Advances in Celiac Disease
Advances in Celiac Disease
Coeliac disease is a polygenic and multifactorial disease with involvement of genes in the human leukocyte antigen (HLA) locus and many other genes as well as gluten and possible other environmental factors. Over the last decade, genome-wide association studies (GWAS) have identified many novel non-HLA loci in coeliac disease, and it has become clear that the HLA locus has by far the largest effect size, whereas the non-HLA loci have small effects with odd ratios typically ranging between 0.7 and 1.3. During the last 2 years, studies from Spain replicated the GWAS results, and investigated the coeliac disease association and intestinal expression of two genes (THEMIS and PTPRK) located in GWAS peak region on chromosome 6. Studies from Finland reported the association of coeliac disease with FUT2 polymorphisms coupled to nonsecretor status, and genome-wide analysis of extended pedigrees confirmed the linkage to the IL2–IL21 gene region first identified in GWAS. Coeliac disease was the first trait for which 'Immunochip' data were reported. The 'Immunochip' was designed to represent the polymorphisms of relevance to 11 autoimmune and inflammatory diseases. This analysis gave 13 new risk loci of genome-wide significance (P < 5 × 10), so the total number of known coeliac disease loci reached 40. Most of these loci contain candidate genes of immune function. For one-third of the loci, multiple independent signals were found, resulting in 57 independent coeliac disease association signals. The 39 non-HLA loci could explain nearly 14% of the genetic variance of coeliac disease, with the HLA locus accounting for another 40%. Thus, much heritability in coeliac disease still remains unexplained.
The missing heritability could relate to the contribution of rare variants not covered by the GWAS single-nucleotide polymorphism (SNP) arrays or by common variant loci that have effect sizes too small to be detected using a conventional significance threshold and the sample sizes analysed. Alternatively, it could be that the size of the missing heritability has been overestimated because of extensive gene–gene interactions. Recently, massive sequencing and genotyping for coding exons of 25 GWAS risk genes in many thousand controls and patients suffering from coeliac disease or five other autoimmune diseases revealed that the rare coding-region variants at known loci have a negligible role in autoimmune disease susceptibility. Rather, evidence has been obtained for the involvement of many common-variant loci of weak effects. Advanced statistical analysis indicated that thousands of SNPs from coeliac disease GWAS not reaching significance could explain an additional 43% of the disease risk. Thus, the current state of play suggests that continued GWAS in very large size cohorts might be rewarding.
Many of the risk loci in coeliac disease are shared with other autoimmune conditions. Comparative analysis of 'Immunochip' data from coeliac disease and five other immune-mediated conditions revealed that coeliac disease not only shares loci with ankylosing spondylitis and psoriasis, but also with rheumatoid arthritis and type 1 diabetes, and to lesser extent with inflammatory bowel disease.
Most of the SNPs associated with coeliac disease do not locate to the coding regions of genes, but rather to the regulatory DNA marked by DNase I hypersensitive sites. These SNPs seem to affect the recognition of gene enhancers by transcription factors, resulting in differences in gene expression. The transcription factors are part of the regulatory networks that are shared between coeliac disease and other autoimmune diseases.
The function of the coeliac disease susceptibility loci BACH2 and RGS1 have been studied. The transcription factor BACH2 plays a key role in the development of regulatory T cells, which could possibly explain its involvement in coeliac disease, but it also has key regulatory roles in B-cell function.
HLA is by far the strongest susceptibility locus in coeliac disease. Strong genetic and functional data pinpoint the involvement and role of DQ2.5 and DQ8. Recently, studies have emerged that underpin the involvement of DQ2.2 as well. A Dutch study of 155 coeliac disease children showed 5.8% carried HLA-DQ2.2 without the expression of DQ2.5 or DQ8. Another study demonstrated that DQ2.2-positive coeliac disease patients, not carrying DQ2.5 or DQ8, have gluten-reactive CD4 T cells in their intestinal lesions. These T cells do not respond to the common DQ2.5-restricted T-cell epitopes. Rather, they respond to distinct epitopes that are not recognized by the T cells of DQ2.5-expressing patients. Other studies reported that MHC genes apart from DQ genes may contribute to coeliac disease susceptibility.
The insight in novel coeliac disease genes provides some improvement in helping diagnose coeliac disease. Adding 57 non-HLA variants raised the area under the receiver operator characteristic curve score from 0.823 with HLA alone to 0.854.
The existence of a large number of non-HLA genes that are only partly shared by each individual patient suggests that the disease can be more heterogeneous than previously considered.
Genetics
Coeliac disease is a polygenic and multifactorial disease with involvement of genes in the human leukocyte antigen (HLA) locus and many other genes as well as gluten and possible other environmental factors. Over the last decade, genome-wide association studies (GWAS) have identified many novel non-HLA loci in coeliac disease, and it has become clear that the HLA locus has by far the largest effect size, whereas the non-HLA loci have small effects with odd ratios typically ranging between 0.7 and 1.3. During the last 2 years, studies from Spain replicated the GWAS results, and investigated the coeliac disease association and intestinal expression of two genes (THEMIS and PTPRK) located in GWAS peak region on chromosome 6. Studies from Finland reported the association of coeliac disease with FUT2 polymorphisms coupled to nonsecretor status, and genome-wide analysis of extended pedigrees confirmed the linkage to the IL2–IL21 gene region first identified in GWAS. Coeliac disease was the first trait for which 'Immunochip' data were reported. The 'Immunochip' was designed to represent the polymorphisms of relevance to 11 autoimmune and inflammatory diseases. This analysis gave 13 new risk loci of genome-wide significance (P < 5 × 10), so the total number of known coeliac disease loci reached 40. Most of these loci contain candidate genes of immune function. For one-third of the loci, multiple independent signals were found, resulting in 57 independent coeliac disease association signals. The 39 non-HLA loci could explain nearly 14% of the genetic variance of coeliac disease, with the HLA locus accounting for another 40%. Thus, much heritability in coeliac disease still remains unexplained.
The missing heritability could relate to the contribution of rare variants not covered by the GWAS single-nucleotide polymorphism (SNP) arrays or by common variant loci that have effect sizes too small to be detected using a conventional significance threshold and the sample sizes analysed. Alternatively, it could be that the size of the missing heritability has been overestimated because of extensive gene–gene interactions. Recently, massive sequencing and genotyping for coding exons of 25 GWAS risk genes in many thousand controls and patients suffering from coeliac disease or five other autoimmune diseases revealed that the rare coding-region variants at known loci have a negligible role in autoimmune disease susceptibility. Rather, evidence has been obtained for the involvement of many common-variant loci of weak effects. Advanced statistical analysis indicated that thousands of SNPs from coeliac disease GWAS not reaching significance could explain an additional 43% of the disease risk. Thus, the current state of play suggests that continued GWAS in very large size cohorts might be rewarding.
Many of the risk loci in coeliac disease are shared with other autoimmune conditions. Comparative analysis of 'Immunochip' data from coeliac disease and five other immune-mediated conditions revealed that coeliac disease not only shares loci with ankylosing spondylitis and psoriasis, but also with rheumatoid arthritis and type 1 diabetes, and to lesser extent with inflammatory bowel disease.
Most of the SNPs associated with coeliac disease do not locate to the coding regions of genes, but rather to the regulatory DNA marked by DNase I hypersensitive sites. These SNPs seem to affect the recognition of gene enhancers by transcription factors, resulting in differences in gene expression. The transcription factors are part of the regulatory networks that are shared between coeliac disease and other autoimmune diseases.
The function of the coeliac disease susceptibility loci BACH2 and RGS1 have been studied. The transcription factor BACH2 plays a key role in the development of regulatory T cells, which could possibly explain its involvement in coeliac disease, but it also has key regulatory roles in B-cell function.
HLA is by far the strongest susceptibility locus in coeliac disease. Strong genetic and functional data pinpoint the involvement and role of DQ2.5 and DQ8. Recently, studies have emerged that underpin the involvement of DQ2.2 as well. A Dutch study of 155 coeliac disease children showed 5.8% carried HLA-DQ2.2 without the expression of DQ2.5 or DQ8. Another study demonstrated that DQ2.2-positive coeliac disease patients, not carrying DQ2.5 or DQ8, have gluten-reactive CD4 T cells in their intestinal lesions. These T cells do not respond to the common DQ2.5-restricted T-cell epitopes. Rather, they respond to distinct epitopes that are not recognized by the T cells of DQ2.5-expressing patients. Other studies reported that MHC genes apart from DQ genes may contribute to coeliac disease susceptibility.
The insight in novel coeliac disease genes provides some improvement in helping diagnose coeliac disease. Adding 57 non-HLA variants raised the area under the receiver operator characteristic curve score from 0.823 with HLA alone to 0.854.
The existence of a large number of non-HLA genes that are only partly shared by each individual patient suggests that the disease can be more heterogeneous than previously considered.
