5-Fluorouracil: Mechanisms of Resistance and Reversal Strategies

The purpose of this work is to review the published studies on the mechanisms of action and resistance of 5-fluorouracil. The review is divided into three main sections: mechanisms of anti-tumor action, studies of the resistance to the drug, and procedures for the identification of new genes involved in resistance with microarray techniques. The details of the induction and reversal of the drug resistance are also described.

Introduction 5-Fluorouracil (5-FU) is still a widely used anticancer drug. Since 1957, it has played an important role in the treatment of colon cancer and is used for patients with breast and other cancers, like those of the head and neck [1].
5-FU is a heterocyclic aromatic organic compound with a structure similar to that of the pyrimidine molecules of DNA and RNA; it is an analogue of uracil with a fluorine atom at the C-5 position in place of hydrogen [2]. Only one crystal structure is reported in the literature for pure 5-FU, in which the compound crystallizes with four molecules in the asymmetric unit and the molecule adopts a OPEN ACCESS hydrogen-bonded sheet structure [3,4]. Due to its structure, 5-FU interferes with nucleoside metabolism and can be incorporated into RNA and DNA, leading to cytotoxicity and cell death [5,6].
Over the past 50 years, despite its many advantages, clinical applications have been greatly limited due to drug resistance. The overall response rate for advanced colorectal cancer of 5-FU alone is still only 10-15% [7], and the combination of 5-FU with other anti-tumor drugs has merely improved the response rates to 40-50% [8]. Therefore, new strategies for therapy and resistance reversal are urgently needed. Meanwhile, understanding the mechanisms by which tumors become resistant to 5-FU is an essential step towards predicting or overcoming that resistance. Fortunately, the development of microarray techniques offers us a chance to identify new genes which have key roles in drug resistance. Now, we can move forward to investigate the mechanism of these molecules, which might contribute to clinical chemotherapy in the future.

Mechanism of action
In mammalian cells, 5-FU is converted to fluorodeoxyuridine monophosphate (FdUMP), which forms a stable complex with thymidylate synthase (TS), and thus inhibits deoxythymidine monophosphate (dTMP) production. dTMP is essential for DNA replication and repair and its depletion therefore causes cytotoxicity [9,10]. Dihydropyrimidine dehydrogenase (DPD)-mediated conversion of 5-FU to dihydrofluorouracil (DHFU) is the rate-limiting step of 5-FU catabolism in normal and tumor cells. Up to 80% of administered 5-FU is broken down by DPD in the liver [11].

TS inhibition
TS, an essential enzyme for catalyzing the biosynthesis of thymidylate, is implicated in the regulation of protein synthesis and apoptotic processes [12,13]. TS catalyzes the methylation of deoxyuridine monophosphate (dUMP) to dTMP, for which 5,10-methylenetetrahydrofolate (CH 2 THF) is the methyl donor, and finally provides with the reaction thymidylate to maintain DNA replication and repair [14]. The reaction has a seriatim binding sequence, and dUMP binds at the active site before CH 2 THF does. Then the reaction is initiated by the nucleophilic addition of the active site Cys 146 (numbering of amino acid residues used is according to the sequence of EcTS) to the pyrimidine C (6)atom of dUMP. Specifically, at the onset of catalysis, the binding position and orientation of the substrate, if correctly adopted, support an efficient binding of the cofactor, and then allow the formation of the ternary TS-dUMP-CH 2 THF complex, and the subsequent reaction [15].
Research has indicated that 5-FU exerts its anticancer effects mainly through inhibition of TS, for which the pathways have not been fully interpreted. Santi has pointed out that the formation of the ternary TS-FdUMP-CH 2 THF complex is time-dependent, and the reaction stops as the fluorine substituent fails to dissociate from the pyrimidine ring, resulting in a slowly reversible inactivation of the enzyme [16]. Reduction of dTMP leads to downstream depletion of deoxythymidine triphosphate (dTTP), which induces perturbations in the levels of the other deoxynucleotides (dATP, dGTP and dCTP). Finally the imbalances (the ATP/dTTP ratio specifically) are thought to severely disrupt DNA synthesis and repair, resulting in lethal DNA damage [17,18]. Accumulation of dUMP, which might subsequently lead to increased levels of deoxyuri-dine triphosphate (dUTP), can be misincorporated into DNA, and FdUTP, the metabolic product of 5-FU has the same action [19]. Furthermore, repairing enzyme uracil-DNA-glycosy-lase (UDG) is suggested to be useless in the presence of high (F) dUTP/dTTP ratios and only results in further false DNA repair [20].

DNA and RNA misincorporation
5-FU is a pyrimidine analogue that can be misincorporated into RNA and DNA in place of uracil or thymine. The interference with the normal biosynthesis or function of nucleic acids is therefore another possible mechanism of action for 5-FU. 5-FU can be misincorporated into DNA of drugtreated cells, and accumulation of 5-FU in the genome, rather than uracil excision, is correlated with 5-FU cytotoxicity in mammalian cells [21]. It can also be misincorporated into RNA, and evidence suggests that RNA-based effects play a significant role in its cytotoxicity. Experiments in yeast show that defects in the nuclear RNA exosome subunit Rrp6p could cause hypersensitivity to 5-FU. Genetic analyses suggest that while a DNA repair mutation, apn1-Δ causes sensitivity to 5-FU-induced DNA damage, and an rrp6-Δ mutation causes hypersensitivity, due to the RNA-based effects of 5-FU [22].
The action of 5-FU is not only inhibiting the processing of pre-rRNA into mature rRNA, but also disrupting post-transcriptional modification of tRNAs and the assembly activity of snRNA/protein complexes, thereby inhibiting splicing of pre-mRNA [28]. 5-FU-containing RNA can also inhibit pseudouridylation, the most abundant post-transcriptional modification of noncoding RNA. Results suggest that Cbf5p binds tightly to substrates containing 5-FU, causing their degradation by the TRAMP/exosome-mediated RNA surveillance pathway. And the RNA-based 5-FU toxicity requires the pseudouridylation activity of Cbf5p, and sequestration of Cbf5p to a particular guide RNA reduced Cbf5p-dependent 5-FU toxicity [29].

Assisted modulation
Leucovorin Leucovorin (LV), which enters the cell via the reduced folate carriers and is anabolized to CH 2 THF, increases the intracellular CH 2 THF pool, thereby enhancing TS inhibition by FdUMP [30]. During the procedure, CH 2 THF is polyglutamated by folylpolyglutamate synthetase. Polyglutamation not only increases the cellular retention of CH 2 THF, but also enhances the stabilization of its ternary complex with TS and FdUMP [31,32]. However, resistance to modulation of 5-FU by LV exists and decreased stability of ternary complexes appears to be the mechanism of acquired resistance to the LV modulation of fluoropyrimidine cytotoxicity, possibly due to mutation(s) of TS [33][34][35].

Interferons
Interferons (IFNs) are pleiotropic cytokines that exert negative regulatory effects on the growth of normal and malignant cells in vitro and in vivo. Studies have reported that IFNs might enhance the cytotoxicity of 5-FU in various cancer cell lines. 5-FU might induce secretion of proteolytically processed mature and degraded IL-18 species, by inducing Caspase-1 and Caspase-3 activation. Conditioned medium from 5-FU-treated induces IFN-g production by activated T cells in an IL-18dependent manner. Carbone et al. [36] suggest treatment of pancreatic cancer cells with 5-FU induces caspase-dependent processing of pro-IL18 leading to the secretion of biologically active IL-18. Results show that combination of IFN-A plus 5-FU strongly induces cell growth inhibition of human hepatocellular carcinoma cells and indicates that one of the direct mechanisms of combination therapy may in part be attributable to alterations in induction of apoptosis through IFN-A/BR [37]. Wada et al. [38] suggest IFN-alpha and 5-FU combination therapy has anti-proliferative and anti-angiogenic effects and can induce apoptosis in vivo.

Apoptosis
Enhancing the sensitization of cancer cells to drug-induced apoptosis has become an important strategy for 5-FU. 5-FU might trigger the cancer cell apoptosis by activating caspase-6. RSV synergistically promotes 5-FU-mediated apoptosis at its higher concentration irrespective of p53 [39]. The 5-FU also generates mitochondrial ROS in the p53-dependent pathway [40]. In addition, BMI-1 depletion enhances the chemosensitivity of NPC cells by inducing apoptosis; which is associated with inhibition of the PI3K/AKT pathway [41].

Cell cycle
Results suggest that 5-FU could induce changes in cell cycle regulation, which might associate with an alteration of G1 cyclins expression [42]. The in vitro 5-FU treatment of oral cancer cells results in an increase in G1/S phase cells, and p21 is remarkably up-regulated. A remarkable up-regulation of cyclin E and a concomitant down-regulation of cyclin D are observed. There is also speculation that its activity might be cancelled by increased binding to CDK4.

L-Arginine
5-FU combined with L-arginine (L-Arg) could inhibit the growth of tumor in nude mice. The effect may be related to inducing the synthesis and increasing the activity of endogenous nitric oxide (NOS). The production of NO is increased, and it can enhance the expression of apoptosis-related gene and antioncogene [43].

Others
Ooyama et al. [45] suggested that 5-FU could also induce cytotoxic activity by inhibition of angiogenesis through the induction of thrombospondin-1 (TSP-1). Hwang et al. [46] proposed that the combination of 5-FU and genistein exert a novel chemotherapeutic effect on colon cancers, and AMPK might be a novel regulatory molecule of COX-2 expression, further implying its involvement in cytotoxicity caused by genistein.

Resistance
Anti-cancer drug resistance can result from various causes including alteration of drug influx and efflux, enhancement of drug inactivation and mutation of the drug target [47]. High-level expression of TS [48], increased activity of deoxyuridine triphosphatase [49], methylation of the MLH1 gene [50], and overexpression of Bcl-2 [51], Bcl-XL [51,52], and Mcl-1 [53] proteins have all been reported to lead to resistance to 5-FU, which suggests that multiple factors might contribute to 5-FU resistance [54,55].

TS
There are several possible aspects for TS induction: decreased accumulation of activated metabolites, target-associated resistance and pharmacokinetic resistance [56]. The stability of the ternary complex is highly dependent on the availability of CH 2 THF or one of its polyglutamates [57,58]. Thymidylate can be salvaged from thymidine through the action of thymidine kinase, thereby alleviating the effects of TS deficiency. This salvage pathway represents a potential mechanism of resistance to 5-FU [59]. The binding of ligands to the TS molecule leads to dramatic changes in the conformation of the enzyme, particularly within the C-terminal domain. Stabilization of the enzyme and an increase in its intracellular level are associated with ligand binding and may be important in cellular response to TS-directed drugs. The C-terminal conformational shift is not required for ligandmediated stabilization of the enzyme TS, while the N-terminus of the TS polypeptide, which is extended in the mammalian enzyme and is disordered in crystal structures, is a primary determinant of the enzyme's half-life, and TS turnover is carried out by the 26S proteasome in a ubiquitin-independent manner [60][61][62]. Further, the penultimate amino acid Pro2, which is capable on its own of destabilizing an evolutionarily distinct TS molecule, plays an important role in governing the half-life of the enzyme [63].
TS over-expression is widely accepted as a major molecular mechanism responsible for 5-FU resistance. The stability of the ternary complex is highly dependent on the availability of CH 2 THF, and in the absence of CH 2 THF or one of its polyglutamates, FdUMP forms an unstable binary complex, which results in poor inhibition [64,65]. Disturbed folate pools and a high level of enzyme before treatment lead to intrinsic resistance [66,67]. Gene amplification of TS and mutations in the gene lead to acquired resistance [67,68]. In addition, thymidylate can be salvaged from thymidine through the action of thymidine kinase, thereby alleviating the effects of TS deficiency. This salvage pathway represents a potential mechanism of resistance to 5-FU [69]. The acute induction in TS levels following therapy with inhibitors of this enzyme is also one of the critical mechanisms of resistance to 5-FU. This mechanism is based on a novel autoregulatory feedback pathway wherein the TS protein regulates its own translational efficiency [70]. Wang et al. have elucidated that TS-independent molecular events might play a key role in 5-FU resistance [71].

DPD
Increased DPD activity and the corresponding catabolism of 5-FU may lead to 5-FU resistance. Studies show that low-DPD tumors could be sensitive to 5-FU [44], and Danenberg et al. have revealed that DPD mRNA expression in colorectal metastatic or disseminated tumors is related to the anti-tumor effect of 5-FU [112]. At least 39 different alleles in the DPYD gene encoding DPD protein have been identified, and a potential prominent mutation affecting the splicing donor consensus sequence of intron 14 has been reported, which results in the deletion of 55 amino acids in the native protein [113]. However, no clear correlation between 5-FU-associated toxicity phenotype and genotype has been established thus far [11].

DNA and RNA misincorporation
Tajima et al. [72] suggested that MMR complex hMutS alpha might be specifically recognized and bound to 5-FU-modified DNA. Evidence suggests that tumor cells with MSI (caused by defective DNA mismatch repair) are more resistant to 5-FU in culture compared with microsatellite stable cells, despite similar amounts of 5-FU misincorporation into the cell's DNA. The reaction is specific as added ATP dissociates the hMutS alpha complex from the 5-FU-modified strand. And there was greater binding between hMutS alpha and 5-FU-modified DNA compared with complementary DNA or DNA containing a C/T mismatch. Qian An et al. [21] suggested that Smug1, but not UNG (uracil-DNA glycosylase), could excise 5-FU from DNA and protect against cell killing, as a predictive biomarker of drug response and a mechanism for acquired resistance in tumors.

Anti-apoptosis
The oncogene B-cell-specific Moloney murine leukemia virus insertion site 1 (BMI-1) has been shown to be involved in the protection of cancer cells from apoptosis. The expression of phospho-AKT and the anti-apoptotic protein BCL-2 are downregulated in the cells in which BMI-1 expression is inhibited, whereas the apoptosis-inducer BAX is observed to be upregulated. Abrogation of AKT pathway by a PI3K inhibitor could not further increase the sensitivity to 5-FU in the cells with reduced BMI-1 expression [41].
Human Ring-Finger, homologous to Inhibitor of Apoptosis protein type (hRFI) has been shown to inhibit death receptor mediated apoptosis. Evidence suggests that the modulation of Bcl-2 family proteins seen in 5-FU treatment plays an important role in the anti-apoptotic function. This might also clarify hRFI overexpression and the manner in which hRFI upregulates Bcl-2 and Bcl-XL and elevates the relative ratio of Bcl-2 to Bax or to Bak during 5-FU treatment [73]. Furthermore, hRFI overexpression results in the activation of nuclear factor-jB (NF-jB). Inhibition of NF-jB might effectively reverse the resistance to apoptosis as well as the upregulation of Bcl-2 and Bcl-XL in the hRFI transfectant, indicating that the activation of NF-jB is the key mechanism [74].
Tseng et al. [75] suggest that Ras, Bcl-2, as well as Raf-1 and PI3K pathways play pivotal roles in 5-FU-induced apoptosis under Ha-ras-overexpressed condition. Aberrant levels of cyclin E and p21Cip/WAF-1 expression, as well as Cdc 2 phosphorylation at Tyrosine 15, suggest that perturbation of G1/S and G2/M transitions in cell cycle might be responsible for 5-FU triggered apoptosis. Besides, the resistance could be related to an increase in the expression of IAP survivin, which can decrease cell response to the treatment or even switch the type of death from apoptosis to another kind, making therapy less efficient [76].

Cell cycle
Cell cycle perturbation may be involved in acquired 5-FU resistance. There might be a slow down in cell cycle traverse preventing incorporation of 5-FU metabolites into DNA, providing cancer cells with sufficient time to correct the misincorporated nucleotides. The resistant cell lines show significantly lower labelling indexes and cell cycle delays in G1 and G1/S boundary and prolonged DNA synthesis time. Meanwhile, the resistant cell lines demonstrate significantly prolonged potential doubling time (Tpot). CDK2 protein, Thr-160 phosphorylated CDK2, cyclin D3 and cyclin A might be remarkably reduced in the resistant cell lines [77]. NO 5-FU is shown to inhibit NO production, indicating that 5-FU inhibits high levels of NO output in activated macrophages. High levels of NO generated by activated macrophages play an important role on anti-tumor activity in tumor immunity [78]. Therefore, 5-FU is very possible to attenuate anti-tumor activity of activated macrophages. 5-FU might prevent LPS-induced NO production via inactivation of the Akt-dependent NF-κB signal pathway. The possibility that 5-FU inhibits the NO production via its cytotoxic action is excluded. The studies show that 5-FU inhibits the expression of iNOS mRNA and protein, such as signal transduction in LPS stimulation. Moreover, 5-FU might inhibit the phosphorylation of Akt that regulates NF-κB activation as an upstream molecule through IKK activation [79], and the failure of NF-κB activation results in attenuated expression of iNOS protein and subsequent NO production [80].

Mitochondria
ATP synthase down-regulation may lead to cellular events responsible for 5-FU resistance. The studies show there is lower expression of the A subunit of mitochondrial F1F0-ATP synthase (ATP synthase) and other ATP synthase complex subunits in 5-FU-resistant cells. Thus ATP synthase activity is decreased and intracellular ATP content reduced. The ATP synthase inhibitor oligomycin A strongly antagonizes 5-FU-induced suppression of cell proliferation [81].

Oxidative stress
Cellular adaptive response to ROS is another mechanism of drug resistance to 5-FU. While acute oxidative stress triggers cell apoptosis or necrosis, persistent oxidative stress induces genomic instability and has been implicated in tumor progression and drug resistance. Studies show that tumor cells which adapt to oxidative stress by increasing manganese superoxide dismutase (MnSOD), Prx I and Bcl-2, show drug resistance to 5-FU. Romo1 siRNA treatment efficiently blocks 5-FU-induced ROS generation, demonstrating that 5-FU treatment stimulates ROS production through Romo1 induction [82].

Interferons
Induction of TSt might represent a relevant mechanism of resistance to 5-FU and this mechanism can be circumvented in the presence of IFNγ. Studies suggest IFNγ is potent in prolonging the antitumor effect of 5-FU by suppressing an acute overexpression of TSt. The triple combination therapy and combination therapy (5-FU and IFNγ) show significant growth suppression of cancer cells when compared with 5-FU [83][84][85].

Other molecules
The results indicate regulation of midkine gene expression in cancer cells (MDK) appears to modulate sensitivities to anticancer drugs. De novo expression of MDK in cancer cells not normally expressing MDK confers a multi-drug resistance, while knockdown of MDK in cancer cells that normally expressed the protein leads to chemosensitization [86]. In addition, Chu et al. [87] suggest that increased metabotropic glutamate receptor 4 (mGluR4) expression is a mechanism underlying 5-FU resistance. The ATP-binding cassette (ABC) proteins play an important role in drug resistance of 5-FU, and ABCC5 is one of the multidrug resistance (MDR) subfamily of ABC proteins [88,89]. Fanciullino et al. [90] propose that nuclear expression of rTS beta could be a novel 5-FU resistance marker in patients with primary breast cancer. And results suggest that hENT1 plays an important role in 5-FU resistance and that hENT1 mRNA levels might be a useful marker to predict 5-FU sensitivity in pancreatic cancer [91].

Gene arrays
Microarray technology, developed in recent years, has enabled analysis of pan-genomic expression profiles in cells or tissues of interest. A combination of microarray and traditional molecular technologies will enable us to functionally characterize genes related to anticancer drug resistance and identify novel molecular targets for anticancer drug development.
Genetic factors might play an important role in resistance towards 5-FU, and genetic variation in any of these may contribute to anti-tumor response. Gene expression data suggest that altered regulation of nucleotide metabolism, amino acid metabolism, cytoskeleton organization, transport, and oxygen metabolism may underlie the differential resistance to 5-FU seen in cell lines.
Several highly significant associations have been observed between genotypes and expression levels of 5-FU metabolizing genes. In a study, Nordgard et al. [97] found a SNP in codon 72 of TP53 is a key regulator of 5-FU metabolizing genes such as DHFR and MTHFR. These data suggest that three copies of the TYMS 50UTR repeat might give a treatment specific reduced survival in breast cancer patients, and that TP53 might have a direct, allele specific, role in 5-FU mediated response. Another analysis confirms that the SNP showing significant associations with drug sensitivity are concentrated in some cytogenetic regions (18p, 17p13.2, 17p12, 11q14.1, 11q11 and 11p11.12). Among these regions, 18p11.32 at the location of the thymidylate synthase gene (TYMS) is strongly associated with resistance to 5-FU-based drugs. The results suggest that amplification of the TYMS gene is associated with innate resistance, supporting the possibility that TYMS copy number might be a predictive marker of drug sensitivity to fluoropyrimidines [98].
Different stages of FU resistance of low-, intermediate-and high-resistance phenotypes are also studied [99]. Some research has also used statistics tools, like the Java program TOUCAN, to identify a consensus gene list associated with 5-FU resistance, perform an in silico comparative promoter analysis, and highlight the potential implication of some TFs in the development of chemoresistance [100].

Future perspectives
Although 5-FU and its derivatives [101,102] have been important anticancer agents and have been widely applied in patients, the overall response rate of 5-FU alone is still low. Therefore, seeking the better therapeutic strategies, increase 5-FU sensitivity and reverse the resistance to drug are the key tasks in the future.

Therapeutic strategies
Studies suggest that resistance to 5-FU could be overcome through a better control of its intratumoural activation and the use of an encapsulated formulation [90]. As drug deliverers, liposomes could congregate drugs in certain tissues, decrease poisonous effects and increase the curative effects. Lecithoid material is the most primary liposome commonly used. Since Glycoprotein C (GC) is a cerebroside containing galactose, liposome made of GC also possesses galactose. Compared with lecithoid, GC has several advantages such as stable chemical characteristics, long retention in blood, anti-oxide ability and unique directional trait. It may be developed into a new kind of medicine carriers.
Combination chemotherapy regimens, including FOLFIRI and FOLFOX, have prolonged the survival of advanced or metastatic cancer patients compared with BSC alone. Additional monoclonal antibody agents provide little additive utility at high cost. New agents, including new macromolecule agents, small molecule agents and vaccines, will be introduced in the chemotherapy against colorectal cancer. Subsequently, clinical researchers will have to consider the cost-utility of these agents using QALY [35].

Reversal of resistance
Some of the enzymes involved in the metabolic process of 5-FU, including thymidylate synthase and dihydropyrimidine dehydrogenase, have been shown to predict sensitivity to 5-FU and/or prognosis [103,104]. Based on the fact that the inhibitor of DPD has been shown to enhance the anticancer effects of 5-FU [105], DPD-inhibitory fluoropyrimidines (DIF) have been developed to enhance anticancer effects. To date, DIFs have played a major role in neoadjuvant and/or adjuvant chemotherapy for patients with advanced gastric carcinoma [106]. And the measurement of orotate phosphoribosyltransferase (OPRT) in cancer tissue may be useful for the prediction and monitoring of the anticancer effects of S-1-based anticancer chemotherapy for patients with advanced cancers [107].
YSV could effectively reverse MDR associated with the down-regulation of MDR1, MRP1 and LRP expression, as well as the inhibition of P-gp function [108]. Studies also show that the Smug1, but not the UNG, excises FU from DNA and protects against cell killing [21]. Wild-type p53 gene has a remarkable reversal activity for the high expression of MDR1 gene in colorectal cancers. The reversal effects seem to be in a time dependent manner [109]. Zhu et al. [110] suggest knockdown of Bcl-XL protein levels by small interfering RNA (siRNA) inhibits the proliferation more effectively in 5-FUresistant cells than in 5-FU-sensitive cells, and down-regulation of Bcl-XL protein expression might provide a new treatment strategy for human 5-FU-resistant colon cancer therapy. Besides, they also [111] suggest that colon cancer cells resistant to tumor necrosis factor-related apoptosis-inducing ligand (TRAIL) could be resensitized by a combination therapy of TRAIL plus 5-FU.

Conclusions
As described above, there are two main anti-tumor mechanisms of 5-FU which have been proven. 5-FU is an effective chemotherapeutic drug developed as an inhibitor of TS, which leads to a thymineless cell death. And it is also a pyrimidine analogue misincorporated into RNA and DNA in place of uracil or thymine. However, its clinical application is greatly limited due to drug resistance, which could result from various causes, including alteration of drug influx and efflux, enhancement of drug inactivation and mutations of the drug target. Surely there could be still many mechanisms of 5-FU anti-tumor action and drug resistance, which have not been demonstrated yet. The new technologies, such as microarrays, might enable us to functionally characterize genes related to these mechanisms more effectually.