DNA demethylation in normal colon tissue predicts predisposition to multiple cancers


Some colon cancer (CC) patients present synchronous cancers at diagnosis and others develop metachronous neoplasms, but the risk factors are unclear for non-hereditary CC. We showed previously that global DNA demethylation increased with aging and correlated with genomic damage in CC, and we show now that preferentially associates to CCs with wild-type p53. This study aimed to elucidate the extent of DNA hypomethylation in patients with single and multiple CC, its relationship with aging, and its potential as predictive tool. We compared by real-time methylation-specific PCR the relative demethylation level (RDL) of long interspersed nucleotide element-1 (LINE-1) sequences in matched cancer tissues and non-cancerous colonic mucosa (NCM) from patients with single and multiple right-sided CCs. Although no RDL difference was found in NCM from single CC patients and healthy volunteers (P=0.5), there was more demethylation (higher RDL) in NCM from synchronous cancer patients (P=1.1 × 10−5) multiple CCs also were more demethylated than single CCs (P=0.0014). High NCM demethylation was predictive for metachronous neoplasms (P=0.003). In multivariate logistic regression analyses RDL was the only independent predictor for metachronous (P=0.02) and multiple (P=4.9 × 10−5) tumors. The higher LINE-1 demethylation in NCM from patients with multiple (synchronous and metachronous) tumors (P=9.6 × 10−7) was also very significant in patients with tumors without (P=3.8 × 10−6), but not with (P=0.16) microsatellite instability. NCM demethylation increased with aging in patients with single tumors, but decreased in those with multiple tumors. Moreover, the demethylation difference between patients with single vs multiple tumors appeared higher in younger (P=3.6 × 10−4) than in older (P=0.0016) patients. These results predict that LINE-1 hypomethylation in NCM can be used as an epigenetic predictive biomarker for multiple CC risk. The stronger association of demethylation in NCM with multiple CC risk from younger patients also suggests an inherited predisposition for the apparent field cancerization effect of somatic demethylation.


Patients undergoing surgical resection for colon cancer (CC) often present synchronous tumors and are at higher risk for developing metachronous neoplasms in the residual colon than those without a history of cancer.1 Although patients with hereditary syndromes such as non-polyposis colorectal cancer (HNPCC) have a high risk for developing synchronous and metachronous tumors,2 the underlying mechanisms in apparently non-hereditary (that is, ‘sporadic’) multiple colorectal cancer have not been elucidated.3 The first concluding unanswered key research question in the Rex et al., review article was ‘What clinical, genetical, or biological markers predict development of metachronous cancers (that is, stratify risk) in colorectal cancer patients without hereditary non-polyposis colorectal cancer?’.

Multiple cancers may arise simultaneously from regions of normal tissue containing certain genomic alterations. This phenomenon is exemplified by the ‘field cancerization’ caused by carcinogen exposure.4, 5 In contrast with genetic alterations found in isolated cases and in only minor fractions of normal tissue cells,4, 6 somatic epigenetic alterations are common in normal tissues adjacent to various cancers.7, 8, 9, 10, 11, 12, 13 Thus, DNA hypermethylation, associated to inactivation of tumor suppressors, has been proposed to be an indicator of epigenetic field cancerization.12, 14, 15

A lower level of DNA methylation in tumors compared with normal tissue is another epigenetic alteration that has been found in many human cancers.16, 17 Evidence from mice model systems links DNA hypomethylation with chromosomal instability and cancer.18, 19 Association of somatic demethylation in repetitive elements such as long interspersed nucleotide element-1 (LINE-1) as marker for global demethylation and colorectal cancer progression has been recently described.20, 21, 22, 23

We previously showed that in human gastrointestinal cancers, DNA hypomethylation was an independent prognostic indicator of poor survival for gastrointestinal cancer, increased with aging and correlated with age-independent genomic damage.24, 25 This led to the proposal of a ‘wear and tear’ model for cancer development linking aging and cancer through the accumulation of DNA demethylation and its causal relationship with genomic damage.

Demethylation in normal colonic mucosa of some cancer patients has been also described previously26 and lower methylation of individual genes was found in normal mucosa of patients with colon adenomas and carcinomas.8, 11, 13, 27 Recent studies also reported concordant DNA methylation alterations between synchronous colorectal cancers.9, 28

The aim of this study was to elucidate whether the extent of DNA hypomethylation in non-cancerous colonic mucosa (NCM) also increases with aging, and whether demethylation in normal tissue differs between healthy individuals and patients with single and multiple colon carcinoma (CC). We also explored the potential application of the degree of DNA demethylation in normal colon epithelium as a predictive tool for multiple CC. None of these questions have been previously addressed.



The patients recruited in this study and the clinico-histological features of the corresponding colon tumors are presented in Table 1. Additional clinicopathological and demographic information is found in Materials and methods and Supplementary Information. From 694 total CC patients we selected those with right-sided CC (n=270). We recruited only right-sided CC patients because left-sided sporadic CCs have different molecular characteristics compared with right-sided sporadic CCs. In addition, the larger distance requirement for surgical margins in right-sided CCs leads to low rates of local recurrence at anastomotic sites. Therefore, we were able to distinguish metachronous colonic neoplasms from local recurrences at anastomotic sites in right-sided CC patients. A diagram of the study design is illustrated in Figure 1.

Table 1 Clinicopathological and demographic data
Figure 1

Diagram of the study design. For details see text and Materials and methods (patients and samples).

DNA hypomethylation and a ‘wear and tear’ model for CC

In our studies using DNA fingerprinting as diagnostic methods, we found a correlation between increased overall demethylation levels in CC with aging and genomic damage.25, 29 We report now that the tendency of demethylation to increase with aging was more evident in those CC patients harboring wild-type p53 (Figure 2). The association between patient age and hypomethylation alterations did not exhibit substantial changes when the samples were stratified according to the mutational status of KRAS or BRAF (not shown).

Figure 2

DNA hypomethylation correlates with patient age (a), especially from patients with tumors with wild-type p53 (b). Hypomethylation alterations were identified by methylation-sensitive amplification length polymorphism (MS-AFLP) in a panel of 77 colorectal tumors and matching normal tissues.25 The mutational status of several genes (TP53, KRAS and BRAF) was determined in these tumor samples as described.24 P-values were calculated using the Pearson's product-moment correlation test.

We next analyzed the extent of methylation in the normal tissue of CC patients, as well as in healthy volunteers, to determine whether the degree of demethylation in normal tissues could distinguish healthy individuals from CC patients. We used a relative demethylation level (RDL) assay of LINE-1 repetitive sequences using the MethyLight approach30 as a surrogate for global demethylation levels. The higher RDL means therefore less methylation. We designed a prospective study with a cohort of CC patients with a 7-year follow-up. Of the 196 patients eligible for follow-up, 99 were analyzed for RDL in NCM. When we correlated RDL and patient age, demethylation of NCM appeared higher in older CC patients when the total patient sample was considered (P=0.017) in support of the ‘wear and tear’ model (Supplementary Figure 1b).

DNA demethylation in NCM associates with synchronous CC

Next, we compared the RDL of NCM from healthy individuals with that of CC patients. Although no difference in methylation was found in NCM from single CC patients and healthy volunteers, the NCM from synchronous cancer patients showed higher RDL (that is, more demethylation) than from single CCs. Synchronous CCs also displayed more demethylation (higher RDL) than single CCs (Table 2a). We tested next whether the higher demethylation in NCM from patients with synchronous CC also extended to those patients who developed later new (metachronous) tumors after colectomy. This prospective study revealed that the level of demethylation of the NCMs from the initial surgery correlated with the risk to develop metachronous tumors (Table 2b).

Table 2 LINE-1 methylation data in right-sided colon cancer patients

DNA demethylation in NCM and microsatellite instability (MSI)

Once the result with synchronous cancers was validated with metachronous tumors, we combined the data (that is, ‘multiple’ tumors, Figure 3). The differences in demethylation level between single and multiple tumors (P=9.6 × 10−7) remained extremely significant (P=3.8 × 10−6) after filtering out the patients with MSI cancer31, 32 both in tumor (Figure 3a) and in NCM (Figure 3b). MSI-positive cancers displayed less demethylation than those without MSI (Figure 3a, right) in agreement with previous reports.23, 33 In contrast, no differences in methylation levels were apparent in NCMs from patients with CC with or without MSI (Figure 3b, right). Therefore, the results exclude the association of multiple tumors with MSI and hence with HNPCC.

Figure 3

LINE-1 RDL in tumor and normal colonic mucosa of right-side CC patients. (a) LINE-1 RDL in the tumor samples from 32 patients with single right-side CC (single, in light gray) and 14 patients with right-side double synchronous cancers (multiple, in dark gray). (b) LINE-1 RDL in the NCM samples from 75 patients with right-side single CC (single, in light gray) and 24 patients who either presented right-side double synchronous CCs, or developed a second metachronous lesion within the following 8 years after surgical resection of the first cancer (multiple, in dark gray). In all cases, the NCM LINE-1 RDL value refers to the normal tissue sample resected at the time of the initial surgery of the primary tumor. Differences between groups were analyzed by the non-parametric Mann–Whitney–Wilcoxon rank-sum test. Horizontal lines indicate the median value for each subgroup. The primers, probes and PCR conditions, as well as the MSI analysis are described in detail in Supplementary Table 1. P-values <0.01 are indicated in bold type.

DNA demethylation in NCM predicts development of metachronous colon tumors

During the follow-up period 10 patients developed a metachronous tumor and 69 patients did not. Of the 10 patients with metachronous neoplasms, 4 developed carcinomas and 6 adenomas (Supplementary Table 2). As only four carcinomas were insuficient to perform statistical comparisons, we grouped the four carcinomas with the six adenomas (Table 2), which is justified as advanced adenomas exhibit a high risk of progressing into carcinomas,34 and are regularly removed after detection, especially in patients with a history of CC.3

First, we performed univariate logistic regression analyses to determine the association between several factors (patient age, gender, tumor location, invasiveness, MSI status and RDL levels in NCM) with the probability of developing a metachronous tumor independently. The univariate logistic regression analyses revealed that the only factors with statistical significance were patient age (P=0.026) and RDL levels in the NCM (P=0.010). A multivariate logistic regression analysis including all these factors together demonstrated that the only factor that retained statistical significance was demethylation in NCM (P=0.0254). A backward-elimination stepwise logistic regression yielded a best-fitting model containing only patient age and demethylation as predictors of metachronous tumor development. In this model the only factor maintaining the statistical significance (P=0.028) was demethylation of normal mucosa (Table 3a).

Table 3 Univariate and multivariate logistic regression analysis

In a similar approach, we analyzed the association between these factors and the probability of developing a multiple tumor, either synchronous or metachronous. For this analysis, we used the data from 75 patients with single CC, 14 patients with double synchronous CC and 10 patients who had a single CC and developed metachronous tumors in the follow-up period. In the univariate logistic regression analyses, three of these factors were statistically significant: patient age (P=0.036), invasiveness (P=0.043) and RDL in NCM (P=1.9 × 10−5). The multivariate logistic regression analysis showed that, as it happened in the previous study, demethylation of the NCM was the only factor retaining statistical significance (P=5.8 × 10−5). The best-fitting model obtained after backward-elimination stepwise logistic regression contained patient age, invasiveness and RDL in NCM as predictors of multiple tumors. In this model, once again, RDL in NCM was the only factor maintaining statistical significance (P=4.9 × 10−5) associated to the development of multiple tumors (Table 3b).

Patients with more demethylation (higher RDL) in NCM at time of surgery developed more metachronous tumors during the follow-up period (Figure 4 and Supplementary Table 2).

Figure 4

Case–control study of metachronous tumor incidence in 76 patients with a single right-side CC who underwent curative surgery. Ten of them developed a right-side metachronous neoplasm in the follow-up period (6–100 months). Patients were stratified into two groups (high and low) according to the LINE-1 RDL in the NCM, using three different thresholds: RDL>4, RDL>5 and RDL>6 (ac, respectively), and metachronous tumor incidence was studied using the log rank test. Threshold 5 represents approximately the median of RDL values in the entire patient sample. The P-values, the number of cases above and below the cutoff, as well as the sensitivity and specificity of the classification are indicated for every used cutoff. Cases with LINE-1 RDL higher than the stratification threshold (‘high risk’) are represented with a solid black line, while cases with LINE-1 RDL lower than the stratification threshold (‘low risk’) are represented by a dashed line. X-axis indicates the time after the surgical resection of the first lesion. Y-axis represents the proportion of patients who developed a second (metachronous) lesion in the right-side colon. Points in the curves correspond to censored cases. The results still retain statistically significance by applying Fisher's exact test (under the alternative hypothesis that true odds ratio is >1) with an event at 36 months set as the endpoint: P=0.05 for cutoff of 4; P=0.015 for the RDL cutoff of 5 and P=0.045 for cutoff of 6.

DNA demethylation and age of patients with multiple colon tumors

Patients with multiple tumors were older than patients with single tumors (68.3±8.3 vs 72.5±7.8, P=0.03). When considered into groups of single and multiple tumors, there was more demethylation (higher RDL) in older patients in the single tumor group (Figure 5a), but less demethylation in older patients with multiple neoplasms (Figure 5b). The difference in demethylation between patients with single vs multiple tumors seemed to be more significant in younger (Figure 5c) than older (Figure 5d) patients. Furthermore, the slopes of the RDL values in relation to aging were inverse, showing that demethylation in NCM increased with aging in patients with single tumors, but decreased in those with multiple tumors (Supplementary Figure 2).

Figure 5

Association between NCM LINE-1 RDL of single or multiple CC and patient age. LINE-1 RDL was analyzed in the NCM from 75 patients with single (in light gray) and 24 patients with multiple (in dark gray) right-side neoplasms. This latter group comprised 14 patients with synchronous double CCs and 10 patients who had a single cancer and developed a metachronous neoplasm in the following 8 years after surgical resection of the first lesion. Cases were categorized into younger (circles) and older (squares) according to the median age of the compared groups (68 for single tumor patients, 72.5 for multiple tumor patients and 70 for the entire sample population) (a) NCM RDL values of younger vs older patients with a single neoplasm. (b) NCM RDL values of younger vs older patients with multiple neoplasms. (c) NCM RDL values of younger patients with a single neoplasm vs younger patients with multiple neoplasms. (d) NCM RDL values of older patients with a single neoplasm vs older patients with multiple neoplasms. Horizontal lines indicate the median value for each subgroup. P-values were calculated using the non-parametric Mann–Whitney–Wilcoxon rank sum test. P-values <0.01 are represented in bold type.


The results showing that normal colon mucosa also accumulates DNA demethylation with aging (Supplementary Figure 1b) add further support to the ‘wear and tear’ model.25 The findings that the age-associated accumulation of demethylation appears more pronounced in those CC patients with wild-type p53 (Figure 1) suggest the existence of a demethylation-driven genomic instability oncogenic pathway for CC independent of p53 signaling inactivation. We postulate that in a significant fraction of patients without p53 mutations, CC development can be explained by a stochastic accumulation of epigenetic demethylation errors in the continuously replicating stem cells of the colon crypts. These errors increase the risk of mitotic mistakes, eventually leading to chromosomal instability and cancer. One possible mechanism linking DNA hypomethylation and subsequent genetic alterations is illustrated by the observation that demethylation of sensitive pericentromeric repetitive sequences can trigger errors in chromosomal replication and segregation, leading to DNA copy number alterations during mitosis.19, 35, 36, 37

In several studies, a correlation has been observed between the degree of hypomethylation of satellite or LINE-1 sequences and the frequency of chromosomal aberrations in human cancer, in agreement with results from mouse models.18 LINE-1 hypomethylation is significantly associated with patient age and has been shown to be a prognostic factor for overall and progression-free survival of CC patients.21, 28 Association between early cancer DNA hypomethylation, nuclear morphometric changes and chromosomal ploidy status was also found in a study of ovarian cancer.38

MSI is the landmark of HNPCC wherein the affected members are at high risk for developing synchronous and metachronous CC.2, 3, 39 Therefore, we excluded all patients with family history of CC. In addition, the negative association of high demethylation with MSI adds further evidence that the described phenomenon linking demethylation with risk for synchronous and metachronous neoplasms is not because of undetected HNPCC patients. MSI CC due to defective DNA mismatch repair has distinct clinicopathological features, including proximal colon predominance, quasi-diploid DNA content, poor differentiation, less advanced stage of progression and better survival.31, 32, 40, 41, 42 Our results provide additional evidence that the colorectal cancer pathway followed by MSI tumors differs from the rest of cancers without MSI: while no difference was found in methylation levels of NCM from patients with or without MSI (Figure 3b), MSI cancers displayed more demethylation (lower RDL) than cancers without MSI (Figure 3a). This lower degree of demethylation in MSI cancers compared with MSS cancers was, however, higher than the level in normal tissue, showing that extensive demethylation in precursor CC cells also occurs before the manifestation of the mutator phenotype. This speculation is based on the fact that the somatic changes taking place after neoplastic transformation are relatively minor compared with those occurring before, in the normal precursor cells of MSI tumors.43 The staggering amount of somatic mutations in microsatellite sequences (hundred of thousands) accumulated by these MSI tumors also argues strongly that these clonal mutations accumulate before transformation.31

Multivariate analysis showed that demethylation was the only independent significant predictor of outcome. When introducing age and RDL as continuous variables in stepwise logistic regression analyses the best-fitting models retained RDL and age, with P-values of P=0.028 (RDL) and P=0.079 (age) for metachronous tumor development, and P=4.2 × 10−5 (RDL), P=0.099 (age) and P=0.106 (stage) for synchronous plus metachronous tumors. However, in other regression analyses whereby age and RDL were considered as categorical variables, age was excluded using different cutoffs and instead anatomical location remained in the best-fitting model together with RDL, although again, demethylation was the only significant factor (data not shown). Therefore, we conclude that age is not a risk factor for future neoplasm development in the proximal colon. That conclusion is consistent with the data: while demethylation increased with age in patients with single cancers, the cases with multiple tumors (metachronous and synchronous), while older overall than the entire patient population, were relatively ‘younger’—if <70 years of age can be so considered (Figure 5 and Supplementary Figure 2a).

We interpret these results as suggesting that the individuals with demethylation levels above the continuous accumulation of background demethylation errors during aging represent a distinct subgroup of relatively younger individuals with a propensity to develop metachronous neoplasms in the proximal colon. The obvious, highly speculative, but parsimonious hypothesis postulates that this may be due to an inherited predisposition.

The association of demethylation not only in tumor tissue, but also in normal colon tissue of CC patients with multiple tumors provides mechanistic insights for the origins of the enhanced demethylation and is consistent with a role of demethylation in a long range field cancerization (Supplementary Table 3). Obvious candidates for the origin of demethylation are genetic or epigenetic defects in nuclear factors involved in the replication of somatic methylation, including the methyl transferases and their associated proteins. Whether these factors, defective or functional, are the same or different in the two types of progressive demethylation in the normal colon epithelium disclosed by our results, that of patients with single or multiple tumors in the proximal colon, remains to be identified and characterized.

Regardless of the underlying mechanisms, our results show that DNA hypomethylation assessed by LINE-1 RDL in NCM may be useful for stratification of non-HNPCC patients into ‘low’ and ‘high’ risk groups for subsequent development of metachronous neoplasms with high sensitivity (Figure 4). To be precise, our data only provide evidence for metachronous tumors (carcinomas and adenomas) developing in the right-sided colon remnant following resection of right-sided cancers, and furthermore, in patients that did not develop left-sided metachronous tumors as these patients were excluded from the study, together with those patients who developed metastases from the primary CC or other extracolonic neoplasms (Figure 1 and Materials and methods). Whether development of distal metachronous neoplasms from proximal CC or proximal or distal metachronous neoplasms from distal CC, also associates with high demethylation of the NCM resected during surgery of the primary cancer remains to be determined. Nevertheless, for at least proximal metachronous tumors from proximal CC, patients at high risk should be placed under more frequent surveillance, while patients at lower risk should be counseled, but undergo less frequent surveillance. Second generation diagnostic tests with higher specificity can be envisioned to enhance the clinical application of DNA demethylation in LINE-1 and possibly other genomic sequences from NCMs and body fluids.

In conclusion, our results are consistent with the existence of at least two distinct roles of DNA demethylation in CC development: (1) a gradual demethylation associated to single cancers in old patients generally without mutant p53, and (2) a more drastic demethylation undergone by relatively younger patients possibly underlying a field cancerization effect for the development of multiple neoplasms. These results also suggest that the enhanced demethylation observed in NCM of patients with multiple CC not only may underlie a field cancerization effect that predisposes to cancer development, but also that this defect may be due to an inherited susceptibility.

Materials and methods

Patients and samples

Tumor tissues and corresponding NCM were obtained from patients with CC who underwent surgery from May 2001 to February 2007 at the Saitama Medical Center, Jichi Medical University. The samples of normal-appearing right-sided colonic mucosa from 11 patients without malignant disease (healthy volunteers) were obtained from biopsy specimens collected during screening colonoscopies and from resected specimens from patients with diverticulitis and appendicitis. This study was approved by the Research Ethics Committee and written informed consent was obtained from each participant. A diagram of the study design is shown in Figure 1. From 694 total CC patients we selected those with right-sided CCs (n=270). Of the right side CCs, 211 underwent curative surgery after excluding 49 patients: 22 patients with insufficient clinical information, 15 patients with three or more CC and 3 patients with 1CC+1ADE, 8 patients enrolled in a different study, and 1 patient suspected to have HNPCC (45 years old female MSI positive). To minimize the confounding factor of multiple cancers due to hereditary predisposition factors such as familial polyposis (FAP) or HNPCC or inflammatory bowel disease, we also excluded 11 cases. Of the final 196 eligible patients, 14 had double cancers, and for comparison we selected a random set of 32 patients with single cancers, in addition to 11 healthy volunteers. The 32 patients with single cancers matched the overall population of single CC bearing patients in regards to age (68.3±7.7 vs 66.7±11.3, P=0.31, Student's t-test), in addition to location.

Next, we designed a prospective study of patients that developed metachronous tumors, to independently validate the findings with the synchronous cancer patients and to test the predictive value of the NCMs RDL. 196 initially eligible patients from the entire series yielded 126 with informative follow-up. 27 were excluded because of development of metastases from the primary tumor, or death, and 14 and 6 were excluded because of presence of left-side metachronous or extra colonic neoplasms, respectively. Of the remaining sample 10 patients developed right-side metachronous neoplasms in the follow-up period, while 69 patients showed no evidence of metachronous tumors.

MS-AFLP DNA fingerprinting

Hypomethylation alterations were identified by methylation-sensitive amplification length polymorphism (MS-AFLP) in a panel of 77 colorectal tumors and matching normal tissues.25 The mutational status of several genes (TP53, KRAS and BRAF) was determined in these tumor samples as described.24

RDL DNA methylation assay

Samples from CC patients were obtained from the Saitama Medical Center, Jichi Medical University, Saitama, Japan. After surgical resection, tissue specimens were immediately soaked in RNAlater (Ambion; Austin, TX) and stored at −80 °C after the RNAlater solution was removed. Before DNA preparation, the dissected tissue was placed in buffered proteinase K solution at 56 °C for 3 h. Genomic DNA was isolated and purified using a BioRobot EZ1 workstation and an EZ1 DNA tissue kit (Qiagen, Hilden, Germany) according to the manufacturer's instructions. FAM: 6-carboxyfluorescein (fluorophore). MGB: dihydrocyclopyrroloindole tripeptide minor groove binder (quencher).

RDL was estimated by quantitative PCR of demethylated LINE-1 elements amplified from bisulfite-treated genomic DNA.30 LINE-1 RDL represents the proportion of completely demethylated LINE-1 elements in the analyzed sample relative to a reference sample. ALU sequences are used for normalization. Primers are listed in Supplementary Table 1 and conditions were essentially as previously published,30 with some modifications in the LINE-1 probe sequence, that was re-designed using primer express software (Applied Biosystems, Foster City, CA). Human DNA from lymphocytes was used as reference. Precision estimations were calculated independently for the LINE-1 and ALU reactions as the ratio of the standard deviations of the threshold cycles (Ct) divided by the mean of the Ct using 3–5 examinations. MSI was determined as described.25, 29

Statistical analyses

Statistical analyses were performed using R Environment for Statistical Computing software.44 Correlation between continuous variables was analyzed using the Pearson's product-moment correlation test. Differences between groups were analyzed by the non-parametric Mann–Whitney–Wilcoxon rank-sum test for those variables not following a normal distribution (RDL) or by Student's t-test for those variables following a normal distribution (patient age). Deviation from normality was assessed using the Shapiro–Wilk normality test. Differences in metachronous tumor incidence during the follow-up period was studied using the log rank test. The significance of associations was analyzed using Fisher's exact tests for 2 × 2 contingency tables or by χ2 test for larger dimension contingency tables.


  1. 1

    Bouvier AM, Latournerie M, Jooste V, Lepage C, Cottet V, Faivre J . The lifelong risk of metachronous colorectal cancer justifies long-term colonoscopic follow-up. Eur J Cancer 2008; 44: 522–527.

  2. 2

    Lynch HT, Smyrk TC, Watson P, Lanspa SJ, Lynch JF, Lynch PM et al. Genetics, natural history, tumor spectrum, and pathology of hereditary nonpolyposis colorectal cancer: an updated review. Gastroenterology 1993; 104: 1535–1549.

  3. 3

    Rex DK, Kahi CJ, Levin B, Smith RA, Bond JH, Brooks D et al. Guidelines for colonoscopy surveillance after cancer resection: a consensus update by the American Cancer Society and the US Multi-Society Task Force on Colorectal Cancer. Gastroenterology 2006; 130: 1865–1871.

  4. 4

    Braakhuis BJ, Tabor MP, Kummer JA, Leemans CR, Brakenhoff RH . A genetic explanation of Slaughter's concept of field cancerization: evidence and clinical implications. Cancer Res 2003; 63: 1727–1730.

  5. 5

    Slaughter DP, Southwick HW, Smejkal W . Field cancerization in oral stratified squamous epithelium; clinical implications of multicentric origin. Cancer 1953; 6: 963–968.

  6. 6

    Deng G, Lu Y, Zlotnikov G, Thor AD, Smith HS . Loss of heterozygosity in normal tissue adjacent to breast carcinomas. Science 1996; 274: 2057–2059.

  7. 7

    Ahlquist T, Lind GE, Costa VL, Meling GI, Vatn M, Hoff GS et al. Gene methylation profiles of normal mucosa, and benign and malignant colorectal tumors identify early onset markers. Mol Cancer 2008; 7: 94.

  8. 8

    Belshaw NJ, Pal N, Tapp HS, Dainty JR, Lewis MP, Williams MR et al. Patterns of DNA methylation in individual colonic crypts reveal aging and cancer-related field defects in the morphologically normal mucosa. Carcinogenesis 2010; 31: 1158–1163.

  9. 9

    Konishi K, Shen L, Jelinek J, Watanabe Y, Ahmed S, Kaneko K et al. Concordant DNA methylation in synchronous colorectal carcinomas. Cancer Prev Res (Phila) 2009; 2: 814–822.

  10. 10

    Menigatti M, Truninger K, Gebbers JO, Marbet U, Marra G, Schar P . Normal colorectal mucosa exhibits sex- and segment-specific susceptibility to DNA methylation at the hMLH1 and MGMT promoters. Oncogene 2009; 28: 899–909.

  11. 11

    Paun BC, Kukuruga D, Jin Z, Mori Y, Cheng Y, Duncan M et al. Relation between normal rectal methylation, smoking status, and the presence or absence of colorectal adenomas. Cancer 2010; 116: 4495–4501.

  12. 12

    Ushijima T . Epigenetic field for cancerization. J Biochem Mol Biol 2007; 40: 142–150.

  13. 13

    Worthley DL, Whitehall VL, Buttenshaw RL, Irahara N, Greco SA, Ramsnes I et al. DNA methylation within the normal colorectal mucosa is associated with pathway-specific predisposition to cancer. Oncogene 2010; 29: 1653–1662.

  14. 14

    Svrcek M, Buhard O, Colas C, Coulet F, Dumont S, Massaoudi I et al. Methylation tolerance due to an O6-methylguanine DNA methyltransferase (MGMT) field defect in the colonic mucosa: an initiating step in the development of mismatch repair-deficient colorectal cancers. Gut 2010; 59: 1516–1526.

  15. 15

    Gonzalo V, Lozano JJ, Munoz J, Balaguer F, Pellise M, Rodriguez de Miguel C et al. Aberrant gene promoter methylation associated with sporadic multiple colorectal cancer. PLoS One 2010; 5: e8777.

  16. 16

    Feinberg AP, Vogelstein B . Hypomethylation distinguishes genes of some human cancers from their normal counterparts. Nature 1983; 301: 89–92.

  17. 17

    Feinberg AP, Tycko B . The history of cancer epigenetics. Nat Rev Cancer 2004; 4: 143–153.

  18. 18

    Eden A, Gaudet F, Waghmare A, Jaenisch R . Chromosomal instability and tumors promoted by DNA hypomethylation. Science 2003; 300: 455.

  19. 19

    Gaudet F, Hodgson JG, Eden A, Jackson-Grusby L, Dausman J, Gray JW et al. Induction of tumors in mice by genomic hypomethylation. Science 2003; 300: 489–492.

  20. 20

    Kawakami K, Matsunoki A, Kaneko M, Saito K, Watanabe G, Minamoto T . Long interspersed nuclear element-1 hypomethylation is a potential biomarker for the prediction of response to oral fluoropyrimidines in microsatellite stable and CpG island methylator phenotype-negative colorectal cancer. Cancer Sci 2011; 102: 166–174.

  21. 21

    Sunami E, de Maat M, Vu A, Turner RR, Hoon DS . LINE-1 hypomethylation during primary colon cancer progression. PLoS One 2011; 6: e18884.

  22. 22

    Ibrahim AE, Arends MJ, Silva AL, Wyllie AH, Greger L, Ito Y et al. Sequential DNA methylation changes are associated with DNMT3B overexpression in colorectal neoplastic progression. Gut 2011; 60: 499–508.

  23. 23

    Ogino S, Nosho K, Kirkner GJ, Kawasaki T, Chan AT, Schernhammer ES et al. A cohort study of tumoral LINE-1 hypomethylation and prognosis in colon cancer. J Natl Cancer Inst 2008; 100: 1734–1738.

  24. 24

    Yamashita K, Dai T, Dai Y, Yamamoto F, Perucho M . Genetics supersedes epigenetics in colon cancer phenotype. Cancer Cell 2003; 4: 121–131.

  25. 25

    Suzuki K, Suzuki I, Leodolter A, Alonso S, Horiuchi S, Yamashita K et al. Global DNA demethylation in gastrointestinal cancer is age dependent and precedes genomic damage. Cancer Cell 2006; 9: 199–207.

  26. 26

    Suter CM, Martin DI, Ward RL . Hypomethylation of L1 retrotransposons in colorectal cancer and adjacent normal tissue. Int J Colorectal Dis 2004; 19: 95–101.

  27. 27

    Horii J, Hiraoka S, Kato J, Harada K, Kuwaki K, Fujita H et al. Age-related methylation in normal colon mucosa differs between the proximal and distal colon in patients who underwent colonoscopy. Clin Biochem 2008; 41: 1440–1448.

  28. 28

    Nosho K, Kure S, Irahara N, Shima K, Baba Y, Spiegelman D et al. A prospective cohort study shows unique epigenetic, genetic, and prognostic features of synchronous colorectal cancers. Gastroenterology 2009; 137: 1609–1620, e1–3.

  29. 29

    Suzuki K, Ohnami S, Tanabe C, Sasaki H, Yasuda J, Katai H et al. The genomic damage estimated by arbitrarily primed PCR DNA fingerprinting is useful for the prognosis of gastric cancer. Gastroenterology 2003; 125: 1330–1340.

  30. 30

    Weisenberger DJ, Campan M, Long TI, Kim M, Woods C, Fiala E et al. Analysis of repetitive element DNA methylation by MethyLight. Nucleic Acids Res 2005; 33: 6823–6836.

  31. 31

    Ionov Y, Peinado MA, Malkhosyan S, Shibata D, Perucho M . Ubiquitous somatic mutations in simple repeated sequences reveal a new mechanism for colonic carcinogenesis. Nature 1993; 363: 558–561.

  32. 32

    Vilar E, Gruber SB . Microsatellite instability in colorectal cancer-the stable evidence. Nat Rev Clin Oncol 2010; 7: 153–162.

  33. 33

    Estecio MR, Gharibyan V, Shen L, Ibrahim AE, Doshi K, He R et al. LINE-1 hypomethylation in cancer is highly variable and inversely correlated with microsatellite instability. PLoS One 2007; 2: e399.

  34. 34

    Brenner H, Hoffmeister M, Stegmaier C, Brenner G, Altenhofen L, Haug U . Risk of progression of advanced adenomas to colorectal cancer by age and sex: estimates based on 840 149 screening colonoscopies. Gut 2007; 56: 1585–1589.

  35. 35

    Ehrlich M . DNA methylation in cancer: too much, but also too little. Oncogene 2002; 21: 5400–5413.

  36. 36

    Qu GZ, Grundy PE, Narayan A, Ehrlich M . Frequent hypomethylation in Wilms tumors of pericentromeric DNA in chromosomes 1 and 16. Cancer Genet Cytogenet 1999; 109: 34–39.

  37. 37

    Xu GL, Bestor TH, Bourc’his D, Hsieh CL, Tommerup N, Bugge M et al. Chromosome instability and immunodeficiency syndrome caused by mutations in a DNA methyltransferase gene. Nature 1999; 402: 187–191.

  38. 38

    Zeimet AG, Fiegl H, Goebel G, Kopp F, Allasia C, Reimer D et al. DNA ploidy, nuclear size, proliferation index and DNA-hypomethylation in ovarian cancer. Gynecol Oncol 2011; 121: 24–31.

  39. 39

    Umar A, Risinger JI, Hawk ET, Barrett JC . Testing guidelines for hereditary non-polyposis colorectal cancer. Nat Rev Cancer 2004; 4: 153–158.

  40. 40

    Kim H, Jen J, Vogelstein B, Hamilton SR . Clinical and pathological characteristics of sporadic colorectal carcinomas with DNA replication errors in microsatellite sequences. Am J Pathol 1994; 145: 148–156.

  41. 41

    Thibodeau SN, Bren G, Schaid D . Microsatellite instability in cancer of the proximal colon. Science 1993; 260: 816–819.

  42. 42

    Perucho M . Cancer of the microsatellite mutator phenotype. Biol Chem 1996; 377: 675–684.

  43. 43

    Tsao JL, Yatabe Y, Salovaara R, Jarvinen HJ, Mecklin JP, Aaltonen LA et al. Genetic reconstruction of individual colorectal tumor histories. Proc Natl Acad Sci USA 2000; 97: 1236–1241.

  44. 44

    R Development Core Team. R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing: Vienna, Austria, 2009.

Download references


This work was supported in part by grants from the Ministry of Education, Culture, Science, Sports, and Technology of Japan, NIH grant R37 CA63585 and from the Spanish Ministry of Science and Innovation (FIS PI09/02444).

Author information

Correspondence to K Suzuki or M Perucho.

Ethics declarations

Competing interests

The authors declare no conflict of interest.

Additional information

Supplementary Information accompanies the paper on the Oncogene website

Supplementary information

Rights and permissions

Reprints and Permissions

About this article

Cite this article

Kamiyama, H., Suzuki, K., Maeda, T. et al. DNA demethylation in normal colon tissue predicts predisposition to multiple cancers. Oncogene 31, 5029–5037 (2012). https://doi.org/10.1038/onc.2011.652

Download citation


  • DNA demethylation
  • colon cancer
  • synchronous
  • metachronous
  • epigenetics

Further reading