Products of aqueous chlorination of 17beta-estradiol and their estrogenic activities.

To assess the estrogenic activity potentially stemming from 17beta-estradiol (E2) in drinking water, ESI-LC-MS was used to identifythe products of its aqueous chlorination under the following conditions: 50 microg/L E2, 1.46 mg/L sodium hypochlorite, pH 7.5, 25 degrees C. Seven products, including 2,4-dichloro-17beta-estradiol, monochloroestrone, 2,4-dichloroestrone, and the four byproducts such as 4-[2-(2,6-dichloro-3-hydroxyphenyl)ethyl]-7alpha-methyloctahydroinden-5-one (product C in the text) were identified in chlorinated E2 solution. The estrogenic activities of the aqueous chlorinated E2 solution at 10, 30, 60, 120, and 180 min contact time were assessed by a yeast two-hybrid system based on the ligand-dependent interaction of two proteins, a human estrogen receptor (ER) and a coactivator. All five solutions elicited transcriptional activation induction. The maximal beta-galactosidase activities induced by the chlorinated solution at 10, 30, and 60 min were similar and slightly lower than those before chlorination, while the activities of the chlorinated solution at 120 and 180 min were about 40% of those before chlorination. Finally, 4-chloro-17beta-estradiol (4-chloro-E2) (we failed to synthesize the 2-chloroestrone (2-chloro-E1)), 2,4-dichloro-17beta-estradiol (2,4-dichloro-E2), and 2,4-dichloroestrone (2,4-dichloro-E1) were synthesized, and product C was fractionated by HPLC. It was found that 4-chloro-E2 elicited strong estrogenic activity, at almost the same level as that of estrone (EC50 = 10(2) nM), while 2,4-dichloro-E2 elicited weaker beta-galactosidase activity compared with that of 4-chloro-E2. The EC50 was ca. 10(3) nM. The maximal beta-galactosidase activity for 2,4-dichloro-E1 was lower than that of 2,4-dichloro-E2, while its EC50 was similar to that of 2,4-dichloro-E2. In addition, product C, 4-[2-(2,6-dichloro-3-hydroxyphenyl)ethyl]-7alpha-methyloctahydroinden-5-one, induced high beta-galactosidase activity at the relatively higher concentration of 3.5 x 10(5) nM. On the basis of the dose-response curve of a single byproduct of chlorinated E2, the estrogenic activity at 120 and 180 min appears to be induced mainly by 2,4-dichloro-E2 and 2,4-dichloro-E1.