Preventing Colitis-Associated Colon Cancer With Antioxidants: A Systematic Review

Inflammatory bowel disease (IBD) patients have an increased risk of developing colitis-associated colon cancer (CAC); however, the basis for inflammation-induced genetic damage requisite for neoplasia is unclear. Several studies have shown that IBD patients have signs of increased oxidative damage, which could be a result of genetic and environmental factors such as an excess in oxidant molecules released during chronic inflammation, mitochondrial dysfunction, a failure in antioxidant capacity, or oxidant promoting diets. It has been suggested that chronic oxidative environment in the intestine leads to the DNA lesions that precipitate colon carcinogenesis in IBD patients. Indeed, several preclinical and clinical studies show that different endogenous and exogenous antioxidant molecules are effective at reducing oxidation in the intestine. However, most clinical studies have focused on the short-term effects of antioxidants in IBD patients but not in CAC. This review article examines the role of oxidative DNA damage as a possible precipitating event in CAC in the context of chronic intestinal inflammation and the potential role of exogenous antioxidants to prevent these cancers.

I nflammatory bowel disease (IBD) is a multifactorial chronic inflammatory disorder associated with dysregulation in the interaction between the host's immune system and the environment within the gastrointestinal tract. Chronic inflammation of the intestinal epithelium is positively associated with cancer development, and although there are several mechanisms by which inflammation could induce epithelial damage, only a few of those point to a direct source of the DNA lesions necessary for cellular transformation and cancer initiation. The oxidant environment created by activated inflammatory cells in the intestinal epithelium has been associated with carcinogenesis. [1][2][3][4] Although linked to cancer initiation, reactive oxygen species (ROS) also act as signalling molecules that regulate multiple signaling pathways associated with mitogenesis, immune and stress response, and autophagy and are therefore required to maintain homeostasis. 5 In this review, we will discuss the evidence that supports the notion that chronic intestinal oxidation is one of the main factors leading to DNA lesions that promote carcinogenesis in IBD patients ( Figure 1). This review will also discuss how antioxidants could be used to suppress tumor development within the inflamed intestinal tissue.
What Is the Source of Oxidative Molecules and DNA Lesions in the Intestinal Epithelium?
There are several sources of oxidative molecules in the intestinal epithelium. Classically activated macrophages, infiltrating neutrophils, and intestinal epithelial cells are all equipped with enzymes that produce ROS and reactive nitrogen intermediaries (RNI) in response to the gut microbiota, specific gut pathogens, or other stimuli. ROS and RNI are normally produced to keep microbes in line and maintain homeostasis within the intestine. For example, the nicotinamide adenine dinucleotide phosphate oxidase (NOX) 2, which produces superoxide (O 2 -), is expressed in macrophages, dendritic cells, and neutrophils that infiltrate the lamina propria. 6 Intestinal epithelial cells express the O 2 producing enzyme NOX1 and the hydrogen peroxide (H 2 O 2 ) producing enzyme dual oxidase 2 (DUOX2). Usually O 2 is further converted into the more stable molecule H 2 O 2 by the enzyme superoxide dismutase (SOD). In phagocytes, H 2 O 2 is further transformed into hypochlorous acid, which has antimicrobial properties. In addition, macrophages, epithelial cells, and neutrophils also express the enzyme inducible nitric oxide synthase, which produces the highly diffusible molecule NO . 6,7 Although ROS is necessary for various cellular functions, [8][9][10][11][12] excessive accumulation causes damage to biological molecules. Consequently, tissues are armed with several defense mechanisms including an intricate antioxidant defense system. The antioxidant defense system primarily functions through (1) limiting the excessive production of ROS/RNI, (2) scavenging free radicals, and (3) converting toxic free radicals into less toxic molecules. 13 An incompetent or dysregulated antioxidant system is associated with inflammatory diseases. Accordingly, mice deficient in nuclear factor erythroid-2-related factor 2, a master regulator of antioxidant responses in tissue through transcriptional regulation of antioxidant genes, develop colitis and colitis-associated cancer (CAC). 14 During chronic inflammation, ROS and RNI surpass the physiological antioxidant detoxifying capacity of cells, leading to the generation of high amounts of oxidant molecules. H 2 O 2 can react with transition metals via Fenton reaction and produce the highly reactive hydroxyl radical (HO ). 7 On the other hand, NO can react with O 2 to generate the highly reactive molecule peroxynitrite (ONOO -). 15 When these oxidant molecules are not neutralized by intracellular antioxidants, these agents induce cell membrane damage and cancer-causing DNA lesions. 1,3,16 In addition, inflamed colons in IBD patients and mice with colitis have decreased expression of antioxidant enzymes such as glutathione-Stransferase theta 1 (GSTT1). GSTT1 is not only necessary for the detoxification of oxidative free radicals, but it is also necessary to induce goblet cell differentiation and mucin production in response to triggers such as interleukin 22 and H 2 O 2 . 17 Because mucus produced by goblet cells is the first barrier that protects intestinal cells from the microbiome, this Figure  1. Unrepaired oxidative DNA damage can lead to cancerpromoting accumulation of mutations. Several stimuli from both endogenous and environmental sources induce increased production of ROS and RNI that can directly produce mutagenic DNA lesions.
Oxidative DNA damage is primarily repaired by MMR repair and base excision repair pathways involving the excision of modified bases followed by repairing of the gaps. Oxidative DNA damage could exceed the repair capacity of these DNA repair pathways, or deficiency in these specific DNA repair pathways can lead to the accumulation of oncogenic mutations, precipitating cellular transformation and ultimately tumor initiation and/or progression. study suggests that GSTT1 deficiency would disrupt the epithelial barrier, generating a positive loop for oxidative epithelial damage. 17

Inflammation, Diets, and Oxidative Environment
Several innate immune cells produce O 2 -, H 2 O 2 , and NO as a defense mechanism. However, some studies suggest that an increase in oxidative molecules can precede inflammation. For example, an inhibitor of the DNA repair enzyme OGG1 (8-oxoguanine DNA glycosylase 1) has been shown to prevent proinflammatory gene expression and cell recruitment in mouse lungs, 18 suggesting that repair of oxidative DNA damage induces inflammation. Other studies showed that mitochondrial ROS (mtROS), generated when electrons leak from Complex I and III during oxidative phosphorylation and react with oxygen to form O 2 -, can act as a signal-transducing molecule that either activates the NLRP3 inflammasome inducing the production of proinflammatory cytokines 19,20 or mediates an increased mitogen-activated protein kinase signaling that induces inflammatory cytokine production after TLR4 activation. 21 Furthermore, gene transfer of the mitochondrial antioxidant enzyme GSTT1 into the colon of mice confers protection against colitis, 17 suggesting that in some cases, an increase in mtROS could precede chronic intestinal inflammation. However, mtROS can have anti-inflammatory properties as well because it can protect the intestine from inflammation by inducing polarization of alternatively activated macrophages and a reduction in the production of proinflammatory cytokines. 22 Overall, these data suggest that mtROS must surpass a physiological threshold to lead to activation of inflammatory pathways in the intestine.
Another factor that can induce an oxidative environment that results in intestinal permeability and inflammation are high-fat diets (HFDs). 23,24 Non-esterified long chain saturated fatty acids present in HFD can increase expression of Nos2 and endoplasmic reticulum stress in goblet cells, which in turn triggers a reduction in the production of the mucus barrier and creates a positive feedback loop for inflammation. 23 In addition, mice fed a HFD had decreased expression of tight junctions proteins and MUC2 and increased expression of the enzymes NOX1, NOX4, and NOS2, which produce O 2 -, H 2 O 2, and NO, respectively. 24 However, none of these molecules or their oxidative effects were directly measured. Although the mechanism by which HFD induces the expression of these ROS-producing enzymes is not clear, it is possible that intestinal permeability associated with deficiencies in tight junctions and/or reduced mucus layer lead to the penetration of molecules such as lipopolysaccharide through the epithelial layer of the gastrointestinal tract. This in turn can lead to the induction of ROSproducing enzymes in immune cells. Interestingly, flavonoid anthocyanins could revert HFD-induced intestinal permeabilization and endotoxemia in part by modulating NOX expression and preventing the production of RNI. 24 This finding supports the idea that HFD-induced oxidative environment can lead to a positive feedback loop for intestinal permeability and intestinal oxidation. Anthocyanins also showed a promising anti-inflammatory potential in a small trial in ulcerative colitis (UC) patients, 21,25 and because HFD is a risk factor for IBD, 26 it is possible that anthocyanins could counteract the initial ROS-related processes that precede chronic intestinal inflammation.
Overall, it seems that the source of oxidative damage in the intestine could be dependent or independent of inflammatory cells. Mitochondrial dysfunction and certain diets can mediate an increase in the production of O 2 and NO in the intestinal epithelium, 20,21,[23][24][25]27 which in turn can activate inflammatory pathways and may lead to mutations that perpetuate inflammation and/or initiate cancer. However, in the majority of cases, it is likely that inflammation is itself caused by a response to microbial stimuli that causes an oxidative environment that can precipitate cancer.
How Is Oxidative DNA Damage Repaired?
HO and ONOOdirectly damage DNA via strand breakage or nucleotide oxidation. Guanine is the nucleotide with the highest oxidation potential, 15 and oxidized guanine is commonly used to detect oxidative DNA damage. Elevated levels of oxidized guanine indicate that the oxidized environment has superseded the capacity of the cell to repair lesions caused by oxidation and therefore indicates potential oxidative damage.
Guanine oxidation leads to the formation of 8-hydroxy-2 0deoyguanosine or 8-oxo-7,8-dihydro-2 0 -deoxyguanosine (8-oxoG), which can lead to mutations if not repaired ( Figure 2). If 8-oxoG is not removed from the DNA before replication and because 8-oxoG preferentially pairs with an adenine, unrepaired 8-oxoG will ultimately lead to G / T and C / A transversion mutations ( Figure 2). Multiple DNA repair systems repair oxidative DNA lesions at different stages of the cell cycle. 28-31 C:8-oxoG pairs in DNA are recognized and repaired by the base excision repair and the nucleotide excision repair systems. 29,31 In addition, 8-oxoG nucleotides from the dNTP pool are usually removed by MTH1, and failure to do so can lead to incorporation of this oxidized base into the nascent DNA strand during DNA replication ( Figure 2D). [29][30][31] The mismatch repair (MMR) system has an especially important role in the repair of 8-oxoG lesions in highly proliferative tissues such as the intestinal epithelium because MMR-deficient human and mouse colonic tissue have exceptionally high levels of this DNA lesion. 3 In the next sections, we discuss the role of endogenous and exogenous antioxidants on IBD and how these agents might prevent cellular transformation in inflamed intestinal tissue.

Examining the Roles of Endogenous Antioxidants in IBD and CAC
The endogenous antioxidant defense system comprises both enzymatic antioxidants such as SOD, glutathione peroxidase (GPx), catalase (CAT), peroxiredoxin (PRDX), and thioredoxin as well as nonenzymatic antioxidants such as glutathione, alpha-lipoic acid, uric acid, melatonin, bilirubin, and ferritin. Multiple studies have found altered expression and/or activities of these antioxidant proteins in IBD and CAC patients, which suggest a role in disease pathology. SOD 32 However, CAT activity remained permanently inhibited and was independent of disease activity. 32 In a different study, a reduction in CAT or total SOD activity was found to be associated with increased risk of colorectal cancer (CRC) and gastric adenocarcinoma, respectively. 33 GPx, PRDX, and thioredoxin, the thiol-dependent proteins that catalyze the reduction of H 2 O 2 , lipid peroxides, and peroxynitrite, are found to be up-regulated in colonic mucosa of IBD and CRC patients compared with healthy subjects. [34][35][36][37][38][39][40] The Gpx isoforms Gpx1 to Gpx4 are expressed in healthy gastrointestinal mucosa; however, their However, if this mismatch is left unrepaired, a second round of replication will lead to C:G/A:T transversion mutation in one daughter cell. (D) Deoxyguanosine triphosphate (dGTP) in the nucleotide pool can be oxidized and incorporated into the nascent DNA strand opposite an A during replication, which can be repaired by MMR. However, A:8-oxo-G mispairs also can be processed through inappropriate MUTYH-initiated base excision repair, leading to the formation of C:8-oxo-G pairs, which could be further repaired by OGG1, generating an A:T/C:G transversion mutation. To avoid this, it is believed that cells avoid MUTYH activity during replication, giving preference to the MMR system. Therefore, lack of MMR activity is particularly an issue for highly proliferative tissues under oxidative environment such as gastrointestinal tract.
deficiency or genetic variants that affect their functions have been associated with increased mucosal damage and risk of developing CRC. [39][40][41] Mice deficient in Gpx1, Gpx2, or Gpx3 develop normally but are susceptible to develop IBD and CAC upon Salmonella infection or azoxymethane (AOM)/ dextran sodium sulfate (DSS) treatment. 42,43 Compartmentalized expression of Gpx1 and Gpx2 is reported along the crypt-villus axis in the intestine; Gpx1 expression is predominantly found in the villus, whereas Gpx2 is localized mainly in the crypt. 44 Gpx2 knockout mice have increased apoptosis and mitosis of crypt cells under selenium restriction. However, upon selenium supplementation in Gpx2 -/mice, Gpx1 expression in the crypt increases, partially compensating for Gpx2 loss of expression and protecting from crypt cell apoptosis. This result suggests an overlapping complementary role of Gpx1 and Gpx2 in maintaining intestinal homeostasis. 44 Accordingly, Gpx1 and Gpx2 double-knockout mice develop spontaneous colitis, dependent on excessive production of ROS by NOX1 and DUOX2. [45][46][47] PRDXs are highly reactive peroxidases that account for the reduction of more than 90% of total cellular peroxides, while also crucial for maintaining physiological levels of cellular peroxides for vital cellular functions. 48 All mammalian PRDXs, PRDX1-6, are overexpressed in the mucosa of active colitis and CRC patients, and their level in mucosa is positively correlated with disease severity and cancer metastasis. [34][35][36][37]49,50 Increased expression of PRDXs in diseased mucosa seems to be a host antioxidant defense response because PRDX4 -/mice have higher disease severity and endoplasmic reticulum stress after DSS treatment. 51 Studies suggest a dual role of PRDXs in cancer. PRDXs can either inhibit ROS-induced DNA damage and carcinogenesis or potentiate cancer progression through inhibition of ROS-mediated cell death in cancerous tissues. 49,50 However, their role has not been investigated in CAC.
Researchers have examined nonenzymatic endogenous antioxidants and their roles in IBD and CAC. Some of these antioxidants such as glutathione, bilirubin, and uric acid are produced during normal metabolism, whereas melatonin is a hormone that is secreted from enterochromaffin cells in the intestine. Glutathione, the most important intracellular nonenzymatic antioxidant, is the substrate for glutathione-Stransferase (GST) that catalyzes the step of reduced glutathione conjugation with reactive electrophiles in the reduction of peroxides by glutathione peroxidase. The cellular level of glutathione was found to be reduced in intestinal mucosa of IBD patients, whereas reduced mucosal expression and/or activity of GST was also observed in IBD and CAC patients. 52,53 Moreover, the serum levels of bilirubin, uric acid, and melatonin are found to be negatively associated with disease severity in IBD patients. [54][55][56] The above findings not only imply the pathophysiological role of these endogenous antioxidants in IBD and CAC development but also suggest that the altered regulation of their expression in a diseased state is a compensatory response of the host to limit disease progression and severity.

Testing Endogenous Antioxidants in IBD and CAC
Because epidemiologic studies have recognized the association of endogenous antioxidants with IBD and CAC pathophysiology, investigators have tested their therapeutic potential in these diseases. However, the short life span of recombinant enzymatic antioxidants in the gastrointestinal tract remains a barrier for therapeutic evaluation in intestinal diseases. Accordingly, attempts have been made to produce either stable proteins using genetic engineering or transgenic probiotic strains. Supplementation of genetically engineered Lactobacillus fermentum expressing recombinant SOD, hyperthermostable SOD from Thermus thermophilus HB27, or SOD mimics having enhanced stability and activity and ameliorated colitis severity in both mouse and human models. [57][58][59][60] In addition, treatment with genetically engineered Lactobacillus casei BL23 or Streptococcus thermophilus CRL807 expressing recombinant CAT or SOD restored endogenous antioxidant pools and reduced disease severity in colitis models. 61,62 Remarkably, Ishihara et al 57 observed a bell-shaped dose-response of SOD in colitis, demonstrating that a protective effect of SOD at lower doses is through reduction in colonic ROS level and ineffectiveness at higher doses is due to accumulation of H 2 O 2 . Accordingly, simultaneous administration of CAT restored the protective effect at higher doses of SOD. 57 Recently, multiple studies demonstrate that boosting colonic H 2 O 2 with probiotics can improve mucosal barrier integrity, increase colonization resistance, and suppress inflammatory responses in the colon, whereas exceeding the physiological levels of H 2 O 2 could be detrimental. 10 Transgenic overexpression of another enzymatic antioxidant thioredoxin in mice led to reduced levels of tumor necrosis factor-a and interferon-g upon DSS treatment compared with controls, suggesting an anti-inflammatory action of this enzyme. 63 Accordingly, administration of recombinant human thioredoxin significantly ameliorated DSS-induced colitis and colonic inflammation in interleukin 10 KO mice. 63 Thioredoxin can modulate the DNA binding properties of multiple transcriptional factors to regulate expression of inflammatory mediators. 63,64 Overall, these studies report that restoring mucosal enzymatic antioxidants could be protective in IBD.
Nonenzymatic antioxidants as a therapeutic intervention in IBD have also been evaluated. Because IBD patients are depleted of glutathione in their gastrointestinal tracts, Ardite et al 65 found that treating colitis-induced mice with glutathione attenuated acute colitis. In addition, ectopic expression of GSTTT1 1 in DSS-treated mice attenuated colitis severity via interleukin 22-dependent restoration of epithelial cell functions. 17 Similar to glutathione, supplementing another thiol-containing endogenous antioxidant alpha-lipoic acid also reduced colitis and ileitis in animal models. 66,67 Melatonin is another nonenzymatic compound recognized to have enteroprotective activity through its antioxidant and anti-inflammatory action. 68 Exogenous administration of melatonin in experimental colitis models improved the disease pathology by reducing inflammation and epithelial damage. [69][70][71] Melatonin also reduced the levels of oxidative DNA damage in colonic mucosa of IBD patients 72 ; however, the effects on CAC were not evaluated.
Collectively, these studies suggest the therapeutic potential of endogenous antioxidants in IBD. However, few have evaluated the role of these agents in CAC in both preclinical models and in the clinic.

Exogenous Antioxidants as Therapy in IBD and CAC
The current pathophysiological understanding of chronic inflammatory diseases and their association with endogenous antioxidants has encouraged researchers to develop therapeutics for IBD and CAC by using exogenous antioxidants. Exogenous antioxidants are substances that our body cannot produce and therefore must be provided as supplements from natural or synthetic sources. Synthetic antioxidants include compounds with antioxidant activities, precursors or mimics of endogenous antioxidants, and derivatives of amino acids such as propionyl-L-carnitine. Natural antioxidants consist of vitamins, polyphenolic compounds, polyunsaturated fatty acids, and trace metals. Numerous exogenous antioxidants have been investigated for their therapeutic potential in IBD, CAC, and CRC with promising results in preclinical models. However, most of the clinical trials do not validate the preclinical findings. Antioxidant supplementation was generally found to be ineffective or detrimental for cancer in most of the clinical studies [73][74][75] (Tables 1 and 2), although some of the compounds used as antioxidants in these clinical studies do not have strict ROS-specific effects. Excessive ROS promotes mutagenesis through oxidative DNA damage and can trigger cancer development. However, it also has inhibitory roles on cancer progression through oxidation-induced cytotoxicity in cancer cells. Cytoplasmic ROS levels are significantly higher in cancer cells because of their increased metabolic activity compared with normal cells. 76,77 To cope with oxidative damage-induced cytotoxicity, cancer cells depend on various mechanisms including an increase in their antioxidant pool for their survival. Thus, the concept of a negative correlation between antioxidant levels and cancer initiation/progression is now not universally valid, 77 and elevating oxidation in tumors by using compounds with prooxidant activity is developing as a new chemopreventive therapy in cancer. 76 Exogenous antioxidants not only indiscriminately block indispensable physiological redoxmediated cellular functions; these can also prevent cancer cells from oxidation-induced death. 77 Indeed, antioxidants such N-acetylcysteine (NAC) or vitamin E (vitE) accelerate tumor progression in mouse models of B-RAF-and K-RAS-induced lung cancer by inactivating p53. 78 NAC or vitE also potentiates disease progression in melanoma patients by promoting metastasis dependent on NADPH-generating folate pathway 79 or activation of small guanosine triphosphate RHOA. 80 However, studies evaluating antioxidants in cancer initiation, particularly in CAC, are scarce.

Testing Exogenous Antioxidants in IBD
Investigators have examined the protective role of synthetic compounds such as inhibitors of pro-oxidant enzymes and precursor of endogenous antioxidants in IBD. Among several drugs, inhibitors of angiotensin II type 1 (AT1) and hydroxymethylglutaryl coenzyme A reductase are reported to have both antioxidant and anti-inflammatory activities. AT1 increases mitochondrial production of O 2 and H 2 O 2 through NADPH oxidase and inflammation through nuclear factor kappa B. Accordingly, the AT1 antagonist telmisartan was protective in DSS-induced colitis. 81 Hydroxymethylglutaryl coenzyme A inhibitors such as simvastatin, rosuvastatin, and pravastatin are primarily lipid-lowering drugs, but they ameliorate disease severity in colitis models by reducing inflammation and inducing endogenous antioxidants such as SOD and glutathione. 82,83 As discussed earlier, glutathione is depleted in IBD patients. Administration of NAC, a synthetic precursor of glutathione, in colitis models ameliorated colitis severity, [84][85][86] but not all studies are in agreement. 3 NAC treatment in UC patients resulted in a significant improvement in clinical features and a reduction in serum proinflammatory cytokines. 87 Similarly, restoring colonic SOD level by administering lecithinized SOD, a synthetic SOD mimic, improved colitis severity in preclinical and human studies. 57,58 Other synthetic antioxidants such as propionyl-L-carnitine, an ester derivative of L-carnitine, improved disease in mild to moderate UC patients. 88,89 Natural exogenous antioxidants have also been examined in IBD. VitE is a lipid-soluble vitamin primarily involved in protecting cell membrane from oxidative damage. Supplementation of vitE in preclinical models of colitis ameliorated colitis severity. 90,91 However, results from clinical studies are inconclusive. 92,93 Vitamin C (vitC) is a water-soluble vitamin that acts as a potent antioxidant because of its ability to donate electrons. Low or high doses of vitC reduce inflammation in animal models 94,95 ; however, not all studies are in agreement. 3 Clinical studies using vitC in IBD patients are also inconsistent. 93,96 Altogether, the use of exogenous antioxidants to treat IBD needs further work because many studies are preliminary especially in clinical trials, and there are many conflicting findings (Table 1).

Testing Exogenous Antioxidants in CAC
Despite some promising albeit conflicting results with exogenous antioxidants in treating IBD, it is possible that many antioxidants have little to no anti-inflammatory activity and thus may not be useful in treating IBD. On the other hand, exogenous antioxidants could be used to protect from CAC by reducing oxidative DNA damage. However, only a few studies have evaluated antioxidants in preclinical models of CAC.
Long-term administration of NAC reduces oxidative damage (nitrotyrosine and 8-oxoG) in colonic mucosa and     protects from CAC development. 3,97,98 In addition to its role in cellular redox signaling, peroxynitrite, which is a coupling product of nitric oxide and superoxide, can oxidize DNA and produce DNA lesions. Accordingly, L-NIL, an inducible nitric oxide synthase inhibitor, reduced 8-oxoG levels and colonic polyps in multiple mouse models of CAC, despite having no significant effect on mucosal inflammation. 3 In agreement with this study, a derivative of L-NIL (SC-51) reduced inducible nitric oxide synthase and COX-2 activities and lessened the incidence of AOM-induced colonic aberrant crypt foci in rats. 99 These findings suggest that limiting nitrosative DNA damage might curb CAC development.
Other synthetic compounds such as GL-V9, a flavonoid derivative with strong antioxidant and anti-inflammatory activities, protect against tumorigenesis in a CAC model through NLRP3 inflammasome degradation. 100 Statins were found to reduce CAC in IBD patients in one study 101 but not in another study. 102 Among natural antioxidants, vitC reduces oxidative DNA damage by neutralizing mutagenic ROS and RNI 3,103 and protects from inflammation-associated tumorigenesis in different animal models of CAC. 3 Paradoxically, a prooxidant role for vitC has also been reported at high doses or in presence of transition metals. 104 High doses of vitC induce cytotoxicity in cancer cells, 105 and it was thus evaluated as a therapeutic agent in CRC patients. However, the results are inconsistent 106 (Table 2). Although studies on therapeutic evaluation of curcumin in CAC are limited, some preclinical studies have found a reduction in colonic tumor burden in CAC models. 107,108 The effects of resveratrol on CAC have been evaluated in one study that reported a reduction in tumor incidence in AOM/DSS model. 109,110 Although there have been numerous trials investigating the effects of antioxidants on disease pathology in IBD patients with mixed results (Table 1), few have used CAC as an endpoint. In light of the findings in preclinical models, serious consideration should be taken to test the role of antioxidants to prevent CAC in IBD patients.

Concluding Thoughts
Several translational studies have shown that antioxidants are effective at reducing both an overt oxidative environment and oxidative DNA lesions in the intestine and other tissues. 3,93,95,103 For years these studies have supported the belief that antioxidants can protect DNA from oxidative damage that could precipitate cancer. However, clinical studies that have tested this hypothesis have not reached consistent results. [111][112][113][114][115] One factor that could explain these contradictory results is that antioxidants have been promoted and tested as the panacea for all cancers. Although an excess in oxidative molecules could theoretically induce tumor initiating DNA lesions in any cell, susceptibility to oxidative DNA damage is expected to vary widely in different tissues. 78,116,117 Differences in cell proliferation, gene expression, and the cell's oxidative environment are expected to influence the probability of acquiring oxidative DNA mutations. For example, because oxidative DNA lesions that occur during S-phase of the cell cycle are more likely to result in mutations (Figure 2), highly proliferative tissues such as the intestine are more susceptible to acquire tumor-initiating mutations in oxidative environments. In addition, some tissues such as the intestine are in close contact with microbes and as a result are in a harsh oxidative environment produced by immune cells to keep microbes in check. It is therefore expected that oxidative DNA lesions only promote certain types of cancers. 116 Indeed, an analysis of mutational signatures in more than 40 different cancers found that most of colorectal and stomach adenocarcinomas have a ROS mutational signature, whereas other cancers do not. 116 The path of genetic mutations that are required for cellular transformation will be different in different tissues and cells. This depends on a number of factors such as the cellular environment and the type of cell being transformed. In the case for CRC and CAC, the cell type that is transformed is similar, but the environments where the cancers arise are different, which might explain the different genetic mutations associated with each of these cancers. For example, whereas mutations that affect the Wnt pathway occur in 85% of sporadic CRCs 118 and are considered to be the first step that leads to CRC initiation, CAC tumors first acquire mutations in p53, followed by KRAS mutations. 119 p53 is a transcription factor that controls the DNA damage response by inducing cell cycle arrest and apoptosis. 120 However, p53 is also involved in other cellular processes such as the antioxidant response, and its down-regulation results in increased DNA oxidation and mutation rates in lymphoma models. 120,121 Hence, it is tempting to speculate that inactivation of genes that regulate the antioxidant response is more important for the development of CAC than CRC possibly because of the high oxidative environment of the inflamed gut. This notion is supported by findings that antioxidants only reduced tumorigenesis in CAC models but not in a familial model of CRC (ie, MMRdeficient Lynch syndrome). 3 This result could be explained by the fact that most mutations in MMR-deficient cells are due to replication errors and spontaneous cytidine deamination, with only 20% of mutations potentially attributed to oxidative DNA lesions. 122 Hence, most mutations that appear in MMR-deficient cells cannot be prevented with antioxidant treatment. In contrast, antioxidants reduced tumorigenesis by 50% in all CAC models tested, 3 suggesting that a larger fraction of genetic lesions in inflamed colons is a consequence of oxidative DNA damage. This argument might provide an explanation for the conflicting results in clinical trials that tested antioxidants in CRC and suggests that clinical trials using antioxidants should stratify patients according to genetic susceptibility to acquire oxidative DNA lesions.
Importantly, because ROS are required to maintain homeostasis in the intestinal epithelium, antioxidants should be administered with precaution. Molecules such as O 2 and H 2 O 2 have both proliferative and antiproliferative effects and can regulate cell differentiation, intestinal repair, and antimicrobial defense. [8][9][10][11][12] Therefore, completely shutting off Redox signaling could potentially disrupt homeostasis and cause disease. Indeed, NOX1, NOX2, and DUOX2 deficiencies have been associated with a higher risk of developing pediatric 123 and very early onset IBD. [124][125][126] Furthermore, patients suffering from chronic granulomatous disease, a rare disorder characterized by deficiency in phagocytic NOX function, have a high risk to develop IBD, 127,128 suggesting that defective O 2 production can lead to IBD and therefore antioxidant doses should be carefully adjusted for these patients.
In conclusion, both in vitro and in vivo studies suggest the potential role of ROS and RNI in the pathophysiology of IBD and CAC. Epidemiologic studies show altered levels and activity of endogenous antioxidants in IBD and CAC patients. However, further studies are required to confirm their association with disease pathophysiology. There are some promising results from preclinical studies that the use of exogenous compounds (natural or synthetic) with antioxidant activity prevents oxidative DNA damage and CAC. However, this treatment strategy needs to be confirmed in the clinic.