Bioactivation and genotoxicity of the herbal constituents safrole, estragole and methyleugenol

The herbal constituents safrole, estragole, and methyleugenol, belonging to the chemical class of the alkenylbenzenes, are genotoxic and carcinogenic compounds. The genotoxicity of these alkenylbenzenes proceeds via electrophilic metabolites generated bycytochromeP450 enzymes (P450) andsulfotransferases(SULT). Carcinogenicity of these compounds was demonstrated in animal experiments using relatively high doses of single compounds. Human exposure to these compounds is much lower than the doses used in the animal experiments and humans are exposed toalkenylbenzenesin a complex food matrix. The first aim of this thesis was to identify the human P450 enzymes that are responsible for thebioactivationof thealkenylbenzenesinto their proximate carcinogenic 1¢-hydroxymetabolites. Several in vitro studies using recombinant P450 enzymes and human livermicrosomeswere undertaken to identify the main enzymes involved in the 1¢-hydroxylation of thealkenylbenzenesand to determine their kinetics. These studies showed that at low substrate concentrations, P450 1A2 is the major enzyme in thebioactivationofmethyleugenol, P450 1A2 and P450 2A6 are the main enzymes in thebioactivationofestragole, and P450 2A6 is the main enzyme in thebioactivationofsafrole. The second objective of this thesis was to study the influence of other herbal constituents on thebioactivationand thegenotoxicityof herb-basedalkenylbenzenes. An on-line high-performance liquid chromatography detection system was developed for the detection of P450 1A2 inhibitors in herbal extracts. The presence of P450 1A2 inhibitors in basil,aherb that containsmethyleugenolandestragole, was demonstrated using this on-line system. In addition to these P450 1A2 inhibitors, also thealkenylbenzenesthemselves may act as inhibitors competing for the active site of P450 1A2 (estragoleandmethyleugenol) or P450 2A6 (estragoleandsafrole). Furthermore it was demonstrated that basil extract is able to strongly inhibitsulfationand subsequent DNA adduct formation of 1¢-hydroxyestragolein incubations with rat and human S9 homogenates and in the humanhepatomaHepG2 cell line. These in vitro results suggest that P450- and SULT-catalyzedbioactivationofmethyleugenoland/orestragoleand subsequent adverse effects may be lower in a matrix of other herbal components than what would be expected on the basis of experiments using single compounds. In vivo experiments have to point out whether the protective effects that are found in these in vitro studies can be extrapolated to the in vivo situation. It may turn out that rodent carcinogenicity data onestragoleandmethyleugenolconsiderably overestimate the risks posed when humans are exposed to these compounds inaherbal matrix.