|Home | About | Journals | Submit | Contact Us | Français|
JC virus has a transforming gene encoding JC virus T-antigen (JCVT). JCVT may inactivate wild-type p53, cause chromosomal instability (CIN), and stabilize β-catenin. A link between JCVT and CpG island methylator phenotype (CIMP) has been suggested. However, no large-scale study has examined the relations of JCVT with molecular alterations, clinical outcome, or prognosis in colon cancer. We detected JCVT expression (by immunohistochemistry) in 271 (35%) of 766 colorectal cancers. We quantified DNA methylation in eight CIMP-specific promoters (CACNA1G, CDKN2A, CRABP1, IGF2, MLH1, NEUROG1, RUNX3, and SOCS1) and eight other loci (CHFR, HIC1, IGFBP3, MGMT, MINT1, MINT31, p14, WRN) by MethyLight. We examined loss of heterozygosity in 2p, 5q, 17q, and 18q. JCVT was significantly associated with p53 expression (P < .0001), p21 loss (P < .0001), CIN (≥2 chromosomal segments with LOH; P < .0001), nuclear β-catenin (P = .006), LINE-1 hypomethylation (P = .002), and inversely with CIMP-high (P = .0005) and microsatellite instability (MSI) (P < .0001), but not with PIK3CA mutation. In multivariate logistic regression analysis, the associations of JCVT with p53 [adjusted odds ratio (OR), 8.45; P < .0001], CIN (adjusted OR, 2.53; P = .003), cyclin D1 (adjusted OR, 1.57; P = .02), LINE-1 hypomethylation (adjusted OR, 1.97 for a 30% decline as a unit; P = .03), BRAF mutation (adjusted OR, 2.20; P = .04), and family history of colorectal cancer (adjusted OR, 0.64; P = .04) remained statistically significant. However, JCVT was no longer significantly associated with CIMP, MSI, β-catenin, or cyclooxygenase-2 expression in multivariate analysis. JCVT was unrelated with patient survival. In conclusion, JCVT expression in colorectal cancer is independently associated with p53 expression and CIN, which may lead to uncontrolled cell proliferation.
The Polyomavirus family includes the simian virus 40 (SV40), JC virus, and BK virus. JC virus is a 5.12-kb, double-stranded, circular, negatively supercoiled DNA virus that commonly infects human cells [1,2]. It has a transforming gene encoding JC virus T-antigen (JCVT), which is believed to mediate the oncogenic potential of the virus [1–10]. Previous studies have suggested the link between JCVT and various human cancers [1,3–9,11,12], including colon cancers [5–7,9,12]. JC virus T-antigen has been reported to bind and inactivate wild type p53 [5,11] and cause chromosomal instability (CIN) [5–7], as well as the stabilization of β-catenin [8,9]. In addition, a previous study has suggested a link between JCVT and promoter methylation leading to the CpG island methylator phenotype (CIMP) in colorectal cancer .
CpG island methylator phenotype is characterized by a widespread, concordant CpG island methylation pattern [13–16]. CpG island methylator phenotype-high in colorectal cancer has been associated with older age, female sex, proximal tumor location, BRAF mutation, microsatellite instability (MSI), wild type TP53, inactive WNT/β-catenin, high-level long interspersed nucleotide element 1 (LINE-1) methylation, and stable chromosomes [17–30]. However, to our knowledge, no large-scale study has been conducted to examine the relationship of JCVT with genetic/epigenetic alterations or clinical outcome in colorectal cancer.
In this study, using a large number (n = 766) of stage I to IV colorectal cancers in two independent cohort studies, we examined tumoral JCVT expression in relation to clinical, pathologic, and molecular features in colorectal cancers. We have found that JCVT expression is significantly associated with p53 expression and CIN but not independently with CIMP or patient survival.
We used the databases of two large prospective cohort studies; the Nurses' Health Study (n = 121,700 women observed since 1976) [31,32], and the Health Professional Follow-up Study (n = 51,500 men observed since 1986) . Data on height and weight were obtained by biennial questionnaire. A subset of the cohort participants developed colorectal cancers during prospective follow-up. Previous studies on the Nurses' Health Study and Health Professionals Follow-up Study have described baseline characteristics of cohort participants and incident colorectal cancer cases and confirmed that our colorectal cancers were a good representative as a population-based sample [31,32]. Data on tumor location and stage were obtained through medical record review. We collected paraffin-embedded tissue blocks from hospitals where cohort participants with colorectal cancers had undergone resections of primary tumors. On the basis of availability of adequate tissue specimens and results, a total of 766 colorectal cancers were included. Written informed consent was obtained from all study subjects. Among our cohort studies, there was no significant difference in the demographic features between cases with tissue available and those without available tissue . This current analysis represents a new analysis of JCVT on the existing colorectal cancer database that have been previously characterized for CIMP, MSI, p53, KRAS, or BRAF , which is analogous to novel studies using the well-described cell lines or animal models. In any of our previous studies, we have not examined JCVT expression or the relations between JCVT and other molecular events. This study represents a unique novel study about the following: 1) a large sample size analyzed for JCVT; 2) the validated set of CIMP-specific methylation markers; and 3) a number of other molecular events analyzed, including eight CpG islands other than the CIMP-specific markers, MSI, CIN, KRAS, BRAF, PIK3CA, p53, LINE-1 methylation, cyclin D1, p21, cyclooxygenase 2 (COX-2), and β-catenin. Tissue collection and analyses were approved by the Harvard School of Public Health and Brigham and Women's Hospital Institutional Review Boards.
Hematoxylin and eosin-stained tissue sections were examined by a pathologist (S.O.) unaware of other data. The tumor grade was categorized as low (≥50% gland formation) or high (<50% gland formation). The presence and extent of extracellular mucin were categorized as 0% (no mucin) or =1% of the tumor volume. The presence and extent of signet ring cells were categorized as 0% (no signet ring cells) or ≥1% of the tumor volume.
Genomic DNA was extracted from dissected tumor tissue sections, and whole-genome amplification was performed by polymerase chain reaction (PCR) using random 15-mer primers . Polymerase chain reaction and Pyrosequencing targeted for KRAS (codons 12 and 13), BRAF (codon 600), and PIK3CA (exons 9 and 20) were performed as previously described . Microsatellite instability analysis was performed, using 10 microsatellite markers (D2S123, D5S346, D17S250, BAT25, BAT26, BAT40, D18S55, D18S56, D18S67, and D18S487) . Microsatellite instability-high was defined as the presence of instability in ≥30% of the markers. Microsatellite instability-low was defined as instability in <30% of the markers, and microsatellite stable (MSS) tumors were defined as tumors without an unstable marker.
For loss of heterozygosity (LOH) analysis using microsatellite markers (D2S123, D5S346, D17S250, D18S55, D18S56, D18S67, and D18S487), we duplicated the PCR in each sample to exclude allele dropouts of one of two alleles. Loss of heterozygosity at each locus was defined as ≥40% reduction of one of two allele peaks in tumor DNA relative to normal DNA. We obtained informative CIN results in 514 tumors (67%). The overall CIN score was defined as the number of chromosomal segments (among 2p, 5q, 17q, and 18q) that were positive for LOH.
Sodium bisulfite treatment on genomic DNA and subsequent real-time PCR (MethyLight)  were validated and performed as previously described . We quantified DNA methylation in eight CIMP-specific promoters [CACNA1G, CDKN2A (p16), CRABP1, IGF2, MLH1, NEUROG1, RUNX3, and SOCS1] [24,33], all of which were selected from screening of 195 CpG islands [24,33]. CpG island methylator phenotype-high was defined as the presence of ≥6 of 8 methylated promoters, CIMP-low as the presence of 1/8 to 5/8 methylated promoters, and CIMP-0 as the absence (0/8) of methylated promoters, according to the previously established criteria . In addition, we quantified DNA methylation in eight other CpG islands (not in the CIMP panel), including CHFR, HIC1, IGFBP3, MGMT, MINT1, MINT31, p14, and WRN. Primers and probes were previously described , except for HIC1, IGFBP3, p14, and WRN: HIC1-F, 5′-TTC GTT ACG GTA GTC GTT GTT TTC-3′ (GenBank no. L41919, nucleotide nos. 43–66); HIC1-R, 5′-GAA AAC TAT CAA CCC TCG ATC GA-3′ (nucleotide nos. 94–116); HIC1-probe, 6FAM-5′-TCG CGC GGT CGT CGT TCG-3′-BHQ-1 (nucleotide nos. 72–89); IGFBP3-F, 5′-GTT TCG GGC GTG AGT ACG A-3′ (GenBank no. M35878, nucleotide nos. 1692–1710); IGFBP3-R, 5′-GAA TCG ACG CAA ACA CGA CTA C-3′ (nucleotide nos. 1789–1810) and IGFBP3-probe, 6FAM-TCG GT T GT T TAG GGC GAA GTA CGG G-BHQ-1 (nucleotide nos. 1760–1784; bisulfite-converted nucleotides are highlighted by bold face and italics); P14 (CDKN2A/ARF)-F, 5′-T TG GAG GCG GCG AGA ATA T-3′ (GenBank no. L41934, nucleotide nos. 238–256); P14-R, 5′-CCC CGT AAA CCG CGA AAT A-3′ (nucleotide nos. 332–350); P14-probe, 6FAM-5′-CGG TTC GTC GCG AGT GAG GGT T-3′ -BHQ-1 (nucleotide nos. 299–320); WRN-F, 5′-GTA TCG TTC GCG GCG TTT AT-3′ (GenBank no. AY442327, nucleotide nos. 1827–1846); WRN-R, 5′-ACG AAA CCG ATA TCC GAA ATC A -3′ (nucleotide nos. 1887–1908) and WRN-probe, 6FAM-TTT TTT TTG CGG TCG TTG CGG G-BHQ-1 (nucleotide nos. 1855–1876). The PCR condition for all markers was initial denaturation at 95°C for 10 minutes followed by 45 cycles at 95°C for 15 seconds and 60°C for 1 minute .
To accurately quantify relatively high methylation levels in LINE-1 repetitive elements, we used Pyrosequencing as previously described . The LINE-1 methylation level measured by Pyrosequencing has been shown to correlate well with overall 5-methylcytosine level (i.e., genome-wide DNA methylation level) in tumor cells [38,39].
Tissue microarrays (TMAs) were constructed as previously described . Methods of immunohistochemical procedures and interpretations were previously described for p53 , p21 (CDKN1A) [41,42], β-catenin , COX-2 , and cyclin D1 .
For JCVT, antigen retrieval was performed, and deparaffinized tissue sections in Antigen Retrieval Citra Solution (Biogenex Laboratories, San Ramon, CA) were treated with microwave in a pressure cooker for 20 minutes. Tissue sections were incubated with 3% H2O2 (10 minutes) to block endogenous peroxidase (Dako Cytomation, Carpinteria, CA). Primary antibody against JCVT [mouse monoclonal anti-SV40 Tantigen (clone PAb416), 1:60 dilution; Oncogene Research Products, San Diego, CA] was applied, and the slides were maintained overnight at room temperature. Next, we applied an antimouse IgG antibody (Vector Laboratories, Burlingame, CA) for 30 minutes, followed by an avidin-biotin complex conjugate (Vector Laboratories) for 30 minutes, diaminobenzidine (5 minutes) and methyl-green counterstain. Nuclear JCVT expression was recorded as no expression, weak expression, or moderate/strong expression (Figure 1). JC virus T-antigen positivity (i.e., expression) was defined as the presence of at least weak nuclear staining. Appropriate positive and negative controls were included in each run of immunohistochemistry. All immunohistochemically stained slides were interpreted by one of the investigators (JCVT, cyclin D1, and β-catenin by K.N.; p53, p21, and COX-2 by S.O.) unaware of other data. A random selection of 147 cases was examined for JCVT by a second observer (Y.B.) unaware of other data, and concordance between the two observers was 0.87 (κ = 0.74, P < .0001), indicating substantial agreement.
All statistical analyses used SAS program (Version 9.1; SAS Institute, Cary, NC). All P values were two-sided, and statistical significance was set at P ≤ .05; however, P values were conservatively interpreted, considering multiple hypotheses testing. For categorical data, the χ2 test (or Fisher exact test when any expected cell count was less than 5) was performed and odds ratio (OR) with 95% confidence interval (CI) was computed. To compare mean LINE-1 methylation levels, the t test assuming unequal variances was performed. The κ coefficient was calculated to assess an agreement between the two interpreters in immunohistochemistry. To assess independent relations of JCVT with a number of variables, a multivariate logistic regression analysis was performed. Odds ratio was adjusted for age (<65 vs ≥65 years), sex, tumor location (proximal vs distal), body mass index (BMI, ≥30 vs <30 kg/m2), tumor stage (I–II vs III–IV), grade (low vs high), family history (present vs absent), mucin (present vs absent), signet ring cells (present vs absent), CIMP status (high vs low/0), MSI status (high vs low/MSS), LINE-1 methylation (as a continuous variable), β-catenin, COX-2, cyclin D1, p53, p21, BRAF, KRAS, and PIK3CA.
For survival analysis, Kaplan-Meier analysis was performed to assess survival time distributions according to JCVT status, and logrank test was performed. We also constructed a multivariate, stage-matched conditional Cox proportional hazard model to compute hazard ratios (HRs) according to tumoral JCVT status, adjusted for age, sex, year of diagnosis, tumor location, stage, grade, CIMP, MSI, KRAS, BRAF, PIK3CA, β-catenin, COX-2, cyclin D1 p53, p21, and LINE-1 methylation. In addition, a univariate Cox proportional hazard model was used to assess the main effect of JCVT on patient mortality. For the analyses of colorectal cancer-specific mortality, death because of colorectal cancer was the primary end point and deaths because of other causes were censored. The proportionality of hazards assumption was satisfied by evaluating time-dependent variables, which were the cross product of the JCVT variable and survival time (P = .30 for colon cancer-specific mortality; P = .48 for overall mortality). To adjust for potential confounding, age, year of diagnosis, and LINE-1 methylation were used as continuous variables, and all of the other covariates were used as categorical variables. An interaction was assessed by including the cross product of the JCVT variable and another variable of interest in a multivariate Cox model, and the likelihood ratio test was performed.
Among the 766 colorectal cancers assessed by immunohistochemistry, 271 (35%) tumors overexpressed JCVT. Table 1 summarizes the frequencies of JCVT expression according to various clinical and pathologic features. JC virus T-antigen expression was inversely associated with proximal location (P = .002), high grade (P = .046), and mucinous component (P = .0004).
We determined CIMP status using MethyLight assays on a panel of eight CIMP-specific promoters (CACNA1G, CDKN2A, CRABP1, IGF2, MLH1, NEUROG1, RUNX3, and SOCS1) [24,33]. Table 2 summarizes the frequencies of JCVT expression according to CIMP and other molecular features in colorectal cancer. JC virus T-antigen expression was inversely associated with CIMP-high (≥6/8 methylated promoters; OR, 0.40; 95% CI, 0.24–0.68, compared to CIMP-0; P = .0005) and MSI-high (OR, 0.25; 95% CI, 0.14–0.44, compared to MSS; P < .0001). To examine the combined effect of MSI and CIMP on JCVT expression, we classified tumors into four subtypes according to CIMP and MSI status (Table 2). JC virus T-antigen overexpression was significantly less common in CIMP-high MSI-high [14% (10/70), P < .0001) and CIMP-low/0 MSI-high [13% (4/31), P = .002] than in CIMP-low/0 MSI-low/MSS tumors [42% (238/573)].
We determined the CIN score in MSS/MSI-low tumors as the number of chromosomal segments (among 2p, 5q, 17q, and 18q) positive for LOH. JC virus T-antigen was significantly associated with high CIN score (≥2+; OR, 3.02; 95% CI, 1.82–5.00, compared to CIN score 0; P < .0001; Table 2). JC virus T-antigen expression was also significantly associated with p53 expression (P < .0001), loss of p21 expression (P < .0001), nuclear β-catenin expression (P = .006), and COX-2 expression (P = .02). In contrast, JCVT expression was not significantly associated with alterations in KRAS, BRAF, PIK3CA, or cyclin D1.
Because JCVT has been implicated in CpG island methylation , we examined the relationship of JCVT with methylation in 16 individual CpG islands, including the 8 CIMP-specific promoters (Table 3). In the univariate analysis, JCVT was inversely associated with methylation in CACNA1G, CRABP1, IGF2, MLH1, NEUROG1, RUNX3, SOCS1, HIC1, MINT31, p14, and WRN. However, after adjusting for CIMP, all of these associations were markedly attenuated, and no association was considered to be highly significant, given multiple hypotheses testing. These results suggest that none of these 16 methylation markers were directly related with JCVT.
We performed multivariate logistic regression analysis, to examine which variables were independently associated with JCVT (Table 4). JC virus T-antigen was significantly associated with p53 expression (multivariate OR, 8.45; 95% CI, 5.72–12.5; P < .0001) and high CIN score (multivariate OR, 2.53; 95% CI, 1.38–4.62; P = .003) and inversely with high tumor grade (multivariate OR, 0.44; 95% CI, 0.24–0.79; P = .006). JC virus T-antigen might also be associated with cyclin D1 expression, LINE-1 hypomethylation, BRAF mutation, and signet ring cells and inversely with family history of colorectal cancer (all P values between 0.05 and 0.01); however, given multiple hypotheses testing, any of these associations might be a chance event.
We assessed the influence of JCVT expression on survival of patients with stage I to IV colorectal cancers. We have previously shown that clinical outcome data in our two independent cohort studies are valid and reliable to detect significant molecular predictors of patient survival [44–46]. In the Kaplan-Meier analysis, JCVT expression was not significantly associated with patient survival (log-rank, P = .31 for colorectal cancer-specific mortality; log-rank, P = .67 for overall mortality). We performed Cox regression analysis to assess mortalities according to JCVT status (Table 5). For both cancer-specific and overall mortalities, JCVT was not significantly related with patient outcome in univariate analysis, stage-matched analysis, or multivariate analysis. When we limited cases to only colon cancers, JCVT remained unrelated with patient outcome. We examined whether JCVT was associated with patient mortality in any of the strata of clinical or molecular variables (such as age, sex, tumor stage, location, CIMP, MSI, BRAF, LINE-1, etc.). However, there was no evidence for the significant relation between JCVT and clinical outcome in any of the strata, and there was no evidence for significant interaction between JCVT and any of the variables in survival analysis (data not shown).
We conducted this study to examine the relations of JCVT expression with clinical, pathologic, and molecular characteristics and patient survival in colorectal cancers. Molecular correlates with JCVT may be important for a better understanding of genetic and epigenetic alterations during the colorectal carcinogenic process. We have found that JCVT is independently associated with p53 expression and CIN. In contrast, JCVT is inversely related with the CIMP and MSI in univariate analysis but not in multivariate analysis. Our data support the hypothesis that JCVT may affect p53 expression and CIN rather than CIMP and MSI in colorectal cancer.
Studying molecular changes is important in cancer research [47–64], and classification of colorectal cancer based on MSI and CIMP is increasingly important because it reflects genomic and epigenomic alterations, respectively, in tumor cells and largely determines clinical, pathologic, and molecular characteristics . To measure DNA methylation, we used real-time PCR (MethyLight Technology) for DNA methylation in eight CIMP-specific loci  as well as in eight other CpG islands. We also used Pyrosequencing to measure LINE-1 methylation, which has been correlated well with cellular 5-methylcytosine level (i.e., genome-wide DNA methylation level) [29,38,39]. Our resource of a large number of colorectal cancers derived from the two prospective cohort studies has enabled us to precisely estimate the frequency of colorectal cancers with a specific molecular feature (such as JCVT expression, CIMP-high, p53 expression, etc.). The large number of cases has also provided a sufficient power in our multivariate logistic regression analysis and survival analysis.
Previous studies have reported the relationship of JC virus with CIN and LOH [6,7]. Introduction of JC virus into a diploid cell line can lead to CIN . In addition, JCVT is strongly associated with LOH in colorectal cancers . In the current study, we have shown that JCVT is associated with CIN and LINE-1 hypomethylation, independent of other variables. These data collectively support a possible role of JCVT in the development of CIN and genome-wide DNA hypomethylation in colorectal cancers.
Regarding the relation between JCVT and p53, a previous study has reported that JCVT can bind and inactivate both p53 and phospho-RB proteins . Especially, JCVT may affect other regulatory mechanisms for p53, which have been implicated in cancer development . Together with our current data of the independent association between JCVT and p53 expression, accumulating data suggest that JCVT may dysregulate the p53 pathways, which can lead to uncontrolled proliferation of colorectal cancer cells.
We have demonstrated that JCVT expression is associated with p53 expression in colorectal cancer. One possible explanation for this phenomenon is that there might be some cases with poor antigenicity of JCVT and p53. Poor quality of tissue for immunohistochemistry would have yielded a false-negative result in either the JCVT or p53 immunoassay, driving the overall relationship between JCVT and p53 expression toward a concordant pattern. However, we have shown that JCVT expression is inversely associated with p21 expression, which cannot be explained by the presence of poor-quality specimens. p21 has been known to be induced by wild-type p53, and indeed, p53 expression (a surrogate of p53 mutation) was inversely associated with p21 expression in our cohorts. These results imply that the strong relation between JCVT and p53 expression we have observed is not simply caused by the presence of poor-quality specimens, leaving a possibility of a molecular interaction or other molecular correlates between JCVT and p53. Interestingly, the relation between JCVT and p21 loss did not persist after adjusting for p53, suggesting that the link between JCVT and p21 loss was mediated by p53 expression.
Previous studies have reported that methylation of host cell gene is not unique to JC virus and occurs with other oncogenic viruses [55,57,61,66]. Associations have been shown between methylation of multiple genes and Epstein-Barr virus in gastric cancer [55,57] and between promoter methylation and hepatitis B virus/hepatitis C virus in hepatocellular cancer [61,66]. A significant association has been found between the presence of SV40 T-antigen and methylation of multiple genes in non-Hodgkin lymphomas and mesotheliomas [67,68]. In addition, a previous study has reported that JCVT may induce CIMP in colorectal cancers through multiple mechanisms of epigenetic alterations . However, our data do not support the relation of JCVT with CIMP (determined by the validated panel of eight CIMP-specific promoters [24,33]) or methylation in any of the 16 CpG islands we examined, using a large number of colorectal cancers. We have shown an “inverse association” between JCVT and CIMP in univariate analysis (P = .0005), which became insignificant in multivariate analysis, indicating no independent association between JCVT and CIMP. In addition, none of the 16 CpG islands seemed to be specifically related with JCVT after adjusting for CIMP. This discrepancy is likely caused by the differences in the sample sizes (n = 100 in reference  vs n = 766 in our current study), the methylation markers examined (MLH1, APC, CDKN2A, p14, PTEN, TIMP3, RUNX3, HIC1, and RARB in reference  vs CACNA1G, CDKN2A, CRABP1, IGF2, MLH1, NEUROG1, RUNX3, SOCS1, and eight other CpG islands in our current study), the methods to detect DNA methylation (nonquantitative methylation-specific PCR in reference  vs quantitative MethyLight in our current study) and the criteria for methylator type or CIMP-high (no clear definition in reference  vs the presence of ≥6 of 8 methylated CIMP-specific promoters in our current study), and the statistical methods (no multivariate analysis in reference  vs both multivariate and univariate analyses performed in our current study). Considering that there is considerable heterogeneity of tumors with regard to CpG island methylation and that CpG islands are methylated in a different manner, the difference in the methylation markers between the two studies may explain discrepancies, at least in part. However, some CpG islands (MLH1, RUNX3, CDKN2A, p14, and HIC1) were used in both studies, and results on the same markers seemed to be discordant. We have conducted rigorous statistical analysis for each marker and performed multivariate analysis to assess independent associations and significant confounding. In addition, we have comprehensively examined the relation between JCVT and CIMP using the validated CIMP marker panel [24,33] and a large number (n = 766) of colorectal cancers with robust statistics. Our results suggest that JCVT may not contribute to CIMP in colorectal cancer. On the basis of our current results and data in the literature, Figure 2 represents hypothetical relations with JCVT in colorectal cancer.
In conclusion, using a large number of colorectal cancers, we have shown that JCVT is independently associated with p53 expression and CIN. Conversely, JCVT seems to be unrelated with CIMP, MSI, or patient outcome. Our data suggest that JCVT may contribute to CIN and dysregulation of the p53 pathways, which may lead to uncontrolled proliferation of colorectal cancer cells.
The authors thank the Nurses' Health Study and Health Professionals Follow-up Study cohort participants who have generously agreed to provide the biological specimens and information through responses to questionnaires. The authors thank Frank Speizer, Walter Willett, Susan Hankinson, Graham Colditz, Meir Stampfer, and many other staff members who implemented and have maintained the cohort studies.
No conflicts of interest exist.
1This work was supported by US National Institutes of Health (NIH) grants P01 CA87969, P01 CA55075, P50 CA127003, and K07 CA122826 (to S.O.) and in part by grants from the Bennett Family Fund and from the Entertainment Industry Foundation through the National Colorectal Cancer Research Alliance. K.N. was supported by a fellowship grant from the Japan Society for Promotion of Science. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Cancer Institute or NIH. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.