The largest-ever genetic study uncovers 15 new genetic risk factors for celiac disease, opening the door to earlier detection and personalized treatment strategies.
Study: Population screening of adults identifies novel genetic variants associated with celiac disease. Image credit: Marian Weyo/Shutterstock.com
In a recent study published in the journal Scientific Reports, researchers conducted the largest genome-wide association study (GWAS) to unravel the genetic loci associated with celiac disease (CeD) incidence and progression.
Background
CeD is an autoimmune disorder triggered by gluten consumption, a protein commonly found in wheat, barley, and rye. Characterized by the inflammation of patients' small intestine lining, the disease forces patients to adhere to a strict gluten-free diet or risk diarrhea, bloating, fatigue, anemia, and osteoporosis.
Decades of genetic research have established the role of the human leukocyte antigen (HLA) in CeD incidence. Specifically, GWASs have demonstrated that HLA-DQ allotypes DQ2.2, DQ2.5, and DQ8 are necessary for disease development.
Unfortunately, HLA-DQ allotypes are extremely common, with approximately 55% of the human population possessing them. However, CeD incidence is comparatively rare, and only around 3% of HLA-DQ carriers develop CeD. This suggests that while HLA-DQs have a role in the development of CeD, they are insufficient to explain the risk alone.
Researchers estimate that the 42 previously identified CeD-associated genetic loci explain only 48% of the disease's heritability, highlighting the need for further risk assessments. Previous GWASs suffer from insufficient sample sizes, sampling bias (only diagnosed patients screened), and underrepresentation of non-coding genetic variants. Furthermore, a growing body of evidence reveals global increases in CeD incidence, many of which remain undiagnosed.
About the study
The study leveraged almost 25 million genetic variants from 52,342 Norwegian adults with diagnosed CeD patients, undiagnosed individuals with CeD, and healthy participants, making it the most extensive and in-depth GWAS on this topic.
Notably, the study included diagnosed and previously undiagnosed cases of CeD, reducing information and selection biases present in previous research. All adults in the region were invited to participate, achieving a comparatively high 54% response rate. However, the study was limited to individuals of European ancestry and did not include children or young adults.
Study data were obtained from adults over 20 who participated in the Trøndelag Health Study round four (HUNT4) between 2017 and 2019. Data collection comprised detailed questionnaires (for sociodemographic and medical history data), clinical measurements, and blood sample collections.
In-depth GWAS sequencing spanned coding and non-coding regions of DNA and incorporated population-wide representation to comprehensively understand the disease's genetic underpinnings.
A sequential cascade of antibody screening (TG2, IgA, IgG), endoscopic examinations, serological testing, and Marsh grading was used to classify participants into CeD cases (Marsh grade 3) and potential cases (Marsh grade 0-2). Cases were further classified into 'previously diagnosed' and 'novel cases.'
Participant blood genotyping was achieved using Illumina HumanCoreExome arrays augmented with Positional Burrows Wheeler Transform (PBWT) imputation, allowing for more efficient haplotype phasing across coding and non-coding DNA regions. Logistic mixed models implemented in the SAIGE tool were used to conduct GWAS analyses, revealing single-nucleotide polymorphisms (SNPs) associated with CeD risk. Models were adjusted for covariates such as age, sex, and genetic variants.
The study identified 15 previously unknown genetic associations across 12 loci, 11 of which were entirely novel to science. The most significant finding was the LINC01019 gene at the 5p15.33 locus, a gene previously implicated in rheumatoid arthritis, which suggests shared autoimmune pathways.
These associations were identified in non-coding genome regions that had previously been unexplored in CeD investigations. Together, these findings significantly increase the scientific understanding of CeD's genetic architecture, emphasizing the importance of comprehensive population screening in exploratory disease research.
However, the authors emphasize that further studies are warranted to replicate these findings and clarify their biological impact.
Study findings
Of the 103,800 Norwegian adults invited to participate in the study, 54% provided complete data and were included in subsequent analyses. Serological assays identified 2.1% seropositive cases and 1.5% demonstrating CeD.
In addition to validating the 41 previously reported non-HLA loci, association testing identified 15 previously unknown SNPs across 12 loci, 11 of these loci never linked to CeD. These associations were revealed to have genome-wide significance (P < 5×10⁻⁸), with the 5p15.33 locus, long non-coding RNA gene LINC01019 exhibiting the strongest association with CeD.
Functional mapping (SNP2GENE analyses) highlighted three risk loci, 2q35, 5p15.33, 6p21.2, and nine genes with a high probability of CeD-modulating biological impacts. Expression analyses (GENE2FUNC) explained the expression of these genes across circulatory tissues, neural tissues, whole blood, digestive tissues, renal glands, and lymphatic ducts, totaling 54 unique tissues.
Notably, the 5p15.33 locus has been previously linked to rheumatoid arthritis, suggesting a shared autoimmune susceptibility region and the potential for a multi-disease-targeting intervention.
Association tests also revealed three HLA loci significantly associated with CeD, 6p22.1 is novel to science. However, imputation coverage of the HLA region was limited, so results should be interpreted cautiously.
Finally, the estimated genome-wide heritability (h2) for confirmed CeD patients was 23%, which is lower than previous estimates. The h2 increased from 7% for previously diagnosed cases to 11% for new cases detected through screening, indicating improved detection over present-day clinical diagnostic methodologies.
Despite these advances, the study acknowledges limitations, including potential difficulty detecting rare variants due to the relatively small number of CeD cases, restriction to individuals of European ancestry, and technical challenges of analyzing the highly complex HLA genomic region.
Conclusions
This large-scale GWAS is the most comprehensive investigation of CeD genetics to date. It revealed 15 previously unknown genetic variants, mainly in non-coding DNA regions, that contribute to disease risk.
Identifying the LINC01019 gene at the 5p15.33 locus is a promising lead for future research and potential clinical applications, including improved risk prediction and earlier diagnosis.
While the study overcomes key limitations of earlier research, such as sampling bias and limited genetic coverage, it is restricted to an adult, predominantly European population. It may not detect rare variants in smaller subgroups.
To build on these findings, future research should focus on replicating results in diverse and younger populations, applying more refined methods for complex genomic regions, and investigating the functional impact of the newly discovered variants to clarify their clinical significance.
Download your PDF copy now!
Breakthrough study finds genes behind calcium pyrophosphate deposition disease