NEW YORK (GenomeWeb) — Researchers have used a next-generation sequencing-based technique for characterizing long-range structural information to link disease-associated SNPs with functional genes and promoters.
Reporting this week in Nature Communications, researchers from the University of Manchester in the UK and their collaborators modified a previously developed NGS technique called Hi-C — which detects chromatin interactions — to piece together links between variants associated with four autoimmune disorders and their functional targets in B and T cell lines.
They found that interactions may be cell-type specific and that some disease-associated SNPs interact not with the nearest gene, but genes that are often several megabases away. There may also be some gene targets in common among autoimmune diseases.
"Our results provide new insights into complex disease genetics and change the way we view the causal genes in disease, with obvious implications for pathway analysis and identification of therapeutic targets," the authors wrote.
Looking at data from genome-wide association studies that pointed to disease-associated SNPs in intergenic regions and enhancers, the researchers wanted to see how those variants were related to functional elements. They also looked at interactions of GWAS confirmed susceptibility loci with genes and functional targets for four autoimmune disorders — rheumatoid arthritis, type 1 diabetes, psoriatic arthritis, and juvenile idiopathic arthritis — in order to link the disease-associated SNPs with disease-causing genes.
The team used the Hi-C approach, which was originally developed in 2009 by researchers from the Broad Institute and the University of Massachusetts to generate 1-megabase resolution whole-genome maps in order to study their 3D architecture.
They generated Hi-C sequencing libraries for T-cell and B-cell cell lines, then used a targeted capture approach to generate a library composed of gene regions associated with the diseases as well as a library of promoters within 500 kb of disease-associated SNPs. Then they sequenced the capture Hi-C libraries on the Illumina HiSeq 2500.
Comparing with publicly available chromatin interaction data, the researchers identified well-established interactions. They also found novel interactions between disease-associated SNPs and promoters that were far away, suggesting that the disease-associated SNP was related to a gene that was not the closest in proximity.
For instance, the researchers found evidence that SNPs associated with rheumatoid arthritis interact with a promoter of the gene AZI2, which is 640 kb away from the SNP. AZI2 is involved in nuclear factor-kappa beta signaling, a process that regulates genes involved in adaptive immunity, inflammation, and stress responses. In addition, SNPs located in intronic regions that are associated with both rheumatoid arthritis and juvenile idiopathic arthritis were found to interact with the promoter of the FOXO1 gene, which plays a role in the survival of fibroblast-like synoviocytes in rheumatoid arthritis. In addition, FOXO1 is hypermethylated in rheumatoid arthritis compared to osteoarthritis, "providing strong supporting functional evidence as to gene candidature," the authors wrote.
They also found that in some cases, the lead disease-associated SNP for each disease mapped to the same gene promoter, despite being found in different genomic locations.
For instance, SNPs associated with psoriatic arthritis within the DENND1B gene make contact with a region associated with rheumatoid arthritis that is in the PTPRC gene, which is responsible for T- and B-cell receptor signaling and is located over 1 megabase away from those SNPs.
The findings challenge "the assumption that disease-associated SNPs have to be in close linkage disequilibrium to have a disease related effect on the same gene," the authors wrote. In fact, around 80 percent of significant interactions were at distances of more than 500 kb.
The authors noted that while the results were intriguing, "detailed examination to confirm these long-range interactions is now required." In addition, further work will need to be done to functionally characterize the interactions to "determine how disease-associated SNPs influence the risk of disease."