The theme of these GEITP pages (in a broad sense) is gene-environment interactions, and how they can be understood, and further elucidated, by using genome-wide association studies (GWAS). This article [see attached] reminds me of a story: [The drunk fellow was searching for his car-keys under a street lamp at night, when a passerby asked if he was sure he’d lost his keys there; he said ‘No, but it’s easier to look here, where there’s more light’.] In other words, GWAS are designed to look for single-nucleotide variants (SNVs) throughout the genome that are statistically significantly associated with a phenotype (i.e. a trait such as obesity, type-2 diabetes, response to a drug, response to an environmental toxicant or mixture of toxicants). The (germ line) DNA used in GWAS is typically derived from white blood cells or cells from skin or sputum. In contrast, this study [see attached article] searches for gene expression uniqueness — specifically occurring in adipose tissue. 😊
Excess adipose tissue, especially around the hips and waist, is associated with increased cardiometabolic risk and mortality. This subcutaneous adipose tissue expands to store additional lipids, and serves as a buffering system for lipid energy balance — especially for fatty acids — providing a protective role in metabolic risk. Interestingly, expansion of adipocyte size (i.e. fat cells getting larger), rather than formation of new adipocytes (new fat cells), has also been linked to insulin resistance. Identification of genetic variants associated with gene expression quantitative trait loci (eQTLs) in relevant tissues has proven useful to correlate non-coding GWAS variants (SNVs that do not participate in transcription to RNA or translation into protein) to plausible candidate genes that may influence complex traits. Whereas 94% of eQTLs are shared across at least two tissues, some eQTLs are specific to one tissue (or a subset of tissues) — necessitating scientists to study tissues that potentially contribute to GWAS traits, in order to identify credible candidate genes.
Recently, eQTL studies have begun to identify other eQTL signals through conditional analysis (this is where one adjusts for the original SNV (primary eQTL) in the model; then one tests to see if the association between a 2nd SNV (secondary eQTL) is independent of the primary eQTL, or whether it is just because they are ‘geographically correlated’ via linkage disequilibrium (LD), i.e. the non-random association of alleles at different loci along the same chromosome, which occurs in any given population). In addition, eQTL studies examine the more commonly reported “primary eQTLs”. These other conditionally distinct secondary eQTL signals are widespread, and located more distal than primary signals from the transcription start-sites (TSS) of their associated genes. These other eQTL signals have also been shown to co-localize with GWAS loci, suggesting they can detect additional candidate genes.
Authors [see attached article] used subcutaneous adipose tissue RNA-seq data from 434 Finnish men [I can’t figure out why researchers always tend to ‘pick on Finnish men’ for these studies? ☹] — from the METSIM study to identify 9,687 primary and 2,785 secondary eQTLs (within 1 million base pairs of the transcription start-site, and false discovery rate less than 1%). Compared to primary eQTL signals, secondary eQTL signals were located further from transcription start-sites, had smaller effect-sizes, and were less enriched in adipose tissue regulatory elements, compared to primary eQTL signals. Among 2,843 cardiometabolic GWAS signals, 262 co-localized by LD; conditional analysis confirmed 318 eQTL transcripts found as primary eQTLs and conditionally distinct secondary eQTLs.
Among the cardiometabolic traits examined for adipose tissue eQTL co-localizations, waist-hip ratio and circulating lipid levels (two quantitative traits) contained the highest percentage of co-localized eQTLs (15% and 14%, respectively). Among alleles associated with increased cardiometabolic GWAS risk, approximately half (~53%) were associated with decreased gene expression levels. Mediation analysis (this seeks to identify and explain the mechanism, or process, that underlies an observed relationship between an independent variable and a dependent variable — via inclusion of a third hypothetical variable, known as a ‘mediator variable’) of co-localized genes and cardiometabolic traits within the 434 individuals — provided further evidence that gene expression influences variant-trait associations. In summary, these results have identified hundreds of candidate genes that appear to act in adipose tissue to influence cardiometabolic traits.
Am J Hum Genet Oct 2019; 105: 773-787