CDX2 VDR Polymorphism and Colorectal Cancer Free

Polymorphisms of the vitamin D receptor (VDR) gene have been associated with several forms of cancer and other chronic diseases (1). VDR, located on chromosome 12 at q12-14 consists of the noncoding 5′ promoter region, the 5′ end of the gene that includes exons 1A, 1B, and 1C, exons 2 and 3 which encode the DNA-binding domain, and exons 4 to 5 which encode the ligand-binding region (2). Different domains of VDR are involved in different functions including DNA binding, receptor dimerization, gene transactivation, and cofactor activation (3).

Prior studies of VDR polymorphisms in relation to colorectal cancer and adenomas have focused on five polymorphisms at the polymorphic 3′-untranslated region: rs1544410 (detected by Bsm1) in intron 8, rs731236 (detected by Taq1) in exon 9, rs7975232 (detected by Apa1) in intron 8, rs757343 (detected by Tru91) in intron 8, and a polyadenylic acid [poly(A)] mononucleotide repeat (4-9). These polymorphisms seem to be in strong linkage disequilibrium, although there are racial variations (9). Another polymorphism examined in relation to risk of colorectal cancer (7, 9) is rs10735810 in exon 2, detected by Fok1, located near the 5′-untranslated region of the gene within the DNA binding domain and exhibits almost no linkage disequilibrium with the 3′ polymorphisms. The CDX2 VDR polymorphism (rs11568820) is in the promoter region of the VDR gene in exon 1; the G allele decreases VDR transcriptional activity relative to the A allele (10) but has not been examined in relation to colorectal cancer. Because different polymorphisms may function differently depending on its location (2), we took the opportunity to broaden our understanding of the VDR gene by examining the CDX2 polymorphism in two large multicenter case-control studies of colon and rectal cancers.

Data for the study came from two case-control studies conducted in Utah, the Northern California Kaiser Permanente Medical Care Program (KPMCP), and the Twin Cities Metropolitan area of Minnesota (colon cancer study only). Eligibility criteria included being between 30 and 79 years of age at the time of diagnosis, English speaking, mentally competent to complete the interview, with no previous history of colorectal cancer, and with no known (as indicated on the pathology report or by self-report) familial adenomatous polyposis, ulcerative colitis, or Crohn's disease. Controls were frequency matched to cases by sex and 5-year age groups. Controls were randomly selected from membership lists (KPMCP) or from driver's license lists (Minnesota). In Utah, controls 65 years of age or older were randomly selected from lists provided by the Centers for Medicare and Medicaid Services, whereas controls younger than 65 years of age were randomly selected from driver's license lists. Study eligibility and recruitment details of the study have been published previously (11, 12). Cooperation rates were 83% at KPMCP, 76% at Utah, and 67% for Minnesota (colon cases); 73% at KPMCP, 53% at Minnesota, and 69% at Utah (colon controls); 75.4% at KPCMP, 69.7% at Utah (rectal cases), and 69.9% at KPMCP and 67.2% at Utah (rectal controls).

The current analysis is restricted to subjects who provided a blood sample from which the VDR CDX2 polymorphism could be genotyped. The colon cancer study population consists of non–Hispanic white cases (n = 1444) and controls (n = 1841), Hispanic or American Indian cases (n = 60) and controls (n = 75), and African American cases (n = 70) and controls (n = 54). The rectal cancer study population consists of non–Hispanic white cases (n = 657) and controls (n = 856), Hispanic or American Indian cases (n = 63) and controls (n = 69), African American cases (n = 31) and controls (n = 44), and Asian cases (n = 40) and controls (n = 30). Ten subjects were excluded due to missing race and ethnicity data.

Trained and certified interviewers collected diet and lifestyle data as previously described (13, 14). The referent year for the study was the calendar year ∼2 years prior to date of diagnosis (cases) or selection (controls). Information was collected on demographic factors such as age, sex, and study center, and on exposures including diet, physical activity, aspirin and nonsteroidal drug use, body size, and other lifestyle factors including medical, family, and reproductive history.

DNA was extracted from blood drawn from study participants. A TaqMan assay was used to detect the single nucleotide polymorphism rs11568820 (CDX2). Genotyping was done in 5 μL reactions on 20 ng genomic DNA, in the presence of assay-specific primers and probes, 1× TaqMan universal PCR master mix (containing AmpErase UNG, AmpliTaq Gold enzyme, and reaction buffer). Control samples representing all possible genotypes were included at four positions in every 384-well tray. Internal replicates representing >1% of the sample set were blinded and included. The methods used for the intron 8 Bsm1 (rs154410), the poly(A), and the Fok1 (rs10735810) VDR polymorphisms have been described previously (9, 15).

The SAS statistical package, version 9.1 (SAS Institute) was used to conduct the analyses. Allele, genotype, and haplotype frequencies were estimated from the control population stratifying by race or ethnicity. In controls, the absolute value of Lewontin's D′ and the square of the correlation coefficient were used to determine if VDR CDX2 was in linkage disequilibrium with the other VDR polymorphisms. The individual haplotype probability estimates are based on the estimates of the controls within each racial or ethnic strata using the expectation-maximization algorithm (SAS Genetics package), resulting in each case and control having a value for each haplotype. Genotypes and haplotypes were examined in relation to colon and rectal cancer in multivariate logistic regression models. Models were adjusted for age at selection or diagnosis, study center, race or ethnicity, sex, energy intake, dietary calcium, dietary vitamin D, total dietary fats, and sunshine exposure, all of which have been shown to influence associations with other VDR polymorphisms. Odds ratios (OR) and 95% confidence intervals (CI) were used to report associations obtained from the multivariate logistic regression models. Trend is assessed by comparing the log-likelihood of a logistic regression model with the variable of interest entered as an ordered categorical variable to the log-likelihood of a model without the variable of interest using a χ² test with 1 df. The global omnibus test for case-control differences over all haplotypes was calculated. Effect modification between genotypes and exposure variables were evaluated by likelihood ratio test for a multiplicative interaction term in the logistic regression model.

The frequency of the CDX2 G allele varied by ethnic group (Table 1), with an allele frequency of 81% among non–Hispanic white participants but only 24% among African American participants. The frequencies of haplotypes derived from the four VDR polymorphisms also varied by ethnicity (Table 2); the bLFG haplotype was most common in non–Hispanic white and Hispanic populations, whereas the bLFA haplotype was most common in African American and Asian populations.

The VDR CDX2 polymorphism was not associated with increased risk of either colon or rectal cancer (Table 3); associations were similar for men and women and further adjustment for the Fok1 VDR polymorphism did not alter the association, although the Fok1 polymorphism remained statistically significantly associated with colon cancer after adjusting for the CDX2 polymorphism. The overall omnibus test for case-control differences among haplotype categories was 0.007. The bLFA haplotype was associated with a statistically significantly increased risk of colon cancer (OR, 2.45; 95% CI, 1.38-4.38) but not rectal cancer. The BSFA haplotype was associated with a significantly reduced risk of rectal cancer (OR, 0.71; 95% CI, 0.52-0.97 for each copy of the haplotype) and the BSfG haplotype was associated with increased risk of rectal cancer (OR, 1.61; 95% CI, 1.05-2.49 per copy of the haplotype). Although not statistically significant, the bLFG haplotype was also associated with reduced risk of colon cancer (OR, 0.62; 95% CI, 0.27-1.41) and the bLfG was associated with an increased risk of colon cancer (OR, 2.02; 95% CI, 0.71-5.75). These associations were not observed for rectal cancer.

There were no statistically significant interactions of the CDX2 VDR polymorphisms with calcium, vitamin D, sunshine exposure, and regular use of nonsteroidal anti-inflammatory drugs/aspirin. Additionally, CDX2 VDR did not interact with other VDR polymorphisms to alter risk of colon or rectal cancer (data not shown in the table).

The CDX2 VDR polymorphism has been shown to activate VDR gene transcription in the intestinal tract (10), and therefore, an association with colon or rectal cancer is plausible. Polymorphisms of the VDR gene have been associated with colorectal adenomas and cancer with individuals carrying the B Bsm1, f Fok1, and Tru91 alleles having a reduced adenoma or cancer risk ranging from 20% to 80% (4, 5, 7, 8, 15, 16). Although not all studies have shown consistent associations with the Fok1 polymorphism, it has been associated with colon cancer in some studies, including our own previously published work (4, 7, 17). In the present study, we describe large variation in the CDX2 polymorphism allele frequency by racial and ethnic group, and show that CDX2 exhibits only limited linkage disequilibrium with other putative functional polymorphisms of the VDR. We did not, however, observe an association between the CDX2 VDR polymorphism genotype and colon or rectal cancer.

We have evaluated haplotypes based on the Bsm1, poly(A), Fok1, and CDX2 polymorphisms. Our previous colon cancer haplotype analysis (9) noted that the bLF haplotype was associated with a statistically significantly increased colon cancer risk, with an OR of 1.15 (95% CI, 1.03-1.28). Addition of the CDX2 polymorphisms suggests that those with the bLFG are at a statistically nonsignificantly reduced risk of colon cancer, with the increased risk limited to those with bLF and the A allele at CDX2. It is noteworthy that whereas 6.5% of non–Hispanic white populations have the bLFA haplotype, 41.2% of African Americans carry this haplotype. Although we were limited in power to evaluate the associations between this haplotype and colon cancer by race, it would be of interest to determine if this haplotype contributes to the colon cancer risk among African American populations given their high incidence rates of colon cancer (18). For rectal cancer, a previous analysis of haplotypes based on Bsm1, poly(A), and Fok1 showed no statistically significant associations, but the addition of the CDX2 polymorphism indicated that those with the BSFA haplotype had a reduced risk compared with those with the BSFG haplotype. The omnibus test for case-control differences over haplotypes that included CDX2 was statistically significant (χ² value of 31.65 with 15 df; this compares to a χ² value of 26.57 for the seven haplotypes based on the poly(A), Bsm1, and Fok1 markers).

Several studies, including our own, have shown that associations between VDR Bsm1 and Fok1 polymorphisms are influenced by levels of dietary calcium, fat, and vitamin D as well as with sunshine (4, 6-8, 17). We did not observe similar interactions with the CDX2 VDR polymorphism. However, as in our previous reports (8, 9, 17), we did observe different trends in associations for colon and rectal cancer in our examination of haplotypes. The reason for these site-specific differences in association can only be speculated upon. Given the many biological functions of VDR and its influence on multiple disease pathways, along with the fact that VDR polymorphisms in different domains might influence VDR activity differently, it is plausible that different haplotypes also influence cancer sites differently. We can only speculate on how specific haplotypes function. Because the CDX2 polymorphism is in the DNA binding portion of the gene, it is possible that it influences other transcriptional processes.

In summary, although we did not observe significant associations between the CDX2 VDR polymorphism genotypes and risk of colon or rectal cancer, haplotypes encompassing the CDX2 variant added information, although only slightly more so than haplotypes consisting of only the Bsm1, poly(A), and Fok1 VDR polymorphisms. Although we did not have power to examine the extent to which the VDR haplotypes contributed to colorectal cancer in various racial and ethnic groups, they are of potential importance given the differences in the prevalence of specific haplotypes according to racial and ethnic groups.

We acknowledge the contributions of Michael Hoffman and Thao Tran for genotyping; and Sandra Edwards, Karen Curtin, Roger Edwards, Leslie Palmer, Donna Schaffer, Dr. Kristin Anderson, and Judy Morse for data management and collection.