CARD15 and HLA DRB1 Alleles Influence Susceptibility and Disease
Objectives: Crohn's disease (CD) is a chronic inflammatory disease of the gut associated with allelic variants of CARD15 and HLA-DRB1 genes. We investigated the prevalence and effects of these variants in a Canadian CD cohort.
Methods: 507 unrelated CD patients were genotyped for the three major CD-associated variants (Arg702Trp, Gly908Arg, and Leu1007fsinsC) and for thirteen HLA-DRB1 alleles.
Results: At least one CARD15 variant was present in 32.5% of the CD patients compared with 20% of controls. The prevalence of CARD15 mutation was similar in both sporadic and familial and Jewish and non-Jewish CD patients. The Gly908Arg variant was significantly higher and the Arg702Trp variant significantly lower in Jewish compared to non-Jewish patients. A positive association between the HLA-DRB1*0103 allele and CD was detected in non-Jewish, familial cases (p= 0.0002), with risk for CD increased by 6.7 fold by the presence of an HLA-DRB1*0103 allele as compared to 1.9 fold and 19 fold by a single or two CARD15 variant alleles, respectively. We show a significant association of ileal involvement with CARD15 variants (OR = 1.8; p= 0.02), HLA-DRB1*0701 (OR = 1.9; p= 0.006) and DRB1*04 (OR = 1.7; p= 0.02) alleles and demonstrate the capacity of combined CARD15 and HLA-DRB1 genotyping to predict ileal disease in CD patients. By contrast, the HLA-DRB1*0103 allele was associated with later age of diagnosis (p= 0.02) and pure colonic disease (p= 0.000013).
Conclusions: These observations confirm the influence of CARD15 and HLA-DRB1 alleles on both CD susceptibility and site of disease and identify genotyping of these variants as a potential tool for improved diagnosis and risk prediction in CD.

Crohn's disease (CD) is a chronic inflammatory bowel disease (IBD) affecting approximately 0.1-0.3% of the population in developed countries. Environmental factors such as infection are thought to play an important role in the development of CD, but epidemiologic data from twin and family studies suggest that genetic factors also contribute significantly to the etiology. This latter possibility is strongly supported by recent data from genome-wide linkage studies of independently collected IBD families which have demonstrated possible IBD susceptibility loci on chromosomal regions 1p, 3p, 3q, 4q, 5q31-33, 6p, 7q21, 12q14, 14q11-q12, 16q13, 16p12, and 19p13. These linkage data have not only confirmed the multigenic etiology of IBD, but have also provided the framework for the refinement and ultimately the positional cloning of a CD susceptibility gene mapping to the pericentromeric region of chromosome 16. Originally designated as IBD1/NOD2, this gene is now known to encode the CARD15 protein, a member of the CED4/APAF1 superfamily of apoptosis regulatory proteins expressed primarily in myelomonocytic and dendritic cells. In addition to two caspase recruitment domains (CARDs), CARD15 contains a nucleotide-binding domain and ten leucine-rich repeats, the latter of which have been shown to mediate binding to bacterial components, including lipopolysaccharides and muramyl dipeptides. Thus CARD15 has potential capacity to modulate host response to bacterial pathogens and its mutation and dysfunction in CD patients is therefore consistent with evidence implicating infectious microorganisms in the pathogenesis of IBD.

Studies of CARD15 mutations among different CD populations have revealed 25-50% of patients to carry at least one CARD15 mutation. In the majority of these studies, the analysis has been restricted to the examination of three specific mutations, including two single nucleotide polymorphisms, Arg702Trp, Gly908Arg, and a frameshift mutation, Leu1007fsinsC, which together appear to account for 81% of the CD-associated variation in CARD15. The remainder of the CARD15 contribution to CD risk is derived from a heterogeneous group of rare missense mutations. Genotyping of the three major CD-associated CARD15 variants in various CD populations has also provided opportunities to compare CARD15 mutation status with clinical phenotype. Such analyses have revealed CARD15 mutation to be particularly associated with ileal disease and suggested that risk and severity of disease are substantively increased by biallelic mutation. Thus, while the molecular mechanisms coupling CARD15 mutation to development of CD require further analysis, the available data identify roles for CARD15 in modulating risk, presentation, and severity of CD.

In addition to CARD15, the major histocompatibility complex (MHC) genes, which encode human leukocyte antigen (HLA) proteins, have also been implicated in IBD susceptibility. Relevance of these genes to IBD has been demonstrated both by association studies wherein selected HLA specialities have been shown to be correlated with UC and/or CD and also by data from genome-wide scans revealing linkage of IBD to the MHC-containing region on chromosome 6p, the so-called IBD3 locus. While the specific alleles found to be associated with IBD are not identical in all studies, an association between class II HLA molecules and IBD is strongly supported by the results of a recent meta-analysis of the published data on the HLA-IBD relationship. As has been observed in relation to the other immunological diseases, HLA alleles are thought to contribute not only to risk for IBD, but also to phenotypic variations such as extraintestinal manifestations and disease severity. Thus, allelic variants of both of the two genes currently recognized as being associated with CD (CARD15 and HLA), appear to impact on clinical presentation as well as disease susceptibility.

In investigating a cohort of IBD patients for the relevance of HLA class II alleles to disease, our group has recently demonstrated a significant positive association of the HLA DRB1*0103 allele with both UC and CD and also with the presence of pure colonic disease in CD patients. These data are consistent with our detection of suggestive linkage between IBD and the MHC-containing region on chromosome 6p (lod score = 2.3) in a cohort of 158 Canadian IBD sib-pair families and also with another report of HLA DRB1*0103 association with IBD. However, the association of the HLA DRB1*0103 allele with colonic disease in CD suggests that HLA and CARD15 alleles differentially influence site of disease in CD and that their coincident characterization may therefore provide a valuable approach to the classification and potential management of these patients. To investigate this possibility, we have now examined a larger cohort of both familial and sporadic cases of CD with respect to both CARD15 and HLA DRB1 genotypes. The results of this analysis confirm a significant association between CD and the three major CARD15 variants as well as the HLA DRB1*0103 alleles. The data also reveal a significant influence of both CARD15 and HLA DRB1 alleles on disease localization and suggest that combined CARD15 and HLA DRB1 genotypes provide markers of both risk and site of disease in CD.