NEW YORK (GenomeWeb) – A new Nature Genetics study suggests some common variants involved in regulating transcript splicing in the blood overlap with SNP sites identified through genome-wide association studies of various human traits and diseases.
Researchers from the National Institutes of Health and Harvard Medical School used microarrays to profile gene expression and SNP genotyping patterns in blood samples from thousands of individuals enrolled in the Framingham Heart Study to come up with a set of candidate splicing quantitative trait loci, including so-called cis-sQTLs that turn up near the transcripts they regulate.
By folding in data from the National Human Genome Research Institute's catalog of GWAS SNPs, the team saw hundreds of sQTLs that overlapped with variants implicated in human traits or diseases, hinting that sQTLs may play an under-recognized role in disease risk.
"[W]e found 4.5 [percent] of published GWAS SNPs with cis-sQTL but not gene-level cis-[expression QTL] evidence, suggesting that sQTL analysis provides additional insights into the functional mechanisms underlying GWAS results," senior author Christopher O'Donnell, a researcher affiliated with the National Heart, Lung, and Blood Institute and Harvard Medical School, and his co-authors wrote.
"Our findings lay the groundwork for studies to define the role of [messenger RNA] splicing in the prevention and treatment of common diseases."
The team set out to tally up sQTLs in blood samples from a large study cohort, looking at the extent to which these sites overlapped with GWAS SNPs.
To catalog variants influencing transcript isoform patterns, the researchers combined two types of data for whole blood samples from 5,257 Framingham Heart Study participants: array-based gene expression profiles and SNP patterns based on variants that were directly genotyped or imputed with 1000 Genomes Project data.
To identify sQTLs, they then compared sets of eQTL variants associated with gene-level expression and eQTLs showing associations at the exon level, but not the gene level.
Using this approach, the team narrowed in on 572,333 cis-sQTL candidates associated with 2,650 genes, including more than 258,100 unique cis-sQTLs.
The set included 1,464 of the 1,763 cis-sQTLs previously described in an RNA sequencing study of 922 individuals that was published online by a Stanford University-led group in Genome Research in late 2013.
In particular, O'Donnell and colleagues saw an over-representation of sQTLs associated with genes coding for phosphoproteins and/or proteins involved in nuclear functions, acetylation, or alternative splicing and RNA-related pathways.
Folding in information on variants falling out of past GWAS indicated that some 528 cis-sQTLs overlapped with sites implicated in 238 different diseases or phenotypes. Of those, 304 cis sQTLs — regulating an estimated 202 genes — showed genome-wide significant ties to both traits or disease and transcript splicing.
At least some of the cis-sQTLs and related genes appear to have roles in processes that could rely on tissue-specific transcript splicing in the blood, the researchers noted. Among them: a cis-sQTL variant with apparent ties to a dozen lipid, blood, or cardiovascular disease-related traits or conditions.
"The impact of these genetic associations on the development of cardiovascular disease has yet to be investigated," the study's authors emphasized. Still, they noted that "[r]esults from our whole-blood sQTL study could help identify and functionally characterize susceptibility variants for cardiovascular diseases and related risk factors."