inferiority of S-1 to cisplatin group and one was adenocarcinoma, vs CS 5-FU demonstrated.cisplatin). Non-inferiority Irinocetan + recorded in AEsthe S-1 vs group5-FU Western cisplatin demonstratednot superior to 5-FU FLAGSSPIRITS AdvancedChemotherapy-naive gastric Phase Phase III, open-label III 5-FU +S-1 cisplatin vs CS305 OS 8.6 mRR (CS) 31 vs % 7.9 (S-1) m (5-FUvs 54 + % (CS);S-1 PFS causes fewerMore haematologicalgrade 3 or 4 AEs 40including 46 adenocarcinoma, vs CS cisplatin). Non-inferiority AEs vs 5-FU patients with 4.0 m (S1) vs 6.0 m (CS); OS leukopenia, neutropenia, Western demonstrated advanced gastric 11.0 m (S-1) vs 13.0 m (CS) anaemia, nausea and anorexia, SPIRITS Chemotherapy-naive Phase III S-1 vs CS305 RR 31 % (S-1) vs 54 % (CS); PFS More grade 3 or 4 AEs including 46 patientscancer, with n=305, Japan 4.0 m (S1) vs 6.0 m (CS); OS leukopenia, inneutropenia, CS vs S-1 START advancedAdvanced gastric GC, n=628, Phase III S-1 vs DS 11.0 m (S-1)RR 27vs 13.0% (S-1) m (CS) vs 39 % (DS);anaemia, PFS nauseaNeutropenia and anorexia, was more frequent in 47 cancer,Japan n=305, and Korea Japan 4.2 m (S1) vs 5.3 m (DS); inOS CS vs S-1 the DS group and one patient died START Advanced GC, n=628, Phase III S-1 vs DS RR 27 % (S-1) vs 39 % (DS); PFS Neutropenia was more frequent in 47 10.8 m (S-1) vs 12.5 m (DS) by Grade 4 thrombocytopenia in the Japan and Korea 4.2 m (S1) vs 5.3 m (DS); OS the DS group and one patient died DS group 10.8 m (S-1) vs 12.5 m (DS) by Grade 4 thrombocytopenia in the 5-FU = 5 fuorouracil; AE = adverse effect; CS = S-1 + cisplatin; DS = docetaxel + S-1; GC = gastric cancer; GEM = DSgemcitabine; group GI = gastrointestinal; LV = leucovorin; m = months; 5-FUmGC = = 5 metastatic fuorouracil; gastric AE = adverse cancer; effect; PC = CS pancreatic = S-1 + cisplatin; cancer; DS PFS = docetaxel = progression-free + S-1; GC = survival;gastric cancer; RFS = GEM relapse-free = gemcitabine; survival; GI = RR gastrointestinal; = response rate; LV = UFTleucovorin; = uracil/tegafur. m = months; mGC = metastatic gastric cancer; PC = pancreatic cancer; PFS = progression-free survival; RFS = relapse-free survival; RR = response rate; UFT = uracil/tegafur.

chemotherapy.29 This is in marked contrast to Asian data, which show Figure 2: 2: The The FLAGs FLAGs Study Study – S-1 – S-1 is is chemotherapy.29 This is in marked contrast to Asian data, which show Non-inferior to to 5-FU 5-FU Plus Plus Cisplatin Cisplatin substantialsubstantial benefts forbene adjuvantfts for chemotherapyadjuvant chemotherapy compared with compared surgery with surgery Source: Proceedings of a Satellite Symposiumalone. 30,31Heldalone. The 30,31 Adjuvantat Thethe Adjuvant Chemotherapy European Chemotherapy Trial Society of S-1 Trial For Gastric offor S-1 Medical Cancer For Gastric Cancer 100100 (ACTS-GC)(ACTS-GC) phase III phase trial in III Japan trial randomised in Japan randomised patients (n=1,059) patients with (n=1,059) with HazardHazard ratio ratio = 0.92 = (950.92 % (95con %fdence conf denceinterval interval 0.80–1.05) 0.80–1.05) Oncology 15th MedianWorld overall Congress survival: of Gastrointestinalresected stage Cancer II and III GC into S-1 Barcelona, monotherapy or surgery3 July only, 2013 and found Median overall survival: resected stage II and III GC to S-1 monotherapy or surgery only, and found 80 S-1 + cisplatin 8.6 months versus 5-fuorouracil + 30,32 80 that 3-year OS was 80.1 % (S-1) versus 70.1 % (surgery only). cisplatinS-1 + 7.9 cisplatin months 8.6 months versus 5-fuorouracil + that 3-year OS was 80.1 % (S-1) versus 70.1 % (surgery only).30,32 Non-inferiority test on 1.10 margin; p=0.0068 60 cisplatin 7.9 months Non-inferiority test on 1.10 margin; p=0.0068 60 Benefit of Triplet Regimen in 40 AdvancedBenefit Gastric of Triplet Cancer Regimen for Patients in

Overall survival (%) 40 Able to Withstand Increased Toxicity 20 Advanced Gastric Cancer for Patients117 Overall survival (%) In inoperableAble or tometastatic Withstand GC, chemotherapy Increased prolongs OSToxicity in the frst- 20 line setting. Furthermore, chemotherapy can preserve and even improve 02468101214 16 18 20 22 24 26 28 30 32 In inoperable or metastatic GC, chemotherapy prolongs OS in the frst- QoL in some patients. 5-FU is the mainstay of GC therapy. The following Months from randomisation line setting. Furthermore, chemotherapy can preserve and even improve 5-fuorouracil + cisplatin combined regimens are commonly used in Europe: CF (cisplatin + 0246S-1 + cisplatin8101214 16 18 20 22 24 26 28 30 32 5-FU), DCFQoL (docetaxel in some patients.+ cisplatin 5-FU + 5-FU), is the PLF mainstay (leucovorin of [LV],GC therapy.cisplatin The following Months from randomisation survival (PFS) and OS in GC compared with surgery alone. This has become + 5-FU, FLO (5-FU + LV + oxaliplatin) and FLOT (5-FU + LV + oxaliplatin S-1 + cisplatin 5-fuorouracil + cisplatin combined regimens are commonly used in Europe: CF (cisplatin + the most important clinical strategy in Europe; adjuvant chemotherapy docetaxel),5-FU), ECF DCF (epirubicin, (docetaxel cisplatin + cisplatin and 5-FU), + ECX5-FU), (epirubicin, PLF (leucovorin cisplatin [LV], cisplatin is employed to a lesser extent. Individual clinical trials of adjuvant + capecitabine) and EOX (epirubicin, oxaliplatin and capecitabine). Recent survival (PFS) and OS in GC compared with surgery alone. This has become + 5-FU, FLO (5-FU + LV + oxaliplatin) and FLOT (5-FU + LV + oxaliplatin chemotherapy have yielded negative results although a meta-analysis advances in research have helped to optimise therapeutic decision- showedthe most modest important improvements clinical strategy in effcacy: in adjuvantEurope; chemotherapyadjuvant chemotherapy was making. docetaxel),It has been foundECF (epirubicin, that oxaliplatin cisplatin is as effective and 5-FU), as cisplatin, ECX (epirubicin, with cisplatin associatedis employed with toa statistically a lesser extent.signifcant Individual beneft in clinicalterms of trials OS (hazard of adjuvant signi fcantly+ capecitabine) reduced toxicity. and33,34 EOX Capecitabine (epirubicin, has oxaliplatin also been andshown capecitabine). to be Recent ratiochemotherapy [HR] 0.82); and have fve-year yielded OS negative increased results from 49.6 although % to 55.3 a meta-analysis % with as effective advances as 5-FU. in34,35 researchTriplet therapy have may helped also be to preferable optimise to therapeuticdoublet: decision- showed modest improvements in effcacy: adjuvant chemotherapy was making. It has been found that oxaliplatin is as effective as cisplatin, with associated with a statistically signifcant beneft in terms of OS (hazard signifcantly reduced toxicity.33,34 Capecitabine has also been shown to be 96ratio [HR] 0.82); and fve-year OS increased from 49.6 % to 55.3 % with as effective as 5-FU.34,35 TripletEUROPEAN therapy ONCOLOGY may also & HAEMATOLOGYbe preferable to doublet:

96 EUROPEAN ONCOLOGY & HAEMATOLOGY A number of studies have shown a benefit in reducing the toxicity when DPYD is targeted as part of S-1 (Table 5-1). Particularly neurotoxicity and cardiotoxicity has been reduced.

This is because the major toxicity associated with 5-FU regimes arise from toxic catabolite of 5-FU such as Fluoro-beta alanine (FBAL) (397). Thus, inhibition of DPYD activity will provide may spare toxicities such as hand foot syndrome (Figure 5-2)

A recent meta-analysis in a 2182 Asian patient population revealed that S-1 was far superior to 5-FU for treatment of advanced gastrointestinal cancer in enhancing overall survival and achieving objective response rates (398). However, based on observation made in chapter 4, it should be noted that for certain tumors such as those carrying the p53R175H mutation, caution is warranted in using S-1 based therapies. Our surprising and unexplained result highlights that p53R175H tumors may be more susceptible to 5-

FU when DPYD is not inhibited. Recent analysis for failure of 5-FU + eniluracil (a potent inhibitor of DPYD) in breast cancer was attributed to excess eniluracil ratio to 5-FU administered (399). Though the mechanistic details remained elusive, our observations that certain mutants such as p53R175H may acquire resistance subsequent to DPYD inhibition could be certainly be one of the reasons. Thus, further studies need to conducted in this regard.

118 Gastrointestinal Oncology

Figure 1: Impact of the Multi-step Modulation The pivotal role of DPD in the metabolism of 5-FU presents a major of FU in Teysuno challenge to clinical oncologists. Available options to circumvent the DPD hurdle include using 5-FU prodrugs such as capecitabine (Xeloda®), or

H to inhibit DPD. Among the oralfuoropyrimidines recently developed, two OH O Oteracil N KO2C N O Gimeracil Tegafur N O potassium 1 were DPD inhibitors: EU-5-FU (eniluracil//FU) and UFT (tegafur [ftorafur] 0.4 ++1 N NH = S-1 (CDHP) CI (FT) F NH (OXO) OH O + uracil at a 1:4 molar ratio; the excess uracil competes with 5-FU for O Liver DPD), but these have not demonstrated signifcant effcacy advantages CYP2A6 in clinical trials (please see Box for defnitions of trials discussed in Tumour OPRT DPD OPRT this article).7–9 Capecitabine is sequentially converted to 5-FU by three FBAL 5-FU cells +++ Normal OPRT enzymes located in the liver and in tumours.10 After oral administration, cells + Activated capecitabine is rapidly and extensively absorbed from the GI tract and Myelodepression Activated 5-FU +++ 5-FU + has a relatively short elimination half-life.11 Hand–foot sydrome GI tract toxicities Cardiotoxicity (stomatitis, mucositis, Neurotoxicity diarrhoea) Alopecia Anti-tumour activity S-1 contains tegafur, gimeracil (5 chloro 2,4 dihydroxypyridine – CDHP), (TS-FdUMP + F-RNA) a strong DPD inhibitor that prolongs the half-life of 5-FU, and oteracil (OXO, potassium oxonate), OXO inhibits the phosphorylation of 5-FU Better general Better safety on Same effcacy safety fast-dividing cells as IV 5-FU to fuorouridine monophosphate, an active intermediary metabolite of 5-FU, by orotate phosphoribosyltransferase (OPRT) in the GI

5-FU = 5-fuorouracil; CYP2A6 = cytochrome P450 2A6; DPD = dihydropyrimidine tract, thereby reducing GI-related toxicity of 5-FU. Tegafur [R,S-1- dehydrogenase; FBAL = fuoro-beta-alanine; F-RNA = 5-FU incorporated in RNA; 1(tetrahydrofuran-2-yl)-5-FU] is a prodrug that is mainly converted GI = gastrointestinal; IV = intravenous; OPRT = orotate phosphoribosyltransferase; TS- FdUMP = thymidine synthase fuorodeoxyuridine monophosphate. by cytochrome P450 2A6 (CYP2A6) to 5-FU.12 CYP2A6 shows large interindividual and interethnic variations in its expression levels and Figurethe tumour’s 5- 1: inherited Benefits genome outlining variation S-1 plustherapy acquired genome variation. conversion activities, which are mainly attributed to CYP2A6 genetic Additional acquired genome variations may occur in metastatic or recurrent polymorphisms, and are more common in Asian than Caucasian Source:tumour Proceedings cells that inf ofuence a Satellite drug response Symposium and treatmentHeld at the outcomes. European A Societypopulations. for Medical The higher effcacy of CYP2A6 results in the more rapid Oncologycomprehensive 15th World pharmacogenomic Congress of strategyGastrointestinal should encompass Cancer in multiple Barcelona, conversion 3 July 2013 of tegafur. to 5-FU in Caucasian subjects, who achieve a 2 mechanisms of genome variation. higher area under the curve of 5-FU than Asians, and therefore have a different optimal dosage level.13 The CYP2A6 genotype correlated TargetingFluoropyrimidines DPYD through have inhibitors formed such the as mainstayGimeracil (partof GI ofcancer S1) can therapy reduce forthe toxicitywith bythe limiting treatment Hand effcacy of S-1-based chemotherapy in previously footmany syndrome years., cardiotoxicity, Prior to 1980, Neurotoxicity. 5-FU was primarilyFurther Oteracil used potassiumas monotherapy. (OXO) block untreated the enzyme metastatic OPRT GC (mGC) patients; patients with fewer variant thusIn limiting the 1990s toxicity it began to the GI.to beBut used cancer in cellscombination anabolism with through newer OPRT cytotoxic is not inhibited alleles as had they signi expressfcantly better response rates.14 verytherapies high levels. such as irinotecan and oxaliplatin, and after 2000, targeted therapies were added to combined treatment regimens, resulting The S-1 formulation of 5-FU may also confer a pharmacokinetic in positive trends in median survival rates, which now exceed 20 advantage. A pharmacokinetic study compared continuous infusion months.3 The continued evolution of 5-FU has led to the development of 5-FU with oral administration. Individual 5-FU concentrations in the Exploitingof oral forms ATM of the-Chk2 drug.-p53 Resistance signaling and toxicity to prevent have been dose signi limitinfcantg toxicityblood during of TOP continuous2 infusion were highly variable, whereas those limitations to the clinical use of 5-FU. Increased understanding of the after oral administration were reproducible.15 In a comparison of 5-FU poisonmechanism by targeting of action Chk2 of 5-FU has led to the development of strategies pharmacokinetics in patients receiving continuous 5-FU infusion and oral to overcome these limitations. 5-FU is regulated via a complex UFT, the maximum 5-FU concentrations generated from oral UFT were The importance of inhibition of Chk2 has been widely established for radioprotection. It is network of anabolic and catabolic genes. The cytotoxicity of 5-FU higher than the steady-state levels during continuous infusion. A further unclearis a result so far of whetheranabolism Chk2 to nucleotides. targeting can The also 5-FU offer active protection metabolite to normalstudy investigatedcells during the impact of DPD inhibition on the pharmacokinetics fuorodeoxyuridine monophosphate (FdUMP) binds to the nucleotide- of 5-FU by comparing the pharmacokinetic profle of S-1 to that of binding site of thymidine synthase (TS) and forms a stable ternary tegafur alone. Exposure to 5-FU was signifcantly greater following S-1

16 complex with TS and CH2THF, blocking the conversion of deoxyuridine administration compared119 with tegafur administration. monophosphate (dUMP) to deoxythymidine monophosphate (dTMP). This results in deoxynucleotide (dNTP) imbalances and increased There is a need to optimise the dosage and scheduling of oral levels of deoxyuridine triphosphate (dUTP), both of which cause fuoropyrimidines on the basis of their pharmacological characteristics. DNA damage.4 In addition to CYP2A6, tumoural thymidine phosphorylase plays a critical role in 5-FU activity by being involved in neoangiogenesis and by 5-FU Pharmacokinetics and Metabolism activating 5-FU prodrugs at the target site.17 In studies of capecitabine, Pharmacokinetics studies have demonstrated that 80 % of clinically predictors of resistance include high DPD expression18,19 and a high TP/ administered 5-FU is inactivated and eliminated through the catabolic DPD ratio.20 In studies of S-1, predictive factors for resistance include pathway. The underlying mechanism is catabolism by dihydropyrimidine TS, DPD and OPRT.21 In conclusion, capecitabine and S-1 represent dehydrogenase (DPD), which is found in liver tissue and enterocytes. a successful shift from IV to oral administration in GI oncology and Variability in DPD activity in the normal population accounts for the pharmacological studies have suggested a potential role for personalised observed differences in the pharmacokinetics and oral bioavailability of treatment, though further research is required to optimise treatment 5-FU;5 a defciency of DPD has been recognised as an important risk factor, and the impact of pharmacogenetics on drug activity should be further predisposing patients to the development of severe 5-FU-associated investigated. The multistep modulation of S-1 may improve its effcacy toxicity.6 Therefore, reducing the catabolism of 5-FU by DPD should and safety ratio and represents a signifcant advance in the treatment of increase the availability of the drug. GI cancer (see Figure 1). n

94 EUROPEAN ONCOLOGY & HAEMATOLOGY chemotherapy. However, since chemotherapy involves various classes of drugs the mechanism of action of these drugs downstream of DNA damage are diverse.

Myelosuppression is a major outcome of chemotherapeutic regimens. It is known that the

ATM-Chk2-p53 pathway plays a major role in apoptotic cell death during myelosuppression. However, in order to fully exploit the targeting of Chk2 along with chemotherapy, establishing the context, i.e. which chemotherapeutic agents elicit robust

Chk2 signaling becomes very important. In our study, here we have identified that Chk2 is a very important determinant of toxicity following treatment with TOP2 poisons.

Conversely, genetic ablation and pharmacological inhibin of Chk2 prevents dose limiting toxicity following TOP2 poison but not Top1. So far Chk2 inhibition have only been evaluated in terms of improving treatment efficacy where genetic loss or transcriptional suppression of Chk2 has not been shown to provide any benefit to chemotherapy.

However recent evidence suggests that certain cancer overexpress Chk2 and targeting them could be beneficial. Understanding the molecular context where Chk2 targeting could be beneficial would provide an improved rationale for targeting Chk2. For example, recent evidence suggests that stem cells in nasopharyngeal tumors overexpressing Myc rely on Chk2 for radioresistance (400). In such cases targeting Chk2 would not only eliminate radio-resistance but also offers protection for normal cells limiting toxicity.

TOP2A amplification has been observed in HCC and Chk2 in HCC has been shown to promote chromosomal instability and progression of HCC (401,402). Thus, our study provides a rationale to use TOP2 poisons like etoposide in combination with Chk2 targeting for improved efficacy and toxicity.

120 Identification of a novel pharmacological Chk2 inhibitor with in-vivo activity

Novel Chk2 inhibitors currently are being developed in the clinic. None of the pure Chk2

inhibitors currently being developed have reached clinical stage. The available inhibitors

have poor in-vivo efficacy as seen in our studies. Therefore, we set out to identify new

Chk2 inhibitors with enhanced in-vivo efficacy. Our approach involved combination of in-

silico molecular modeling, cell free Chk2 kinase activity assay and functional analysis for

protection of cells following TOP2 poison. By combined approach we were able to identify

a new compound NSC105171 with significant in-vitro and in-vivo activity superior than

current inhibitors.

A. 5-FU B. Etoposide

Figure 5-2: Different outcomes of p53 signaling following DNA damage therapy

(A) Model depicting the repression of DPYD by p53 following thymidylate synthase (TS) inhibition, dTTP

imbalance and DNA damage. P53 transcriptionally represses DPYD expression and negatively impacts the

catabolism of pyrimidines following inhibition of TS. In extension, this results in an imbalance in the

dTTP/dUTP ratio and a need to salvage cellular levels of dTTP to maintain uninterrupted DNA replication

121 and repair. However, in the presence of 5-FU this might cause increased cell death due to reduced catabolism and increased incorporation of the 5-FU metabolites FdUTP and FUTP into DNA and RNA respectively. (B) Effect of Chk2 inhibition in normal and cancer cell. Chk2 inhibition following DNA damage in normal cells prevent cell cycle arrest and apoptosis whereas in cancer cell which have aberrant Chk2 signaling or Chk2 over expression leads to chromosomal instability and mitotic catastrophe.

In summary, the p53 response to DNA damage is central in determining cellular fates.

P53 response to damage varies depending on the nature of the damage, the duration of the damage, cell and tissue type and finally the status of p53. In Chapter 2 we have explored p53 function in metabolism and identified a key target DPYD whose modification is central to 5-FU induced toxicity and efficacy. In fact, loss of p53 in kind of universal feature of cancer and hence inhibition of DPYD in p53-deficient tumors can enhance chemotherapeutic efficacy. In Chapter 3 we have closely followed the strategy of cyclotherapy. Cyclotherapy is therapeutic strategy to protect normal cells from side effects of chemotherapy. It relies on minimizing the chemotherapeutic effect on normal cells by utilizing the functional p53 pathway to minimize toxicity by arresting or preventing apoptotic cell death while cancer cells deficient in p53 undergo G2/M damage subsequently leading to mitotic catastrophe. To this extent we have utilized the ATM-

Chk2-p53 pathway to limit toxicity identifying Chk2 inhibitors which can reduce myelosuppression in response to TOP2 poisons whereas cancer cells are not protected but in fact is expected to be sensitized due to deregulated p53 singling and aberrant DNA repair mediated by Chk2.

122 Future directions

Mechanistic basis of p53-dependent DPYD repression

One of the outstanding questions arising for the work in chapters 2 and 4 is the mechanistic basis of DPYD repression by p53. There are several ways to address this.

First the p53DBS could be cloned ahead of a reporter gene to assess the functional binding of p53 and its impact on reporter expression in a heterologous system. Second, a chromosome confirmation capture assay could be performed by pulling down p53 and assessing for interaction between the promoter regions of the DPYD gene with a wild- type or mutated p53DBS. An extension of this assay would be to analyze the immunoprecipitated p53 for interaction with known negative regulators of transcription such as various HDAC’s and mSin3A.

Characterization of DPYD expression in different p53-mutant expressing cells

Preliminary work in isogenic cell lines with different p53 status i.e. p53WT, p53-null, p53R175H and p53R273H has identified differential expression of DPYD. In particular, the p53R175H mutant expressed high DPYD and formed robust spheres. Preparations of stable knockdown of DPYD in these cells and assessment for proliferation, invasion and sphere formation could further yield mutant p53-specific vulnerabilities. Future work may also involve xenograft studies to assess the in-vivo impact on therapeutic efficacy.

Use of novel mouse models (Albcre;p53fl/fl;DPYDfl/fl mice) and p53 R72P to study the direct effects of manipulating DPYD in liver.

123 We have developed a novel liver-specific DPYD knockout mouse (Albcre;p53fl/fl;DPYDfl/fl mice) to directly confirm the effect of DPYD loss on systemic pharmacokinetics of 5-FU.

This model could also be used to study the effect of 5-FU on growth of syngeneic tumors.

P53-Ras-Myc tumors can serve as an excellent model.

Currently we have also acquired p53 polymorphic mice carrying the codon 72 polymorphism. Since we observed that the p53P/P allele represses DPYD expression better than p53 R/R allele, our future studies could be aimed at a more detailed mechanistic understanding of the observation and its impact of pharmacokinetics of 5-

FU.

Assessment of patient PBMC’s for DPYD expression and activity to predict toxicity to 5-FU.

Blood samples can be collected from patients undergoing or who will undergo 5-FU therapy and DPYD expression analysis with respect to p53 status (i.e., germ-line polymorphism if any) could be assessed to correlate with severity of toxicity observed in the clinic. Alternatively, a DPYD expression score could be assigned based on various levels of toxicity.

Identifying NSC105171 analogues

Molecular modeling efforts based on the results in chapter 3 could enhance NSC105171 efficacy and potency. New analogues of NSC105171 could be tested in-vitro and in-vivo for Chk2 inhibition.

124

Combination of NSC105171 with etoposide in-vivo in tumor-bearing mice to study efficacy and toxicity

Syngeneic mouse tumor studies can be conducted for evaluating the potential clinical efficacy of NSC105171 in limiting toxicity when combined with etoposide. Briefly blood from these mice can be analyzed for hematological toxicity through BM staining and CBC analysis. Tumor measurement in the same mice will reveal potential clinical efficacy and confirm that the combination therapy will protect from toxicity while not reducing efficacy.

Further, syngeneic mouse tumor cells carrying high levels of Chk2 could be tested for feasibility of dose escalation in the presence of Chk2 inhibition in order to improve efficacy while simultaneously limiting toxicity.

Impact of post-translational modification on Chk2 elicited by TOP2 poisons versus other chemotherapeutic agents

Mechanistic studies of post-translational modification of ChK2 in response to TOP2 poisons and other chemotherapeutic drugs could reveal unique insights into downstream molecular events mediating toxicity.

Role of p21 loss in the survival of mice with respect to etoposide-induced toxicity.

Our p21-/- mouse model showed enhanced survival in the face of etoposide treatment similar to Chk2-/- mice. p21 protects the cells from toxicity as cell cycle arrest allows for the cells to repair damage, maintain genome stability and improve cell survival. Thus, the survival of our p21-/-mice is counterintuitive to what one might expect. p21-/- mice survive

125 about an average of 16 months before the onset of spontaneous tumors. The loss of p21 contributes to enhanced apoptosis (p53-dependent and -independent) in severely damaged cells through mitotic catastrophe and thus might delay tumor initiation and progression. This might explain the enhanced survival of p21-/- compared to p53-/- mice.

However, in normal cells, the role of p21 is ambiguous and tissue dependent. In case of

GI, a highly proliferative tissue, p21 loss allows for the immediate expansion of the crypt cells but ultimately leads to persistent DNA damage and enhanced apoptosis, albeit, this response is delayed (403). In case of bone marrow HSC, particularly in C57BL/6 mice, the loss of p21 does not show increased sensitivity to ionizing radiation (404). In fact, a recent study highlighted that a heterozygous loss of the p21 gene partially rescued the radio-lethality of p537KR mice, where p537KR allele induces higher p21 expression than apoptotic genes in bone marrow (405). This suggests that p21 contributes to hematopoietic sensitivity by blocking G1/S transition. Thus, tissue-specific p53 activities through different downstream effectors coordinate to bring about cell death. In our model, etoposide mainly causes myelosuppression which is the dose-limiting toxicity, hence it could be possible that the loss of p21 did allow the HSC compartment to expand and survive better in the short term, preventing toxicity. However, this hypothesis could be tested by assessing the cell cycle of BM-HSC from p21-/- mice treated with etoposide in the same way as in chapter 3. Further, overall HSC number can be analyzed by FACS.

Additionally, p21-/- mice can be given a higher dose of etoposide and observed over months (>4months) to assess for the possibility of tumors from this compartment which could indicate the survival of damaged HSC cells.

126 References

1. Linzer DI, Levine AJ. Characterization of a 54K dalton cellular SV40 tumor antigen present in SV40-transformed cells and uninfected embryonal carcinoma cells. Cell 1979;17:43-52. 2. DeLeo AB, Jay G, Appella E, Dubois GC, Law LW, Old LJ. Detection of a transformation- related antigen in chemically induced sarcomas and other transformed cells of the mouse. Proceedings of the National Academy of Sciences of the United States of America 1979;76:2420-4. 3. Kress M, May E, Cassingena R, May P. Simian virus 40-transformed cells express new species of proteins precipitable by anti-simian virus 40 tumor serum. J Virol 1979;31:472- 83. 4. Lane DP, Crawford LV. T antigen is bound to a host protein in SV40-transformed cells. Nature 1979;278:261-3. 5. Melero JA, Stitt DT, Mangel WF, Carroll RB. Identification of new polypeptide species (48- 55K) immunoprecipitable by antiserum to purified large T antigen and present in SV40- infected and -transformed cells. Virology 1979;93:466-80. 6. Smith AE, Smith R, Paucha E. Characterization of different tumor antigens present in cells transformed by simian virus 40. Cell 1979;18:335-46. 7. Chumakov PM, Iotsova VS, Georgiev GP. [Isolation of a plasmid clone containing the mRNA sequence for mouse nonviral T-antigen]. Dokl Akad Nauk SSSR 1982;267:1272-5. 8. Harlow E, Williamson NM, Ralston R, Helfman DM, Adams TE. Molecular cloning and in- vitro expression of a cDNA clone for human cellular tumor antigen p53. Molecular and cellular biology 1985;5:1601-10. 9. Jenkins JR, Rudge K, Currie GA. Cellular immortalization by a cDNA clone encoding the transformation-associated phosphoprotein p53. Nature 1984;312:651-4. 10. Leppard K, Totty N, Waterfield M, Harlow E, Jenkins J, Crawford L. Purification and partial amino acid sequence analysis of the cellular tumour antigen, p53, from mouse SV40- transformed cells. The EMBO journal 1983;2:1993-9. 11. Oren M, Levine AJ. Molecular cloning of a cDNA specific for the murine p53 cellular tumor antigen. Proceedings of the National Academy of Sciences of the United States of America 1983;80:56-9. 12. Pennica D, Goeddel DV, Hayflick JS, Reich NC, Anderson CW, Levine AJ. The amino acid sequence of murine p53 determined from a c-DNA clone. Virology 1984;134:477-82. 13. Ben David Y, Prideaux VR, Chow V, Benchimol S, Bernstein A. Inactivation of the p53 oncogene by internal deletion or retroviral integration in erythroleukemic cell lines induced by Friend leukemia virus. Oncogene 1988;3:179-85. 14. Mowat M, Cheng A, Kimura N, Bernstein A, Benchimol S. Rearrangements of the cellular p53 gene in erythroleukaemic cells transformed by Friend virus. Nature 1985;314:633-6. 15. Wolf D, Rotter V. Inactivation of p53 gene expression by an insertion of Moloney murine leukemia virus-like DNA sequences. Molecular and cellular biology 1984;4:1402-10. 16. Wolf D, Rotter V. Major deletions in the gene encoding the p53 tumor antigen cause lack of p53 expression in HL-60 cells. Proceedings of the National Academy of Sciences of the United States of America 1985;82:790-4.

127 17. Baker SJ, Fearon ER, Nigro JM, Hamilton SR, Preisinger AC, Jessup JM, et al. Chromosome 17 deletions and p53 gene mutations in colorectal carcinomas. Science 1989;244:217-21. 18. Eliyahu D, Michalovitz D, Eliyahu S, Pinhasi-Kimhi O, Oren M. Wild-type p53 can inhibit oncogene-mediated focus formation. Proceedings of the National Academy of Sciences of the United States of America 1989;86:8763-7. 19. Finlay CA, Hinds PW, Levine AJ. The p53 proto-oncogene can act as a suppressor of transformation. Cell 1989;57:1083-93. 20. Cancer Genome Atlas Research N. Integrated genomic analyses of ovarian carcinoma. Nature 2011;474:609-15. 21. Belyi VA, Ak P, Markert E, Wang H, Hu W, Puzio-Kuter A, et al. The origins and evolution of the p53 family of genes. Cold Spring Harb Perspect Biol 2010;2:a001198. 22. Polyak K, Waldman T, He TC, Kinzler KW, Vogelstein B. Genetic determinants of p53- induced apoptosis and growth arrest. Genes & development 1996;10:1945-52. 23. Waldman T, Kinzler KW, Vogelstein B. p21 is necessary for the p53-mediated G1 arrest in human cancer cells. Cancer research 1995;55:5187-90. 24. el-Deiry WS, Kern SE, Pietenpol JA, Kinzler KW, Vogelstein B. Definition of a consensus binding site for p53. Nat Genet 1992;1:45-9. 25. Harper JW, Adami GR, Wei N, Keyomarsi K, Elledge SJ. The p21 Cdk-interacting protein Cip1 is a potent inhibitor of G1 cyclin-dependent kinases. Cell 1993;75:805-16. 26. Laronga C, Yang HY, Neal C, Lee MH. Association of the cyclin-dependent kinases and 14- 3-3 sigma negatively regulates cell cycle progression. The Journal of biological chemistry 2000;275:23106-12. 27. Zhan Q, Antinore MJ, Wang XW, Carrier F, Smith ML, Harris CC, et al. Association with Cdc2 and inhibition of Cdc2/Cyclin B1 kinase activity by the p53-regulated protein Gadd45. Oncogene 1999;18:2892-900. 28. Wei JS, Song YK, Durinck S, Chen QR, Cheuk AT, Tsang P, et al. The MYCN oncogene is a direct target of miR-34a. Oncogene 2008;27:5204-13. 29. Sun F, Fu H, Liu Q, Tie Y, Zhu J, Xing R, et al. Downregulation of CCND1 and CDK6 by miR- 34a induces cell cycle arrest. FEBS Lett 2008;582:1564-8. 30. He L, He X, Lim LP, de Stanchina E, Xuan Z, Liang Y, et al. A microRNA component of the p53 tumour suppressor network. Nature 2007;447:1130-4. 31. Brodsky MH, Weinert BT, Tsang G, Rong YS, McGinnis NM, Golic KG, et al. Drosophila melanogaster MNK/Chk2 and p53 regulate multiple DNA repair and apoptotic pathways following DNA damage. Molecular and cellular biology 2004;24:1219-31. 32. Sutcliffe JE, Brehm A. Of flies and men; p53, a tumour suppressor. FEBS Lett 2004;567:86- 91. 33. Jiang P, Du W, Heese K, Wu M. The Bad guy cooperates with good cop p53: Bad is transcriptionally up-regulated by p53 and forms a Bad/p53 complex at the mitochondria to induce apoptosis. Molecular and cellular biology 2006;26:9071-82. 34. Miyashita T, Reed JC. Tumor suppressor p53 is a direct transcriptional activator of the human bax gene. Cell 1995;80:293-9. 35. Nakano K, Vousden KH. PUMA, a novel proapoptotic gene, is induced by p53. Mol Cell 2001;7:683-94.

128 36. Oda E, Ohki R, Murasawa H, Nemoto J, Shibue T, Yamashita T, et al. Noxa, a BH3-only member of the Bcl-2 family and candidate mediator of p53-induced apoptosis. Science 2000;288:1053-8. 37. Oda K, Arakawa H, Tanaka T, Matsuda K, Tanikawa C, Mori T, et al. p53AIP1, a potential mediator of p53-dependent apoptosis, and its regulation by Ser-46-phosphorylated p53. Cell 2000;102:849-62. 38. Tait SW, Green DR. Mitochondria and cell death: outer membrane permeabilization and beyond. Nat Rev Mol Cell Biol 2010;11:621-32. 39. Yu J, Zhang L, Hwang PM, Kinzler KW, Vogelstein B. PUMA induces the rapid apoptosis of colorectal cancer cells. Mol Cell 2001;7:673-82. 40. Chipuk JE, Kuwana T, Bouchier-Hayes L, Droin NM, Newmeyer DD, Schuler M, et al. Direct activation of Bax by p53 mediates mitochondrial membrane permeabilization and apoptosis. Science 2004;303:1010-4. 41. Munsch D, Watanabe-Fukunaga R, Bourdon JC, Nagata S, May E, Yonish-Rouach E, et al. Human and mouse Fas (APO-1/CD95) death receptor genes each contain a p53- responsive element that is activated by p53 mutants unable to induce apoptosis. The Journal of biological chemistry 2000;275:3867-72. 42. Wu GS, Burns TF, McDonald ER, 3rd, Jiang W, Meng R, Krantz ID, et al. KILLER/DR5 is a DNA damage-inducible p53-regulated death receptor gene. Nat Genet 1997;17:141-3. 43. Ashkenazi A. Directing cancer cells to self-destruct with pro-apoptotic receptor agonists. Nat Rev Drug Discov 2008;7:1001-12. 44. Burns TF, Bernhard EJ, El-Deiry WS. Tissue specific expression of p53 target genes suggests a key role for KILLER/DR5 in p53-dependent apoptosis in-vivo. Oncogene 2001;20:4601- 12. 45. Finnberg NK, Gokare P, Navaraj A, Lang Kuhs KA, Cerniglia G, Yagita H, et al. Agonists of the TRAIL Death Receptor DR5 Sensitize Intestinal Stem Cells to Chemotherapy-Induced Cell Death and Trigger Gastrointestinal Toxicity. Cancer research 2016;76:700-12. 46. Wang S, El-Deiry WS. Inducible silencing of KILLER/DR5 in-vivo promotes bioluminescent colon tumor xenograft growth and confers resistance to chemotherapeutic agent 5- fluorouracil. Cancer research 2004;64:6666-72. 47. Marchenko ND, Zaika A, Moll UM. Death signal-induced localization of p53 protein to mitochondria. A potential role in apoptotic signaling. The Journal of biological chemistry 2000;275:16202-12. 48. Mihara M, Erster S, Zaika A, Petrenko O, Chittenden T, Pancoska P, et al. p53 has a direct apoptogenic role at the mitochondria. Mol Cell 2003;11:577-90. 49. Chipuk JE, Bouchier-Hayes L, Kuwana T, Newmeyer DD, Green DR. PUMA couples the nuclear and cytoplasmic proapoptotic function of p53. Science 2005;309:1732-5. 50. Leu JI, Dumont P, Hafey M, Murphy ME, George DL. Mitochondrial p53 activates Bak and causes disruption of a Bak-Mcl1 complex. Nature cell biology 2004;6:443-50. 51. Eischen CM, Weber JD, Roussel MF, Sherr CJ, Cleveland JL. Disruption of the ARF-Mdm2- p53 tumor suppressor pathway in Myc-induced lymphomagenesis. Genes & development 1999;13:2658-69.

129 52. Michalak EM, Jansen ES, Happo L, Cragg MS, Tai L, Smyth GK, et al. Puma and to a lesser extent Noxa are suppressors of Myc-induced lymphomagenesis. Cell Death Differ 2009;16:684-96. 53. Schmitt CA, McCurrach ME, de Stanchina E, Wallace-Brodeur RR, Lowe SW. INK4a/ARF mutations accelerate lymphomagenesis and promote chemoresistance by disabling p53. Genes & development 1999;13:2670-7. 54. Aubrey BJ, Kelly GL, Kueh AJ, Brennan MS, O'Connor L, Milla L, et al. An inducible lentiviral guide RNA platform enables the identification of tumor-essential genes and tumor- promoting mutations in-vivo. Cell Rep 2015;10:1422-32. 55. Schmitt CA, Fridman JS, Yang M, Lee S, Baranov E, Hoffman RM, et al. A senescence program controlled by p53 and p16INK4a contributes to the outcome of cancer therapy. Cell 2002;109:335-46. 56. Garrison SP, Jeffers JR, Yang C, Nilsson JA, Hall MA, Rehg JE, et al. Selection against PUMA gene expression in Myc-driven B-cell lymphomagenesis. Molecular and cellular biology 2008;28:5391-402. 57. Hemann MT, Zilfou JT, Zhao Z, Burgess DJ, Hannon GJ, Lowe SW. Suppression of tumorigenesis by the p53 target PUMA. Proceedings of the National Academy of Sciences of the United States of America 2004;101:9333-8. 58. Clarke AR, Purdie CA, Harrison DJ, Morris RG, Bird CC, Hooper ML, et al. Thymocyte apoptosis induced by p53-dependent and independent pathways. Nature 1993;362:849- 52. 59. Symonds H, Krall L, Remington L, Saenz-Robles M, Lowe S, Jacks T, et al. p53-dependent apoptosis suppresses tumor growth and progression in-vivo. Cell 1994;78:703-11. 60. Xue W, Zender L, Miething C, Dickins RA, Hernando E, Krizhanovsky V, et al. Senescence and tumour clearance is triggered by p53 restoration in murine liver carcinomas. Nature 2007;445:656-60. 61. Liu G, Parant JM, Lang G, Chau P, Chavez-Reyes A, El-Naggar AK, et al. Chromosome stability, in the absence of apoptosis, is critical for suppression of tumorigenesis in Trp53 mutant mice. Nat Genet 2004;36:63-8. 62. Timofeev O, Schlereth K, Wanzel M, Braun A, Nieswandt B, Pagenstecher A, et al. p53 DNA binding cooperativity is essential for apoptosis and tumor suppression in-vivo. Cell Rep 2013;3:1512-25. 63. Li T, Kon N, Jiang L, Tan M, Ludwig T, Zhao Y, et al. Tumor suppression in the absence of p53-mediated cell-cycle arrest, apoptosis, and senescence. Cell 2012;149:1269-83. 64. Brady CA, Jiang D, Mello SS, Johnson TM, Jarvis LA, Kozak MM, et al. Distinct p53 transcriptional programs dictate acute DNA-damage responses and tumor suppression. Cell 2011;145:571-83. 65. Valente LJ, Gray DH, Michalak EM, Pinon-Hofbauer J, Egle A, Scott CL, et al. p53 efficiently suppresses tumor development in the complete absence of its cell-cycle inhibitory and proapoptotic effectors p21, Puma, and Noxa. Cell Rep 2013;3:1339-45. 66. Sax JK, Fei P, Murphy ME, Bernhard E, Korsmeyer SJ, El-Deiry WS. BID regulation by p53 contributes to chemosensitivity. Nature cell biology 2002;4:842-9. 67. Rikhof B, Corn PG, El-Deiry WS. Caspase 10 levels are increased following DNA damage in a p53-dependent manner. Cancer biology & therapy 2003;2:707-12.

130 68. Sheikh MS, Burns TF, Huang Y, Wu GS, Amundson S, Brooks KS, et al. p53-dependent and -independent regulation of the death receptor KILLER/DR5 gene expression in response to genotoxic stress and tumor necrosis factor alpha. Cancer research 1998;58:1593-8. 69. Adimoolam S, Ford JM. p53 and DNA damage-inducible expression of the xeroderma pigmentosum group C gene. Proceedings of the National Academy of Sciences of the United States of America 2002;99:12985-90. 70. Hastak K, Adimoolam S, Trinklein ND, Myers RM, Ford JM. Identification of a Functional In Vivo p53 Response Element in the Coding Sequence of the Xeroderma Pigmentosum Group C Gene. Genes Cancer 2012;3:131-40. 71. Hwang BJ, Ford JM, Hanawalt PC, Chu G. Expression of the p48 xeroderma pigmentosum gene is p53-dependent and is involved in global genomic repair. Proceedings of the National Academy of Sciences of the United States of America 1999;96:424-8. 72. Leveillard T, Andera L, Bissonnette N, Schaeffer L, Bracco L, Egly JM, et al. Functional interactions between p53 and the TFIIH complex are affected by tumour-associated mutations. The EMBO journal 1996;15:1615-24. 73. Wang XW, Yeh H, Schaeffer L, Roy R, Moncollin V, Egly JM, et al. p53 modulation of TFIIH- associated nucleotide excision repair activity. Nat Genet 1995;10:188-95. 74. Rubbi CP, Milner J. p53 is a chromatin accessibility factor for nucleotide excision repair of DNA damage. The EMBO journal 2003;22:975-86. 75. Hanson S, Kim E, Deppert W. Redox factor 1 (Ref-1) enhances specific DNA binding of p53 by promoting p53 tetramerization. Oncogene 2005;24:1641-7. 76. Cun Y, Dai N, Li M, Xiong C, Zhang Q, Sui J, et al. APE1/Ref-1 enhances DNA binding activity of mutant p53 in a redox-dependent manner. Oncol Rep 2014;31:901-9. 77. Zaky A, Busso C, Izumi T, Chattopadhyay R, Bassiouny A, Mitra S, et al. Regulation of the human AP-endonuclease (APE1/Ref-1) expression by the tumor suppressor p53 in response to DNA damage. Nucleic acids research 2008;36:1555-66. 78. Offer H, Erez N, Zurer I, Tang X, Milyavsky M, Goldfinger N, et al. The onset of p53- dependent DNA repair or apoptosis is determined by the level of accumulated damaged DNA. Carcinogenesis 2002;23:1025-32. 79. Zurer I, Hofseth LJ, Cohen Y, Xu-Welliver M, Hussain SP, Harris CC, et al. The role of p53 in base excision repair following genotoxic stress. Carcinogenesis 2004;25:11-9. 80. Chatterjee A, Mambo E, Osada M, Upadhyay S, Sidransky D. The effect of p53-RNAi and p53 knockout on human 8-oxoguanine DNA glycosylase (hOgg1) activity. FASEB J 2006;20:112-4. 81. Oka S, Leon J, Tsuchimoto D, Sakumi K, Nakabeppu Y. MUTYH, an adenine DNA glycosylase, mediates p53 tumor suppression via PARP-dependent cell death. Oncogenesis 2014;3:e121. 82. Achanta G, Huang P. Role of p53 in sensing oxidative DNA damage in response to reactive oxygen species-generating agents. Cancer research 2004;64:6233-9. 83. Zhou J, Ahn J, Wilson SH, Prives C. A role for p53 in base excision repair. The EMBO journal 2001;20:914-23. 84. Floter J, Kaymak I, Schulze A. Regulation of Metabolic Activity by p53. Metabolites 2017;7. 85. Ö W. Über den stoffwechsel der carcinomzelle. Die Naturwissenschaften 1924 ;12:1131-7.

131 86. Schwartzenberg-Bar-Yoseph F, Armoni M, Karnieli E. The tumor suppressor p53 down- regulates glucose transporters GLUT1 and GLUT4 gene expression. Cancer research 2004;64:2627-33. 87. Kawauchi K, Araki K, Tobiume K, Tanaka N. p53 regulates glucose metabolism through an IKK-NF-kappaB pathway and inhibits cell transformation. Nature cell biology 2008;10:611- 8. 88. Ganapathy S, Xiao S, Seo SJ, Lall R, Yang M, Xu T, et al. Low-dose arsenic induces chemotherapy protection via p53/NF-kappaB-mediated metabolic regulation. Oncogene 2014;33:1359-66. 89. Mathupala SP, Heese C, Pedersen PL. Glucose catabolism in cancer cells. The type II hexokinase promoter contains functionally active response elements for the tumor suppressor p53. The Journal of biological chemistry 1997;272:22776-80. 90. Bensaad K, Cheung EC, Vousden KH. Modulation of intracellular ROS levels by TIGAR controls autophagy. The EMBO journal 2009;28:3015-26. 91. Li H, Jogl G. Structural and biochemical studies of TIGAR (TP53-induced glycolysis and apoptosis regulator). The Journal of biological chemistry 2009;284:1748-54. 92. Endo F, Nishizuka SS, Kume K, Ishida K, Katagiri H, Ishida K, et al. A compensatory role of NF-kappaB to p53 in response to 5-FU-based chemotherapy for gastric cancer cell lines. PLoS One 2014;9:e90155. 93. Franklin DA, He Y, Leslie PL, Tikunov AP, Fenger N, Macdonald JM, et al. p53 coordinates DNA repair with nucleotide synthesis by suppressing PFKFB3 expression and promoting the pentose phosphate pathway. Sci Rep 2016;6:38067. 94. Jiang P, Du W, Wang X, Mancuso A, Gao X, Wu M, et al. p53 regulates biosynthesis through direct inactivation of glucose-6-phosphate dehydrogenase. Nature cell biology 2011;13:310-6. 95. Matoba S, Kang JG, Patino WD, Wragg A, Boehm M, Gavrilova O, et al. p53 regulates mitochondrial respiration. Science 2006;312:1650-3. 96. Kulawiec M, Ayyasamy V, Singh KK. p53 regulates mtDNA copy number and mitocheckpoint pathway. J Carcinog 2009;8:8. 97. Lebedeva MA, Eaton JS, Shadel GS. Loss of p53 causes mitochondrial DNA depletion and altered mitochondrial reactive oxygen species homeostasis. Biochim Biophys Acta 2009;1787:328-34. 98. Kitamura N, Nakamura Y, Miyamoto Y, Miyamoto T, Kabu K, Yoshida M, et al. Mieap, a p53-inducible protein, controls mitochondrial quality by repairing or eliminating unhealthy mitochondria. PLoS One 2011;6:e16060. 99. Boidot R, Vegran F, Meulle A, Le Breton A, Dessy C, Sonveaux P, et al. Regulation of monocarboxylate transporter MCT1 expression by p53 mediates inward and outward lactate fluxes in tumors. Cancer research 2012;72:939-48. 100. Contractor T, Harris CR. p53 negatively regulates transcription of the pyruvate dehydrogenase kinase Pdk2. Cancer research 2012;72:560-7. 101. Hu W, Zhang C, Wu R, Sun Y, Levine A, Feng Z. Glutaminase 2, a novel p53 target gene regulating energy metabolism and antioxidant function. Proceedings of the National Academy of Sciences of the United States of America 2010;107:7455-60.

132 102. Suzuki S, Tanaka T, Poyurovsky MV, Nagano H, Mayama T, Ohkubo S, et al. Phosphate- activated glutaminase (GLS2), a p53-inducible regulator of glutamine metabolism and reactive oxygen species. Proceedings of the National Academy of Sciences of the United States of America 2010;107:7461-6. 103. Polyak K, Xia Y, Zweier JL, Kinzler KW, Vogelstein B. A model for p53-induced apoptosis. Nature 1997;389:300-5. 104. Forrester K, Ambs S, Lupold SE, Kapust RB, Spillare EA, Weinberg WC, et al. Nitric oxide- induced p53 accumulation and regulation of inducible nitric oxide synthase expression by wild-type p53. Proceedings of the National Academy of Sciences of the United States of America 1996;93:2442-7. 105. Subbaramaiah K, Chung WJ, Michaluart P, Telang N, Tanabe T, Inoue H, et al. Resveratrol inhibits cyclooxygenase-2 transcription and activity in phorbol ester-treated human mammary epithelial cells. The Journal of biological chemistry 1998;273:21875-82. 106. Budanov AV, Sablina AA, Feinstein E, Koonin EV, Chumakov PM. Regeneration of peroxiredoxins by p53-regulated sestrins, homologs of bacterial AhpD. Science 2004;304:596-600. 107. Budanov AV, Karin M. p53 target genes sestrin1 and sestrin2 connect genotoxic stress and mTOR signaling. Cell 2008;134:451-60. 108. Cano CE, Gommeaux J, Pietri S, Culcasi M, Garcia S, Seux M, et al. Tumor protein 53- induced nuclear protein 1 is a major mediator of p53 antioxidant function. Cancer research 2009;69:219-26. 109. Chen W, Sun Z, Wang XJ, Jiang T, Huang Z, Fang D, et al. Direct interaction between Nrf2 and p21(Cip1/WAF1) upregulates the Nrf2-mediated antioxidant response. Mol Cell 2009;34:663-73. 110. Zhao Y, Chaiswing L, Velez JM, Batinic-Haberle I, Colburn NH, Oberley TD, et al. p53 translocation to mitochondria precedes its nuclear translocation and targets mitochondrial oxidative defense protein-manganese superoxide dismutase. Cancer research 2005;65:3745-50. 111. Jiang P, Du W, Mancuso A, Wellen KE, Yang X. Reciprocal regulation of p53 and malic enzymes modulates metabolism and senescence. Nature 2013;493:689-93. 112. Jiang L, Kon N, Li T, Wang SJ, Su T, Hibshoosh H, et al. Ferroptosis as a p53-mediated activity during tumour suppression. Nature 2015;520:57-62. 113. Dixon SJ, Lemberg KM, Lamprecht MR, Skouta R, Zaitsev EM, Gleason CE, et al. Ferroptosis: an iron-dependent form of nonapoptotic cell death. Cell 2012;149:1060-72. 114. Lowe SW, Ruley HE, Jacks T, Housman DE. p53-dependent apoptosis modulates the cytotoxicity of anticancer agents. Cell 1993;74:957-67. 115. Linke SP, Clarkin KC, Di Leonardo A, Tsou A, Wahl GM. A reversible, p53-dependent G0/G1 cell cycle arrest induced by ribonucleotide depletion in the absence of detectable DNA damage. Genes & development 1996;10:934-47. 116. Bunz F, Hwang PM, Torrance C, Waldman T, Zhang Y, Dillehay L, et al. Disruption of p53 in human cancer cells alters the responses to therapeutic agents. The Journal of clinical investigation 1999;104:263-9.

133 117. Tanaka H, Arakawa H, Yamaguchi T, Shiraishi K, Fukuda S, Matsui K, et al. A ribonucleotide reductase gene involved in a p53-dependent cell-cycle checkpoint for DNA damage. Nature 2000;404:42-9. 118. Chu E, Copur SM, Ju J, Chen TM, Khleif S, Voeller DM, et al. Thymidylate synthase protein and p53 mRNA form an in-vivo ribonucleoprotein complex. Molecular and cellular biology 1999;19:1582-94. 119. Reddy BA, van der Knaap JA, Bot AG, Mohd-Sarip A, Dekkers DH, Timmermans MA, et al. Nucleotide biosynthetic enzyme GMP synthase is a TRIM21-controlled relay of p53 stabilization. Mol Cell 2014;53:458-70. 120. Tsuchiya N, Izumiya M, Ogata-Kawata H, Okamoto K, Fujiwara Y, Nakai M, et al. Tumor suppressor miR-22 determines p53-dependent cellular fate through post-transcriptional regulation of p21. Cancer research 2011;71:4628-39. 121. Song B, Wang Y, Kudo K, Gavin EJ, Xi Y, Ju J. miR-192 Regulates dihydrofolate reductase and cellular proliferation through the p53-microRNA circuit. Clin Cancer Res 2008;14:8080-6. 122. Kim HR, Roe JS, Lee JE, Hwang IY, Cho EJ, Youn HD. A p53-inducible microRNA-34a downregulates Ras signaling by targeting IMPDH. Biochem Biophys Res Commun 2012;418:682-8. 123. Hong B, van den Heuvel AP, Prabhu VV, Zhang S, El-Deiry WS. Targeting tumor suppressor p53 for cancer therapy: strategies, challenges and opportunities. Curr Drug Targets 2014;15:80-9. 124. Levesque AA, Kohn EA, Bresnick E, Eastman A. Distinct roles for p53 transactivation and repression in preventing UCN-01-mediated abrogation of DNA damage-induced arrest at S and G2 cell cycle checkpoints. Oncogene 2005;24:3786-96. 125. Ma CX, Cai S, Li S, Ryan CE, Guo Z, Schaiff WT, et al. Targeting Chk1 in p53-deficient triple- negative breast cancer is therapeutically beneficial in human-in-mouse tumor models. The Journal of clinical investigation 2012;122:1541-52. 126. Gilad O, Nabet BY, Ragland RL, Schoppy DW, Smith KD, Durham AC, et al. Combining ATR suppression with oncogenic Ras synergistically increases genomic instability, causing synthetic lethality or tumorigenesis in a dosage-dependent manner. Cancer research 2010;70:9693-702. 127. Murga M, Campaner S, Lopez-Contreras AJ, Toledo LI, Soria R, Montana MF, et al. Exploiting oncogene-induced replicative stress for the selective killing of Myc-driven tumors. Nat Struct Mol Biol 2011;18:1331-5. 128. Schoppy DW, Ragland RL, Gilad O, Shastri N, Peters AA, Murga M, et al. Oncogenic stress sensitizes murine cancers to hypomorphic suppression of ATR. The Journal of clinical investigation 2012;122:241-52. 129. Fedier A, Schlamminger M, Schwarz VA, Haller U, Howell SB, Fink D. Loss of atm sensitises p53-deficient cells to topoisomerase poisons and antimetabolites. Ann Oncol 2003;14:938-45. 130. Jiang H, Reinhardt HC, Bartkova J, Tommiska J, Blomqvist C, Nevanlinna H, et al. The combined status of ATM and p53 link tumor development with therapeutic response. Genes & development 2009;23:1895-909.

134 131. Morandell S, Reinhardt HC, Cannell IG, Kim JS, Ruf DM, Mitra T, et al. A reversible gene- targeting strategy identifies synthetic lethal interactions between MK2 and p53 in the DNA damage response in-vivo. Cell Rep 2013;5:868-77. 132. Sur S, Pagliarini R, Bunz F, Rago C, Diaz LA, Jr., Kinzler KW, et al. A panel of isogenic human cancer cells suggests a therapeutic approach for cancers with inactivated p53. Proceedings of the National Academy of Sciences of the United States of America 2009;106:3964-9. 133. Maiyar AC, Phu PT, Huang AJ, Firestone GL. Repression of glucocorticoid receptor transactivation and DNA binding of a glucocorticoid response element within the serum/glucocorticoid-inducible protein kinase (sgk) gene promoter by the p53 tumor suppressor protein. Mol Endocrinol 1997;11:312-29. 134. Kavurma MM, Figg N, Bennett MR, Mercer J, Khachigian LM, Littlewood TD. Oxidative stress regulates IGF1R expression in vascular smooth-muscle cells via p53 and HDAC recruitment. Biochem J 2007;407:79-87. 135. Ohlsson C, Kley N, Werner H, LeRoith D. p53 regulates insulin-like growth factor-I (IGF-I) receptor expression and IGF-I-induced tyrosine phosphorylation in an osteosarcoma cell line: interaction between p53 and Sp1. Endocrinology 1998;139:1101-7. 136. Werner H, Karnieli E, Rauscher FJ, LeRoith D. Wild-type and mutant p53 differentially regulate transcription of the insulin-like growth factor I receptor gene. Proceedings of the National Academy of Sciences of the United States of America 1996;93:8318-23. 137. Kanaya T, Kyo S, Hamada K, Takakura M, Kitagawa Y, Harada H, et al. Adenoviral expression of p53 represses telomerase activity through down-regulation of human telomerase reverse transcriptase transcription. Clin Cancer Res 2000;6:1239-47. 138. Chun AC, Jin DY. Transcriptional regulation of mitotic checkpoint gene MAD1 by p53. The Journal of biological chemistry 2003;278:37439-50. 139. Hoffman WH, Biade S, Zilfou JT, Chen J, Murphy M. Transcriptional repression of the anti- apoptotic survivin gene by wild type p53. The Journal of biological chemistry 2002;277:3247-57. 140. Lin T, Chao C, Saito S, Mazur SJ, Murphy ME, Appella E, et al. p53 induces differentiation of mouse embryonic stem cells by suppressing Nanog expression. Nature cell biology 2005;7:165-71. 141. Ho JS, Ma W, Mao DY, Benchimol S. p53-Dependent transcriptional repression of c-myc is required for G1 cell cycle arrest. Molecular and cellular biology 2005;25:7423-31. 142. Nguyen TT, Cho K, Stratton SA, Barton MC. Transcription factor interactions and chromatin modifications associated with p53-mediated, developmental repression of the alpha-fetoprotein gene. Molecular and cellular biology 2005;25:2147-57. 143. Lee KC, Crowe AJ, Barton MC. p53-mediated repression of alpha-fetoprotein gene expression by specific DNA binding. Molecular and cellular biology 1999;19:1279-88. 144. Li B, Lee MY. Transcriptional regulation of the human DNA polymerase delta catalytic subunit gene POLD1 by p53 tumor suppressor and Sp1. The Journal of biological chemistry 2001;276:29729-39. 145. Ori A, Zauberman A, Doitsh G, Paran N, Oren M, Shaul Y. p53 binds and represses the HBV enhancer: an adjacent enhancer element can reverse the transcription effect of p53. The EMBO journal 1998;17:544-53.

135 146. Engeland K. Simplify p53: just an activator. Oncotarget 2015;6:3-4. 147. Fischer M, Quaas M, Nickel A, Engeland K. Indirect p53-dependent transcriptional repression of Survivin, CDC25C, and PLK1 genes requires the cyclin-dependent kinase inhibitor p21/CDKN1A and CDE/CHR promoter sites binding the DREAM complex. Oncotarget 2015;6:41402-17. 148. Tarasov V, Jung P, Verdoodt B, Lodygin D, Epanchintsev A, Menssen A, et al. Differential regulation of microRNAs by p53 revealed by massively parallel sequencing: miR-34a is a p53 target that induces apoptosis and G1-arrest. Cell cycle (Georgetown, Tex) 2007;6:1586-93. 149. Yamakuchi M, Ferlito M, Lowenstein CJ. miR-34a repression of SIRT1 regulates apoptosis. Proceedings of the National Academy of Sciences of the United States of America 2008;105:13421-6. 150. Pichiorri F, Suh SS, Rocci A, De Luca L, Taccioli C, Santhanam R, et al. Downregulation of p53-inducible microRNAs 192, 194, and 215 impairs the p53/MDM2 autoregulatory loop in multiple myeloma development. Cancer cell 2010;18:367-81. 151. Sachdeva M, Zhu S, Wu F, Wu H, Walia V, Kumar S, et al. p53 represses c-Myc through induction of the tumor suppressor miR-145. Proceedings of the National Academy of Sciences of the United States of America 2009;106:3207-12. 152. Suzuki HI, Yamagata K, Sugimoto K, Iwamoto T, Kato S, Miyazono K. Modulation of microRNA processing by p53. Nature 2009;460:529-33. 153. Huarte M, Guttman M, Feldser D, Garber M, Koziol MJ, Kenzelmann-Broz D, et al. A large intergenic noncoding RNA induced by p53 mediates global gene repression in the p53 response. Cell 2010;142:409-19. 154. Riley T, Sontag E, Chen P, Levine A. Transcriptional control of human p53-regulated genes. Nat Rev Mol Cell Biol 2008;9:402-12. 155. Ahn J, Murphy M, Kratowicz S, Wang A, Levine AJ, George DL. Down-regulation of the stathmin/Op18 and FKBP25 genes following p53 induction. Oncogene 1999;18:5954-8. 156. Zhang L, Yu D, Hu M, Xiong S, Lang A, Ellis LM, et al. Wild-type p53 suppresses angiogenesis in human leiomyosarcoma and synovial sarcoma by transcriptional suppression of vascular endothelial growth factor expression. Cancer research 2000;60:3655-61. 157. St Clair S, Manfredi JJ. The dual specificity phosphatase Cdc25C is a direct target for transcriptional repression by the tumor suppressor p53. Cell cycle (Georgetown, Tex) 2006;5:709-13. 158. St Clair S, Giono L, Varmeh-Ziaie S, Resnick-Silverman L, Liu WJ, Padi A, et al. DNA damage- induced downregulation of Cdc25C is mediated by p53 via two independent mechanisms: one involves direct binding to the cdc25C promoter. Mol Cell 2004;16:725-36. 159. McKenzie L, King S, Marcar L, Nicol S, Dias SS, Schumm K, et al. p53-dependent repression of polo-like kinase-1 (PLK1). Cell cycle (Georgetown, Tex) 2010;9:4200-12. 160. Johnson RA, Ince TA, Scotto KW. Transcriptional repression by p53 through direct binding to a novel DNA element. The Journal of biological chemistry 2001;276:27716-20. 161. Godar S, Ince TA, Bell GW, Feldser D, Donaher JL, Bergh J, et al. Growth-inhibitory and tumor- suppressive functions of p53 depend on its repression of CD44 expression. Cell 2008;134:62-73.

136 162. Heidelberger C, Chaudhuri NK, Danneberg P, Mooren D, Griesbach L, Duschinsky R, et al. Fluorinated pyrimidines, a new class of tumour-inhibitory compounds. Nature 1957;179:663-6. 163. Rutman RJ, Cantarow A, Paschkis KE. Studies in 2-acetylaminofluorene carcinogenesis. III. The utilization of uracil-2-C14 by preneoplastic rat liver and rat hepatoma. Cancer research 1954;14:119-23. 164. Jaffe JJ, Handschumacher RE, Welch AD. Studies on the carcinostatic activity in mice of 6- azauracil riboside (azauridine), in comparison with that of 6-azauracil. Yale J Biol Med 1957;30:168-75. 165. Diasio RB, Harris BE. Clinical pharmacology of 5-fluorouracil. Clin Pharmacokinet 1989;16:215-37. 166. Johnston PG, Kaye S. Capecitabine: a novel agent for the treatment of solid tumors. Anticancer Drugs 2001;12:639-46. 167. Schuller J, Cassidy J, Dumont E, Roos B, Durston S, Banken L, et al. Preferential activation of capecitabine in tumor following oral administration to colorectal cancer patients. Cancer Chemother Pharmacol 2000;45:291-7. 168. Cao D, Russell RL, Zhang D, Leffert JJ, Pizzorno G. Uridine phosphorylase (-/-) murine embryonic stem cells clarify the key role of this enzyme in the regulation of the pyrimidine salvage pathway and in the activation of fluoropyrimidines. Cancer research 2002;62:2313-7. 169. Johnston PG, Drake JC, Trepel J, Allegra CJ. Immunological quantitation of thymidylate synthase using the monoclonal antibody TS 106 in 5-fluorouracil-sensitive and -resistant human cancer cell lines. Cancer research 1992;52:4306-12. 170. Copur S, Aiba K, Drake JC, Allegra CJ, Chu E. Thymidylate synthase gene amplification in human colon cancer cell lines resistant to 5-fluorouracil. Biochem Pharmacol 1995;49:1419-26. 171. Lenz HJ, Hayashi K, Salonga D, Danenberg KD, Danenberg PV, Metzger R, et al. p53 point mutations and thymidylate synthase messenger RNA levels in disseminated colorectal cancer: an analysis of response and survival. Clin Cancer Res 1998;4:1243-50. 172. Johnston PG, Lenz HJ, Leichman CG, Danenberg KD, Allegra CJ, Danenberg PV, et al. Thymidylate synthase gene and protein expression correlate and are associated with response to 5-fluorouracil in human colorectal and gastric tumors. Cancer research 1995;55:1407-12. 173. Edler D, Blomgren H, Allegra CJ, Johnston PG, Lagerstedt U, Magnusson I, et al. Immunohistochemical determination of thymidylate synthase in colorectal cancer-- methodological studies. Eur J Cancer 1997;33:2278-81. 174. Houghton JA, Houghton PJ. Elucidation of pathways of 5-fluorouracil metabolism in xenografts of human colorectal adenocarcinoma. Eur J Cancer Clin Oncol 1983;19:807-15. 175. Peters GJ, Backus HH, Freemantle S, van Triest B, Codacci-Pisanelli G, van der Wilt CL, et al. Induction of thymidylate synthase as a 5-fluorouracil resistance mechanism. Biochim Biophys Acta 2002;1587:194-205. 176. Chu E, Voeller D, Koeller DM, Drake JC, Takimoto CH, Maley GF, et al. Identification of an RNA binding site for human thymidylate synthase. Proceedings of the National Academy of Sciences of the United States of America 1993;90:517-21.

137 177. Chu E, Voeller DM, Jones KL, Takechi T, Maley GF, Maley F, et al. Identification of a thymidylate synthase ribonucleoprotein complex in human colon cancer cells. Molecular and cellular biology 1994;14:207-13. 178. 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 Struct Funct 1995;20:191-7. 179. Marsh S, McKay JA, Cassidy J, McLeod HL. Polymorphism in the thymidylate synthase promoter enhancer region in colorectal cancer. Int J Oncol 2001;19:383-6. 180. Pullarkat ST, Stoehlmacher J, Ghaderi V, Xiong YP, Ingles SA, Sherrod A, et al. Thymidylate synthase gene polymorphism determines response and toxicity of 5-FU chemotherapy. Pharmacogenomics J 2001;1:65-70. 181. 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 Res 1999;19:3249-52. 182. Grem JL, Fischer PH. Enhancement of 5-fluorouracil's anticancer activity by dipyridamole. Pharmacol Ther 1989;40:349-71. 183. Fischer F, Baerenfaller K, Jiricny J. 5-Fluorouracil is efficiently removed from DNA by the base excision and mismatch repair systems. Gastroenterology 2007;133:1858-68. 184. 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;67:940-5. 185. Meyers M, Wagner MW, Mazurek A, Schmutte C, Fishel R, Boothman DA. DNA mismatch repair-dependent response to fluoropyrimidine-generated damage. The Journal of biological chemistry 2005;280:5516-26. 186. Carethers JM, Chauhan DP, Fink D, Nebel S, Bresalier RS, Howell SB, et al. Mismatch repair proficiency and in-vitro response to 5-fluorouracil. Gastroenterology 1999;117:123-31. 187. 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;61:5193-201. 188. Saridaki Z, Souglakos J, Georgoulias V. Prognostic and predictive significance of MSI in stages II/III colon cancer. World J Gastroenterol 2014;20:6809-14. 189. Kawakami H, Zaanan A, Sinicrope FA. Microsatellite instability testing and its role in the management of colorectal cancer. Curr Treat Options Oncol 2015;16:30. 190. Kanamaru R, Kakuta H, Sato T, Ishioka C, Wakui A. The inhibitory effects of 5-fluorouracil on the metabolism of preribosomal and ribosomal RNA in L-1210 cells in-vitro. Cancer Chemother Pharmacol 1986;17:43-6. 191. Ghoshal K, Jacob ST. Specific inhibition of pre-ribosomal RNA processing in extracts from the lymphosarcoma cells treated with 5-fluorouracil. Cancer research 1994;54:632-6. 192. Santi DV, Hardy LW. Catalytic mechanism and inhibition of tRNA (uracil-5- )methyltransferase: evidence for covalent catalysis. Biochemistry 1987;26:8599-606. 193. Randerath K, Tseng WC, Harris JS, Lu LJ. Specific effects of 5-fluoropyrimidines and 5- azapyrimidines on modification of the 5 position of pyrimidines, in particular the synthesis of 5-methyluracil and 5-methylcytosine in nucleic acids. Recent Results Cancer Res 1983;84:283-97.

138 194. Samuelsson T. Interactions of transfer RNA pseudouridine synthases with RNAs substituted with fluorouracil. Nucleic acids research 1991;19:6139-44. 195. Patton JR. Ribonucleoprotein particle assembly and modification of U2 small nuclear RNA containing 5-fluorouridine. Biochemistry 1993;32:8939-44. 196. Kaehler C, Isensee J, Hucho T, Lehrach H, Krobitsch S. 5-Fluorouracil affects assembly of stress granules based on RNA incorporation. Nucleic acids research 2014;42:6436-47. 197. Salonga D, Danenberg KD, Johnson M, Metzger R, Groshen S, Tsao-Wei DD, et al. Colorectal tumors responding to 5-fluorouracil have low gene expression levels of dihydropyrimidine dehydrogenase, thymidylate synthase, and thymidine phosphorylase. Clin Cancer Res 2000;6:1322-7. 198. Takebe N, Zhao SC, Ural AU, Johnson MR, Banerjee D, Diasio RB, et al. Retroviral transduction of human dihydropyrimidine dehydrogenase cDNA confers resistance to 5- fluorouracil in murine hematopoietic progenitor cells and human CD34+-enriched peripheral blood progenitor cells. Cancer Gene Ther 2001;8:966-73. 199. Fukuda H, Takiguchi N, Koda K, Oda K, Seike K, Miyazaki M. Thymidylate synthase and dihydropyrimidine dehydrogenase are related to histological effects of 5-fluorouracil and cisplatin neoadjuvant chemotherapy for primary gastric cancer patients. Cancer Invest 2006;24:235-41. 200. Milano G, Etienne MC, Pierrefite V, Barberi-Heyob M, Deporte-Fety R, Renee N. Dihydropyrimidine dehydrogenase deficiency and fluorouracil-related toxicity. Br J Cancer 1999;79:627-30. 201. Ahnen DJ, Feigl P, Quan G, Fenoglio-Preiser C, Lovato LC, Bunn PA, Jr., et al. Ki-ras mutation and p53 overexpression predict the clinical behavior of colorectal cancer: a Southwest Oncology Group study. Cancer research 1998;58:1149-58. 202. Elsaleh H, Powell B, McCaul K, Grieu F, Grant R, Joseph D, et al. P53 alteration and microsatellite instability have predictive value for survival benefit from chemotherapy in stage III colorectal carcinoma. Clin Cancer Res 2001;7:1343-9. 203. Liang JT, Huang KC, Cheng YM, Hsu HC, Cheng AL, Hsu CH, et al. P53 overexpression predicts poor chemosensitivity to high-dose 5-fluorouracil plus leucovorin chemotherapy for stage IV colorectal cancers after palliative bowel resection. Int J Cancer 2002;97:451- 7. 204. Calascibetta A, Martorana A, Cabibi D, Aragona F, Sanguedolce R. Relationship between thymidylate synthase and p53 and response to FEC versus taxane adjuvant chemotherapy for breast carcinoma. J Chemother 2011;23:354-7. 205. Paradiso A, Simone G, Petroni S, Leone B, Vallejo C, Lacava J, et al. Thymidilate synthase and p53 primary tumour expression as predictive factors for advanced colorectal cancer patients. Br J Cancer 2000;82:560-7. 206. Sjogren S, Inganas M, Norberg T, Lindgren A, Nordgren H, Holmberg L, et al. The p53 gene in breast cancer: prognostic value of complementary DNA sequencing versus immunohistochemistry. J Natl Cancer Inst 1996;88:173-82. 207. Ladner RD, Lynch FJ, Groshen S, Xiong YP, Sherrod A, Caradonna SJ, et al. dUTP nucleotidohydrolase isoform expression in normal and neoplastic tissues: association with survival and response to 5-fluorouracil in colorectal cancer. Cancer research 2000;60:3493-503.

139 208. Kawahara A, Akagi Y, Hattori S, Mizobe T, Shirouzu K, Ono M, et al. Higher expression of deoxyuridine triphosphatase (dUTPase) may predict the metastasis potential of colorectal cancer. J Clin Pathol 2009;62:364-9. 209. Nitiss JL. Targeting DNA topoisomerase II in cancer chemotherapy. Nat Rev Cancer 2009;9:338-50. 210. Burgess DJ, Doles J, Zender L, Xue W, Ma B, McCombie WR, et al. Topoisomerase levels determine chemotherapy response in-vitro and in-vivo. Proceedings of the National Academy of Sciences of the United States of America 2008;105:9053-8. 211. Cortes Ledesma F, El Khamisy SF, Zuma MC, Osborn K, Caldecott KW. A human 5'-tyrosyl DNA phosphodiesterase that repairs topoisomerase-mediated DNA damage. Nature 2009;461:674-8. 212. Malik M, Nitiss JL. DNA repair functions that control sensitivity to topoisomerase- targeting drugs. Eukaryot Cell 2004;3:82-90. 213. Zoppoli G, Regairaz M, Leo E, Reinhold WC, Varma S, Ballestrero A, et al. Putative DNA/RNA helicase Schlafen-11 (SLFN11) sensitizes cancer cells to DNA-damaging agents. Proceedings of the National Academy of Sciences of the United States of America 2012;109:15030-5. 214. Biard DS. Untangling the relationships between DNA repair pathways by silencing more than 20 DNA repair genes in human stable clones. Nucleic acids research 2007;35:3535- 50. 215. Jin S, Inoue S, Weaver DT. Differential etoposide sensitivity of cells deficient in the Ku and DNA-PKcs components of the DNA-dependent protein kinase. Carcinogenesis 1998;19:965-71. 216. Willmore E, de Caux S, Sunter NJ, Tilby MJ, Jackson GH, Austin CA, et al. A novel DNA- dependent protein kinase inhibitor, NU7026, potentiates the cytotoxicity of topoisomerase II poisons used in the treatment of leukemia. Blood 2004;103:4659-65. 217. Beck WT, Danks MK, Wolverton JS, Kim R, Chen M. Drug resistance associated with altered DNA topoisomerase II. Adv Enzyme Regul 1993;33:113-27. 218. Nitiss JL, Beck WT. Antitopoisomerase drug action and resistance. Eur J Cancer 1996;32A:958-66. 219. Coutts J, Plumb JA, Brown R, Keith WN. Expression of topoisomerase II alpha and beta in an adenocarcinoma cell line carrying amplified topoisomerase II alpha and retinoic acid receptor alpha genes. Br J Cancer 1993;68:793-800. 220. Keith WN, Douglas F, Wishart GC, McCallum HM, George WD, Kaye SB, et al. Co- amplification of erbB2, topoisomerase II alpha and retinoic acid receptor alpha genes in breast cancer and allelic loss at topoisomerase I on chromosome 20. Eur J Cancer 1993;29A:1469-75. 221. Dawkins HJ, Robbins PD, Smith KL, Sarna M, Harvey JM, Sterrett GF, et al. What's new in breast cancer? Molecular perspectives of cancer development and the role of the oncogene c-erbB-2 in prognosis and disease. Pathol Res Pract 1993;189:1233-52. 222. Azarova AM, Lyu YL, Lin CP, Tsai YC, Lau JY, Wang JC, et al. Roles of DNA topoisomerase II isozymes in chemotherapy and secondary malignancies. Proceedings of the National Academy of Sciences of the United States of America 2007;104:11014-9.

140 223. Cowell IG, Sondka Z, Smith K, Lee KC, Manville CM, Sidorczuk-Lesthuruge M, et al. Model for MLL translocations in therapy-related leukemia involving topoisomerase IIbeta- mediated DNA strand breaks and gene proximity. Proceedings of the National Academy of Sciences of the United States of America 2012;109:8989-94. 224. Haffner MC, Aryee MJ, Toubaji A, Esopi DM, Albadine R, Gurel B, et al. Androgen-induced TOP2B-mediated double-strand breaks and prostate cancer gene rearrangements. Nat Genet 2010;42:668-75. 225. Canela A, Maman Y, Jung S, Wong N, Callen E, Day A, et al. Genome Organization Drives Chromosome Fragility. Cell 2017;170:507-21 e18. 226. Zhang S, Liu X, Bawa-Khalfe T, Lu LS, Lyu YL, Liu LF, et al. Identification of the molecular basis of doxorubicin-induced cardiotoxicity. Nat Med 2012;18:1639-42. 227. Kobayashi K, Ratain MJ. Pharmacodynamics and long-term toxicity of etoposide. Cancer Chemother Pharmacol 1994;34 Suppl:S64-8. 228. Johnson MR, Wang K, Tillmanns S, Albin N, Diasio RB. Structural organization of the human dihydropyrimidine dehydrogenase gene. Cancer research 1997;57:1660-3. 229. Wei X, Elizondo G, Sapone A, McLeod HL, Raunio H, Fernandez-Salguero P, et al. Characterization of the human dihydropyrimidine dehydrogenase gene. Genomics 1998;51:391-400. 230. Yokota H, Fernandez-Salguero P, Furuya H, Lin K, McBride OW, Podschun B, et al. cDNA cloning and chromosome mapping of human dihydropyrimidine dehydrogenase, an enzyme associated with 5-fluorouracil toxicity and congenital thymine uraciluria. The Journal of biological chemistry 1994;269:23192-6. 231. Kent WJ, Sugnet CW, Furey TS, Roskin KM, Pringle TH, Zahler AM, et al. The human genome browser at UCSC. Genome Res 2002;12:996-1006. 232. Dobritzsch D, Ricagno S, Schneider G, Schnackerz KD, Lindqvist Y. Crystal structure of the productive ternary complex of dihydropyrimidine dehydrogenase with NADPH and 5- iodouracil. Implications for mechanism of inhibition and electron transfer. The Journal of biological chemistry 2002;277:13155-66. 233. Sykes DB, Kfoury YS, Mercier FE, Wawer MJ, Law JM, Haynes MK, et al. Inhibition of Dihydroorotate Dehydrogenase Overcomes Differentiation Blockade in Acute Myeloid Leukemia. Cell 2016;167:171-86 e15. 234. Ukon K, Tanimoto K, Shimokuni T, Noguchi T, Hiyama K, Tsujimoto H, et al. Activator protein accelerates dihydropyrimidine dehydrogenase gene transcription in cancer cells. Cancer research 2005;65:1055-62. 235. Zhang X, Li L, Fourie J, Davie JR, Guarcello V, Diasio RB. The role of Sp1 and Sp3 in the constitutive DPYD gene expression. Biochim Biophys Acta 2006;1759:247-56. 236. Harris BE, Song R, Soong SJ, Diasio RB. Relationship between dihydropyrimidine dehydrogenase activity and plasma 5-fluorouracil levels with evidence for circadian variation of enzyme activity and plasma drug levels in cancer patients receiving 5- fluorouracil by protracted continuous infusion. Cancer research 1990;50:197-201. 237. Harris BE, Song RL, He YJ, Soong SJ, Diasio RB. Circadian rhythm of rat liver dihydropyrimidine dehydrogenase. Possible relevance to fluoropyrimidine chemotherapy. Biochem Pharmacol 1988;37:4759-62.

141 238. Raida M, Kliche KO, Schwabe W, Hausler P, Clement JH, Behnke D, et al. Circadian variation of dihydropyrimidine dehydrogenase mRNA expression in leukocytes and serum cortisol levels in patients with advanced gastrointestinal carcinomas compared to healthy controls. J Cancer Res Clin Oncol 2002;128:96-102. 239. Offer SM, Butterfield GL, Jerde CR, Fossum CC, Wegner NJ, Diasio RB. microRNAs miR-27a and miR-27b directly regulate liver dihydropyrimidine dehydrogenase expression through two conserved binding sites. Mol Cancer Ther 2014;13:742-51. 240. Chai J, Dong W, Xie C, Wang L, Han DL, Wang S, et al. MicroRNA-494 sensitizes colon cancer cells to fluorouracil through regulation of DPYD. IUBMB Life 2015;67:191-201. 241. Cai D, He K, Chang S, Tong D, Huang C. MicroRNA-302b Enhances the Sensitivity of Hepatocellular Carcinoma Cell Lines to 5-FU via Targeting Mcl-1 and DPYD. Int J Mol Sci 2015;16:23668-82. 242. Zhang X, Soong R, Wang K, Li L, Davie JR, Guarcello V, et al. Suppression of DPYD expression in RKO cells via DNA methylation in the regulatory region of the DPYD promoter: a potentially important epigenetic mechanism regulating DPYD expression. Biochem Cell Biol 2007;85:337-46. 243. Noguchi T, Tanimoto K, Shimokuni T, Ukon K, Tsujimoto H, Fukushima M, et al. Aberrant methylation of DPYD promoter, DPYD expression, and cellular sensitivity to 5-fluorouracil in cancer cells. Clin Cancer Res 2004;10:7100-7. 244. Ezzeldin HH, Lee AM, Mattison LK, Diasio RB. Methylation of the DPYD promoter: an alternative mechanism for dihydropyrimidine dehydrogenase deficiency in cancer patients. Clin Cancer Res 2005;11:8699-705. 245. Yu J, McLeod HL, Ezzeldin HH, Diasio RB. Methylation of the DPYD promoter and dihydropyrimidine dehydrogenase deficiency. Clin Cancer Res 2006;12:3864; author reply 246. Jubb AM, Quirke P, Oates AJ. DNA methylation, a biomarker for colorectal cancer: implications for screening and pathological utility. Ann N Y Acad Sci 2003;983:251-67. 247. Berger R, Stoker-de Vries SA, Wadman SK, Duran M, Beemer FA, de Bree PK, et al. Dihydropyrimidine dehydrogenase deficiency leading to thymine-uraciluria. An inborn error of pyrimidine metabolism. Clin Chim Acta 1984;141:227-34. 248. Diasio RB, Beavers TL, Carpenter JT. Familial deficiency of dihydropyrimidine dehydrogenase. Biochemical basis for familial pyrimidinemia and severe 5-fluorouracil- induced toxicity. The Journal of clinical investigation 1988;81:47-51. 249. Meulendijks D, Cats A, Beijnen JH, Schellens JH. Improving safety of fluoropyrimidine chemotherapy by individualizing treatment based on dihydropyrimidine dehydrogenase activity - Ready for clinical practice? Cancer Treat Rev 2016;50:23-34. 250. Caudle KE, Thorn CF, Klein TE, Swen JJ, McLeod HL, Diasio RB, et al. Clinical Pharmacogenetics Implementation Consortium guidelines for dihydropyrimidine dehydrogenase genotype and fluoropyrimidine dosing. Clin Pharmacol Ther 2013;94:640- 5. 251. Etienne-Grimaldi MC, Boyer JC, Beroud C, Mbatchi L, van Kuilenburg A, Bobin-Dubigeon C, et al. New advances in DPYD genotype and risk of severe toxicity under capecitabine. PLoS One 2017;12:e0175998.

142 252. Chazal M, Etienne MC, Renee N, Bourgeon A, Richelme H, Milano G. Link between dihydropyrimidine dehydrogenase activity in peripheral blood mononuclear cells and liver. Clin Cancer Res 1996;2:507-10. 253. Kristensen MH, Pedersen P, Mejer J. The value of dihydrouracil/uracil plasma ratios in predicting 5-fluorouracil-related toxicity in colorectal cancer patients. J Int Med Res 2010;38:1313-23. 254. Ciccolini J, Mercier C, Blachon MF, Favre R, Durand A, Lacarelle B. A simple and rapid high- performance liquid chromatographic (HPLC) method for 5-fluorouracil (5-FU) assay in plasma and possible detection of patients with impaired dihydropyrimidine dehydrogenase (DPD) activity. J Clin Pharm Ther 2004;29:307-15. 255. Gamelin E, Boisdron-Celle M, Guerin-Meyer V, Delva R, Lortholary A, Genevieve F, et al. Correlation between uracil and dihydrouracil plasma ratio, fluorouracil (5-FU) pharmacokinetic parameters, and tolerance in patients with advanced colorectal cancer: A potential interest for predicting 5-FU toxicity and determining optimal 5-FU dosage. J Clin Oncol 1999;17:1105. 256. Jiang H, Lu J, Jiang J, Hu P. Important role of the dihydrouracil/uracil ratio in marked interpatient variations of fluoropyrimidine pharmacokinetics and pharmacodynamics. J Clin Pharmacol 2004;44:1260-72. 257. Carlsson G, Odin E, Gustavsson B, Wettergren Y. Pretherapeutic uracil and dihydrouracil levels in saliva of colorectal cancer patients are associated with toxicity during adjuvant 5-fluorouracil-based chemotherapy. Cancer Chemother Pharmacol 2014;74:757-63. 258. Wettergren Y, Carlsson G, Odin E, Gustavsson B. Pretherapeutic uracil and dihydrouracil levels of colorectal cancer patients are associated with sex and toxic side effects during adjuvant 5-fluorouracil-based chemotherapy. Cancer 2012;118:2935-43. 259. Meulendijks D, Henricks LM, Jacobs BAW, Aliev A, Deenen MJ, de Vries N, et al. Pretreatment serum uracil concentration as a predictor of severe and fatal fluoropyrimidine-associated toxicity. Br J Cancer 2017;116:1415-24. 260. Cunha-Junior GF, De Marco L, Bastos-Rodrigues L, Bolina MB, Martins FL, Pianetti GA, et al. (13)C-uracil breath test to predict 5-fluorouracil toxicity in gastrointestinal cancer patients. Cancer Chemother Pharmacol 2013;72:1273-82. 261. van Staveren MC, Theeuwes-Oonk B, Guchelaar HJ, van Kuilenburg AB, Maring JG. Pharmacokinetics of orally administered uracil in healthy volunteers and in DPD-deficient patients, a possible tool for screening of DPD deficiency. Cancer Chemother Pharmacol 2011;68:1611-7. 262. Matsuoka S, Huang M, Elledge SJ. Linkage of ATM to cell cycle regulation by the Chk2 protein kinase. Science 1998;282:1893-7. 263. Zannini L, Delia D, Buscemi G. CHK2 kinase in the DNA damage response and beyond. J Mol Cell Biol 2014;6:442-57. 264. Buscemi G, Carlessi L, Zannini L, Lisanti S, Fontanella E, Canevari S, et al. DNA damage- induced cell cycle regulation and function of novel Chk2 phosphoresidues. Molecular and cellular biology 2006;26:7832-45. 265. Matsuoka S, Ballif BA, Smogorzewska A, McDonald ER, 3rd, Hurov KE, Luo J, et al. ATM and ATR substrate analysis reveals extensive protein networks responsive to DNA damage. Science 2007;316:1160-6.

143 266. Ahn JY, Schwarz JK, Piwnica-Worms H, Canman CE. Threonine 68 phosphorylation by ataxia telangiectasia mutated is required for efficient activation of Chk2 in response to ionizing radiation. Cancer research 2000;60:5934-6. 267. Lee CH, Chung JH. The hCds1 (Chk2)-FHA domain is essential for a chain of phosphorylation events on hCds1 that is induced by ionizing radiation. The Journal of biological chemistry 2001;276:30537-41. 268. Bahassi el M, Myer DL, McKenney RJ, Hennigan RF, Stambrook PJ. Priming phosphorylation of Chk2 by polo-like kinase 3 (Plk3) mediates its full activation by ATM and a downstream checkpoint in response to DNA damage. Mutat Res 2006;596:166-76. 269. Adamson AW, Beardsley DI, Kim WJ, Gao Y, Baskaran R, Brown KD. Methylator-induced, mismatch repair-dependent G2 arrest is activated through Chk1 and Chk2. Molecular biology of the cell 2005;16:1513-26. 270. Antoni L, Sodha N, Collins I, Garrett MD. CHK2 kinase: cancer susceptibility and cancer therapy - two sides of the same coin? Nat Rev Cancer 2007;7:925-36. 271. Yang S, Jeong JH, Brown AL, Lee CH, Pandolfi PP, Chung JH, et al. Promyelocytic leukemia activates Chk2 by mediating Chk2 autophosphorylation. The Journal of biological chemistry 2006;281:26645-54. 272. Freeman AK, Dapic V, Monteiro AN. Negative regulation of CHK2 activity by protein phosphatase 2A is modulated by DNA damage. Cell cycle (Georgetown, Tex) 2010;9:736- 47. 273. Carlessi L, Buscemi G, Fontanella E, Delia D. A protein phosphatase feedback mechanism regulates the basal phosphorylation of Chk2 kinase in the absence of DNA damage. Biochim Biophys Acta 2010;1803:1213-23. 274. Fujimoto H, Onishi N, Kato N, Takekawa M, Xu XZ, Kosugi A, et al. Regulation of the antioncogenic Chk2 kinase by the oncogenic Wip1 phosphatase. Cell Death Differ 2006;13:1170-80. 275. van Vugt MA, Gardino AK, Linding R, Ostheimer GJ, Reinhardt HC, Ong SE, et al. A mitotic phosphorylation feedback network connects Cdk1, Plk1, 53BP1, and Chk2 to inactivate the G(2)/M DNA damage checkpoint. PLoS Biol 2010;8:e1000287. 276. Bohgaki M, Hakem A, Halaby MJ, Bohgaki T, Li Q, Bissey PA, et al. The E3 ligase PIRH2 polyubiquitylates CHK2 and regulates its turnover. Cell Death Differ 2013;20:812-22. 277. Lee JS, Collins KM, Brown AL, Lee CH, Chung JH. hCds1-mediated phosphorylation of BRCA1 regulates the DNA damage response. Nature 2000;404:201-4. 278. Bahassi EM, Ovesen JL, Riesenberg AL, Bernstein WZ, Hasty PE, Stambrook PJ. The checkpoint kinases Chk1 and Chk2 regulate the functional associations between hBRCA2 and Rad51 in response to DNA damage. Oncogene 2008;27:3977-85. 279. Tan Y, Raychaudhuri P, Costa RH. Chk2 mediates stabilization of the FoxM1 transcription factor to stimulate expression of DNA repair genes. Molecular and cellular biology 2007;27:1007-16. 280. Zhang J, Willers H, Feng Z, Ghosh JC, Kim S, Weaver DT, et al. Chk2 phosphorylation of BRCA1 regulates DNA double-strand break repair. Molecular and cellular biology 2004;24:708-18.

144 281. Goodarzi AA, Kurka T, Jeggo PA. KAP-1 phosphorylation regulates CHD3 nucleosome remodeling during the DNA double-strand break response. Nat Struct Mol Biol 2011;18:831-9. 282. Bolderson E, Savage KI, Mahen R, Pisupati V, Graham ME, Richard DJ, et al. Kruppel- associated Box (KRAB)-associated co-repressor (KAP-1) Ser-473 phosphorylation regulates heterochromatin protein 1beta (HP1-beta) mobilization and DNA repair in heterochromatin. The Journal of biological chemistry 2012;287:28122-31. 283. Hu C, Zhang S, Gao X, Gao X, Xu X, Lv Y, et al. Roles of Kruppel-associated Box (KRAB)- associated Co-repressor KAP1 Ser-473 Phosphorylation in DNA Damage Response. The Journal of biological chemistry 2012;287:18937-52. 284. Falck J, Mailand N, Syljuasen RG, Bartek J, Lukas J. The ATM-Chk2-Cdc25A checkpoint pathway guards against radioresistant DNA synthesis. Nature 2001;410:842-7. 285. Sorensen CS, Syljuasen RG, Falck J, Schroeder T, Ronnstrand L, Khanna KK, et al. Chk1 regulates the S phase checkpoint by coupling the physiological turnover and ionizing radiation-induced accelerated proteolysis of Cdc25A. Cancer cell 2003;3:247-58. 286. Hirao A, Kong YY, Matsuoka S, Wakeham A, Ruland J, Yoshida H, et al. DNA damage- induced activation of p53 by the checkpoint kinase Chk2. Science 2000;287:1824-7. 287. Chehab NH, Malikzay A, Appel M, Halazonetis TD. Chk2/hCds1 functions as a DNA damage checkpoint in G(1) by stabilizing p53. Genes & development 2000;14:278-88. 288. Dornan D, Shimizu H, Perkins ND, Hupp TR. DNA-dependent acetylation of p53 by the transcription coactivator p300. The Journal of biological chemistry 2003;278:13431-41. 289. Ou YH, Chung PH, Sun TP, Shieh SY. p53 C-terminal phosphorylation by CHK1 and CHK2 participates in the regulation of DNA-damage-induced C-terminal acetylation. Molecular biology of the cell 2005;16:1684-95. 290. Inoue Y, Kitagawa M, Taya Y. Phosphorylation of pRB at Ser612 by Chk1/2 leads to a complex between pRB and E2F-1 after DNA damage. The EMBO journal 2007;26:2083-93. 291. Okada N, Yabuta N, Suzuki H, Aylon Y, Oren M, Nojima H. A novel Chk1/2-Lats2-14-3-3 signaling pathway regulates P-body formation in response to UV damage. J Cell Sci 2011;124:57-67. 292. Blasina A, de Weyer IV, Laus MC, Luyten WH, Parker AE, McGowan CH. A human homologue of the checkpoint kinase Cds1 directly inhibits Cdc25 phosphatase. Curr Biol 1999;9:1-10. 293. Chaturvedi P, Eng WK, Zhu Y, Mattern MR, Mishra R, Hurle MR, et al. Mammalian Chk2 is a downstream effector of the ATM-dependent DNA damage checkpoint pathway. Oncogene 1999;18:4047-54. 294. Bruno T, De Nicola F, Iezzi S, Lecis D, D'Angelo C, Di Padova M, et al. Che-1 phosphorylation by ATM/ATR and Chk2 kinases activates p53 transcription and the G2/M checkpoint. Cancer cell 2006;10:473-86. 295. Chen L, Gilkes DM, Pan Y, Lane WS, Chen J. ATM and Chk2-dependent phosphorylation of MDMX contribute to p53 activation after DNA damage. The EMBO journal 2005;24:3411- 22. 296. Yeh YH, Huang YF, Lin TY, Shieh SY. The cell cycle checkpoint kinase CHK2 mediates DNA damage-induced stabilization of TTK/hMps1. Oncogene 2009;28:1366-78.

145 297. Jack MT, Woo RA, Motoyama N, Takai H, Lee PW. DNA-dependent protein kinase and checkpoint kinase 2 synergistically activate a latent population of p53 upon DNA damage. The Journal of biological chemistry 2004;279:15269-73. 298. Li J, Stern DF. Regulation of CHK2 by DNA-dependent protein kinase. The Journal of biological chemistry 2005;280:12041-50. 299. Stevens C, Smith L, La Thangue NB. Chk2 activates E2F-1 in response to DNA damage. Nature cell biology 2003;5:401-9. 300. Ghosh JC, Dohi T, Raskett CM, Kowalik TF, Altieri DC. Activated checkpoint kinase 2 provides a survival signal for tumor cells. Cancer research 2006;66:11576-9. 301. Yang S, Kuo C, Bisi JE, Kim MK. PML-dependent apoptosis after DNA damage is regulated by the checkpoint kinase hCds1/Chk2. Nature cell biology 2002;4:865-70. 302. Masuda K, Abdelmohsen K, Kim MM, Srikantan S, Lee EK, Tominaga K, et al. Global dissociation of HuR-mRNA complexes promotes cell survival after ionizing radiation. The EMBO journal 2011;30:1040-53. 303. Bahassi el M, Conn CW, Myer DL, Hennigan RF, McGowan CH, Sanchez Y, et al. Mammalian Polo-like kinase 3 (Plk3) is a multifunctional protein involved in stress response pathways. Oncogene 2002;21:6633-40. 304. Tsvetkov L, Xu X, Li J, Stern DF. Polo-like kinase 1 and Chk2 interact and co-localize to centrosomes and the midbody. The Journal of biological chemistry 2003;278:8468-75. 305. Zhang S, Hemmerich P, Grosse F. Centrosomal localization of DNA damage checkpoint proteins. J Cell Biochem 2007;101:451-65. 306. Shang ZF, Huang B, Xu QZ, Zhang SM, Fan R, Liu XD, et al. Inactivation of DNA-dependent protein kinase leads to spindle disruption and mitotic catastrophe with attenuated checkpoint protein 2 Phosphorylation in response to DNA damage. Cancer research 2010;70:3657-66. 307. Yu Q, Rose JH, Zhang H, Pommier Y. Antisense inhibition of Chk2/hCds1 expression attenuates DNA damage-induced S and G2 checkpoints and enhances apoptotic activity in HEK-293 cells. FEBS Lett 2001;505:7-12. 308. Larson G, Yan S, Chen H, Rong F, Hong Z, Wu JZ. Identification of novel, selective and potent Chk2 inhibitors. Bioorg Med Chem Lett 2007;17:172-5. 309. Carlessi L, Buscemi G, Larson G, Hong Z, Wu JZ, Delia D. Biochemical and cellular characterization of VRX0466617, a novel and selective inhibitor for the checkpoint kinase Chk2. Mol Cancer Ther 2007;6:935-44. 310. Velic D, Couturier AM, Ferreira MT, Rodrigue A, Poirier GG, Fleury F, et al. DNA Damage Signalling and Repair Inhibitors: The Long-Sought-After Achilles' Heel of Cancer. Biomolecules 2015;5:3204-59. 311. Pires IM, Ward TH, Dive C. Oxaliplatin responses in colorectal cancer cells are modulated by CHK2 kinase inhibitors. Br J Pharmacol 2010;159:1326-38. 312. Anderson VE, Walton MI, Eve PD, Boxall KJ, Antoni L, Caldwell JJ, et al. CCT241533 is a potent and selective inhibitor of CHK2 that potentiates the cytotoxicity of PARP inhibitors. Cancer research 2011;71:463-72. 313. Wasternack C. Degradation of pyrimidines and pyrimidine analogs--pathways and mutual influences. Pharmacol Ther 1980;8:629-51.

146 314. Shaul YD, Freinkman E, Comb WC, Cantor JR, Tam WL, Thiru P, et al. Dihydropyrimidine accumulation is required for the epithelial-mesenchymal transition. Cell 2014;158:1094- 109. 315. Danneberg PB, Montag BJ, Heidelberger C. Studies on fluorinated pyrimidines. IV. Effects on nucleic acid metabolism in-vivo. Cancer research 1958;18:329-34. 316. Kline CL, Sheikh HS, Scicchitano A, Gingrich R, Beachler C, Finnberg NK, et al. Preliminary observations indicate variable patterns of plasma 5-fluorouracil (5-FU) levels during dose optimization of infusional 5-FU in colorectal cancer patients. Cancer biology & therapy 2011;12:557-68. 317. Santi DV, McHenry CS. 5-Fluoro-2'-deoxyuridylate: covalent complex with thymidylate synthetase. Proceedings of the National Academy of Sciences of the United States of America 1972;69:1855-7. 318. Santi DV, McHenry CS, Sommer H. Mechanism of interaction of thymidylate synthetase with 5-fluorodeoxyuridylate. Biochemistry 1974;13:471-81. 319. Deenen MJ, Tol J, Burylo AM, Doodeman VD, de Boer A, Vincent A, et al. Relationship between single nucleotide polymorphisms and haplotypes in DPYD and toxicity and efficacy of capecitabine in advanced colorectal cancer. Clin Cancer Res 2011;17:3455-68. 320. Meinsma R, Fernandez-Salguero P, Van Kuilenburg AB, Van Gennip AH, Gonzalez FJ. Human polymorphism in drug metabolism: mutation in the dihydropyrimidine dehydrogenase gene results in exon skipping and thymine uracilurea. DNA Cell Biol 1995;14:1-6. 321. Offer SM, Fossum CC, Wegner NJ, Stuflesser AJ, Butterfield GL, Diasio RB. Comparative functional analysis of DPYD variants of potential clinical relevance to dihydropyrimidine dehydrogenase activity. Cancer research 2014;74:2545-54. 322. Saif MW, Rosen LS, Saito K, Zergebel C, Ravage-Mass L, Mendelson DS. A phase I study evaluating the effect of CDHP as a component of S-1 on the pharmacokinetics of 5- fluorouracil. Anticancer Res 2011;31:625-32. 323. Rivera E, Chang JC, Semiglazov V, Burdaeva O, Kirby MG, Spector T. Eniluracil plus 5- fluorouracil and leucovorin: treatment for metastatic breast cancer patients in whom capecitabine treatment rapidly failed. Clin Breast Cancer 2014;14:26-30. 324. Schilsky RL, Levin J, West WH, Wong A, Colwell B, Thirlwell MP, et al. Randomized, open- label, phase III study of a 28-day oral regimen of eniluracil plus fluorouracil versus intravenous fluorouracil plus leucovorin as first-line therapy in patients with metastatic/advanced colorectal cancer. J Clin Oncol 2002;20:1519-26. 325. Van Cutsem E, Twelves C, Cassidy J, Allman D, Bajetta E, Boyer M, et al. Oral capecitabine compared with intravenous fluorouracil plus leucovorin in patients with metastatic colorectal cancer: results of a large phase III study. J Clin Oncol 2001;19:4097-106. 326. Wilson PM, Danenberg PV, Johnston PG, Lenz HJ, Ladner RD. Standing the test of time: targeting thymidylate biosynthesis in cancer therapy. Nat Rev Clin Oncol 2014;11:282-98. 327. Toscano F, Parmentier B, Fajoui ZE, Estornes Y, Chayvialle JA, Saurin JC, et al. p53 dependent and independent sensitivity to oxaliplatin of colon cancer cells. Biochem Pharmacol 2007;74:392-406. 328. Zaanan A, Cuilliere-Dartigues P, Guilloux A, Parc Y, Louvet C, de Gramont A, et al. Impact of p53 expression and microsatellite instability on stage III colon cancer disease-free

147 survival in patients treated by 5-fluorouracil and leucovorin with or without oxaliplatin. Ann Oncol 2010;21:772-80. 329. Finnberg N, Wambi C, Ware JH, Kennedy AR, El-Deiry WS. Gamma-radiation (GR) triggers a unique gene expression profile associated with cell death compared to proton radiation (PR) in mice in-vivo. Cancer biology & therapy 2008;7:2023-33. 330. Wilson PM, Labonte MJ, Russell J, Louie S, Ghobrial AA, Ladner RD. A novel fluorescence- based assay for the rapid detection and quantification of cellular deoxyribonucleoside triphosphates. Nucleic acids research 2011;39:e112. 331. Wei CL, Wu Q, Vega VB, Chiu KP, Ng P, Zhang T, et al. A global map of p53 transcription- factor binding sites in the human genome. Cell 2006;124:207-19. 332. Dumont P, Leu JI, Della Pietra AC, 3rd, George DL, Murphy M. The codon 72 polymorphic variants of p53 have markedly different apoptotic potential. Nat Genet 2003;33:357-65. 333. Kuribayashi K, Finnberg N, Jeffers JR, Zambetti GP, El-Deiry WS. The relative contribution of pro-apoptotic p53-target genes in the triggering of apoptosis following DNA damage in-vitro and in-vivo. Cell cycle (Georgetown, Tex) 2011;10:2380-9. 334. Johnson MR, Wang K, Smith JB, Heslin MJ, Diasio RB. Quantitation of dihydropyrimidine dehydrogenase expression by real-time reverse transcription polymerase chain reaction. Anal Biochem 2000;278:175-84. 335. Johnston SJ, Ridge SA, Cassidy J, McLeod HL. Regulation of dihydropyrimidine dehydrogenase in colorectal cancer. Clin Cancer Res 1999;5:2566-70. 336. Takabayashi A, Iwata S, Kawai Y, Kanai M, Taki Y, Takechi T, et al. Dihydropyrimidine dehydrogenase activity and mRNA expression in advanced gastric cancer analyzed in relation to effectiveness of preoperative 5-fluorouracil-based chemotherapy. Int J Oncol 2000;17:889-95. 337. Fujiwara H, Terashima M, Irinoda T, Takagane A, Abe K, Kashiwaba M, et al. Quantitative measurement of thymidylate synthase and dihydropyrimidine dehydrogenase mRNA level in gastric cancer by real-time RT-PCR. Jpn J Cancer Res 2002;93:1342-50. 338. Postic C, Shiota M, Niswender KD, Jetton TL, Chen Y, Moates JM, et al. Dual roles for glucokinase in glucose homeostasis as determined by liver and pancreatic beta cell- specific gene knock-outs using Cre recombinase. The Journal of biological chemistry 1999;274:305-15. 339. Heggie GD, Sommadossi JP, Cross DS, Huster WJ, Diasio RB. Clinical pharmacokinetics of 5-fluorouracil and its metabolites in plasma, urine, and bile. Cancer research 1987;47:2203-6. 340. Peters GJ, Lankelma J, Kok RM, Noordhuis P, van Groeningen CJ, van der Wilt CL, et al. Prolonged retention of high concentrations of 5-fluorouracil in human and murine tumors as compared with plasma. Cancer Chemother Pharmacol 1993;31:269-76. 341. Finnberg N, Klein-Szanto AJ, El-Deiry WS. TRAIL-R deficiency in mice promotes susceptibility to chronic inflammation and tumorigenesis. The Journal of clinical investigation 2008;118:111-23. 342. Peters GJ, van Triest B, Backus HH, Kuiper CM, van der Wilt CL, Pinedo HM. Molecular downstream events and induction of thymidylate synthase in mutant and wild-type p53 colon cancer cell lines after treatment with 5-fluorouracil and the thymidylate synthase inhibitor raltitrexed. Eur J Cancer 2000;36:916-24.

148 343. Finnberg N, Gruber JJ, Fei P, Rudolph D, Bric A, Kim SH, et al. DR5 knockout mice are compromised in radiation-induced apoptosis. Molecular and cellular biology 2005;25:2000-13. 344. Brunn ND, Dibrov SM, Kao MB, Ghassemian M, Hermann T. Analysis of mRNA recognition by human thymidylate synthase. Biosci Rep 2014;34:e00168. 345. Lin X, Mizunuma N, Chen T, Copur SM, Maley GF, Liu J, et al. In vitro selection of an RNA sequence that interacts with high affinity with thymidylate synthase. Nucleic acids research 2000;28:4266-74. 346. Jorissen RN, Gibbs P, Christie M, Prakash S, Lipton L, Desai J, et al. Metastasis-Associated Gene Expression Changes Predict Poor Outcomes in Patients with Dukes Stage B and C Colorectal Cancer. Clin Cancer Res 2009;15:7642-51. 347. Pritchard DM, Potten CS, Hickman JA. The relationships between p53-dependent apoptosis, inhibition of proliferation, and 5-fluorouracil-induced histopathology in murine intestinal epithelia. Cancer research 1998;58:5453-65. 348. Longley DB, Harkin DP, Johnston PG. 5-fluorouracil: mechanisms of action and clinical strategies. Nat Rev Cancer 2003;3:330-8. 349. van Kuilenburg AB. Dihydropyrimidine dehydrogenase and the efficacy and toxicity of 5- fluorouracil. Eur J Cancer 2004;40:939-50. 350. Schulte-Eistrup S, Ashizawa S, Nonaka K, Ichikawa S, Motomura T, Murai N, et al. Combination of therapeutic apheresis and therapeutic ventricular assistance for end- stage heart failure patients. Ther Apher 2002;6:247-54. 351. Kornmann M, Schwabe W, Sander S, Kron M, Strater J, Polat S, et al. Thymidylate synthase and dihydropyrimidine dehydrogenase mRNA expression levels: predictors for survival in colorectal cancer patients receiving adjuvant 5-fluorouracil. Clin Cancer Res 2003;9:4116- 24. 352. Panczyk M. Pharmacogenetics research on chemotherapy resistance in colorectal cancer over the last 20 years. World J Gastroenterol 2014;20:9775-827. 353. Kadota K, Huang CL, Liu D, Yokomise H, Haba R, Wada H. Combined therapy with a thymidylate synthase-inhibiting vector and S-1 has effective antitumor activity against 5- FU-resistant tumors. Int J Oncol 2011;38:355-63. 354. Noro R, Miyanaga A, Minegishi Y, Okano T, Seike M, Soeno C, et al. Histone deacetylase inhibitor enhances sensitivity of non-small-cell lung cancer cells to 5-FU/S-1 via down- regulation of thymidylate synthase expression and up-regulation of p21(waf1/cip1) expression. Cancer Sci 2010;101:1424-30. 355. Kline CL, Schiccitano A, Zhu J, Beachler C, Sheikh H, Harvey HA, et al. Personalized dosing via pharmacokinetic monitoring of 5-fluorouracil might reduce toxicity in early- or late- stage colorectal cancer patients treated with infusional 5-fluorouracil-based chemotherapy regimens. Clin Colorectal Cancer 2014;13:119-26. 356. Ahn J, Prives C. Checkpoint kinase 2 (Chk2) monomers or dimers phosphorylate Cdc25C after DNA damage regardless of threonine 68 phosphorylation. The Journal of biological chemistry 2002;277:48418-26. 357. Wu X, Chen J. Autophosphorylation of checkpoint kinase 2 at serine 516 is required for radiation-induced apoptosis. The Journal of biological chemistry 2003;278:36163-8.

149 358. Blagosklonny MV. Sequential activation and inactivation of G2 checkpoints for selective killing of p53-deficient cells by microtubule-active drugs. Oncogene 2002;21:6249-54. 359. Blagosklonny MV, Pardee AB. Exploiting cancer cell cycling for selective protection of normal cells. Cancer research 2001;61:4301-5. 360. Sato K, Ohta T, Venkitaraman AR. A mitotic role for the DNA damage-responsive CHK2 kinase. Nature cell biology 2010;12:424-5. 361. Arienti KL, Brunmark A, Axe FU, McClure K, Lee A, Blevitt J, et al. Checkpoint kinase inhibitors: SAR and radioprotective properties of a series of 2-arylbenzimidazoles. Journal of medicinal chemistry 2005;48:1873-85. 362. Jobson AG, Lountos GT, Lorenzi PL, Llamas J, Connelly J, Cerna D, et al. Cellular inhibition of checkpoint kinase 2 (Chk2) and potentiation of camptothecins and radiation by the novel Chk2 inhibitor PV1019 [7-nitro-1H-indole-2-carboxylic acid {4-[1- (guanidinohydrazone)-ethyl]-phenyl}-amide]. The Journal of pharmacology and experimental therapeutics 2009;331:816-26. 363. Nguyen TN, Saleem RS, Luderer MJ, Hovde S, Henry RW, Tepe JJ. Radioprotection by hymenialdisine-derived checkpoint kinase 2 inhibitors. ACS chemical biology 2012;7:172- 84. 364. Garrett MD, Collins I. Anticancer therapy with checkpoint inhibitors: what, where and when? Trends in pharmacological sciences 2011;32:308-16. 365. Takai H, Naka K, Okada Y, Watanabe M, Harada N, Saito S, et al. Chk2-deficient mice exhibit radioresistance and defective p53-mediated transcription. The EMBO journal 2002;21:5195-205. 366. Oliver AW, Paul A, Boxall KJ, Barrie SE, Aherne GW, Garrett MD, et al. Trans-activation of the DNA-damage signalling protein kinase Chk2 by T-loop exchange. The EMBO journal 2006;25:3179-90. 367. Vanquelef E, Simon S, Marquant G, Garcia E, Klimerak G, Delepine JC, et al. R.E.D. Server: a web service for deriving RESP and ESP charges and building force field libraries for new molecules and molecular fragments. Nucleic acids research 2011;39:W511-7. 368. Friesner RA, Banks JL, Murphy RB, Halgren TA, Klicic JJ, Mainz DT, et al. Glide: a new approach for rapid, accurate docking and scoring. 1. Method and assessment of docking accuracy. Journal of medicinal chemistry 2004;47:1739-49. 369. Navaraj A, Finnberg N, Dicker DT, Yang W, Matthew EM, El-Deiry WS. Reduced cell death, invasive and angiogenic features conferred by BRCA1-deficiency in mammary epithelial cells transformed with H-Ras. Cancer biology & therapy 2009;8:2417-44. 370. Cheng H, Hong B, Zhou L, Allen JE, Tai G, Humphreys R, et al. Mitomycin C potentiates TRAIL-induced apoptosis through p53-independent upregulation of death receptors: evidence for the role of c-Jun N-terminal kinase activation. Cell cycle (Georgetown, Tex) 2012;11:3312-23. 371. Debbas M, White E. Wild-type p53 mediates apoptosis by E1A, which is inhibited by E1B. Genes & development 1993;7:546-54. 372. Jack MT, Woo RA, Hirao A, Cheung A, Mak TW, Lee PW. Chk2 is dispensable for p53- mediated G1 arrest but is required for a latent p53-mediated apoptotic response. Proceedings of the National Academy of Sciences of the United States of America 2002;99:9825-9.

150 373. Kim KH, Yoo HY, Joo KM, Jung Y, Jin J, Kim Y, et al. Time-course analysis of DNA damage response-related genes after in-vitro radiation in H460 and H1229 lung cancer cell lines. Experimental & molecular medicine 2011;43:419-26. 374. Hayat MJ, Howlader N, Reichman ME, Edwards BK. Cancer statistics, trends, and multiple primary cancer analyses from the Surveillance, Epidemiology, and End Results (SEER) Program. Oncologist 2007;12:20-37. 375. Freireich EJ, Gehan EA, Rall DP, Schmidt LH, Skipper HE. Quantitative comparison of toxicity of anticancer agents in mouse, rat, hamster, dog, monkey, and man. Cancer chemotherapy reports Part 1 1966;50:219-44. 376. Zhang JH, Chung TD, Oldenburg KR. A Simple Statistical Parameter for Use in Evaluation and Validation of High Throughput Screening Assays. Journal of biomolecular screening 1999;4:67-73. 377. Jobson AG, Cardellina JH, 2nd, Scudiero D, Kondapaka S, Zhang H, Kim H, et al. Identification of a Bis-guanylhydrazone [4,4'-Diacetyldiphenylurea-bis(guanylhydrazone); NSC 109555] as a novel chemotype for inhibition of Chk2 kinase. Molecular pharmacology 2007;72:876-84. 378. Zoppoli G, Solier S, Reinhold WC, Liu H, Connelly JW, Monks A, et al. CHEK2 genomic and proteomic analyses reveal genetic inactivation or endogenous activation across the 60 cell lines of the US National Cancer Institute. Oncogene 2012;31:403-18. 379. Choong ML, Yang H, Lee MA, Lane DP. Specific activation of the p53 pathway by low dose actinomycin D: a new route to p53 based cyclotherapy. Cell cycle (Georgetown, Tex) 2009;8:2810-8. 380. Apontes P, Leontieva OV, Demidenko ZN, Li F, Blagosklonny MV. Exploring long-term protection of normal human fibroblasts and epithelial cells from chemotherapy in cell culture. Oncotarget 2011;2:222-33. 381. Mao Y, Yu C, Hsieh TS, Nitiss JL, Liu AA, Wang H, et al. Mutations of human topoisomerase II alpha affecting multidrug resistance and sensitivity. Biochemistry 1999;38:10793-800. 382. Li Z, Sun B, Clewell RA, Adeleye Y, Andersen ME, Zhang Q. Dose Response Modeling of Etoposide-Induced DNA Damage Response. Toxicological sciences : an official journal of the Society of Toxicology 2013. 383. Conti C, Seiler JA, Pommier Y. The mammalian DNA replication elongation checkpoint: implication of Chk1 and relationship with origin firing as determined by single DNA molecule and single cell analyses. Cell cycle (Georgetown, Tex) 2007;6:2760-7. 384. Bartek J, Lukas J. Chk1 and Chk2 kinases in checkpoint control and cancer. Cancer cell 2003;3:421-9. 385. Nikolaeva-Glomb L, Galabov AS. Synergistic drug combinations against the in-vitro replication of Coxsackie B1 virus. Antiviral research 2004;62:9-19. 386. Galabov AS, Velichkova EH. Antiviral activity of N-phenyl-N'-aryl- or alkylthiourea derivatives in Coxsackie virus infections in mice. Antimicrobial agents and chemotherapy 1974;5:1-8. 387. Huang da W, Sherman BT, Lempicki RA. Systematic and integrative analysis of large gene lists using DAVID bioinformatics resources. Nat Protoc 2009;4:44-57.

151 388. Huang DW, Sherman BT, Tan Q, Kir J, Liu D, Bryant D, et al. DAVID Bioinformatics Resources: expanded annotation database and novel algorithms to better extract biology from large gene lists. Nucleic acids research 2007;35:W169-75. 389. Iwanicki MP, Chen HY, Iavarone C, Zervantonakis IK, Muranen T, Novak M, et al. Mutant p53 regulates ovarian cancer transformed phenotypes through autocrine matrix deposition. JCI Insight 2016;1. 390. Ciaparrone M, Quirino M, Schinzari G, Zannoni G, Corsi DC, Vecchio FM, et al. Predictive role of thymidylate synthase, dihydropyrimidine dehydrogenase and thymidine phosphorylase expression in colorectal cancer patients receiving adjuvant 5-fluorouracil. Oncology 2006;70:366-77. 391. Goto T, Shinmura K, Yokomizo K, Sakuraba K, Kitamura Y, Shirahata A, et al. Expression levels of thymidylate synthase, dihydropyrimidine dehydrogenase, and thymidine phosphorylase in patients with colorectal cancer. Anticancer Res 2012;32:1757-62. 392. Liu J, Zhang C, Hu W, Feng Z. Tumor suppressor p53 and its mutants in cancer metabolism. Cancer Lett 2015;356:197-203. 393. Kollareddy M, Dimitrova E, Vallabhaneni KC, Chan A, Le T, Chauhan KM, et al. Regulation of nucleotide metabolism by mutant p53 contributes to its gain-of-function activities. Nat Commun 2015;6:7389. 394. Burnstock G. Purine and pyrimidine receptors. Cell Mol Life Sci 2007;64:1471-83. 395. Nam H, Campodonico M, Bordbar A, Hyduke DR, Kim S, Zielinski DC, et al. A systems approach to predict oncometabolites via context-specific genome-scale metabolic networks. PLoS Comput Biol 2014;10:e1003837. 396. Muhale FA, Wetmore BA, Thomas RS, McLeod HL. Systems pharmacology assessment of the 5-fluorouracil pathway. Pharmacogenomics 2011;12:341-50. 397. Davis ST, Joyner SS, Baccanari DP, Spector T. 5-Ethynyluracil (776C85): protection from 5- fluorouracil-induced neurotoxicity in dogs. Biochem Pharmacol 1994;48:233-6. 398. Cao C, Zhang X, Kuang M, Gu D, He M, Chen J, et al. Survival benefit from S-1 as compared to Fluorouracil in Asian patients with advanced gastrointestinal cancer: a meta-analysis. Cancer Sci 2014;105:1008-14. 399. Spector T, Cao S. A possible cause and remedy for the clinical failure of 5-fluorouracil plus eniluracil. Clin Colorectal Cancer 2010;9:52-4. 400. Wang WJ, Wu SP, Liu JB, Shi YS, Huang X, Zhang QB, et al. MYC regulation of CHK1 and CHK2 promotes radioresistance in a stem cell-like population of nasopharyngeal carcinoma cells. Cancer research 2013;73:1219-31. 401. Panvichian R, Tantiwetrueangdet A, Angkathunyakul N, Leelaudomlipi S. TOP2A amplification and overexpression in hepatocellular carcinoma tissues. Biomed Res Int 2015;2015:381602. 402. Carloni V, Lulli M, Madiai S, Mello T, Hall A, Luong TV, et al. CHK2 overexpression and mislocalisation within mitotic structures enhances chromosomal instability and hepatocellular carcinoma progression. Gut 2017. 403. Leibowitz BJ, Qiu W, Liu H, Cheng T, Zhang L, Yu J. Uncoupling p53 functions in radiation- induced intestinal damage via PUMA and p21. Mol Cancer Res 2011;9:616-25. 404. Pant V, Xiong S, Chau G, Tsai K, Shetty G, Lozano G. Distinct downstream targets manifest p53-dependent pathologies in mice. Oncogene 2016;35:5713-21.

152 405. van Os R, Kamminga LM, Ausema A, Bystrykh LV, Draijer DP, van Pelt K, et al. A Limited role for p21Cip1/Waf1 in maintaining normal hematopoietic stem cell functioning. Stem Cells 2007;25:836-43.

153 Appendix: Letters of Permission

My dissertation includes published work from published research articles from on which I am first author. The corresponding articles are listed below along with the permission from the respective sources according to the journal policies

Article 1. Part of chapter 2.

P53 represses pyrimidine catabolic gene dihydropyrimidine dehydrogenase (DPYD) expression in response to thymidylate synthase (TS) targeting. Gokare P, Finnberg NK, Abbosh PH, Dai J, Murphy ME, El-Deiry WS. Sci Rep. 2017 Aug 29;7(1):9711.

According to Journal policy.

Pre-Publicity

“Contributions being prepared for or submitted to Scientific Reports can be posted on recognized preprint servers (such as ArXiv), and on collaborative websites such as wikis or the author's blog. The website and URL must be identified in the cover letter accompanying submission of the paper, and the content of the paper must not be advertised to the media by virtue of being on the website or preprint server. Material in a contribution submitted to Scientific Reports may also have been published as part of a PhD or other academic thesis.”

Article 2. Part of Chapter 3.

Targeting of Chk2 as a countermeasure to dose-limiting toxicity triggered by topoisomerase-II (TOP2) poisons. Gokare P, Navaraj A, Zhang S, Motoyama N, Sung SS, Finnberg NK.Oncotarget. 2016 May 17;7(20):29520-30

According to Journal policy

Oncotarget applies the Creative Commons Attribution 3.0 License (CC BY 3.0) to all works we publish (read the human-readable summary or the full license legal code). Under the CC BY, authors retain ownership of the copyright for their article, but authors allow anyone to download, reuse, reprint, modify, distribute, and/or copy articles in Oncotarget, so long as the original authors and source are cited.

No permission is required from the authors or the publishers. In most cases, appropriate attribution can be provided by simply citing the original article. If the item you plan to reuse is not part of a published article (e.g., a featured issue image), then please indicate the originator of the work, and the volume, issue, and date of the journal in which the item appeared. For any reuse or redistribution of a work, you must also make clear the license terms under which the work was published. This broad license was developed to facilitate open access to, and free use of, original works of all types. Applying this standard license to your own work will ensure your right to make your work freely and openly available.

155 VITA

EDUCATION

2017 Doctor of Philosophy (Ph.D.) Penn State University 2010 Master of Science (MS)- University of Houston, Clear Lake, 2008- Bachelor of Engineering (B.E) – Visveshwariah Technological University, Mysore India

. 06/16-09/17- FUNDED GRANTS AS PRINCIPAL INVESTIGATOR- . . Recipient of the Peer Reviewed Cancer Research Program Horizon award (PRCRP) from Department of Defense office of Congressionally Directed Medical Research Program (CDRMP). . PUBLICATIONS (2 out of 11) . 1) Gokare P, Finnberg NK, Abbosh PH, Dai J, Murphy ME, El-Deiry WS; “p53 inhibits the expression of 5-FU catabolism gene Dihydropyrimidine Dehydrogenase following TS inhibition” Sci Rep. 2017 Aug 29;7(1):9711. . 2) Gokare P, Navaraj A, Zhang S, Motoyama N, Sung SS, Finnberg NK. “Chk2-targeting prevents toxicity from TOP2-poisons”, Oncotarget. 2016;7(20).

AWARDS/ RECOGNITIONS . ➢ AbbVie Scholar-in-Training award from the American Association of Cancer Research (AACR) (April 2016). . ➢ “Scholar-in-Training award” from the American Association of Cancer Research (AACR) (June 2015). . ➢ “Invited Speaker” at AACR conference on Metabolism and Cancer (June 2015). . ➢ Second place in National conference on Health Care Biotechnology held at J.N. Tata Auditorium, Indian Institute of Science Complex, Bangalore during 20-21st April, 2007, for the poster entitled “Nanotechnology-A multi-pronged tool for cancer therapy”. . ➢ First place in the Quiz event “Gene hunt” during February 2008 conducted by Visveswaraiah Technological University (VTU), Belgaum, Karnataka, India. This event was open to, over 15 colleges, under the VTU system. . ➢ Second Place in the “Bio-Quiz” conducted in July 2009 conducted by Visveswaraiah Technological University (VTU), Belgaum Karnataka, India.

. MEMBERSHIP IN PROFESSIONAL SOCIETIES . ➢ American Association for Cancer research (AACR). . ➢ American Society for Biochemistry and Molecular Biology (ASBMB). . ➢ Association of Microbiologists of India (AMI). . ➢ Society of Biological Chemists (SBC), India. . ➢ Association of Food Scientists Technologists of India, AFST (I). . ➢ Biotech club “Mutagens” UHCL, Texas. . . PEER REVIEW BOARD/JOURNAL REVIEW . Molecular and Cellular Oncology (MCO) journal published by labland biosciences. . Cancer chemotherapy and Biology (CBT) Taylor and Francis publication group. . Journal of Clinical Investigation (JCI). . Oncotarget journal. . Cancer Research and Cell death & differentiation.