NEW YORK (GenomeWeb) – A team from the University of Cambridge, Merck Research Laboratories, Addenbrooke's Hospital, and elsewhere has published an atlas of genetic variants that appear to impact blood plasma levels of specific proteins.
"Compared to genes, proteins have been relatively understudied in human blood, even though they are the 'effectors' of human biology, are disrupted in many diseases, and are the targets of most medicines," co-senior author Adam Butterworth, a public health and primary care researcher at the University of Cambridge, said in a statement. "Novel technologies are now allowing us to start addressing this gap in our knowledge."
As part of the UK's Interval study, which involves about 50,000 participant, Butterworth and colleagues systematically quantified levels for thousands of proteins in plasma samples from 3,301 seemingly healthy, genotyped individuals. With these data, they uncovered more than 1,900 interactions between almost 800 genomic regions and nearly 1,500 proteins — collected into a resource described online today in Nature.
Bringing in expression quantitative trait locus (QTL) and protein QTL data, biological pathway clues, drug database insights, and variants identified in prior genome-wide association studies, the team subsequently searched for plasma proteins contributing to common conditions, such as inflammatory bowel disease, as well as for potential drugs for altering these pathways.
"By linking genetic factors to diseases via specific proteins, our analysis highlights potential therapeutic targets, opportunities for matching existing drugs with new disease indications, and potential safety concerns for drugs under development," Butterworth and colleagues wrote.
Moreover, the team suggested that there will likely be additional settings that may benefit from insights into the genetic and plasma proteome interactions identified in the study.
"We’ve given some examples in this study of how it might be used, but now it’s over to the research community to begin using it and finding new applications," co-first author Benjamin Sun, also a public health and primary care researcher at the University of Cambridge, said in a statement.
Using SomaLogic's SOMAscan, an aptamer-based multiplex protein assay, the researchers quantified plasma levels of 3,622 proteins in blood samples from 3,301 healthy donors. They set those proteome data alongside genetic profiles for the participants, searching for associations between plasma proteins and 10.6 million autosomal SNPs that were imputed or directly assessed using Affymetrix Axiom UK Biobank arrays.
The team's analysis uncovered 1,927 associations involving 1,478 proteins and 764 regions in the genome. Most of those associations — 89 percent — had not been described previously, the group reported.
The researchers noted that 502 of the protein-associated loci appeared to act locally, or in cis, while 228 had trans effects on plasma proteins. The remaining 34 loci appeared to have both cis and trans protein interactions.
The team validated 106 of 163 proposed protein QTLs using an Olink protein assay on samples from another 4,998 individuals, noting that the cis pQTLs appeared more apt to replicate than those involving longer-range trans interactions.
After exploring the overlap between the proposed pQTLs and expression QTLs reported in the past, the researchers incorporated information from prior GWAS to identify pQTLs coinciding or in linkage disequilibrium with common disease-associated SNPs.
For example, they noted that an inflammatory bowel disease-associated variant in the chromosome 3 gene MST1 was not only linked to lower-than-usual MST1 protein levels in the blood, but also to altered levels for another eight proteins with apparent trans interactions.
"[B]y combining our database with what we know about associations between genetic variants and disease, we are able to say a lot more about the biology of disease," co-first author James Peters, a public health and primary care researcher affiliated with the University of Cambridge and Addenbrooke's Hospital, said in a statement.
The genotype-proteome interactions also provided clues to causal protein contributors to disease as well as potential drug targets, teased out by considering pQTLs in combination with drug-targeted proteins documented in the past.