GWAS identifies Dapp1 as a determinant of air pollution-induced airway hyperreactivity !!

This topic obviously fits the gene-environment theme of these GEITP pages. The environmental signal is air pollutants, especially particulate matter with diameter less than 2.5 μM (PM2.5) — which has been well studied as an ambient air pollutant linked to diminished reduced lung development in children and increased risk of asthma. Genes in the host genome respond to this signal, producing via genetic networks a response of airway hyperreactivity, which of course can vary depending on interindividual differences in genetic susceptibility. This publication is a great disappointment, as detailed below.

Estimates for heritability of asthma range between 35% and 95% — suggesting a strong genetic component to risk. Genome-wide association studies (GWAS) in humans have identified ~200 loci for asthma and related pulmonary phenotypes; interestingly, these genes — along with bioinformatic analyses — provide compelling genetic evidence for a strong inflammatory component to asthma susceptibility. Despite the large numbers of loci identified to date, the risk alleles, most of which are common in the population, still only explain a small fraction (~7%) of asthma’s overall heritability. This observation implies either the existence of additional variants with smaller effect-sizes, rare susceptibility alleles, and/or higher-order interactions between genes and environmental (GxE) factors.

Authors state that “interactions between genetic risk factors and environmental triggers” is compounded by inherent difficulties of carrying out GxE interaction studies in humans (e.g. accurate exposure assessments, adequately powered sample sizes in which genetic, phenotypic, and exposure data are all available, and the heterogeneous nature of asthma itself pose significant practical and technical hurdles that have yet to be overcome). To address these challenges, authors [see attached article] used the Hybrid Mouse Diversity Panel (HMDP) to elucidate the genetic architecture of asthma-related phenotypes in mice and identify loci associated with airway hyperreactivity under control exposure conditions and in response to diesel exhaust particles (DEP), as a model traffic-related air pollutant. In the absence of exposure, authors identified “two loci on chromosomes (Chr) 2 and 19” for airway hyperreactivity. The interleukin-33 gene (Il33) on Chr 19 is syntenic to the IL33 locus in humans (one of the ~200 asthma-related genes in clinical GWAS).

In response to DEP exposure, authors mapped airway hyperreactivity to a region on chromosome 3 and used a genetically modified mouse model to functionally demonstrate that Dapp1 (dual adaptor of phosphotyrosine and 3-phosphoinositides-1) is one of the genes underlying the GxE association at this locus. Authors conclude (not surprisingly) that “collectively, our results support the concept that some of the genetic determinants for asthma-related phenotypes may be shared between mice and humans, as well as the existence of GxE interactions in mice that modulate lung function in response to air pollution exposures relevant to humans.”

These GEITP pages are astounded that this PLoS Genet paper is so poorly-written and poorly-reviewed. First, there is no mention in the title that this GWAS was performed in mice, not humans. Second, the authors abbreviated “airway hyperreactivity” as “AHR” — which is not only a well-known gene name for aryl hydrocarbon receptor, but presumably AHR function (regulation of polycyclic aromatic hydrocarbon-inducible genes) would likely be among the ~200 genes associated with risk of PM2.5-induced lung disease in humans and mice. Third, “101 strains from the HMDP” were studied, but there was no mention (in Summary/Abstract) of the total number of mice used in the GWAS; at one point in the text it was mentioned they studied “4-8 mice per strain” (thus, 101 x 4 = 404; 101 x 8 = 808, so the total number in the GWAS is presumably between 400 and 800 DEP-treated and control mice). In this day and age for a GWAS, N = 400-800 is a very, very small number, compared with clinical GWAS with N’s of 100,000 and over 1 million. Fourth, P-values of 3.0e-06, 5.6e-07, and 2.5e06 (in Summary/Abstract) do not rise to the level of statistical significance in a GWAS (P-values must be 5.0e–08 or lower), but obviously these values are not sufficiently low due to the small N in this study. Fifth, (in the Summary) Chr’s 2 and 19 were emphasized for control exposure mice, then further information (after DEP exposure) was given only about Dapp1 on Chr 3 and Il33 on Chr 19. Why was the Chr 2 locus not further described? ☹


PLoS Genet Dec 2019; 15: e1008528

This entry was posted in Center for Environmental Genetics. Bookmark the permalink.