Application of Percellome Toxicogenomics to Food Safety
When an item of food is found to be non-toxic by routine toxicity testing at the maximal applicable dose of 10 g kg−1, the tolerable daily intake is calculated, using a safety factor of 1/100, to be 100 mg kg−1. Thus such a study cannot guarantee the safety of a food which is consumed by the grams or even tens of grams per kg body weight. For other reasons, flavorings are also not subject to routine toxicity testing. In both cases, they normally do not exert detectable histopathological effects on experimental animals at applicable dose ranges. One strategy to cope with this problem is to detect early molecular changes that occur prior to morphological alterations. Our Percellome Toxicogenomics is designed to identify dynamic and extensive networks of genes whose time- and dose-dependent patterns of expression allows its toxic effects to be predicted.
Here, we show that estragole induces genes known to be regulated by PPAR-alpha. The program, which automatically compares over 100 chemicals in our database, identifies clofibrate and di(2-ethylhexyl) phthalate (DEHP) as the two most similar chemicals. Further analysis indicates that DEHP activates both PPAR-alpha and the constitutive androstane receptor (CAR), whereas estragole and clofibrate activate PPAR-alpha but not CAR. Our findings may help elucidate the mechanism(s) of the reported estragole hepatocarcinogenesis.
An advantage of the Percellome method is that the responses obtained in different studies can be compared directly. Judging from the magnitude of the commonly up-regulated genes, estragole appears to be as potent as clofibrate in activating PPAR-alpha signaling.