Disinfection of drinking water generates halogenated by-products, whose potential biological effects have important public health implications. Several disinfection by-products are carcinogenic when administered to experimental animals. The extrapolation of such findings to human health, however, is unclear, since these tests generally involved exposures that were several orders of magnitude greater than likely human exposures. In addition, the mechanisms underlying the carcinogenic effects are poorly understood. Carcinogenesis may involve covalent modification of DNA, either by direct chemical reaction or by modulation of endogenous processes to produce reactive electrophiles. We will test this hypothesis by the following specific aims:
Specific Aim 1. Using a battery of well-characterized and ultrasensitive endpoints, we will investigate whether disinfection by-products enhance endogenous DNA damage (i.e., DNA modification derived from reactive oxygen species).
Specific Aim 2. The formation of direct DNA adducts that derive from the disinfection by-product, and 3-chloro-4(dichloromethyl)-5-hydroxy-2(5H)-furanone (MX) will be compared with those derived from reactive oxygen species.
Specific Aim 3. Dose-response relationships of DNA modification will be determined, and those findings will be related to other mechanistic data, such as cell proliferation and to the outcomes of 2-year bioassays (tumor incidence). We propose to investigate three disinfection by-products, bromodichloromethane, dibromoacetic acid, and MX in Specific Aims 1 and 3, and MX in Specific Aim 2. A strength of this proposal will be the ability to relate DNA adducts to other mechanistic data and to tumors that occur under the same exposure conditions. Analysis of different types of DNA damage will result in a better understanding of the mechanisms underlying the toxicity and carcinogenicity of disinfection by-products. In addition, DNA adducts will provide a mechanistic basis for conducting extrapolations across doses and between species, which will improve the prediction of human health risks from animal studies.
