RT Journal Article
SR Electronic
T1 Genetic control of the human brain proteome
JF bioRxiv
FD Cold Spring Harbor Laboratory
SP 816652
DO 10.1101/816652
A1 Chloe Robins
A1 Aliza P. Wingo
A1 Wen Fan
A1 Duc M. Duong
A1 Jacob Meigs
A1 Ekaterina S. Gerasimov
A1 Eric B. Dammer
A1 David J. Cutler
A1 Philip L. De Jager
A1 David A. Bennett
A1 James J. Lah
A1 Allan I. Levey
A1 Nicholas T Seyfried
A1 Thomas S. Wingo
YR 2019
UL http://biorxiv.org/content/early/2019/10/24/816652.abstract
AB Alteration of protein confirmation and abundance is widely believed to be the hallmark of neurodegenerative diseases. Yet relatively little is known about the genetic variation that controls protein abundance in the healthy human brain. Genetic control of protein abundance is generally thought to parallel that of RNA expression, but heretofore there has been little direct evidence to support that belief. Here, we directly assessed single nucleotide variants (SNVs) that are associated with variation in brain protein abundance in healthy humans. We performed protein quantitative trait loci (pQTL) analyses using tandem mass spectrometry-based proteomic quantification of proteins from dorsolateral prefrontal cortex (dPFC) that identified 12,691 unique proteins (7,901 after quality control) and whole genome sequencing of 144 cognitively unimpaired older participants of the Religious Order Study (ROS) and Memory and Aging Project (MAP). Linear regression was used to test whether SNVs within a 100-kb window around the protein-coding sequence were associated with protein abundance. We identified 28,211 SNVs that were significantly associated with the abundance of 864 proteins (i.e. pQTLs), and the complete results are searchable at http://brainqtl.org. Brain pQTL sites were compared to expression quantitative trait loci (eQTL) analyses performed using RNA-sequencing from the dPFC of 169 cognitively unimpaired ROS/MAP participants, of which 81 were the exact same individuals whose protein abundance was measured. We found that strong pQTLs are generally only weak eQTLs, and that the majority of strong eQTLs are not detectable pQTLs. These results suggest that the genetic control of mRNA and protein abundance may be substantially distinct and provide additional evidence that inference concerning protein abundance made from mRNA studies should be treated with caution.