Background: Genetic variants in DLG5, which encodes a scaffolding protein on chromosome 10q23, and tumor necrosis factor (TNF)-alpha, encoding a proinflammatory cytokine on chromosome 6p, have recently been reported to be associated with inflammatory bowel disease (IBD). We studied these variants to seek evidence of association with IBD in a large independent dataset.
Methods: We genotyped 1104 unrelated white IBD subjects-496 with Crohn's disease, 512 with ulcerative colitis, and 96 with indeterminate colitis from the Cambridge/Eastern (UK) panel-and 760 healthy control subjects for DLG5_113G/A, DLG5_4136C/A, TNF-857C/T, and TNF-1031T/C polymorphisms. Known Crohn's disease-predisposing variants in CARD15/NOD2 were also genotyped to permit analysis for reported epistatic interactions.
Results: : TNF-857 was shown to be associated with IBD overall (P = 0.0079). A formal interaction test showed that TNF-857 is associated equally with ulcerative colitis and Crohn's disease. Neither of the DLG5 alleles, however, was associated with IBD (P = 0.32 and 0.35). Subgroup analysis also failed to show evidence of association between either DLG5 allele or genotype frequencies and ulcerative colitis or Crohn's disease. Stratification of TNF-alpha and DLG5 cases by CARD15 genotype made no significant difference in the strength of associations.
Conclusions: We have confirmed an association between the TNF-857 promoter polymorphism and IBD in a large independent UK dataset but were unable to replicate an association at the previously reported loci within DLG5. This may reflect heterogeneity between the populations, a smaller effect size than originally predicted, or possibly a false-positive result in the original study. Further fine mapping studies of the TNF promoter region and studies assessing functional consequences of TNF promoter polymorphisms are now required in IBD.