Relative contribution of DNA strand breaks and DNA adducts to the genotoxicity of benzo[a]pyrene as a pure compound and in complex mixtures

Polycyclic aromatic hydrocarbons (PAH) produced upon incomplete combustion of organic matter are suspected to be carcinogenic to humans. In the present work, we especially studied the genotoxicity of benzo[a]pyrene (B[a]P), pure or in mixtures, with emphasis placed on the contribution of oxidative stress and alkylation. A comparison was made between the extent of DNA strand breaks as determined by the Comet assay and the number of DNA adducts to the diol epoxide metabolite of B[a]P measured by HPLC-mass spectrometry. HepG2 cultured human hepatocytes were treated with either pure B[a]P or particulate matter extracted from air samples collected in an urban peri-industrial site or in a metallurgic plant. Treatment with pure B[a]P did not induce increase in Comet measurements below a concentration of 1 μM whereas adducts were observed for concentrations as low as 0.025 μM. Very different results were obtained with environmental samples. Increase in the Comet score was observed with both urban and industrial mixtures containing 0.16 μM of B[a]P, especially for samples of urban origin. Comparison with the effect of the reconstituted PAH fraction of the mixtures allowed us to conclude that the induction of strand breaks results from the action of other components of the samples. In addition, a 30% potentialization and a 90% inhibition in the level of DNA adducts with respect to exposure to 0.16 μM pure B[a]P were observed for cells exposed to industrial and urban mixtures, respectively. These results contrast with the 6-fold enhancement in the yield of BPDE adducts in cells exposed to the reconstituted PAH fraction with respect to pure BaP. Altogether, our data emphasize that (i) a combination of analytical approaches is required to assess the genotoxicity of complex mixtures and (ii) risk assessment based on additivity consideration such as toxic equivalent factors may be misleading.