pathway determinants of 5-fluorouracil sensitivity · pathway determinants of 5-fluorouracil...

PATHWAY DETERMINANTS OF 5-FLUOROURACIL SENSITIVITY

TAN WEN LEE B.Sc. (Pharm)(Hons.), NUS

A THESIS SUBMITTED

FOR THE DEGREE OF MASTER OF SCIENCE

DEPARTMENT OF PATHOLOGY

NATIONAL UNIVERSITY OF SINGAPORE

2009

1

ACKNOWLEDGEMENTS First of all, I wish to thank Prof. Teh Ming for giving me an opportunity to study in the Department of Pathology. I would like to express my heartfelt gratitude to my supervisor – A/Prof Richie Soong for giving me this opportunity to undertake this project. I am indebted to him for his invaluable understanding, guidance and support throughout this course. I would also like to thank Prof. Limsoon Wong and Mr. Difeng Dong from School of Computing, National University of Singapore, and Ms. Marie Loh from Oncology Research Institute (Singapore) for their invaluable advice and inputs for the statistical analyses; Dr. Ross Soo from Department of Haematology Oncology, National University Hospital (Singapore) for his expert opinions in clinical research; Dr. Iduna Fichtner from Max Delbrück Centre for Molecular Medicine, Berlin-Buch, (Germany) for the provision of tumor xenograft materials; Dr. Robert Hewitt, Dr. Rajeev Singh, Dr. Chon Boon Eng, Ms. Almeda and Ms Wentong Yang from NUH-NUS Tissue Repository (Singapore) for the provision of human tissue samples and clinical information and, Ms. Cecilia Lee from Institute of Molecular and Cell Biology (Singapore) for the provision of human cancer cell lines. Lastly, I would like to thank my family members, my hubby, Eng Lee and all my lab colleagues from Translational Interface, Cancer Science Institute (Singapore) for their companionships and their help in one way or another.

2

CONTENTS

List of Tables ..................................................................................................................... 3 List of Figures .................................................................................................................... 4 ABSTRACT ....................................................................................................................... 5 CHAPTER 1: LITERATURE REVIEW ....................................................................... 6

1.1 Colorectal Cancer ............................................................................................... 6 1.2 5-fluorouracil ...................................................................................................... 7

1.2.1 Transport ....................................................................................................... 9 1.2.2 Metabolism ................................................................................................. 10 1.2.3 Mechanism of Action .................................................................................. 14

1.3 Fluorouracil-based therapies ............................................................................. 15 1. 4 Molecular determinants of 5FU sensitivity ...................................................... 18

1. 4.1 Candidate gene approach ............................................................................ 18 1.4.2 Genome-wide approach .............................................................................. 24

1.5 Study approaches .............................................................................................. 30 1.5.1 Quantification of mRNA expression using real-time PCR ......................... 31 1.5.2 Regulation patterns of genes in response to 5FU treatment ....................... 35 1.5.3 RNA interference (RNAi) ........................................................................... 35

1.6 Scope of Study .................................................................................................. 38 CHAPTER 2: MATERIALS AND METHODS .......................................................... 41

2.1 Cell lines ........................................................................................................... 41 2.2 Cell sensitivity analysis .................................................................................... 41 2.3 Xenografts ......................................................................................................... 42 2.4 Primary Tumors ................................................................................................ 42 2.5 Gene expression analysis .................................................................................. 42 2.6 Small interfering RNA (siRNA) transfections .................................................. 49 2.7 Statistical analysis ............................................................................................. 49

CHAPTER 3: RESULTS ............................................................................................... 52 3.1 Correlation between gene expression and 5FU sensitivity ............................... 52 3.2 Identification of baseline gene expression associated with 5FU sensitivity ..... 60 3.3 Identification of post-treatment gene expression changes associated with 5FU sensitivity ...................................................................................................................... 65 3.4 Modulation of 5FU sensitivity by CTPS2 siRNA ............................................ 69

CHAPTER 4: DISCUSSION ......................................................................................... 73 CONCLUSION ............................................................................................................... 79 REFERENCES ................................................................................................................ 80 APPENDIX ...................................................................................................................... 94 PUBLICATIONS ............................................................................................................ 95

3

List of Tables

Table 1: Overview of studies on gene/protein expression levels in relation to 5FU

chemosensitivity.

Table 2: Summary of genome-wide and proteomic studies to identify determinants for

5FU-based chemosensitivities.

Table 3: List of 96 genes selected for LDA analysis.

Table 4: Characterization of 5FU sensitivity: (A) IC50 values of 21 colorectal cancer cell

lines; (B) % of tumor shrinkage in 14 xenografts treated by 5FU; (C) Recurrence data of

25 stage II or III colorectal cancer patients treated with standard 5FU and leucovorin

therapy.

Table 5: Associations (χ2 test) between 5FU sensitivity and clinicopathological

parameters

Table 6: Candidate genes with differential gene expression (p<0.05) identified by

unsupervised clustering analysis for xenografts and primary tumors.

Table 7: Differential expression of genes in xenograft and patient series (p<0.05).

4

List of Figures

Figure 1: Chemical structure of uracil and 5-fluorouracil (5FU).

Figure 2: 5FU metabolism.

Figure 3: Pathways of fluoropyrimdine and folate metabolism and downstream events.

Figure 4: Activation of DNA repair mechanisms and stress response in relation to 5FU

effects.

Figure 5: Diagram illustrating the concepts for the TaqMan probe assay.

Figure 6: TaqMan® LDA.

Figure 7: The RNA Interference pathway.

Figure 8: Genes analyzed in this study for their roles in transport and metabolism of 5FU

and its downstream events.

Figure 9: Unsupervised hierarchical clustering of samples based on the expression of 96

genes for (A) cell lines, (B) xenografts and (C) primary tumors.

Figure 10: Difference in expression levels of CTPS2 and NME4 between clusters in (A)

xenografts and (B) patients.

Figure 11: Patterns of gene expression at 0, 4, 8, 12, 24 and 48 hours after 10 μM 5FU

treatment for (A) RKO and (B) SW837.

Figure 12: CTPS2 and GAPDH mRNA expression levels in SW620 cells transfected with

(A) CTPS2 siRNA and (B) GAPDH siRNA 48 and 120 hours after transfection. Growth

curves of cells either treated or untreated with 10 μM 5FU after 48 hours of transfection

of (C) CTPS2 siRNA, (D) GAPDH siRNA and (E) negative control siRNA.

LITERATURE REVIEW

5

ABSTRACT

Purpose: Large inter-patient variability in patient response to the chemotherapeutic

agent, 5-Fluorouracil (5FU), makes markers for predicting 5FU response highly

desirable. We describe here a pathway-based gene expression analysis of the relative

roles of all known components of 5FU transport, metabolism, its downstream effects and

co-factor (folate) metabolism in determining 5FU sensitivity in cell lines, xenografts and

tumours. Experimental design: The expression of 96 rationally-selected genes was measured by

low-density array real-time PCR in 21 colorectal cancer cell lines, 14 xenografts and 25

tumors. Samples were grouped by unsupervised hierarchical clustering and the groups

tested for associations with respective sensitivity, response and patient survival data.

Patterns of expression changes of the 96 genes in 5FU sensitive and resistant cells

following a time course of 5FU treatment were also clustered to identify the most

defining expression changes, and their component genes. siRNA knockdown was

performed to verify the influence of the top candidate on 5FU sensitivity in-vitro.

Results: 5FU sensitivity associated with clustered groups for xenografts and tumours,

but not cell lines. CTPS2 and NME4 were the genes common to the significantly

differentially regulated genes in xenografts and tumors. CTPS2 was also in the group of

genes with the most distinct expression change following 5FU treatment in sensitive and

resistant cell lines. siRNA knockdown of CTPS2 reduced sensitivity to 5FU.

Conclusion: Pathway-based analysis in multiple cancer models identifies CTPS2

expression as a major determinant of 5FU sensitivity.

LITERATURE REVIEW

6

CHAPTER 1: LITERATURE REVIEW

1.1 Colorectal Cancer

Colorectal cancer (CRC) is one of the leading causes of cancer deaths worldwide

(Greenlee et al., 2000). In Singapore, CRC is the most prevalent cancer in both men and

women combined and the incidence is increasing very rapidly with about 1000 new cases

diagnosed annually (Singapore Cancer Registry Report No. 6). The probability of

developing CRC increases with age, with 1 in 124 males and 1 in 149 females aged

between 40 to 59 years old developing CRC, increases to 1 in 29 males and 1 in 33

females for age between 60 to 79 (Greenlee et al., 2000).

Surgical resection remains the primary treatment for patients with CRC. Resected

samples are then sent for pathological evaluation which provides prognostic information

based on tumor staging and facilitates treatment decision. Chemotherapy and/or

radiotherapy is often given in adjuvant settings, and less commonly as neoadjuvant

treatment. For patients whose surgery is deemed non-curative, for example, in metastatic

colorectal cancer, chemotherapy or radiotherapy may be used as primary treatment

(palliative). In colon cancer, whereby recurrence to distant sites like liver, bone and lungs

is not uncommon, systemic chemotherapy given in adjuvant settings is more preferred to

radiotherapy (Midgley et al., 1999). In contrast, about 50% of recurrences of rectal

cancer localize in the pelvic region and hence, local radiotherapy is a preferred treatment

for relapse prevention.

LITERATURE REVIEW

7

Forty to 60% of the CRC patient population is diagnosed with stage II or stage III disease

(Aranha et al., 2007). After curative surgery, stage III patients experience 50-60% chance

of developing disease recurrence. Published clinical trials have proved the beneficial

effects of adjuvant chemotherapy in improving curative rates in resected stage III cancer

(Wolmark et al., 1993; Moertel et al., 1995; Bleiberg et al., 2000). As such, the use of

adjuvant chemotherapy for stage III cancer has since become the standard care. In

contrast, patients with stage II colorectal cancer, the use of adjuvant therapy remains

controversial and is not routinely recommended (Benson et al., 2004). However, the

recurrence rate for stage II patients is about 20%, which is considerably high. This poses

a dilemma for clinicians in deciding the appropriateness of adjuvant chemotherapy in

stage II disease, so not to expose patients to unnecessary side effects while ensuring

clinical benefits.

1.2 5-fluorouracil

In 1957, fluoropyrimidines were rationally designed based on the observation that rat

hepatomas utilize radiolabelled uracil more rapidly than normal tissues, indicating that

uracil metabolism was a potential target for chemotherapy (Heidelberger et al., 1957). 5-

fluorouracil (5FU), then synthesized with the presence of a fluorine atom on the fifth

carbon of naturally occurring uracil is readily taken up and metabolized by cells, thus

interfering with uracil metabolism (Figure 1). The more rapid uracil metabolism in tumor

cells provides 5FU the advantage of being tumor selective as the rate of 5FU uptake goes

hand-in-hand with uracil. With well demonstrated cytotoxicity, 5FU has since been

widely used in the treatment of a number of cancer types, including colorectal, breast,

LITERATURE REVIEW

8

head and neck cancer (Longley et al., 2003). Though being tumor selective, 5FU also

causes normal tissues toxicities like nausea, vomiting, diarrhea, myelosuppression,

mucositis, hand-foot disease, and, more rarely cardiac and neurotoxicities (Grem et al.,

2000).

Figure 1: Chemical structure of uracil and 5-fluorouracil (5FU). The hydrogen atom at the carbon-5 position of the pyrimidine ring is replaced by a fluorine atom in 5FU.

LITERATURE REVIEW

9

1.2.1 Transport

As a synthetic nucleoside, uptake of 5FU into cells has been shown to occur via the same

non-concentrative, saturable mechanism as uracil (Wohlhueter et al., 1980), via an active,

concentrative mechanism (Yamamoto et al., 1981), or via nonfacilitated diffusion (Kessel

et al., 1967). It was later reported that 5FU enters human erythrocytes via same carriers

that transport purine nucleobases and nonfacilitated diffusion (Domin et al., 1993).

It is now recognized that uptake transporters responsible for translocation of nucleoside

and nucleobases are members of the equilibrative nucleoside transporters (ENT,

SLC29A) family and concentrative nucleoside transporter (CNT, SLC28A) family

(Pastor-Aglada et al., 1998). The ENTs (comprises 4 isoforms, ENT1-4) transport

hydrophilic nucleosides and nucleoside analogs by bi-directional facilitated diffusion

(Baldwin et al., 2004). They accept both purine and pyrimidine nucleosides but differ in

their sensitivity to inhibition by nitrobenzylthioinosine (NBMPR). Human ENT1

(hENT1, an equilibrative sensitive, es-type transporter) is inhibited by nanomolar

concentration of NBMPR, whereas hENT2 (an equilibrative insensitive, ei-type

transporter) is unaffected by low concentrations of NBPMR. Besides purine and

pyrimidine analogs, ei-type transporter also transports nucleobases (Ritzel et al., 2001).

On the other hand, CNT transporters are Na+-dependent and show different substrate

specificity and pharmacological properties (Pastor-Aglada et al., 1998). A number of

isoforms (CNT1-3) are known to exist: hCNT1 (a concentrative NBMPR-insensitive

thymidine selective or cit-type transporter) is pyrimidine-preferring; hCNT2 (a

LITERATURE REVIEW

10

concentrative insensitive and formycin B-selective or cif-type transporter) is purine-

preferring; hCNT3 (a concentrative NBMPR-insensitive or cib-type tranporter) is broadly

selective, transporting both purine and pyrimidine nucleosides and also nucleobases

(Pastor-Aglada et al., 1998, Ritzel et al., 2001). In another study investigating the role of

hCNT1 in cytotoxicity of 5’-deoxy-5-fluorouridine (5’-DFUR), a 5FU prodrug, it was

found that 5’-DFUR is an hCNT1 substrate, but 5FU is not (Mata et al., 2001). Efflux

transporters for 5FU cytotoxic intracellular metabolite has first been identified by Guo et

al. in an in-vitro transport assay (Guo et al., 2003). Like drug uptake mechanism, efflux

transport could have important implication in chemotherapeutic cytotoxicity as efflux

mechanism also determines drug availability in tissues.

1.2.2 Metabolism

Biotransformation of 5FU is important for its clinical effectiveness. The balance between

anabolism and catabolism determines the availability of 5FU in tissues (Heggie et al.,

1987). As a uracil analogue, 5FU undergoes the de novo anabolic and catabolic pathways

used by uracil. The anabolism of 5FU to nucleotides occurs through several pathways

(Figure 2 and 3):

(1) a reversible conversion of 5FU to 5-fluoro-2’-deoxyuridine (FdUrd) by thymidine

phosphorylase (TP), followed by phosphorylation to 5-fluoro-2’-deoxyuridine-5’-

monophosphate (FdUMP) by thymidine kinase (TK) (Hartmann and Heidelberger et al.,

1961);

LITERATURE REVIEW

11

(2) a transformation of 5FU to 5-fluorouridine (FUrd) by uridine phosphorylase (UP),

followed by phosphorylation to 5-fluorouridine monophosphate (FUMP) by uridine

kinase (UK) (Skold, 1958);

(3) a direct conversion of 5FU to FUMP by orotate phosphoribosyltransferase

(OPRT) which directly transfers a ribose phosphate from phosphoribosyl pyrophosphate

(PRPP) to 5FU (Reyes et al., 1969).

Another pathway for FdUMP formation is via dephosphorylation of fluorodeoxyuridine-

5’-diphosphate (FdUDP), which is converted from fluoridine-5-diphosphate (FUDP) by

ribonucleotide reductase. FdUMP and FdUDP are substrates for thymidine kinases,

resulting in the formation of fluorodeoxyuridine-5’-triphosphate (FdUTP). FUMP formed

from pathways described in (2) and (3) is sequentially phosphorylated to FUDP and

fluorouridine-5’-triphosphate (FUTP).

Dihydropyrimidine dehydrogenase (DPD) is the first and rate-limiting enzyme of 5FU

degradation, catabolising approximately 80% of 5FU to dihydrofluorouracil (DHFU)

(Diasio and Harris, 1989). Dihydropyrimidinase (DHP) subsequently converts DHFU to

5-fluoroureido-propionic acid (FUPA), followed by conversion to fluoro-β-alanine

(FBAL) by β-alanine synthase.

LITERATURE REVIEW

12

Figure 2: 5-fluorouracil (5FU) metabolism. 5FU is converted to three main active metabolites: fluorodeoxyuridine monophosphate (FdUMP), fluorodeoxyuridine triphosphate (FdUTP) and fluorouridine triphosphate (FUTP). 5FU is directly converted to FUMP by orotate phosphoribosyltransferase (OPRT) phosphoribosyl pyrophosphate (PRPP) as the co-factor, or indirectly by fluorouridine (FUR), acting through uridine phosphorylase (UP) followed by uridine kinase (UK). FUMP is sequentially phosphorylated to fluorouridine diphosphate (FUDP) and fluorodeoxyuridine-5’-diphosphate (FdUDP), which can either be phosphorylated or dephosphorylated to FdUTP and FdUMP, respectively. Alternatively, 5FU is converted via thymidine phosphorylase (TP) to flurodeoxyuridine (FUDR), which is then phosphorylated by thymidine kinase (TK) to FdUMP. Dihydropyrimidine dehydrogenase (DPD) catabolizes 5FU to dihydrofluorouracil (DHFU) (Adapted from Longley et al., 2003).

LITERATURE REVIEW

13

Figure 3: Pathways of fluoropyrimdine and folate metabolism and downstream events. The anti-tumor activity of 5FU is via inhibition of thymidylate synthase (TS), incorporation into RNA and DNA. The prominent candidate fluoropyrimidine predictive indicators are highlighted in large font and bold (Adapted from Soong and Diasio, 2005). 3UP: 3-ureidoproionoic acid; 3UP: 3-ureidopropionase; 5FU: 5-fluorouracil; BAL: b alanine; CDA: Cytidine deaminase; CE: Carboxylesterase; CK: Cytidylate kinase; CYP450: Cytochrome P450; DHF: Dihydrofolate; DHFR: Dihydrofolate reductase; DHP: Dihydropyrimidinase; DHU: Dihydrouracil; DNA: Deoxyribonucleic acid; DNAPol: DNA polymerase; DNMT: DNA methyltransferase; DPD: Dihydropyrimidine dehydrogenase; dUDP: Deoxyuridine diphosphate; dUMP: Deoxyuridine monophosphate; dUrd: Deoxyuridine; dUTP: Deoxyuridine triphosphate; dUTPPP: Deoxy-UTP-pyrophosphatase; F: fluoro; MS: Methionine synthase; NDK: Nucleotide diphosphate kinase; NDPase: Nucleoside dihposphatase; NMK: Nucleotide monophosphate kinase; NSase: Nucleosidase; NTase: 5’-nucleotidase; OPRT: Orotate phosphoribosyltransferase; RDR: Ribonucleotide diphosphate reductase; RNA Ribonucleic acid; RNAPol: RNA polymerase; SAH: S-adenosylhomocysteine; SAM: S-adenosylmethionine; THF: Tetrahydrofolate; TK: Thymidine kinase; TP: Thymidine phosphorylase; TS: Thymidylate synthase; UDP: Uridine diphosphate; UKL Uridine kinase; UMP: Uridine monophosphate; UNSases: Uridine nucleosidase; UP: Uridine phosphorylase; Urd: Uridine; UTP: Uridine triphosphate.

LITERATURE REVIEW

14

1.2.3 Mechanism of Action

The antitumor activity of 5FU is exerted by its metabolites, mainly FdUMP, FdUTP and

FUTP. At least three mechanisms of action have been identified for 5FU toxicity:

inhibition of thymidylate synthase (TS), incorporation into RNA and also DNA (Figure

3).

FdUMP inhibits TS enzyme through the formation of a stable ternary complex with 5,10-

methylene-tetrahydrofolate (CH2THF), thereby blocking the binding of natural substrate

dUMP (Sommer et al., 1974). The failure of dUMP binding prevents deoxythymidine

monophosphate (dTMP) synthesis, which is important for deoxythymidine triphosphate

(dTTP) synthesis. Disruption of dTTP synthesis leads to deoxynucleotide pool

imbalances, thus interrupting DNA synthesis.

CH2THF, the reduced-folate cofactor essential for the tight binding of FdUMP to TS is

derived from folic acid in diet. Figure 3 (Soong et al., 2005) summarizes the key enzymes

involved in folate metabolism and illustrates the role of CH2THF in TS inhibition. Folate

deficiency has been shown to incur resistance to 5FU therapy (Branda et al., 1998).

Besides disrupting the dTMP and dTTP pools, TS inhibition also leads to accumulation

of dUMP which in turn results in increased levels of dUTP. Together with FdUTP, dUTP

can be misincorporated into DNA by DNA polymerase (Longley et al., 2003). There are

two enzymes which prevent the misincoporation of FdUTP and dUTP into DNA: (1)

dUTP pyrophosphatase (dUTPase) which hydrolyses FdUTP to FdUMP and; (2) uracil-

LITERATURE REVIEW

15

DNA-glycosylase, a DNA repair enzyme which hydrolyses the fluorouracil-deoxyribose

glycosyl bond of (F)dUMP on DNA, thereby leaving an apyrimidinic site in DNA

(Mauro et al., 1993). The lesion in DNA is recognized by an apurinic/apyrimidinic

endonuclease, then resulting in single strand break, which is subsequently repaired. The

efficiency of DNA repair mechanism is further weakened by the depletion of dTTP.

Furthermore, 5FU substituted DNA is a weaker substrate for uracil-DNA-glycosylase as

compared to uracil.

The incorporation of FUTP into RNA by the action of RNA polymerase also contributes

to 5FU cytotoxicity. Various effects that occur as a result of 5FU-RNA incorporation

include decrease in net RNA synthesis, disruption of normal RNA processing and

function, inhibition of mRNA polyadenylation and alteration of secondary structure of

RNA (Longley et al., 2003).

1.3 Fluorouracil-based therapies

For the past few decades, 5FU has been administered intravenously in various forms of

different schedules, and in both adjuvant and palliative settings. Firstly used as a single

agent, 5FU has an overall response rate (RR) of 10-15% (Advanced Colorectal Cancer

Meta-Analysis Project, 1992; Thirion et al., 2004). In view of the poor clinical outcome,

many strategies have been developed to improve survival benefits and overcome drug

resistance. Few of the early biochemical modulators used to increase 5FU cytotoxicity

includes leucovorin (LV), levamisole, methotrexate (MTX) and interferon. Levamisole

has been found to be effective in treating colon cancer when combined with 5FU

LITERATURE REVIEW

16

(Moertel et al., 1990). However, the long term clinical outcome of this combination

remains unclear (Cassidy 1994; Chlebowski et al., 1994). The use of MTX and interferon

has not been translated into clinical practice today as the results from randomized trials

failed to support their clinical benefit (Thirion et al., 2001; Punt et al., 2002). LV, a

folinic acid, helps to increase the intracellular reduced folate pools, thus increasing the

stability of folate-FdUMP-TS ternary complex. The maintenance of a stable ternary

complex is essential for 5FU cytotoxicity. A meta-analysis of 9 randomized clinical trials

demonstrated a response rate in favor of 5FU/ LV over 5FU alone (23% vs 11%) but did

not improve overall survival (OS) (Advanced Colorectal Cancer Meta-Analysis Project,

1992).

As rapid degradation by DPD greatly affects 5FU bioavailability, the oral Uracil/Ftorafur

(UFT) formulation have been developed to reduce 5FU degradation by saturating DPD

activity with uracil. It has been shown that UFT/LV offered similar response rates as

5FU/LV with fewer side effects (Douillard et al., 2002). In the late 1990s, capecitabine, a

5FU oral prodrug, was designed to by-pass DPD degradation in the liver. It has been

shown that capecitabine produced higher response rate than 5FU/LV (25% vs 16%), but

time to disease progression and survival were not significantly different for both

treatment groups (Hoff et al., 2001). Fewer side effects were observed in patients treated

with capecitabine, except hand-foot syndrome which develops in 25% of these patients

(van Cutsem et al., 2004).

LITERATURE REVIEW

17

Novel combinations of 5FU and its analogs with agents that have different mechanisms

of actions like irinotecan and oxaliplatin further improves clinical outcome in colorectal

cancer. The MOSAIC trial showed that stage II or III colon cancer patients treated with

5FU, LV and oxaliplatin (FOLFOX) achieved better disease-free survival at 3 years as

compared to infusional 5FU/LV therapy (78.2% vs 72.9%), although no difference was

noted for OS (Andre et al., 2004). Randomized trials have shown that irinotecan given in

combination with 5FU/LV (FOLFIRI) to be superior to 5FU/LV regimen (Saltz et al.,

2000; Douillard et al., 2000). Saltz and colleagues reported that metastatic colorectal

cancer patients treated with irinotecan, 5FU and LV had a significantly higher RR, longer

progression-free survival (PFS) and longer OS than those treated with 5FU and LV alone

(Saltz et al., 2000). Similar observations were also reported by Douillard et al. (2000).

Other newer FDA-approved chemotherapeutic agents including monoclonal antibodies

like bevacizmab and cetuximab have been used in combination for the treatment of

metastatic colorectal cancer. Bevacizumab, targeting the vascular endothelial growth

factor, was the first targeted drug to show clinical benefit in metastatic CRC (Fakih

In a combined analysis of 3 randomized studies, metastatic colorectal cancer patients

treated with 5FU/LV/bevacizumab had improved RR, longer PFS and OS (Fairooz 2005)

compared to 5FU/LV alone. Cetuximab, which targets the epidermal growth factor

receptor provide good clinical efficacy when used in combinations with 5-fluorouracil

and/or irinotecan (Reynolds et al., 2004). In the CRYSTAL phase III trial, patients who

received FOLFIRI and cetuximab had a significantly longer progression-free survival and

higher response rate than patients who had FOLFIRI only (van Cutsem et al., 2007).

LITERATURE REVIEW

18

1. 4 Molecular determinants of 5FU sensitivity

1. 4.1 Candidate gene approach

The ability to predict 5FU response from its metabolism and cytotoxic mechanisms was

first demonstrated using candidate gene approaches. Early studies have associated

resistance to 5FU with the loss in activity of uridine kinase (Reichard et al., 1962),

uridine phosphorylase (Goldberg et al., 1966) and pyrimidine phosphoribosyltransferase

(Kasbekar et al., 1963), leading to changes in 5FU metabolite levels. Soon after TS

inhibition by 5-FdUMP was identified as an important determinant for 5FU cytotoxicity,

leading the degree of TS inhibition and TS levels to be deemed as potential indicators for

response. Many studies have demonstrated the relationship between elevated TS

expression/activity and 5FU resistance (Berger et al., 1985; Clark et al., 1987; Johnston

et al., 1992; Johnston et al., 1995). The general consensus associated low TS

expression/activity with improved response to 5FU (Johnston et al., 1995; Leichman et

al., 1995; Bathe et al., 1999; Wong et al., 2001; Fernandez-Contreras et al., 2006).

However, some reports have shown no relationship between these parameters (Findlay et

al., 1997; Allegra et al., 2003; Johnston et al., 2003) or reverse association, i.e. high TS

expression provides disease-free survival benefit (Berglund et al., 2002; Edler et al.,

2002). Polymorphism studies have revealed the presence of variable number tandemly

repeated polymorphic regulatory sequences (VNTR) in the TS enhancer region (TSER)

correlates with TS expression, with triple repeat (3R) having higher TS expression than

double repeat (2R) (Horie et al., 1995; Pullarkat et al., 2001). Clinical studies later

showed that homozygous 3R/3R patients had a lower clinical response as compared to

LITERATURE REVIEW

19

others (Pullarkat et al., 2001; Etienne et al., 2002), or did not obtain a survival benefit

from 5FU (Iacopetta et al., 2001). In the second repeat within the VNTR, the presence of

a functional G>C single nucleotide polymorphism (SNP) at bp 58 was associated with

reduced TS expression (Kawakami et al., 2003). Another TS polymorphism consisting of

a 6 bp deletion at bp 1494 in the 3’-untraslated region is associated with reduced TS

expression in colorectal tumors (Ulrich et al., 2000).

The catabolic enzyme, DPD, is another important determinant for responsiveness to 5FU

since its level and activity greatly affects 5FU availability. DPD activity is inversely

associated with 5FU sensitivity: 5FU resistance being observed in patients having high

DPD level/activity (Ishikawa et al., 1999) whilst patients with DPD deficiency suffering

from severe toxicities (Tuchman et al., 1985; Milano et al., 1999). DPD deficiency has

been linked with a DPD exon 14 deletion variant and thus identified as a potential

indicator of 5FU response (Meinsma et al., 1995). Nevertheless, there are few studies

which failed to observe such inverse relationship between DPD expression/activity and

5FU response (Jakob et al., 2005; Kinoshita et al., 2007; Yamada et al., 2008).

Other genes or proteins involved in the metabolic pathway identified as molecular

markers for 5FU response includes TP, OPRT and TK. The role of TP in determining

5FU sensitivity is unclear. Metzger et al. reported that high TP level in colorectal tumors

is associated with nonresponse to 5FU (Metzger et al., 1999) but opposite findings have

also been reported (Schwartz et al., 1995; Yasuno et al., 2005). Colorectal cancer

patients with high OPRT activity were reported to have better 5FU efficacy (Sakamoto et

LITERATURE REVIEW

20

al., 2007) and better disease-free and overall survival when treated with 5FU based

chemotherapy (Ochiai et al., 2006). However, no relationship between OPRT level and

5FU efficacy was observed by Ishida et al. (2005). Decreased TK activity has been

shown to be associated with 5FU insensitivity (van der Wilt et al., 1998).

As CH2THF is crucial in maintaining a stable ternary complex with TS and 5-FdUMP,

folate metabolism becomes the focus for 5FU response prediction. Intracellular CH2THF

concentration is mainly controlled by MTHFR. Several polymorphisms are commonly

found in the MTHFR gene, with 677C>T polymorphism in exon 4 and 1298A>C

polymorphism in exon 7 being most commonly linked with altered TS activity (Etienne

et al., 2007). The 677C>T polymorphism is frequently associated with enhanced 5FU

activity due to increased CH2THF levels (Cohen et al., 2003; Etienne et al., 2004).

Higher 5FU sensitivity was observed in mutated A1298C variants of MTHFR (Etienne et

al., 2004).

A promising molecular marker related to the downstream mechanisms of 5FU is p53.

Loss of p53 function may contribute to 5FU resistance as p53-deficient cells fail to

undergo p53-mediated apoptosis (Longley et al., 2002). Mutations in p53 occur in more

than 50% of colorectal tumors, often resulting in p53 overexpression (Brett et al., 1996).

Tumor overexpression of p53 was associated with lower response rate and higher rate of

deterioration radiologically and clinically. Others have also provided evidence for the

correlation between mutations/deletions of p53 with 5FU resistance (Lowe et al., 1993;

Bunz et al., 1999; Liang et al., 2002).

LITERATURE REVIEW

21

Some studies have investigated these determinants in minimal combinations, as

summarized in Table 1.

LITERATURE REVIEW

22

Table 1: Overview of studies on gene/protein expression levels in relation to 5FU chemosensitivity. Gene/protein Sample Outcome References RNA UP, TP 8 colorectal

cancer cell lines

UP and TP expression were predictive for 5FU cytotoxicity in 5/7 cell lines

Mader et al., 1997

DPD, TP, TS 33 colorectal tumors

Colorectal tumors responding to 5FU had low DPD, TS and TP

Salonga et al., 2000

DPD, TS 5 xenografts, 85 gastric tumors

Low TS and DPD was associated with 5FU sensitivity

Fujiwara et al., 2002

DPD, TS 37 colorectal tumors

Patients with low TS and DPD had better response rate and longer median survival time

Ichikawa et al., 2003

DPD, OPRT, UP

37 colorectal tumors

UFT-responding tumors had a statistically higher OPRT/DPD ratio survived longer than those with low ratio

Ichikawa et al., 2003

DPD, TS 23 colorectal tumors

Increase in TS in 5FU-sensitive colorectal cancers

Inoue et al., 2005

DPD, TP, TS 4 colorectal cancer cell lines

Higher sensitivity to 5FU of cell lines with the lowest TS expression

Amatori et al., 2006

DPD, OPRT, TP, TS, UP

HCT116 xenotransplant

DPD and TP levels were inversely correlated with 5FU sensitivity

Okumura et al., 2006

DPD, OPRT, TP, TS


DPD levels and OPRT/DPD ratio is weakly correlated with 5FU sensitivity

Kinoshita et al., 2007

DPD, OPRT, TP, TS


Patients with high OPRT had longer disease-free and overall survival

Yamada et al., 2008

Protein OPRT, UP NUGC-

3/5FU/L cell line

5FU resistance was due to reduced activities of 5FU anabolizing enzymes

Inaba et al., 1996

DPD, TP, TS 41 colon tumors

Intratumoral TS protein expression was inversely correlated with response to chemotherapy

Cascinu et al., 1999

DPD, TP, TS 40 rectal tumors

High TS and low TP was associated with nonresponse to 5FU

Jakob et al., 2005

LITERATURE REVIEW

23


Patients with low TS and DPD had longer disease-free survival and overall survival

Ciaparrone et al., 2006


Low TS and DPD were prognostic for worse outcome in patients treated by surgery alone; low DPD, TP and TS were prognostics for better outcome in patients treated with 5FU

Soong et al., 2008

p53, TS 122 colorectal tumors

No relationship between TS and p53 expression with 5FU response

Berglund et al., 2002

p53, TS 706 colorectal tumors

None of the markers could be used to predict benefit from 5FU treatment

Allegra et al., 2003

p53, TS 967 colorectal cancer

None of the markers was significantly associated with survival

Popat et al., 2006

LITERATURE REVIEW

24

1.4.2 Genome-wide approach

As 5FU metabolism is a complex event and its biochemical action involves multiple

targets, multiple genes involved in pyrimidine metabolism and activity may be associated

with 5FU sensitivity. Approximately 50 genes identified from Kyoto Encyclopedia of

Genes and Genomes (KEGG) pathway database are found to be involved in the normal

pyrimidine metabolism which is important for thymidine and cytidine synthesis

(http://www.genome.ad.jp/kegg/pathway.html). Insufficient folate supply due to impaired

folate metabolism may also affect 5FU action as MTHF is crucial for stabilizing the

tertiary complex formed with 5FU and thymidylate synthase. Also, there is growing body

of evidence which suggests the involvement of many other genes/proteins in the transport

mechanism, stress response and DNA repair in influencing 5FU sensitivity.

The role of efflux transporters in determining FP response has been investigated by Guo

and coworkers in an in-vitro transport assay (Guo et al., 2003). Using MRP8-transfected

LLC-PK1 cells, the investigators demonstrated that MRP8 (ABCC11) functions as a

resistance factor for 5FU by mediating the efflux of 5-FdUMP, the intracellular cytotoxic

metabolite of 5FU. The role of MRP8 in 5FU resistance is also confirmed by Oguri and

colleagues (Oguri et al., 2007). The high degree of structural resemblance between

MRP8 and MRP9 (Tammur et al., 2001, Yabuuchi et al., 2001) suggests that MRP9 may

participate in 5FU efflux. MRP5-transfected cells were found to be resistant to 5FU (Pratt

et al., 2005). However, a review by Borst and coworkers on MRP4 and MRP5 found no

evidence that the transport of nucleotide analogs results in resistance in vivo (Borst et al.,

LITERATURE REVIEW

25

2007). Further investigation on the role of 5FU transport in clinical resistance to 5FU is

clearly needed.

Overexpression of DNA repair genes gives rise to tumor resistance to DNA-damaging

agents (Yu et al., 2006). Deficiency in DNA mismatch repair (MMR) genes MLH1,

MSH2, MSH6 and PMS2 was shown to have a negative impact on overall survival in

CRC, suggesting that MMR status may contribute to decision making regarding treatment

approaches (Gologan et al., 2005). Ide et al. has reported an association between low

MLH1 expression with better disease-free survival rate in 94 colorectal cancer patients

(Ide et al., 2008). Cell lines deficient of Smug-1, a DNA glycosylase in base excision

repair mechanism, were found to be highly sensitive to 5FU cytotoxicity. Figure 4 shows

the genes involved in the DNA repair mechanisms and stress response in relation to 5FU

effects. However to date, their influence has not been well characterized.

LITERATURE REVIEW

26

Figure 4: Activation of DNA repair mechanisms and stress response in relation to 5FU effects. 5FU-induced DNA damage leads to activation of ataxia telangiectasia mutated (ATM) and ataxia telangiectasia related (ATR). ATM phosphorylates Chk2 while ATR phosphorylates Chk1. Both Chk2 and Chk1 phosphorylates p53, leading to cell cycle arrest, apoptosis and enhanced DNA repair. Poly(ADP-ribose) polymerase (PARP)-1 and PARP-2 proteins are also play an important role in DNA repair.

DDNNAA rreeppaaiirr UNG, SMUG1, XRCC1, HMGB1, MLH1, MSH2, MSH6, PMS2, Exo1

DNA damage PARP1 & 2

ATR/ATM

CHK1

CCeellll ccyyccllee iinnhhiibbiittiioonn

AAppooppttoossiiss Bcl-2

CHK2

p53

p53

5FU

LITERATURE REVIEW

27

With the advent of microarray technology, gene expression profiling has been applied to

identifying determinants of 5FU response at a genome/proteome-wide level (Table 2).

Several model systems have been investigated for 5FU-based therapies in the treatment

of colorectal cancer, including the cancer cell lines, tumor xenografts and primary

as reviewed by Mariadason et al. (Mariadason et al., 2004). These studies analyzed basal

gene expression profiles and identified determinants for pre-treatment drug response.

Genome-wide expression analysis has also been applied to identifying determinants from

after-drug treatment by identifying genes that were altered following drug exposure (De

Angelis et al., 2006; Boyer et al., 2006).

LITERATURE REVIEW

28

Table 2: Summary of genome-wide and proteomic studies to identify determinants for 5FU-based chemosensitivities. Sample Compounds Outcome References RNA 60 human cancer cell lines

1400 compounds including 5FU and analogues

Negative correlation between DPYD expression and 5FU potency

Scherf et al., 2000

39 human cancer cell lines

55 anticancer agents

Survivin was highly expressed in 5FU resistant cells while members of the aldo-keto reductase and aldehyde dehydrogenase families and galectin 4 were highly expressed in 5FU responsive cells

Dan et al., 2002

85 human cancer xenografts

5FU 20 genes of the highest correlation to 5FU chemosensitivity were not previously reported

Zembutsu et al., 2002

30 colon cancer cell lines

5FU and camptothecin

Expression of 50 genes not previously reported were correlated with 5FU-induced-apoptosis and significantly predicted response to 5FU

Mariadason et al., 2003

23 rectal cancer patients

5FU and radiation

Expression levels of 54 genes not previously reported were significantly different between 5FU responders and non-responders

Ghadimi et al., 2005

MIP101 colorectal cancer cells

5FU Low SPARC expression in resistant cell line

Tai et al., 2005

3 colorectal cancer and corresponding resistant sublines

5FU 81 genes were identified to be related to 5FU sensitivity

Shimizu et al., 2005

22 colorectal cancer patients

5FU and analogues

Expression levels of TNFRSF1B, SLC35F5 and OPRT were significantly different in 5FU responders and non-responders

Matsuyama et al., 2006

30 human cancer xenografts

5FU analogues

Tumors with low DPYD expression had high sensitivity to 5FU based therapy, except S-1

Ooyama et al., 2006

Protein Colon cancer cell lines

5FU Metabotropic glutamate receptor 4 was overexpressed in 5FU resistant cells

Yoo et al., 2004

LITERATURE REVIEW

29

Colon cancer cell lines

5FU Lower expression of the alpha subunit of mitochondrial F(1)F(0)-ATP synthase in 5FU-resistant cells

Shin et al., 2005

SW480 cells 5FU After 5FU treatment, cyclophilin A, cytokeratin (CK) 19, CK8, ras-related nuclear protein, heat shock protein (hsp) 27 and peroxiredoxin 6 (Prx 6) were upregulated whereas hsp60, CK18, CK9, carbamoylphosphate synthetase I, alpha-enolase, hsp70, nm23 and beta-actin were down-regulated

Wong et al., 2008

LITERATURE REVIEW

30

1.5 Study approaches

In recognizing the value of identifying determinants for 5FU pre-treatment response

prediction and the lack of promising markers for clinical implementation, a multiple-gene

approach based on real-time quantitative RT-PCR was employed in this study, focusing

on genes from fluoropyridimine- and closely related cellular pathways. This pathway

based analysis of best characterized genes enables us to better investigate the combined

effects of gene expression relevant to 5FU mechanism as well as limits the potential

genes of interest and avoids too many irrelevant, false positive or negative markers.

Furthermore, genome-wide analyses require a reasonably large sample size to accurately

determine the predictive value of a gene and this requirement is often unmet.

In order to identify factors robustly related to 5FU sensitivity, we have investigated cell

lines, tumor xenografts and clinical samples. Cell lines and tumor xenografts allows

assessment of tumor-directed effects, which is less achievable in humans. Furthermore, it

was reported that antitumor effects in human tumor xenograft models correlate well with

clinical effects (Tashiro et al., 1989).

Examination on regulation patterns of 5FU pathway genes in response to 5FU timepoint

treatment in resistant and sensitive cell lines was used as another approach to identifying

potential determinants for 5FU sensitivity. RNAi technology was used to validate the

determinants identified from gene expression studies of pre- and post-drug treatment.

LITERATURE REVIEW

31

1.5.1 Quantification of mRNA expression using real-time PCR

Quantitative RT-PCR (Q-PCR) is sensible method for quantifying gene expression levels

and could be realized with small amount of tissue. However, quantification of many

genes laborious as primers and probes specific for each gene need to be individually

designed. Although high-throughput gene expression profiling with microarray is a

powerful alternative and allows more genes to be investigated, it requires more sample

tissues and there is a higher probability of having false positive or negative markers that

still require validation with Q-PCR.

The introduction of low-density array (LDA) based on the principle of TaqMan Q-PCR

has made the quantification of multiple genes easier and faster than before, requiring

much less template cDNA than conventional Q-PCR. The chemistry employed in the

real-time Q-PCR is to use a Taq polymerase that has a 5′exonuclease activity. The

principle of the process was first described by Holland et al. (Holland et al., 1991) and it

was utilized with fluorescent and quencher dyes by Livak et al. (Livak et al., 1995). In

the TaqMan probe-based approach, the probe contains one fluorophore (6-carboxy-

fluoroscein) at its 5′ terminus and a quencher (6-carboxytetramethyl- rhodamine;

TAMRA) at its 3′ terminus. When in close proximity, the two dyes formed a quenched

system and no fluorescence was observed. However, when the probe hybridized to its

template, the 5′exonuclease activity of the Taq polymerase would hydrolyse the probe

during PCR amplification. Once the fluorophores were separated from the quencher, the

emissions of the fluorophore were no longer being quenched and fluorescence was

detected (Figure 5). The cycle number at which the real-time fluorescence signals is

LITERATURE REVIEW

32

detected represents the progression of the PCR reaction and is used as an indicator of

successful target amplification (Wilhelm et al., 2001). This cycle number, also known as

the threshold cycle (CT) is defined as the PCR cycle in which the increase in fluorescence

generated by the accumulating PCR amplicons exceeds the mean baseline fluorescence

and is proportional to the number of target copies present in the sample (Gibson et al.,

1996; Jung et al., 2000). Thus, the development of real-time PCR also allows true

quantitation of the target nucleic acids. Quantitation of real-time PCR can be performed in

two ways - relative quantitation and absolute quantitation. Relative quantitation describes

changes in the amount of a target sequence compared with its level in a related matrix.

Generally, relative quantitation provides sufficient information and is simpler to develop.

On the other hand, absolute quantitation states the exact number of nucleic acid targets

present in the sample, based on the CT of the sample which can then be compared with

similar data collected from a series of standards by the calculation of a standard curve

(Freeman et al., 1999).

LDA is a 384-well micro fluidic card that enables 384 simultaneous real-time PCR

reactions, determining the expression of multiple, user-defined gene clusters (Figure 6).

This array allows for 1 to 8 samples to be run in parallel against 12 to 384 gene

expression targets that are pre-loaded into each well on the card. The reproducibility of

LDA have also been demonstrated and results were compatible with the conventional

single gene based Q-PCR (Abruzzo et al., 2005; Steg et al., 2006; Lu et al., 2007).

LITERATURE REVIEW

33

Figure 5: Diagram illustrating the concepts for the TaqMan probe assay. When in close proximity, the reporter dye (R) and the quencher (Q) formed a quenched system and no fluorescence is observed. Upon exonuclease digestion during the PCR reaction, the fluorigenic dye is released from the close proximity of the quencher dye and is hence able to fluoresce and be detected by the fluorimeter (Adapted from Applied Biosystems, USA).

LITERATURE REVIEW

34

A.

B.

Figure 6: TaqMan® LDA. (A) Plate layout; (B) LDA format of 96 genes (Format 96a).

LITERATURE REVIEW

35

1.5.2 Regulation patterns of genes in response to 5FU treatment

Tumor resistance to 5FU may manifest at the start of chemotherapy or develop during

treatment. Alteration in gene regulation after exposure to 5FU could be the key to 5FU

resistance. It has been shown that 5FU treatment for 24 hours in resistant cell lines

resulted in p53 accumulation, up-regulation of p53-target genes on DNA damage

response, cell cycle regulation and apoptotic pathway (De Angelis et al., 2006). Another

gene expression analysis identified genes like PDF, CCNG2, AVEN, SSAT and JAG1 to

be consitutively dysregulated in 5FU resistant cells and transiently altered following 5FU

exposure for 0, 6, 12 and 24 hours (Boyer et al., 2006).

Multi-timepoint temporal gene expression analysis following 5FU treatment could offer

insights into understanding the 5FU effects on gene regulation and identifying sets of

genes responsible for these changes. By comparing the regulation patterns of genes in

5FU resistant and sensitive samples, we could possibly identify genes that govern 5FU

resistance if there is distinctive difference in their respective regulation patterns.

1.5.3 RNA interference (RNAi)

RNA interference (RNAi) is a defense strategy which was first described in

Caenorhabditis elegans (Fire et al., 1998). It is a sequence-specific, post-transcriptional

gene silencing process initiated by double-stranded RNA (dsRNA) molecules (Fire et al.,

1998; Figure 7). These dsRNAs are first cleaved by Dicer to produce functional siRNAs

(19-23 nucleotides long). Subsequently, these siRNAs are loaded into the RNA-induced

silencing complex (RISC), which contains helicase that unwinds the duplex (Nykanen et

LITERATURE REVIEW

36

al., 2001). The siRNA strands direct the RISC to the target mRNA and cleave the

complementary mRNA. The mRNA is rapidly degraded, resulting in reduction or

blocking of protein expression.

RNAi technology has been widely applied in cancer research and treatment (Martin et al.,

2007, Bruserud et al., 2007). There are many advantages of using siRNA over other

methods, such as chemical inhibitors and dominant negative mutants, to knockdown gene

expression. First of all, virtually any gene can be silenced by siRNA which is highly

specific. On the contrary, there are not many chemical inhibitors available for gene

silencing and most of them are non-specific. Direct comparison of an optimized

phosphorothioate-modified ASO with a siRNA directed against the same target mRNA

site found that the siRNA was approximately 100 to 1000-fold more efficient (Bertrand et

al., 2002). Second, the synthesis of siRNA is more straight-forward and less time

consuming than constructing dominant negative mutants. Lastly, siRNA silencing is more

specific and has longer sustained silencing than chemical inhibitors and dominant

negative mutants (Bertrand et al., 2002). This could be due to the protection of the

siRNAs from intracellular degradation by its incorporation into the RISC.

Although siRNA is highly specific, there are off-target effects which need to be

minimized in order to have a meaningful siRNA application. Off-target effects can be

divided into two types of responses: (a) the induction of nonsequence-specific silencing

pathways via Type 1 Interferon pathway and (b) the silencing of targets that have partial

complementarity to the siRNA.

LITERATURE REVIEW

37

Figure 7: The RNA Interference pathway. Long double-stranded RNA (dsRNA) is cleaved by the RNase III family member, Dicer, into small interfering RNAs (siRNAs) in an ATP-dependent reaction. These siRNAs are then incorporated into the RNA-induced silencing complex (RISC). The single-stranded antisense strand guides RISC to messenger RNA that has a complementary sequence, which results in the endonucleolytic cleavage of the target mRNA.

LITERATURE REVIEW

38

1.6 Scope of Study

The overall aim of this study is to identify and validate determinants of 5FU sensitivity

based on a comprehensive examination of genes involved in its transport, metabolism and

activity (Figure 8). The specific aims are:

1. To rank genes (or combinations thereof) involved in 5FU transport, metabolism and

activity according to their association with 5FU sensitivity in cell lines, xenografts and

human tissue samples

Hypothesis: Expression levels of genes involved in 5FU transport, metabolism and

activity influence overall cellular 5FU sensitivity. By ranking the genes (or combinations

thereof) according to strength and consistency of association with 5FU sensitivity in cell

lines, xenografts and human tissue samples, the strongest determinants of 5FU sensitivity

can be prioritized for further investigation.

2. To characterize how modulation of expression of prioritized genes affects 5FU

sensitivity in vitro

Hypothesis: Modulation of the expression of genes that have a determining role in 5FU

sensitivity will change 5FU sensitivity and reveal mechanisms for this change.

F Figure 8: Geness analyzed in thiis study for theirr roles in transpoort and metabollism of 5FU andd its downstream

LITERATUR

m events.

RE REVIEW

39

LITERATURE REVIEW

40

5FU: 5-fluorouracil; 5-Fo-THF: 5-Formimino-THF; 5-Fm-THF: 5-formyl-THF; 10-Fm-THF: 10-formyl-THF; 5-met-THF: 5-methyl-THF; 5,10-met-THF: 5,10-methenyl-THF; 5,10-metn-THF: 5,10-methylene-THF; (F): Fluoro; ABCC4: ATP-binding cassette, sub-family C, member 4; ABCC5: ATP-binding cassette, sub-family C, member 5; ABCC11: ATP-binding cassette, sub-family C, member 11; ABCC12: ATP-binding cassette, sub-family C, member 12; AK3: Adenylate kinase 3; ALDH1L1: Aldehyde dehydrogenase I family, member L1; AMT: Aminomethyltransferase; ATIC: 5’-aminoimidazole-4-carboxamide ribonucleotide formyltransferase; ATR: Ataxia telangiectasia and Rad3 related; ATM: Ataxia telangiectasia mutated homolog; BAL: β-alanine; bcl-2: B-cell CLL/lymphoma 2; CANT1: Calcium activated nucleotidase I; CDA: Cytidine deaminase; CHEK1: CHK1 checkpoint homolog; CHEK2: CHK2 checkpoint homolog; CMP: Cytidine monophosphate CDP: Cytidine diphosphate CTP: Cytidine triphosphate; CTPS: CTP synthase; Cyd: Cytidine; DHF: Dihydrofolate; DHFR: DHF reductase; DPYS: Dihydropyrimidinase; DPYD: Dihydropyrimidine dehydrogenase; dUDP: Deoxyuridine diphosphate; dUMP: Deoxyuridine monophosphate; DUT: dUTP pyrophosphatase; dUTP: Deoxyuridine triphosphate; dUrd: Deoxyuridine; dTMP: Deoxyuridine triphosphate; DTYMK: Deoxythymidylate kinase; dUMP: Deoxyuridine monophosphate; ENTPD: Ectonucleoside triphosphate diphosphohydrolase; Exo1: Exonuclease 1; FTCD: Formiminotransferase cyclodeaminase; GART: Phosphoribosylglycinamide formyltransferase; HMGB1: High-mobility group box 1; MLH1: mutL homolog 1; MSH2: mutS homolog 2; MSH6: mutS homolog 6; MTFMT: Mitochondrial methionyl-tRNA formyltransferase; MTHFD: Methylenetetrahydrofolate dehydrogenase; MTHFD1L: Methylenetetrahydrofolate dehydrogenase (NADP+ dependent) I-like; MTHFR: 5,10-methylenetetrahydrofolate reductase (NADPH); MTHFS: 5,10-methenyltetrahydrofolate synthetase; MTR: 5-methyltetrahydrofolate-homocysteine methyltransferase; NME: Non-metastatic cells; NT: Nucleotidase; NUDT2: Nudix (nucleoside diphosphate linked moiety X)-type motif 2; PARP-1: Poly (ADP-ribose) polymerase 1; PARP-2: Poly (ADP-ribose) polymerase 2; Postmeiotic segregation increased 2; RRM: Ribonucleotide reductase; SHMT: Serine hydroxymethyltransferase; SLC28A1: Solute carrier family 28, member 1; SLC28A2: Solute carrier family 28, member 2; SLC28A3: Solute carrier family 28, member 3; SLC29A1: Solute carrier family 29, member 1; SLC29A2: Solute carrier family 29, member 2; SLC29A3: Solute carrier family 29, member 3; SLC29A4: Solute carrier family 29, member 4; SMUG1: Single-strand-selective monofunctional uracil-DNA glycosylase 1; THF: Tetrahydrofolate; TK: Thymidine kinase; TP: Thymidine phosphorylase; TP53: Tumor protein p53; TS: Thymidylate synthase; TXNRD: Thioredoxin reductase; TYMP: Thymidine phosphorylase; UCK: Uridine-cytidine kinase; Urd: Uridine; UDP: Uridine diphosphate; UMP: Uridine monophosphate; UMPS: Uridine monophosphate synthetase; UNG: Uracil-DNA glycosylase; UTP: Uridine triphosphate; UPB1: β-ureidopropinase; UppppU: P1,P4-Bis(5'-uridyl) tetraphosphate; XRCC1: X-ray repair complementing defective repair in Chinese hamster cells 1

MATERIALS AND METHODS

41

CHAPTER 2: MATERIALS AND METHODS 2.1 Cell lines

Twenty one colorectal cancer cell lines of human origin were investigated. Caco-2,

Colo201, Colo205, Colo320DM, DLD-1, HCT116, HT29, LoVo, LS513, LS174T, RKO,

SW403, SW480, SW620, SW837 and WiDR were obtained from American Type Culture

Collection (ATCC, Rockville, MD). CCK-81, CoCM-1, HCC56, OUMS23 and RCM1

were obtained from Japanese Collection of Research Bioresources (Osaka, Japan). Cells

were routinely cultured in DMEM or RPMI1740 supplemented with 10% fetal bovine

serum (Invitrogen, Madison), 50,000 units penicillin and 50mg streptomycin (Sigma, St

Louis, MO) at 37°C in a humidified atmosphere containing 5% CO2.

2.2 Cell sensitivity analysis

Cells were seeded in 100 µl medium onto 96-well microtitration plates (Nunc, Rochester,

NY) with an initial cell density of 5000 per well. After 24 hours, 100 µl of medium

containing 5FU (Sigma) with graded concentrations ranging from 0.01 µM to 1000 µM

was added to the wells and cells were incubated for another 48 hours. Cytotoxicity was

assessed by methyl thiazolyltetrazolium (MTT)-based CellTiter 96® Non-Radioactive

Cell Proliferation assay (Promega, Madison, WI) according to manufacturer’s

instructions and analyzed at 570 nm. All investigations were performed in triplicates. IC50

values were determined from curves plotted according to the dose-response function of

GraphPad Prism software (GraphPad Software Inc., San Diego, CA).


42

2.3 Xenografts

Fourteen colon carcinoma xenografts were kindly provided by Dr. Iduna Fichtner at the

Max-Delbrück-Center for Molecular Medicine Berlin-Buch (Berlin, Germany). The

antitumor activity of 5FU in the 14 xenografts has been previously characterized

(Fitchner et al., 2004). Non-response to 5FU was defined as tumor shrinkage of less than

50% after drug treatment.

2.4 Primary Tumors

Total RNA and clinicopathological data of 25 stage II/III colorectal frozen tumors of

patients treated with 5FU-based chemotherapy was kindly provided by NUH-NUS Tissue

Repository, Singapore under protocols approved by the Institutional Review Board,

National University of Singapore. Tumor RNA provided was extracted from whole

human tissue of tumor content ≥50% using the TRI reagent (MRC, Cincinnati, OH, USA)

according to manufacturer’s instructions. Twelve patients were females and 13 were

males, with ages ranging from 45 to 82 year-old (median 62). Four cases were stage II

disease. Treatment consisted of standard 5FU and leucovorin administration for at least 1

cycle after tumor resection. Lack of response to 5FU treatment is defined as recurrence

within 5 years.

2.5 Gene expression analysis

A total of 92 genes involved in the pyrimidine and folate metabolism were identified

from Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway database

(http://www.genome.ad.jp/kegg/pathway.html), whereas those in the transport and


43

downstream mechanisms were obtained from literature review (Table 3). Four

housekeeping genes (18s, ACTB, B2M and GAPDH) were also included into the assay.


44

Table 3: List of 96 genes selected for LDA analysis.

Gene symbol Gene name Alias GenBank Accession No.

References

Housekeeping genes

18S 18S ribosomal RNA - X03205.1 Banda et al., 2008

ACTB Actin, beta - NM_001101.3 Banda et al., 2008

B2M Beta-2-microglobulin - NM_004048.2 Banda et al., 2008

GAPDH Glyceraldehyde-3-phosphate dehydrogenase

- NM_002046.3 Banda et al., 2008

Transporters

ABCC4 ATP-binding cassette, sub-family C (CFTR/MRP), member 4

MRP4 NM_005845.3

Borst et al., 2007


MRP5 NM_005688.2 Pratt et al., 2005


MRP8 NM_032583.3

Guo et al., 2003

ABCC12 ATP-binding cassette protein C12; ATP-binding cassette transporter sub-family C member 12

MRP9 NM_033226.2 Yabuuchi et al., 2001

SLC28A1 Solute carrier family 28 (sodium-coupled nucleoside transporter), member 1

CNT1 NM_004213.3

Griffith and Jarvis, 1996


CNT2 NM_004212.2



CNT3 NM_022127.1

Ritzel et al., 2001

SLC29A1 Solute carrier family 29 (nucleoside transporters), member 1

ENT1 NM_004955.2



ENT2 NM_001532.2



ENT3 NM_018344.4

Baldwin et al., 2005


ENT4 NM_153247.2

Metabolism (pyrimidine)

AK3 Adenylate kinase 3 AK6 NM_016282.2 Kegg (pyrimidine)

CANT1 Calcium activated nucleotidase I ENTPD8 NM_138793.2 Kegg (pyrimidine)

CDA Cytidine deaminase CDD NM_001785.2 Kegg (pyrimidine)

CES1 Carboxylesterase 1 ACAT NM_001266.4 Kegg (pyrimidine)

CES2 Carboxylesterase 2 CE-2 NM_003869.4 Kegg (pyrimidine)

CTPS CTP synthase - NM_001905.2 Kegg (pyrimidine)


45

CTPS2 CTP synthase II - NM_019857.3 Kegg (pyrimidine)

DPYD Dihydropyrimidine dehydrogenase DPD NM_000110.3 Kegg (pyrimidine)

DPYS Dihydropyrimidinase DHPase NM_001385.2 Kegg (pyrimidine)

DTYMK Deoxythymidylate kinase TYMK NM_012145.2 Kegg (pyrimidine)

DUT dUTP pyrophosphatase dUTPase NM_001948.3 Kegg (pyrimidine)

ECGF1 Thymidine phosphorylase NTPDase-1

NM_012145.2 Kegg (pyrimidine)

ENTPD1 Ectonucleoside triphosphate diphosphohydrolase 1

NTPDase-3



NTPDase-4



NTPDase-5



NTPDase-6



ITPase NM_001247.2 Kegg (pyrimidine)

ITPA Inosine triphosphatase NM23-H1 NM_181493.1 Kegg (pyrimidine)

NME1 Non-metastatic cells 1 NM23-H2 NM_000269.2 Kegg (pyrimidine)






NME7 Non-metastatic cells 7 PNP NM_013330.3 Kegg (pyrimidine)

NP Nucleoside phosphorylase DNT NM_000270.2 Kegg (pyrimidine)

NT5C 5’,3’-nucleotidase, cytosolic CN1A NM_014595.1 Kegg (pyrimidine)

NT5C1A 5’-nucleotidase, cytosolic IA CN1B NM_032526.1 Kegg (pyrimidine)

NT5C1B 5’-nucleotidase, cytosolic IB GMP NM_033253.2 Kegg (pyrimidine)

NT5C2 5’-nucleotidase, cytosolic II UMPH NM_012229.2 Kegg (pyrimidine)

NT5C3 5’-nucleotidase, cytosolic III CD73 NM_016489.11

Kegg (pyrimidine)

NT5E 5’-nucleotidase, ecto (CD73) dNT-2 NM_002526.1 Kegg (pyrimidine)

NT5M 5’,3’-nucleotidase, mitochondrial APAH1 NM_020201.3 Kegg (pyrimidine)

NUDT2 Nudix (nucleoside diphosphate linked moiety X)-type motif 2

PNPASE NM_001161.3 Kegg (pyrimidine)


46

PNPT1 Polyribonucleotide nucleotidyltransferase 1

R1 NM_033109.2 Kegg (pyrimidine)

RRM1 Ribonucleotide reductase M1 polypeptide R2 NM_001033.3 Kegg (pyrimidine)

RRM2 Ribonucleotide reductase M2 polypeptide p53R2 NM_001034.1 Kegg (pyrimidine)

RRM2B Ribonucleotide reductase M2 B - NM_015713.3 Kegg (pyrimidine)

TK1 Thymidine kinase 1 - NM_003258.4 Kegg (pyrimidine)

TK2 Thymidine kinase 2 TR1 NM_004614.3 Kegg (pyrimidine)

TXNRD1 Thioredoxin reductase 2 TR3 NM_003330.2 Kegg (pyrimidine)

TXNRD2 Thioredoxin reductase 2 TP NM_006440.3 Kegg (pyrimidine)

TYMS Thymidylate synthase TS NM_001071.1 Kegg (pyrimidine)

UCK1 Uridine-cytidine kinase 1 URK1 NM_031432.1 Kegg (pyrimidine)

UCK2 Uridine-cytidine kinase 2 UK NM_012474.3 Kegg (pyrimidine)

UMPS Uridine monophosphate synthetase OPRT NM_000373.2 Kegg (pyrimidine)

UPB1 Ureidopropinase, beta BUP1 NM_016327.2 Kegg (pyrimidine)

UPP1 Uridine phosphorylase 1 UP NM_181597.1 Kegg (pyrimidine)

UPP2 Uridine phosphorylase 2 UP2 NM_173355.2 Kegg (pyrimidine)

Metabolism

(folate)

ALDH1L1 Aldehyde dehydrogenase I family, member L1

FTHFD NM_012190.2 Kegg (folate)

AMT Aminomethyltransferase GCE NM_000481.2 Kegg (folate)

ATIC 5’-aminoimidazole-4-carboxamide ribonucleotide formyltransferase

AICAR NM_004044.4 Kegg (folate)

DHFR Dihydrofolate reductase - NM_000791.3 Kegg (folate)

FTCD Formiminotransferase cyclodeaminase LCHC1 NM_006657.2 Kegg (folate)

GART Phosphoribosylglycinamide formyltransferase

PRGS NM_000819.3 Kegg (folate)

MTFMT Mitochondrial methionyl-tRNA formyltransferase

FMT1 NM_139242.3 Kegg (folate)

MTHFD1 Methylenetetrahydrofolate dehydrogenase DCS NM_005956.2 Kegg (folate)

MTHFD1L Methylenetetrahydrofolate dehydrogenase (NADP+ dependent) I-like

MTC1THFS

NM_015440.3 Kegg (folate)

MTHFD2 Methylenetetrahydrofolate dehydrogenase (NADP+ dependent)

NMDMC NM_006636.3 Kegg (folate)

MTHFR 5,10-methylenetetrahydrofolate reductase (NADPH)

- NM_005957.3 Kegg (folate)


47

MTHFS 5,10-methenyltetrahydrofolate synthetase - NM_006441.2 Kegg (folate)

MTR 5-methyltetrahydrofolate-homocysteine methyltransferase

MS NM_000254.2 Kegg (folate)

SHMT1 Serine hydroxymethyltransferase 1 (soluble)

CSHMT NM_004169.3 Kegg (folate)

SHMT2 Serine hydroxymethyltransferase 2 GYLA NM_005412.4 Kegg (folate)

DNA repair/apoptosis/cell cycle regulation

ATM Ataxia telangiectasia mutated homolog AT1 NM_000051.3 Yu et al., 2006

ATR Ataxia telangiectasia and Rad3 related FRP1 NM_001184.3 Yu et al., 2006

bcl2 B-cell CLL/lymphoma 2 Bcl-2 NM_000633.2 Christmann et al., 2003

CHEK1 CHK1 checkpoint homolog Chk1 NM_001274.4 Christmann et al., 2003

CHEK2 CHK2 checkpoint homolog Chk2 NM_007194.3 Christmann et al., 2003

Exo1 Exonuclease 1 HEX1 NM_003686.3 Christmann et al., 2003

HMGB1 High-mobility group box 1 HMG1 NM_002128.4 Yu et al., 2006

MLH1 mutL homolog 1, colon cancer, nonpolyposis type 2

HNPCC2 NM_000249.2 Yu et al., 2006

MSH2 mutS homolog 2, colon cancer, nonpolyposis type 1

HNPCC1 NM_000251.1 Yu et al., 2006

MSH6 mutS homolog 6 HNPCC5 NM_000179.2 Yu et al., 2006

PARP-1 Poly (ADP-ribose) polymerase 1 ADPRT NM_001618.3 Christmann et al., 2003

PARP-2 Poly (ADP-ribose) polymerase 2 ADPRT2 NM_005484.3 Christmann et al., 2003

PMS2 Postmeiotic segregation increased 2 HNPCC4 NM_000535.4 Christmann et al., 2003

SMUG1 Single-strand-selective monofunctional uracil-DNA glycosylase 1

UNG3 NM_014311.1 Christmann et al., 2003

TP53 Tumor protein p53 p53 NM_000546.3 Yu et al., 2006

UNG Uracil-DNA glycosylase UDG NM_003362.2 Christmann et al., 2003

XRCC1 X-ray repair complementing defective repair in Chinese hamster cells 1

RCC NM_006297.1 Yu et al., 2006


48

Total RNA was extracted from the respective cell lines and whole tumor xenograft

tissues using the TRI reagent (MRC) according to manufacturer’s instructions. Total

RNA yields and purity were determined spectrophotometrically at 260 and 280 nm. Real-

time RT-PCR was carried out using the Low Density Array (LDA) system (Applied

Biosystems, Foster City, CA), preloaded with inventoried TaqMan® Gene Expression

Assays into 384-well micro-fluidic cards, according to the manufacturer’s instructions. In

brief, 1µg of total RNA was reverse transcribed using the High Capacity cDNA Reverse

Transcription Kit (Applied Biosystems). The reaction mixtures were incubated at 25°C

for 10 min, followed by 37°C for 120 min and 95°C for 5 min. Nine µl of cDNA samples

(equivalent to 400 ng of total RNA), along with 41 µl of nuclease-free water and 50 µl of

2 × PCR master mix, were loaded into each sample port on the micro-fluidic card.

Thereafter, the card was centrifuged twice for 1 min at 331 x g at room temperature and

then sealed. Real-time PCR and relative quantification were performed using an ABI

PRISM 7900 HT Sequence Detection System. The expression values of target gene were

normalized to GAPDH levels using SDS 2.2.1 software (Applied Biosystems).


49

2.6 Small interfering RNA (siRNA) transfections

Pre-designed Silencer Select siRNA (CTPS2, sense strand: 5’-

GGUUCGAGGUAAACCCUAATT-3’ and antisense strand: 5’-

UUAGGGUUUACCUCGAACCGA-3’; GAPDH, sense strand: 5'-

GGUCAUCCAUGACAACUUUTT-3' and antisense strand: 5'-

AAAGUUGUCAUGGAUGACCTT-3; Negative Controls, sense strand: 5’-

UAACGACGCGACGACGUAATT-3’ and antisense strand: 5’-

UUACGUCGUCGCGUCGUUATT-3’) were purchased from Applied Biosystems.

SW620 cells were transfected with CTPS2 siRNA (5 nM) or GAPDH siRNA (10 nM) or

negative controls using siPORT NeoFX transfection agent (Applied Biosystems) at 50%

cell confluence incubated in serum-free Opti-MEM I medium (Invitrogen) according to

the manufacturer’s instructions. Cells were treated with 10 μM of 5FU 48 hours after

transfection. To measure the number of viable cells, cells were stained with Trypan blue

solution (Promega) at 24, 48 and 72 hours after 5FU treatment. Experiments were done in

duplicate.

2.7 Statistical analysis

Using statistical software R version 2.7.1 (R Foundation for Statistical Computing,

Vienna, Austria), unsupervised hierarchical clustering analysis was carried out to

examine gene expression relationships and to identify sample groupings in each series.

The function heatmap.2 in library gplots within R version 2.7.1 was used for the heatmap

and cluster generation. The default distance/dissimilarity matrix was used, which equates


50

to the Euclidean distance measure. Multiple testing correction was performed by

Bonferroni correction.

Two-sample t-test was used to identify differentially expressed gene between sample

groupings with significance of p<0.05. Gene expression analysis over time was carried

out for RKO and SW837 using Short Time-series Expression Miner (STEM) software

(Ernst et al., 2006) to identify significant temporal expression profiles and the genes

associated with these profiles. A schema of analyses is shown in Figure 9.

RESULTS

51

Figure 9: A schema of analyses done in the 3 series (cell lines, xenografts and patients) in identifying determinants of 5FU sensitivity.

RESULTS

52

CHAPTER 3: RESULTS

3.1 Correlation between gene expression and 5FU sensitivity

To determine if 5FU sensitivity correlates with gene expression in the first instance,

unsupervised hierarchical clustering analysis was performed separately on cell line,

xenograft and primary tumor RNA levels. For cell lines, a wide range of 5FU IC50 values

(0.7-32.7 µM) were observed (Table 4a), however there was no correlation between 5FU

sensitivity and gene expression (Figure 10A). This was the case either using a cut-off for

sensitivity of 10 µM (Rossi et al. 2007), 7.91 µM (median) or 8.95 µM (mean).

Nevertheless, related cell lines such as HT29 and WiDR (ATCC), SW480 and SW620

(ATCC), as well as COLO201 and COLO205 (ATCC), grouped together, supporting the

accuracy of the clustering analysis.

The characterization of 5FU sensitivity in xenografts is reported in Table 4b. Xenograft

cases clustered into three major groups (Figure 10B), for which one group (I) of 4 cases

consisted entirely (100%) of responders, while only 2/9 (22%) cases in the other group

were responders (p=0.02). Xenograft sample 5854 clustered alone, indicating it to be an

outlier. As a result, this case was not included in subsequent analysis.

5FU sensitivity in primary tumors characterized as tumor recurrence within 5 years is

shown in Table 4c. In primary tumors, gene expression patterns also displayed a trend for

correlation with 5FU response (Figure 10C). In one major group (I), all 4/4 (100%) cases

were responders, while only 17/21 (71%) cases were responders in the other major group

RESULTS

53

(II) (p>0.05). There was no significant difference in age, sex and tumor stage between

the groups (Table 5).

RESULTS

54

Table 4: Characterization of 5FU sensitivity: (A) IC50 values of 21 colorectal cancer cell

lines; (B) % of tumor shrinkage in 14 xenografts treated by 5FU; (C) Recurrence data of

25 stage II or III colorectal cancer patients treated with standard 5FU and leucovorin

therapy. ‘R’ refers to responder and ‘NR’ refers to non-responder.

A. Cell Line IC50 ± S.D. (μM) Caco-2 5.6 ± 1.6 CCK81 5.0 ± 1.3 CoCMI 9.8 ± 1.5 Colo201 8.5 ± 1.5 Colo205 12.2 ± 1.5 Colo320 3.4 ± 1.4 DLD-1 2.2 ± 1.4 HCC56 17.6 ± 1.3 HCT116 8.0 ± 1.1 HT29 7.9 ± 1.3 LoVo 3.1 ± 1.4 LS174T 4.4 ± 1.4 LS513 13.1 ± 1.5 OUMS23 0.7 ± 2.7 RCM1 9.5 ± 1.6 RKO 19.5 ± 1.4 SW403 32.7 ± 1.3 SW480 7.2 ± 1.5 SW620 9.3 ± 1.3 SW837 0.7 ± 2.1 WiDR 7.4 ± 1.2

RESULTS

55

B.

Xenograft Tumor shrinkage 5FU sensitivity 5676 + R 5677 – NR 5679 – NR 5682 + R 5734 + R 5735 + R 5771 + R 5776 – NR 5841 – NR 5854 – NR 5896 – NR 6044 – NR 6228 – NR 7271 + R

a + indicates ≥50% tumor shrinkage – indicates <50% tumor shrinkage

RESULTS

56

C.

Tumor Recurrence 5FU sensitivity NT00/0008 – R NT00/0025 + NR NT00/0031 – R NT00/0068 – R NT00/0076 – R NT00/0126 – R NT00/0137 – R NT00/0160 – R NT00/0189 – R NT01/0005 – R NT01/0007 – R NT01/0019 – R NT01/0035 – R NT01/0057 – R NT02/0012 – R NT02/0019 – R NT02/0028 – R NT02/0048 – R NT02/0049 – R NT02/0060 – R NT02/0062 – R NT02/0094 + NR NT03/0014 – R NT03/0048 + NR NT04/0099 + NR

a + indicates tumor recurrence within 5 years – indicates no tumor recurrence within 5 years

A.

B.

RESULT

5

TS

57

C.

Figurgenesexpre(blue

re 10: Unsups for (A) cellession, and ge) to 5FU are

pervised hierl lines, (B) x

green indicate indicated in

rarchical cluxenografts antes increasedn the bar abo

ustering of sand (C) primad expressionove the heatm

amples basedary tumors. R. Cases sensmaps.

d on the exprRed indicateitive (red) an

RESULT

5

ression of 96es decreased nd resistant

TS

58

6

RESULTS

59

Table 5: Associations (χ2 test) between 5FU sensitivity and clinicopathological parameters. ‘R’ refers to responder and ‘NR’ refers to non-responder.

Total (%) 5FU sensitivity (%) p R NR

All cases 25 (100%) 21 (84%) 4 (16%) Age, y 1.0

<50 3 (12%) 3 (100%) 0 (0%) ≥50 22 (68%) 18 (82%) 4 (18%)

Sex 1.0Female 12 (48%) 10 (83%) 2 (17%) Male 13 (52%) 11 (85%) 2 (15%)

Stage 0.5

II 4 (16%) 3 (75%) 1 (25%) III 21 (84%) 18 (86%) 3 (14%)

RESULTS

60

3.2 Identification of baseline gene expression associated with 5FU sensitivity

Since the cell line groupings did not correspond with their sensitivity to 5FU, thus, no

candidate gene was derived from this series. Using a two-sample t-test, 8 out of 96 genes

showed significant difference in expression between the 2 main clusters in xenografts

(Table 6). For primary tumors, the expression of 41 genes was significantly different.

Two genes, cytidine triphosphate synthase type II (CTPS2) and non-metastatic cells 4

(NME4) were present in both lists and hence were considered the top priority candidate

determinants of 5FU sensitivity. Indeed, CTPS2 and NME4 expression was higher in

responders than non-responders in both xenografts and primary tumors and the difference

in CTPS2 expression is significant (Figure 11).

RESULTS

61

Table 6: Candidate genes with differential gene expression (p<0.05) identified by two sample t-test for xenografts and primary tumors. Genes that occur in both lists of xenografts and primary tumors are highlighted in bold.

Gene Code Gene t statistic p-value

Fold change (Sensitive/ resistant)

Xenografts ACTB Actin, beta 5.6 <0.01 1.98 18S 18S ribosomal RNA 3.2 0.01 4.57 CANT1 Calcium activated nucleotidase I 3.0 0.01 0.54 MSH6 mutS homolog 2, colon cancer,

nonpolyposis type 1 2.7 0.02 1.29 NME4 Non-metastatic cells 4 2.6 0.02 1.69 CTPS2 CTP synthase II 2.4 0.03 2.25 NT5C 5’,3’-nucleotidase, cytosolic 2.3 0.04 2.19 PARP1 Poly (ADP-ribose) polymerase 1 2.4 0.04 1.98

Primary Tumours

B2M Beta-2-microglobulin 11.5 <0.01 0.17 TXNRD1 Thioredoxin reductase 1 8.7 <0.01 0.31 NT5E 5’-nucleotidase, ecto (CD73) 6.3 <0.01 0.13 ECGF1 Thymidine phosphorylase 6.0 <0.01 0.25 UCK1 Uridine-cytidine kinase 1 5.5 <0.01 0.32 ABCC4 ATP-binding cassette, sub-

family C (CFTR/MRP), member 4 4.7 <0.01 0.26

RRM1 Ribonucleotide reductase M1 polypeptide 4.6 <0.01 0.37

ENTPD1 Ectonucleoside triphosphate diphosphohydrolase 1 4.2 <0.01 0.13

MTFMT Mitochondrial methionyl-tRNA formyltransferase 4.2 <0.01 0.40

UPB1 Ureidopropinase, beta 4.1 <0.01 0.15 ATM Ataxia telangiectasia mutated

homolog 4.0 <0.01 0.21 MTHFR 5,10-methylenetetrahydrofolate

reductase (NADPH) 4.0 <0.01 0.29 CES1 Carboxylesterase 1 4.0 <0.01 0.16 DPYS Dihydropyrimidinase 3.8 <0.01 0.07

RESULTS

62

ENTPD3 Ectonucleoside triphosphate diphosphohydrolase 3 3.8 <0.01 0.16

BCL2 B-cell CLL/lymphoma 2 3.8 <0.01 0.26 PARP2 Poly (ADP-ribose) polymerase 2 3.7 <0.01 0.32 DPYD Dihydropyrimidine

dehydrogenase 3.6 <0.01 0.16 NT5C3 5’-nucleotidase, cytosolic III 3.6 <0.01 0.23 RRM2B Ribonucleotide reductase M2 B 3.5 <0.01 0.44 NME3 Non-metastatic cells 3 3.4 <0.01 0.28 ALDH1L1 Aldehyde dehydrogenase I

family, member L1 3.4 <0.01 0.34 NUDT2 Nudix (nucleoside diphosphate

linked moiety X)-type motif 2 3.4 <0.01 0.30 MTHFD2 Methylenetetrahydrofolate

dehydrogenase (NADP+ dependent) 3.2 <0.01 0.38

CTPS2 CTP synthase II 3.1 <0.01 0.40 NME5 Non-metastatic cells 5 6.3 0.01 0.29 SLC28A3 Solute carrier family 28

(sodium-coupled nucleoside transporter), member 3 3.0 0.01 0.34

ENTPD4 Ectonucleoside triphosphate diphosphohydrolase 4 3.0 0.01 0.32

CHEK2 CHK2 checkpoint homolog 2.9 0.01 0.45 DHFR Dihydrofolate reductase 2.7 0.01 0.55 NME7 Non-metastatic cells 7 2.7 0.01 0.41 MSH2 mutS homolog 2, colon cancer,

nonpolyposis type 1 2.7 0.02 0.50 NME4 Non-metastatic cells 4 2.6 0.02 0.34 NME6 Non-metastatic cells 6 2.5 0.02 0.47 TK2 Thymidine kinase 2 2.5 0.02 0.33 UPP2 Uridine phosphorylase 2 2.4 0.03 0.06 ATR Ataxia telangiectasia and Rad3

related 2.4 0.03 0.36 MLH1 mutL homolog 1, colon cancer,

nonpolyposis type 2 2.3 0.03 0.47 CES2 Carboxylesterase 2 2.3 0.03 0.19

RESULTS

63

A.

p<0.01

Non-response Response0.00

0.01

0.02

0.03

0.04

0.05

CT

PS2

mR

NA

p=0.02


0.05

0.10

0.15

NM

E4

mR

NA

RESULTS

64

B. p>0.05


0.05

0.10

0.15

0.20

CT

PS2

mR

NA

p>0.05


0.5

1.0

1.5

2.0

NM

E4

mR

NA

Figure 11: Difference in expression levels of CTPS2 and NME4 between clusters in (A) xenografts and (B) patients.

RESULTS

65

3.3 Identification of post-treatment gene expression changes associated with 5FU

sensitivity

To identify genes for which expression changes after 5FU treatment best discriminated

5FU sensitivity from resistance, SW837 (5FU sensitive) and RKO (5FU resistant) cells

were seeded at an initial density of 1 x 105/ml for 24 hours. Cells were treated with 10

μM of 5FU for 0, 4, 8, 12, 24 and 48 hours, after which cells were harvested for RNA and

analyzed for the expression of the selected 96 genes.

Genes with similar expression change patterns were grouped based on Pearson

correlation coefficient, and these groups ranked according to degree of similarity in

change patterns (Figure 12). The model profiles (patterns) are generated by the software

in two steps. Firstly, the system exhaustedly listed all possible patterns according to the

user parameter c. “c” represents possible stages that a gene can be regulated between time

points. For example, if c=1, it means that a gene can be up-regulated, down-regulated, or

unchanged, and if c=2, it means that a gene can be up-regulated for 1 or 2 stages, down-

regulated for 1 or 2 stages or unchanged. Thus, there are in total (2*c+1)n-1, where n is the

number of time points. The first time point will always be regarded as unchanged. In our

experiment, we have set: c=2, n=6. Secondly, the software selected a subset of patterns

from the exhaustedly generated patterns, promising that the selected patterns are

representative and distinct from each other, that is, to maximize the minimal distance

between any of the two patterns. Since the optimal solution to the second step is Non-

deterministic Polynomial-time Hard (NP-Hard), a heuristic algorithm is proposed by

Ernst and colleagues (Ernst et al. 2005).

RESULTS

66

The top-ranked patterns were surmised to be the ones that defined sensitivity and

resistance in the respective cell lines. Interestingly, CTPS2 was identified in the top

groups of both RKO and SW837, making it the highest priority candidate gene of

A.

RESULT

6

TS

67

B.

Figurtreatmfor (Aregul

re 12: Patternment accordiA) RKO andlated groups

ns of gene exing to reduci

d (B) SW837are displaye

xpression ating statistica

7. The compoed on the left

t 0, 4, 8, 12, al significanconent genes

ft.

24 and 48 hoce from left of the top th

ours after 10to right, top

hree concurr

RESULT

6

0 μM 5FU to bottom ently

TS

68

RESULTS

69

3.4 Modulation of 5FU sensitivity by CTPS2 siRNA

To verify the direct influence of CTPS2 on 5FU sensitivity, the effects of siRNA

knockdown of CTPS2 expression were examined, with GAPDH siRNA treatment as a

control. 5FU-sensitive cell line (<10 µM) with relatively high CTPS2 levels would be an

ideal choice for CTPS2 knockdown and measuring the change in 5FU sensitivity. SW620

as an adherent cell line was chosen for this investigation as its CTPS2 level was fourth

highest among all 5FU sensitive cell lines. Using chemical transfection agents, siRNA

transfections of the top 2 cell lines in rank (CCK81, CoCM1) were unsuccessful while

transfection of Colo201 (third in rank for its CTPS2 level) was not attempted as these

cells are in suspension. Cell lines grown in suspension culture remain challenging

transfection targets.

More than 50% reduction of CTPS2 and GAPDH RNA levels was achievable in SW620

cells following 48 hours treatment with respective siRNA (Figure 13A and 13B),

although full CTPS2 expression were restored after 120 hours treatment. Cells transfected

for 48 hours with siRNA were then treated with 10 μM 5FU (IC50 of SW620 at 48 hours)

and cell viability was measured at 24, 48 and 72 hours. At 48 hours post-5FU treatment,

viability of CTPS2-knockdown cells was the same as that of transfected cells without

5FU treatment (Figure 13C), indicating a resistance to 5FU of CTPS2-knockdown cells.

This result contrasts with that of cells transfected with negative control or GAPDH

siRNA (Figure 13D and 13E). For these cells, 40-60% reduction in cell growth was

observed at 48 hours post-5FU treatment, consistent with reduction expected from using

an IC50 dosage. At 72 hours post 5-FU treatment (120 hours post-siRNA treatment), the

RESULTS

70

number of viable CTPS2-knockdown cells became 53% of non-treated cells. This is

likely due to the lower CTPS2 knockdown at 120 hours post-siRNA treatment (Figure

5A), and the known transient nature of siRNA knockdown.

RESULTS

71

A.

B.

0

0.2

0.4

0.6

0.8

1

1.2

Negative control CTPS2 (48 hours) CTPS2 (120 hours)

CT

PS2

mR

NA

siRNA

0

0.2

0.4

0.6

0.8

1

1.2

Negative control GAPDH (48 hours) GAPDH (120 hours)

GA

PDH

mR

NA

siRNA

C.

D.

E.

Figur(A) Ccurve5FU negat

re 13: CTPS2CTPS2 siRNes of cells eafter 48 hou

tive control s

2 and GAPDNA and (B) Geither treatedurs of transfsiRNA.

DH mRNA eGAPDH siRd (unfilled cfection of (C

expression leRNA 48 and circles) or uC) CTPS2 s

evels in SW6120 hours a

untreated (filsiRNA, (D)

620 cells traafter transfeclled circles) GAPDH siR

RESULT

7

ansfected witction. Growt

with 10 μMRNA and (E

TS

72

th th M E)

CONCLUSION

73

CHAPTER 4: DISCUSSION

Most investigations on determinants of 5FU sensitivity until now have been candidate

gene approaches. These have mostly been on genes/proteins involved in the metabolism

and cytotoxic mechanisms of 5FU, with thymidylate synthase (TS) (Peters et al., 1995)

and dihydropyrimidine dehydrogenase (DPD) (Diasio et al., 2001) being the most

prominent. TS has been studied as a major target of 5FU activity, and numerous reports

have associated elevated TS expression/activity with 5FU resistance (Johnston et al.,

1995, Berger et al., 1985, Clarke et al., 1987, Johnston et al., 1992). DPD is the rate-

limiting step in 5FU clearance, and its deficiency has been correlated to 5FU toxicity and

increased sensitivity (Diasio et al., 1988; Ishikawa et al., 1999; Milano et al., 1999).

However, not all studies on TS and DPD have been in agreement (Soong and Diasio

Findlay et al., 1997; Berglund et al., 2002; Allegra et al., 2003; Johnston et al., 2003;

Kinoshita et al., 2007; Yamada et al., 2008), indicating other factors are involved.

Gene expression profiling has also been applied to identifying determinants of 5FU

response at a genome-wide level (Scherf et al., 2000; Zembutsu et al., 2002; Mariadason

et al., 2003; Shimizu et al., 2005; Matsuyama et al., 2006). Testing response to multiple

drugs on the NCI panel of 60 cell lines, Scherf et al. identified that DPD is highly

negatively correlated with 5FU response (Scherf et al., 2000). Similar to the NCI panel of

60 cell lines, the cell lines used in our study were exposed to 5FU for 48 hours. The IC50

values for the overlapped cell lines from both experiments were similar except for SW620

which is shown to be more 5FU resistant by Scherf et al. Mariadason et al. studied a

panel of 30 colon cancer cell lines and identified 420 genes, enriched in DNA replication

CONCLUSION

74

and repair and protein processing/targeting, to be significantly associated with 5FU

sensitivity (Mariadason et al., 2003). The 30 colon cancer cell lines described by

Mariadason et al. were exposed to 5FU for 72 hours. Most of the IC50 values of

overlapping cell lines reported by Mariadason et al. were similar except for SW480 and

SW620 which seemed to be more resistant, whilst RKO and SW403 seemed to be more

sensitive to 5FU. Moreover, based on the IC50 values, 2 of the four cell lines tested by

Scherf et al. (Colo205, HCT116) were found to be more sensitive to 5FU than that of

Mariadason et al. Zembutsu and colleagues examined 9 anticancer agents in 85 cancer

xenografts and identified 20 novel genes related to 5FU sensitivity (Zembutsu et al.,

2002). Shimizu et al. analyzed 3 colorectal cancer cell lines and corresponding resistant

sublines and identified 81 genes related to 5FU sensitivity (Shimizu et al., 2005). A

derivative panel of these genes has since been validated with TNFRS1B, SLC35F5 and

OPRT identified as a “Response Index” (Matsuyama et al., 2006).

In this study, CTPS2 was identified as a determinant of 5FU sensitivity through an

integration of pathway-based real-time PCR analysis of three cancer models, an

examination of both basal and temporal gene expression levels and siRNA knockdown. A

pathway-based approach, comprising examination of 92 genes in 5FU transport,

metabolic and downstream activity (Figure 1), was adopted as the candidates identified

were more likely to have a rational basis for influencing 5FU sensitivity. Real-time PCR

was used to achieve optimal detection sensitivity, dynamic range and analytical precision,

as opposed to a hybridization array approach (Wong et al., 2005). Interrogation of cell

lines, xenografts and primary tumours were used to best identify candidates that were

CONCLUSION

75

robustly linked to 5FU sensitivity (Figure 10). However, candidate genes were not

derived from the cell lines as their chemosensitivity did not correspond well with sample

groupings. The lack of correlation between cell lines and clinical specimens has been

described previously (Andreotti et al., 1994). The specific 5FU/LV regimen undertaken

by the patients in this study makes it more valuable as studies which combine patients

receiving different types of 5FU administration introduce greater variation between

patients and thus, their findings do not strictly correspond to drug effect (Gustavsson et

al., 2008). Finally, siRNA knockdowns were used to confirm that reduced CTPS2

expression directly leads to 5FU resistance (Figure 13). Although the use of one CTPS2

siRNA sequence might raise concern with off-target effects, it was shown using BLAST

analysis that the designed siRNA sequence has 100% specificity to the gene of interest,

i.e. CTPS2, thus minimizing off-target effects. Not more than one siRNA sequence is

introduced as high concentration of siRNA could potentially lead to induction of alpha-

and beta-interferon pathway via protein kinase K pathway (Khabar et al., 2003; Garcia et

al., 2006). Taken together, these numerous lines of evidence suggest a role for CTPS2 in

determining 5FU sensitivity.

A major feature of this study is that we have also identified determinants through

temporal gene expression analysis following 5FU treatment. On the premise that drug

sensitivity is likely to be influenced by cell responses to treatment, we characterized the

expression patterns of the 92 pathway-based genes at 6 timepoints over 48 hours in

extremely sensitive (SW837) and resistant (RKO) cell lines. Multiple timepoints were

measured as we considered significant information could be missed between traditional

CONCLUSION

76

two (pre- and post-treatment) timepoint assessments. STEM software has been useful for

clustering genes of similar regulation patterns across a short time-series (Lulko et al.,

2007; Martinez et al., 2007), and we took this further by identifying the most tightly

regulated expression patterns (and component genes) in respective cell lines as

determinants of 5FU sensitivity and resistance. CTPS2 was one of four genes (the others

were DTYMK, ITPA, UPP1) in the top gene groups of both sensitive and resistant cell

lines (Figure 12). Interestingly, the level of CTPS2 went up in the sensitive cell line, and

decreased in the resistant line, consistent with the baseline expression trends.

CTPS2 is one of 2 genes (the other is CTPS) encoding human cytidine triphosphate

synthase (van Kuilenberg et al., 2000). The enzyme catalyses the rate-limiting conversion

of uridine triphosphate (UTP) to CTP, that is important for biosynthesis of RNA, DNA

and lipids. Hence, one possible mechanism for 5FU resistance observed with low CTPS2

expression could be UTP accumulation occurring due to reduced conversion to CTP. The

UTP may compete with the incorporation of FUTP into RNA, thereby reducing the RNA-

directed cytotoxic effects of 5FU. However, other scarce data on CTPS2 neither supports

nor refutes this mechanism, even more so its relationship to 5FU sensitivity. Increased

activity of human CTP synthase has been observed in leukemias and tumors of liver, lung

and colon (Kizaki et al., 1980). Whelan et al. showed treatment of CHO cells with ultra-

violet radiation and ethylmethanesulfonate generates CTPS mutations, eliminating its

feedback regulation by CTP, altering its synthase activity and increasing intracellular

pools of CTP and dCTP and resistance to 5FU (Whelan et al., 1993). Somatic mutation of

neither CTPS nor CTPS2 in tumors has been reported so far. It is indeed possible that

CONCLUSION

77

other mechanisms like loss of heterozygosity or methylation-induced transcriptional

silencing of CTPS2 to cause inactivation of this gene and lead to tumor resistance to 5FU.

The second potential candidate, NME4, is identified from the basal gene expression

analysis but not from the timepoint temporal analysis. NME4 is one of the isoforms of

nucleoside diphosphate kinase (NDPK/Nm23) and encodes for human mitochondrial

NDPK-D. NDPK is important for the regulation of intracellular nucleotide homeostasis.

NME4 was found to be overexpressed in colorectal cancer (Hayer et al., 2001).

Furthermore, high expression of NME4 has been shown to be associated with poor

prognosis in patients with acute myeloid leukemia (AML) (Kracmarova et al., 2008).

However, here we observed that relatively high NME4 expression was associated with

5FU response. This conflicting observation could be explained by the differences in

treatment and tumor characteristics of colorectal cancer and AML and the different

endpoint of response and survival. To further validate the role of NME4, functional

investigations will be carried out.

As compared to earlier studies, CTPS2 has not been identified as a determinant of 5FU

response, though its role in 5FU sensitivity has been suggested for via the gene

expression analyses and in-vitro experiments in our study. This difference in observations

could be due to the use of different types of samples (cell lines/ xenografts/ human

paraffin/ frozen tissues; tumor content) and various experimental approaches (DNA/

RNA/ protein analyses; RNA analysis using microarray/ real-time PCR). One limitation

of our study is sample bias, with only few 5FU-resistant patient tumors being analyzed.

CONCLUSION

78

Continual effort in developing tumor banks of fresh frozen tissue and maintain proper

documentation of clinical follow up is of utmost importance in order to complement the

currently available profiling technologies. When more tumors become available, these

samples could contribute to the validation set for future profiling experiments to validate

the role of CTPS2.

As in some previous studies, TYMS expression did not associated with 5FU sensitivity in

our study. In the xenograft series, DPYD expression was observed to be higher among

non-responders. This is in concordance with the findings by Yoshinare et al. that high

DPYD expression is associated with low 5FU sensitivity (Yoshinare et al., 2003).

However, this trend was not observed in the primary tumors. This could be possibly due

to the under representation of the non responders in this series.

In summary, this is the first study which identified CTPS2 as a pharmacologically

relevant marker which could be predictive for 5FU response, through a convergence of

pathway-based, baseline and temporal, and in-vitro siRNA investigations. Future

investigation of CTPS2 protein levels and activity will help further elucidate the potential

role of CTPS2 in determining sensitivity to 5FU.

CONCLUSION

79

CONCLUSION

The antimetabolite, 5-fluorouracil (5FU) is the mainstay for the treatment of many cancer

types. Inter-patient response to 5FU is however extremely variable, making it desirable to

identify factors that influence 5FU sensitivity. Here, we describe the identification of

cytidine triphosphate synthase II (CTPS2) as a significant determinant of 5FU sensitivity

through an exhaustive search, involving real-time PCR analysis of 92 genes involved in

5FU and folate transport, metabolism and downstream effects in three cancer models

(cell lines, xenografts and primary tumors), with examination of both baseline and

temporal data, and siRNA functional assessment. The identification of CTPS2 presents a

potentially sensitive factor that could be used in the clinic for predicting response to 5FU,

or devising strategies to potentiate drug activity. In addition, the pathway-based, multi-

model data integration, and multi-timepoint temporal analysis approaches used in this

study presents useful alternative strategies to discovering determinants of drug

80

REFERENCES

Abruzzo LV, Lee KY, Fuller A, et al. Validation of oligonucleotide microarray data using microfluidic low-density arrays: a new statistical method to normalize real-time RT-PCR data. Biotechniques 2005; 38:785-92. Advanced Colorectal Cancer Meta-Analysis Project. Modulation of fluorouracil by leucovorin in patients with advanced colorectal cancer: evidence in terms of response rate. J Clin Oncol 1992; 10:896-903. Allegra CJ, Paik S, Colangelo LH, et al. Prognostic value of thymidylate synthase, Ki-67, and p53 in patients with Dukes' B and C colon cancer: a National Cancer Institute-National Surgical Adjuvant Breast and Bowel Project collaborative study. J Clin Oncol 2003; 21:241-50. Amatori F, Di Paolo A, Del Tacca M, et al. Thymidylate synthase, dihydropyrimidine dehydrogenase and thymidine phosphorylase expression in colorectal cancer and normal mucosa in patients. Pharmacogenet Genomics 2006; 16:809-16. Andre T, Boni C, Mounedji-Boudiaf L, et al. Oxaliplatin, fluorouracil, and leucovorin as adjuvant treatment for colon cancer. N Engl J Med 2004; 350:2343-51. Andreotti PE, Linder D, Hartmann DM, Cree IA, Pazzagli M, Bruckner HW. TCA-100 tumour chemosensitivity assay: differences in sensitivity between cultured tumour cell lines and clinical studies. J Biolumin Chemilumin 1994; 9:373-8. Aranha O, Benson AB, 3rd. Adjuvant therapy for colon cancer. Curr Gastroenterol Rep 2007; 9:415-21. Baldwin SA, Beal PR, Yao SY, King AE, Cass CE, Young JD. The equilibrative nucleoside transporter family, SLC29. Pflugers Arch 2004; 447:735-43. Baldwin SA, Yao SY, Hyde RJ, et al. Functional characterization of novel human and mouse equilibrative nucleoside transporters (hENT3 and mENT3) located in intracellular membranes. J Biol Chem 2005; 280:15880-7. Banda M, Bommineni A, Thomas RA, Luckinbill LS, Tucker JD. Evaluation and validation of housekeeping genes in response to ionizing radiation and chemical exposure for normalizing RNA expression in real-time PCR. Mutat Res 2008; 649:126-34. Bathe OF, Franceschi D, Livingstone AS, Moffat FL, Tian E, Ardalan B. Increased thymidylate synthase gene expression in liver metastases from colorectal carcinoma: implications for chemotherapeutic options and survival. Cancer J Sci Am 1999; 5:34-40.

81

Benson AB, 3rd, Schrag D, Somerfield MR, et al. American Society of Clinical Oncology recommendations on adjuvant chemotherapy for stage II colon cancer. J Clin Oncol 2004; 22:3408-19. Berger SH, Jenh CH, Johnson LF, Berger FG. Thymidylate synthase overproduction and gene amplification in fluorodeoxyuridine-resistant human cells. Mol Pharmacol 1985; 28:461-7. Berglund A, Edler D, Molin D, Nordlinder H, Graf W, Glimelius B. Thymidylate synthase and p53 expression in primary tumor do not predict chemotherapy outcome in metastatic colorectal carcinoma. Anticancer Res 2002; 22:3653-9. Bertrand JR, Pottier M, Vekris A, Opolon P, Maksimenko A, Malvy C. Comparison of antisense oligonucleotides and siRNAs in cell culture and in vivo. Biochem Biophys Res Commun 2002; 296:1000-4. Bleiberg H. Adjuvant therapy in high-risk colon cancer. Semin Oncol 2000; 27:48-59. Borst P, de Wolf C, van de Wetering K. Multidrug resistance-associated proteins 3, 4, and 5. Pflugers Arch 2007; 453:661-73. Boyer J, Allen WL, McLean EG, et al. Pharmacogenomic identification of novel determinants of response to chemotherapy in colon cancer. Cancer Res 2006; 66:2765-77. Branda RF, Nigels E, Lafayette AR, Hacker M. Nutritional folate status influences the efficacy and toxicity of chemotherapy in rats. Blood 1998; 92:2471-6. Brett MC, Pickard M, Green B, et al. p53 protein overexpression and response to biomodulated 5-fluorouracil chemotherapy in patients with advanced colorectal cancer. Eur J Surg Oncol 1996; 22:182-5. Bruserud O. Introduction: RNA and the treatment of human cancer. Curr Pharm Biotechnol 2007; 8:318-9. Bunz F, Hwang PM, Torrance C, et al. Disruption of p53 in human cancer cells alters the responses to therapeutic agents. J Clin Invest 1999; 104:263-9. Cassidy J. Adjuvant 5-fluorouracil plus levamisole in colon cancer: the plot thickens? Br J Cancer 1994; 69:986-7. Cascinu S, Aschele C, Barni S, et al. Thymidylate synthase protein expression in advanced colon cancer: correlation with the site of metastasis and the clinical response to leucovorin-modulated bolus 5-fluorouracil. Clin Cancer Res 1999; 5:1996-9. Chang JC, Wooten EC, Tsimelzon A, et al. Gene expression profiling for the prediction of therapeutic response to docetaxel in patients with breast cancer. Lancet 2003; 362:362-9.

82

Chlebowski RT, Lillington L, Nystrom JS, Sayre J. Late mortality and levamisole adjuvant therapy in colorectal cancer. Br J Cancer 1994; 69:1094-7. Ciaparrone M, Quirino M, Schinzari G, 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. Clark JL, Berger SH, Mittelman A, Berger FG. Thymidylate synthase gene amplification in a colon tumor resistant to fluoropyrimidine chemotherapy. Cancer Treat Rep 1987; 71:261-5. Clarke PA, George ML, Easdale S, et al. Molecular pharmacology of cancer therapy in human colorectal cancer by gene expression profiling. Cancer Res 2003; 63:6855-63. Cohen V, Panet-Raymond V, Sabbaghian N, Morin I, Batist G, Rozen R. Methylenetetrahydrofolate reductase polymorphism in advanced colorectal cancer: a novel genomic predictor of clinical response to fluoropyrimidine-based chemotherapy. Clin Cancer Res 2003; 9:1611-5. Dan S, Tsunoda T, Kitahara O, et al. An integrated database of chemosensitivity to 55 anticancer drugs and gene expression profiles of 39 human cancer cell lines. Cancer Res 2002; 62:1139-47. De Angelis PM, Svendsrud DH, Kravik KL, Stokke T. Cellular response to 5-fluorouracil (5-FU) in 5-FU-resistant colon cancer cell lines during treatment and recovery. Mol Cancer 2006; 5:20. Diasio RB, Harris BE. Clinical pharmacology of 5-fluorouracil. Clin Pharmacokinet 1989; 16:215-37. Domin BA, Mahony WB, Zimmerman TP. Transport of 5-fluorouracil and uracil into human erythrocytes. Biochem Pharmacol 1993; 46:503-10. Douillard JY, Cunningham D, Roth AD, et al. Irinotecan combined with fluorouracil compared with fluorouracil alone as first-line treatment for metastatic colorectal cancer: a multicentre randomised trial. Lancet 2000; 355:1041-7. Douillard JY, Hoff PM, Skillings JR, et al. Multicenter phase III study of uracil/tegafur and oral leucovorin versus fluorouracil and leucovorin in patients with previously untreated metastatic colorectal cancer. J Clin Oncol 2002; 20:3605-16. Edler D, Glimelius B, Hallstrom M, et al. Thymidylate synthase expression in colorectal cancer: a prognostic and predictive marker of benefit from adjuvant fluorouracil-based chemotherapy. J Clin Oncol 2002; 20:1721-8.

83

Ernst J, Nau G, Bar-Joseph Z. Clustering Short Time Series Gene Expression Data. Bioinformatics 2005; 21 (Suppl 1): i159-i168.

Ernst J. STEM: a tool for the analysis of short time series gene expression data. BMC Bioinformatics, 7:191, 2006. Etienne MC, Chazal M, Laurent-Puig P, et al. Prognostic value of tumoral thymidylate synthase and p53 in metastatic colorectal cancer patients receiving fluorouracil-based chemotherapy: phenotypic and genotypic analyses. J Clin Oncol 2002; 20:2832-43. Etienne MC, Formento JL, Chazal M, et al. Methylenetetrahydrofolate reductase gene polymorphisms and response to fluorouracil-based treatment in advanced colorectal cancer patients. Pharmacogenetics 2004; 14:785-92. Etienne-Grimaldi MC, Francoual M, Formento JL, Milano G. Methylenetetrahydrofolate reductase (MTHFR) variants and fluorouracil-based treatments in colorectal cancer. Pharmacogenomics 2007; 8:1561-6. Evans WE, Relling MV. Moving towards individualized medicine with pharmacogenomics. Nature 2004; 429:464-8. Fakih M. The Role of Targeted Therapy in the Treatment of Advanced Colorectal Cancer. Curr Treat Options Oncol. 2009 Feb 24. [Epub ahead of print] Fernandez-Contreras ME, Sanchez-Prudencio S, Sanchez-Hernandez JJ, et al. Thymidylate synthase expression pattern, expression level and single nucleotide polymorphism are predictors for disease-free survival in patients of colorectal cancer treated with 5-fluorouracil. Int J Oncol 2006; 28:1303-10. Fichtner I, Slisow W, Gill J, et al. Anticancer drug response and expression of molecular markers in early-passage xenotransplanted colon carcinomas. Eur J Cancer 2004; 40:298-307. Findlay MP, Cunningham D, Morgan G, Clinton S, Hardcastle A, Aherne GW. Lack of correlation between thymidylate synthase levels in primary colorectal tumours and subsequent response to chemotherapy. Br J Cancer 1997; 75:903-9. Fire A, Xu S, Montgomery MK, Kostas SA, Driver SE, Mello CC. Potent and specific genetic interference by double-stranded RNA in Caenorhabditis elegans. Nature 1998; 391:806-11. Freeman WM, Walker SJ, Vrana KE. Quantitative RT-PCR: pitfalls and potential. Biotechniques 1999; 26:112-22, 24-5.

84

Fujiwara H, Terashima M, Irinoda T, 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. Garcia MA, Gil J, Ventoso I, et al. Impact of Protein Kinase PKR in Cell Biology: from Antiviral to Antiproliferative Action. Microbiol Mol Biol Rev 2006; 70(4):1032-1060 Ghadimi BM, Grade M, Difilippantonio MJ, et al. Effectiveness of gene expression profiling for response prediction of rectal adenocarcinomas to preoperative chemoradiotherapy. J Clin Oncol 2005; 23:1826-38. Gibson UE, Heid CA, Williams PM. A novel method for real time quantitative RT-PCR. Genome Res 1996; 6:995-1001. Goldberg AR, Machledt JH, Jr., Pardee AB. On the action of fluorouracil on leukemia cells. Cancer Res 1966; 26:1611-5. Gologan A, Sepulveda AR. Microsatellite instability and DNA mismatch repair deficiency testing in hereditary and sporadic gastrointestinal cancers. Clin Lab Med 2005; 25:179-96. Greenlee RT, Murray T, Bolden S, Wingo PA. Cancer statistics, 2000. CA Cancer J Clin 2000; 50:7-33. Grem JL. 5-Fluorouracil: forty-plus and still ticking. A review of its preclinical and clinical development. Invest New Drugs 2000; 18:299-313. Griffith DA, Jarvis SM. Nucleoside and nucleobase transport systems of mammalian cells. Biochim Biophys Acta 1996; 1286:153-81. Guo Y, Kotova E, Chen ZS, et al. MRP8, ATP-binding cassette C11 (ABCC11), is a cyclic nucleotide efflux pump and a resistance factor for fluoropyrimidines 2',3'-dideoxycytidine and 9'-(2'-phosphonylmethoxyethyl)adenine. J Biol Chem 2003; 278:29509-14. Gustavsson B, Kaiser C, Carlsson G, et al. Molecular determinants of efficacy for 5-FU-based treatments in advanced colorectal cancer: mRNA expression for 18 chemotherapy-related genes. Int J Cancer 2009; 124:1220-6. Hartmann KU, Heidelberger C. Studies on fluorinated pyrimidines. XIII. Inhibition of thymidylate synthetase. J Biol Chem 1961; 236:3006-13. Hayer J, Engel M, Seifert M, Seitz G, Welter C. Overexpression of nm23-H4 RNA in colorectal and renal tumours. Anticancer Res 2001; 21:2821-5.

85

Heggie GD, Sommadossi JP, Cross DS, Huster WJ, Diasio RB. Clinical pharmacokinetics of 5-fluorouracil and its metabolites in plasma, urine, and bile. Cancer Res 1987; 47:2203-6. Heidelberger C, Chaudhuri NK, Danneberg P, et al. Fluorinated pyrimidines, a new class of tumour-inhibitory compounds. Nature 1957; 179:663-6. Hoff PM, Ansari R, Batist G, et al. Comparison of oral capecitabine versus intravenous fluorouracil plus leucovorin as first-line treatment in 605 patients with metastatic colorectal cancer: results of a randomized phase III study. J Clin Oncol 2001; 19:2282-92. Holland PM, Abramson RD, Watson R, Gelfand DH. Detection of specific polymerase chain reaction product by utilizing the 5'----3' exonuclease activity of Thermus aquaticus DNA polymerase. Proc Natl Acad Sci U S A 1991; 88:7276-80. 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. Iacopetta B, Grieu F, Joseph D, Elsaleh H. A polymorphism in the enhancer region of the thymidylate synthase promoter influences the survival of colorectal cancer patients treated with 5-fluorouracil. Br J Cancer 2001; 85:827-30. Ichikawa W, Uetake H, Shirota Y, et al. Both gene expression for orotate phosphoribosyltransferase and its ratio to dihydropyrimidine dehydrogenase influence outcome following fluoropyrimidine-based chemotherapy for metastatic colorectal cancer. Br J Cancer 2003; 89:1486-92. Ichikawa W, Uetake H, Shirota Y, et al. Combination of dihydropyrimidine dehydrogenase and thymidylate synthase gene expressions in primary tumors as predictive parameters for the efficacy of fluoropyrimidine-based chemotherapy for metastatic colorectal cancer. Clin Cancer Res 2003; 9:786-91. Ide T, Kitajima Y, Ohtaka K, Mitsuno M, Nakafusa Y, Miyazaki K. Expression of the hMLH1 gene is a possible predictor for the clinical response to 5-fluorouracil after a surgical resection in colorectal cancer. Oncol Rep 2008; 19:1571-6. Inaba M, Mitsuhashi J, Sawada H, et al. Reduced activity of anabolizing enzymes in 5-fluorouracil-resistant human stomach cancer cells. Jpn J Cancer Res 1996; 87:212-20. Inoue T, Hibi K, Nakayama G, et al. Expression level of thymidylate synthase is a good predictor of chemosensitivity to 5-fluorouracil in colorectal cancer. J Gastroenterol 2005; 40:143-7. Ishida H, Shirakawa K, Ohsawa T, et al. [Expression of mRNA levels of thymidylate synthase, dihydropyrimidine dehydrogenase, and orotate phosphoribosyltransferase of

86

colorectal cancer--relationships among mRNA levels in association with response to 5-FU based treatment]. Gan To Kagaku Ryoho 2005; 32:1929-34. Ishikawa Y, Kubota T, Otani Y, et al. Dihydropyrimidine dehydrogenase activity and messenger RNA level may be related to the antitumor effect of 5-fluorouracil on human tumor xenografts in nude mice. Clin Cancer Res 1999; 5:883-9. Jakob C, Liersch T, Meyer W, et al. Immunohistochemical analysis of thymidylate synthase, thymidine phosphorylase, and dihydropyrimidine dehydrogenase in rectal cancer (cUICC II/III): correlation with histopathologic tumor regression after 5-fluorouracil-based long-term neoadjuvant chemoradiotherapy. Am J Surg Pathol 2005; 29:1304-9. 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 Res 1992; 52:4306-12. Johnston PG, Lenz HJ, Leichman CG, et al. Thymidylate synthase gene and protein expression correlate and are associated with response to 5-fluorouracil in human colorectal and gastric tumors. Cancer Res 1995; 55:1407-12. Johnston PG, Benson AB, 3rd, Catalano P, Rao MS, O'Dwyer PJ, Allegra CJ. Thymidylate synthase protein expression in primary colorectal cancer: lack of correlation with outcome and response to fluorouracil in metastatic disease sites. J Clin Oncol 2003; 21:815-9. Jung R, Soondrum K, Neumaier M. Quantitative PCR. Clin Chem Lab Med 2000; 38:833-6. Kasbekar DK, Greenberg DM. Studies on Tumor Resistance to 5-Fluorouracil. Cancer Res 1963; 23:818-24. Kawakami K, Watanabe G. Identification and functional analysis of single nucleotide polymorphism in the tandem repeat sequence of thymidylate synthase gene. Cancer Res 2003; 63:6004-7. Khabar KS, Siddiqui YM, al-Zoqhaibi F, et al. RNase L mediates transient control of the interferon response through modulation of the double-stranded RNA-dependent protein kinase PKR. J Biol Chem 2003; 278(22):20124-20132 Kessel D, Hall TC. Studies on drug transport by normal human leukocytes. Biochem Pharmacol 1967; 16:2395-403. Kinoshita M, Kodera Y, Hibi K, et al. Gene expression profile of 5-fluorouracil metabolic enzymes in primary colorectal cancer: potential as predictive parameters for response to fluorouracil-based chemotherapy. Anticancer Res 2007; 27:851-6.

87

Kizaki H, Williams JC, Morris HP, Weber G. Increased cytidine 5'-triphosphate synthetase activity in rat and human tumors. Cancer Res 1980; 40:3921-7. Kracmarova A, Cermak J, Brdicka R, Bruchova H. High expression of ERCC1, FLT1, NME4 and PCNA associated with poor prognosis and advanced stages in myelodysplastic syndrome. Leuk Lymphoma 2008; 49:1297-305. Leichman L, Lenz HJ, Leichman CG, et al. Quantitation of intratumoral thymidylate synthase expression predicts for resistance to protracted infusion of 5-fluorouracil and weekly leucovorin in disseminated colorectal cancers: preliminary report from an ongoing trial. Eur J Cancer 1995; 31A:1306-10. Liang JT, Huang KC, Lai HS, et al. High-frequency microsatellite instability predicts better chemosensitivity to high-dose 5-fluorouracil plus leucovorin chemotherapy for stage IV sporadic colorectal cancer after palliative bowel resection. Int J Cancer 2002; 101:519-25. Livak KJ, Flood SJ, Marmaro J, Giusti W, Deetz K. Oligonucleotides with fluorescent dyes at opposite ends provide a quenched probe system useful for detecting PCR product and nucleic acid hybridization. PCR Methods Appl 1995; 4:357-62. Longley DB, Boyer J, Allen WL, et al. The role of thymidylate synthase induction in modulating p53-regulated gene expression in response to 5-fluorouracil and antifolates. Cancer Res 2002; 62:2644-9. Longley DB, Harkin DP, Johnston PG. 5-fluorouracil: mechanisms of action and clinical strategies. Nat Rev Cancer 2003; 3:330-8. Lowe SW, Ruley HE, Jacks T, Housman DE. p53-dependent apoptosis modulates the cytotoxicity of anticancer agents. Cell 1993; 74:957-67. Lu B, Xu J, Zhu Y, Zhang H, Lai M. Systemic analysis of the differential gene expression profile in a colonic adenoma-normal SSH library. Clin Chim Acta 2007; 378:42-7. Lulko AT, Buist G, Kok J, Kuipers OP. Transcriptome analysis of temporal regulation of carbon metabolism by CcpA in Bacillus subtilis reveals additional target genes. J Mol Microbiol Biotechnol 2007; 12:82-95. Mader RM, Sieder AE, Braun J, et al. Transcription and activity of 5-fluorouracil converting enzymes in fluoropyrimidine resistance in colon cancer in vitro. Biochem Pharmacol 1997; 54:1233-42. Mariadason JM, Arango D, Shi Q, et al. Gene expression profiling-based prediction of response of colon carcinoma cells to 5-fluorouracil and camptothecin. Cancer Res 2003; 63:8791-812.

88

Mariadason JM, Arango D, Augenlicht LH. Customizing chemotherapy for colon cancer: the potential of gene expression profiling. Drug Resist Updat 2004; 7:209-18. Martin SE, Caplen NJ. Applications of RNA interference in mammalian systems. Annu Rev Genomics Hum Genet 2007; 8:81-108.Martinez I, Lombardia L, Garcia-Barreno B, Dominguez O, Melero JA. Distinct gene subsets are induced at different time points after human respiratory syncytial virus infection of A549 cells. J Gen Virol 2007; 88:570-81. Mata JF, Garcia-Manteiga JM, Lostao MP, et al. Role of the human concentrative nucleoside transporter (hCNT1) in the cytotoxic action of 5[Prime]-deoxy-5-fluorouridine, an active intermediate metabolite of capecitabine, a novel oral anticancer drug. Mol Pharmacol 2001; 59:1542-8. Matsuyama R, Togo S, Shimizu D, et al. Predicting 5-fluorouracil chemosensitivity of liver metastases from colorectal cancer using primary tumor specimens: three-gene expression model predicts clinical response. Int J Cancer 2006; 119:406-13. Mauro DJ, De Riel JK, Tallarida RJ, Sirover MA. Mechanisms of excision of 5-fluorouracil by uracil DNA glycosylase in normal human cells. Mol Pharmacol 1993; 43:854-7. 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. Metzger R, Danenberg K, Leichman CG, et al. High basal level gene expression of thymidine phosphorylase (platelet-derived endothelial cell growth factor) in colorectal tumors is associated with nonresponse to 5-fluorouracil. Clin Cancer Res 1998; 4:2371-6. Midgley R, Kerr D. Colorectal cancer. Lancet 1999; 353:391-9.

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. Moertel CG, Fleming TR, Macdonald JS, et al. Levamisole and fluorouracil for adjuvant therapy of resected colon carcinoma. N Engl J Med 1990; 322:352-8. Moertel CG, Fleming TR, Macdonald JS, et al. Fluorouracil plus levamisole as effective adjuvant therapy after resection of stage III colon carcinoma: a final report. Ann Intern Med 1995; 122:321-6. Nykanen A, Haley B, Zamore PD. ATP requirements and small interfering RNA structure in the RNA interference pathway. Cell 2001; 107:309-21.

89

Ochiai T, Nishimura K, Noguchi H, et al. Prognostic impact of orotate phosphoribosyl transferase activity in resectable colorectal cancers treated by 5-fluorouracil-based adjuvant chemotherapy. J Surg Oncol 2006; 94:45-50. Oguri T, Bessho Y, Achiwa H, et al. MRP8/ABCC11 directly confers resistance to 5-fluorouracil. Mol Cancer Ther 2007; 6:122-7. Okumura K, Shiomi H, Mekata E, et al. Correlation between chemosensitivity and mRNA expression level of 5-fluorouracil-related metabolic enzymes during liver metastasis of colorectal cancer. Oncol Rep 2006; 15:875-82. Ooyama A, Takechi T, Toda E, et al. Gene expression analysis using human cancer xenografts to identify novel predictive marker genes for the efficacy of 5-fluorouracil-based drugs. Cancer Sci 2006; 97:510-22. Pastor-Anglada M, Felipe A, Casado FJ. Transport and mode of action of nucleoside derivatives used in chemical and antiviral therapies. Trends Pharmacol Sci 1998; 19:424-30. Popat S, Chen Z, Zhao D, et al. A prospective, blinded analysis of thymidylate synthase and p53 expression as prognostic markers in the adjuvant treatment of colorectal cancer. Ann Oncol 2006; 17:1810-7. Pratt S, Shepard RL, Kandasamy RA, Johnston PA, Perry W, 3rd, Dantzig AH. The multidrug resistance protein 5 (ABCC5) confers resistance to 5-fluorouracil and transports its monophosphorylated metabolites. Mol Cancer Ther 2005; 4:855-63. Punt CJ, Blanke CD, Zhang J, Hammershaimb L. Integrated analysis of overall survival in two randomised studies comparing 5-fluorouracil/leucovorin with or without trimetrexate in advanced colorectal cancer. Ann Oncol 2002; 13:92-4. Reichard P, Skold O, Klein G, Revesz L, Magnusson PH. Studies on resistance against 5-fluorouracil. I. Enzymes of the uracil pathway during development of resistance. Cancer Res 1962; 22:235-43. Reyes P, Hall TC. Synthesis of 5-fluorouridine 5'-phosphate by a pyrimidine phosphoribosyltransferase of mammalian origin. II. Correlation between the tumor levels of the enzyme and the 5-fluorouracil-promoted increase in survival of tumor-bearing mice. Biochem Pharmacol 1969; 18:2587-90. Reynolds NA, Wagstaff AJ. Cetuximab: in the treatment of metastatic colorectal cancer. Drugs 2004; 64:109-18; discussion 19-21. Ritzel MW, Ng AM, Yao SY, et al. Molecular identification and characterization of novel human and mouse concentrative Na+-nucleoside cotransporter proteins (hCNT3 and mCNT3) broadly selective for purine and pyrimidine nucleosides (system cib). J Biol Chem 2001; 276:2914-27.

90

Roden DM, George AL, Jr. The genetic basis of variability in drug responses. Nat Rev Drug Discov 2002; 1:37-44. Sakamoto E, Nagase H, Kobunai T, et al. Orotate phosphoribosyltransferase expression level in tumors is a potential determinant of the efficacy of 5-fluorouracil. Biochem Biophys Res Commun 2007; 363:216-22. Salonga D, Danenberg KD, Johnson M, 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. Saltz LB, Cox JV, Blanke C, et al. Irinotecan plus fluorouracil and leucovorin for metastatic colorectal cancer. Irinotecan Study Group. N Engl J Med 2000; 343:905-14. Scherf U, Ross DT, Waltham M, et al. A gene expression database for the molecular pharmacology of cancer. Nat Genet 2000; 24:236-44. Schwartz EL, Baptiste N, Wadler S, Makower D. Thymidine phosphorylase mediates the sensitivity of human colon carcinoma cells to 5-fluorouracil. J Biol Chem 1995; 270:19073-7. Shimizu D, Ishikawa T, Ichikawa Y, et al. Prediction of chemosensitivity of colorectal cancer to 5-fluorouracil by gene expression profiling with cDNA microarrays. Int J Oncol 2005; 27:371-6. Shin YK, Yoo BC, Chang HJ, et al. Down-regulation of mitochondrial F1F0-ATP synthase in human colon cancer cells with induced 5-fluorouracil resistance. Cancer Res 2005; 65:3162-70. Singapore Cancer Registry Report No. 6 Skold O. Enzymic ribosidation and ribotidation of 5-fluorouracil by extracts of the Ehrlich-ascites tumor. Biochim Biophys Acta 1958; 29:651. Sommer H, Santi DV. Purification and amino acid analysis of an active site peptide from thymidylate synthetase containing covalently bound 5-fluoro-2'-deoxyuridylate and methylenetetrahydrofolate. Biochem Biophys Res Commun 1974; 57:689-95. Soong R, Diasio RB. Advances and challenges in fluoropyrimidine pharmacogenomics and pharmacogenetics. Pharmacogenomics 2005; 6:835-47. Soong R, Shah N, Salto-Tellez M, et al. Prognostic significance of thymidylate synthase, dihydropyrimidine dehydrogenase and thymidine phosphorylase protein expression in colorectal cancer patients treated with or without 5-fluorouracil-based chemotherapy. Ann Oncol 2008; 19:915-9.

91

Staunton JE, Slonim DK, Coller HA, et al. Chemosensitivity prediction by transcriptional profiling. Proc Natl Acad Sci U S A 2001; 98:10787-92. Steg A, Wang W, Blanquicett C, et al. Multiple gene expression analyses in paraffin-embedded tissues by TaqMan low-density array: Application to hedgehog and Wnt pathway analysis in ovarian endometrioid adenocarcinoma. J Mol Diagn 2006; 8:76-83. Tai IT, Dai M, Owen DA, Chen LB. Genome-wide expression analysis of therapy-resistant tumors reveals SPARC as a novel target for cancer therapy. J Clin Invest 2005; 115:1492-502. Tammur J, Prades C, Arnould I, et al. Two new genes from the human ATP-binding cassette transporter superfamily, ABCC11 and ABCC12, tandemly duplicated on chromosome 16q12. Gene 2001; 273:89-96. Tashiro T, Inaba M, Kobayashi T, et al. Responsiveness of human lung cancer/nude mouse to antitumor agents in a model using clinically equivalent doses. Cancer Chemother Pharmacol 1989; 24:187-92. Thirion P, Piedbois P, Buyse M, et al. Alpha-interferon does not increase the efficacy of 5-fluorouracil in advanced colorectal cancer. Br J Cancer 2001; 84:611-20. Thirion P, Michiels S, Pignon JP, et al. Modulation of fluorouracil by leucovorin in patients with advanced colorectal cancer: an updated meta-analysis. J Clin Oncol 2004; 22:3766-75. Tuchman M, Stoeckeler JS, Kiang DT, O'Dea RF, Ramnaraine ML, Mirkin BL. Familial pyrimidinemia and pyrimidinuria associated with severe fluorouracil toxicity. N Engl J Med 1985; 313:245-9. Ulrich CM, Bigler J, Velicer CM, Greene EA, Farin FM, Potter JD. Searching expressed sequence tag databases: discovery and confirmation of a common polymorphism in the thymidylate synthase gene. Cancer Epidemiol Biomarkers Prev 2000; 9:1381-5. Van Cutsem E, Hoff PM, Harper P, et al. Oral capecitabine vs intravenous 5-fluorouracil and leucovorin: integrated efficacy data and novel analyses from two large, randomised, phase III trials. Br J Cancer 2004; 90:1190-7. Van Cutsem E, Nowacki M, Lang I et al. Randomized phase III trial of irinotecan and 5-FU/FA with or without cetuximab in the first-line treatment of patients with metastatic colorectal cancer: the CRYSTAL trial, ASCO Annual Meeting 2007 http://www.asco.org abstract 4000. van der Wilt CL, Smid K, Veerman G, Peters GJ. The role of thymidine kinase activity in murine colon tumours treated with 5-fluorouracil. Adv Exp Med Biol 1998; 431:653-6. van Kuilenburg AB, Meinsma R, Vreken P, Waterham HR, van Gennip AH. Isoforms of human CTP synthetase. Adv Exp Med Biol 2000; 486:257-61.

92

Whelan J, Phear G, Yamauchi M, Meuth M. Clustered base substitutions in CTP synthetase conferring drug resistance in Chinese hamster ovary cells. Nat Genet 1993; 3:317-22. Wilhelm J, Hahn M, Pingoud A. Influence of DNA target melting behavior on real-time PCR quantification. Clin Chem 2000; 46:1738-43. Wohlhueter RM, McIvor RS, Plagemann PG. Facilitated transport of uracil and 5-fluorouracil, and permeation of orotic acid into cultured mammalian cells. J Cell Physiol 1980; 104:309-19. Wolmark N, Rockette H, Fisher B, et al. The benefit of leucovorin-modulated fluorouracil as postoperative adjuvant therapy for primary colon cancer: results from National Surgical Adjuvant Breast and Bowel Project protocol C-03. J Clin Oncol 1993; 11:1879-87. Wong NA, Brett L, Stewart M, et al. Nuclear thymidylate synthase expression, p53 expression and 5FU response in colorectal carcinoma. Br J Cancer 2001; 85:1937-43. Wong ML, Medrano JF. Real-time PCR for mRNA quantitation. Biotechniques 2005; 39:75-85. Wong CS, Wong VW, Chan CM, et al. Identification of 5-fluorouracil response proteins in colorectal carcinoma cell line SW480 by two-dimensional electrophoresis and MALDI-TOF mass spectrometry. Oncol Rep 2008; 20:89-98. Yabuuchi H, Shimizu H, Takayanagi S, Ishikawa T. Multiple splicing variants of two new human ATP-binding cassette transporters, ABCC11 and ABCC12. Biochem Biophys Res Commun 2001; 288:933-9. Yamada H, Iinuma H, Watanabe T. Prognostic value of 5-fluorouracil metabolic enzyme genes in Dukes' stage B and C colorectal cancer patients treated with oral 5-fluorouracil-based adjuvant chemotherapy. Oncol Rep 2008; 19:729-35. Yamamoto S, Kawasaki T. Active transport of 5-fluorouracil and its energy coupling in Ehrlich ascites tumor cells. J Biochem 1981; 90:635-42. Yasuno M, Mori T, Koike M, et al. Importance of thymidine phosphorylase expression in tumor stroma as a prognostic factor in patients with advanced colorectal carcinoma. Oncol Rep 2005; 13:405-12. Yoo BC, Jeon E, Hong SH, Shin YK, Chang HJ, Park JG. Metabotropic glutamate receptor 4-mediated 5-Fluorouracil resistance in a human colon cancer cell line. Clin Cancer Res 2004; 10:4176-84.

93

Yoshinare K, Kubota T, Watanabe M, et al. Gene expression in colorectal cancer and in vitro chemosensitivity to 5-fluorouracil: a study of 88 surgical specimens. Cancer Sci 2003; 94:633-8. Yu J, Shannon WD, Watson MA, McLeod HL. Gene expression profiling of the irinotecan pathway in colorectal cancer. Clin Cancer Res 2005; 11:2053-62. Yu J, Mallon MA, Zhang W, et al. DNA repair pathway profiling and microsatellite instability in colorectal cancer. Clin Cancer Res 2006; 12:5104-11. Zembutsu H, Ohnishi Y, Tsunoda T, et al. Genome-wide cDNA microarray screening to correlate gene expression profiles with sensitivity of 85 human cancer xenografts to anticancer drugs. Cancer Res 2002; 62:518-27.

94

APPENDIX This research was supported by an academic research fund from National Medical

Research Council (1123/2007), Singapore.

The data in this manuscript was presented in the Poster Discussion session of the 20th

EORTC-NCI-AACR Symposium, Switzerland, October 2008.

95

PUBLICATIONS Soong R, Shah N, Salto-Tellez M, Tai BC, Soo RA, Han HC, Ng SS, Tan WL, Zep N, Joseph D, Diasio RB, Iacopetta B. Prognostic significance of thymidylate synthase, dihydropyrimidine dehydrogenase and thymidine phosphorylase protein expression in colorectal cancer patients treated with or without 5-fluorouracil-based chemotherapy. Annals of Oncology 19: 915-919 (2008). Tan WL, Dong DF, Loh M, Soo R, Fichtner I, Wong LS, Soong R. Pathway determinants of 5-fluorouracil Sensitivity. (In submission)

pathway determinants of 5-fluorouracil sensitivity · pathway determinants of 5-fluorouracil...

Documents