Search in the Abstract Database

Search Abstracts 2015

* = Presenting author

P065 Identification of DSG3, MAGI1 and TFF1 as functionally important genes in inflammatory bowel disease pathogenesis

M. Vancamelbeke*, T. Vanuytsel, R. Farré, M. Ferrante, K. Verbeke, P. Rutgeerts, S. Vermeire, I. Arijs, I. Cleynen

KU Leuven, Clinical and Experimental Medicine, Leuven, Belgium

Background

Genetic and functional studies have implicated an intestinal epithelial barrier dysfunction ('leaky mucosal barrier') in inflammatory bowel disease (IBD). However, it remains unclear whether this dysfunction is a causal event in IBD or rather a consequence of mucosal inflammation. In this study, we investigated the role of intestinal epithelial barrier genes in IBD.

Methods

The mRNA expression of 128 genes involved in different aspects of intestinal epithelial barrier function was studied in 116 colonic mucosal biopsies (74 active ulcerative colitis (UC), 23 inactive UC, 8 active Crohn's disease (CD) and 11 controls) and in 78 ileal biopsies (51 active CD, 16 inactive CD and 11 controls). Disease activity was based on endoscopic findings. Total RNA from biopsies was used to analyse the gene expression with Affymetrix Human Gene 1.0 ST arrays (false discovery rate <0.05 and >2-fold change). We also compared allele frequencies of 3220 SNPs within 104 of the selected genes in our cohort of IBD patients (n=2804; 1856 CD, 948 UC) and 1013 healthy controls using immunochip data. In addition, we tested if these SNPs influenced expression of the differentially expressed genes (eQTL).

Results

No significant difference in colonic barrier gene expression was seen between active UC and active CD. In active IBD, the colonic expression of MUC1, MUC5B, EMCN, MCAM, TFF1, CLDN1, JAM2, DSG3, LAMA4, LAMC1, TCF4 and F2RL2 was upregulated compared to controls, whereas the colonic expression of RETNLB, CLDN8, OCLN, MAGI1 and MEP1A was downregulated. In inactive colonic IBD, no significant changes were observed compared to controls.

The ileal expression of MUC1, MUC4, MUC5B, MUC6, TFF1, CLDN1, CLDN18 and F2RL2 was significantly upregulated in active CD compared to controls, while only the ileal expression of CLDN8 was significantly downregulated. In inactive CD, MUC1 and MUC4 ileal expression remained upregulated, and CLDN8 downregulated compared to controls.

Forty-six genes showed a significant association (≥1 SNP with puncorrected<0.05) with IBD, of which 8 belonged to the differentially expressed genes (MUC1, MUC4, TFF1, CLDN8, DSG3, MAGI1, TCF4, MEP1A). Interestingly, SNPs in DSG3, MAGI1, and TFF1 did not only confer risk to IBD, but also were eQTLs for their expression in inflamed colon and ileum respectively.

Conclusion

Our data show a dysregulated expression of several barrier genes in active IBD patients, while only in inactive ileal CD patients few barrier genes remained dysregulated, suggesting a primary barrier defect in these patients. The expression data, but also genetic data and eQTL analysis point to DSG3, MAGI1 and TFF1 as possibly important and functional candidate genes for IBD pathogenesis.