HOME
Return to Table of Contents
A CANDIDATE GENE ASSOCIATION STUDY IDENTIFIES A NEW SUSCEPTIBILITY GENE FOR CROHN’S DISEASE INVOLVED IN IL-1beta PROCESSING
AC Villani1, M Lemire2, E Louis3, MS Silverberg4, C Collette1, G Fortin1, C Libioulle3, A Bitton1, D Gaudet5, A Cohen1, D Langelier6, JD Rioux7, P Rutgeerts8, S Vermeire8, TJ Hudson2,9, D Franchimont11
1McGill University Health Center; 2McGill University & Genome Quebec Innovation Centre; 3CHU of Liège, Belgium; 4Mount Sinai Hospital IBD Center; 5Université de Montréal, Canada; 6Centre Hospitalier de Sherbrooke; 7KUL, Belgium; 8Ontario Institute for Cancer Research
We used a candidate gene approach to identify novel Crohn’s disease (CD) susceptibility loci, and to complement the list of CD candidates recently reported from whole genome association (WGA) studies. We focussed on genes involved in IL-1 processing and signalling, a cytokine known to play a pivotal role in inflammation.
METHODS: 710 CD trios, 239 CD cases and 107 controls were assembled for this study. 62 tagging SNPs were selected within a 72kb interval and genotyped in all samples. Association testing was done using FBAT and UNPHASED softwares.
RESULTS: In the exploratory phase, the major allele of three SNPs was significantly associated with CD (p as low as 1.68×10–3) in the Leuven cohort (296 CD trios). Two of these signals, in strong LD with each other, were replicated in four independent cohorts: Liege trios (155 CD trios) (p=4.31×10–3), Liege case-controls (239 CD and 107 controls) (p=1.20×10–4), Quebec trios (130 CD trios) (p=0.0208), and Toronto trios (129 CD trios) (p=7.27×10–5). We selected 24 individuals based on carrier status of risk alleles and sequenced the 9kb associated region. We observed 81 SNPs in LD with the associated region and performed comprehensive genotyping to localize the most disease-associated polymorphism. No coding SNP could explain the signals. Yet, we uncovered two SNPs located on putative functional sites, which were genotyped across all sample sets. The major allele of both SNPs was significantly associated with CD in all 5 sample sets, with associations as low as pcomb=1.87×10–4. A pooled analysis comprising all sample sets evaluated the strongest association to pcomb=1.66×10–9 (OR: 1.809; CI: 1.486-2.201). We are currently conducting functional studies.
CONCLUSION: We have uncovered a novel CD susceptibility gene using a candidate gene approach, emphasizing the importance of IL-1 in the pathogenesis of CD. These results highlight the utility of a hypothesis driven methodology to complement WGA studies.