Dual modulatory role of folate in carcinogenesis: cancer develops over decades, if not lifetime, through different stages of premalignant lesions in the target organ. Folate deficiency in normal tissues predispose them to neoplastic transformation, and modest supplemental levels suppress, whereas supraphysiologic doses of supplementation enhances, the development of tumors in normal tissues. In contrast, folate deficiency has an inhibitory effect whereas folate supplementation has a promoting effect on the progression of established neoplasms. It is unknown at present the effect of folate deficiency and supplementation on the progression of early precursor or preneoplastic lesions of CRC ( e. g. , aberrant crypt foci, ACF) to adenoma and to frank cancer. The mechanisms by which folate exerts dual modulatory effects on carcinogenesis depending on the timing and dose of folate intervention relate to its essential role in one-carbon transfer reactions involved in DNA synthesis and biological methylation reactions. 

Contexts in source publication

Context 1

... from animal studies and clinical observations suggest that folate possesses dual modulatory effects on CRC devel-opment and progression depending on the timing and dose of folate intervention (Fig. 3) [5,6,58,84,169]. Folate defi- ciency has an inhibitory effect whereas folate supplementa- tion has a promoting effect on the progression of estab- lished colorectal neoplasms (Fig. 3). In contrast, folate defi- ciency in normal colorectal mucosa appears to predispose it to neoplastic transformation, and modest levels of folic acid ...

Context 2

... observations suggest that folate possesses dual modulatory effects on CRC devel-opment and progression depending on the timing and dose of folate intervention (Fig. 3) [5,6,58,84,169]. Folate defi- ciency has an inhibitory effect whereas folate supplementa- tion has a promoting effect on the progression of estab- lished colorectal neoplasms (Fig. 3). In contrast, folate defi- ciency in normal colorectal mucosa appears to predispose it to neoplastic transformation, and modest levels of folic acid supplementation (4 -10 times above the BDR) suppress, whereas supraphysiological supplemental doses enhance the development of CRC in normal colorectal mucosa (Fig. 3). Are there ...

Context 3

... lished colorectal neoplasms (Fig. 3). In contrast, folate defi- ciency in normal colorectal mucosa appears to predispose it to neoplastic transformation, and modest levels of folic acid supplementation (4 -10 times above the BDR) suppress, whereas supraphysiological supplemental doses enhance the development of CRC in normal colorectal mucosa (Fig. 3). Are there biologically plausible explanations for these seemingly paradoxical and contradictory epidemiolo- gic, animal, and clinical observations concerning the dual role of folate in CRC development and ...

Context 4

... 186]. Therefore, the effect of folate deficiency and supple- mentation on the DNA synthesis pathway in the normal col- orectum have been generally considered to be the primary mechanism by which folate deficiency predisposes it to neo- plastic transformation and folate supplementation prevents or suppresses neoplastic transformation, respectively (Fig. 3) [58, 59, 121 -123, 186]. Another proposed mechanism by which folate deficiency enhances the development of cancer in the colorectum is through an induction of genomic DNA hypomethylation [90]. It has been proposed that a mechanism by which folate supplementation may protect against the development of cancer in the colorectum is through ...

Context 5

... appears to be able to reverse pre-existing genomic DNA hypomethylation and to increase the extent of genomic DNA methylation above the pre-existing level [49,50]. Therefore, prevention or rever- sal of genomic DNA hypomethylation may be a mechanism by which folate supplementation suppresses neoplastic transformation in the colorectum (Fig. ...

Context 6

... preneoplastic and neoplastic cells where DNA replica- tion and cell division are occurring at an accelerated rate, folate depletion causes ineffective DNA synthesis, resulting in inhibition of tumor growth and progression (Fig. 3), which is the basis for cancer chemotherapy using antifolate agents (e. g., methotrexate) and 5-fluorouracil [58,59,123]. Thus, this is the most likely mechanism by which folate deficiency inhibits the progression of the established pre- neoplastic neoplastic foci in the colorectum. Another possible mechanism is that folate deficiency ...

Context 7

... possible mechanism by which folic acid supplementation may promote the progres- sion of preneoplastic or neoplastic foci in the colorectum may be de novo methylation of promoter CpG islands of tumor suppressor genes and other critical genes involved in colorectal carcinogenesis with consequent gene inactiva- tion leading to tumor progression (Fig. 3). This potential epigenetic mechanism of tumor progression is supported by recent animal studies using viable yellow agouti mice that unequivocally have demonstrated that maternal dietary methyl group supplementation containing folic acid perma- nently alters phenotypic coat color of the offspring via increased methylation at the ...

Context 8

... possible means by which folic acid supplemen- tation may promote colorectal carcinogenesis may be through hypermutability of methylated cytosines in CpG dinucleotides (Fig. 3). Methylated CpG sites are mutational hot spots for human cancer as described earlier [194]. The majority of mutations observed in CpG sites are cytosine- to-thymine transitions mediated by the spontaneous deami- nation of 5-methylcytosine to thymine and by the enzymatic deamination of 5-methylcytosine to thymine by DNMT [194]. ...

Context 9

... folate status and CRC risk. Definitive answers to questions about folate and CRC are probably beyond the reach of both observational epidemiologic studies and ran- domized controlled trials [204]. It is clear that folate appears to possess dual modulatory effects on colorectal carcinogenesis depending on the timing and dose of folate intervention (Fig. 3). Folate deficiency has an inhibitory effect whereas folate supplementation has a promoting effect on the progression of established colorectal neo- plasms (Fig. 3). In contrast, folate deficiency in normal col- orectal mucosa appears to predispose it to neoplastic trans- formation, and modest levels of folic acid supplementation ...

Context 10

... domized controlled trials [204]. It is clear that folate appears to possess dual modulatory effects on colorectal carcinogenesis depending on the timing and dose of folate intervention (Fig. 3). Folate deficiency has an inhibitory effect whereas folate supplementation has a promoting effect on the progression of established colorectal neo- plasms (Fig. 3). In contrast, folate deficiency in normal col- orectal mucosa appears to predispose it to neoplastic trans- formation, and modest levels of folic acid supplementation suppress, whereas supraphysiologic supplemental doses enhance the development of CRC in normal colorectal mucosa (Fig. 3). Several potential mechanisms relating to the ...

Context 11

... on the progression of established colorectal neo- plasms (Fig. 3). In contrast, folate deficiency in normal col- orectal mucosa appears to predispose it to neoplastic trans- formation, and modest levels of folic acid supplementation suppress, whereas supraphysiologic supplemental doses enhance the development of CRC in normal colorectal mucosa (Fig. 3). Several potential mechanisms relating to the disruption of the known biochemical function of folate (mediating the transfer of one-carbon moieties and conse- quent DNA synthesis and methylation) exist to support the dual modulatory role of folate in colorectal carcinogenesis (Fig. ...

Context 12

... enhance the development of CRC in normal colorectal mucosa (Fig. 3). Several potential mechanisms relating to the disruption of the known biochemical function of folate (mediating the transfer of one-carbon moieties and conse- quent DNA synthesis and methylation) exist to support the dual modulatory role of folate in colorectal carcinogenesis (Fig. ...

Context 13

... clearly established in humans. An obvious inference from the above discussion is that for folate to be a safe and effec- tive chemopreventive agent against CRC, modest doses of folic acid supplementation should be implemented before the development of precursor lesions in the colorectum or in individuals free of any evidence of neoplastic foci (Fig. 3). However, determining the presence of neoplastic foci in the general population is an almost impossible task. Furthermore, folate might prevent the progression of cer- tain precursor or preneoplastic lesions to frank malignancy but promote the progression of other lesions. What constitu- tes safe precursor or preneoplastic lesions on ...

Context 14

... DNA methylation in lymphocytes (by 31%; p = 0.05) and in colonic mucosa (by 25%; p = 0.09) in patients with colorectal adenomas [152]. Collectively, currently available evidence indicates that folate supplementation appears to be able to reverse pre-existing genomic DNA hypomethylation and to increase the extent of genomic DNA methylation above the pre-existing level [49, 50]. Therefore, prevention or rever- sal of genomic DNA hypomethylation may be a mechanism by which folate supplementation suppresses neoplastic transformation in the colorectum (Fig. 3). In preneoplastic and neoplastic cells where DNA replication and cell division are occurring at an accelerated rate, folate depletion causes ineffective DNA synthesis, resulting in inhibition of tumor growth and progression (Fig. 3), which is the basis for cancer chemotherapy using antifolate agents ( e. g. , methotrexate) and 5-fluorouracil [58, 59, 123]. Thus, this is the most likely mechanism by which folate deficiency inhibits the progression of the established preneoplastic neoplastic foci in the colorectum. Another possible mechanism is that folate deficiency may reverse CpG promoter methylation of tumor suppressor and other anticancer genes involved in colorectal carcinogenesis, thereby reactivating these genes. However, there is currently no experimental evidence to support this theo- retical possibility. As discussed earlier, folate deficiency appears to be unable to induce genomic and gene-specific DNA hypomethylation in the colorectum [49, 50]. Furthermore, in a recent in vitro study, folate deficiency induced a significant reduction in genomic and site-specific DNA methylation in untransformed NIH/3T3 and CHO-K1 mammalian cells but not in HCT116 and Caco2 human colon adenocarcinoma cells [196]. In this study, folate deficiency did not produce significant changes in the promoter CpG island methylation of the p16 tumor suppressor gene and the MLH1 mismatch repair gene in HCT116 and Caco- 2 cells [196]. However, certain sites in the promoter CpG island of the ER gene were associated with modest, albeit statistically significant, changes in CpG methylation in response to folate deficiency, which were not associated with significant functional consequences [196]. In line with this observation, another study showed that HCT116 cells lacking DNMT1 exhibited only a modest 20% decrease in the overall genomic DNA methylation despite the markedly decreased cellular DNMT activity [197]. In this model, although juxtacentromeric satellites became significantly demethylated, centromeric satellite loci, and the promoter CpG island of the p16 gene remained fully methylated [197]. Only when both the DNMT1 and DNMT3b genes were disrupted, genomic DNA methylation was reduced by A 95% and significant hypomethylation of satellite sequences and several promoter CpG islands, including that of the p16 gene, was observed [198]. These observations suggest that it may be extremely difficult to reverse DNA methylation in cancer cell lines such as HCT116. The fact that an almost complete abolishment of DNMT activity by disruption of both the DNMT1 and DNMT3b genes is required to produce significant DNA hypomethylation in HCT116 cells [198] suggest that folate deficiency alone is unlikely to be a sufficient predisposing condition to produce significant DNA hypomethylation in colon cancer cells. Mechanistically, the most likely mechanism by which folic acid supplementation may promote the progression of established preneoplastic and neoplastic lesions in the col- orecutm is provision of nucleotide precursors to rapidly replicating neoplastic cells for accelerated proliferation and growth [58, 59, 123]. Another possible mechanism by which folic acid supplementation may promote the progression of preneoplastic or neoplastic foci in the colorectum may be de novo methylation of promoter CpG islands of tumor suppressor genes and other critical genes involved in colorectal carcinogenesis with consequent gene inactiva- tion leading to tumor progression (Fig. 3). This potential epigenetic mechanism of tumor progression is supported by recent animal studies using viable yellow agouti mice that unequivocally have demonstrated that maternal dietary methyl group supplementation containing folic acid permanently alters phenotypic coat color of the offspring via increased methylation at the promoter CpG site of the agouti gene [199 – 201]. However, it is unknown at present whether this de novo methylation of promoter CpG islands can happen with folic acid supplementation alone, whether it is operative in normal or neoplastic tissues or both, whether this effect is associated with folic acid supplementation provided in utero only, in postpartum period, or in adulthood, and whether it is tissue and gene-specific. Another possible means by which folic acid supplementation may promote colorectal carcinogenesis may be through hypermutability of methylated cytosines in CpG dinucleotides (Fig. 3). Methylated CpG sites are mutational hot spots for human cancer as described earlier [194]. The majority of mutations observed in CpG sites are cytosine- to-thymine transitions mediated by the spontaneous deamination of 5-methylcytosine to thymine and by the enzymatic deamination of 5-methylcytosine to thymine by DNMT [194]. Therefore, there is a possibility that de novo methylation of cytosines in CpG sites in critical genes involved in colorectal carcinogenesis may create mutational hot spots, leading to inactivating mutations of these genes. A cause-and-effect relationship between folate and CRC is difficult to establish. Because of inherent limitations associated with study design, the results from epidemiologic, animal, and interventional studies examining this relationship have been inconsistent and conflicting. In clinical medicine, the best evidence has been considered to come from well designed and executed double-blind randomized controls trials, which minimize a variety of biases. This has resulted in a clear hierarchy of evidence that is weighted heavily toward randomized controlled trials. Evidence from randomized controlled trials is thought to supersede evidence from other sources such as observational studies. The field of nutritional epidemiology has also followed this traditional approach and considered correlation, case-control, and prospective observational epidemiologic studies and intervention trials as a spectrum of increasing weight of evidence for or against a relationship between dietary factors and cancer risk [202]. Thus, general conclusions and recommendations regarding the effect of dietary factors on cancer risk have relied heavily on data from large prospective studies and randomized, controlled intervention human trials [202]. This traditional approach to grading epidemiologic evidence concerning the relationship between dietary factors and cancer risk has recently been challenged [203, 204]. As cancer develops over decades, if not a lifetime, single clinical trials, which normally last up to 5 years, cannot address the whole span of cancer development. In addition, randomized controlled trials tend to use uncharacteristic levels of exposure. Furthermore, the dietary, nutritional, and physi- cal activity exposures involved are complex and interre- lated, making them difficult to manipulate in a controlled fashion. Even if a difference in outcome followed such a clinical intervention, it would not necessarily indicate that reproducing the intervention under other conditions would cause similar outcomes. Thus, it has been argued that draw- ing a definitive conclusion concerning the effect of dietary factors on cancer risk mainly from randomized, controlled intervention human trials is probably not the right paradigm of nutritional epidemiology [203, 204]. Rather, it has been articulated that the totality or “portfolio” of evidence from observational and intervention studies as well as animal and in vitro experiments must be analyzed for this purpose [203, 204]. The portfolio approach does not set out a hierarchy of evidence. Instead, it recognizes that all types of evidence have advantages and disadvantages. This means that no single kind of study is considered to be definitive. Instead, all of the different types of studies that are used to investigate the link between nutrition and cancer are considered along- side each other, without favoring evidence from one type over another. In support of the portfolio approach, systema- tic comparisons of the results of randomized intervention studies with observational evidence in several clinical situations have shown that observational data from well-con- ducted studies do not appear to produce biased results compared to randomized interventions [205, 206]. Furthermore, the importance of experimental studies that contribute to understanding mechanisms that might under- lie any observed association between a dietary factor and cancer and might bear on the inference of causation has been increasingly recognized and appreciated in the field of nutrition and cancer. Epidemiologic and experimental evidence indicating a causal association between a dietary factor and cancer is strengthened when a biologic pathway or mechanism by which colorectal carcinogenesis may be modified is identified and when this mechanism is biologically plausible [204]. It can be argued that epidemiologic data, however strong and consistent, are an inadequate basis for any definite judgment of causality unless supported by mechanistic evidence [204]. Recent advances in molecular epidemiology have added another dimension to the already complex field of nutrition and cancer. Recently identified and characterized single nucleotide polymorphisms and other genetic and epigenetic variants of genes that are involved in absorption, transport, metabolism, and excretion of nutrients have been shown not only to modify cancer risk but also to significantly ...

Context 15

... portfolio approach, systema- tic comparisons of the results of randomized intervention studies with observational evidence in several clinical situations have shown that observational data from well-con- ducted studies do not appear to produce biased results compared to randomized interventions [205, 206]. Furthermore, the importance of experimental studies that contribute to understanding mechanisms that might under- lie any observed association between a dietary factor and cancer and might bear on the inference of causation has been increasingly recognized and appreciated in the field of nutrition and cancer. Epidemiologic and experimental evidence indicating a causal association between a dietary factor and cancer is strengthened when a biologic pathway or mechanism by which colorectal carcinogenesis may be modified is identified and when this mechanism is biologically plausible [204]. It can be argued that epidemiologic data, however strong and consistent, are an inadequate basis for any definite judgment of causality unless supported by mechanistic evidence [204]. Recent advances in molecular epidemiology have added another dimension to the already complex field of nutrition and cancer. Recently identified and characterized single nucleotide polymorphisms and other genetic and epigenetic variants of genes that are involved in absorption, transport, metabolism, and excretion of nutrients have been shown not only to modify cancer risk but also to significantly modulate the effect of nutrients and related compounds on cancer risk [207]. This emerging important topic in the field of nutrition and cancer, termed “gene – nutrient interactions” in carcinogenesis, has a very significant implication in designing and interpreting data from observational epidemiologic and intervention studies. Although individuals are subjected to the same level of nutritional exposure, systemic, and target tissue bioavailability of nutrients and their metabolites, as well as their functional effects in the target tissue, might be vastly different because of genetic and epigenetic variations. Genetic and epigenetic susceptibility to cancer and their interaction with diets and other environmental exposures have not been incorporated into the study design of and interpretation of data from previously published epidemiologic and intervention studies. The precise nature and magnitude of gene – nutrient interactions in carcinogenesis are yet to be clearly defined. It appears that overall diet, rather than individual factors, plays the more important role in the development of CRC, thus underscoring the importance of as yet undetermined interactions among dietary components in the development of cancer. It is likely that dietary factors or components do not act in isolation but as part of a biological action package [208]. The major difficulty in establishing a relationship between diet and cancer and in translating observations from nutritional epidemiology into progress in cancer prevention has been due to inability to identify all relevant dietary components that act coordinately to modulate cancer risk and due to inability to identify the other relevant non-nutritional factors that interact with dietary components to modify cancer risk [208]. What can we conclude about the role of folate in CRC development and progression from the seemingly paradoxical and contradictory epidemiologic, animal, and clinical studies? Currently available evidence from epidemiologic, laboratory animal, and intervention studies does not unequivocally support the role of folate in the development and progression of CRC. Furthermore, the precise nature and magnitude of the relationship of CRC with folate have not been clearly defined. However, when the whole body or portfolio of evidence from these studies is analyzed critically, the overall conclusion supports the inverse association between folate status and CRC risk. Definitive answers to questions about folate and CRC are probably beyond the reach of both observational epidemiologic studies and randomized controlled trials [204]. It is clear that folate appears to possess dual modulatory effects on colorectal carcinogenesis depending on the timing and dose of folate intervention (Fig. 3). Folate deficiency has an inhibitory effect whereas folate supplementation has a promoting effect on the progression of established colorectal neoplasms (Fig. 3). In contrast, folate deficiency in normal col- orectal mucosa appears to predispose it to neoplastic transformation, and modest levels of folic acid supplementation suppress, whereas supraphysiologic supplemental doses enhance the development of CRC in normal colorectal mucosa (Fig. 3). Several potential mechanisms relating to the disruption of the known biochemical function of folate (mediating the transfer of one-carbon moieties and consequent DNA synthesis and methylation) exist to support the dual modulatory role of folate in colorectal carcinogenesis (Fig. 3). As discussed briefly above, evidence for a protective effect of folate supplementation on NTD [7, 8] was considered to be sufficiently conclusive and led to mandatory folic acid fortification in the US [11] and Canada [12] in 1998. Folic acid fortification has already significantly improved folate status and has had a substantial beneficial effect on the original target, NTD, in the US and Canada [21 – 26]. Mandatory folic acid fortification is probably the most important science-drive intervention in nutrition and public health in decades [209]. However, the possibility remains that certain segments of the exposed population may benefit less and may even experience some adverse effects from an increased folic acid intake. Over the past few years, the US and Canadian populations have been exposed to a significant increase in folate intake, for which essentially no data on safety exist [13]. No studies have been done to look directly or even indirectly for the adverse effects of greatly increased folate intakes [13]. In addition to the drastic increase in dietary folate intake from mandatory folic acid fortification, 30 – 40% of the North American population consume supplemental folic acid for several possible but as yet unproven health benefits [35]. Whether or not possible deleterious effects of folic acid supplementation ( e. g. , cancer-promoting effect on established preneoplastic and neoplastic lesions) outweigh the known and potential health benefits ( e. g. , prevention of atherosclerosis and NTD; improvement of cognitive function; cancer prevention in normal tissues free of preneoplastic and neoplastic foci) is largely unknown at present. Folate is generally regarded as safe [210] and may become the ultimate functional food component for disease prevention [211]. The potential masking effect of folic acid on vitamin B 12 deficiency, especially in the elderly, has been the only major concern of folic acid fortification and supplementation [13]. However, an emerging body of evidence suggests that folate supplementation may be associated with other potentially serious adverse effects [6]. These include: the occurrence of resistance or tolerance to antifolate-based chemotherapy and anti-inflammatory and antiseizure drugs; decreased natural killer cell cytotoxicty; accelerated cognitive decline in older subjects; increased twin pregnancies; and genetic selections of disease alleles ( e. g. , MTHFR C677T) that predispose individuals to chronic diseases if exposed to low folate status [6, 211 – 220]. Folic acid, the synthetic, fully oxidized form of folate used in fortification and supplementation, is normally reduced and methylated by the intestine before it is released into the circulation as 5-methylTHF; consequently, the latter form is the sole circulating form of folate under normal conditions [2]. However, studies show that this absorption and biotransformation process is saturated at doses in the region of 400 l g folic acid or less [221]. At higher doses, synthetic folic acid is also transported into the blood and may enter in large quantities. Consumption of folic acid A 200 l g have shown to lead to the appearance of unmetabolized folic acid in the serum. Although compelling data about possible antagonistic activities of this fully oxidized form of folate in tissues is lacking, there nevertheless exist concerns about the effect of long-term exposure of cells to unmetabolized folic acid [222]. In this regard, Troen et al. [212] have recently reported that 78% of 104 postmenopausal women 60 – 75 years of age had detectable levels of folic acid in plasma, which was associated with an approxi- mately 23% decrease in natural killer cell cytotoxicity inde- pendent of plasma 5-methylTHF and total folate concentrations. Among participants in a large ( n = 2928) trial of folic acid supplementation during pregnancy, women who received 5 mg folic acid/day had a 70% increased risk of total cancer mortality compared with those not on supplementation (HR = 1.70; 95% CI = 1.06 – 2.72) [168]. In this study, the risk of death from breast cancer in women taking 5 mg folic acid/day was twice that of women taking no supplementation, albeit nonsignificant (HR = 2.02; 95% CI = 0.88 – 4.72) [168]. Furthermore, in line with this observation, a recent the Prostate, Lung, Colorectal, and Ovarian Cancer Screening Trial ( n = 25, 400 postmenopausal women) has reported a significant positive association between total folate and supplemental folic acid intakes, but not food folate, and breast cancer risk [85]. Given the new evidence that cells of the absorptive mucosa in the small intestine may not be the primary site for folic acid biotransformation and that humans have an extremely low activity of DHFR, which converts folic acid through dihy- drofolic acid to tetrahydrofolic acid, a form that can then enter the main folate metabolic cycle [177], potential adverse effects of the appearance of large quantities ...

Context 16

... are critical in providing safe and effective chemoprevention [5, 58, 169]. Two animal studies using the well-established chemical carcinogen (dimethylhydrazine, DMH) rat model of CRC showed that a moderate degree of folate deficiency promoted, whereas modest levels of folic acid supplementation up to 4 6 the basal daily requirement (BDR) for rodents inhibited, the development of CRC [90, 93]. There was a suggestion that a very high supplemental dose of folic acid (20 6 the BDR] might promote the progression of microscopic neoplastic foci to CRC [93]. In support of this latter finding, animal studies using a metabolite of DMH, azoxy- methane (AOM), showed that folic acid supplementation exceeding the BDR by 1000 – 10 000 6 enhanced colorectal carcinogenesis in rats [170 – 174]. Therefore, it appears that folate modulates colorectal carcinogenesis in chemical carcinogen rodent models over a wide range of dietary intakes. Folate deficiency of a moderate degree enhances colorectal carcinogenesis whereas modest levels of folate supplementation above the BDR suppress colorectal tumorigenesis. Supraphysiologic levels of folate supplementation do not appear to confer additional protection and, in some cases, may enhance colorectal carcinogenesis. The implication of this issue is important because the optimal dose of folate supplementation must be determined for folate chemoprevention to be effective and safe in humans. Although some similarities do exist, tumor development in chemical rodent models of CRC differs in several important histological, clinical, and molecular genetic aspects from that observed in humans [175, 176]. Therefore, any extrapolation of the observations from these models to human situations should be made very cautiously. Whether or not the supplemental doses of folic acid used in these animal studies can be directly extrapolated to intake levels in humans is a highly contentious and controversial issue at present because of inherent differences in folate absorption and metabolism between rodents and humans [177, 178]. Recent evidence suggests that animals, unlike humans, have a comparatively high DHFR activity [177, 178]. Consequently, if assessing the impact of systemic exposure of unmetabolized folic acid, animals would have to be orally dosed with a much greater than prorata amount of folic acid in order to elicit the same circulating serum/plasma concentrations of unmetabolized folic acid [177, 178]. Therefore, it may be a gross mistake to dismiss the effects of folic acid exposure in animal studies on the grounds that the experimental folic acid intake (multiple of animal BDR) would translate to an unlikely intake in humans [177, 178]. Therefore, arguably, a 10 – 20 6 exposure in a small animal model may turn out to hardly equate to an extra 1 – 2 6 RDA in humans [177, 178]. In a recent study using the Apc Min mouse model of CRC, both semisynthetic diets with low and high vitamin contents (1/3 of the BDR and 5 6 BDR, respectively), which also contained 1/3 of the BDR and 2 6 BDR folic acid, respectively, significantly increased the number of small intestinal polyps [179]. Furthermore, in two genetic models of CRC ( Apc Min and Apc +/ – 6 Msh2 – / – mice), moderate dietary folate deficiency enhanced, whereas modest levels of folic acid supplementation (four to ten times the BDR) suppressed, the development and progression of CRC, if folate intervention was started before the establishment of neoplastic foci in the intestine [180, 181]. If, however, folate intervention was started after the establishment of neoplastic foci, the same degree of folate deficiency inhibited the progression and induced regression of the established neoplastic foci [180, 181]. A potential tumor promoting effect of folic acid supplementation on the established neoplastic foci could not be clearly determined in these studies because of the inherent limitations associated with these genetic models [180, 181]. Therefore, these observations suggest that the timing of folate intervention is critical in providing an effective and safe chemopreventive effect on colorectal carcinogenesis. Folate deficiency has an inhibitory effect whereas folate supplementation has a promoting effect on the progression of established colorectal neoplasms. In contrast, folate deficiency in the normal colorectum appears predispose it to neoplastic transformation, and modest supplemental levels of folate suppress the development of neoplasms in the normal colorectum. Some animal studies have also shown that dietary folate deficiency inhibits, and not suppresses, the development of breast cancer in rats [182 – 184] in contrast to the inverse association between folate status and breast cancer risk observed in epidemiologic studies [185]. Data from animal studies and clinical observations suggest that folate possesses dual modulatory effects on CRC devel- opment and progression depending on the timing and dose of folate intervention (Fig. 3) [5, 6, 58, 84, 169]. Folate deficiency has an inhibitory effect whereas folate supplementation has a promoting effect on the progression of established colorectal neoplasms (Fig. 3). In contrast, folate deficiency in normal colorectal mucosa appears to predispose it to neoplastic transformation, and modest levels of folic acid supplementation (4 – 10 times above the BDR) suppress, whereas supraphysiological supplemental doses enhance the development of CRC in normal colorectal mucosa (Fig. 3). Are there biologically plausible explanations for these seemingly paradoxical and contradictory epidemiologic, animal, and clinical observations concerning the dual role of folate in CRC development and progression? There exist several biologically plausible mechanisms by which folate deficiency increases, whereas folate supplementation reduces, the risk of CRC in normal colorectal epithelial cells [50, 58, 59, 123]. As an essential cofactor for the de novo biosynthesis of purines and thymidylate (Fig. 2), folate plays an important role in DNA synthesis, stability and integrity, and repair, aberrations of which have been impli- cated in colorectal carcinogenesis [58, 59, 123]. Indeed, a large body of evidence from in vitro , animal and human studies indicates that folate deficiency is associated with DNA strand breaks, chromosomal and genomic instability, uracil misincorporation, impaired DNA repair, and increased mutations [58, 59, 121 – 123, 186]. Furthermore, this body of evidence indicates that folate supplementation can correct some of these defects induced by folate deficiency, and ensures DNA fidelity, maintains DNA integrity and stability, and optimizes DNA repair by providing nucleotide precursors for DNA synthesis and replication [58, 59, 121 – 123, 186]. Therefore, the effect of folate deficiency and supplementation on the DNA synthesis pathway in the normal colorectum have been generally considered to be the primary mechanism by which folate deficiency predisposes it to neoplastic transformation and folate supplementation prevents or suppresses neoplastic transformation, respectively (Fig. 3) [58, 59, 121 – 123, 186]. Another proposed mechanism by which folate deficiency enhances the development of cancer in the colorectum is through an induction of genomic DNA hypomethylation [90]. It has been proposed that a mechanism by which folate supplementation may protect against the development of cancer in the colorectum is through a protection against genomic DNA hypomethylation [90]. This mechanism is based on the biochemical function of folate in mediating one-carbon transfer for the provision of SAM, the primary methyl group donor for most biological methylation reactions, including that of DNA (Fig. 2) and on evidence from animal experiments that demonstrated that diets deficient in different combinations of methyl group donors (choline, folate, methionine, and vitamin B 12 ) consistently induce genomic and site and gene-specific DNA hypomethylation [49, 50]. DNA methylation of cytosine located within the cyto- sine-guanine (CpG) dinucleotide sequences is an important epigenetic determinant in gene expression (an inverse relation), in the maintenance of DNA integrity and stability, in chromosomal modifications, and in the development of mutations [124, 187]. In contrast to methylated CpG sites in the CpG-poor bulk of the genome and unmethylated CpG islands in normal cells, cancer cells simultaneously harbor widespread loss of methylation in the CpG-depleted regions where most CpG dinucleotides should be methylated and gains in methylation of CpG islands in gene promoter regions (Fig. 4) [124, 187]. Global hypomethylation is an early, and consistent, event in colorectal carcinogenesis [124, 187]. Global hypomethylation of the coding and noncoding regions and demethylation of repetitive DNA sequences contribute to the development of cancer through the following mechanisms: chromosomal instability; increased mutations; reactivation of intra- genomic parasitic sequences that could be transcribed and moved to other sites where they could disrupt normal cellular genes; mitotic recombination leading to loss of heterozygosity and promotion of rearrangements; aneuploidy; loss of imprinting; and up-regulation of protooncogenes (Fig. 4) [188]. However, animal studies have shown that genomic demethylation may protect against some cancers ( e. g. , intestinal tumors) [189, 190] but may promote chromosomal instability and increase the risk of cancer in other tissues ( e. g. , lymphoma, sarcoma) [191, 192]. Methylation at promoter CpG islands is an important mechanism of silencing transcription in carcinogenesis; the affected genes are silenced and their function is stably lost in a clonally propagated fashion (Fig. 4) [187, 188, 193]. Many genes inactivated by promoter CpG methylation in carcinogenesis have classic tumor-suppressor function or play critical roles in cell cycle control, repair of DNA damage, ...

Context 17

... a comparatively high DHFR activity [177, 178]. Consequently, if assessing the impact of systemic exposure of unmetabolized folic acid, animals would have to be orally dosed with a much greater than prorata amount of folic acid in order to elicit the same circulating serum/plasma concentrations of unmetabolized folic acid [177, 178]. Therefore, it may be a gross mistake to dismiss the effects of folic acid exposure in animal studies on the grounds that the experimental folic acid intake (multiple of animal BDR) would translate to an unlikely intake in humans [177, 178]. Therefore, arguably, a 10 – 20 6 exposure in a small animal model may turn out to hardly equate to an extra 1 – 2 6 RDA in humans [177, 178]. In a recent study using the Apc Min mouse model of CRC, both semisynthetic diets with low and high vitamin contents (1/3 of the BDR and 5 6 BDR, respectively), which also contained 1/3 of the BDR and 2 6 BDR folic acid, respectively, significantly increased the number of small intestinal polyps [179]. Furthermore, in two genetic models of CRC ( Apc Min and Apc +/ – 6 Msh2 – / – mice), moderate dietary folate deficiency enhanced, whereas modest levels of folic acid supplementation (four to ten times the BDR) suppressed, the development and progression of CRC, if folate intervention was started before the establishment of neoplastic foci in the intestine [180, 181]. If, however, folate intervention was started after the establishment of neoplastic foci, the same degree of folate deficiency inhibited the progression and induced regression of the established neoplastic foci [180, 181]. A potential tumor promoting effect of folic acid supplementation on the established neoplastic foci could not be clearly determined in these studies because of the inherent limitations associated with these genetic models [180, 181]. Therefore, these observations suggest that the timing of folate intervention is critical in providing an effective and safe chemopreventive effect on colorectal carcinogenesis. Folate deficiency has an inhibitory effect whereas folate supplementation has a promoting effect on the progression of established colorectal neoplasms. In contrast, folate deficiency in the normal colorectum appears predispose it to neoplastic transformation, and modest supplemental levels of folate suppress the development of neoplasms in the normal colorectum. Some animal studies have also shown that dietary folate deficiency inhibits, and not suppresses, the development of breast cancer in rats [182 – 184] in contrast to the inverse association between folate status and breast cancer risk observed in epidemiologic studies [185]. Data from animal studies and clinical observations suggest that folate possesses dual modulatory effects on CRC devel- opment and progression depending on the timing and dose of folate intervention (Fig. 3) [5, 6, 58, 84, 169]. Folate deficiency has an inhibitory effect whereas folate supplementation has a promoting effect on the progression of established colorectal neoplasms (Fig. 3). In contrast, folate deficiency in normal colorectal mucosa appears to predispose it to neoplastic transformation, and modest levels of folic acid supplementation (4 – 10 times above the BDR) suppress, whereas supraphysiological supplemental doses enhance the development of CRC in normal colorectal mucosa (Fig. 3). Are there biologically plausible explanations for these seemingly paradoxical and contradictory epidemiologic, animal, and clinical observations concerning the dual role of folate in CRC development and progression? There exist several biologically plausible mechanisms by which folate deficiency increases, whereas folate supplementation reduces, the risk of CRC in normal colorectal epithelial cells [50, 58, 59, 123]. As an essential cofactor for the de novo biosynthesis of purines and thymidylate (Fig. 2), folate plays an important role in DNA synthesis, stability and integrity, and repair, aberrations of which have been impli- cated in colorectal carcinogenesis [58, 59, 123]. Indeed, a large body of evidence from in vitro , animal and human studies indicates that folate deficiency is associated with DNA strand breaks, chromosomal and genomic instability, uracil misincorporation, impaired DNA repair, and increased mutations [58, 59, 121 – 123, 186]. Furthermore, this body of evidence indicates that folate supplementation can correct some of these defects induced by folate deficiency, and ensures DNA fidelity, maintains DNA integrity and stability, and optimizes DNA repair by providing nucleotide precursors for DNA synthesis and replication [58, 59, 121 – 123, 186]. Therefore, the effect of folate deficiency and supplementation on the DNA synthesis pathway in the normal colorectum have been generally considered to be the primary mechanism by which folate deficiency predisposes it to neoplastic transformation and folate supplementation prevents or suppresses neoplastic transformation, respectively (Fig. 3) [58, 59, 121 – 123, 186]. Another proposed mechanism by which folate deficiency enhances the development of cancer in the colorectum is through an induction of genomic DNA hypomethylation [90]. It has been proposed that a mechanism by which folate supplementation may protect against the development of cancer in the colorectum is through a protection against genomic DNA hypomethylation [90]. This mechanism is based on the biochemical function of folate in mediating one-carbon transfer for the provision of SAM, the primary methyl group donor for most biological methylation reactions, including that of DNA (Fig. 2) and on evidence from animal experiments that demonstrated that diets deficient in different combinations of methyl group donors (choline, folate, methionine, and vitamin B 12 ) consistently induce genomic and site and gene-specific DNA hypomethylation [49, 50]. DNA methylation of cytosine located within the cyto- sine-guanine (CpG) dinucleotide sequences is an important epigenetic determinant in gene expression (an inverse relation), in the maintenance of DNA integrity and stability, in chromosomal modifications, and in the development of mutations [124, 187]. In contrast to methylated CpG sites in the CpG-poor bulk of the genome and unmethylated CpG islands in normal cells, cancer cells simultaneously harbor widespread loss of methylation in the CpG-depleted regions where most CpG dinucleotides should be methylated and gains in methylation of CpG islands in gene promoter regions (Fig. 4) [124, 187]. Global hypomethylation is an early, and consistent, event in colorectal carcinogenesis [124, 187]. Global hypomethylation of the coding and noncoding regions and demethylation of repetitive DNA sequences contribute to the development of cancer through the following mechanisms: chromosomal instability; increased mutations; reactivation of intra- genomic parasitic sequences that could be transcribed and moved to other sites where they could disrupt normal cellular genes; mitotic recombination leading to loss of heterozygosity and promotion of rearrangements; aneuploidy; loss of imprinting; and up-regulation of protooncogenes (Fig. 4) [188]. However, animal studies have shown that genomic demethylation may protect against some cancers ( e. g. , intestinal tumors) [189, 190] but may promote chromosomal instability and increase the risk of cancer in other tissues ( e. g. , lymphoma, sarcoma) [191, 192]. Methylation at promoter CpG islands is an important mechanism of silencing transcription in carcinogenesis; the affected genes are silenced and their function is stably lost in a clonally propagated fashion (Fig. 4) [187, 188, 193]. Many genes inactivated by promoter CpG methylation in carcinogenesis have classic tumor-suppressor function or play critical roles in cell cycle control, repair of DNA damage, apoptosis, differentiation, angiogenesis, metastasis, growth – factor response, drug resistance, and detoxifi- cation [187]. Promoter CpG islands of over 60% of tumor suppressor and mismatch repair genes have been observed to be methylated in cancer [187]. Another means by which CpG methylation may contribute to carcinogenesis is the hypermutability of methylated cytosine. CpG dinucleotides within certain genes are not only the sites of DNA methylation but also mutational hot spots for human cancers [194]. The majority of mutations observed in CpG sites are cytosine-to-thymine transitions mediated by the spontaneous deamination of 5-methylcyto- sine to thymine, by the enzymatic deamination of 5-methyl- cytosine to thymine by DNA methyltransferase (DNMT), and by the enzymatic deamination of unmethylated cytosine to uracil and subsequent methylation of uracil to thymine by DNMT [194]. CpG sites have been shown to act as hot spots for germline mutations, contributing to 30% of all point mutations in the germ line, and for acquired somatic mutations that lead to cancer [195]. For example, methylated CpG sites in the p53 tumor suppressor coding region contribute to as many as 50% of all inactivating mutations in CRC and to 25% of cancers in general [195]. The portfolio of evidence from animal, human, and in vitro studies collectively suggests that the effect of folate deficiency on DNA methylation is highly complex and vari- able. It appears to be gene and site-specific and depends on species, cell type, target organ, and stage of transformation as well as on the degree and duration of folate depletion [49, 50]. In particular, folate deficiency appears to be unable to induce genomic and gene-specific DNA hypomethylation in the colorectum on a consistent and predict- able manner [49, 50]. This may be related to the fact that modulation of SAM and SAH in the colorectum is particularly resistant to folate depletion [49, 50]. Collectively, currently available evidence indicates that genomic DNA hypomethylation in the colorectum is not a probable mechanism by which folate deficiency enhances colorectal ...

Context 18

... that no single kind of study is considered to be definitive. Instead, all of the different types of studies that are used to investigate the link between nutrition and cancer are considered along- side each other, without favoring evidence from one type over another. In support of the portfolio approach, systema- tic comparisons of the results of randomized intervention studies with observational evidence in several clinical situations have shown that observational data from well-con- ducted studies do not appear to produce biased results compared to randomized interventions [205, 206]. Furthermore, the importance of experimental studies that contribute to understanding mechanisms that might under- lie any observed association between a dietary factor and cancer and might bear on the inference of causation has been increasingly recognized and appreciated in the field of nutrition and cancer. Epidemiologic and experimental evidence indicating a causal association between a dietary factor and cancer is strengthened when a biologic pathway or mechanism by which colorectal carcinogenesis may be modified is identified and when this mechanism is biologically plausible [204]. It can be argued that epidemiologic data, however strong and consistent, are an inadequate basis for any definite judgment of causality unless supported by mechanistic evidence [204]. Recent advances in molecular epidemiology have added another dimension to the already complex field of nutrition and cancer. Recently identified and characterized single nucleotide polymorphisms and other genetic and epigenetic variants of genes that are involved in absorption, transport, metabolism, and excretion of nutrients have been shown not only to modify cancer risk but also to significantly modulate the effect of nutrients and related compounds on cancer risk [207]. This emerging important topic in the field of nutrition and cancer, termed “gene – nutrient interactions” in carcinogenesis, has a very significant implication in designing and interpreting data from observational epidemiologic and intervention studies. Although individuals are subjected to the same level of nutritional exposure, systemic, and target tissue bioavailability of nutrients and their metabolites, as well as their functional effects in the target tissue, might be vastly different because of genetic and epigenetic variations. Genetic and epigenetic susceptibility to cancer and their interaction with diets and other environmental exposures have not been incorporated into the study design of and interpretation of data from previously published epidemiologic and intervention studies. The precise nature and magnitude of gene – nutrient interactions in carcinogenesis are yet to be clearly defined. It appears that overall diet, rather than individual factors, plays the more important role in the development of CRC, thus underscoring the importance of as yet undetermined interactions among dietary components in the development of cancer. It is likely that dietary factors or components do not act in isolation but as part of a biological action package [208]. The major difficulty in establishing a relationship between diet and cancer and in translating observations from nutritional epidemiology into progress in cancer prevention has been due to inability to identify all relevant dietary components that act coordinately to modulate cancer risk and due to inability to identify the other relevant non-nutritional factors that interact with dietary components to modify cancer risk [208]. What can we conclude about the role of folate in CRC development and progression from the seemingly paradoxical and contradictory epidemiologic, animal, and clinical studies? Currently available evidence from epidemiologic, laboratory animal, and intervention studies does not unequivocally support the role of folate in the development and progression of CRC. Furthermore, the precise nature and magnitude of the relationship of CRC with folate have not been clearly defined. However, when the whole body or portfolio of evidence from these studies is analyzed critically, the overall conclusion supports the inverse association between folate status and CRC risk. Definitive answers to questions about folate and CRC are probably beyond the reach of both observational epidemiologic studies and randomized controlled trials [204]. It is clear that folate appears to possess dual modulatory effects on colorectal carcinogenesis depending on the timing and dose of folate intervention (Fig. 3). Folate deficiency has an inhibitory effect whereas folate supplementation has a promoting effect on the progression of established colorectal neoplasms (Fig. 3). In contrast, folate deficiency in normal col- orectal mucosa appears to predispose it to neoplastic transformation, and modest levels of folic acid supplementation suppress, whereas supraphysiologic supplemental doses enhance the development of CRC in normal colorectal mucosa (Fig. 3). Several potential mechanisms relating to the disruption of the known biochemical function of folate (mediating the transfer of one-carbon moieties and consequent DNA synthesis and methylation) exist to support the dual modulatory role of folate in colorectal carcinogenesis (Fig. 3). As discussed briefly above, evidence for a protective effect of folate supplementation on NTD [7, 8] was considered to be sufficiently conclusive and led to mandatory folic acid fortification in the US [11] and Canada [12] in 1998. Folic acid fortification has already significantly improved folate status and has had a substantial beneficial effect on the original target, NTD, in the US and Canada [21 – 26]. Mandatory folic acid fortification is probably the most important science-drive intervention in nutrition and public health in decades [209]. However, the possibility remains that certain segments of the exposed population may benefit less and may even experience some adverse effects from an increased folic acid intake. Over the past few years, the US and Canadian populations have been exposed to a significant increase in folate intake, for which essentially no data on safety exist [13]. No studies have been done to look directly or even indirectly for the adverse effects of greatly increased folate intakes [13]. In addition to the drastic increase in dietary folate intake from mandatory folic acid fortification, 30 – 40% of the North American population consume supplemental folic acid for several possible but as yet unproven health benefits [35]. Whether or not possible deleterious effects of folic acid supplementation ( e. g. , cancer-promoting effect on established preneoplastic and neoplastic lesions) outweigh the known and potential health benefits ( e. g. , prevention of atherosclerosis and NTD; improvement of cognitive function; cancer prevention in normal tissues free of preneoplastic and neoplastic foci) is largely unknown at present. Folate is generally regarded as safe [210] and may become the ultimate functional food component for disease prevention [211]. The potential masking effect of folic acid on vitamin B 12 deficiency, especially in the elderly, has been the only major concern of folic acid fortification and supplementation [13]. However, an emerging body of evidence suggests that folate supplementation may be associated with other potentially serious adverse effects [6]. These include: the occurrence of resistance or tolerance to antifolate-based chemotherapy and anti-inflammatory and antiseizure drugs; decreased natural killer cell cytotoxicty; accelerated cognitive decline in older subjects; increased twin pregnancies; and genetic selections of disease alleles ( e. g. , MTHFR C677T) that predispose individuals to chronic diseases if exposed to low folate status [6, 211 – 220]. Folic acid, the synthetic, fully oxidized form of folate used in fortification and supplementation, is normally reduced and methylated by the intestine before it is released into the circulation as 5-methylTHF; consequently, the latter form is the sole circulating form of folate under normal conditions [2]. However, studies show that this absorption and biotransformation process is saturated at doses in the region of 400 l g folic acid or less [221]. At higher doses, synthetic folic acid is also transported into the blood and may enter in large quantities. Consumption of folic acid A 200 l g have shown to lead to the appearance of unmetabolized folic acid in the serum. Although compelling data about possible antagonistic activities of this fully oxidized form of folate in tissues is lacking, there nevertheless exist concerns about the effect of long-term exposure of cells to unmetabolized folic acid [222]. In this regard, Troen et al. [212] have recently reported that 78% of 104 postmenopausal women 60 – 75 years of age had detectable levels of folic acid in plasma, which was associated with an approxi- mately 23% decrease in natural killer cell cytotoxicity inde- pendent of plasma 5-methylTHF and total folate concentrations. Among participants in a large ( n = 2928) trial of folic acid supplementation during pregnancy, women who received 5 mg folic acid/day had a 70% increased risk of total cancer mortality compared with those not on supplementation (HR = 1.70; 95% CI = 1.06 – 2.72) [168]. In this study, the risk of death from breast cancer in women taking 5 mg folic acid/day was twice that of women taking no supplementation, albeit nonsignificant (HR = 2.02; 95% CI = 0.88 – 4.72) [168]. Furthermore, in line with this observation, a recent the Prostate, Lung, Colorectal, and Ovarian Cancer Screening Trial ( n = 25, 400 postmenopausal women) has reported a significant positive association between total folate and supplemental folic acid intakes, but not food folate, and breast cancer risk [85]. Given the new evidence that cells of the absorptive mucosa in the small intestine may not be the primary site for folic acid ...

Context 19

... regions (Fig. 4) [124, 187]. Global hypomethylation is an early, and consistent, event in colorectal carcinogenesis [124, 187]. Global hypomethylation of the coding and noncoding regions and demethylation of repetitive DNA sequences contribute to the development of cancer through the following mechanisms: chromosomal instability; increased mutations; reactivation of intra- genomic parasitic sequences that could be transcribed and moved to other sites where they could disrupt normal cellular genes; mitotic recombination leading to loss of heterozygosity and promotion of rearrangements; aneuploidy; loss of imprinting; and up-regulation of protooncogenes (Fig. 4) [188]. However, animal studies have shown that genomic demethylation may protect against some cancers ( e. g. , intestinal tumors) [189, 190] but may promote chromosomal instability and increase the risk of cancer in other tissues ( e. g. , lymphoma, sarcoma) [191, 192]. Methylation at promoter CpG islands is an important mechanism of silencing transcription in carcinogenesis; the affected genes are silenced and their function is stably lost in a clonally propagated fashion (Fig. 4) [187, 188, 193]. Many genes inactivated by promoter CpG methylation in carcinogenesis have classic tumor-suppressor function or play critical roles in cell cycle control, repair of DNA damage, apoptosis, differentiation, angiogenesis, metastasis, growth – factor response, drug resistance, and detoxifi- cation [187]. Promoter CpG islands of over 60% of tumor suppressor and mismatch repair genes have been observed to be methylated in cancer [187]. Another means by which CpG methylation may contribute to carcinogenesis is the hypermutability of methylated cytosine. CpG dinucleotides within certain genes are not only the sites of DNA methylation but also mutational hot spots for human cancers [194]. The majority of mutations observed in CpG sites are cytosine-to-thymine transitions mediated by the spontaneous deamination of 5-methylcyto- sine to thymine, by the enzymatic deamination of 5-methyl- cytosine to thymine by DNA methyltransferase (DNMT), and by the enzymatic deamination of unmethylated cytosine to uracil and subsequent methylation of uracil to thymine by DNMT [194]. CpG sites have been shown to act as hot spots for germline mutations, contributing to 30% of all point mutations in the germ line, and for acquired somatic mutations that lead to cancer [195]. For example, methylated CpG sites in the p53 tumor suppressor coding region contribute to as many as 50% of all inactivating mutations in CRC and to 25% of cancers in general [195]. The portfolio of evidence from animal, human, and in vitro studies collectively suggests that the effect of folate deficiency on DNA methylation is highly complex and vari- able. It appears to be gene and site-specific and depends on species, cell type, target organ, and stage of transformation as well as on the degree and duration of folate depletion [49, 50]. In particular, folate deficiency appears to be unable to induce genomic and gene-specific DNA hypomethylation in the colorectum on a consistent and predict- able manner [49, 50]. This may be related to the fact that modulation of SAM and SAH in the colorectum is particularly resistant to folate depletion [49, 50]. Collectively, currently available evidence indicates that genomic DNA hypomethylation in the colorectum is not a probable mechanism by which folate deficiency enhances colorectal carcinogenesis [49, 50]. In contrast, folate supplementation appears to significantly increase the extent of genomic and site-specific DNA methylation in animal and human studies [49, 50]. Dietary folic acid supplementation up to 20 6 BDR significantly reversed DMH-induced DNA hypomethylation within a coding region of the p53 tumor suppressor gene in rat colon in a dose-dependent and site-specific manner [92] in the absence of change in genomic DNA methylation [93]. Folic acid supplementation (286 – 516 l g/day 6 3 wk) [142] or 5-methylTHF (15 mg/day 6 8 wk) [145] was able to normalize pre-existing genomic DNA hypomethylation in peripheral leukocytes in humans. Folic acid supplementation at 12.5 – 25 6 BDR for 3 – 12 months significantly increased the extent of colonic mucosal genomic DNA methylation in subjects with resected colorectal adenomas or cancer [146 – 148]. Even a physiological dose of folic acid (400 l g/day) for 10 wk increased genomic DNA methylation in lymphocytes (by 31%; p = 0.05) and in colonic mucosa (by 25%; p = 0.09) in patients with colorectal adenomas [152]. Collectively, currently available evidence indicates that folate supplementation appears to be able to reverse pre-existing genomic DNA hypomethylation and to increase the extent of genomic DNA methylation above the pre-existing level [49, 50]. Therefore, prevention or rever- sal of genomic DNA hypomethylation may be a mechanism by which folate supplementation suppresses neoplastic transformation in the colorectum (Fig. 3). In preneoplastic and neoplastic cells where DNA replication and cell division are occurring at an accelerated rate, folate depletion causes ineffective DNA synthesis, resulting in inhibition of tumor growth and progression (Fig. 3), which is the basis for cancer chemotherapy using antifolate agents ( e. g. , methotrexate) and 5-fluorouracil [58, 59, 123]. Thus, this is the most likely mechanism by which folate deficiency inhibits the progression of the established preneoplastic neoplastic foci in the colorectum. Another possible mechanism is that folate deficiency may reverse CpG promoter methylation of tumor suppressor and other anticancer genes involved in colorectal carcinogenesis, thereby reactivating these genes. However, there is currently no experimental evidence to support this theo- retical possibility. As discussed earlier, folate deficiency appears to be unable to induce genomic and gene-specific DNA hypomethylation in the colorectum [49, 50]. Furthermore, in a recent in vitro study, folate deficiency induced a significant reduction in genomic and site-specific DNA methylation in untransformed NIH/3T3 and CHO-K1 mammalian cells but not in HCT116 and Caco2 human colon adenocarcinoma cells [196]. In this study, folate deficiency did not produce significant changes in the promoter CpG island methylation of the p16 tumor suppressor gene and the MLH1 mismatch repair gene in HCT116 and Caco- 2 cells [196]. However, certain sites in the promoter CpG island of the ER gene were associated with modest, albeit statistically significant, changes in CpG methylation in response to folate deficiency, which were not associated with significant functional consequences [196]. In line with this observation, another study showed that HCT116 cells lacking DNMT1 exhibited only a modest 20% decrease in the overall genomic DNA methylation despite the markedly decreased cellular DNMT activity [197]. In this model, although juxtacentromeric satellites became significantly demethylated, centromeric satellite loci, and the promoter CpG island of the p16 gene remained fully methylated [197]. Only when both the DNMT1 and DNMT3b genes were disrupted, genomic DNA methylation was reduced by A 95% and significant hypomethylation of satellite sequences and several promoter CpG islands, including that of the p16 gene, was observed [198]. These observations suggest that it may be extremely difficult to reverse DNA methylation in cancer cell lines such as HCT116. The fact that an almost complete abolishment of DNMT activity by disruption of both the DNMT1 and DNMT3b genes is required to produce significant DNA hypomethylation in HCT116 cells [198] suggest that folate deficiency alone is unlikely to be a sufficient predisposing condition to produce significant DNA hypomethylation in colon cancer cells. Mechanistically, the most likely mechanism by which folic acid supplementation may promote the progression of established preneoplastic and neoplastic lesions in the col- orecutm is provision of nucleotide precursors to rapidly replicating neoplastic cells for accelerated proliferation and growth [58, 59, 123]. Another possible mechanism by which folic acid supplementation may promote the progression of preneoplastic or neoplastic foci in the colorectum may be de novo methylation of promoter CpG islands of tumor suppressor genes and other critical genes involved in colorectal carcinogenesis with consequent gene inactiva- tion leading to tumor progression (Fig. 3). This potential epigenetic mechanism of tumor progression is supported by recent animal studies using viable yellow agouti mice that unequivocally have demonstrated that maternal dietary methyl group supplementation containing folic acid permanently alters phenotypic coat color of the offspring via increased methylation at the promoter CpG site of the agouti gene [199 – 201]. However, it is unknown at present whether this de novo methylation of promoter CpG islands can happen with folic acid supplementation alone, whether it is operative in normal or neoplastic tissues or both, whether this effect is associated with folic acid supplementation provided in utero only, in postpartum period, or in adulthood, and whether it is tissue and gene-specific. Another possible means by which folic acid supplementation may promote colorectal carcinogenesis may be through hypermutability of methylated cytosines in CpG dinucleotides (Fig. 3). Methylated CpG sites are mutational hot spots for human cancer as described earlier [194]. The majority of mutations observed in CpG sites are cytosine- to-thymine transitions mediated by the spontaneous deamination of 5-methylcytosine to thymine and by the enzymatic deamination of 5-methylcytosine to thymine by DNMT [194]. Therefore, there is a possibility that de novo methylation of cytosines in CpG sites in critical genes involved in colorectal carcinogenesis may create mutational hot spots, leading to inactivating mutations of these ...

Context 20

... 50]. In contrast, folate supplementation appears to significantly increase the extent of genomic and site-specific DNA methylation in animal and human studies [49, 50]. Dietary folic acid supplementation up to 20 6 BDR significantly reversed DMH-induced DNA hypomethylation within a coding region of the p53 tumor suppressor gene in rat colon in a dose-dependent and site-specific manner [92] in the absence of change in genomic DNA methylation [93]. Folic acid supplementation (286 – 516 l g/day 6 3 wk) [142] or 5-methylTHF (15 mg/day 6 8 wk) [145] was able to normalize pre-existing genomic DNA hypomethylation in peripheral leukocytes in humans. Folic acid supplementation at 12.5 – 25 6 BDR for 3 – 12 months significantly increased the extent of colonic mucosal genomic DNA methylation in subjects with resected colorectal adenomas or cancer [146 – 148]. Even a physiological dose of folic acid (400 l g/day) for 10 wk increased genomic DNA methylation in lymphocytes (by 31%; p = 0.05) and in colonic mucosa (by 25%; p = 0.09) in patients with colorectal adenomas [152]. Collectively, currently available evidence indicates that folate supplementation appears to be able to reverse pre-existing genomic DNA hypomethylation and to increase the extent of genomic DNA methylation above the pre-existing level [49, 50]. Therefore, prevention or rever- sal of genomic DNA hypomethylation may be a mechanism by which folate supplementation suppresses neoplastic transformation in the colorectum (Fig. 3). In preneoplastic and neoplastic cells where DNA replication and cell division are occurring at an accelerated rate, folate depletion causes ineffective DNA synthesis, resulting in inhibition of tumor growth and progression (Fig. 3), which is the basis for cancer chemotherapy using antifolate agents ( e. g. , methotrexate) and 5-fluorouracil [58, 59, 123]. Thus, this is the most likely mechanism by which folate deficiency inhibits the progression of the established preneoplastic neoplastic foci in the colorectum. Another possible mechanism is that folate deficiency may reverse CpG promoter methylation of tumor suppressor and other anticancer genes involved in colorectal carcinogenesis, thereby reactivating these genes. However, there is currently no experimental evidence to support this theo- retical possibility. As discussed earlier, folate deficiency appears to be unable to induce genomic and gene-specific DNA hypomethylation in the colorectum [49, 50]. Furthermore, in a recent in vitro study, folate deficiency induced a significant reduction in genomic and site-specific DNA methylation in untransformed NIH/3T3 and CHO-K1 mammalian cells but not in HCT116 and Caco2 human colon adenocarcinoma cells [196]. In this study, folate deficiency did not produce significant changes in the promoter CpG island methylation of the p16 tumor suppressor gene and the MLH1 mismatch repair gene in HCT116 and Caco- 2 cells [196]. However, certain sites in the promoter CpG island of the ER gene were associated with modest, albeit statistically significant, changes in CpG methylation in response to folate deficiency, which were not associated with significant functional consequences [196]. In line with this observation, another study showed that HCT116 cells lacking DNMT1 exhibited only a modest 20% decrease in the overall genomic DNA methylation despite the markedly decreased cellular DNMT activity [197]. In this model, although juxtacentromeric satellites became significantly demethylated, centromeric satellite loci, and the promoter CpG island of the p16 gene remained fully methylated [197]. Only when both the DNMT1 and DNMT3b genes were disrupted, genomic DNA methylation was reduced by A 95% and significant hypomethylation of satellite sequences and several promoter CpG islands, including that of the p16 gene, was observed [198]. These observations suggest that it may be extremely difficult to reverse DNA methylation in cancer cell lines such as HCT116. The fact that an almost complete abolishment of DNMT activity by disruption of both the DNMT1 and DNMT3b genes is required to produce significant DNA hypomethylation in HCT116 cells [198] suggest that folate deficiency alone is unlikely to be a sufficient predisposing condition to produce significant DNA hypomethylation in colon cancer cells. Mechanistically, the most likely mechanism by which folic acid supplementation may promote the progression of established preneoplastic and neoplastic lesions in the col- orecutm is provision of nucleotide precursors to rapidly replicating neoplastic cells for accelerated proliferation and growth [58, 59, 123]. Another possible mechanism by which folic acid supplementation may promote the progression of preneoplastic or neoplastic foci in the colorectum may be de novo methylation of promoter CpG islands of tumor suppressor genes and other critical genes involved in colorectal carcinogenesis with consequent gene inactiva- tion leading to tumor progression (Fig. 3). This potential epigenetic mechanism of tumor progression is supported by recent animal studies using viable yellow agouti mice that unequivocally have demonstrated that maternal dietary methyl group supplementation containing folic acid permanently alters phenotypic coat color of the offspring via increased methylation at the promoter CpG site of the agouti gene [199 – 201]. However, it is unknown at present whether this de novo methylation of promoter CpG islands can happen with folic acid supplementation alone, whether it is operative in normal or neoplastic tissues or both, whether this effect is associated with folic acid supplementation provided in utero only, in postpartum period, or in adulthood, and whether it is tissue and gene-specific. Another possible means by which folic acid supplementation may promote colorectal carcinogenesis may be through hypermutability of methylated cytosines in CpG dinucleotides (Fig. 3). Methylated CpG sites are mutational hot spots for human cancer as described earlier [194]. The majority of mutations observed in CpG sites are cytosine- to-thymine transitions mediated by the spontaneous deamination of 5-methylcytosine to thymine and by the enzymatic deamination of 5-methylcytosine to thymine by DNMT [194]. Therefore, there is a possibility that de novo methylation of cytosines in CpG sites in critical genes involved in colorectal carcinogenesis may create mutational hot spots, leading to inactivating mutations of these genes. A cause-and-effect relationship between folate and CRC is difficult to establish. Because of inherent limitations associated with study design, the results from epidemiologic, animal, and interventional studies examining this relationship have been inconsistent and conflicting. In clinical medicine, the best evidence has been considered to come from well designed and executed double-blind randomized controls trials, which minimize a variety of biases. This has resulted in a clear hierarchy of evidence that is weighted heavily toward randomized controlled trials. Evidence from randomized controlled trials is thought to supersede evidence from other sources such as observational studies. The field of nutritional epidemiology has also followed this traditional approach and considered correlation, case-control, and prospective observational epidemiologic studies and intervention trials as a spectrum of increasing weight of evidence for or against a relationship between dietary factors and cancer risk [202]. Thus, general conclusions and recommendations regarding the effect of dietary factors on cancer risk have relied heavily on data from large prospective studies and randomized, controlled intervention human trials [202]. This traditional approach to grading epidemiologic evidence concerning the relationship between dietary factors and cancer risk has recently been challenged [203, 204]. As cancer develops over decades, if not a lifetime, single clinical trials, which normally last up to 5 years, cannot address the whole span of cancer development. In addition, randomized controlled trials tend to use uncharacteristic levels of exposure. Furthermore, the dietary, nutritional, and physi- cal activity exposures involved are complex and interre- lated, making them difficult to manipulate in a controlled fashion. Even if a difference in outcome followed such a clinical intervention, it would not necessarily indicate that reproducing the intervention under other conditions would cause similar outcomes. Thus, it has been argued that draw- ing a definitive conclusion concerning the effect of dietary factors on cancer risk mainly from randomized, controlled intervention human trials is probably not the right paradigm of nutritional epidemiology [203, 204]. Rather, it has been articulated that the totality or “portfolio” of evidence from observational and intervention studies as well as animal and in vitro experiments must be analyzed for this purpose [203, 204]. The portfolio approach does not set out a hierarchy of evidence. Instead, it recognizes that all types of evidence have advantages and disadvantages. This means that no single kind of study is considered to be definitive. Instead, all of the different types of studies that are used to investigate the link between nutrition and cancer are considered along- side each other, without favoring evidence from one type over another. In support of the portfolio approach, systema- tic comparisons of the results of randomized intervention studies with observational evidence in several clinical situations have shown that observational data from well-con- ducted studies do not appear to produce biased results compared to randomized interventions [205, 206]. Furthermore, the importance of experimental studies that contribute to understanding mechanisms that might under- lie any observed association between a dietary factor and cancer and might bear on the inference of causation has been ...

Context 21

... of CRC showed that a moderate degree of folate deficiency promoted, whereas modest levels of folic acid supplementation up to 4 6 the basal daily requirement (BDR) for rodents inhibited, the development of CRC [90, 93]. There was a suggestion that a very high supplemental dose of folic acid (20 6 the BDR] might promote the progression of microscopic neoplastic foci to CRC [93]. In support of this latter finding, animal studies using a metabolite of DMH, azoxy- methane (AOM), showed that folic acid supplementation exceeding the BDR by 1000 – 10 000 6 enhanced colorectal carcinogenesis in rats [170 – 174]. Therefore, it appears that folate modulates colorectal carcinogenesis in chemical carcinogen rodent models over a wide range of dietary intakes. Folate deficiency of a moderate degree enhances colorectal carcinogenesis whereas modest levels of folate supplementation above the BDR suppress colorectal tumorigenesis. Supraphysiologic levels of folate supplementation do not appear to confer additional protection and, in some cases, may enhance colorectal carcinogenesis. The implication of this issue is important because the optimal dose of folate supplementation must be determined for folate chemoprevention to be effective and safe in humans. Although some similarities do exist, tumor development in chemical rodent models of CRC differs in several important histological, clinical, and molecular genetic aspects from that observed in humans [175, 176]. Therefore, any extrapolation of the observations from these models to human situations should be made very cautiously. Whether or not the supplemental doses of folic acid used in these animal studies can be directly extrapolated to intake levels in humans is a highly contentious and controversial issue at present because of inherent differences in folate absorption and metabolism between rodents and humans [177, 178]. Recent evidence suggests that animals, unlike humans, have a comparatively high DHFR activity [177, 178]. Consequently, if assessing the impact of systemic exposure of unmetabolized folic acid, animals would have to be orally dosed with a much greater than prorata amount of folic acid in order to elicit the same circulating serum/plasma concentrations of unmetabolized folic acid [177, 178]. Therefore, it may be a gross mistake to dismiss the effects of folic acid exposure in animal studies on the grounds that the experimental folic acid intake (multiple of animal BDR) would translate to an unlikely intake in humans [177, 178]. Therefore, arguably, a 10 – 20 6 exposure in a small animal model may turn out to hardly equate to an extra 1 – 2 6 RDA in humans [177, 178]. In a recent study using the Apc Min mouse model of CRC, both semisynthetic diets with low and high vitamin contents (1/3 of the BDR and 5 6 BDR, respectively), which also contained 1/3 of the BDR and 2 6 BDR folic acid, respectively, significantly increased the number of small intestinal polyps [179]. Furthermore, in two genetic models of CRC ( Apc Min and Apc +/ – 6 Msh2 – / – mice), moderate dietary folate deficiency enhanced, whereas modest levels of folic acid supplementation (four to ten times the BDR) suppressed, the development and progression of CRC, if folate intervention was started before the establishment of neoplastic foci in the intestine [180, 181]. If, however, folate intervention was started after the establishment of neoplastic foci, the same degree of folate deficiency inhibited the progression and induced regression of the established neoplastic foci [180, 181]. A potential tumor promoting effect of folic acid supplementation on the established neoplastic foci could not be clearly determined in these studies because of the inherent limitations associated with these genetic models [180, 181]. Therefore, these observations suggest that the timing of folate intervention is critical in providing an effective and safe chemopreventive effect on colorectal carcinogenesis. Folate deficiency has an inhibitory effect whereas folate supplementation has a promoting effect on the progression of established colorectal neoplasms. In contrast, folate deficiency in the normal colorectum appears predispose it to neoplastic transformation, and modest supplemental levels of folate suppress the development of neoplasms in the normal colorectum. Some animal studies have also shown that dietary folate deficiency inhibits, and not suppresses, the development of breast cancer in rats [182 – 184] in contrast to the inverse association between folate status and breast cancer risk observed in epidemiologic studies [185]. Data from animal studies and clinical observations suggest that folate possesses dual modulatory effects on CRC devel- opment and progression depending on the timing and dose of folate intervention (Fig. 3) [5, 6, 58, 84, 169]. Folate deficiency has an inhibitory effect whereas folate supplementation has a promoting effect on the progression of established colorectal neoplasms (Fig. 3). In contrast, folate deficiency in normal colorectal mucosa appears to predispose it to neoplastic transformation, and modest levels of folic acid supplementation (4 – 10 times above the BDR) suppress, whereas supraphysiological supplemental doses enhance the development of CRC in normal colorectal mucosa (Fig. 3). Are there biologically plausible explanations for these seemingly paradoxical and contradictory epidemiologic, animal, and clinical observations concerning the dual role of folate in CRC development and progression? There exist several biologically plausible mechanisms by which folate deficiency increases, whereas folate supplementation reduces, the risk of CRC in normal colorectal epithelial cells [50, 58, 59, 123]. As an essential cofactor for the de novo biosynthesis of purines and thymidylate (Fig. 2), folate plays an important role in DNA synthesis, stability and integrity, and repair, aberrations of which have been impli- cated in colorectal carcinogenesis [58, 59, 123]. Indeed, a large body of evidence from in vitro , animal and human studies indicates that folate deficiency is associated with DNA strand breaks, chromosomal and genomic instability, uracil misincorporation, impaired DNA repair, and increased mutations [58, 59, 121 – 123, 186]. Furthermore, this body of evidence indicates that folate supplementation can correct some of these defects induced by folate deficiency, and ensures DNA fidelity, maintains DNA integrity and stability, and optimizes DNA repair by providing nucleotide precursors for DNA synthesis and replication [58, 59, 121 – 123, 186]. Therefore, the effect of folate deficiency and supplementation on the DNA synthesis pathway in the normal colorectum have been generally considered to be the primary mechanism by which folate deficiency predisposes it to neoplastic transformation and folate supplementation prevents or suppresses neoplastic transformation, respectively (Fig. 3) [58, 59, 121 – 123, 186]. Another proposed mechanism by which folate deficiency enhances the development of cancer in the colorectum is through an induction of genomic DNA hypomethylation [90]. It has been proposed that a mechanism by which folate supplementation may protect against the development of cancer in the colorectum is through a protection against genomic DNA hypomethylation [90]. This mechanism is based on the biochemical function of folate in mediating one-carbon transfer for the provision of SAM, the primary methyl group donor for most biological methylation reactions, including that of DNA (Fig. 2) and on evidence from animal experiments that demonstrated that diets deficient in different combinations of methyl group donors (choline, folate, methionine, and vitamin B 12 ) consistently induce genomic and site and gene-specific DNA hypomethylation [49, 50]. DNA methylation of cytosine located within the cyto- sine-guanine (CpG) dinucleotide sequences is an important epigenetic determinant in gene expression (an inverse relation), in the maintenance of DNA integrity and stability, in chromosomal modifications, and in the development of mutations [124, 187]. In contrast to methylated CpG sites in the CpG-poor bulk of the genome and unmethylated CpG islands in normal cells, cancer cells simultaneously harbor widespread loss of methylation in the CpG-depleted regions where most CpG dinucleotides should be methylated and gains in methylation of CpG islands in gene promoter regions (Fig. 4) [124, 187]. Global hypomethylation is an early, and consistent, event in colorectal carcinogenesis [124, 187]. Global hypomethylation of the coding and noncoding regions and demethylation of repetitive DNA sequences contribute to the development of cancer through the following mechanisms: chromosomal instability; increased mutations; reactivation of intra- genomic parasitic sequences that could be transcribed and moved to other sites where they could disrupt normal cellular genes; mitotic recombination leading to loss of heterozygosity and promotion of rearrangements; aneuploidy; loss of imprinting; and up-regulation of protooncogenes (Fig. 4) [188]. However, animal studies have shown that genomic demethylation may protect against some cancers ( e. g. , intestinal tumors) [189, 190] but may promote chromosomal instability and increase the risk of cancer in other tissues ( e. g. , lymphoma, sarcoma) [191, 192]. Methylation at promoter CpG islands is an important mechanism of silencing transcription in carcinogenesis; the affected genes are silenced and their function is stably lost in a clonally propagated fashion (Fig. 4) [187, 188, 193]. Many genes inactivated by promoter CpG methylation in carcinogenesis have classic tumor-suppressor function or play critical roles in cell cycle control, repair of DNA damage, apoptosis, differentiation, angiogenesis, metastasis, growth – factor response, drug resistance, and detoxifi- cation [187]. Promoter CpG islands of over 60% of tumor suppressor and ...

Context 22

... is probably not the right paradigm of nutritional epidemiology [203, 204]. Rather, it has been articulated that the totality or “portfolio” of evidence from observational and intervention studies as well as animal and in vitro experiments must be analyzed for this purpose [203, 204]. The portfolio approach does not set out a hierarchy of evidence. Instead, it recognizes that all types of evidence have advantages and disadvantages. This means that no single kind of study is considered to be definitive. Instead, all of the different types of studies that are used to investigate the link between nutrition and cancer are considered along- side each other, without favoring evidence from one type over another. In support of the portfolio approach, systema- tic comparisons of the results of randomized intervention studies with observational evidence in several clinical situations have shown that observational data from well-con- ducted studies do not appear to produce biased results compared to randomized interventions [205, 206]. Furthermore, the importance of experimental studies that contribute to understanding mechanisms that might under- lie any observed association between a dietary factor and cancer and might bear on the inference of causation has been increasingly recognized and appreciated in the field of nutrition and cancer. Epidemiologic and experimental evidence indicating a causal association between a dietary factor and cancer is strengthened when a biologic pathway or mechanism by which colorectal carcinogenesis may be modified is identified and when this mechanism is biologically plausible [204]. It can be argued that epidemiologic data, however strong and consistent, are an inadequate basis for any definite judgment of causality unless supported by mechanistic evidence [204]. Recent advances in molecular epidemiology have added another dimension to the already complex field of nutrition and cancer. Recently identified and characterized single nucleotide polymorphisms and other genetic and epigenetic variants of genes that are involved in absorption, transport, metabolism, and excretion of nutrients have been shown not only to modify cancer risk but also to significantly modulate the effect of nutrients and related compounds on cancer risk [207]. This emerging important topic in the field of nutrition and cancer, termed “gene – nutrient interactions” in carcinogenesis, has a very significant implication in designing and interpreting data from observational epidemiologic and intervention studies. Although individuals are subjected to the same level of nutritional exposure, systemic, and target tissue bioavailability of nutrients and their metabolites, as well as their functional effects in the target tissue, might be vastly different because of genetic and epigenetic variations. Genetic and epigenetic susceptibility to cancer and their interaction with diets and other environmental exposures have not been incorporated into the study design of and interpretation of data from previously published epidemiologic and intervention studies. The precise nature and magnitude of gene – nutrient interactions in carcinogenesis are yet to be clearly defined. It appears that overall diet, rather than individual factors, plays the more important role in the development of CRC, thus underscoring the importance of as yet undetermined interactions among dietary components in the development of cancer. It is likely that dietary factors or components do not act in isolation but as part of a biological action package [208]. The major difficulty in establishing a relationship between diet and cancer and in translating observations from nutritional epidemiology into progress in cancer prevention has been due to inability to identify all relevant dietary components that act coordinately to modulate cancer risk and due to inability to identify the other relevant non-nutritional factors that interact with dietary components to modify cancer risk [208]. What can we conclude about the role of folate in CRC development and progression from the seemingly paradoxical and contradictory epidemiologic, animal, and clinical studies? Currently available evidence from epidemiologic, laboratory animal, and intervention studies does not unequivocally support the role of folate in the development and progression of CRC. Furthermore, the precise nature and magnitude of the relationship of CRC with folate have not been clearly defined. However, when the whole body or portfolio of evidence from these studies is analyzed critically, the overall conclusion supports the inverse association between folate status and CRC risk. Definitive answers to questions about folate and CRC are probably beyond the reach of both observational epidemiologic studies and randomized controlled trials [204]. It is clear that folate appears to possess dual modulatory effects on colorectal carcinogenesis depending on the timing and dose of folate intervention (Fig. 3). Folate deficiency has an inhibitory effect whereas folate supplementation has a promoting effect on the progression of established colorectal neoplasms (Fig. 3). In contrast, folate deficiency in normal col- orectal mucosa appears to predispose it to neoplastic transformation, and modest levels of folic acid supplementation suppress, whereas supraphysiologic supplemental doses enhance the development of CRC in normal colorectal mucosa (Fig. 3). Several potential mechanisms relating to the disruption of the known biochemical function of folate (mediating the transfer of one-carbon moieties and consequent DNA synthesis and methylation) exist to support the dual modulatory role of folate in colorectal carcinogenesis (Fig. 3). As discussed briefly above, evidence for a protective effect of folate supplementation on NTD [7, 8] was considered to be sufficiently conclusive and led to mandatory folic acid fortification in the US [11] and Canada [12] in 1998. Folic acid fortification has already significantly improved folate status and has had a substantial beneficial effect on the original target, NTD, in the US and Canada [21 – 26]. Mandatory folic acid fortification is probably the most important science-drive intervention in nutrition and public health in decades [209]. However, the possibility remains that certain segments of the exposed population may benefit less and may even experience some adverse effects from an increased folic acid intake. Over the past few years, the US and Canadian populations have been exposed to a significant increase in folate intake, for which essentially no data on safety exist [13]. No studies have been done to look directly or even indirectly for the adverse effects of greatly increased folate intakes [13]. In addition to the drastic increase in dietary folate intake from mandatory folic acid fortification, 30 – 40% of the North American population consume supplemental folic acid for several possible but as yet unproven health benefits [35]. Whether or not possible deleterious effects of folic acid supplementation ( e. g. , cancer-promoting effect on established preneoplastic and neoplastic lesions) outweigh the known and potential health benefits ( e. g. , prevention of atherosclerosis and NTD; improvement of cognitive function; cancer prevention in normal tissues free of preneoplastic and neoplastic foci) is largely unknown at present. Folate is generally regarded as safe [210] and may become the ultimate functional food component for disease prevention [211]. The potential masking effect of folic acid on vitamin B 12 deficiency, especially in the elderly, has been the only major concern of folic acid fortification and supplementation [13]. However, an emerging body of evidence suggests that folate supplementation may be associated with other potentially serious adverse effects [6]. These include: the occurrence of resistance or tolerance to antifolate-based chemotherapy and anti-inflammatory and antiseizure drugs; decreased natural killer cell cytotoxicty; accelerated cognitive decline in older subjects; increased twin pregnancies; and genetic selections of disease alleles ( e. g. , MTHFR C677T) that predispose individuals to chronic diseases if exposed to low folate status [6, 211 – 220]. Folic acid, the synthetic, fully oxidized form of folate used in fortification and supplementation, is normally reduced and methylated by the intestine before it is released into the circulation as 5-methylTHF; consequently, the latter form is the sole circulating form of folate under normal conditions [2]. However, studies show that this absorption and biotransformation process is saturated at doses in the region of 400 l g folic acid or less [221]. At higher doses, synthetic folic acid is also transported into the blood and may enter in large quantities. Consumption of folic acid A 200 l g have shown to lead to the appearance of unmetabolized folic acid in the serum. Although compelling data about possible antagonistic activities of this fully oxidized form of folate in tissues is lacking, there nevertheless exist concerns about the effect of long-term exposure of cells to unmetabolized folic acid [222]. In this regard, Troen et al. [212] have recently reported that 78% of 104 postmenopausal women 60 – 75 years of age had detectable levels of folic acid in plasma, which was associated with an approxi- mately 23% decrease in natural killer cell cytotoxicity inde- pendent of plasma 5-methylTHF and total folate concentrations. Among participants in a large ( n = 2928) trial of folic acid supplementation during pregnancy, women who received 5 mg folic acid/day had a 70% increased risk of total cancer mortality compared with those not on supplementation (HR = 1.70; 95% CI = 1.06 – 2.72) [168]. In this study, the risk of death from breast cancer in women taking 5 mg folic acid/day was twice that of women taking no supplementation, albeit nonsignificant (HR = 2.02; 95% CI = 0.88 – ...

Context 23

... promoter of the Axin Fu gene [226]. These investi- gators speculated that “population-based supplementation with folic acid, intended to reduce the incidence of NTD and long presumed to be purely beneficial, may have unin- tended deleterious influences on the establishment of epigenetic gene-regulatory mechanisms during human embryo- nic development.” [199] It has been recently shown that these diet-induced epigenetic changes can be transmitted to future generations [227, 228]. The possibility that folic acid fortification and supplementation during embryogenesis may establish and maintain “hypermethylated CpG islands” DNA methylation pattern in the offspring, leading to silencing of critical tumor suppressor genes is an important issue in the field of folate and cancer and should be investigated. Furthermore, folic acid supplementaion during embryogenesis may methylate cytosines within CpG sequences, ren- dering them mutational hot spots. However, the epigenetic effect of folic acid supplementation during embryogenesis may not all be detrimental because increased DNA methylation of CpG sites present in the coding and noncoding regions and in repetitive DNA sequences may protect against the development of cancer by genomic and chromosomal stability and by suppressing reactivation of intrage- nomic parasitic sequences (Fig. 4) [188]. The predominant epigenetic effect of folic acid supplementation on the developing colorectum is unknown at present. Perhaps the most concerning potential adverse effect of folic acid fortification and supplementation is the cancer- promoting effect. Although folic acid fortification and supplementation may prevent the development of new cancers in persons without preexisting premalignant or neoplastic lesions, it may promote the progression of already existing, undiagnosed premalignant and malignant lesions including those in the colorectum [5, 6, 84, 169, 185]. Population- based folic acid fortification, intended to prevent NTD, and folic acid supplementation, long presumed to be purely beneficial and believed to provide several health benefits, may promote the development and progression of already existing, undiagnosed premalignant lesions (ACF, adenomas) in the colorectum to CRC in the vast majority of the US and Canadian populations, who are not at risk of NTD but have been unintentionally exposed to high amounts of folic acid [6, 13]. In Canada, CRC is the fourth most frequently diagnosed cancer and the second most common cause of can- cer-specific death [229]. In 2004 alone, 19 100 new cases of CRC were diagnosed, and l 40% of these are expected to die within 5 years [229]. In 2004, 8300 deaths were caused by CRC[229]. The lifetime risk of developing CRC is l 6% [161], and treatment costs nearly $6 billion annually in the US [230]. Colorectal adenomas are found in l 25 – 50% of people by 50 years of age in the US and Canada, and the prevalence increases with age [161]. It has been estimated that l 25% of adenomas progress to CRC over 5 – 10 years [161]. In contrast, NTD occur in l 1 of every 1000 births in the US and Canada [231], and spina bifida and anencephaly, the most common NTD, together affect l 4000 pregnancies resulting in 2500 – 3000 US births annually [231]. It is evident from these statistics that the potential effect of folic acid fortification and supplementation on adenoma progression to CRC and on CRC progression to metastasis far outweighs the effect on NTD risk reduction. Thus, the potential cancer-promoting effect of folic acid fortification and supplementation is a legitimate public health concern and needs a careful monitoring. The role of folate has greatly evolved over the past two decades from the prevention of anemia to the prevention of cardiovascular disease and NTD. A large body of evidence suggests that folate may also play a role in the development and progression of cancer. In particular, the portfolio of evidence suggests an inverse association between folate status and the risk of CRC. Given the incidence and mortality of CRC in North America, determining the overall benefits of folic acid fortification and supplementation has major public health implications. As such, preclinical and population- based studies are needed to determine the efficacy, safety, and potential deleterious effects of folic acid fortification and supplementation on CRC and other health outcomes. Although folate appears to be an ideal candidate for CRC chemoprevention given its proven safety and cost [210], the safe and effective dose range of folate supplementation and optimal timing of folate chemoprevention have not been clearly established in humans. An obvious inference from the above discussion is that for folate to be a safe and effective chemopreventive agent against CRC, modest doses of folic acid supplementation should be implemented before the development of precursor lesions in the colorectum or in individuals free of any evidence of neoplastic foci (Fig. 3). However, determining the presence of neoplastic foci in the general population is an almost impossible task. Furthermore, folate might prevent the progression of cer- tain precursor or preneoplastic lesions to frank malignancy but promote the progression of other lesions. What constitu- tes safe precursor or preneoplastic lesions on which folate may exert a protective effect has not yet been established. For example, should folate chemoprevention be started before there is evidence of established premalignant lesions, such as ACF or microscopic adenomas in the colorectum or should folate chemoprevention be started even after these lesions are present? In this regard, animal studies investigating the effects of folic acid supplementation on the progression of ACF, microscopic adenomas, and adenomas are urgently needed. Furthermore, careful dose – response studies are warranted in both animal models and humans to determine safe dose and effective doses of folic acid supplementation. Based on the lack of compelling supportive evidence and on the potential tumor-promoting effect, routine folic acid supplementation should not be recommended as a chemopreventive measure against CRC at present. A more logical approach to folate chemoprevention might be that of targeted chemoprevention in individuals at high risk of developing CRC without evidence of preexisting premalignant lesions or neoplastic foci. For instance, individuals with the MTHFR 677 TT genotype with inadequate folate intake or with significant alcohol consumption have been shown to have an increased risk of CRC [119, 120, 232, 233]. These individuals may therefore benefit from targeted folate chemoprevention, provided that they are free of preneoplastic or neoplastic foci in the colorectum and other target ...

Context 24

... contribute to the development of cancer through the following mechanisms: chromosomal instability; increased mutations; reactivation of intra- genomic parasitic sequences that could be transcribed and moved to other sites where they could disrupt normal cellular genes; mitotic recombination leading to loss of heterozygosity and promotion of rearrangements; aneuploidy; loss of imprinting; and up-regulation of protooncogenes (Fig. 4) [188]. However, animal studies have shown that genomic demethylation may protect against some cancers ( e. g. , intestinal tumors) [189, 190] but may promote chromosomal instability and increase the risk of cancer in other tissues ( e. g. , lymphoma, sarcoma) [191, 192]. Methylation at promoter CpG islands is an important mechanism of silencing transcription in carcinogenesis; the affected genes are silenced and their function is stably lost in a clonally propagated fashion (Fig. 4) [187, 188, 193]. Many genes inactivated by promoter CpG methylation in carcinogenesis have classic tumor-suppressor function or play critical roles in cell cycle control, repair of DNA damage, apoptosis, differentiation, angiogenesis, metastasis, growth – factor response, drug resistance, and detoxifi- cation [187]. Promoter CpG islands of over 60% of tumor suppressor and mismatch repair genes have been observed to be methylated in cancer [187]. Another means by which CpG methylation may contribute to carcinogenesis is the hypermutability of methylated cytosine. CpG dinucleotides within certain genes are not only the sites of DNA methylation but also mutational hot spots for human cancers [194]. The majority of mutations observed in CpG sites are cytosine-to-thymine transitions mediated by the spontaneous deamination of 5-methylcyto- sine to thymine, by the enzymatic deamination of 5-methyl- cytosine to thymine by DNA methyltransferase (DNMT), and by the enzymatic deamination of unmethylated cytosine to uracil and subsequent methylation of uracil to thymine by DNMT [194]. CpG sites have been shown to act as hot spots for germline mutations, contributing to 30% of all point mutations in the germ line, and for acquired somatic mutations that lead to cancer [195]. For example, methylated CpG sites in the p53 tumor suppressor coding region contribute to as many as 50% of all inactivating mutations in CRC and to 25% of cancers in general [195]. The portfolio of evidence from animal, human, and in vitro studies collectively suggests that the effect of folate deficiency on DNA methylation is highly complex and vari- able. It appears to be gene and site-specific and depends on species, cell type, target organ, and stage of transformation as well as on the degree and duration of folate depletion [49, 50]. In particular, folate deficiency appears to be unable to induce genomic and gene-specific DNA hypomethylation in the colorectum on a consistent and predict- able manner [49, 50]. This may be related to the fact that modulation of SAM and SAH in the colorectum is particularly resistant to folate depletion [49, 50]. Collectively, currently available evidence indicates that genomic DNA hypomethylation in the colorectum is not a probable mechanism by which folate deficiency enhances colorectal carcinogenesis [49, 50]. In contrast, folate supplementation appears to significantly increase the extent of genomic and site-specific DNA methylation in animal and human studies [49, 50]. Dietary folic acid supplementation up to 20 6 BDR significantly reversed DMH-induced DNA hypomethylation within a coding region of the p53 tumor suppressor gene in rat colon in a dose-dependent and site-specific manner [92] in the absence of change in genomic DNA methylation [93]. Folic acid supplementation (286 – 516 l g/day 6 3 wk) [142] or 5-methylTHF (15 mg/day 6 8 wk) [145] was able to normalize pre-existing genomic DNA hypomethylation in peripheral leukocytes in humans. Folic acid supplementation at 12.5 – 25 6 BDR for 3 – 12 months significantly increased the extent of colonic mucosal genomic DNA methylation in subjects with resected colorectal adenomas or cancer [146 – 148]. Even a physiological dose of folic acid (400 l g/day) for 10 wk increased genomic DNA methylation in lymphocytes (by 31%; p = 0.05) and in colonic mucosa (by 25%; p = 0.09) in patients with colorectal adenomas [152]. Collectively, currently available evidence indicates that folate supplementation appears to be able to reverse pre-existing genomic DNA hypomethylation and to increase the extent of genomic DNA methylation above the pre-existing level [49, 50]. Therefore, prevention or rever- sal of genomic DNA hypomethylation may be a mechanism by which folate supplementation suppresses neoplastic transformation in the colorectum (Fig. 3). In preneoplastic and neoplastic cells where DNA replication and cell division are occurring at an accelerated rate, folate depletion causes ineffective DNA synthesis, resulting in inhibition of tumor growth and progression (Fig. 3), which is the basis for cancer chemotherapy using antifolate agents ( e. g. , methotrexate) and 5-fluorouracil [58, 59, 123]. Thus, this is the most likely mechanism by which folate deficiency inhibits the progression of the established preneoplastic neoplastic foci in the colorectum. Another possible mechanism is that folate deficiency may reverse CpG promoter methylation of tumor suppressor and other anticancer genes involved in colorectal carcinogenesis, thereby reactivating these genes. However, there is currently no experimental evidence to support this theo- retical possibility. As discussed earlier, folate deficiency appears to be unable to induce genomic and gene-specific DNA hypomethylation in the colorectum [49, 50]. Furthermore, in a recent in vitro study, folate deficiency induced a significant reduction in genomic and site-specific DNA methylation in untransformed NIH/3T3 and CHO-K1 mammalian cells but not in HCT116 and Caco2 human colon adenocarcinoma cells [196]. In this study, folate deficiency did not produce significant changes in the promoter CpG island methylation of the p16 tumor suppressor gene and the MLH1 mismatch repair gene in HCT116 and Caco- 2 cells [196]. However, certain sites in the promoter CpG island of the ER gene were associated with modest, albeit statistically significant, changes in CpG methylation in response to folate deficiency, which were not associated with significant functional consequences [196]. In line with this observation, another study showed that HCT116 cells lacking DNMT1 exhibited only a modest 20% decrease in the overall genomic DNA methylation despite the markedly decreased cellular DNMT activity [197]. In this model, although juxtacentromeric satellites became significantly demethylated, centromeric satellite loci, and the promoter CpG island of the p16 gene remained fully methylated [197]. Only when both the DNMT1 and DNMT3b genes were disrupted, genomic DNA methylation was reduced by A 95% and significant hypomethylation of satellite sequences and several promoter CpG islands, including that of the p16 gene, was observed [198]. These observations suggest that it may be extremely difficult to reverse DNA methylation in cancer cell lines such as HCT116. The fact that an almost complete abolishment of DNMT activity by disruption of both the DNMT1 and DNMT3b genes is required to produce significant DNA hypomethylation in HCT116 cells [198] suggest that folate deficiency alone is unlikely to be a sufficient predisposing condition to produce significant DNA hypomethylation in colon cancer cells. Mechanistically, the most likely mechanism by which folic acid supplementation may promote the progression of established preneoplastic and neoplastic lesions in the col- orecutm is provision of nucleotide precursors to rapidly replicating neoplastic cells for accelerated proliferation and growth [58, 59, 123]. Another possible mechanism by which folic acid supplementation may promote the progression of preneoplastic or neoplastic foci in the colorectum may be de novo methylation of promoter CpG islands of tumor suppressor genes and other critical genes involved in colorectal carcinogenesis with consequent gene inactiva- tion leading to tumor progression (Fig. 3). This potential epigenetic mechanism of tumor progression is supported by recent animal studies using viable yellow agouti mice that unequivocally have demonstrated that maternal dietary methyl group supplementation containing folic acid permanently alters phenotypic coat color of the offspring via increased methylation at the promoter CpG site of the agouti gene [199 – 201]. However, it is unknown at present whether this de novo methylation of promoter CpG islands can happen with folic acid supplementation alone, whether it is operative in normal or neoplastic tissues or both, whether this effect is associated with folic acid supplementation provided in utero only, in postpartum period, or in adulthood, and whether it is tissue and gene-specific. Another possible means by which folic acid supplementation may promote colorectal carcinogenesis may be through hypermutability of methylated cytosines in CpG dinucleotides (Fig. 3). Methylated CpG sites are mutational hot spots for human cancer as described earlier [194]. The majority of mutations observed in CpG sites are cytosine- to-thymine transitions mediated by the spontaneous deamination of 5-methylcytosine to thymine and by the enzymatic deamination of 5-methylcytosine to thymine by DNMT [194]. Therefore, there is a possibility that de novo methylation of cytosines in CpG sites in critical genes involved in colorectal carcinogenesis may create mutational hot spots, leading to inactivating mutations of these genes. A cause-and-effect relationship between folate and CRC is difficult to establish. Because of inherent limitations associated with study design, the results from epidemiologic, animal, and interventional studies examining this ...

Context 25

... the progression of microscopic neoplastic foci to CRC [93]. In support of this latter finding, animal studies using a metabolite of DMH, azoxy- methane (AOM), showed that folic acid supplementation exceeding the BDR by 1000 – 10 000 6 enhanced colorectal carcinogenesis in rats [170 – 174]. Therefore, it appears that folate modulates colorectal carcinogenesis in chemical carcinogen rodent models over a wide range of dietary intakes. Folate deficiency of a moderate degree enhances colorectal carcinogenesis whereas modest levels of folate supplementation above the BDR suppress colorectal tumorigenesis. Supraphysiologic levels of folate supplementation do not appear to confer additional protection and, in some cases, may enhance colorectal carcinogenesis. The implication of this issue is important because the optimal dose of folate supplementation must be determined for folate chemoprevention to be effective and safe in humans. Although some similarities do exist, tumor development in chemical rodent models of CRC differs in several important histological, clinical, and molecular genetic aspects from that observed in humans [175, 176]. Therefore, any extrapolation of the observations from these models to human situations should be made very cautiously. Whether or not the supplemental doses of folic acid used in these animal studies can be directly extrapolated to intake levels in humans is a highly contentious and controversial issue at present because of inherent differences in folate absorption and metabolism between rodents and humans [177, 178]. Recent evidence suggests that animals, unlike humans, have a comparatively high DHFR activity [177, 178]. Consequently, if assessing the impact of systemic exposure of unmetabolized folic acid, animals would have to be orally dosed with a much greater than prorata amount of folic acid in order to elicit the same circulating serum/plasma concentrations of unmetabolized folic acid [177, 178]. Therefore, it may be a gross mistake to dismiss the effects of folic acid exposure in animal studies on the grounds that the experimental folic acid intake (multiple of animal BDR) would translate to an unlikely intake in humans [177, 178]. Therefore, arguably, a 10 – 20 6 exposure in a small animal model may turn out to hardly equate to an extra 1 – 2 6 RDA in humans [177, 178]. In a recent study using the Apc Min mouse model of CRC, both semisynthetic diets with low and high vitamin contents (1/3 of the BDR and 5 6 BDR, respectively), which also contained 1/3 of the BDR and 2 6 BDR folic acid, respectively, significantly increased the number of small intestinal polyps [179]. Furthermore, in two genetic models of CRC ( Apc Min and Apc +/ – 6 Msh2 – / – mice), moderate dietary folate deficiency enhanced, whereas modest levels of folic acid supplementation (four to ten times the BDR) suppressed, the development and progression of CRC, if folate intervention was started before the establishment of neoplastic foci in the intestine [180, 181]. If, however, folate intervention was started after the establishment of neoplastic foci, the same degree of folate deficiency inhibited the progression and induced regression of the established neoplastic foci [180, 181]. A potential tumor promoting effect of folic acid supplementation on the established neoplastic foci could not be clearly determined in these studies because of the inherent limitations associated with these genetic models [180, 181]. Therefore, these observations suggest that the timing of folate intervention is critical in providing an effective and safe chemopreventive effect on colorectal carcinogenesis. Folate deficiency has an inhibitory effect whereas folate supplementation has a promoting effect on the progression of established colorectal neoplasms. In contrast, folate deficiency in the normal colorectum appears predispose it to neoplastic transformation, and modest supplemental levels of folate suppress the development of neoplasms in the normal colorectum. Some animal studies have also shown that dietary folate deficiency inhibits, and not suppresses, the development of breast cancer in rats [182 – 184] in contrast to the inverse association between folate status and breast cancer risk observed in epidemiologic studies [185]. Data from animal studies and clinical observations suggest that folate possesses dual modulatory effects on CRC devel- opment and progression depending on the timing and dose of folate intervention (Fig. 3) [5, 6, 58, 84, 169]. Folate deficiency has an inhibitory effect whereas folate supplementation has a promoting effect on the progression of established colorectal neoplasms (Fig. 3). In contrast, folate deficiency in normal colorectal mucosa appears to predispose it to neoplastic transformation, and modest levels of folic acid supplementation (4 – 10 times above the BDR) suppress, whereas supraphysiological supplemental doses enhance the development of CRC in normal colorectal mucosa (Fig. 3). Are there biologically plausible explanations for these seemingly paradoxical and contradictory epidemiologic, animal, and clinical observations concerning the dual role of folate in CRC development and progression? There exist several biologically plausible mechanisms by which folate deficiency increases, whereas folate supplementation reduces, the risk of CRC in normal colorectal epithelial cells [50, 58, 59, 123]. As an essential cofactor for the de novo biosynthesis of purines and thymidylate (Fig. 2), folate plays an important role in DNA synthesis, stability and integrity, and repair, aberrations of which have been impli- cated in colorectal carcinogenesis [58, 59, 123]. Indeed, a large body of evidence from in vitro , animal and human studies indicates that folate deficiency is associated with DNA strand breaks, chromosomal and genomic instability, uracil misincorporation, impaired DNA repair, and increased mutations [58, 59, 121 – 123, 186]. Furthermore, this body of evidence indicates that folate supplementation can correct some of these defects induced by folate deficiency, and ensures DNA fidelity, maintains DNA integrity and stability, and optimizes DNA repair by providing nucleotide precursors for DNA synthesis and replication [58, 59, 121 – 123, 186]. Therefore, the effect of folate deficiency and supplementation on the DNA synthesis pathway in the normal colorectum have been generally considered to be the primary mechanism by which folate deficiency predisposes it to neoplastic transformation and folate supplementation prevents or suppresses neoplastic transformation, respectively (Fig. 3) [58, 59, 121 – 123, 186]. Another proposed mechanism by which folate deficiency enhances the development of cancer in the colorectum is through an induction of genomic DNA hypomethylation [90]. It has been proposed that a mechanism by which folate supplementation may protect against the development of cancer in the colorectum is through a protection against genomic DNA hypomethylation [90]. This mechanism is based on the biochemical function of folate in mediating one-carbon transfer for the provision of SAM, the primary methyl group donor for most biological methylation reactions, including that of DNA (Fig. 2) and on evidence from animal experiments that demonstrated that diets deficient in different combinations of methyl group donors (choline, folate, methionine, and vitamin B 12 ) consistently induce genomic and site and gene-specific DNA hypomethylation [49, 50]. DNA methylation of cytosine located within the cyto- sine-guanine (CpG) dinucleotide sequences is an important epigenetic determinant in gene expression (an inverse relation), in the maintenance of DNA integrity and stability, in chromosomal modifications, and in the development of mutations [124, 187]. In contrast to methylated CpG sites in the CpG-poor bulk of the genome and unmethylated CpG islands in normal cells, cancer cells simultaneously harbor widespread loss of methylation in the CpG-depleted regions where most CpG dinucleotides should be methylated and gains in methylation of CpG islands in gene promoter regions (Fig. 4) [124, 187]. Global hypomethylation is an early, and consistent, event in colorectal carcinogenesis [124, 187]. Global hypomethylation of the coding and noncoding regions and demethylation of repetitive DNA sequences contribute to the development of cancer through the following mechanisms: chromosomal instability; increased mutations; reactivation of intra- genomic parasitic sequences that could be transcribed and moved to other sites where they could disrupt normal cellular genes; mitotic recombination leading to loss of heterozygosity and promotion of rearrangements; aneuploidy; loss of imprinting; and up-regulation of protooncogenes (Fig. 4) [188]. However, animal studies have shown that genomic demethylation may protect against some cancers ( e. g. , intestinal tumors) [189, 190] but may promote chromosomal instability and increase the risk of cancer in other tissues ( e. g. , lymphoma, sarcoma) [191, 192]. Methylation at promoter CpG islands is an important mechanism of silencing transcription in carcinogenesis; the affected genes are silenced and their function is stably lost in a clonally propagated fashion (Fig. 4) [187, 188, 193]. Many genes inactivated by promoter CpG methylation in carcinogenesis have classic tumor-suppressor function or play critical roles in cell cycle control, repair of DNA damage, apoptosis, differentiation, angiogenesis, metastasis, growth – factor response, drug resistance, and detoxifi- cation [187]. Promoter CpG islands of over 60% of tumor suppressor and mismatch repair genes have been observed to be methylated in cancer [187]. Another means by which CpG methylation may contribute to carcinogenesis is the hypermutability of methylated cytosine. CpG dinucleotides within certain genes are not only the sites of DNA methylation but also mutational hot spots for human ...

Context 26

... and intervention studies as well as animal and in vitro experiments must be analyzed for this purpose [203, 204]. The portfolio approach does not set out a hierarchy of evidence. Instead, it recognizes that all types of evidence have advantages and disadvantages. This means that no single kind of study is considered to be definitive. Instead, all of the different types of studies that are used to investigate the link between nutrition and cancer are considered along- side each other, without favoring evidence from one type over another. In support of the portfolio approach, systema- tic comparisons of the results of randomized intervention studies with observational evidence in several clinical situations have shown that observational data from well-con- ducted studies do not appear to produce biased results compared to randomized interventions [205, 206]. Furthermore, the importance of experimental studies that contribute to understanding mechanisms that might under- lie any observed association between a dietary factor and cancer and might bear on the inference of causation has been increasingly recognized and appreciated in the field of nutrition and cancer. Epidemiologic and experimental evidence indicating a causal association between a dietary factor and cancer is strengthened when a biologic pathway or mechanism by which colorectal carcinogenesis may be modified is identified and when this mechanism is biologically plausible [204]. It can be argued that epidemiologic data, however strong and consistent, are an inadequate basis for any definite judgment of causality unless supported by mechanistic evidence [204]. Recent advances in molecular epidemiology have added another dimension to the already complex field of nutrition and cancer. Recently identified and characterized single nucleotide polymorphisms and other genetic and epigenetic variants of genes that are involved in absorption, transport, metabolism, and excretion of nutrients have been shown not only to modify cancer risk but also to significantly modulate the effect of nutrients and related compounds on cancer risk [207]. This emerging important topic in the field of nutrition and cancer, termed “gene – nutrient interactions” in carcinogenesis, has a very significant implication in designing and interpreting data from observational epidemiologic and intervention studies. Although individuals are subjected to the same level of nutritional exposure, systemic, and target tissue bioavailability of nutrients and their metabolites, as well as their functional effects in the target tissue, might be vastly different because of genetic and epigenetic variations. Genetic and epigenetic susceptibility to cancer and their interaction with diets and other environmental exposures have not been incorporated into the study design of and interpretation of data from previously published epidemiologic and intervention studies. The precise nature and magnitude of gene – nutrient interactions in carcinogenesis are yet to be clearly defined. It appears that overall diet, rather than individual factors, plays the more important role in the development of CRC, thus underscoring the importance of as yet undetermined interactions among dietary components in the development of cancer. It is likely that dietary factors or components do not act in isolation but as part of a biological action package [208]. The major difficulty in establishing a relationship between diet and cancer and in translating observations from nutritional epidemiology into progress in cancer prevention has been due to inability to identify all relevant dietary components that act coordinately to modulate cancer risk and due to inability to identify the other relevant non-nutritional factors that interact with dietary components to modify cancer risk [208]. What can we conclude about the role of folate in CRC development and progression from the seemingly paradoxical and contradictory epidemiologic, animal, and clinical studies? Currently available evidence from epidemiologic, laboratory animal, and intervention studies does not unequivocally support the role of folate in the development and progression of CRC. Furthermore, the precise nature and magnitude of the relationship of CRC with folate have not been clearly defined. However, when the whole body or portfolio of evidence from these studies is analyzed critically, the overall conclusion supports the inverse association between folate status and CRC risk. Definitive answers to questions about folate and CRC are probably beyond the reach of both observational epidemiologic studies and randomized controlled trials [204]. It is clear that folate appears to possess dual modulatory effects on colorectal carcinogenesis depending on the timing and dose of folate intervention (Fig. 3). Folate deficiency has an inhibitory effect whereas folate supplementation has a promoting effect on the progression of established colorectal neoplasms (Fig. 3). In contrast, folate deficiency in normal col- orectal mucosa appears to predispose it to neoplastic transformation, and modest levels of folic acid supplementation suppress, whereas supraphysiologic supplemental doses enhance the development of CRC in normal colorectal mucosa (Fig. 3). Several potential mechanisms relating to the disruption of the known biochemical function of folate (mediating the transfer of one-carbon moieties and consequent DNA synthesis and methylation) exist to support the dual modulatory role of folate in colorectal carcinogenesis (Fig. 3). As discussed briefly above, evidence for a protective effect of folate supplementation on NTD [7, 8] was considered to be sufficiently conclusive and led to mandatory folic acid fortification in the US [11] and Canada [12] in 1998. Folic acid fortification has already significantly improved folate status and has had a substantial beneficial effect on the original target, NTD, in the US and Canada [21 – 26]. Mandatory folic acid fortification is probably the most important science-drive intervention in nutrition and public health in decades [209]. However, the possibility remains that certain segments of the exposed population may benefit less and may even experience some adverse effects from an increased folic acid intake. Over the past few years, the US and Canadian populations have been exposed to a significant increase in folate intake, for which essentially no data on safety exist [13]. No studies have been done to look directly or even indirectly for the adverse effects of greatly increased folate intakes [13]. In addition to the drastic increase in dietary folate intake from mandatory folic acid fortification, 30 – 40% of the North American population consume supplemental folic acid for several possible but as yet unproven health benefits [35]. Whether or not possible deleterious effects of folic acid supplementation ( e. g. , cancer-promoting effect on established preneoplastic and neoplastic lesions) outweigh the known and potential health benefits ( e. g. , prevention of atherosclerosis and NTD; improvement of cognitive function; cancer prevention in normal tissues free of preneoplastic and neoplastic foci) is largely unknown at present. Folate is generally regarded as safe [210] and may become the ultimate functional food component for disease prevention [211]. The potential masking effect of folic acid on vitamin B 12 deficiency, especially in the elderly, has been the only major concern of folic acid fortification and supplementation [13]. However, an emerging body of evidence suggests that folate supplementation may be associated with other potentially serious adverse effects [6]. These include: the occurrence of resistance or tolerance to antifolate-based chemotherapy and anti-inflammatory and antiseizure drugs; decreased natural killer cell cytotoxicty; accelerated cognitive decline in older subjects; increased twin pregnancies; and genetic selections of disease alleles ( e. g. , MTHFR C677T) that predispose individuals to chronic diseases if exposed to low folate status [6, 211 – 220]. Folic acid, the synthetic, fully oxidized form of folate used in fortification and supplementation, is normally reduced and methylated by the intestine before it is released into the circulation as 5-methylTHF; consequently, the latter form is the sole circulating form of folate under normal conditions [2]. However, studies show that this absorption and biotransformation process is saturated at doses in the region of 400 l g folic acid or less [221]. At higher doses, synthetic folic acid is also transported into the blood and may enter in large quantities. Consumption of folic acid A 200 l g have shown to lead to the appearance of unmetabolized folic acid in the serum. Although compelling data about possible antagonistic activities of this fully oxidized form of folate in tissues is lacking, there nevertheless exist concerns about the effect of long-term exposure of cells to unmetabolized folic acid [222]. In this regard, Troen et al. [212] have recently reported that 78% of 104 postmenopausal women 60 – 75 years of age had detectable levels of folic acid in plasma, which was associated with an approxi- mately 23% decrease in natural killer cell cytotoxicity inde- pendent of plasma 5-methylTHF and total folate concentrations. Among participants in a large ( n = 2928) trial of folic acid supplementation during pregnancy, women who received 5 mg folic acid/day had a 70% increased risk of total cancer mortality compared with those not on supplementation (HR = 1.70; 95% CI = 1.06 – 2.72) [168]. In this study, the risk of death from breast cancer in women taking 5 mg folic acid/day was twice that of women taking no supplementation, albeit nonsignificant (HR = 2.02; 95% CI = 0.88 – 4.72) [168]. Furthermore, in line with this observation, a recent the Prostate, Lung, Colorectal, and Ovarian Cancer Screening Trial ( n = 25, 400 postmenopausal ...