CEACAM6 Gene Variants in Inflammatory Bowel Disease

Background The carcinoembryonic antigen-related cell adhesion molecule 6 (CEACAM6) acts as a receptor for adherent-invasive E. coli (AIEC) and its ileal expression is increased in patients with Crohn's disease (CD). Given its contribution to the pathogenesis of CD, we aimed to investigate the role of genetic variants in the CEACAM6 region in patients with inflammatory bowel diseases (IBD). Methodology In this study, a total of 2,683 genomic DNA samples (including DNA from 858 CD patients, 475 patients with ulcerative colitis (UC), and 1,350 healthy, unrelated controls) was analyzed for eight CEACAM6 SNPs (rs10415946, rs1805223 = p.Pro42Pro, rs4803507, rs4803508, rs11548735 = p.Gly239Val, rs7246116 = pHis260His, rs2701, rs10416839). In addition, a detailed haplotype analysis and genotype-phenotype analysis were performed. Overall, our genotype analysis did not reveal any significant association of the investigated CEACAM6 SNPs and haplotypes with CD or UC susceptibility, although certain CEACAM6 SNPs modulated CEACAM6 expression in intestinal epithelial cell lines. Despite its function as receptor of AIEC in ileal CD, we found no association of the CEACAM6 SNPs with ileal or ileocolonic CD. Moreover, there was no evidence of epistasis between the analyzed CEACAM6 variants and the main CD-associated NOD2, IL23R and ATG16L1 variants. Conclusions This study represents the first detailed analysis of CEACAM6 variants in IBD patients. Despite its important role in bacterial attachment in ileal CD, we could not demonstrate a role for CEACAM6 variants in IBD susceptibility or regarding an ileal CD phenotype. Further functional studies are required to analyze if these gene variants modulate ileal bacterial attachment.

While a specific causative pathogen in IBD has not been found so far [13,14], investigations of the regulatory mechanisms operating at the mucosal level suggest that regulatory cells reactive to the commensal intestinal microflora might play a role in cross-reactive protection toward different bacterial antigens [15]. Moreover, there is raising evidence for a major role of certain bacteria such as adherent-invasive E. coli (AIEC) in ileal CD [16,17,18]. Interestingly, the carcinoembryonic antigen-related cell adhesion molecule 6 (CEACAM6) has recently shown to act as a receptor for AIEC, supporting ileal bacterial colonization as a major pathomechanism in CD [19]. The carcinoembryonic antigen (CEA) family consists of two subfamilies, the CEACAM subgroup and the pregnancy specific glycoprotein (PSG) subgroup [20,21]. CEACAM family members were found to be expressed in epithelial, endothelial, and hematopoietic cells, including T-lymphocytes, natural killer (NK) cells, dendritic cells (DC) and neutrophils. They may also be useful as biomarkers in cancer since they are often over-expressed in ovarian, endometrial, breast, lung, and colon carcinomas [21,22,23]. Depending on the tissue involved, CEACAMs are transmitting signals that result in a variety of effects including regulation of the cell cycle, tumor suppression, angiogenesis, lymphocyte activation and adhesion [22,23,24,25,26,27,28,29]. CEACAM1, CEACAM5, and CEACAM6 represent three of the CEACAM subfamily members expressed in intestinal epithelial cells. There is increased expression of CEACAM5 and CEACAM6 at the apical surface of the ileal epithelium in CD patients [19]. Moreover, ileal lesions in CD patients were found to be colonized by pathogenic AIEC [19], strengthening the hypothesis that an abnormal intestinal expression of CEACAM6 in CD patients is associated with an increased colonization of AIEC via type 1 pili expression inducing gut inflammation [18]. AIEC adhere to and invade intestinal epithelial cells [30] resulting in AIEC accumulation in macrophages leading to high amounts of TNF-a [31], thereby perpetuating intestinal inflammation.
Given the potential implication of dysfunctional CEACAM6 expression in the pathogenesis of IBD, we aimed to analyze the role of CEACAM6 SNPs in IBD susceptibility. A total of eight single nucleotide polymorphisms (SNPs) were analyzed in a large German cohort of CD and UC patients. Five SNPs in the CEACAM6 region (rs10415946, rs4803507, rs4803508, rs2701, rs10416839) were selected from the data of the international HapMap project covering the CEACAM6 gene plus 10 kB flanking the centromeric and telomeric end of the gene, respectively. Additional selection criteria for the SNPs were a minor allele frequency of at least 5% and a r 2 of 1. The SNPs rs4803507 and rs4803508 are localized in intron 2, rs2701 is localized within exon 6 encoding the 39untranslated region, while the SNPs rs10415946 and rs10416839 are within the 59-and the 39-flanking region, respectively. Additionally, the coding variants rs1805223 = p.Pro42Pro (exon 2), rs11548735 = p.Gly239Val and rs7246116 = pHis260His (exon 4) were investigated for which allele frequencies are published und which display a minor allele frequency of at least 5% in the Caucasian population. The structure of the CEACAM6 gene and the localization of the SNPs investigated in the presented study are shown in figure 1. Considering the abnormal expression of CEACAM6 in the ileal epithelium of CD patients and its role as receptor for ileal AIEC [19], we also analyzed for a potential association with an ileal CD phenotype and investigated potential gene-gene interactions with the NOD2 gene, which has been shown to be a strong predictor of ileal CD, as well as with other CD susceptibility genes such as IL23R and ATG16L1.

Ethics statement
The study was approved by the local Ethics committee of the Ludwig-Maximilians-University of Munich (Department of Medicine) and adhered to the ethical principles for medical research involving human subjects of the Helsinki Declaration. Prior to the study, we obtained written, informed consent from all patients included.

Study population and characterization of disease phenotype
The study population comprised 858 patients with CD, 475 patients with UC, and 1350 healthy, unrelated controls of Caucasian origin. The study was approved by the local Ethics committee of the Ludwig-Maximilians-University of Munich (Department of Medicine) and adhered to the ethical principles for medical research involving human subjects of the Helsinki Declaration. Prior to the study, we obtained written, informed consent from all patients included. The phenotypic assessment was performed blind to the results of the genotypic data and included demographic data and clinical parameters (behaviour and anatomic location of IBD, disease-related complications, previous surgery or immunosuppressive therapy) which were recorded by investigation of patient charts and a detailed questionnaire. The diagnosis of CD or UC was based on established international guidelines including endoscopic, radiological, and histopathological criteria [32,33]. Patients with CD were assessed according to the Montreal classification [33] based on age at diagnosis (A), location (L), and behaviour (B) of disease. In patients with UC, anatomic location was also assessed in accordance to the Montreal classification [33], using the criteria ulcerative proctitis (E1), leftsided UC (distal UC; E2), and extensive UC (pancolitis; E3). Patients with indeterminate colitis were excluded from the study. The demographic and phenotypic data of the IBD study population are summarized in Table 1.

DNA extraction and genotyping of the CEACAM6 variants
Blood samples were taken from all participants of the study and genomic DNA was isolated from peripheral blood leukocytes using the DNA blood mini kit from Qiagen (Hilden, Germany) according to the manufacturer's guidelines. Eight CEACAM6 SNPs (rs10415946, rs1805223 = p.Pro42Pro, rs4803507, rs4803508, rs11548735 = p.Gly239Val, rs7246116 = pHis260His, rs2701, rs10416839) were genotyped by PCR and melting curve analysis using a pair of fluorescence resonance energy transfer (FRET) probes in a LightCyclerH 480 Instrument (Roche Diagnostics, Mannheim, Germany) as described in previous studies [34,35,36]. The total volume of the PCR was 5 ml containing 25 ng of genomic DNA, 16 Light Cycler 480 Genotyping Master Mix (Roche Diagnostics), 2.5 pmol of each primer and 0.75 pmol of each FRET probe (TIB MOLBIOL, Berlin, Germany). In the case of rs1805223, rs4803507 and rs4803508, the concentration of the forward primer, and in the case of rs10415946 and rs2701, the concentration of the reverse primer, were reduced to 1.25 pmol.

Statistical analyses
Each genetic marker was tested for Hardy-Weinberg equilibrium in the three subgroups of the study population. Fisher's exact test was used for comparison between categorical variables, while Student's t test was applied for quantitative variables. Singlemarker allelic tests were performed with Pearson's x 2 test. All tests were two-tailed, considering p-values,0.05 as significant. Odds ratios were calculated for the minor allele at each SNP. For multiple comparisons, Bonferroni correction was applied where      harvard.edu/,purcell/plink/) was used running a sliding window approach with variation of the window size from 2 to 8 included markers and using the option ''hap-logistic''. Linkage disequilibrium (LD) was also analyzed using PLINK.

CEACAM6 variants are not associated with IBD susceptibility
The genotyping success rates were at least 99% for all eight SNPs tested and were comparable between the controls and the CD and UC patients groups. In all three subgroups (CD, UC, and controls), the allele frequencies of the CEACAM6 SNPs rs10415946, rs1805223 = p.Pro42Pro, rs4803507, rs4803508, rs11548735 = p.Gly239Val, rs7246116 = p.His260His, rs2701, rs10416839 were in accordance with the predicted Hardy-Weinberg equilibrium (Table 4). Overall, we observed no significant differences in the frequencies of the investigated CEACAM6 SNPs in CD and UC patients compared to healthy controls ( Table 4) implicating no significant association of CEACAM6 variants and IBD susceptibility. Only two patients (both with UC) were minor allele carriers of the rare CEACAM6 SNP rs7246116 = p.His260His, therefore not allowing a comparative analysis of this SNP regarding CD susceptibility.

CEACAM6 haplotypes are not associated with CD and UC susceptibility
Considering recent evidence showing that certain CEACAM6 haplotypes modulate susceptibility to bacterial infections [38], we next performed a detailed haplotype analysis in our IBD cohort. However, as shown in Tables 5 and 6, we could not demonstrated significant associations of CEACAM6 haplotypes with CD and UC susceptibility.
The CEACAM6 variants are not associated with an ileal disease phenotype in CD patients Since CEACAM6 has recently shown to act as a receptor for AIEC, thereby promoting bacterial colonization in ileal CD [19], we further investigated whether CEACAM6 SNPs are associated with ileal disease in CD patients. Based on a phenotype analysis according to the Montreal classification of IBD [33], the detailed phenotypic data available from a subcohort of 667 CD patients was analyzed for disease localization. None of the investigated CEACAM6 SNPs was associated with ileal or ileocolonic CD     (Table 7). However, we have to acknowledge that the sample size has limited power to detect weak disease associations. For example, based on 667 patients with a L1/L3 disease phenotype, a minor allele frequency of 0.40 and an OR of 1.1, the power is 28.88% to detect an effect on a significance level of 5% (Genetic Power Calculator, http://pngu.mgh.harvard.edu/,purcell/gpc/).
Analysis for gene-gene interaction with CD-associated NOD2, IL23R and ATG16L1 variants Given the raising evidence for a key role of CEACAM6 in the complex interaction of the mucosal immune system and intestinal bacteria, we next analyzed for potential epistasis between CEACAM6 SNPs (rs10415946, rs1805223 = p.Pro42Pro, rs4803507, rs4803508, rs11548735 = p.Gly239Val, rs7246116 = pHis260His, rs2701, rs10-416839) and the three main CD-associated NOD2/CARD15 variants p.Arg702Trp (rs2066844), p.Gly908Arg (rs2066845), and p.Leu1007fsX1008 (rs2066847) which have previously shown to be strongly associated with CD and ileal disease localization. However, there was no evidence for epistasis between the CEACAM6 SNPs and the three analyzed NOD2/CARD15 variants (Table 8).
Recently, we demonstrated an association of the IL23R SNP rs1004819 (TT homozygous carriers) with ileal CD [34]. Therefore, we also analyzed for potential gene-gene interaction between CEACAM6 SNPs and the major CD-associated IL23R variants. However, we did not find evidence for epistasis between CEACAM6 and IL23R regarding CD susceptibility (Table 9).
In addition, novel findings indicate a major role for ATG16L1 in Paneth cell development in the terminal ileum. Therefore, we also analyzed for potential epistasis between the CEACAM6 SNPs and the major CD-associated ATG16L1 SNPs. However, we were also unable to demonstrate evidence for epistasis between these two genes (Table 10).
p-values for epistasis between CEACAM6 and ATGT16L1 SNPs in the CD case-control sample. After Bonferroni correction, the association highlighted in bold did not remain significant. * There were no carriers of the minor allele of rs7146116 in the CD and control cohort.

CEACAM6 genotypes modulate CEACAM6 expression in intestinal epithelial cell lines
To analyze a potential influence of CEACAM6 gene variants on CEACAM6 gene expression, we determined CEACAM6 mRNA levels in five intestinal epithelial cell (IEC) lines DLD-1, HCT116, HT-29, SW480 and T84 by quantitative PCR. This analysis revealed considerable differences in CEACAM6 expression depending on the cell line. While T84 cells showed the highest expression, the expression in HCT116 cells was four orders of magnitude smaller and close to the detection limit ( Fig. 2A).
SW480 and DLD-1 cells showed similar, intermediate expression ( Fig. 2A). Interestingly, when these cell lines were analyzed for CEACAM6 gene variants, only T84 and HCT116 cells, the two cell lines with the highest and lowest CEACAM6 expression, respectively, had unique genotype variants when compared with the other cell lines (Fig. 2B). While T84 cells were the only cells that had a unique genotype for rs10415946 and rs1805223 = p.Pro42Pro, HCT116 had a unique genotype in SNPs rs11548735 = p.Gly239Val, rs2701 and rs10416839. SW480 and DLD-1 cells had identical CEACAM6 genotypes and their CEACAM6 expression was nearly identical ( Fig. 2A and 2B). A detailed analysis regarding linkage disequilibrium between the investigated CEACAM6 SNPs stratified for CD, UC and controls is given in Tables 11, 12 and 13.

Discussion
In summary, our study represents the first detailed analysis of CEACAM6 SNPs regarding disease susceptibility and phenotypic consequences in IBD patients. Compared to previous GWAS, our study had a more complete coverage of the CEACAM6 gene region (see details in Table 14). Overall, we did not observe a significant influence of the investigated CEACAM6 SNPs on CD and UC susceptibility. Moreover, a detailed haplotype analysis did not reveal significant associations with IBD susceptibility. CEACAM6 has recently shown to act as a receptor for AIEC suggesting an important role in bacterial colonization of the ileal mucosa in CD patients [19]. However, none of the investigated CEACAM6 SNPs was associated with ileal or ileocolonic CD.
Interestingly, a recent study demonstrated that the defect in CEACAM family members in intestinal epithelial cells isolated from CD patients appears to be related to the aberrant nuclear localization of the transcription factor SOX9 [39] which regulates cell proliferation and is required for Paneth cell differentiation in  the intestinal epithelium [40,41]. However, ileal CD is characterized by a specific decrease in Paneth cell alpha-defensins and defective Paneth cell-mediated host defense [42] which has been linked to the NOD2 genotype [43], although this finding is opposed by the results of a recent study [44], and additional modifiers of Paneth cell function such as XBP1 are involved [45]. Therefore, one might speculate whether the role of CEACAM6 in ileal bacterial colonization is regulated via SOX9 expression implicating defective Paneth cell function in patients with small bowel CD. Given the association of defensin secretion with the NOD2 genotype [43] and the findings of numerous previous studies including studies from our IBD center demonstrating ileal disease localization in CD patients with NOD2 mutations [37,46,47], we also tested for potential gene-gene interaction of CEACAM6 and NOD2. However, we found no evidence for epistasis between these two genes regarding CD susceptibility. Further functional studies analyzing the complex interaction of intestinal CEACAM6 expression and bacterial adherence in the gut particularly of CD patients carrying CEACAM6 variants will therefore be required. Given the important role of ATG16L1 in Paneth cell development of the terminal ileum and the role of IL23R in the development of proinflammatory Th17 cells, we also analyzed for epistasis of these two genes with CEACAM6 but were unable to find evidence for significant epistasis of these genes regarding CD susceptibility. Interestingly, a recent study indicated that CEACAM6 and a regulatory element near the 39 end of CEACAM3 are associated with disease severity in patients with cystic fibrosis [48]. However, a previous study in IBD patients suggested that heterozygous carriers of the DF508 mutation in the CFTR gene, the main susceptibility gene for patients with cystic fibrosis, might exert a protective effect in CD [49], suggesting opposing effects of genetic risk loci for cystic fibrosis and IBD.
In the meta-analysis of Barrett et al. [50], a SNP (rs4807569) within the chromosomal region 19q13, in which the CEACAM6 gene is located, was weakly associated with CD, but this association could not be confirmed in a replication cohort. In the recent meta-analysis of Franke et al. [51], two SNPs (rs 736289 and rs281376) within this region were strongly associated with CD. However, the distance between these SNPs and the CEACAM6 gene is 9 and 7 megabases, respectively, and thus, the disease causing variant within this region remains to be identified.
CEACAM6 is also a major target gene for Smad3-mediated TGF-b signaling [52]. Since Smad3 differentially regulates the induction of regulatory and inflammatory Th17 cell differentiation [53], which are key players in the IBD pathogenesis [54], further investigations analyzing Th17 cell differentiation in IBD patients carrying CEACAM6 variants might also be of high interest. Moreover, very recent evidence from studies in mice demonstrated that colonization of the small intestine with a single commensal microbe, segmented filamentous bacterium (SFB), is sufficient to induce Th17 cells in the lamina propria [55]. These SFB adhere tightly to the surface of epithelial cells in the terminal ileum of mice with Th17 cells but are absent from mice that have few Th17 cells [55]. Further studies will have to characterize if SFB adherence is mediated (similar to AIEC adherence) by CEACAM family members.
In summary, we performed the first systemic analysis of CEACAM6 gene variants in IBD patients. Despite the great importance of CEACAM6 as receptor for AIEC on the ileal mucosa of CD patients, we were unable to demonstrate a specific role of CEACAM6 variants in IBD susceptibility. Furthermore, there was no evidence for an association with ileal CD or for epistasis with NOD2, IL23R, and ATG16L1 variants in CD susceptibility. Further functional studies will be necessary to elucidate how CEACAM6 gene variants may modulate bacterial colonization in IBD patients. Even if this study was unable to find a role for CEACAM6 gene variants in IBD susceptibility, the CEACAM6 protein is likely to be an important mediator of the pathogenesis of CD [56].

Acknowledgments
This work contains parts of the unpublished degree theses of C. Fries.