The 5,10-methylenetetrahydrofolate reductase (MTHFR) gene plays a critical role in folate metabolism. Studies on the association between MTHFR polymorphisms and length changes in short tandem repeat DNA sequences [microsatellite instability (MSI)] are inconsistent. Using data from colon cancer cases (n = 503) enrolled as part of an existing population-based case-control study, we investigated the association between MTHFR 677 and MTHFR 1298 polymorphisms and MSI. We also examined whether the association was modified by folate intake. Participants were case subjects enrolled as part of the North Carolina Colon Cancer Study. Consenting cases provided information about lifestyle and diet during in-home interviews as well as blood specimens and permission to obtain tumor blocks. DNA from normal and tumor tissue sections was used to determine microsatellite status (MSI). Tumors were classified as MSI if two or more microsatellite markers (BAT25, BAT26, D5S346, D2S123, and D17S250) had changes in the number of DNA sequence repeats compared with matched nontumor tissue. Tumors with one positive marker (MSI-low) or no positive markers (microsatellite stable) were grouped together as non-MSI tumors (microsatellite stable). MTHFR 677 and MTHFR 1298 genotypes were determined by real-time PCR using the 5′ exonuclease (Taqman) assay. Logistic regression was used to calculate odds ratio (OR) and 95% confidence intervals (95% CI). MSI was more common in proximal tumors (OR, 3.8; 95% CI, 1.7-8.4) and in current smokers (OR, 4.0; 95% CI, 1.6-9.7). Compared with MTHFR 677 CC referent, MTHFR 677 CT/TT genotype was inversely associated with MSI among White cases (OR, 0.36; 95% CI, 0.16-0.81) but not significant among African Americans. Although not statistically significant, a similar inverse association was observed between MTHFR 677 CT/TT genotype and MSI among the entire case subjects (OR, 0.54; 95% CI, 0.26-1.10). Among those with adequate folate intake (>400 μg total folate), we found strong inverse associations between combined MTHFR genotypes and MSI (677 CC + 1298 AC/CC, OR, 0.09; 95% CI, 0.01-0.59; 677 CT/TT + 1298 AA, OR, 0.13; 95% CI, 0.02-0.85) compared with the combined wild-type genotypes (677 CC + 1298 AA). This protective effect was not evident among those with low folate (<400 μg total folate) intake. Our results suggest that MTHFR variant genotypes are associated with reduced risk of MSI tumors under conditions of adequate folate intake, although the data are imprecise due to small numbers. These results indicate that the relationship between MTHFR genotypes and MSI is influenced by folate status.

Colon cancer, a common cancer in the United States and other developed countries, is thought to have a strong environmental component. Among environmental exposures, diet has received a great deal of attention. There are at least three pathways to colon cancer: chromosomal instability, microsatellite instability (MSI), and CpG island methylator phenotype. Some sporadic colon cancers are associated with chromosomal instability, which involves allelic loss of tumor suppressor genes, such as p53 or APC. Microsatellites are short tandem repeat DNA sequences. The change in length of these tandem repeats is termed MSI and is often associated with defects in DNA repair genes (1, 2). MSI is associated with 10% to 15% of sporadic colorectal carcinomas (3). Colon tumors harboring MSI more frequently present with proximal location, large tumor size, decreased likelihood of metastasis, less advanced stage at diagnosis, and better prognosis compared with microsatellite-stable (MSS) colon cancers (4, 5). Some colon cancers exhibit aberrant DNA methylation or CpG island methylator phenotype, which involves inactivation or silencing of genes by hypermethylation of promoter cytosine-guanosine (CpG) residues. Interestingly, MSI arises through aberrant methylation of the hMLH1 promoter in sporadic colon cancers (6, 7).

Folate is ingested through diet (e.g., dark green vegetables, eggs, fish, and wheat) or supplements. Evidence from epidemiologic studies suggest that adequate folate intake is associated with decreased risk of colorectal cancer (8-11), although not all studies support this association (12, 13). Folate is integral in the processes of both DNA synthesis and DNA methylation. The 5,10-methylenetetrahydrofolate reductase (MTHFR) enzyme plays an important role in folate metabolism and determines the balance between the different forms of folate for DNA synthesis and DNA methylation (14). MTHFR catalyzes the reduction of 5,10-methylenetetrahydrofolate to 5-methyltetrahydrofolate, a substrate for the conversion of homocysteine to methionine. The latter is a precursor to S-adenosylmethionine, which is an important methyl donor for DNA methylation. In addition, 5,10-methylenetetrahydrofolate, a substrate of MTHFR, is also required for thymidylate and purine synthesis. However, low folate may lead to uracil misincorporation into DNA chromosomal breaks and increase the potential for premutagenic lesions (15, 16). Folate deficiency has also been linked to increased chromosomal aberrations in mammalian cells during mitosis (17). Altered folate metabolism lead to shifts in the balance between availability of 5-methyltetrahydrofolate for methylation reactions and 5,10-methylenetetrahydrofolate for DNA synthesis and repair (18, 19) may influence colorectal cancer risk.

Not surprisingly, polymorphisms in genes related to folate metabolism, specifically MTHFR, are thought to play a role in carcinogenesis of the large bowel. Two polymorphisms in the MTHFR gene that affect the efficiency of folate metabolism have been described (20-22). The MTHFR 677 C>T transition in exon 4 and MTHFR 1298 A>C transversion in exon 7 are associated with reduced enzyme activity resulting in slower folate metabolism. The MTHFR 677 TT genotype results in 30% enzyme activity in vitro compared with the CC wild-type (23), whereas the MTHFR 1298 CC genotype has been found to have 60% of the AA wild-type enzyme activity in vitro (22, 24). We have shown previously a significantly reduced risk of colon cancer among Whites with the MTHFR 1298 CC variant genotype (20). Other studies also reported reduced risk of colorectal cancer among individuals homozygous for MTHFR 677 TT genotype with high folate intake (10, 25, 26). Studies on colorectal adenomas, colorectal cancer precursors, suggest positive association between MTHFR 677 TT genotype and adenoma risk, under inadequate folate status (27, 28). These observations indicate that the effect of MTHFR polymorphisms on colon adenoma and cancer risk is likely modifiable by folate status. MTHFR 677 T or MTHFR 1298 C polymorphisms combined with inadequate folate intake could lead to aberrant DNA methylation. MTHFR polymorphisms associated with reduced enzyme activity limit the conversion of 5,10-methylenetetrahydrofolate to 5-methyltetrahydrofolate, the form of folate required for DNA methylation. Interestingly, MTHFR 677 TT genotype has been associated with reduced DNA methylation under low folate conditions (29, 30). On the other hand, MTHFR polymorphisms that are associated with reduced enzyme activity enhance the accumulation of 5,10-methylenetetrahydrofolate (31) for DNA synthesis by providing more nonmethylated folate for nucleotide synthesis and DNA repair (32). Low folate is associated with increased uracil misincorporation during DNA synthesis (33) and increased frequency of DNA strand breaks (15).

No studies have examined the relationship among MTHFR polymorphisms, MSI, and folate status in relation to colon cancer. The findings from the few studies that examined MSI and MTHFR polymorphisms are not consistent. A recent study reported a modest but not significant inverse association between MTHFR TT genotype and MSI (34), whereas another study observed a higher frequency of MTHFR TT genotype among colorectal cancer cases with MSI tumors (35). A third study found no association between MTHFR TT genotypes and MSI (36). In this study, we hypothesized that in colon cancer subjects MTHFR 677 T and MTHFR 1298 C polymorphisms would be associated with MSI and that the association between MTHFR and MSI would be modified by folate intake. We also hypothesized that there would be differences by race given the disparities in incidence and mortality and the distribution of MTHFR gene polymorphisms by race.

Study Design

Participants in this study were colon cancer cases enrolled as part of a separate study, the North Carolina Colon Cancer Study, a population-based case-control study of colon cancer in 33 counties of central and eastern North Carolina. Details about the North Carolina Colon Cancer Study have been described previously (20). Colon cancer cases were identified through the rapid ascertainment system of the North Carolina Central Cancer Registry. All African Americans and a sample of Whites were recruited between October 1, 1996 and September 1, 2000. Colon cancer patients ages between 40 and 80 years residing in the 33-county area with confirmed invasive adenocarcinoma of the colon were considered eligible. A total of 676 interviews were completed among colon cancer cases in North Carolina Colon Cancer Study. The cooperation rate among cases [interviewed / (interviewed + refused)] was 84%. MSI data were collected on 503 cases. Of these, 486 had data on MTHFR genotype or folate intake and these were the people included in the analyses. Tissue specimens were prospectively collected under institutional review board approval as part of the North Carolina Colon Cancer Study. Formalin-fixed, paraffin-embedded colon blocks containing tumor or normal (tumor-free) tissue were identified from pathology reports and blocks were requested from participating hospitals. The study pathologist evaluated a reference slide from each block that was stained with H&E to confirm the initial diagnosis and also to identify areas of normal and tumor tissues for microdissection on additional serial tissue sections.

Trained nurse interviewers conducted interviews in the participants' homes. The interviews covered lifestyle and dietary habits, physical activity, smoking and alcohol use, family history of colon cancer, personal attributes, and nonsteroidal anti-inflammatory drug (NSAID) usage. Dietary information (frequency and serving size) was collected using a modified version of the semiquantitative Block food frequency questionnaire developed at the National Cancer Institute (NCI; ref. 37). The food questionnaire was modified by adding 29 food items consumed in North Carolina (38). Participants were asked to estimate their usual frequency of intake of various foods and portion sizes for the year before diagnosis. The 1-year period was selected to account for seasonal variations in dietary intake. We used the nutrients database program provided by NCI to compute estimates of usual intake for a variety of nutrients, including calcium and folate intake, for the year before diagnosis. In addition, use of NSAIDs during the previous 5 years was assessed. The University of North Carolina Medical School Institutional Review Board approved the study protocol.

Blood Collection

Blood was collected following the interview after receiving written consent from the participant. The blood specimens were kept at 4°C to 21°C for up to 48 hours during transit to the laboratory. DNA samples were extracted from the buffy coats using the Puregene DNA isolation kit following the manufacturer's recommendations (Gentra Systems, Inc., Minneapolis, MN).

MSI Status

DNA was extracted from normal and tumor tissue sections from formalin-fixed, paraffin-embedded tissue sections. A H&E-stained section was used as a guide for microdissection. Genomic DNA was extracted from the tissue by deparaffinization in xylene, purified with absolute alcohol, and pelleted at maximum speed in a microcentrifuge. The pellets were dried in a DNA SpeedVac (Savant, Inc., Farmingdale, NY) and resuspended in 15 μL Genereleaser (BioVentures, Inc., Murfeesboro, TN) according to the manufacturer's protocol. The resuspended DNA was then incubated overnight at 55°C, with 200 mg/mL proteinase K (Sigma, St. Louis, MO). This product was then used directly in the PCR reactions.

To determine MSI in the pathologic specimens, a reference panel of five NCI-recommended microsatellite markers (BAT25, BAT26, D5S346, D2S123, and D17S250) were used (39). The reaction to end-label one primer from each pair contained the primers, kinase buffer, T4 polynucleotide kinase, and 32P/33P. The PCR reaction contained 0.125 pmol of each primer in the pair, 0.25 units Taq DNA polymerase, 40 mmol/L deoxynucleotide triphosphate mix, and 1.5 to 2.0 mmol/L MgCl2. The PCR products were denatured in 95% formamide and used for electrophoresis on a 6% polyacrylamide/7.5 mol/L urea gel. Mutations, identified by changes in the electrophoretic mobility of the PCR products, were analyzed by autoradiography. Alleles in the tumor DNA were compared with nontumor DNA taken from the same specimen. The sample was defined as having high MSI if two or more of the five markers contained novel alleles in the tumor compared with matched nontumor tissue and as MSS if none of the markers had novel alleles. Specimens with one of five positive markers were defined as MSI-low and combined with the MSS group as non-MSI tumors (MSS). Among the 486 subjects with MSI, folate, and MTHFR genotype data, the frequencies of MSI and MSS were 10.1% and 89.9%, respectively.

Genotyping

Genotyping was done as reported previously (20) using the RFLP assay and the 5′ exonuclease (Taqman, Applied Biosystems, Foster City, CA) assays. Primers for the RFLP assay were MTHFR 677 forward 5′-AGGACGGTGCGGTGAGAGTG-3′ and reverse 5′-TGAAGGAGAAGGTGTCTGCGG-3′ and MTHFR 1298 forward 5′-CTTTGGGGAGCTGAAGGACTACTAC-3′ and reverse 5′-CACTTTGTGACCATTCCGGTTTG-3′. The 5′ exonuclease assay primer and probe sequences are as follows: MTHFR 677 forward 5′-AGGCTGACCTGAAGCACTTGAA-3′ and reverse 5′-CTCAAAGAAAGCTGCGTGATGA-3′; probes end-labeled with the quencher and reporter dyes: VIC-TGTCTGCGGG AGCCGATTTCA for the common allele and FAM-AGGTGTCTGCGGGAGTCGATTTCA for the variant allele; MTHFR 1298 forward 5′-AAGGAGGAGCTGCTGAAGATGT-3′ and reverse 5′-TGTGACCATTCCGGTTTGG-3′; probes: VIC-AAGACACTTTCTTCACTG (A allele) and FAM-AGACACTTGCTTCACT (C allele). At least four negative template controls as well as five positive controls for each allele were included with every assay. A random 10% repeat of samples was done using both methods, with 100% agreement.

Statistical Analysis

The response of interest was MSI. MSI was dichotomized into MSI versus non-MSS (MSI-low + MSS). The main exposures of interest in this study were folate intake and MTHFR genotype. Three different measures of folate intake were evaluated: total folate, dietary folate, and supplemental folate. The Block-NCI program was used to determine the amounts of dietary and supplemental folate intake. Total folate was the summation of the reported microgram amounts of supplemental folate and dietary folate intake. Dichotomous variables were created from the continuous data for both total folate and dietary folate. For total folate, 400 μg was chosen as the cut point because it is the current recommended daily allowance and it was used in previous studies (20). For dietary folate, we used the mean intake of 275 μg. The odds ratios (OR) did not differ when we used the median intake of dietary folate (257 μg) as the cut point (data not shown). Supplemental folate was collapsed into either yes, supplement taken, or no, supplement not taken.

There are three genotypes for each MTHFR locus: a homozygous wild-type or common allele, a homozygous variant allele, and a heterozygote, which has one copy of each allele. For both genotypes, heterozygotes and variant alleles were collapsed into one category to increase precision, thereby creating a dichotomous variable for each locus.

Covariates tested were total calcium intake (quartiles), multivitamin use [use regularly (>3 days a week), use occasionally (>1 day a month but <3 days a week), or do not use], family history (at least one first-degree relative with colorectal cancer: yes/no), alcohol use (nonuser, lower half of consumers, or upper half of consumers), smoking (never, former, or current), NSAID use during the past 5 years before diagnosis [never, occasionally (>1 day a month but <3 days a week), or frequently (≥3 days a week)], stage at time of diagnosis (local, regional, or distal), and tumor location (proximal or distal).

Proc Logistic (SAS 8.2, SAS Institute, Cary, NC) was used to calculate ORs and 95% confidence intervals (95% CI). Covariates were included if they resulted in at least a 10% difference between ORs for the main exposure of interest in the crude and adjusted models. Ps were considered statistically significant at the <0.05 level. Interaction terms were created for the combinations for MTHFR and folate.

For all interaction models, race, sex, and age were included in the adjusted model. To test for multiplicative interaction, the interaction term was added to the model and evaluated using the likelihood ratio test. To test for additive interaction, the ORs and their variances from the model with the interaction term were used to find an interaction contrast ratio and its corresponding 95% CI and from that a P was generated test the null hypothesis that the interaction contrast ratio is equal to 0 (no interaction).

Table 1 provides the ORs with 95% CIs for the association between MSI and general characteristics among colon cancer cases. Age, White race, proximal location of tumor, high alcohol intake, and both former and current smoking were significantly associated with the likelihood of having MSI. Frequent use of NSAIDs versus no use over the past 5 years was associated with lower likelihood of MSI. Gender, family history, stage at the time of diagnosis, alcohol use, occasional NSAID use, calcium intake, and vitamin/supplement use were not significantly associated with MSI.

Table 2 provides the ORs for the association between MSI and total, dietary, and supplemental folate intake. There were no associations between MSI and the various forms of folate. We examined the relationship between MTHFR polymorphisms (MTHFR 677 C>T and MTHFR 1298 A>C) and MSI among the entire study subjects and by race. However, we had limited power due to small numbers. To increase precision, we combined the MTHFR heterozygotes and homozygous variant genotypes (677 CT + TT; 1298 AC + CC) for each polymorphism separately. Among the entire case subjects, there was a nonsignificant inverse association between MTHFR CT/TT genotype and MSI (677 CT/TT, OR, 0.54; 95% CI, 0.26-1.10 versus 677 CC wild-type referent). Among White subjects, those with MTHFR CT/TT genotype were significantly less likely to have MSI (OR, 0.36; 95% CI, 0.16-0.81) compared with the MTHFR 677 CC wild-type referent. Among African Americans, there was no significant association between MTHFR CT/TT genotype and MSI. These data should be interpreted with caution because of unstable estimates due to small numbers. There was no association between MTHFR 1298 AC/CC genotype and MSI among entire subjects and in stratified analysis by race. For MTHFR 677 and MTHFR 1298 haplotypes (677 CT/TT + 1298 AC/CC compared with wild-type 677 CC + 1298 AA) and MSI, the results were suggestive of inverse relation to MSI, but the results were not significant (OR, 0.51; 95% CI, 0.13-1.80).

Results for the relationship among MTHFR genotypes, total, dietary, or supplemental folate, and MSI are found in Table 3. Variables that resulted in at least a 10% difference between ORs for the main exposure of interest in the crude and adjusted models (age, race, tumor location, smoking status, total calcium intake, and vitamin/mineral supplement use) were included in the models. In all combinations, the MTHFR common wild-type genotypes 677 CC or 1298 AA and high total, dietary, or supplemental folate were used as the reference. Compared with the reference, those with MTHFR 1298 AC/CC genotype and adequate total folate (>400 μg) were less likely to have MSI (OR, 0.18; 95% CI, 0.04-0.73). Similarly, those with MTHFR 1298 AC/CC genotype and supplemental folate also showed significant inverse association to MSI (OR, 0.10; 95% CI, 0.01-0.81). Compared with the reference (677 CC + >275 μg dietary folate), subjects with MTHFR 677 CT/TT genotype and dietary folate intake >275 μg were less likely to have MSI (OR, 0.11; 95% CI, 0.52), but there was no association between MTHFR 1298 AC/CC genotype/high dietary folate and MSI (Table 3).

ORs for combined MTHFR genotypes and MSI among those with adequate folate are presented in Table 4A. Among those with high total folate (>400 μg), MTHFR variant genotypes were significantly associated with reduced risk of MSI, although some of the cell sizes are small. Among those with low folate intake, the results suggest a slight positive association between combined MTHFR variants and MSI, although not significant (Table 4B).

Interaction terms were created for the combinations for MTHFR and folate. For all interaction models, race, sex, and age were included in the adjusted model. There was evidence for multiplicative (P = 0.01) and additive (P = 0.001) interaction between MTHFR 1298 genotype and total folate intake. The same was true for MTHFR 1298 AC/CC and supplemental folate (additive P = 0.001, multiplicative P = 0.02) and MTHFR 677 CT/TT and dietary folate (additive P = 0.01, multiplicative P = 0.01). None of the remaining interaction models, whether additive or multiplicative, resulted in a statistically significant result. We were unable to stratify the interactions by race due to low numbers in the African American category, which made the models unreliable.

In this study, we investigated the association of two MTHFR polymorphisms, codons 677 C>T and 1298 A>C, folate intake, and MSI status among colon cancer cases. We hypothesized that MTHFR polymorphisms that are linked with reduced MTHFR enzyme activity would be associated with greater risk of colon tumors exhibiting MSI and that the association would be modified by folate status. We found that the relationship between MTHFR polymorphisms and MSI was influenced by folate status. Among subjects with adequate folate (>400 μg), the combined MTHFR 677 and MTHFR 1298 variant genotypes were associated with reduced risk of MSI tumors. This protective effect was not evident under low folate conditions. We examined the relationship between MTHFR genotypes and MSI among all case subjects and in stratified analysis by race. We observed inverse association between MTHFR 677 CT/TT genotype and MSI for all case subjects, but the association was only significant among Whites. The relationship between MTHFR 1298 AC/CC genotype and MSI was close to the null.

Several studies but not all have shown reduction in colorectal cancer risk among individuals homozygous for MTHFR 677 TT under conditions of adequate folate intake (10, 25, 26, 40). These observations suggest that the protective effects of MTHFR variant genotypes on colon cancer risk may be modified by interactions with folate intake. In this study, combined MTHFR variant genotypes (677 CT/TT + 1298 AA; 677 CC + 1298 AC/CC; Table 4) were inversely related to the risk of MSI tumors in the presence of adequate folate intake. A possible explanation for this finding may relate to more efficient metabolism when folate intake is adequate. For instance, it is known that MTHFR variant genotypes with lower enzyme activity favor increased availability of the nonmethylated form of folate 5,10-methylenetetrahydrofolate for DNA synthesis and decreased levels of 5-methyltetrahydrofolate, the form of folate required for DNA methylation. However, when there is adequate supply of folate, although MTHFR activity is low, presumably, enough folate is being converted to 5-methyltetrahydrofolate for DNA methylation while at the same time shunting 5,10-methylenetetrahydrofolate toward DNA synthesis for the conversion of uracil to thymidine. However, MTHFR 677 TT genotype in combination with low folate disrupts DNA methylation and may contribute to carcinogenesis (29). Indeed, emerging evidence implicates aberrant DNA methylation as an important component in the pathways leading to colorectal cancer. Aberrant DNA methylation, particularly in tumor suppressor and DNA mismatch repair genes, presumably acts through global hypomethylation or hypermethylation of specific CpG islands to promote gene inactivation. Hypermethylation of the hMLH1 promoter has been observed in most sporadic colorectal cancer with MSI (41) particularly when there is no evidence of mutations in DNA MMR genes (35, 42). However, we did not evaluate DNA methylation and mutations in MMR genes in this study.

Contrary to our predictions, we found an inverse relationship between MTHFR variant genotypes and MSI. MTHFR 677 and MTHFR 1298 variants are linked with reduced enzyme activity. The MTHFR 677 TT genotype is associated with 30% reduced enzyme activity (21) compared with CC wild-type; therefore, our findings may be related to lower MTHFR enzyme activity. The increased availability of 5,10-methylenetetrahydrofolate associated with reduced MTHFR activity favors DNA synthesis and repair, thereby resulting in less uracil misincorporation and decreased DNA strand breaks (15) and presumably leading to more MSS tumors. The MTHFR 1298 C variant affects enzyme activity to a lesser degree than MTHFR 677 T variant (22). We did not observe a relationship between MTHFR 1298 C variant and MSI. This finding could be related to higher MTHFR 1298 enzyme activity. Although the two MTHFR variants are in linkage disequilibrium (43), we found no evidence for linkage disequilibrium between the two MTHFR polymorphisms in our study population. A recent study reported that the modest reduction in colorectal cancer risk associated with MTHFR 1298 A>C may be independent of the MTHFR 677 C>T (43).

Very few studies have examined the relationship between MTHFR polymorphisms and MSI. Toffoli et al. (34) observed a borderline inverse association between MTHFR 677 TT genotype and MSI but found no significant association between MTHFR 1298 CC genotype and MSI status. Our findings are similar but stronger than the results of the study by Toffoli et al. but conflict with two other studies. Shannon et al. (35), observed that MSI tumors were more common among older colorectal cancer cases with MTHFR TT genotype compared with those with MSS tumors, whereas Plaschke et al. (36) reported no significant associations between MTHFR genotypes and MSI. The reasons for these conflicting observations are not entirely clear. Our results differ from these two studies and may reflect differences in methodologies. For example, Shannon et al. (35) assessed MSI status using only one microsatellite marker (BAT26), whereas we used the panel of five microsatellite markers recommended by NCI. Differences in the classification of MSI resulting from testing a varying number of loci at different sites may contribute to conflicting reports in the literature (44). In our analysis, we used the common MTHFR 677 wild-type genotypes (CC) as referent and the heterozygote and variant genotypes (CT + TT) were combined to increase precision. Shannon et al. (35) and Plaschke et al. (36) combined the common MTHFR 677 wild-type and heterozygote genotypes (CT) as the referent group.

Not many studies have examined the relationship between dietary factors, such as folate, and colorectal tumors exhibiting MSI. Interestingly, we found no associations between MSI and folate intake (total, dietary, and supplemental). Our results agree with the observations of Slattery et al. (45), who evaluated the relationship between dietary intake of various nutrients and foods in relation to MSI and reported no associations between most dietary factors and MSI, except for alcohol consumption. The inverse associations that we observed between MTHFR haplotypes and MSI would suggest that, under conditions of adequate folate, the DNA synthesis and repair process is favored which likely leads to MSS tumors.

The strengths of this study include the effort to thoroughly collect pathologic specimens for MSI analysis, nutritional data, recruitment of cases from a region that contained both rural and urban areas, and the deliberate recruitment of both African Americans and Whites. To our knowledge, stratification by race for MSI studies has not been addressed before. The weaknesses of our study include the following issues. We estimated folate intake from the Block-NCI food frequency questionnaire. Estimation of nutrient intake from self-reported dietary information collected after diagnosis of disease in cases is subject to recall bias. In addition, we assumed in our study that folate supplement users took daily tablet containing 400 μg, which could have overestimated folate intake in some people. However, the results did not differ when we assumed lower values of folate for supplement users (data not shown). Many of the previous studies tested serum folate, which we were not able to do because blood samples were obtained after diagnosis in cases. One of the objectives of this study was to investigate differences among the Whites and African Americans. Unfortunately, the small numbers of African Americans with MSI were small, which led to imprecise estimates as evidenced by the large 95% CIs. These insufficient data made it difficult to formulate any clear-cut conclusions. Additional weaknesses include missing genotype data due to refusals for blood draw and the need to combine several categories of data due to small numbers. We examined only two polymorphic loci, and according to Little et al. (46), >30 different polymorphic loci are related to folate metabolism, and by looking at only a few of these loci, we may be missing information on the role of other important loci.

In conclusion, we found that MTHFR variant genotypes were inversely associated with MSI when folate intake was adequate. This would imply that, under conditions of reduced MTHFR enzyme activity and abundant folate, a balance in folate pools is maintained for DNA methylation and DNA synthesis. In light of a lack of survival advantage for patients with MSI colon tumors undergoing treatment with 5-fluorouracil (47, 48) whose metabolism involves folate intermediates, our findings could have potential implications for colon cancer treatment and prevention. Our observations need to be confirmed in larger studies.

Grant support: NIH grants R01 CA66635 and R01 CA90231, Center for Gastrointestinal Biology and Disease grant P30 DK34987, Lineberger Comprehensive Cancer Center core grant P30-CA16086, Center for Environmental Health and Susceptibility, National Institute of Environmental Health Sciences grant P30-ES10126, and California Department of Health Services Cancer Research Program.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

We thank the High Throughput Genotyping Core, the laboratory of Dr. Keku, and the laboratory of Dr. Carethers for the excellent work; Chris Martin and Kui Huang for invaluable advice; and the North Carolina Colon Cancer Study nurse interviewers and participants.

1
Dietmaier W, Wallinger S, Bocker T, et al. Diagnostic microsatellite instability: definition and correlation with mismatch repair protein expression.
Cancer Res
1997
;
57
:
4749
–56.
2
Thibodeau SN, French AJ, Cunningham JM, et al. Microsatellite instability in colorectal cancer: different mutator phenotypes and the principal involvement of hMLH1.
Cancer Res
1998
;
58
:
1713
–8.
3
Chapusot C, Martin L, Bouvier AM, et al. Microsatellite instability and intratumoural heterogeneity in 100 right-sided sporadic colon carcinomas.
Br J Cancer
2002
;
87
:
400
–4.
4
Suh JH, Lim SD, Kim JC, Hong SH, Kang GH. Comparison of clinicopathologic characteristics and genetic alterations between microsatellite instability-positive and microsatellite instability-negative sporadic colorectal carcinomas in patients younger than 40 years old.
Dis Colon Rectum
2002
;
45
:
219
–28.
5
Gryfe R, Kim H, Hsieh ET, et al. Tumor microsatellite instability and clinical outcome in young patients with colorectal cancer.
N Engl J Med
2000
;
342
:
69
–77.
6
Veigl ML, Kasturi L, Olechnowicz J, et al. Biallelic inactivation of hMLH1 by epigenetic gene silencing, a novel mechanism causing human MSI cancers.
Proc Natl Acad Sci U S A
1998
;
95
:
8698
–702.
7
Menigatti M, Di Gregorio C, Borghi F, et al. Methylation pattern of different regions of the MLH1 promoter and silencing of gene expression in hereditary and sporadic colorectal cancer.
Genes Chromosomes Cancer
2001
;
31
:
357
–61.
8
Baron JA, Sandler RS, Haile RW, et al. Folate intake, alcohol consumption, cigarette smoking, and risk of colorectal adenomas.
J Natl Cancer Inst
1998
;
90
:
57
–62.
9
Konings EJ, Goldbohm RA, Brants HA, Saris WH, van den Brandt PA. Intake of dietary folate vitamers and risk of colorectal carcinoma: results from The Netherlands Cohort Study.
Cancer
2002
;
95
:
1421
–33.
10
Slattery ML, Potter JD, Samowitz W, Schaffer D, Leppert M. Methylenetetrahydrofolate reductase, diet, and risk of colon cancer.
Cancer Epidemiol Biomarkers Prev
1999
;
8
:
513
–8.
11
Su LJ, Arab L. Nutritional status of folate and colon cancer risk: evidence from NHANES I epidemiologic follow-up study.
Ann Epidemiol
2001
;
11
:
65
–72.
12
Levi F, Pasche C, Lucchini F, La Vecchia C. Selected micronutrients and colorectal cancer. a case-control study from the canton of Vaud, Switzerland.
Eur J Cancer
2000
;
36
:
2115
–9.
13
Harnack L, Jacobs DR Jr, Nicodemus K, et al. Relationship of folate, vitamin B-6, vitamin B-12, and methionine intake to incidence of colorectal cancers.
Nutr Cancer
2002
;
43
:
152
–8.
14
Bailey LB, Gregory JF III. Polymorphisms of methylenetetrahydrofolate reductase and other enzymes: metabolic significance, risks and impact on folate requirement.
J Nutr
1999
;
129
:
919
–22.
15
Blount BC, Mack MM, Wehr CM, et al. Folate deficiency causes uracil misincorporation into human DNA and chromosome breakage: implications for cancer and neuronal damage.
Proc Natl Acad Sci U S A
1997
;
94
:
3290
–5.
16
Pogribny IP, Muskhelishvili L, Miller BJ, James SJ. Presence and consequence of uracil in preneoplastic DNA from folate/methyl-deficient rats.
Carcinogenesis
1997
;
18
:
2071
–6.
17
Libbus BL, Borman LS, Ventrone CH, Branda RF. Nutritional folate-deficiency in Chinese hamster ovary cells. Chromosomal abnormalities associated with perturbations in nucleic acid precursors.
Cancer Genet Cytogenet
1990
;
46
:
231
–42.
18
Choi SW, Mason JB. Folate and carcinogenesis: an integrated scheme.
J Nutr
2000
;
130
:
129
–32.
19
Duthie SJ. Folic acid deficiency and cancer: mechanisms of DNA instability.
Br Med Bull
1999
;
55
:
578
–92.
20
Keku T, Millikan R, Worley K, et al. 5,10-Methylenetetrahydrofolate reductase codon 677 and 1298 polymorphisms and colon cancer in African Americans and whites.
Cancer Epidemiol Biomarkers Prev
2002
;
11
:
1611
–21.
21
Lievers KJ, Boers GH, Verhoef P, et al. A second common variant in the methylenetetrahydrofolate reductase (MTHFR) gene and its relationship to MTHFR enzyme activity, homocysteine, and cardiovascular disease risk.
J Mol Med
2001
;
79
:
522
–8.
22
van der Put NM, Gabreels F, Stevens EM, et al. A second common mutation in the methylenetetrahydrofolate reductase gene: an additional risk factor for neural-tube defects?
Am J Hum Genet
1998
;
62
:
1044
–51.
23
Frosst P, Blom HJ, Milos R, et al. A candidate genetic risk factor for vascular disease: a common mutation in methylenetetrahydrofolate reductase.
Nat Genet
1995
;
10
:
111
–3.
24
Weisberg I, Tran P, Christensen B, Sibani S, Rozen R. A second genetic polymorphism in methylenetetrahydrofolate reductase (MTHFR) associated with decreased enzyme activity.
Mol Genet Metab
1998
;
64
:
169
–72.
25
Chen J, Giovannucci E, Kelsey K, et al. A methylenetetrahydrofolate reductase polymorphism and the risk of colorectal cancer.
Cancer Res
1996
;
56
:
4862
–4.
26
Ma J, Stampfer MJ, Giovannucci E, et al. Methylenetetrahydrofolate reductase polymorphism, dietary interactions, and risk of colorectal cancer.
Cancer Res
1997
;
57
:
1098
–102.
27
Ulvik A, Evensen ET, Lien EA, et al. Smoking, folate and methylenetetrahydrofolate reductase status as interactive determinants of adenomatous and hyperplastic polyps of colorectum.
Am J Med Genet
2001
;
101
:
246
–54.
28
Levine AJ, Siegmund KD, Ervin CM, et al. The methylenetetrahydrofolate reductase 677C->T polymorphism and distal colorectal adenoma risk.
Cancer Epidemiol Biomarkers Prev
2000
;
9
:
657
–63.
29
Stern LL, Mason JB, Selhub J, Choi SW. Genomic DNA hypomethylation, a characteristic of most cancers, is present in peripheral leukocytes of individuals who are homozygous for the C677T polymorphism in the methylenetetrahydrofolate reductase gene.
Cancer Epidemiol Biomarkers Prev
2000
;
9
:
849
–53.
30
Friso S, Choi SW, Girelli D, et al. A common mutation in the 5,10-methylenetetrahydrofolate reductase gene affects genomic DNA methylation through an interaction with folate status.
Proc Natl Acad Sci U S A
2002
;
99
:
5606
–11.
31
Bagley PJ, Selhub J. A common mutation in the methylenetetrahydrofolate reductase gene is associated with an accumulation of formylated tetrahydrofolates in red blood cells.
Proc Natl Acad Sci U S A
1998
;
95
:
13217
–20.
32
James SJ, Basnakian AG, Miller BJ. In vitro folate deficiency induces deoxynucleotide pool imbalance, apoptosis, and mutagenesis in Chinese hamster ovary cells.
Cancer Res
1994
;
54
:
5075
–80.
33
Wickramasinghe SN, Fida S. Bone marrow cells from vitamin B12- and folate-deficient patients misincorporate uracil into DNA.
Blood
1994
;
83
:
1656
–61.
34
Toffoli G, Gafa R, Russo A, et al. Methylenetetrahydrofolate reductase 677 C->T polymorphism and risk of proximal colon cancer in north Italy.
Clin Cancer Res
2003
;
9
:
743
–8.
35
Shannon B, Gnanasampanthan S, Beilby J, Iacopetta B. A polymorphism in the methylenetetrahydrofolate reductase gene predisposes to colorectal cancers with microsatellite instability.
Gut
2002
;
50
:
520
–4.
36
Plaschke J, Schwanebeck U, Pistorius S, Saeger HD, Schackert HK. Methylenetetrahydrofolate reductase polymorphisms and risk of sporadic and hereditary colorectal cancer with or without microsatellite instability.
Cancer Lett
2003
;
191
:
179
–85.
37
Block G, Hartman AM, Dresser CM, et al. A data-based approach to diet questionnaire design and testing.
Am J Epidemiol
1986
;
124
:
453
–69.
38
Gerber AM, James SA, Ammerman AS, et al. Socioeconomic status and electrolyte intake in black adults: the Pitt County Study.
Am J Public Health
1991
;
81
:
1608
–12.
39
Boland CR, Thibodeau SN, Hamilton SR, et al. A National Cancer Institute Workshop on Microsatellite Instability for cancer detection and familial predisposition: development of international criteria for the determination of microsatellite instability in colorectal cancer.
Cancer Res
1998
;
58
:
5248
–57.
40
Curtin K, Bigler J, Slattery ML, et al. MTHFR C677T and A1298C polymorphisms: diet, estrogen, and risk of colon cancer.
Cancer Epidemiol Biomarkers Prev
2004
;
13
:
285
–92.
41
Herman JG, Umar A, Polyak K, et al. Incidence and functional consequences of hMLH1 promoter hypermethylation in colorectal carcinoma.
Proc Natl Acad Sci U S A
1998
;
95
:
6870
–5.
42
van Rijnsoever M, Grieu F, Elsaleh H, Joseph D, Iacopetta B. Characterisation of colorectal cancers showing hypermethylation at multiple CpG islands.
Gut
2002
;
51
:
797
–802.
43
Chen J, Ma J, Stampfer MJ, et al. Linkage disequilibrium between the 677C>T and 1298A>C polymorphisms in human methylenetetrahydrofolate reductase gene and their contributions to risk of colorectal cancer.
Pharmacogenetics
2002
;
12
:
339
–42.
44
Anwar S, Frayling IM, Scott NA, Carlson GL. Systematic review of genetic influences on the prognosis of colorectal cancer.
Br J Surg
2004
;
91
:
1275
–91.
45
Slattery ML, Anderson K, Curtin K, et al. Dietary intake and microsatellite instability in colon tumors.
Int J Cancer
2001
;
93
:
601
–7.
46
Little J, Sharp L, Duthie S, Narayanan S. Colon cancer and genetic variation in folate metabolism: the clinical bottom line.
J Nutr
2003
;
133
:
3758
–66S.
47
Carethers JM, Smith EJ, Behling CA, et al. Use of 5-fluorouracil and survival in patients with microsatellite-unstable colorectal cancer.
Gastroenterology
2004
;
126
:
394
–401.
48
Ribic CM, Sargent DJ, Moore MJ, et al. Tumor microsatellite-instability status as a predictor of benefit from fluorouracil-based adjuvant chemotherapy for colon cancer.
N Engl J Med
2003
;
349
:
247
–57.