|Home | About | Journals | Submit | Contact Us | Français|
The discovery of NOD2/CARD15 as the first susceptibility gene in Crohn's disease has contributed significantly to a fundamental change in the direction of basic research in inflammatory bowel disease (IBD), triggering renewed interest in the integrity of the innate immune response in IBD and appropriate orchestration of a subsequent adaptive immune response.1,2 More widely, in all complex diseases this finding in 2001 provided a much welcomed and needed proof of principle for non‐parametric linkage analysis.
Another study with major implications for the pathogenesis of Crohn's disease as well as for investigation of all complex disorders has recently been published.3 The North American consortium performed an association study testing 308332 markers spanning the entire genome in 567 patients with ileal Crohn's disease and 571 controls of non‐Jewish European ancestry. Of the three markers reported to retain significance after stringent Bonferroni correction, two were located in the NOD2/CARD15 gene. The third marker (rs11209026) was a non‐synonymous variant in the interleukin‐23 receptor (IL23R) gene on chromosome 1p31. Replication was obtained in the index paper in a Jewish ancestry case‐control analysis of patients with Crohn's disease by transmission disequilibrium testing in 883 families with offspring affected by IBD and in a combined case‐control analysis of these three cohorts (IBD, p=6.62E‐19).
IL‐23 is a pivotal cytokine in the differentiation of T helper cells, especially their differentiation into Th17 T cells.4 Although the Th17 T cell subset has been shown to mediate chronic and autoimmune inflammatory conditions in animal models, clear evidence exists for the central role of IL‐23 in the development of intestinal disease.5,6,7
Both Crohn's disease and ulcerative colitis commonly first present during childhood and adolescence and are associated with high disease‐related and treatment‐related morbidity in these young patients.8 Disease incidence is high in our population, and in others in Northern Europe in whom the NOD2/CARD15 contribution is small. These considerations, together with well established epidemiological data suggesting that early‐onset disease has a strong genetic basis, provide a clear scientific rationale for performing molecular studies in this group.
Our aim was to assess the contribution of the Arg381Gln variant (rs11209026) of IL23R in determining susceptibility and phenotype in childhood onset IBD in Scotland. We also sought to investigate the interaction between carriage of any of the three common NOD2/CARD15 variants and carriage of this IL23R variant in determining susceptibility to Crohn's disease.
A total of 1294 subjects comprising 358 with IBD aged <17 years at diagnosis (table 11),), 594 parents and 342 controls were genotyped for rs11209026 G/A using TaqMan (7900HT sequence detection system; Applied Biosystems, Foster City, California, USA). Allelic and genotype frequency comparisons between cases and controls using χ2 and transmission disequilibrium testing were applied to assess the association of IL23R rs11209026 with IBD. The three common NOD2/CARD15 variants were genotyped as previously described.10
In cases and controls, rs11209026 was in Hardy‐Weinberg equilibrium. The allelic frequency of rs11209026*A differed significantly between IBD/Crohn's disease cases and controls (2.9%/3.0% vs 5.5%, p=0.01, OR 0.51 (95% CI 0.30 to 0.88) and p=0.04, OR 0.53 (95% CI 0.28 to 0.98); table 22).
The GG genotype was associated with an increased risk of IBD/Crohn's disease (p=0.01, OR 2.01 (95% CI 1.15 to 3.49) and p=0.03, OR 1.96 (95% CI 1.03 to 3.70)). Analysis by transmission disequilibrium testing showed significant overtransmission of the G allele for IBD (p=0.004) and Crohn's disease (p=0.04), with a trend towards significance in ulcerative colitis hindered by a small number of informative families with ulcerative colitis (table 33).
In Crohn's disease there was no difference (p=0.94) in allelic frequency between NOD2/CARD15 wild type and NOD2/CARD15 variant‐carrying patients. However, owing to the small numbers of cases and controls carrying this IL23R variant, our study was not adequately powered to formally assess epistasis with NOD2/CARD15. Genotype‐phenotype analysis in Crohn's disease and ulcerative colitis based on the Montreal classification did not demonstrate any significant effect of IL23R rs11209026; specifically, we were not able to show a protective effect against ileal Crohn's disease (p=0.21).9
The successful identification of IL23R as a novel IBD susceptibility gene has provided proof of principle for the applicability of genome‐wide association studies in the genetics of complex diseases. We show for the first time that IL23R variation influences susceptibility to IBD and Crohn's disease, but not phenotype, in an exclusively paediatric IBD cohort. The data complement the results of the initial North American study and the replication studies currently underway in the UK adult population.10 However, the contribution of this IL23R allele to IBD is not sufficiently strong to explain the high incidence of childhood IBD in our or other populations with a low NOD2/CARD15 contribution. Other determinants are likely to be involved in Northern Europe.11,12 Genome‐wide association scanning has already provided other candidates requiring rigorous analyses.13,14
The authors acknowledge the contribution of staff at the Medical Research Council Human Genetics Unit Edinburgh, the Wellcome Trust Clinical Research Facility Edinburgh and the referring physicians from Scottish Gastrointestinal Medicine and Surgery services. They also acknowledge the help of all patients and parents who participated in the study together with the specialist nurses, dieticians and secretaries in each of the Scottish paediatric teaching hospitals and the paediatricians, practice nurses and GPs throughout Scotland whose support was invaluable.
JVL is funded by a Research Training Fellowship from Action Medical Research, The Gay‐Ramsay‐Steel‐Maitland or Stafford Trust and the Hazel M Wood Charitable Trust; RKR was funded by a University of Edinburgh Medical Faculty Fellowship and ERN is supported by a Wellcome Trust Programme Grant (072789/Z/03/Z). Financial assistance was also provided by Schering‐Plough and the GI/Nutrition Research Fund, Child Life and Health, University of Edinburgh.
Competing interests: None.