A large, nationwide, case-control study for the association of DLG5, OCTN1/2 and CARD15 with Inflammatory Bowel Diseases in the Netherlands.

Abstract

Introduction: Chronic inflammatory bowel diseases (IBD) comprising Crohn’s disease (CD) and ulcerative colitis (UC) have a complex genetic background. Conflicting results have been reported in association studies for genetic variants in Drosophila Discs Large Homologue 5 (DLG5) and in organic cation transporters 1 and 2 (OCTN1/2) and IBD. Therefore, a large, collaborative, nationwide study was conducted. DNA samples and clinical characteristics of IBD patients from seven University Medical Centers in the Netherlands were studied. Methods: 3061 IBD mainly Dutch Caucasian patients (1818 CD and 1243 UC) and 1518 healthy controls were included. Phenotypic details were available for 2317 patients (1497 CD / 820 UC). For DLG5 the risk associated SNP 113G→A and rs2165047 and for OCTN1/2 SNPs 1672 C→T and -207G→C were analyzed. Additional SNPs (rs272893, rs273900, rs274551) in OCTN1/2 were analyzed in a subset of 1263 (695 CD / 568 UC) patients. Results were stratified the three common CD risk associated alleles for CARD15. Results: For DLG5, 113G→A was associated with IBD (p=0.022), UC (p=0.033) and proctitis in UC (p=0.017) but not with CD. rs2165046 was associated with IBD (p=0.0042) CD (0.0043), familial occurrence in CD (p=0.0050) and proctitis in UC patients (p=0.033). OCTN1/2 SNPs 1672C→T and -207G→C were not associated with IBD, CD or UC except for subsets of stricturing CD (1672C→T, p=0.046) and leftsided UC or colectomy in UC (-207G→C, p=0.02 and p=0.035). rs272893, rs273900, rs274551 in OCTN1/2 were significantly associated (p<0.05) with an increased risk for IBD and CD. rs274551 was associated with UC (p=0.026), particularly with an early age of onset (p=0.008). No interaction with CARD15 was found for either DLG5 or OCTN1/2. Conclusion: In this large, well phenotyped, nationwide study in the Netherlands, we found an association of DLG5 113G→A with IBD, though not with CD but with UC. Furthermore rs2165046 (haplotype B) in DLG5 was associated with IBD and CD. We did not find an association for OCTN1/2 with IBD, CD or UC and SNPs 1672C→T and -207G→C, except for stricturing CD and leftsided UC or need for colectomy in UC. However, significant association was found with IBD, CD and (early onset) UC, for three additional SNPs in OCTN1/2. This supports the view that OCTN1/2 might be the associated gene in the IBD5 locus, but further analysis concerning the haplotype structure of IBD5 is needed. No epistasis between DLG5, OCTN1/2 and CARD15 was observed. Large, nationwide, case-control studies in rigorously phenotyped cohorts of IBD patients are needed to confirm previously described genetic associations.