RESUMO
Chemical analyses and short-term mutagenicity bioassays have revealed the presence of genotoxic disinfection by-products in drinking water. In this study, the influence of the different steps of surface water treatment on drinking water mutagen content was evaluated. Four different samples were collected at a full-scale treatment plant: raw lake water (A), water after pre-disinfection with chlorine dioxide and coagulation (B), water after pre-disinfection, coagulation and granular activated carbon filtration (C) and tap water after post-disinfection with chlorine dioxide just before its distribution (D). Water samples, concentrated by solid phase adsorption on silica C18 columns, were tested in human leukocytes and HepG2 hepatoma cells using the comet assay and in HepG2 cells in the micronuclei test. A significant increase in DNA migration was observed in both cell types after 1 h treatment with filtered and tap water, and, to a lesser extent, chlorine dioxide pre-disinfected water. Similar findings were observed for the induction of "ghost" cells. Overloading of the carbon filter, with a consequent peak release, might explain the high genotoxicity found in water samples C and D. Cell toxicity and DNA damage increases were also detected in metabolically competent HepG2 cells treated with a lower concentration of tap water extract for a longer exposure time (24 h). None of the water extracts significantly increased micronuclei frequencies. Our monitoring approach appears to be able to detect contamination related to the different treatment stages before drinking water consumption and the results suggest the importance of improving the technologies for drinking water treatment to prevent human exposure to potential genotoxic compounds.