Gene-environment interaction analysis in atopic eczema: evidence from large population datasets and modelling in vitro

Read the full article See related articles

Listed in

This article is not in any list yet, why not save it to one of your lists.
Log in to save this article

Abstract

Background

Environmental factors play a role in the pathogenesis of complex traits including atopic eczema (AE) and a greater understanding of gene-environment interactions (G*E) is needed to define pathomechanisms for disease prevention. We analysed data from 16 European studies to test for interaction between the 24 most significant AE-associated loci identified from genome-wide association studies and 18 early-life environmental factors. We tested for replication using a further 10 studies and in vitro modelling to independently assess findings.

Results

The discovery analysis showed suggestive evidence for interaction (p<0.05) between 7 environmental factors (antibiotic use, cat ownership, dog ownership, breastfeeding, elder sibling, smoking and washing practices) and at least one established variant for AE, 14 interactions in total (maxN=25,339). In replication analysis (maxN=252,040) dog exposure*rs10214237 (on chromosome 5p13.2 near IL7R ) was nominally significant (OR interaction =0.91 [0.83-0.99] P=0.025), with a risk effect of the T allele observed only in those not exposed to dogs. A similar interaction with rs10214237 was observed for siblings in the discovery analysis (OR interaction =0.84[0.75-0.94] P=0.003), but replication analysis was under-powered OR interaction =1.09[0.82-1.46]). Rs10214237 homozygous risk genotype is associated with lower IL-7R expression in human keratinocytes, and dog exposure modelled in vitro showed a differential response according to rs10214237 genotype.

Conclusions

Interaction analysis and functional assessment provide evidence that early-life dog exposure may modify the genetic effect of rs10214237 on AE via IL7R , supporting observational epidemiology showing a protective effect for dog ownership. The lack of evidence for other G*E studied here implies that only weak effects are likely to occur.

Article activity feed