TY - JOUR
T1 - A genome-wide association study identifies protein quantitative trait loci (pQTLs)
AU - Melzer, David
AU - Perry, John R.B.
AU - Hernandez, Dena
AU - Corsi, Anna Maria
AU - Stevens, Kara
AU - Rafferty, Ian
AU - Lauretani, Fulvio
AU - Murray, Anna
AU - Gibbs, J. Raphael
AU - Paolisso, Giuseppe
AU - Rafiq, Sajjad
AU - Simon-Sanchez, Javier
AU - Lango, Hana
AU - Scholz, Sonja
AU - Weedon, Michael N.
AU - Arepalli, Sampath
AU - Rice, Neil
AU - Washecka, Nicole
AU - Hurst, Alison
AU - Britton, Angela
AU - Henley, William
AU - Van De Leemput, Joyce
AU - Li, Rongling
AU - Newman, Anne B.
AU - Tranah, Greg
AU - Harris, Tamara
AU - Panicker, Vijay
AU - Dayan, Colin
AU - Bennett, Amanda
AU - McCarthy, Mark I.
AU - Ruokonen, Aimo
AU - Jarvelin, Marjo Riitta
AU - Guralnik, Jack
AU - Bandinelli, Stefania
AU - Frayling, Timothy M.
AU - Singleton, Andrew
AU - Ferrucci, Luigi
PY - 2008/5
Y1 - 2008/5
N2 - There is considerable evidence that human genetic variation influences gene expression. Genome-wide studies have revealed that mRNA levels are associated with genetic variation in or close to the gene coding for those mRNA transcripts - cis effects, and elsewhere in the genome - trans effects. The role of genetic variation in determining protein levels has not been systematically assessed. Using a genome-wide association approach we show that common genetic variation influences levels of clinically relevant proteins in human serum and plasma. We evaluated the role of 496,032 polymorphisms on levels of 42 proteins measured in 1200 fasting individuals from the population based InCHIANTI study. Proteins included insulin, several interleukins, adipokines, chemokines, and liver function markers that are implicated in many common diseases including metabolic, inflammatory, and infectious conditions. We identified eight Cis effects, including variants in or near the IL6R (p = 1.8×10 -57), CCL4L1 (p = 3.9×10-21), IL18 (p = 6.8×10-13), LPA (p = 4.4×10-10), GGT1 (p = 1.5×10-7), SHBG (p = 3.1×10-7), CRP (p = 6.4×10-6) and IL1RN (p = 7.3×10-6) genes, all associated with their respective protein products with effect sizes ranging from 0.19 to 0.69 standard deviations per allele. Mechanisms implicated include altered rates of cleavage of bound to unbound soluble receptor (IL6R), altered secretion rates of different sized proteins (LPA), variation in gene copy number (CCL4L1) and altered transcription (GGT1). We identified one novel trans effect that was an association between ABO blood group and tumour necrosis factor alpha (TNF-alpha) levels (p = 6.8×10-40), but this finding was not present when TNF-alpha was measured using a different assay , or in a second study, suggesting an assay-specific association. Our results show that protein levels share some of the features of the genetics of gene expression. These include the presence of strong genetic effects in cis locations. The identification of protein quantitative trait loci (pQTLs) may be a powerful complementary method of improving our understanding of disease pathways.
AB - There is considerable evidence that human genetic variation influences gene expression. Genome-wide studies have revealed that mRNA levels are associated with genetic variation in or close to the gene coding for those mRNA transcripts - cis effects, and elsewhere in the genome - trans effects. The role of genetic variation in determining protein levels has not been systematically assessed. Using a genome-wide association approach we show that common genetic variation influences levels of clinically relevant proteins in human serum and plasma. We evaluated the role of 496,032 polymorphisms on levels of 42 proteins measured in 1200 fasting individuals from the population based InCHIANTI study. Proteins included insulin, several interleukins, adipokines, chemokines, and liver function markers that are implicated in many common diseases including metabolic, inflammatory, and infectious conditions. We identified eight Cis effects, including variants in or near the IL6R (p = 1.8×10 -57), CCL4L1 (p = 3.9×10-21), IL18 (p = 6.8×10-13), LPA (p = 4.4×10-10), GGT1 (p = 1.5×10-7), SHBG (p = 3.1×10-7), CRP (p = 6.4×10-6) and IL1RN (p = 7.3×10-6) genes, all associated with their respective protein products with effect sizes ranging from 0.19 to 0.69 standard deviations per allele. Mechanisms implicated include altered rates of cleavage of bound to unbound soluble receptor (IL6R), altered secretion rates of different sized proteins (LPA), variation in gene copy number (CCL4L1) and altered transcription (GGT1). We identified one novel trans effect that was an association between ABO blood group and tumour necrosis factor alpha (TNF-alpha) levels (p = 6.8×10-40), but this finding was not present when TNF-alpha was measured using a different assay , or in a second study, suggesting an assay-specific association. Our results show that protein levels share some of the features of the genetics of gene expression. These include the presence of strong genetic effects in cis locations. The identification of protein quantitative trait loci (pQTLs) may be a powerful complementary method of improving our understanding of disease pathways.
UR - https://www.scopus.com/pages/publications/44949201115
U2 - 10.1371/journal.pgen.1000072
DO - 10.1371/journal.pgen.1000072
M3 - Article
C2 - 18464913
AN - SCOPUS:44949201115
SN - 1553-7390
VL - 4
JO - PLoS Genetics
JF - PLoS Genetics
IS - 5
M1 - e1000072
ER -
