The results of our study show modification of the association between endotoxin exposure and asthma by a genetic polymorphism in ACAA1, demonstrating that higher home endotoxin levels appear to confer protection against asthma only in those individuals with this SNP (rs156265). This finding is unique, and offers a potential explanation for some of the inconsistent effects of endotoxin exposure observed across multiple epidemiological studies. In other work, investigators have previously shown markedly different dose response curves in response to endotoxin, depending upon the presence or absence of a SNP in CD14 . A cross sectional study of farm children also found that a polymorphism in TLR4 determined whether or not endotoxin was associated with decreased prevalence of atopy in children . Taken together, these studies (and ours) demonstrate the importance of characterizing the underlying genetic susceptibility of individuals before attempting to determine the impact of environmental exposures on allergic disease.
The majority of investigators examining the impact of genetics on responses to environmental microbial exposures in childhood asthma have focused on polymorphisms in CD14, TLR4, and TLR2. Our candidate gene approach allowed us to scan polymorphisms in and around multiple innate immune genes involved in the endotoxin-TLR signaling pathway. By testing a broader range of polymorphisms in biologically relevant genes, we were able to capture an interaction with a polymorphism in ACAA1 that would not have been observed using the single gene approach (TLR2, TLR4, or CD14) commonly employed in gene by environment interaction studies.
While we were able to detect effect modification of endotoxin exposure by this previously unstudied ACAA1 polymorphism (rs156265) we did not observe significant main effects  or interactions for the classically studied genes (CD14, TLR4, TLR2) related to endotoxin/TLR signaling. One potential explanation for this is that other polymorphisms, such as those in and around the cell signaling molecule MYD88, may play a more important role in endotoxin mediated responses than CD14, TLR2 or TLR4. (Other studies have not queried this many genes in the pathway, and therefore may have overlooked the importance of these other signaling molecules). It is also possible that effect modification by genetic polymorphisms may vary depending upon the endotoxin levels experienced by the cohort. The majority of gene by environment studies on endotoxin exposure and allergic disease have focused on farm children, who are exposed to higher levels of endotoxin than in suburban cohorts like ours.
Although ACAA1 is not an innate immune gene involved in endotoxin TLR signaling, we genotyped the rs156265 SNP in this region because of its proximity to MYD88. MYD88 is an adaptor molecule involved in endotoxin-mediated TLR signaling in innate immune cells. The MYD88 dependent signaling cascade ultimately results in the production of IL-12, a cytokine responsible for differentiation of Th1 cells that down-regulate the asthma-promoting Th2 response .
The ACAA1 SNP that modified the association between endotoxin and asthma risk is found less than 10 kb from MYD88, in a region that may regulate MYD88 transcription. The potential regulatory effect of ACAA1 on MYD88 will require further investigation. In addition to possible regulatory effects, ACAA1 may act as part of a gene cluster. Recently, GWAS analyses have begun to identify gene clusters that may play an important role in allergic disease. For example, genetic polymorphisms within a tumor necrosis factor gene cluster on chromosome 6p (including the genes for TNF-α and lymphotoxin α) have been shown to influence asthma and asthma-related phenotypes . ACAA1 could act as part of a similar gene cluster, responsible for altering innate immune function in response to endotoxin.
The rs156265 SNP was not in linkage disequilibrium with any SNPs in the MYD88 gene, suggesting that this polymorphism is an independent factor in modifying response to endotoxin, and not simply a surrogate for a polymorphism in MYD88. The clear dose response observed for the protective effects of endotoxin exposure in those individuals with the ACAA1 SNP strengthens the evidence that this genetic polymorphism alters response to environmental endotoxin, either through regulation of MYD88 or by an alternative mechanism.
While this study clearly demonstrated a gene by environment interaction for a SNP in ACAA1 and environmental endotoxin exposure, there were some study limitations. In order to increase our power to detect gene by environment interactions, we tested only those genetic polymorphisms significantly associated with asthma or eczema (regardless of environmental endotoxin exposure) in our cohort. This approach allowed us to minimize the problem of multiple comparisons; however, we may have missed interactions between endotoxin exposure and genetic polymorphisms that were not significant as main effects in models for asthma or eczema risk. Even after combining two separate cohorts, the total number of children with complete genotype, endotoxin and health outcome data was relatively small. Although we were able to detect a significant gene by environment interaction, sample size may have reduced our power to detect additional interactions between endotoxin and other genetic polymorphisms. The use of high risk cohorts in this study may limit the generalizability of our results to populations that are not high risk. Lastly, we did not have a replication population for this study, which would have provided additional confirmation of our findings.