Detection of endocrine disruptors: from simple assays to whole genome scanning

Abstract

Endocrine disruptors frequently bear little structural relationship to the hormone whose actions they disrupt. Consequently, the threat of an uninvestigated chemical cannot easily be assessed. Here three different approaches to assessment are discussed. The first presumes an endocrine-disrupting property, following which a cell model capable of responding to such a hormone is used. Although simple and cheap, it provides limited data. A second approach involves multiple assays to detect multiple hormones. Increasing the amount of data increased the difficulty in assessing the significance of results. To meet this problem, cluster analysis based on a simple mathematical matrix was adopted. The matrix was used to determine (i) a limited number of assays to identify a maximum number of endocrine disruptors and (ii) the chemicals with the most wide-ranging effects. A third approach was a whole genome expression analysis based on expression of mRNAs in human TE671 medulloblastoma cells. Expression of individual mRNAs was assessed using the Affymetrix GeneChip((R)) Human Genome U133 Plus 2.0 chip. The significance of differential expressed genes was assessed based on gene ontology and pathways analyses using DAVID and GenMaPP programs. The results illustrated the very wide-ranging effects of these chemicals across the genome.