Coding and regulatory variants affect serum protein levels and common disease

Abstract

Circulating proteins are prognostic for human outcomes including cancer, heart failure, brain trauma and brain amyloid plaque burden. A deep serum proteome survey recently revealed close associations of serum protein networks and common diseases. The present study reveals unprecedented number of individual serum proteins that overlap genetic signatures of diseases emanating from different tissues of the body. Here, 55,932 low-frequency and common exome-array variants were compared with 4782 protein measurements in the serum of 5457 individuals of the deeply annotated AGES Reykjavik cohort. At a Bonferroni adjusted P-value threshold < 2.1610 -10 , 5553 variants affecting levels of 1931 serum proteins were detected. These associated variants overlapped genetic loci for hundreds of complex disease traits, emphasizing the emerging role for serum proteins as biomarkers of and potential causative agents of multiple diseases.Large-scale genome-wide association studies (GWASs) have expanded our knowledge of the genetic basis of complex disease. As of 2018, approximately 5687 GWASs have been published revealing 71,673 DNA variants to phenotype associations 1 . More recently, exomewide genotyping arrays have linked rare and common variants to many complex traits. For example, 444 independent risk variants were identified for lipoprotein fractions across 250 genes 2 . Despite the overall success of GWAS, the common lead SNPs rarely point directly to a clear causative polymorphism, making determination of the underlying disease mechanism difficult 3-6 . Regulatory variants affecting mRNA and/or protein levels and structural variants like missense mutations can point directly to the causal candidate. Alteration of the amino acid sequence may affect protein activity and/or influence transcription, translation, stability, processing and secretion of the protein in question 7-9 . Thus, by integrating intermediate traits

Large Bi-Ethnic Study of Plasma Proteome Leads to Comprehensive Mapping of cis-pQTL and Models for Proteome-wide Association Studies

“…Availability of a large number of proteins or protein complexes in the SOMAScan platform as well as large sample size that almost tripled the number of cis-pGenes, i.e. genes that with at least one identified cis-pQTL, compared to studies conducted in the past 16,17 , and we successfully replicated over 92% (761/820) of the previously identified cis-pGenes. Further, analysis of data from both European American (EA) and African American (AA) samples allowed us to understand consistent genetic regulation of proteins across ethnic groups and better characterize sets of causal pQTLs through bi-ethnic fine-mapping analysis.…”

“…1b). Compared to plasma pQTL studies conducted in the past in European ancestry sample 16,17 , we tripled the number of cis-pGenes, i.e. genes that with at least one identified cis-pQTL, 17,18 and we successfully replicated over 92% (761/820) of previously identified ones (Supplementary Table 4).…”

“…The first was performed in the INTERVAL study with UK blood donors 16 . The other was performed in the AGES-RS cohort 17 . From their publicly available GWAS results of plasma proteins or protein complexes, we filtered out only cis-associations and thus cis-pGenes.…”

A proteome-wide genetic investigation identifies several SARS-CoV-2-exploited host targets of clinical relevance

“…The ABO signal (rs8176719-TC) overlaps several genes. We explored proteome-wide associations in three separate proteomic datasets 15,18,38 . Aside from the ABO protein, the ABO SNP rs8176719-TC showed the strongest association (Sun et al: p = 6.03 × 10 −258 , Emilsson et al: p = 1.00 × 10 −307 , Suhre et al: p = 1.27 × 10 −75 ) with higher plasma concentrations of soluble CD209 in all three datasets (associations from two datasets illustrated in Figure 3 , and associations from all three datasets tabulated in Table S6 ).…”

Integrating Genetics and the Plasma Proteome to Predict the Risk of Type 2 Diabetes

“…A list of pGWAS studies was originally published by Suhre et al [74] and updated in their online web resource. Two of the most comprehensive studies used the SomaLogic platform and have identified 5553 exome array variants affecting 1931 proteins in 5457 individuals [83] and 1927 genomewide variants affecting 1478 proteins in 3301 individuals [82•]. We recently expanded the genetic discovery for a subset of 179 proteins from the SomaLogic platform, including 45 proteins with no previously described pQTLs, in 10,708 individuals [85].…”