Search

Effects of dietary folate and alcohol intake on promoter methylation in sporadic colorectal cancer: the Netherlands cohort study on diet and cancer. van Engeland M, Weijenberg MP, Roemen GM, Brink M, de Bruine AP, Goldbohm RA, van den Brandt PA, Baylin SB, de Goeij AF, Herman JG. The Research Institute GROW, Department of Pathology, University Maastricht, 6200 MD Maastricht, The Netherlands. Sporadic colorectal cancer (CRC) is characterized by genetic and epigenetic changes such as regional DNA hypermethylation and global DNA hypomethylation. Epidemiological and animal studies suggest that aberrant DNA methylation is associated with low dietary folate intake, which is aggravated by high alcohol intake. The relationship between promoter methylation of genes involved in CRC carcinogenesis and folate and alcohol intake was investigated. Methylation of the APC-1A, p14(ARF), p16(INK4A), hMLH1, O(6)-MGMT, and RASSF1A promoters was studied using methylation-specific PCR in 122 sporadic CRCs, derived from patients with folate and alcohol intake at either the lower or the higher quintiles of the distribution. Overall, promoter hypermethylation frequencies observed were: 39% for APC; 33% for p14(ARF); 31% for p16(INK4A); 29% for hMLH1; 41% for O(6)-MGMT; and 20% for RASSF1A. For each of the tested genes, the prevalence of promoter hypermethylation was higher in CRCs derived from patients with low folate/high alcohol intake (n = 61) when compared with CRCs from patients with high folate/low alcohol intake (n = 61), but the differences were not statistically significant. The number of CRCs with at least one gene methylated was higher (84%) in the low folate intake/high alcohol intake group when compared with the high folate intake/low alcohol intake group (70%; P = 0.085). Despite the size limitations of this study, these data suggest that folate and alcohol intake may be associated with changes in promoter hypermethylation in CRC

Human cancer is characterized by genetic and epigenetic alterations. In this study we provide evidence for the interruption of Ras signaling in sporadic colorectal cancer (CRC) by either genetic activation of the K-ras oncogene or epigenetic silencing of the putative tumor suppressor gene RASSF1A. Paraffin embedded tumor tissue samples from 222 sporadic CRC patients were analysed for K-ras codon 12 and codon 13 activating mutations and RASSF1A promoter hypermethylation. Overall, K-ras mutations were observed in 87 of 222 (39%) and RASSF1A methylation was observed in 45 of 222 (20%) of CRCs. Mutation of K-ras alone was detected in 76 of 222 (34%) CRCs. RASSF1A promoter methylation with wild-type K-ras was observed in 34 of 222 (15%) CRCs. In 101 of 222 (46%) CRCs neither K-ras mutations nor RASSF1A methylation was observed and 11 of 222 (5%) CRCs showed both K-ras mutations and RASSF1A methylation. These data show that the majority of the studied CRCs with K-ras mutations lack RASSF1A promoter methylation, an event which occurs predominantly in K-ras wild-type CRCs (P=0.023, Chi-square test).

Insulin-like growth factors (IGFs) have been associated with growth, body size, physical activity, and colorectal cancer (CRC). We hypothesized that variants in IGF-related genes increase the CRC susceptibility associated with a larger body size and a lack of physical activity. We assessed this in the Netherlands Cohort Study (NLCS). Participants (n=120 852) completed a baseline questionnaire on diet and cancer. ~75% returned toenail clippings. Using a case-cohort approach and 16.3 years of follow-up, toenail DNA from 3768 subcohort members and 2580 CRC cases was genotyped. We aggregated unfavorable alleles (potentially increasing CRC risk) for 18 single nucleotide polymorphisms in 8 genes into a sum score. The sum score (in tertiles) and an IGF1 19-CA repeat polymorphism (19/19, 19/non-19, and non-19/non-19 repeats) in combination with body size (mostly in tertiles) and (non-)occupational physical activity (>12, 8-12, and <8 kJ/min in the job and >90, >60-90, <30-60, and

Low mitochondrial DNA (mtDNA) copy number in tumors has been associated with worse prognosis in colorectal cancer (CRC). This study further deciphers the role of mtDNA copy number in CRC by comparing mtDNA copy number between healthy, adenoma and carcinoma tissue, by investigating its association according to several clinicopathological characteristics in CRC, and by relating it to CRC-specific survival in CRC patients. A hospital-based series of samples including cancer, adenoma and adjacent histologically normal tissue from primary CRC patients (n = 56) and recurrent CRC (n = 16) was studied as well as colon mucosa samples from healthy subjects (n = 76). Furthermore, mtDNA copy number was assessed in carcinomas of 693 CRC cases identified from the population-based Netherlands Cohort Study (NLCS). MtDNA copy number was significantly lower in carcinoma tissue (P = 0.011) and adjacent tissue (P <0.001) compared to earlier resected adenoma tissue and in primary CRC tissue compared to recurrent CRC tissue (P = 0.011). Within both study populations, mtDNA copy number was significantly lower in mutated BRAF (P = 0.027 and P = 0.006) and in microsatellite unstable (MSI) tumors (P = 0.033 and P <0.001) and higher in KRAS mutated tumors (P = 0.004). Furthermore, the association between mtDNA and survival seemed to follow an inverse U-shape with the highest HR observed in the second quintile of mtDNA copy number (HR = 1.70, 95% CI = 1.18, 2.44) compared to the first quintile. These results might reflect an association of mtDNA copy number with various malignant processes in cancer cells and warrants further research on tumor energy metabolism in CRC prognosis.

Insulin-like growth factors (IGFs) have been associated with growth, body size, physical activity, and colorectal cancer (CRC). We hypothesized that variants in IGF-related genes increase the CRC susceptibility associated with a larger body size and a lack of physical activity. We assessed this in the Netherlands Cohort Study (NLCS). Participants (n=120 852) completed a baseline questionnaire on diet and cancer. ~75% returned toenail clippings. Using a case-cohort approach and 16.3 years of follow-up, toenail DNA from 3768 subcohort members and 2580 CRC cases was genotyped. We aggregated unfavorable alleles (potentially increasing CRC risk) for 18 single nucleotide polymorphisms in 8 genes into a sum score. The sum score (in tertiles) and an IGF1 19-CA repeat polymorphism (19/19, 19/non-19, and non-19/non-19 repeats) in combination with body size (mostly in tertiles) and (non-)occupational physical activity (>12, 8-12, and <8 kJ/min in the job and >90, >60-90, <30-60, and

Low mitochondrial DNA (mtDNA) copy number in tumors has been associated with worse prognosis in colorectal cancer (CRC). This study further deciphers the role of mtDNA copy number in CRC by comparing mtDNA copy number between healthy, adenoma and carcinoma tissue, by investigating its association according to several clinicopathological characteristics in CRC, and by relating it to CRC-specific survival in CRC patients. A hospital-based series of samples including cancer, adenoma and adjacent histologically normal tissue from primary CRC patients (n = 56) and recurrent CRC (n = 16) was studied as well as colon mucosa samples from healthy subjects (n = 76). Furthermore, mtDNA copy number was assessed in carcinomas of 693 CRC cases identified from the population-based Netherlands Cohort Study (NLCS). MtDNA copy number was significantly lower in carcinoma tissue (P = 0.011) and adjacent tissue (P <0.001) compared to earlier resected adenoma tissue and in primary CRC tissue compared to recurrent CRC tissue (P = 0.011). Within both study populations, mtDNA copy number was significantly lower in mutated BRAF (P = 0.027 and P = 0.006) and in microsatellite unstable (MSI) tumors (P = 0.033 and P <0.001) and higher in KRAS mutated tumors (P = 0.004). Furthermore, the association between mtDNA and survival seemed to follow an inverse U-shape with the highest HR observed in the second quintile of mtDNA copy number (HR = 1.70, 95% CI = 1.18, 2.44) compared to the first quintile. These results might reflect an association of mtDNA copy number with various malignant processes in cancer cells and warrants further research on tumor energy metabolism in CRC prognosis.

Background: We investigated body size, physical activity, and early-life energy restriction in relation to colorectal tumors with and without methylated insulin-like growth factor-binding protein (IGFBP) genes, which are putative tumor-suppressor genes. Methods: We determined IGFBP2, IGFBP3, and IGFBP7 promoter CpG island hypermethylation in tumors of 733 colorectal cancer cases from the Netherlands Cohort Study (N = 120,852). Participants self-reported lifestyle and dietary factors at baseline in 1986. Using a case-cohort approach (N subcohort = 5,000), we estimated hazard ratios (HR) for colorectal cancer by extent of IGFBP methylation. Results: Comparison of the highest versus lowest sex-specific tertiles of adult body mass index (BMI) gave multivariable-adjusted HRs [95% confidence intervals (CI)] for colorectal cancers with 0 (18.7%), 1 (29.5%), 2 (32.4%), and 3 (19.5%) methylated genes of 1.39 (0.88-2.19), 1.11 (0.77-1.62), 1.67 (1.17-2.38), and 2.07 (1.29-3.33), respectively. Other anthropometric measures and physical activity were not associated with colorectal cancer risk by extent of IGFBP methylation, except height in sex-specific analyses for women. Exposure to energy restriction during the Dutch Hunger Winter versus nonexposure gave HRs (95% CIs) for colorectal cancers with 0, 1, 2, and 3 methylated genes of 1.01 (0.67-1.53), 1.03 (0.74-1.44), 0.72 (0.52-0.99), and 0.50 (0.32-0.78), respectively. Conclusions: Adult BMI, height (in women only), and early-life energy restriction were associated with the risk of having a colorectal tumor characterized by IGFBP methylation. Impact: Body size may particularly increase the risk of IGFBP gene-methylated colorectal tumors; this finding might facilitate more targeted approaches to prevent obesity-related colorectal cancers.

Acrylamide, a probable human carcinogen, is present in heat-treated carbohydrate-rich foods. Epidemiological studies have not shown a clear association between acrylamide intake and colorectal cancer (CRC) risk. This may be due to the molecular heterogeneity in colorectal tumors, which was not taken into consideration before. Since the acrylamide metabolite glycidamide induces specific DNA mutations in rodents, we investigated whether acrylamide is associated with CRC risk characterized by mutations in Kirsten-ras (KRAS) and adenomatous polyposis coli (APC); key genes in colorectal carcinogenesis. This case-cohort analysis, within the Netherlands Cohort Study on diet and cancer, was based on 7.3 years of follow-up. Acrylamide intake was assessed with a food frequency questionnaire. Mutation analysis of codons 1286-1520 in exon 15 in APC and codons 12 and 13 in exon 1 in KRAS was performed on tumor tissue of 733 cases. Hazard ratios (HR) were calculated using Cox proportional hazards analysis. Among men, acrylamide intake was statistically significantly associated with an increased risk of particularly tumors with an activating KRAS mutation {HR fourth versus first quartile: 2.12 [95% confidence interval (CI): 1.16-3.87], P trend: 0.01}. Among women, acrylamide intake was statistically significantly associated with a decreased risk of particularly tumors with a truncating APC mutation (fourth versus first quartile: 0.47 (95% CI: 0.23-0.94), P trend: 0.02), but only in the highest quartile of intake. This is the first study to show that acrylamide might be associated with CRC with specific somatic mutations, differentially in men and women. More research is needed to corroborate or refute these findings.

Background: We investigated body size, physical activity, and early-life energy restriction in relation to colorectal tumors with and without methylated insulin-like growth factor-binding protein (IGFBP) genes, which are putative tumor-suppressor genes. Methods: We determined IGFBP2, IGFBP3, and IGFBP7 promoter CpG island hypermethylation in tumors of 733 colorectal cancer cases from the Netherlands Cohort Study (N = 120,852). Participants self-reported lifestyle and dietary factors at baseline in 1986. Using a case-cohort approach (N subcohort = 5,000), we estimated hazard ratios (HR) for colorectal cancer by extent of IGFBP methylation. Results: Comparison of the highest versus lowest sex-specific tertiles of adult body mass index (BMI) gave multivariable-adjusted HRs [95% confidence intervals (CI)] for colorectal cancers with 0 (18.7%), 1 (29.5%), 2 (32.4%), and 3 (19.5%) methylated genes of 1.39 (0.88-2.19), 1.11 (0.77-1.62), 1.67 (1.17-2.38), and 2.07 (1.29-3.33), respectively. Other anthropometric measures and physical activity were not associated with colorectal cancer risk by extent of IGFBP methylation, except height in sex-specific analyses for women. Exposure to energy restriction during the Dutch Hunger Winter versus nonexposure gave HRs (95% CIs) for colorectal cancers with 0, 1, 2, and 3 methylated genes of 1.01 (0.67-1.53), 1.03 (0.74-1.44), 0.72 (0.52-0.99), and 0.50 (0.32-0.78), respectively. Conclusions: Adult BMI, height (in women only), and early-life energy restriction were associated with the risk of having a colorectal tumor characterized by IGFBP methylation. Impact: Body size may particularly increase the risk of IGFBP gene-methylated colorectal tumors; this finding might facilitate more targeted approaches to prevent obesity-related colorectal cancers.

Acrylamide, a probable human carcinogen, is present in heat-treated carbohydrate-rich foods. Epidemiological studies have not shown a clear association between acrylamide intake and colorectal cancer (CRC) risk. This may be due to the molecular heterogeneity in colorectal tumors, which was not taken into consideration before. Since the acrylamide metabolite glycidamide induces specific DNA mutations in rodents, we investigated whether acrylamide is associated with CRC risk characterized by mutations in Kirsten-ras (KRAS) and adenomatous polyposis coli (APC); key genes in colorectal carcinogenesis. This case-cohort analysis, within the Netherlands Cohort Study on diet and cancer, was based on 7.3 years of follow-up. Acrylamide intake was assessed with a food frequency questionnaire. Mutation analysis of codons 1286-1520 in exon 15 in APC and codons 12 and 13 in exon 1 in KRAS was performed on tumor tissue of 733 cases. Hazard ratios (HR) were calculated using Cox proportional hazards analysis. Among men, acrylamide intake was statistically significantly associated with an increased risk of particularly tumors with an activating KRAS mutation {HR fourth versus first quartile: 2.12 [95% confidence interval (CI): 1.16-3.87], P trend: 0.01}. Among women, acrylamide intake was statistically significantly associated with a decreased risk of particularly tumors with a truncating APC mutation (fourth versus first quartile: 0.47 (95% CI: 0.23-0.94), P trend: 0.02), but only in the highest quartile of intake. This is the first study to show that acrylamide might be associated with CRC with specific somatic mutations, differentially in men and women. More research is needed to corroborate or refute these findings.

We studied the overlap between the major (epi)genomic events microsatellite instability (MSI), the CpG island methylator phenotype (CIMP) and chromosomal instability (CIN) in colorectal cancer (CRC), and whether specific (epi)genotypes were associated with CRC-related deaths. Molecular analyses using tumor DNA were successful in 509 CRC cases identified within the Netherlands Cohort Study in the period 1989-1993. Follow-up for the vital status until May 2005 was 100%. MSI (12.6%), CIMP-only (5.3%), CIMP + CIN (13.4%), CIN-only (58.2%) and triple-negative tumors (10.6%) differed significantly regarding tumor localization, differentiation grade, initial adjuvant therapy (AT) use and genetic characteristics (P <0.03). CIMP-only, CIMP + CIN and triple-negative tumors, compared with CIN-only tumors, were significantly associated with a 3.67, 2.44 and 3.78-fold risk of CRC-related deaths after 2-year follow-up (95% confidence intervals, CIs, 1.70-7.91, 1.35-4.41 and 1.97-7.25, respectively), but not after late follow-up. MSI tumors were borderline significantly associated with a 0.40-fold risk of CRC-related deaths after late follow-up (95% CI 0.15-1.03). This is the first study to show that specific (epi)genotypes may hold a differential prognostic value that may vary over time. Although no specific treatment data were available, an explanation for the differential findings over time might be that (epi)genotypes modify therapy response.

We studied the overlap between the major (epi)genomic events microsatellite instability (MSI), the CpG island methylator phenotype (CIMP) and chromosomal instability (CIN) in colorectal cancer (CRC), and whether specific (epi)genotypes were associated with CRC-related deaths. Molecular analyses using tumor DNA were successful in 509 CRC cases identified within the Netherlands Cohort Study in the period 1989-1993. Follow-up for the vital status until May 2005 was 100%. MSI (12.6%), CIMP-only (5.3%), CIMP + CIN (13.4%), CIN-only (58.2%) and triple-negative tumors (10.6%) differed significantly regarding tumor localization, differentiation grade, initial adjuvant therapy (AT) use and genetic characteristics (P <0.03). CIMP-only, CIMP + CIN and triple-negative tumors, compared with CIN-only tumors, were significantly associated with a 3.67, 2.44 and 3.78-fold risk of CRC-related deaths after 2-year follow-up (95% confidence intervals, CIs, 1.70-7.91, 1.35-4.41 and 1.97-7.25, respectively), but not after late follow-up. MSI tumors were borderline significantly associated with a 0.40-fold risk of CRC-related deaths after late follow-up (95% CI 0.15-1.03). This is the first study to show that specific (epi)genotypes may hold a differential prognostic value that may vary over time. Although no specific treatment data were available, an explanation for the differential findings over time might be that (epi)genotypes modify therapy response.

In recent years, attention has focused on the biology and potential clinical importance of the CpG island methylator phenotype (CIMP) in colorectal cancer (CRC). While it is generally well accepted that etiologically and clinically distinct subgroups exist in this disease, a precise definition of CIMP remains to be established. Here, we summarize existing literature that documents the prevalence of CIMP in CRC, with particular attention to the various methods and definitions used to classify a tumor as CIMP positive. Through a systematic review on both case-series and population based studies, we examined only original research articles reporting on sporadic CRC and/or adenomas in unselected cases. Forty-eight papers published between January 1999 and August 2011 met the inclusion criteria. We describe the use of multiple gene panels, marker threshold values, and laboratory techniques which results in a wide range in the prevalence of CIMP. Because there is no universal standard or consensus on quantifying the phenotype, establishing its true prevalence is a challenge. This bottleneck is becoming increasingly evident as molecular pathological epidemiology continues to offer possibilities for clear answers regarding environmental risk factors and disease trends. For the first time, large, unselected series of cases are available for analysis, but comparing populations and pooling data will remain a challenge unless a universal definition of CIMP and a consensus on analysis can be reached, and the primary cause of CIMP identified.

In recent years, attention has focused on the biology and potential clinical importance of the CpG island methylator phenotype (CIMP) in colorectal cancer (CRC). While it is generally well accepted that etiologically and clinically distinct subgroups exist in this disease, a precise definition of CIMP remains to be established. Here, we summarize existing literature that documents the prevalence of CIMP in CRC, with particular attention to the various methods and definitions used to classify a tumor as CIMP positive. Through a systematic review on both case-series and population based studies, we examined only original research articles reporting on sporadic CRC and/or adenomas in unselected cases. Forty-eight papers published between January 1999 and August 2011 met the inclusion criteria. We describe the use of multiple gene panels, marker threshold values, and laboratory techniques which results in a wide range in the prevalence of CIMP. Because there is no universal standard or consensus on quantifying the phenotype, establishing its true prevalence is a challenge. This bottleneck is becoming increasingly evident as molecular pathological epidemiology continues to offer possibilities for clear answers regarding environmental risk factors and disease trends. For the first time, large, unselected series of cases are available for analysis, but comparing populations and pooling data will remain a challenge unless a universal definition of CIMP and a consensus on analysis can be reached, and the primary cause of CIMP identified.

PURPOSE: The transcription factors GATA4 and GATA5 are involved in gastrointestinal development and are inactivated by promoter hypermethylation in colorectal cancer. Here, we evaluated GATA4/5 promoter methylation as potential biomarkers for noninvasive colorectal cancer detection, and investigated the role of GATA4/5 in colorectal cancer. EXPERIMENTAL DESIGN: Promoter methylation of GATA4/5 was analyzed in colorectal tissue and fecal DNA from colorectal cancer patients and healthy controls using methylation-specific PCR. The potential function of GATA4/5 as tumor suppressors was studied by inducing GATA4/5 overexpression in human colorectal cancer cell lines. RESULTS: GATA4/5 methylation was observed in 70% (63/90) and 79% (61/77) of colorectal carcinomas, respectively, and was independent of clinicopathologic features. Methylation frequencies in normal colon tissues from noncancerous controls were 6% (5 of 88, GATA4; P < 0.001) and 13% (13 of 100, GATA5; P < 0.001). GATA4/5 overexpression suppressed colony formation (P < 0.005), proliferation (P < 0.001), migration (P < 0.05), invasion (P < 0.05), and anchorage-independent growth (P < 0.0001) of colorectal cancer cells. Examination of GATA4 methylation in fecal DNA from two independent series of colorectal cancer patients and controls yielded a sensitivity of 71% [95% confidence interval (95% CI), 55-88%] and specificity of 84% (95% CI, 74-95%) for colorectal cancer detection in the training set, and a sensitivity of 51% (95% CI, 37-65%) and specificity of 93% (95% CI, 84-100%) in the validation set. CONCLUSIONS: Methylation of GATA4/5 is a common and specific event in colorectal carcinomas, and GATA4/5 exhibit tumor suppressive effects in colorectal cancer cells in vitro. GATA4 methylation in fecal DNA may be of interest for colorectal cancer detection.

BACKGROUND: Identification of hypermethylated tumor suppressor genes in body fluids is an appealing strategy for the noninvasive detection of colorectal cancer. Here we examined the role of N-Myc downstream-regulated gene 4 (NDRG4) as a novel tumor suppressor and biomarker in colorectal cancer. METHODS: NDRG4 promoter methylation was analyzed in human colorectal cancer cell lines, colorectal tissue, and noncancerous colon mucosa by using methylation-specific polymerase chain reaction (PCR) and bisulfite sequencing. NDRG4 mRNA and protein expression were studied using real-time-PCR and immunohistochemistry, respectively. Tumor suppressor functions of NDRG4 were examined by colony formation, cell proliferation, and migration and invasion assays in colorectal cancer cell lines that were stably transfected with an NDRG4 expression construct. Quantitative methylation-specific PCR was used to examine the utility of NDRG4 promoter methylation as a biomarker in fecal DNA from 75 colorectal cancer patients and 75 control subjects. All P values are two-sided. RESULTS: The prevalence of NDRG4 promoter methylation in two independent series of colorectal cancers was 86% (71/83) and 70% (128/184) compared with 4% (2/48) in noncancerous colon mucosa (P < .001). NDRG4 mRNA and protein expression were decreased in colorectal cancer tissue compared with noncancerous colon mucosa. NDRG4 overexpression in colorectal cancer cell lines suppressed colony formation (P = .014), cell proliferation (P < .001), and invasion (P < .001). NDRG4 promoter methylation analysis in fecal DNA from a training set of colorectal cancer patients and control subjects yielded a sensitivity of 61% (95% confidence interval [CI] = 43% to 79%) and a specificity of 93% (95% CI = 90% to 97%). An independent test set of colorectal cancer patients and control subjects yielded a sensitivity of 53% (95% CI = 39% to 67%) and a specificity of 100% (95% CI = 86% to 100%). CONCLUSIONS: NDRG4 is a candidate tumor

Aberrant DNA methylation affects carcinogenesis of colorectal cancer. Folate metabolizing enzymes may influence the bioavailability of methyl groups, whereas DNA and histone methyltransferases are involved in epigenetic regulation of gene expression. We studied associations of genetic variants of folate metabolizing enzymes (MTHFR, MTR, and MTRR), DNA methyltransferase DNMT3b, and histone methyltransferases (EHMT1, EHMT2, and PRDM2), with colorectal cancers, with or without the CpG island methylator phenotype (CIMP), MLH1 hypermethylation, or microsatellite instability. Incidence rate ratios were calculated in case-cohort analyses, with common homozygotes as reference, among 659 cases and 1,736 subcohort members of the Netherlands Cohort Study on diet and cancer (n = 120,852). Men with the MTHFR 677TT genotype were at decreased colorectal cancer risk (incidence rate ratio, 0.49; P = 0.01), but the T allele was associated with increased risk in women (incidence rate ratio, 1.39; P = 0.02). The MTR 2756GG genotype was associated with increased colorectal cancer risk (incidence rate ratio, 1.58; P = 0.04), and inverse associations were observed among women carrying DNMT3b C-->T (rs406193; incidence rate ratio, 0.72; P = 0.04) or EHMT2 G-->A (rs535586; incidence rate ratio, 0.76; P = 0.05) polymorphisms. Although significantly correlated (P < 0.001), only 41.5% and 33.3% of CIMP tumors harbored MLH1 hypermethylation or microsatellite instability, respectively. We observed inverse associations between MTR A2756G and CIMP among men (incidence rate ratio, 0.58; P = 0.04), and between MTRR A66G and MLH1 hypermethylation among women (incidence rate ratio, 0.55; P = 0.02). In conclusion, MTHFR, MTR, DNMT3b, and EHMT2 polymorphisms are associated with colorectal cancer, and rare variants of MTR and MTRR may reduce promoter hypermethylation. The incomplete overlap between CIMP, MLH1 hypermethylation, and microsatellite instability indicates that these related "methylat

PURPOSE: The transcription factors GATA4 and GATA5 are involved in gastrointestinal development and are inactivated by promoter hypermethylation in colorectal cancer. Here, we evaluated GATA4/5 promoter methylation as potential biomarkers for noninvasive colorectal cancer detection, and investigated the role of GATA4/5 in colorectal cancer. EXPERIMENTAL DESIGN: Promoter methylation of GATA4/5 was analyzed in colorectal tissue and fecal DNA from colorectal cancer patients and healthy controls using methylation-specific PCR. The potential function of GATA4/5 as tumor suppressors was studied by inducing GATA4/5 overexpression in human colorectal cancer cell lines. RESULTS: GATA4/5 methylation was observed in 70% (63/90) and 79% (61/77) of colorectal carcinomas, respectively, and was independent of clinicopathologic features. Methylation frequencies in normal colon tissues from noncancerous controls were 6% (5 of 88, GATA4; P < 0.001) and 13% (13 of 100, GATA5; P < 0.001). GATA4/5 overexpression suppressed colony formation (P < 0.005), proliferation (P < 0.001), migration (P < 0.05), invasion (P < 0.05), and anchorage-independent growth (P < 0.0001) of colorectal cancer cells. Examination of GATA4 methylation in fecal DNA from two independent series of colorectal cancer patients and controls yielded a sensitivity of 71% [95% confidence interval (95% CI), 55-88%] and specificity of 84% (95% CI, 74-95%) for colorectal cancer detection in the training set, and a sensitivity of 51% (95% CI, 37-65%) and specificity of 93% (95% CI, 84-100%) in the validation set. CONCLUSIONS: Methylation of GATA4/5 is a common and specific event in colorectal carcinomas, and GATA4/5 exhibit tumor suppressive effects in colorectal cancer cells in vitro. GATA4 methylation in fecal DNA may be of interest for colorectal cancer detection.

Aberrant DNA methylation affects carcinogenesis of colorectal cancer. Folate metabolizing enzymes may influence the bioavailability of methyl groups, whereas DNA and histone methyltransferases are involved in epigenetic regulation of gene expression. We studied associations of genetic variants of folate metabolizing enzymes (MTHFR, MTR, and MTRR), DNA methyltransferase DNMT3b, and histone methyltransferases (EHMT1, EHMT2, and PRDM2), with colorectal cancers, with or without the CpG island methylator phenotype (CIMP), MLH1 hypermethylation, or microsatellite instability. Incidence rate ratios were calculated in case-cohort analyses, with common homozygotes as reference, among 659 cases and 1,736 subcohort members of the Netherlands Cohort Study on diet and cancer (n = 120,852). Men with the MTHFR 677TT genotype were at decreased colorectal cancer risk (incidence rate ratio, 0.49; P = 0.01), but the T allele was associated with increased risk in women (incidence rate ratio, 1.39; P = 0.02). The MTR 2756GG genotype was associated with increased colorectal cancer risk (incidence rate ratio, 1.58; P = 0.04), and inverse associations were observed among women carrying DNMT3b C-->T (rs406193; incidence rate ratio, 0.72; P = 0.04) or EHMT2 G-->A (rs535586; incidence rate ratio, 0.76; P = 0.05) polymorphisms. Although significantly correlated (P < 0.001), only 41.5% and 33.3% of CIMP tumors harbored MLH1 hypermethylation or microsatellite instability, respectively. We observed inverse associations between MTR A2756G and CIMP among men (incidence rate ratio, 0.58; P = 0.04), and between MTRR A66G and MLH1 hypermethylation among women (incidence rate ratio, 0.55; P = 0.02). In conclusion, MTHFR, MTR, DNMT3b, and EHMT2 polymorphisms are associated with colorectal cancer, and rare variants of MTR and MTRR may reduce promoter hypermethylation. The incomplete overlap between CIMP, MLH1 hypermethylation, and microsatellite instability indicates that these related "methylat

BACKGROUND: Identification of hypermethylated tumor suppressor genes in body fluids is an appealing strategy for the noninvasive detection of colorectal cancer. Here we examined the role of N-Myc downstream-regulated gene 4 (NDRG4) as a novel tumor suppressor and biomarker in colorectal cancer. METHODS: NDRG4 promoter methylation was analyzed in human colorectal cancer cell lines, colorectal tissue, and noncancerous colon mucosa by using methylation-specific polymerase chain reaction (PCR) and bisulfite sequencing. NDRG4 mRNA and protein expression were studied using real-time-PCR and immunohistochemistry, respectively. Tumor suppressor functions of NDRG4 were examined by colony formation, cell proliferation, and migration and invasion assays in colorectal cancer cell lines that were stably transfected with an NDRG4 expression construct. Quantitative methylation-specific PCR was used to examine the utility of NDRG4 promoter methylation as a biomarker in fecal DNA from 75 colorectal cancer patients and 75 control subjects. All P values are two-sided. RESULTS: The prevalence of NDRG4 promoter methylation in two independent series of colorectal cancers was 86% (71/83) and 70% (128/184) compared with 4% (2/48) in noncancerous colon mucosa (P < .001). NDRG4 mRNA and protein expression were decreased in colorectal cancer tissue compared with noncancerous colon mucosa. NDRG4 overexpression in colorectal cancer cell lines suppressed colony formation (P = .014), cell proliferation (P < .001), and invasion (P < .001). NDRG4 promoter methylation analysis in fecal DNA from a training set of colorectal cancer patients and control subjects yielded a sensitivity of 61% (95% confidence interval [CI] = 43% to 79%) and a specificity of 93% (95% CI = 90% to 97%). An independent test set of colorectal cancer patients and control subjects yielded a sensitivity of 53% (95% CI = 39% to 67%) and a specificity of 100% (95% CI = 86% to 100%). CONCLUSIONS: NDRG4 is a candidate tumor

Background: A-type lamins are type V intermediate filament proteins encoded by the gene LMNA. Mutations in LMNA give rise to diverse degenerative diseases related to premature ageing. A-type lamins also influence the activity of the Retinoblastoma protein (pRb) and oncogenes such a beta-catenin. Consequently, it has been speculated that expression of A-type lamins may also influence tumour progression. Methodology/Principal Findings: An archive of colorectal cancer (CRC) and normal colon tissue was screened for expression of A-type lamins. We used the Cox proportional hazard ratio (HR) method to investigate patient survival. Using CRC cell lines we investigated the effects of lamin A expression on other genes by RT-PCR; on cell growth by FACS analysis; and on invasiveness by cell migration assays and siRNA knockdown of targeted genes. We found that lamin A is expressed in colonic stem cells and that patients with A-type lamin-expressing tumours have significantly worse prognosis than patients with A-type lamin negative tumours (HR = 1.85, p = 0.005). To understand this finding, we established a model system based upon expression of GFP-lamin A in CRC cells. We found that expression of GFP-lamin A in these cells did not affect cell proliferation but did promote greatly increased cell motility and invasiveness. The reason for this increased invasiveness was that expression of lamin A promoted up-regulation of the actin bundling protein T-plastin, leading to down regulation of the cell adhesion molecule E-cadherin. Conclusions: Expression of A-type lamins increases the risk of death from CRC because its presence gives rise to increased invasiveness and potentially a more stem cell-like phenotype. This report directly links A-type lamin expression to tumour progression and raises the profile of LMNA from one implicated in multiple but rare genetic conditions to a gene involved in one of the commonest diseases in the Western World.

Adequate intake of folate, methionine, riboflavin, and vitamin B-6 may prevent aberrant DNA methylation and thereby protect against colorectal cancer (CRC). However, previous epidemiological studies investigating associations between dietary intakes of these nutrients and CRC have been inconsistent. We investigated the associations between intakes of folate, methionine, riboflavin, and vitamin B-6 and CRC risk, accounting for the sublocalization of the tumor. Within the Netherlands Cohort Study on diet and cancer (n = 120,852), 2349 cases and 4168 subcohort members were available for data analyses from a follow-up period of 13.3 y after baseline. Gender-specific adjusted incidence rate ratios (RR) were calculated over quintiles of dietary intake in case-cohort analyses. Folate intake was not associated with CRC risk in either men or women. However, methionine was associated with decreased risk of proximal colon cancer among men (RR 0.57 for highest vs. lowest quintile of intake; P-trend = 0.03) and rectal cancer among women (highest vs. lowest quintile; RR = 0.45; P-trend = 0.05). Riboflavin tended to be associated with decreased proximal colon cancer risk among women (RR = 0.61; P-trend = 0.07). Conversely, there was a strong positive association between vitamin B-6 and rectal cancer among women (RR = 3.57; P-trend = 0.01). Our findings suggest that relatively high methionine intake may protect against proximal colon cancer in men and rectal cancer in women but that folate may not have a protective effect. This is the 2nd prospective cohort study in which vitamin B-6 intake was associated with increased risk of rectal tumors in women, which might suggest that this vitamin enhances rectal cancer in women. J. Nutr. 138: 2372-2378, 2008.

Alcohol consumption has been associated with a decreased risk for renal cell cancer in several studies. We investigated whether alcohol is associated with (epi)genetic changes of the von Hippel-Lindau WHO gene in renal cell cancer. The Netherlands Cohort Study (NLCS) on Diet and Cancer started in 1986 (n = 120,852) and uses the case-cohort method. After 11.3 years of follow-up, 314 renal cell cancer cases and 4,511 subcohort members were available for analysis. DNA was isolated from paraffin-embedded tumor tissue from 235 cases. VHL mutations were analyzed by sequencing, whereas VHL promoter methylation was analyzed using methylation-specific PCR. In multivariate analysis, hazard ratios of renal cell cancer for cohort members who consumed up to 5, 15, 30, and >= 30 g of alcohol per day were 0.72, 0.64, 0.81, and 0.69, respectively, compared with nondrinkers [95% confidence interval (95% CI) for the >= 30 category, 0.44-1.07; P for trend, 0.1.7]. Alcohol intake from beer, wine, and liquor was associated with decreased risks for renal cell cancer, although not statistically significant. Hazard ratios were not different for clear-cell renal cell cancer with and without VHL mutations, except for alcohol from beer, which was associated with an increased risk for clear-cell renal cell cancer without VHL mutations (hazard ratio for >= 5 g of alcohol from beer compared with nondrinkers, 2.74; 95% CI, 1.35-5.57). Alcohol was associated with a decreased risk for clear-cell renal cell cancer without VHL gene promoter methylation (hazard ratio for > 15 g compared with nondrinkers, 0.58; 95% CI, 0.34-0.99). In this study, a not statistically significant inverse association was observed between alcohol and renal cell cancer. There was no statistical significant heterogeneity by VHL mutation or methylation status. (Cancer Epidemiol Biomarkers Prev 2008;17(12):3543-50)

Background: A-type lamins are type V intermediate filament proteins encoded by the gene LMNA. Mutations in LMNA give rise to diverse degenerative diseases related to premature ageing. A-type lamins also influence the activity of the Retinoblastoma protein (pRb) and oncogenes such a beta-catenin. Consequently, it has been speculated that expression of A-type lamins may also influence tumour progression. Methodology/Principal Findings: An archive of colorectal cancer (CRC) and normal colon tissue was screened for expression of A-type lamins. We used the Cox proportional hazard ratio (HR) method to investigate patient survival. Using CRC cell lines we investigated the effects of lamin A expression on other genes by RT-PCR; on cell growth by FACS analysis; and on invasiveness by cell migration assays and siRNA knockdown of targeted genes. We found that lamin A is expressed in colonic stem cells and that patients with A-type lamin-expressing tumours have significantly worse prognosis than patients with A-type lamin negative tumours (HR = 1.85, p = 0.005). To understand this finding, we established a model system based upon expression of GFP-lamin A in CRC cells. We found that expression of GFP-lamin A in these cells did not affect cell proliferation but did promote greatly increased cell motility and invasiveness. The reason for this increased invasiveness was that expression of lamin A promoted up-regulation of the actin bundling protein T-plastin, leading to down regulation of the cell adhesion molecule E-cadherin. Conclusions: Expression of A-type lamins increases the risk of death from CRC because its presence gives rise to increased invasiveness and potentially a more stem cell-like phenotype. This report directly links A-type lamin expression to tumour progression and raises the profile of LMNA from one implicated in multiple but rare genetic conditions to a gene involved in one of the commonest diseases in the Western World.

Adequate intake of folate, methionine, riboflavin, and vitamin B-6 may prevent aberrant DNA methylation and thereby protect against colorectal cancer (CRC). However, previous epidemiological studies investigating associations between dietary intakes of these nutrients and CRC have been inconsistent. We investigated the associations between intakes of folate, methionine, riboflavin, and vitamin B-6 and CRC risk, accounting for the sublocalization of the tumor. Within the Netherlands Cohort Study on diet and cancer (n = 120,852), 2349 cases and 4168 subcohort members were available for data analyses from a follow-up period of 13.3 y after baseline. Gender-specific adjusted incidence rate ratios (RR) were calculated over quintiles of dietary intake in case-cohort analyses. Folate intake was not associated with CRC risk in either men or women. However, methionine was associated with decreased risk of proximal colon cancer among men (RR 0.57 for highest vs. lowest quintile of intake; P-trend = 0.03) and rectal cancer among women (highest vs. lowest quintile; RR = 0.45; P-trend = 0.05). Riboflavin tended to be associated with decreased proximal colon cancer risk among women (RR = 0.61; P-trend = 0.07). Conversely, there was a strong positive association between vitamin B-6 and rectal cancer among women (RR = 3.57; P-trend = 0.01). Our findings suggest that relatively high methionine intake may protect against proximal colon cancer in men and rectal cancer in women but that folate may not have a protective effect. This is the 2nd prospective cohort study in which vitamin B-6 intake was associated with increased risk of rectal tumors in women, which might suggest that this vitamin enhances rectal cancer in women. J. Nutr. 138: 2372-2378, 2008.

Alcohol consumption has been associated with a decreased risk for renal cell cancer in several studies. We investigated whether alcohol is associated with (epi)genetic changes of the von Hippel-Lindau WHO gene in renal cell cancer. The Netherlands Cohort Study (NLCS) on Diet and Cancer started in 1986 (n = 120,852) and uses the case-cohort method. After 11.3 years of follow-up, 314 renal cell cancer cases and 4,511 subcohort members were available for analysis. DNA was isolated from paraffin-embedded tumor tissue from 235 cases. VHL mutations were analyzed by sequencing, whereas VHL promoter methylation was analyzed using methylation-specific PCR. In multivariate analysis, hazard ratios of renal cell cancer for cohort members who consumed up to 5, 15, 30, and >= 30 g of alcohol per day were 0.72, 0.64, 0.81, and 0.69, respectively, compared with nondrinkers [95% confidence interval (95% CI) for the >= 30 category, 0.44-1.07; P for trend, 0.1.7]. Alcohol intake from beer, wine, and liquor was associated with decreased risks for renal cell cancer, although not statistically significant. Hazard ratios were not different for clear-cell renal cell cancer with and without VHL mutations, except for alcohol from beer, which was associated with an increased risk for clear-cell renal cell cancer without VHL mutations (hazard ratio for >= 5 g of alcohol from beer compared with nondrinkers, 2.74; 95% CI, 1.35-5.57). Alcohol was associated with a decreased risk for clear-cell renal cell cancer without VHL gene promoter methylation (hazard ratio for > 15 g compared with nondrinkers, 0.58; 95% CI, 0.34-0.99). In this study, a not statistically significant inverse association was observed between alcohol and renal cell cancer. There was no statistical significant heterogeneity by VHL mutation or methylation status. (Cancer Epidemiol Biomarkers Prev 2008;17(12):3543-50)

BACKGROUND: Colorectal cancers are characterized by genetic and epigenetic alterations. This study aimed to explore the timing of promoter methylation and relationship with mutations and chromosomal alterations in colorectal carcinogenesis. METHODS: In a series of 47 nonprogressed adenomas, 41 progressed adenomas (malignant polyps), 38 colorectal carcinomas and 18 paired normal tissues, we evaluated promoter methylation status of hMLH1, O6MGMT, APC, p14ARF, p16INK4A, RASSF1A, GATA-4, GATA-5, and CHFR using methylation-specific PCR. Mutation status of TP53, APC and KRAS were studied by p53 immunohistochemistry and sequencing of the APC and KRAS mutation cluster regions. Chromosomal alterations were evaluated by comparative genomic hybridization. RESULTS: Our data demonstrate that nonprogressed adenomas, progressed adenomas and carcinomas show similar frequencies of promoter methylation for the majority of the genes. Normal tissues showed significantly lower frequencies of promoter methylation of APC, p16INK4A, GATA-4, and GATA-5 (P-values: 0.02, 0.02, 1.1x10(-5) and 0.008 respectively). P53 immunopositivity and chromosomal abnormalities occur predominantly in carcinomas (P values: 1.1x10(-5) and 4.1x10(-10)). CONCLUSIONS: Since promoter methylation was already present in nonprogressed adenomas without chromosomal alterations, we conclude that promoter methylation can be regarded as an early event preceding TP53 mutation and chromosomal abnormalities in colorectal cancer development.

Folate deficiency has been associated with colorectal cancer risk and may be involved in colorectal carcinogenesis through increased chromosome instability, gene mutations, and aberrant DNA methylation. Within the Netherlands Cohort Study on diet and cancer, we investigated the associations between dietary folate intake and colorectal cancer risk with (APC(+)) and without (APC(-)) truncating APC mutations, accounting for hMLH1 expression and K-ras mutations. In total, 528 cases and 4200 subcohort members were available for data analyses of the study cohort (n = 120,852) from a follow-up period between 2.3 and 7.3 y after baseline. Adjusted gender-specific incidence rate ratios (RR) over tertiles of folate intake were calculated in case-cohort analyses for colon and rectal cancer. Although relatively high folate intake was not associated with overall colorectal cancer risk, it reduced the risk of APC(-)colon tumors in men (RR 0.58, 95% CI 0.32-1.05, P(trend) = 0.06 for the highest vs. lowest tertile of folate intake). In contrast, it was positively associated with APC(+) colon tumors in men (highest vs. lowest tertile: RR 2.77, 95% CI 1.29-5.95, P(trend) = 0.008) and was even stronger when the lack of hMLH1 expression and K-ras mutations were excluded (RR 3.99, 95% CI 1.43-11.14, P(trend) = 0.007). Such positive associations were not observed among women; nor was folate intake associated with rectal cancer when APC mutation status was taken into account. Relatively high folate consumption reduced the risk of APC(-) colon tumors, but folate intake was positively associated with APC(+) colon tumors among men. These opposite results may indicate that folate enhances colorectal carcinogenesis through a distinct APC mutated pathway

Folate deficiency has been associated with colorectal cancer risk and may be involved in colorectal carcinogenesis through increased chromosome instability, gene mutations, and aberrant DNA methylation. Within the Netherlands Cohort Study on diet and cancer, we investigated the associations between dietary folate intake and colorectal cancer risk with (APC(+)) and without (APC(-)) truncating APC mutations, accounting for hMLH1 expression and K-ras mutations. In total, 528 cases and 4200 subcohort members were available for data analyses of the study cohort (n = 120,852) from a follow-up period between 2.3 and 7.3 y after baseline. Adjusted gender-specific incidence rate ratios (RR) over tertiles of folate intake were calculated in case-cohort analyses for colon and rectal cancer. Although relatively high folate intake was not associated with overall colorectal cancer risk, it reduced the risk of APC(-)colon tumors in men (RR 0.58, 95% CI 0.32-1.05, P(trend) = 0.06 for the highest vs. lowest tertile of folate intake). In contrast, it was positively associated with APC(+) colon tumors in men (highest vs. lowest tertile: RR 2.77, 95% CI 1.29-5.95, P(trend) = 0.008) and was even stronger when the lack of hMLH1 expression and K-ras mutations were excluded (RR 3.99, 95% CI 1.43-11.14, P(trend) = 0.007). Such positive associations were not observed among women; nor was folate intake associated with rectal cancer when APC mutation status was taken into account. Relatively high folate consumption reduced the risk of APC(-) colon tumors, but folate intake was positively associated with APC(+) colon tumors among men. These opposite results may indicate that folate enhances colorectal carcinogenesis through a distinct APC mutated pathway

BACKGROUND: Colorectal cancers are characterized by genetic and epigenetic alterations. This study aimed to explore the timing of promoter methylation and relationship with mutations and chromosomal alterations in colorectal carcinogenesis. METHODS: In a series of 47 nonprogressed adenomas, 41 progressed adenomas (malignant polyps), 38 colorectal carcinomas and 18 paired normal tissues, we evaluated promoter methylation status of hMLH1, O6MGMT, APC, p14ARF, p16INK4A, RASSF1A, GATA-4, GATA-5, and CHFR using methylation-specific PCR. Mutation status of TP53, APC and KRAS were studied by p53 immunohistochemistry and sequencing of the APC and KRAS mutation cluster regions. Chromosomal alterations were evaluated by comparative genomic hybridization. RESULTS: Our data demonstrate that nonprogressed adenomas, progressed adenomas and carcinomas show similar frequencies of promoter methylation for the majority of the genes. Normal tissues showed significantly lower frequencies of promoter methylation of APC, p16INK4A, GATA-4, and GATA-5 (P-values: 0.02, 0.02, 1.1x10(-5) and 0.008 respectively). P53 immunopositivity and chromosomal abnormalities occur predominantly in carcinomas (P values: 1.1x10(-5) and 4.1x10(-10)). CONCLUSIONS: Since promoter methylation was already present in nonprogressed adenomas without chromosomal alterations, we conclude that promoter methylation can be regarded as an early event preceding TP53 mutation and chromosomal abnormalities in colorectal cancer development.

A subset of sporadic colon cancers has been shown to have microsatellite instability caused by an epigenetic inactivation of the MLH1 gene by hypermethylation of the the CpG island in its promoter region. We report here that in colorectal cancer, inactivation of the MLH1 gene is frequently accompanied by hypermethylation of the CpG island in the promoter of the mitotic gene checkpoint with forkhead and ring finger domains (CHFR). This was first observed in the colon cancer cell lines HCT-116, DLD-1, RKO and HT29. Among the 61 primary colon cancer samples studied, hypermethylation of the MLH1 and the CHFR promoter was found in 31% of the tumors. In 68% of all primary cancers (13/19) with MLH1 promoter hypermethylation, hypermethylation of the CHFR promoter was observed as well (P-value < 0.0001, Fisher's two-sided exact). Hypermethylation of the HLTF, MGMT, RASSF1, APC, p14 and p16 promoter regions were also frequent events, being observed in 48% (28/58), 40% (26/64), 21% (14/64), 50% (31/62), 43% (26/60) and 56% (35/63), respectively. However, methylation of these genes was not associated with methylation of either MLH1 or CHFR. The observed methylation profile was unrelated to Duke's stage. The coordinated loss of both mismatch repair caused by methylation of MLH1 and loss of checkpoint control associated with methylation of CHFR suggests the potential to overcome cell cycle checkpoints, which may lead to an accumulation of mutations.

We studied the association between dietary folate and specific K-ras mutations in colon and rectal cancer in The Netherlands Cohort Study on diet and cancer. After 7.3 years of follow-up, 448 colon and 160 rectal cancer patients and 3,048 sub-cohort members (55-69 years at baseline) were available for data analyses. Mutation analysis of the K-ras gene was carried out on all archival adenocarcinoma specimens. Case-cohort analyses were used to compute adjusted incidence rate ratios (RR) and 95% confidence intervals (CI) for colon and rectal cancer overall and for K-ras mutation status subgroups according to 100 mug/day increased intake in dietary folate. Dietary folate intake was not significantly associated with colon cancer risk for men or women, neither overall nor with K-ras mutation status. For rectal cancer, folate intake was associated with a decreased disease risk in men and was most pronounced for K-ras mutated tumors, whereas an increased association was observed for women. Regarding the K-ras mutation status in women, an increased association was observed for both wild-type and mutated K-ras tumors. Specifically, folate intake was associated with an increased risk of G>T and G>C transversions in rectal tumors (RR = 2.69, 95% CI = 1.43-5.09), but inversely associated with G>A transitions (RR = 0.08, 95% CI = 0.01-0.53). Our data suggest that the effect of folate on rectal cancer risk is different for men and women and depends on the K-ras mutation status of the tumor. (c) 2004 Wiley-Liss, Inc.

We studied the association between dietary folate and specific K-ras mutations in colon and rectal cancer in The Netherlands Cohort Study on diet and cancer. After 7.3 years of follow-up, 448 colon and 160 rectal cancer patients and 3,048 sub-cohort members (55-69 years at baseline) were available for data analyses. Mutation analysis of the K-ras gene was carried out on all archival adenocarcinoma specimens. Case-cohort analyses were used to compute adjusted incidence rate ratios (RR) and 95% confidence intervals (CI) for colon and rectal cancer overall and for K-ras mutation status subgroups according to 100 mug/day increased intake in dietary folate. Dietary folate intake was not significantly associated with colon cancer risk for men or women, neither overall nor with K-ras mutation status. For rectal cancer, folate intake was associated with a decreased disease risk in men and was most pronounced for K-ras mutated tumors, whereas an increased association was observed for women. Regarding the K-ras mutation status in women, an increased association was observed for both wild-type and mutated K-ras tumors. Specifically, folate intake was associated with an increased risk of G>T and G>C transversions in rectal tumors (RR = 2.69, 95% CI = 1.43-5.09), but inversely associated with G>A transitions (RR = 0.08, 95% CI = 0.01-0.53). Our data suggest that the effect of folate on rectal cancer risk is different for men and women and depends on the K-ras mutation status of the tumor. (c) 2004 Wiley-Liss, Inc.

A subset of sporadic colon cancers has been shown to have microsatellite instability caused by an epigenetic inactivation of the MLH1 gene by hypermethylation of the the CpG island in its promoter region. We report here that in colorectal cancer, inactivation of the MLH1 gene is frequently accompanied by hypermethylation of the CpG island in the promoter of the mitotic gene checkpoint with forkhead and ring finger domains (CHFR). This was first observed in the colon cancer cell lines HCT-116, DLD-1, RKO and HT29. Among the 61 primary colon cancer samples studied, hypermethylation of the MLH1 and the CHFR promoter was found in 31% of the tumors. In 68% of all primary cancers (13/19) with MLH1 promoter hypermethylation, hypermethylation of the CHFR promoter was observed as well (P-value < 0.0001, Fisher's two-sided exact). Hypermethylation of the HLTF, MGMT, RASSF1, APC, p14 and p16 promoter regions were also frequent events, being observed in 48% (28/58), 40% (26/64), 21% (14/64), 50% (31/62), 43% (26/60) and 56% (35/63), respectively. However, methylation of these genes was not associated with methylation of either MLH1 or CHFR. The observed methylation profile was unrelated to Duke's stage. The coordinated loss of both mismatch repair caused by methylation of MLH1 and loss of checkpoint control associated with methylation of CHFR suggests the potential to overcome cell cycle checkpoints, which may lead to an accumulation of mutations.

The adenomatous polyposis coli (APC) gene is considered to be a gatekeeper in colorectal tumourigenesis. Inactivating mutations in APC have been reported in 34-70% of sporadic colorectal cancer patients, the majority of which occur in the mutation cluster region (MCR). In this study, tumour tissue from 665 incident colorectal cancer patients, who originate from 120 852 men and women (55-69 years of age at baseline) participating in The Netherlands Cohort Study, was evaluated for the occurrence and type of APC mutations with regard to age at diagnosis, gender, family history of colorectal cancer, Dukes' stage, tumour differentiation and sub-localization. Mutation analysis of the MCR, which spans codons 1286-1513, was performed on archival adenocarcinoma samples using macrodissection, nested PCR and direct sequencing of purified PCR fragments. A large number of genetic aberrations (n = 978), including point mutations (n = 833), deletions (n = 126) and insertions (n = 19) was detected in the MCR in 72% of patients (479/665). In particular, we observed a large number of missense mutations, more than reported previously. This may indicate involvement in colorectal carcinogenesis, although their significance for APC functions is unclear. Truncating mutations were found in 37% of patients (248/665). Patients with rectosigmoid and rectum tumours relatively more frequently harboured C > T nonsense mutations and truncating frameshift mutations as compared with patients with proximal and distal colon tumours (P = 0.009 and P = 0.045, respectively). Differences in occurrence of truncating mutations with regard to tumour sub-localization suggest a different aetiology of tumourigenesis in colon and rectum.

The adenomatous polyposis coli (APC) gene is considered to be a gatekeeper in colorectal tumourigenesis. Inactivating mutations in APC have been reported in 34-70% of sporadic colorectal cancer patients, the majority of which occur in the mutation cluster region (MCR). In this study, tumour tissue from 665 incident colorectal cancer patients, who originate from 120 852 men and women (55-69 years of age at baseline) participating in The Netherlands Cohort Study, was evaluated for the occurrence and type of APC mutations with regard to age at diagnosis, gender, family history of colorectal cancer, Dukes' stage, tumour differentiation and sub-localization. Mutation analysis of the MCR, which spans codons 1286-1513, was performed on archival adenocarcinoma samples using macrodissection, nested PCR and direct sequencing of purified PCR fragments. A large number of genetic aberrations (n = 978), including point mutations (n = 833), deletions (n = 126) and insertions (n = 19) was detected in the MCR in 72% of patients (479/665). In particular, we observed a large number of missense mutations, more than reported previously. This may indicate involvement in colorectal carcinogenesis, although their significance for APC functions is unclear. Truncating mutations were found in 37% of patients (248/665). Patients with rectosigmoid and rectum tumours relatively more frequently harboured C > T nonsense mutations and truncating frameshift mutations as compared with patients with proximal and distal colon tumours (P = 0.009 and P = 0.045, respectively). Differences in occurrence of truncating mutations with regard to tumour sub-localization suggest a different aetiology of tumourigenesis in colon and rectum.

Sporadic colorectal cancer (CRC) is characterized by genetic and epigenetic changes such as regional DNA hypermethylation and global DNA hypomethylation. Epidemiological and animal studies suggest that aberrant DNA methylation is associated with low dietary folate intake, which is aggravated by high alcohol intake. The relationship between promoter methylation of genes involved in CRC carcinogenesis and folate and alcohol intake was investigated. Methylation of the APC-1A, p14(ARF), p16(INK4A), hMLH1, O(6)-MGMT, and RASSF1A promoters was studied using methylation-specific PCR in 122 sporadic CRCs, derived from patients with folate and alcohol intake at either the lower or the higher quintiles of the distribution. Overall, promoter hypermethylation frequencies observed were: 39% for APC; 33% for p14(ARF); 31% for p16(INK4A); 29% for hMLH1; 41% for O(6)-MGMT; and 20% for RASSF1A. For each of the tested genes, the prevalence of promoter hypermethylation was higher in CRCs derived from patients with low folate/high alcohol intake (n = 61) when compared with CRCs from patients with high folate/low alcohol intake (n = 61), but the differences were not statistically significant. The number of CRCs with at least one gene methylated was higher (84%) in the low folate intake/high alcohol intake group when compared with the high folate intake/low alcohol intake group (70%; P = 0.085). Despite the size limitations of this study, these data suggest that folate and alcohol intake may be associated with changes in promoter hypermethylation in CRC

The Sidney Kimmel Comprehensive Cancer Center at Johns Hopkins, Johns Hopkins University School of Medicine, 1650 Orleans Street, Baltimore, Maryland 21231, USA. Aberrant hypermethylation of gene promoters is a major mechanism associated with inactivation of tumor-suppressor genes in cancer. We previously showed this transcriptional silencing to be mediated by both methylation and histone deacetylase activity, with methylation being dominant. Here, we have used cDNA microarray analysis to screen for genes that are epigenetically silenced in human colorectal cancer. By screening over 10,000 genes, we show that our approach can identify a substantial number of genes with promoter hypermethylation in a given cancer; these are distinct from genes with unmethylated promoters, for which increased expression is produced by histone deacetylase inhibition alone. Many of the hypermethylated genes we identified have high potential for roles in tumorigenesis by virtue of their predicted function and chromosome position.We also identified a group of genes that are preferentially hypermethylated in colorectal cancer and gastric cancer. One of these genes, SFRP1, belongs to a gene family; we show that hypermethylation of four genes in this family occurs very frequently in colorectal cancer, providing for (i) a unique potential mechanism for loss of tumor-suppressor gene function and (ii) construction of a molecular marker panel that could detect virtually all colorectal cancer.

The Sidney Kimmel Comprehensive Cancer Center at Johns Hopkins, Johns Hopkins University School of Medicine, 1650 Orleans Street, Baltimore, Maryland 21231, USA. Aberrant hypermethylation of gene promoters is a major mechanism associated with inactivation of tumor-suppressor genes in cancer. We previously showed this transcriptional silencing to be mediated by both methylation and histone deacetylase activity, with methylation being dominant. Here, we have used cDNA microarray analysis to screen for genes that are epigenetically silenced in human colorectal cancer. By screening over 10,000 genes, we show that our approach can identify a substantial number of genes with promoter hypermethylation in a given cancer; these are distinct from genes with unmethylated promoters, for which increased expression is produced by histone deacetylase inhibition alone. Many of the hypermethylated genes we identified have high potential for roles in tumorigenesis by virtue of their predicted function and chromosome position.We also identified a group of genes that are preferentially hypermethylated in colorectal cancer and gastric cancer. One of these genes, SFRP1, belongs to a gene family; we show that hypermethylation of four genes in this family occurs very frequently in colorectal cancer, providing for (i) a unique potential mechanism for loss of tumor-suppressor gene function and (ii) construction of a molecular marker panel that could detect virtually all colorectal cancer.

Human cancer is characterized by genetic and epigenetic alterations. In this study we provide evidence for the interruption of Ras signaling in sporadic colorectal cancer (CRC) by either genetic activation of the K-ras oncogene or epigenetic silencing of the putative tumor suppressor gene RASSF1A. Paraffin embedded tumor tissue samples from 222 sporadic CRC patients were analysed for K-ras codon 12 and codon 13 activating mutations and RASSF1A promoter hypermethylation. Overall, K-ras mutations were observed in 87 of 222 (39%) and RASSF1A methylation was observed in 45 of 222 (20%) of CRCs. Mutation of K-ras alone was detected in 76 of 222 (34%) CRCs. RASSF1A promoter methylation with wild-type K-ras was observed in 34 of 222 (15%) CRCs. In 101 of 222 (46%) CRCs neither K-ras mutations nor RASSF1A methylation was observed and 11 of 222 (5%) CRCs showed both K-ras mutations and RASSF1A methylation. These data show that the majority of the studied CRCs with K-ras mutations lack RASSF1A promoter methylation, an event which occurs predominantly in K-ras wild-type CRCs (P=0.023, Chi-square test).