Docking and QSAR studies of an indirect estrogenic effect of hydroxylated PCBs

Abstract

A quantitative structure-activity relationship (QSAR) of hydroxylated polychlorinated biphenyls (OH-PCBs) potency to inhibit human estrogen sulfotransferase (hEST), was modeled using multivariate methods and calculated semi-empirical descriptors. The QSAR model included 10 selected OH-PCBs, recently reported experimental inhibition potencies of hEST and 17 physico- chemical parameters. The cross-validated explained variance of 0.71 indicates a high predictive capacity of the model. The most important parameters were sub-molecular electronic parameters, such as partial atomic charges and nucleophilic electron densities. The natural substrate estradiol and OH-PCBs have been suggested to inhibit hEST by non-competitive binding to a not yet identified allosteric site. Potential allosteric binding sites in the crystal structure of mouse estrogen sulfotransferase (mEST) and in a homology model of hEST were investigated and discussed using the programs CAST and GRAMM. In general, the most abundant docking sites using GRAMM were in agreement with pockets defined by CAST. These preliminary results should be verified and more detailed studies undertaken when the active dimer of hEST is available. The results indicate that other similar pollutants may well interfere with the sulfonation of estradiol.