Inferring toxicant susceptibility in global populations from gene-environment interactions involving the Aryl Hydrocarbon Receptor

Gene-environment interaction (GxE) studies comprise a very small part of the genetics or environmental epidemiology literature, and most existing studies are in populations of European ancestry. We made predictions about GxE in global populations by combining existing GxE studies with genetic data from the 1000 Genomes Project (1kGP), which captured genetic variation in diverse populations worldwide. We modeled susceptibility of 1kGP populations to 2, 3, 7, 8-tetrachlorodibenzo-p-dioxin (TCDD) exposures based on variation in the aryl hydrocarbon receptor (AHR) gene. The premise of our approach is that the risk variants involved in GxE are shared across global super-populations but vary in frequency by population. We built our model upon GxE estimates from a study in Seveso, Italy, where offspring birthweight was influenced by AHR variants after TCDD exposure. Our simulations predicted that GxE would result in different outcomes across global populations. This framework can be extended to model population susceptibility to a broad range of toxicants that impact public health, including common AHR ligands like the polycyclic aromatic hydrocarbons found in cigarette smoke and diesel exhaust.