Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2026 Mar 12.
Published in final edited form as: Cancer Epidemiol Biomarkers Prev. 2026 Mar 2;35(3):420–428. doi: 10.1158/1055-9965.EPI-25-1006

Germline predisposition to oncogenic alkylating damage in colorectal cancer

Carino Gurjao 1,2,3,4,#,*, Jules Cazaubiel 1,2,#, Chichun Tan 5, Brendan Reardon 1,2, Matan Hofree 6, Tomotaka Ugai 7,8, Jeffrey A Meyerhardt 1, Jonathan A Nowak 7, Edward Giovannucci 8,9, Jeffrey P Townsend 5,10,11,12, Shuji Ogino 2,7,8,13,#, Marios Giannakiss 1,2,#,*
PMCID: PMC12977196  NIHMSID: NIHMS2132990  PMID: 41416871

Abstract

Background:

Red meat consumption is a risk factor for colorectal cancer (CRC) and has been linked to tumor alkylating DNA damage. rs16906252-T is a cis expression quantitative trait locus (eQTL) variant associated with silencing of MGMT, a central alkylating damage repair gene. We hypothesize that rs16906252-T carriers are predisposed to alkylating damage mutations.

Methods:

We conducted mutational signature deconvolution of CRC whole-exome sequencing data from The Cancer Genome Atlas (TCGA, n = 540), the Nurses’ Health Studies/ Health Professional Follow-up Study (NHS/HPFS, n = 900) as well as non-western samples from the Pan-Cancer Analysis of Whole Genomes (COCA-CN, n = 295); and examined the relationship of rs16906252-T with putative alkylation-dependent tumor mutations. Leveraging lifestyle data from NHS/HPFS, we also investigated the interaction between red meat consumption and rs16906252-T.

Results:

Among CRC patients, rs16906252-T carriers exhibited higher tumor alkylating damage compared to non-carriers. In the general population, rs16906252-T is largely absent in individuals with East Asian ancestries, and we consistently find a negligible contribution of alkylating damage in CRC patients with East Asian ancestries. We show that the alkylating mutational signature’s carcinogenicity is mainly mediated by KRAS G12D and G13D mutations. We also observe a synergistic effect of rs16906252-T with high pre-diagnosis red meat intake for tumor alkylating damage.

Conclusions:

MGMT rs16906252-T carriers are predisposed to CRC oncogenic alkylating damage which is potentiated by red meat intake.

Impact:

Our results support a causal relationship between red meat and CRC and may inform tailored dietary and screening guidelines for CRC prevention.

Introduction

Gene variants can have a wide range of effects on cancer initiation and development. High-penetrance variants can cause familial cancers such as those associated with Lynch syndrome or MUTYH-associated polyposis1, but are rare and affect initiation of only a small percentage of all colorectal cancers (CRC). On the other hand, low-penetrance variants can be maintained at higher frequency and affect a larger segment of the population, contributing to cancer risk by synergistic effects with carcinogenic exposures2,3. These gene–environment (G×E) interactions that influence cancer risk have frequently been thought to function by facilitating the acquisition of somatic mutations required for cancer initiation and progression. Such G×E-associated variants can be detected in Genome-Wide Association Studies (GWAS) but have not previously been associated with carcinogenic exposures in CRC and the consequent pathogenic mutations. Recent advances in mutation analysis have made it possible to deconvolute signatures of mutational processes that tumor cells experienced prior to sequencing. These mutational signatures can provide insights into the carcinogenic activity of mutagenic exposures and have made it possible to measure the amount of DNA damage stemming from specific exposures. Linking such mutational signatures to germline variants provides compelling evidence regarding the etiology of exposure and the nature of variation in susceptibility4.

We previously discovered a CRC alkylating mutational signature that was associated with high pre-diagnosis red-meat intake5. O-6-methylguanine-DNA methyltransferase (MGMT) is a major DNA-damage repair protein that removes alkylating damage, notably methylated guanine residues6. The rs16906252-T (NM_002412.2:c.-56C>T) MGMT promoter variant is a cis expression quantitative trait locus (eQTL) associated with MGMT promoter CpG-island hypermethylation and gene silencing7,8. This hypermethylation and gene silencing is observed to occur across diverse normal tissues, including blood samples and healthy colonic tissues from the Genotype-Tissue Expression9 database (GTEx, Figure S1A). This strong association with gene silencing was also previously observed across cancers—including CRC6,10. The rs16906252-T MGMT promoter variant has also been identified as a risk factor for CRC11, supported by a large Genome-Wide Association Study of DNA repair genes7 (n = 11,114 CRCs and 14,659 controls; Figure S1B). Given the strong association of rs16906252-T with silencing of the alkylating damage-repair gene MGMT, we sought to explore the relationship between this variant and the CRC alkylating mutational signature.

Materials and Methods

Ethics Statement

Institutional review board approval and written informed consent were obtained for all patients in the original published studies. All of the tumor data is publicly available via dbGaP (https://dbgap.ncbi.nlm.nih.gov/aa/wga.cgi, RRID:SCR_002709) and the ICGC Data Access Compliance Office (DACO; http://icgc.org/daco).

Study populations

We utilized data from three prospective cohort studies in the U.S., the Nurses’ Health Study I (NHS1, including 121,701 women ages 30 to 55 years at enrollment who had been followed since 1976), the Nurses’ Health Study II (NHS2, including 116,429 women ages 25 to 42 years followed since 1989), and the Health Professionals Follow-up Study (HPFS, including 51,529 men ages 40 to 75 years followed since 1986)12. The study participants were sent questionnaires biennially to update information on lifestyle factors and newly diagnosed diseases including colorectal carcinoma. The follow-up rate was more than 90% for each follow-up questionnaire cycle in the three cohort studies. Archival formalin-fixed paraffin-embedded (FFPE) tissue blocks of tumor and normal colon were collected in a subset of CRCs, using the prospective cohort incident-tumor biobank method13. Whole exome sequencing (WES) on tumor-normal paired DNA from FFPE tissues was performed as described14.

We also utilized data from the Colon Adenocarcinoma (COAD-US, 191 women ages 34 to 90 years and 208 men ages 34 to 90 years) and Rectum Adenocarcinoma (READ-US, 65 women ages 31 to 90 years, 74 men ages 33 to 86 years, and 2 people of unknown sex and age) projects of The Cancer Genome Atlas (TCGA, RRID:SCR_003193), along with data from the Colorectal Adenocarcinoma in China cohort (COCA-CN, 154 women ages 21 to 88 years and 257 men ages 20 to 89 years). Clinical and somatic mutation data for 835 CRCs were downloaded from the Data Coordination Center (DCC) data portal at https://dcc.icgc.org/releases/current/Projects/COCA-CN; https://dcc.icgc.org/releases/current/Projects/COAD-US and https://dcc.icgc.org/releases/current/Projects/READ-US (as of January 2022). For consistency, only WES samples (n = 295 samples in COCA-CN and n = 540 in COAD-US and READ-US) were analyzed.

Dietary data

As previously described15, dietary intake assessment in NHS/HPFS was carried out with food frequency questionnaires (FFQ) – added to the previously mentioned biennial update questionnaires – that were initially collected in 1980 for NHS and in 1986 for HPFS. The NHS relied on a 61-item semiquantitative FFQ used at baseline16 and later expanded to approximately 130 food and beverage items in 1984, 1986, and every 4 years thereafter. The HPFS cohorts’ baseline dietary intake ascertainment relied on a 131-item FFQ that was also used for updates generally every 4 years subsequently17. The validity of this method in assessing dietary intake was robustly established by using diet and plasma nutrient records18,19,20. In this study, the dietary data were based on the most recent pre-diagnosis reported intake for each patient.

Specifically, unprocessed red meat consumption was evaluated based on the intake of “beef or lamb as main dish,” “pork as main dish,” “hamburger,” and “beef, pork, or lamb as a sandwich or mixed dish.”. Processed meat diets included “bacon”; “beef or pork hot dogs”; “salami, bologna, or other processed meat sandwiches”; and “other processed red meats such as sausage, kielbasa, etc.”. Participants reported their usual fruit and vegetable intake of standard portion size (e.g., 0.5 cups of strawberries, 1 banana, and 0.5 cups of cooked spinach) of fruit and vegetables for the preceding year of each FFQ. Frequencies and portions of fruit and vegetable items were converted to the average daily intake for each participant as previously described21. Consumption of red meat, poultry, fish, fruits and vegetables was evaluated in grams per day. Smoking status (never smoking, past smoking, current-smoking) and alcohol consumption (women: 0–<0.15, 0.15–<2.0, 2.0–<7.5, and ≥7.5 g/day; men: 0 to <1, 1–<6, 6–<15, and ≥15 g/day) were both stratified into three categories as previously described5.

Germline data and rs16906252-T detection in the general population and patients with CRC

Germline data for the general population were retrieved (as of January 2022) from the 1000 Genomes Project Phase 3 (RRID:SCR_006828): http://grch37.ensembl.org/Homo_sapiens/Variation/Population?db=core;r=10:131265045-131266045;v=rs16906252;vdb=variation;vf=48812926; and the Genome Aggregation Database (gnomAD, RRID:SCR_014964) v.2.1.1: https://gnomad.broadinstitute.org/variant/10-131265545-C-T?dataset=gnomad_r2_1.

The rs16906252-T variant is in a promoter region. Promoter regions can be subject to low coverage in WES data. We thus discarded tumor/normal pairs with less than four reads covering the rs16906252-T locus in both samples, which roughly corresponds to a 95% statistical power to detect rs16906252-T if present (probability to detect for a heterozygous rs16906252-T patient is (1 – 0.54). This filter resulted in n = 826 passing samples for NHS/HPFS (8% discarded samples) and n = 378 passing samples in TCGA (30% discarded samples). The higher number of discarded pairs in TCGA was mainly attributed to the use of older WES capture kits, such as SureSelect Human All Exon 38 Mb v2.

For the remaining samples, we used DeepVariant (version 1.4.0)22 to detect patients harboring rs16906252-T. For the tumor-normal pairs with no coverage of the rs16906252-T locus in the normal sample, it was not immediately possible to state if the variant was germline or somatic. However, in all instances of both tumor and normal being covered, the variant, if present, was harbored by both. Thus, to infer rs16906252-T status in the samples with no coverage in the paired normal tissue, we used the rs16906252-T variant status in the matched tumor specimen. We also rescued rs16906252-T carriers by manually reviewing this locus in the Integrative Genomics Viewer (RRID:SCR_011793)23. The passing samples (n=1204) and their SNV status from each method are listed in Table S1.

Quantitative DNA methylation and microsatellite status analysis

MGMT promoter methylation analysis in the NHS/HPFS cohorts was carried out using bisulfite conversion and real-time PCR as previously described24. Microsatellite Instability (MSI) status was evaluated using 10 microsatellite markers (D2S123, D5S346, D17S250, BAT25, BAT26, BAT40, D18S55, D18S56, D18S67, and D18S487) as formerly detailed12.

Average signature weights and cancer effect weights

The average signature weight is the average contribution of each mutational process to the total mutational burden in each tumor genome. Average cancer effect weights quantify the average contribution of each mutational process to the total cancer effect of variants classified as drivers in each tumor25. Cancer effects were quantified by estimation of underlying gene-based and site-based mutation rates and analysis compared to variant prevalence to quantify post-mutation cell proliferation and survival, as previously described25. To attribute cancer effects to signatures, effects of individual variants attributed to a signature were normalized to the total cancer effect of variants in each tumor. We then summed up the normalized cancer effects from all tumors and divided the sum by the total number of tumors.

CRC ancestry determination

WES-derived ancestry calls from TCGA colorectal samples were downloaded from the National Cancer Institute Genomic Data Commons https://gdc.cancer.gov/about-data/publications/CCG-AIM-2020. The ancestry of each TCGA patient was determined using EthSeq26, which leverages genotype data from individuals of known ethnicity to infer the ancestry admixture of inputted individual genotypes. We applied this approach to NHS/HPFS samples to annotate their ethnicity (n = 900) based on DeepVariant germline variant calls from the normal samples. EthSeq-inferred results were highly concordant with self-determined ethnicities in NHS/HPFS: 4 out of 4 patients self-determined as Asian American were detected as such by EthSeq. In addition, 11 out of 13 patients who self-identified as African American were detected as such by EthSeq.

GTEX data analysis

The data used for the analyses described in this manuscript were obtained at https://www.gtexportal.org/home/snp/rs16906252 from the GTEx Portal (RRID:SCR_001618) on 10/09/2022. In GTEx, the P values (Figure S1A) were generated by testing the alternative hypothesis that the slope of a linear regression model between genotype and expression deviates from zero. The normalized effect size (NES) of the eQTLs is defined as the slope of the linear regression and is computed as the effect of rs16906252-T relative to the reference allele.

Mutational signature analysis

Mutational signature analysis for NHS/HPFS and TCGA was performed as previously published5,27. For COCA-CN, mutations were classified by their 96 trinucleotide contexts for each sample. We then used the NMF R package (RRID:SCR_023124) and SigProfilerExtractor (RRID:SCR_023121) for the de novo signature extraction. While the NMF package performs a standard Kullback-Leibler-based non-negative matrix factorization directly on the mutational matrix5, SigProfilerExtractor resamples and normalizes it separately for each replicate28. Using their respective quality measures, rank 4 was selected as the best rank for both methods. For each sample, the signature-specific mutational counts were divided by the total mutational count, resulting in normalized signature contribution values.

Of note, all the signatures mentioned in this manuscript were also found in WES. For the refitting, MutationalPatterns (RRID:SCR_024247) and SigProfilerAssignment (RRID:SCR_026899) were used. Both methods solve this non-negative least-squares (NNLS) optimization problem using the pracma package (RRID:SCR_026021) and a custom implementation of the forward stagewise algorithm respectively29,30. Additionally, the decompose_fit function of SigProfilerAssignment was used to perform the decomposition of extracted signatures into their exome renormalized COSMIC components.

When compared to reference COSMIC v3 Single Base Substitution (SBS) signatures (RRID:SCR_002260)31, the four COCA-CN de novo signatures extracted using the NMF package displayed the highest similarity with SBS5 [cosine similarity (cossim) = 0.90], a featureless signature with unknown aetiology; SBS1 (cossim = 0.94), the age signature; SBS10a (cossim = 0.88) a POLE exonuclease deficiency signature; and SBS44 (cossim = 0.93) a defective DNA mismatch repair signature. Similarly, the 4 de novo signatures extracted using SigProfilerExtractor displayed the highest similarity with SBS5 (cossim = 0.888), SBS1 (cossim = 0.95), SBS10a (cossim = 0.873), and SBS44 (cossim = = 0.866).

Genome-Wide Association Study (GWAS) data

The rs16906252 MGMT variant was found to associate with cancer risk in a GWAS study of DNA repair genes7 comprising 11,114 CRCs and 14,659 controls from the Colon Cancer Family Registry (CCFR, RRID:SCR_013162) and the Genetics and Epidemiology of Colorectal Cancer Consortium (GECCO). Briefly, the authors leveraged 15,419 single nucleotide variants (SNVs) from 185 DNA repair genes and evaluated the association with CRC risk with a Binomial Sequential Goodness of Fit (BSGoF) procedure to adjust P values. Here, we replotted the SNVs from DNA repair genes as a Manhattan plot (Figure S1B) where we show the adjusted P values (y-axis) along the genome (position index, x-axis).

Fine-mapping of MGMT (top right inset, Figure S1B) was performed with LDassoc32. Both P values and rs16906252-T linkage disequilibrium patterns are shown for MGMT SNVs across all populations in the 1000 Genome Project.

Statistical analysis, including regression models and interaction analyses

To investigate the interaction between rs16906252-T MGMT SNV and pre-diagnosis total (processed and unprocessed) red meat intake, we used a Tobit regression model as implemented in the R Package ‘AER’ (version 1.2–10, RRID:SCR_027778)33. In addition, we used a standard generalized linear model from R package ‘stats’ (version 4.2.1, RRID:SCR_025968). In the Tobit model, the observed range of the dependent variable is clustered at a lower boundary. We used a left boundary of 0.01 and a right boundary of 0.99.

For the interaction analysis, we assume that rs16906252-T amplify the amount of DNA alkylating damage but does not directly cause it. Interaction contrasts were computed with the emmeans R package (RRID:SCR_018734).

We used R version 4.1.0 (RRID:SCR_001905) to perform statistical analyses. Significance for two-group comparisons was evaluated by one-sided Mann–Whitney U tests unless otherwise indicated. P < 0.05 was considered statistically significant. Multiple hypothesis correction was not performed unless specified.

Data Availability Statement

Publicly available data used in this study was accessed from the Data Coordination Center data portal at COCA-CN, COAD-US, and READ-US; from the 1000 Genomes Project Phase 3 at rs16906252; from the Genome Aggregation Database (gnomAD) v.2.1.1 at 10–131265545-C-T; from the National Cancer Institute Genomic Data Commons at CCG-AIM-2020; and from the GTEx Portal at rs16906252. NHS/HPFS WES data have been deposited in dbGAP at accession number phs000722. All analysis scripts are available upon request.

Results

Prevalence of rs16906252-T in the general population and CRC

We first examined the prevalence of rs16906252 C>T (both CT and TT, hereafter simply referred to as rs16906252-T carriers) in the general population by leveraging data from the 1000 Genomes Project34 and the Genome Aggregation Database35 (gnomAD; Figure S2). We found that rs16906252-T was present in 12% of the study populations: 380 rs16906252-TT and 9174 rs16906252-CT versus 79,068 rs16906252-CC individuals in gnomAD. Strikingly, rs16906252-T was largely absent in individuals with African and East Asian ancestries and more abundant in individuals with European ancestry (Figure S3A and Figure S3B). When stratified by populations, rs16906252-T genotype proportions conformed to Hardy-Weinberg equilibrium (right-tail chi-square P > 0.05 after Bonferroni correction, Table S2).

We then detected (see Methods) and surveyed the distribution of rs16906252-T in two independent WES CRC datasets from US-wide cohorts, namely NHS/HPFS5 (n = 826) and TCGA36 (n = 378). We used DeepVariant, a deep convolutional neural-network approach, to call genetic variants. Out of the 1440 CRC TCGA and NHS/HPFS patients analyzed, 132 were CT carriers and 20 were TT. The small sample size of TT individuals precluded their separate analysis, and we thus analyzed both CT and TT together (both CT and TT are hereafter simply referred to as rs16906252-T carriers). We found that overall, 12.4 % of patients with CRC harbored rs16906252-T (12.3% [= 102/826] of NHS / HPFS and 13.0% [= 49/378] of TCGA). rs16906252-T is significantly associated with MGMT promoter CpG island methylation in CRC8, including in the tumor specimens used in this analysis (Fisher’s exact tests P = 4.9 × 10−8 for TCGA, Table S3, P = 2.0 × 10−13 for NHS/HPFS, Table S4 and P < 2.2 × 10−16 when aggregating TCGA and NHS/HPFS, Figure 1A and Table S5).

Figure 1: The rs16906252-T MGMT promoter variant is associated with alkylating damage in CRC.

Figure 1:

Open in a new tab

(A) Proportion of tumors with hypermethylated MGMT promoter stratified by rs16906252-T status. The Chi-square test P value is shown on top. (B) Proportion of mutations assigned to the alkylating signature in all CRCs and in non-MSI-high CRCs, segregated by rs16906252-T status in NHS/HPFS and TCGA. Mann-Whitney U test P values are shown. Box-plot outliers are not shown. CRC: Colorectal Cancer. Non-MSI-high: Non-microsatellite-high. NHS/HPFS: NHS, Nurses’ Health Studies I and II. HPFS, Health Professionals Follow-up Study. TCGA, The Cancer Genome Atlas.

Tumors harboring rs16906252-T are enriched with alkylating damage

Because MGMT is a crucial gene for alkylating DNA damage repair6, we next investigated whether patients harboring rs16906252-T developed tumors that were enriched with the alkylating mutational signature5. TCGA and NHS/HPFS mutational signature deconvolution was performed as previously published by using a standard Non-negative Matrix Factorization (NMF) approach5,27. After aggregating NHS/HPFS and TCGA, we found that CRC patients harboring rs16906252-T exhibit significantly higher levels of tumor alkylating damage among all CRCs and in non-microsatellite-high (non-MSI-high) CRCs (P = 4.9 × 10−4 and P = 7.1 × 10−4 respectively, Mann–Whitney U test, Figure 1B). In addition, we evaluated the effect size for the Mann–Whitney U tests by calculating the rank-biserial correlation rrb. We observed similar effect sizes for all CRCs and non-MSI-high CRCs only (rrb = 0.16 and 0.18 respectively). Tumors of NHS/HPFS CRC patients that were rs16906252-T carriers were enriched with alkylating damage (rrb = 0.21 and P = 3.2 × 10−4 among all CRCs; rrb = 0.21 and P = 9.2 × 10−4 in non-MSI-high CRCs, Mann–Whitney U test, Figure S4A). Tumors of TCGA CRC patients that were rs16906252-T carriers, as determined by DeepVariant, did not exhibit statistically significantly higher levels of alkylating damage (Figure S4B). However, after rescuing false-negative tumor/normal pairs with the Integrated Genomics Viewer (IGV, see Methods and Figure S4C, Figure S4D, and Figure S4E) by manual review, we observed increased alkylating damage among carriers in TCGA CRCs also (P = 0.075; rrb = 0.10 among all CRCs; and P = 0.043; rrb = 0.15 in non-MSI-high CRCs, Mann–Whitney U test, Figure S4D). Interestingly, we also observed that among tumors with MGMT promoter hypermethylation, there is a significantly higher amount of alkylating damage in tumors from rs16906252-T carriers (Figure S5). When we examined the association of the MGMT SNV with the other CRC mutational signatures [Age, mismatch repair deficiency (dMMR), POLE], we found no significant associations (Figure S6).

Alkylating damage confers significant cancer advantage

The contribution of the mutational signatures to the total mutational burden is not necessarily equivalent to their contribution to oncogenic mutations25, but critically depends on the degree to which the subsequent mutations provide a selective advantage by increasing cancer cell proliferation or survival. Thus, we performed a direct calculation of cancer effect size as previously established25 that accounts for underlying gene- and site-based mutation rates to quantify both the selective advantage of each variant arising from the CRC mutagenic processes, as well as attribute these oncogenic effects to each mutational signature. This calculation quantifies the extent to which a mutational process accounts for carcinogenesis in a particular tumor. We observed that the alkylating signature contributes an average of 7% of the total oncogenic effect among non-MSI-high patients (Figure 2A) and that KRAS c.35G>A (p.G12D), KRAS c.38G>A (p.G13D), and PIK3CA c.1633G>A (p.E545K) were the main oncogenic variants targeted by the CRC alkylating signature (Figure 2B), consistent with our prior findings5. Indeed, our analysis demonstrated that 42% of the total carcinogenic effect of alkylating damage was mediated by key oncogenic mutations in KRAS p.G12D and p.G13D. Thus, the increased alkylating damage burden in rs16906252-T carriers can confer a substantial selective advantage to colorectal cancer cell lineages.

Figure 2: Average signature contribution and cancer effect weight in NHS/HPFS CRCs.

Figure 2:

Open in a new tab

(A) For each of the seven mutational signatures found in NHS/HPFS, the average contribution to the mutational burden (on the left) and the average contribution to the total cancer effect (on the right) is shown. Paired Wilcoxon Rank test P values for the difference of signature weight between TMB and cancer effect are also plotted. (B) The contribution of each signature to the average cancer effect of CRC hotspots. Here, the cancer effect is quantified proportionate to the total cancer effect of all recurrent variants in each tumor. dMMRa and dMMRb are the two signatures for DNA mismatch-repair deficiency. POLEa and POLEb are the two signatures for polymerase-epsilon deficiency. SBS40 is a featureless, flat signature with unknown aetiology.

Alkylating damage and rs16906252-T across ancestries

Because of the wide range of rs16906252-T frequencies across ancestries (Figure 3A, Figure S3A, and Figure S3B), we investigated the contribution of CRC alkylating damage across ethnicities in NHS/HPFS, TCGA as well as COCA-CN, an independent cohort of Chinese patients with CRC (Figure 3B).

Figure 3: Tumor alkylating signature contribution across genetic ancestries.

Figure 3:

Open in a new tab

(A) Proportion of rs16906252-T carriers from the 1000 Genome project Phase 3. (B) Mutational signature deconvolution in US-wide cohorts (TCGA and NHS/HPFS, on the left) as well as one East Asian cohort (COCA-CN, on the right) (C) The proportion of mutations assigned to alkylating damage in CRC, stratified by genetic ancestries in TCGA and NHS/HPFS. (D) Heat map of the similarity scores between colorectal tumor (from TCGA, NHS/HPFS and COCA-CN) signatures, clustered on the y-axis, and reference COSMIC signatures, clustered on the x-axis. Clustering has been performed according to cosine similarity. The table on the right summarizes the presence / absence of each mutagenic process (y axis) in each dataset (x axis).

First, we investigated previously published WES-derived genetic ancestries in TCGA37 (n = 13 patients with East Asian ancestry and n = 42 patients with African ancestry). We similarly inferred genetic ancestries in NHS/HPFS using EthSeq26 (n = 5 patients with East Asian ancestry and n = 13 patients with African ancestry, see Methods). Overall, we observed a significant depletion in the alkylating signature in patients with East Asian (rrb = 0.26, P = 3.0 × 10−3, Mann–Whitney U test) and African ancestries (rrb = 0.38, P = 7.5 × 10−5, Mann–Whitney U test) compared to patients with European ancestry (Figure 3C). Comparable results were observed in TCGA (Figure S7A), but not in NHS/HPFS (Figure S7B) likely because of the relative lack of patients with non-European ancestry in the latter (14% and 2% patients with non-European ancestry in TCGA and NHS/HPFS respectively).

Next, we performed a mutational-signature analysis on COCA-CN (n = 295 samples, see Methods and Figure 3D, Figure S8A, Figure S8B, Figure S9, Figure S10). In contrast to TCGA and NHS/HPFS CRCs, the alkylating signature was absent in COCA-CN. We assessed the robustness of this result with three different methods. First, we extracted seven de novo signatures in COCA-CN and showed that none matched the alkylating signature (see Methods and Figure S11). Next, to demonstrate that the difference in sample size between COCA-CN (n = 295) and NHS/HPFS (n = 900) cannot explain the lack of alkylating signature detection in COCA-CN, we (i) randomly sampled 295 CRCs of the 900 from NHS/HPFS; (ii) extracted seven signatures from the 295 CRCs; and (iii) repeated steps (i) and (ii) a hundred times. In 72% of the simulations, either SBS11 or SBS305 was detected (Figure S12); we previously found these two signatures to be interchangeable in smaller sample-size cohorts because of their similarity (cossim = 0.81) 5. Lastly, we fitted COCA-CN mutational profiles to SBS signatures found in TCGA-CRCs (SBS1, SBS10a, SBS10b, SBS30, SBS15, SBS26 and SBS40) using the MutationalPatterns R package30. The age (SBS1) and noise signatures (SBS40) showed no significant difference between TCGA and COCA-CN. The dMMR (SBS15 and SBS26) and POLE (SBS10a and SBS10b) signature contributions were significantly lower in COCA-CN compared to TCGA, due to the lower proportion of MSI / POLE patients in COCA-CN (0.7% POLE deficient and 3.8% MSI samples as computed by MSIseq38 compared to TCGA (1.8% POLE deficient and 14.0% MSI samples). The alkylating signature (SBS30) was significantly lower in COCA-CN (P < 2.2 × 10−16) compared to TCGA CRCs (Figure S13). We further demonstrated the lack of detectable signature in COCA-CN by decomposing COCA-CN signatures into combination sets of COSMIC reference signatures and showing the lack of an alkylating signature contribution (Figure S14A). In addition, no alkylating signature was found in COCA-CN after deconvoluting signatures for other ranks (rank 2 to 15 shown on Figure S14B). Altogether, these results support the lack of a detectable alkylating mutational signature in COCA-CN, consistent with the lower frequency of rs16906252-T in individuals of East Asian ancestry.

rs16906252-T and red meat intake potentiate alkylating damage

The addition of nitrates (e.g. during processing) and catalysis by heme iron in red meat can lead to the formation of N-nitroso compounds (NOCs), which are alkylating agents15,39,40. We previously demonstrated an association of pre-diagnosis processed and unprocessed red-meat intake with the CRC alkylating mutational signature in NHS/HPFS5. This previous analysis relied on the comparison of extreme eaters (top 10%) to the rest of the cohort. Here, we leveraged a more general approach of regression modeling that did not rely on discretization of a continuous variable into binary data. To find lifestyle/ dietary behaviors associated with a change in alkylating damage, we also further expanded our analysis to include prospectively collected repeated measurements of smoking and alcohol, in addition to dietary variables of fruit, vegetable, chicken, and fish consumption in the NHS and HPFS cohorts (see Methods). Because the distribution of alkylating damage appears to be a truncated normal distribution with zero-inflation (Figure S15), we used a Tobit regression model41 which showed that among all variables studied, only red meat is associated with CRC alkylating damage (P = 0.019, Figure 4A). We observed similar results with a standard Generalized Linear Model (GLM; P = 0.014, Figure S16A).

Figure 4: Effects of lifestyle and rs16906252-T variant for CRC alkylating damage in NHS/HPFS.

Figure 4:

Open in a new tab

(A) Tobit model of dietary variable and alkylating damage in CRC after z-scoring (centering and scaling of all the variables) (B) Interactive effect and 95% confidence intervals of rs16906252-T and pre-diagnosis red meat consumption on CRC alkylating signature contribution using a Tobit model (solid lines). NHS/HPFS: NHS, Nurses’ Health Studies I and II. HPFS, Health Professionals Follow-up Study.

Next, we tested for an interaction between rs16906252-T and pre-diagnosis red-meat intake in a prediction model of alkylating damage. We observed a significant interaction of red meat intake with the tumor alkylating mutational signature for rs16906252-T carriers (interaction-term Wald test P = 0.046 in the Tobit Model, Figure 4B). Similar results were observed in the GLM model (interaction-term Wald test P = 0.053, Figure S16B).

Discussion

Our study demonstrates that the rs16906252-T MGMT promoter germline variant associates with a CRC alkylating signature that has a substantial carcinogenic effect, especially by originating cancer driver mutations KRAS p.G12D and p.G13D5. This result is consistent with prior studies showing a link between MGMT loss and KRAS c.35G>A10,42. In addition, we observe that among tumors with MGMT promoter hypermethylation, there is a significantly higher amount of alkylating damage in tumors from rs16906252-T carriers. We thus postulate that individuals with the rs16906252-T germline variant may have accumulated augmented alkylating damage over the course of their lifetime. While this increase in mutational burden is relatively small, those added mutations are more likely to affect highly oncogenic loci such as KRAS G12D than any other mutational process present in CRC.

Furthermore, our work supports the importance of studying the interactions between MGMT and dietary alkylating agents, not solely for their effects on carcinogenesis but also for implications in treatment. Temozolomide, a therapeutic alkylating agent, has shown clinical activity in CRCs with MGMT promoter hypermethylation in MMR proficient samples and, as demonstrated by Crisafulli et al.43, can also induce MMR deficiency and increase mutational burden, thereby opening potential avenues for immunotherapy in tumors that typically are resistant.

In addition, we demonstrated a negligible contribution of alkylating damage to tumor mutational burden and to mutagenic oncogenesis in an East Asian CRC cohort, which can be partly attributed to the paucity of rs16906252-T in this population. This is consistent with a previously described weaker association between red meat intake and CRC in East Asian populations44, however, large East Asian prospective or retrospective studies combined with genomic datasets would enable confirmation of the contribution of diet, behavioral, or environmental differences to the lack of a detectable alkylating signature. In particular, red meat cooking methods have been shown to affect CRC risk through the formation of heterocyclic amines and polycyclic aromatic hydrocarbons45. Because cooking practices often reflect cultural and ancestral traditions, however, these may be important confounders. Thus, more research into differences in cooking methods between the studied populations is needed to better assess this contribution.

Alkylating damage arises from a wide array of modifiable exposures46,47 including red-meat intake, which in our study was the only variable significantly associated with alkylating damage among all the dietary variables studied. We observe a synergistic gene-by-environment interaction effect of the rs16906252-T germline variant and high pre-diagnosis red-meat intake, leading to increased tumor alkylating damage. Consistently, experimental studies have previously shown that MGMT-null mice develop tumors only in the presence of alkylating exposures48,49. Overall, our results support a causal relationship between red meat intake and CRC and encourage the implementation of precision prevention by enhanced screening and/or dietary modifications for individuals harboring the rs16906252-T variant.

Supplementary Material

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21

Acknowledgements

We thank G. Wang and S.McGrath for technical feedback, as well as S. Gazal and N. Oak for useful comments. The authors would like to acknowledge the contribution to this study from central cancer registries supported through the Centers for Disease Control and Prevention’s National Program of Cancer Registries (NPCR) and/or the National Cancer Institute’s Surveillance, Epidemiology, and End Results (SEER) Program. Central registries may also be supported by state agencies, universities, and cancer centers. Participating central cancer registries include the following: Alabama, Alaska, Arizona, Arkansas, California, Delaware, Colorado, Connecticut, Florida, Georgia, Hawaii, Idaho, Indiana, Iowa, Kentucky, Louisiana, Maine, Maryland, Massachusetts, Michigan, Mississippi, Montana, Nebraska, Nevada, New Hampshire, New Jersey, New Mexico, New York, North Carolina, North Dakota, Ohio, Oklahoma, Oregon, Pennsylvania, Puerto Rico, Rhode Island, Seattle SEER Registry, South Carolina, Tennessee, Texas, Utah, Virginia, West Virginia, Wyoming.

This work was supported by Cancer Research UK Grand Challenge Award (UK C10674/A27140 to M.G. and S.O.); by the U.S. National Institutes of Health (NIH) grants R01 CA151993 (to S.O.) and R35 CA197735 (to S.O.); by the American Cancer Society Clinical Research Professor Award (Grant # CRP-24–1185864-01-PROF to S.O.); and by the Project P and Crush Colon Cancer Funds. M.G. was supported by a High Pointe Investigatorship in Gastrointestinal Oncology. The Nurses’ Health Study was supported by the NIH grants UM1 CA186107 and P01 CA879690. The Nurses’ Health Study II was supported by the NIH grant U01 CA176726. The Health Professionals Follow-up Study was supported by NIH grants P01 CA055075, UM1 CA167552, and U01 CA167552.

Abbreviations:

COAD-US

Colon Adenocarcinoma

COCA-CN

Colorectal Adenocarcinoma in China

CRC

Colorectal Cancer

eQTL

cis expression Quantitative Trait Locus

FFPE

Formalin-Fixed Paraffin-Embedded

FFQ

Food Frequency Questionnaires

GxE

Gene-Environment

GLM

Generalized Linear Model

gnomAD

Genome Aggregation Database

GTEx

Genotype-Tissue Expression

GWAS

Genome-Wide Association Studies

HPFS

Health Professional Follow-up Study

MMR

Mismatch Repair

MSI

Microsatellite Instability

NHS

Nurses’ Health Study

NMF

Non-negative Matrix Factorization

READ-US

Rectal Adenocarcinoma

SBS

Single Base Substitution

SNV

Single Nucleotide Variant

TCGA

The Cancer Genome Atlas

WES

Whole Exome Sequencing

Footnotes

Declarations of interests: M.G. receives research funding from Janssen Pharmaceuticals and Sunbird bio. He has served in a consulting role for Nerviano Medical Sciences and received honoraria from PER and OncLive.

References

  • 1.Seppälä TT, Burkhart RA, Katona BW. Hereditary colorectal, gastric, and pancreatic cancer: comprehensive review. BJS Open. 2023;7(3):zrad023. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2.Szuman M, Kaczmarek-Ryś M, Hryhorowicz S, Kryszczyńska A, Grot N, Pławski A. Low-Penetrance Susceptibility Variants in Colorectal Cancer-Current Outlook in the Field. Int J Mol Sci. 2024;25(15):8338. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3.Gertig DM, Hunter DJ. Genes and environment in the etiology of colorectal cancer. Semin Cancer Biol. 1998;8(4):285–298. [DOI] [PubMed] [Google Scholar]
  • 4.Gabriel AAG, Atkins JR, Penha RCC, Smith-Byrne K, Gaborieau V, Voegele C, et al. Genetic Analysis of Lung Cancer and the Germline Impact on Somatic Mutation Burden. J Natl Cancer Inst. 2022;114(8):1159–1166. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Gurjao C, Zhong R, Haruki K, Li YY, Spurr LF, Lee-Six H, et al. Discovery and Features of an Alkylating Signature in Colorectal Cancer. Cancer Discov. 2021;11(10):2446–2455. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6.Esteller M, Hamilton SR, Burger PC, Baylin SB, Herman JG. Inactivation of the DNA repair gene O6-methylguanine-DNA methyltransferase by promoter hypermethylation is a common event in primary human neoplasia. Cancer Res. 1999;59(4):793–797. [PubMed] [Google Scholar]
  • 7.Pardini B, Corrado A, Paolicchi E, Cugliari G, Berndt SI, Bezieau S, et al. DNA repair and cancer in colon and rectum: Novel players in genetic susceptibility. Int J Cancer. 2020;146(2):363–372. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.Ogino S, Hazra A, Tranah GJ, Kirkner GJ, Kawasaki T, Nosho K, et al. MGMT germline polymorphism is associated with somatic MGMT promoter methylation and gene silencing in colorectal cancer. Carcinogenesis. 2007;28(9):1985–1990. [DOI] [PubMed] [Google Scholar]
  • 9.GTEx Consortium. The GTEx Consortium atlas of genetic regulatory effects across human tissues. Science. 2020;369(6509):1318–1330. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Esteller M, Toyota M, Sanchez-Cespedes M, Capella G, Peinado MA, Watkins DN, et al. Inactivation of the DNA repair gene O6-methylguanine-DNA methyltransferase by promoter hypermethylation is associated with G to A mutations in K-ras in colorectal tumorigenesis. Cancer Res. 2000;60(9):2368–2371. [PubMed] [Google Scholar]
  • 11.Kuroiwa-Trzmielina J, Wang F, Rapkins RW, Ward RL, Buchanan DD, Win AK, et al. SNP rs16906252C>T Is an Expression and Methylation Quantitative Trait Locus Associated with an Increased Risk of Developing MGMT-Methylated Colorectal Cancer. Clin Cancer Res. 2016;22(24):6266–6277. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Liao X, Lochhead P, Nishihara R, Morikawa T, Kuchiba A, Yamauchi M, et al. Aspirin use, tumor PIK3CA mutation, and colorectal-cancer survival. N Engl J Med. 2012;367(17):1596–1606. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Hamada T, Ugai T, Gurjao C, Ugai S, Zhang X, Haruki K, et al. Smoking habit and long-term colorectal cancer incidence by exome-wide mutational and neoantigen loads: evidence based on the prospective cohort incident-tumour biobank method. BMJ Oncol. 2025;4(1):e000787. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Giannakis M, Mu XJ, Shukla SA, Qian ZR, Cohen O, Nishihara R, et al. Genomic Correlates of Immune-Cell Infiltrates in Colorectal Carcinoma. Cell Rep. 2016;15(4):857–865. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Bernstein AM, Song M, Zhang X, Pan A, Wang M, Fuchs CS, et al. Processed and Unprocessed Red Meat and Risk of Colorectal Cancer: Analysis by Tumor Location and Modification by Time. PLoS One. 2015;10(8):e0135959. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Willett WC, Sampson L, Stampfer MJ, Rosner B, Bain C, Witschi J, et al. Reproducibility and validity of a semiquantitative food frequency questionnaire. Am J Epidemiol. 1985;122(1):51–65. [DOI] [PubMed] [Google Scholar]
  • 17.Rimm EB, Giovannucci EL, Stampfer MJ, Colditz GA, Litin LB, Willett WC. Reproducibility and validity of an expanded self-administered semiquantitative food frequency questionnaire among male health professionals. Am J Epidemiol. 1992;135(10):1114–1136. [DOI] [PubMed] [Google Scholar]
  • 18.Feskanich D, Rimm EB, Giovannucci EL, Colditz GA, Stampfer MJ, Litin LB, et al. Reproducibility and validity of food intake measurements from a semiquantitative food frequency questionnaire. J Am Diet Assoc. 1993;93(7):790–796. [DOI] [PubMed] [Google Scholar]
  • 19.Al-Shaar L, Yuan C, Rosner B, Dean SB, Ivey KL, Clowry CM, et al. Reproducibility and Validity of a Semiquantitative Food Frequency Questionnaire in Men Assessed by Multiple Methods. Am J Epidemiol. 2021;190(6):1122–1132. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Chiu YH, Sandoval-Insausti H, Ley SH, Bhupathiraju SN, Hauser R, Rimm EB, et al. Association between intake of fruits and vegetables by pesticide residue status and coronary heart disease risk. Environ Int. 2019;132:105113. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Wang DD, Li Y, Bhupathiraju SN, Rosner BA, Sun Q, Giovannucci EL, et al. Fruit and Vegetable Intake and Mortality: Results From 2 Prospective Cohort Studies of US Men and Women and a Meta-Analysis of 26 Cohort Studies. Circulation. 2021;143(17):1642–1654. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Poplin R, Chang PC, Alexander D, Schwartz S, Colthurst T, Ku A, et al. A universal SNP and small-indel variant caller using deep neural networks. Nat Biotechnol. 2018;36(10):983–987. [DOI] [PubMed] [Google Scholar]
  • 23.Thorvaldsdóttir H, Robinson JT, Mesirov JP. Integrative Genomics Viewer (IGV): high-performance genomics data visualization and exploration. Brief Bioinform. 2013;14(2):178–192. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Ogino S, Kawasaki T, Brahmandam M, Cantor M, Kirkner GJ, Spiegelman D, et al. Precision and performance characteristics of bisulfite conversion and real-time PCR (MethyLight) for quantitative DNA methylation analysis. J Mol Diagn. 2006;8(2):209–217. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Cannataro VL, Mandell JD, Townsend JP. Attribution of Cancer Origins to Endogenous, Exogenous, and Preventable Mutational Processes. Mol Biol Evol. 2022;39(5):msac084. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26.Romanel A, Zhang T, Elemento O, Demichelis F. EthSEQ: ethnicity annotation from whole exome sequencing data. Bioinformatics. 2017;33(15):2402–2404. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27.Alexandrov LB, Nik-Zainal S, Wedge DC, Aparicio SA, Behjati S, Biankin AV, et al. Signatures of mutational processes in human cancer. Nature. 2013;500(7463):415–421. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28.Islam SMA, Díaz-Gay M, Wu Y, Barnes M, Vangara R, Bergstrom EN, et al. Uncovering novel mutational signatures by de novo extraction with SigProfilerExtractor. Cell Genom. 2022;2(11):100179. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29.Díaz-Gay M, Vangara R, Barnes M, Wang X, Islam SMA, Vermes I, et al. Assigning mutational signatures to individual samples and individual somatic mutations with SigProfilerAssignment. Bioinformatics. 2023;39(12):btad756. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Blokzijl F, Janssen R, van Boxtel R, Cuppen E. MutationalPatterns: comprehensive genome-wide analysis of mutational processes. Genome Med. 2018;10(1):33. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.Alexandrov LB, Kim J, Haradhvala NJ, Huang MN, Tian Ng AW, Wu Y, et al. The repertoire of mutational signatures in human cancer. Nature. 2020;578(7793):94–101. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 32.Machiela MJ, Chanock SJ. LDassoc: an online tool for interactively exploring genome-wide association study results and prioritizing variants for functional investigation. Bioinformatics. 2018;34(5):887–889. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33.Kleiber C, Zeileis A. Applied Econometrics with R. New York: Springer-Verlag; 2008: 141–143. [Google Scholar]
  • 34.1000 Genomes Project Consortium, Auton A, Brooks LD, Durbin RM, Garrison EP, Kang HM, Korbel JO, et al. A global reference for human genetic variation. Nature. 2015;526(7571):68–74. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 35.Koch L Exploring human genomic diversity with gnomAD. Nat Rev Genet. 2020;21(8):448. [DOI] [PubMed] [Google Scholar]
  • 36.ICGC/TCGA Pan-Cancer Analysis of Whole Genomes Consortium. Pan-cancer analysis of whole genomes. Nature. 2020;578(7793):82–93. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37.Carrot-Zhang J, Chambwe N, Damrauer JS, Knijnenburg TA, Robertson AG, Yau C, et al. Comprehensive Analysis of Genetic Ancestry and Its Molecular Correlates in Cancer. Cancer Cell. 2020;37(5):639–654.e6. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 38.Ni Huang M, McPherson JR, Cutcutache I, Teh BT, Tan P, Rozen SG. MSISEQ: Software for assessing microsatellite instability from catalogs of somatic mutations. Sci Rep. 2015;5(1):13321. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 39.Fahrer J, Kaina B. Impact of DNA repair on the dose-response of colorectal cancer formation induced by dietary carcinogens. Food Chem Toxicol. 2017;106(Pt B):583–594. [DOI] [PubMed] [Google Scholar]
  • 40.Gilsing AM, Fransen F, de Kok TM, Goldbohm AR, Schouten LJ, de Bruïne AP, et al. Dietary heme iron and the risk of colorectal cancer with specific mutations in KRAS and APC. Carcinogenesis. 2013;34(12):2757–2766. [DOI] [PubMed] [Google Scholar]
  • 41.Long JS. Regression Models for Categorical and Limited Dependent Variables. London: SAGE Publications; 1997: 187–216. [Google Scholar]
  • 42.Shima K, Morikawa T, Baba Y, Nosho K, Suzuki M, Yamauchi M, et al. MGMT promoter methylation, loss of expression and prognosis in 855 colorectal cancers. Cancer Causes Control. 2011;22(2):301–309. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 43.Crisafulli G, Sartore-Bianchi A, Lazzari L, Pietrantonio F, Amatu A, Macagno M, et al. Temozolomide Treatment Alters Mismatch Repair and Boosts Mutational Burden in Tumor and Blood of Colorectal Cancer Patients. Cancer Discov. 2022;12(7):1656–1675. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 44.Cheung HS, Lu LY, So WL, Wong HW, Wong SH, Mak CM. Association of Red Meat Intake and Colorectal Cancer among East-Asians: A Systematic Review and Meta-analysis of Observational Studies Performed between 2011–2021. J Gastrointestin Liver Dis. 2023;32(3):377–383. [DOI] [PubMed] [Google Scholar]
  • 45.Sinha R, Peters U, Cross AJ, Kulldorff M, Weissfeld JL, Pinsky PF, et al. Meat, meat cooking methods and preservation, and risk for colorectal adenoma. Cancer Res. 2005;65(17):8034–8041. [DOI] [PubMed] [Google Scholar]
  • 46.Lijinsky W N-Nitroso compounds in the diet. Mutat Res. 1999;443(1–2):129–138. [DOI] [PubMed] [Google Scholar]
  • 47.Mirvish SS. Role of N-nitroso compounds (NOC) and N-nitrosation in etiology of gastric, esophageal, nasopharyngeal and bladder cancer and contribution to cancer of known exposures to NOC. Cancer Lett. 1995;93(1):17–48. [DOI] [PubMed] [Google Scholar]
  • 48.Meira LB, Calvo JA, Shah D, Klapacz J, Moroski-Erkul CA, Bronson RT, et al. Repair of endogenous DNA base lesions modulate lifespan in mice. DNA Repair (Amst). 2014;21:78–86. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 49.Shiraishi A, Sakumi K, Sekiguchi M. Increased susceptibility to chemotherapeutic alkylating agents of mice deficient in DNA repair methyltransferase. Carcinogenesis. 2000;21(10):1879–1883. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

1
NIHMS2132990-supplement-1.docx (159.3KB, docx)
2
NIHMS2132990-supplement-2.docx (149.6KB, docx)
3
NIHMS2132990-supplement-3.docx (144KB, docx)
4
NIHMS2132990-supplement-4.docx (173.7KB, docx)
5
NIHMS2132990-supplement-5.docx (53.7KB, docx)
6
NIHMS2132990-supplement-6.docx (230KB, docx)
7
NIHMS2132990-supplement-7.docx (97.2KB, docx)
8
NIHMS2132990-supplement-8.docx (458.5KB, docx)
9
NIHMS2132990-supplement-9.docx (137.2KB, docx)
10
NIHMS2132990-supplement-10.docx (124.6KB, docx)
11
NIHMS2132990-supplement-11.docx (138.6KB, docx)
12
NIHMS2132990-supplement-12.docx (65.2KB, docx)
13
NIHMS2132990-supplement-13.docx (115.1KB, docx)
14
NIHMS2132990-supplement-14.docx (301.6KB, docx)
15
NIHMS2132990-supplement-15.docx (21.6KB, docx)
16
NIHMS2132990-supplement-16.docx (198.5KB, docx)
17
NIHMS2132990-supplement-17.xlsx (38.5KB, xlsx)
18
NIHMS2132990-supplement-18.docx (15.5KB, docx)
19
NIHMS2132990-supplement-19.docx (14.3KB, docx)
20
NIHMS2132990-supplement-20.docx (14.3KB, docx)
21
NIHMS2132990-supplement-21.docx (14.3KB, docx)

Data Availability Statement

Publicly available data used in this study was accessed from the Data Coordination Center data portal at COCA-CN, COAD-US, and READ-US; from the 1000 Genomes Project Phase 3 at rs16906252; from the Genome Aggregation Database (gnomAD) v.2.1.1 at 10–131265545-C-T; from the National Cancer Institute Genomic Data Commons at CCG-AIM-2020; and from the GTEx Portal at rs16906252. NHS/HPFS WES data have been deposited in dbGAP at accession number phs000722. All analysis scripts are available upon request.

RESOURCES