A Brief Review of the Controversial Role of Iron in Colorectal Carcinogenesis

Colorectal Cancer (CRC) is the third most common cancer in males and the second in females worldwide with over 1.2 million new cases and over 600,000 estimated deaths in 2008 [1]. In the United States, it is estimated that 142,820 new cases of CRC with over 50,000 deaths occur annually [2]. The development of CRC has been shown to be related to individual’s genetics and environmental factors. The genetic effect on the development of sporadic CRC is reflected by different incidences occurring based on age, sex, and ethnicity [3]. About 10-15 % of CRC are solely attributable to genetic abnormalities. Hereditary CRC syndromes such as familial adenomatous polyposis (FAP) and hereditary nonpolyposis colorectal cancer (HNPCC) are two of these examples [4-7].

The development of sporadic CRCs, comprising the majority of cases, has been linked to many environmental factors. These exogenous factors include high intake of red and processed meats, highly refined grains and starches, sugars, fat and alcohol consumption [8,9]. Certain chemicals are also implicated in carcinogenesis. Among them, high iron consumption has been associated with an increased risk of CRC [10]. One study has demonstrated serum transferrin saturation (a surrogate for iron overload) as a strong independent marker for the presence of colon adenomas, precursors of colon cancer [11]. Other studies, however, showed no such association and that iron deficiency is possibly linked to gastrointestinal carcinogenesis by increased oxidative stress, DNA damage, and impairment of iron-dependent metabolic functions related to genome repair and protection [12]. This review will focus on iron overload as a controversial risk factor in colon carcinogenesis.

Iron is required for proliferation of normal and neoplastic cells [13-17]. The relationship between iron and cancer risk appears to be more obvious in the colon due to its high concentrations of iron. It has been suggested that the protective effect of dietary fiber observed in human colorectal cancer may be due to iron chelation by phytic acid in dietary fiber [18,19]. The Haber-Weiss cycle that results in synthesis of hydroxyl radical and other oxidants has been demonstrated in vitro and in vivo, and for iron delivered parenterally or enterally [18]. Multiple epidemiologic studies have shown a positive correlation between iron and colorectal cancer, including examination of effects of genetics, dietary fiber, dietary iron, and total body iron stores [10,13,14,20,21].

In a meta-analysis of epidemiological studies of the risk of colorectal carcinoma conferred by increased body iron load, approximately three quarters of 33 studies supported the association of iron with colorectal cancer risk [18,22]. The National Health and Nutrition Examination Survey I study (1971 - 1984) showed that iron s

tores were higher in 242 men who developed cancer than in 3113 men who were cancer free [18]. Total iron binding capacity (TIBC) and transferrin saturation (Tf sat.) were measured. Colon, esophagus, lung, and bladder cancers were the neoplasms associated with iron overload. The follow-up study by Wurzelman et al. [10] also showed a positive association between dietary and body iron stores with colorectal cancer [10,22]. Several prospective studies have also shown a positive correlation between various indices of increased body iron stores and colorectal cancer [10,13,14,20,21,23]. More evidence for the role of iron overload in the pathogenesis of colorectal carcinoma comes from the observation that the relative risk of colorectal cancer was substantially increased in hereditary hemochromatosis and in individuals who are either heterozygotes or homozygotes for Hemochromatosis Fe (HFE) gene [10,24,25].

Some studies have found that there is no association between the risk of colorectal cancer and body iron levels or dietary iron uptake [26,27], and that iron supplement or serum ferritin concentration is not linked to the recurrence of colorectal adenoma [27,28]. Negative associations between the risk of colorectal cancer and body iron levels or dietary iron uptake has also been reported, including one nested case–control study in women [29]. Kato et al. [29] found an inverse association with risk of colorectal cancer with one of the biomarkers for body iron stores, although there was a weak correlation with increased iron intake and an increased risk of colorectal cancer in certain subgroups, including subjects with higher fat intake [29]. Most recently, Cross et al., found a significant inverse association between several serum iron indices (serum ferritin, serum iron and transferring saturation) and colon cancer risk in a nested case-control study within the alpha-tocopherol, beta-carotene cancer prevention study cohort which included 130 colorectal cancer cases and based on a 276-item food frequency questionnaire [30]. Cross et al. [30] however, found the serum Unsaturated Iron Binding Capacity (UIBC) associated with an elevated risk for colon cancer.

In addition to association with colorectal cancer, studies have found that iron overload also has a positive relationship with the development of precancerous lesions. There is evidence showing that unabsorbed dietary iron increases free radical production in the colon to a level that could cause damage to the mucosa and crypt cell proliferation [31,32]. Results from animal studies showed that iron supplementation in rats led to cytotoxic events. These cytotoxic events include decreased manganese superoxide dismutase activity, increased lipid peroxidation and free radical generating capacity in the colon and cecum, and elevated colonic aberrant crypt foci [22,33-37]. Cytoplasmic staining for the iron storage protein ferritin was detected in both colonic adenoma and colorectal cancer, and expression of ferritin has been found to be positively associated with the degree of dysplasia and tumor size [38]. Examination of resected polyps and colorectal carcinomas showed an association of colonic adenoma progression to carcinoma with an increased expression of iron import proteins and a decreased expression of iron export proteins [39,40].

Controversial association between body iron load and colorectal carcinogenesis is listed in (Table 1).

Table 1: Summary of association between body iron load and the development of colorectal cancer.

In the colon, iron increases the production of Reactive Oxygen Species (ROS) from peroxidases via the Fenton reaction, which may cause cellular toxicity and pro-mutagenic lesions [41-43]. Lipids, DNA, and proteins are targeted by these ROS. Iron-overload was shown to induce oxidative DNA damage in the form of DNA breaks and oxidized bases in the human colon carcinoma cell line HT29 clone 19A when the tumor cells were incubated with ferric-nitrilotriacetate (Fe-NTA) or with hemoglobin [42]. Hemoglobin was used as a source of physiological iron and was found to be as efficient in inducing DNA damage as Fe-NTA in a clear-cut concentration-effect relationship [42]. DNA damage incubated with Fe-NTA or with hemoglobin was two-six folds of the control [42]. Iron increased damage two-three folds over the control at concentration which have been reported to be physiologically available (200-500μM) [17,42]. It has also been demonstrated that cells localized near the stem cells of the colon crypt were more susceptible to oxidative stress than colon cells located on the luminal part of the crypt [44].

Recent studies also suggest that iron-mediated ROS may target specific genes in Fenton reaction-induced cancer. A Fe-NTA-induced rat Renal Cell Carcinoma (RCC) model showed increased numbers of oxidatively modified DNA bases including 8-oxoguanine and a major lipid peroxidation product, 4-hydroxy-2-nonenal [45-47]. Gpt delta transgenic mice showed deletion and single nucleotide substitutions at G:C sites to be preferred mutations in the kidney treated with Fe-NTA [45,48]. Microsatellite analysis in F1 hybrid rats revealed that p15 and p16 tumor suppressor genes were targeted and either homozygously deleted or methylated at the promoter region [49]. This led to the proposal of ‘fragile’ sites in the genome susceptible to oxidative stress [45]. Gene expression microarray and array-based comparative genome hybridization analyses identified target oncogenes in Fe-NTAinduced RCC. A tyrosine phosphatase, ptprz1, was found to be highly expressed in RCC and at the chromosomal region of amplification (4q22). This model showed that iron-mediated oxidative stress induced genomic amplification of ptprz1 which results in activation of β-catenin pathways during carcinogenesis [45,50].

It has also been proposed that iron can induce early signaling pathways that modulate activities of oxidative-responsive transcription factors such as activator protein 1 (AP-1) and nuclear factor kappa B (NF?B) [22]. AP-1 and NF?B binding sites are found in promoters of many genes such as interleukin-6 (IL-6). Activation of these transcription factors may lead to upregulation of cytokine genes, and over-expression of IL-6 caused by iron may lead to chronic inflammation and carcinogenesis [22,51].

Other possible mechanisms for iron carcinogenesis include effects of iron on the immune system. Tumoricidal activity of macrophages was markedly suppressed by phagocytosed erythrocytes, erythrocyte lysate, hemoglobin, iron salts or iron dextran [22,52]. The ratio of CD4 and CD8 positive lymphocytes were lowered by iron [53]. Studies in mice have also shown iron to reduce both macrophage and tumor cellsderived nitric oxide release, which promotes carcinogenesis [54].

Iron itself may facilitate cancer growth by serving as a nutrient. Iron is required for cell proliferation and is a constituent of proteins that catalyze key reactions including oxygen sensing, energy metabolism, respiration, and DNA synthesis [22,55]. Tumor cells express more transferrin receptors, resulting in higher cellular iron uptake for growth than normal cells [22].

Another interesting aspect of iron carcinogenesis is the synergistic effect iron may have with non-amidated gastrins in the development of CRC. Non-amidated gastrins include progastrin and Ggly. CRCs have been found to express progastrin in greater amounts than normal colon cells, and patients with CRC have increased circulating concentrations of Gamide and total gastrins [40,56,57]. It has also been recognized that ferric ions were essential for the biological activities of non-amidated gastrins with the proposal that gastrins catalyze the loading of transferrin with iron [40]. Although precise mechanisms underlying the involvement of gastrins in iron homeostasis are unknown, this synergistic effect of non-amidated gastrins and iron on CRC development is strongly suggested [40]. Figure 1 summarizes the main mechanisms involved in iron-mediated carcinogenesis.

Figure 1: Overview of the main mechanisms involved in iron-mediated carcinogenesis.

Many epidemiologic and experimental studies have shown a positive association between dietary and body iron stores with the development of colorectal cancer and precancerous lesions. However, the risk of iron intake for the development of CRC is controversial. There are a few studies showing no association between the risk of colorectal cancer and body iron levels or dietary iron uptake and that iron supplement or serum ferritin concentration is not linked to the recurrence of colorectal adenoma. Inverse associations between the risk of colorectal cancer and body iron levels or dietary iron uptake have also been reported.

The mechanisms of possible iron carcinogenesis are not unclear. Based on in vitro evidence, it is proposed that iron can induce early signaling pathways that modulate oxidative-responsive transcription factors, or that iron suppresses tumoricidal activity of macrophages and lowers the ratio of CD4 and CD8 positive lymphocytes. Emerging research shows that certain genomic sequences may be vulnerable to oxidative damage. This illustrates the complex interplay of genetic and metabolic alterations in chemical carcinogenesis.