Monitoring estrogenic activities of waste and surface waters using an in vivo zebrafish embryonic assay: comparison with in vitro cell-based assays and determination of effect-based trigger

Abstract : Monitoring of environmental estrogens has become a major concern to assess the quality of water bodies. In that respect, the use of in vitro reporter gene assays to monitor estrogenic activity has proven relevant as it enables an integrative and quantitative assessment of ER-active contaminants. However, in vitro cell assays may not necessarily reflect the estrogenic activity observed in vivo in fish, raising the question of the relevance of in vitro measurement of estrogenic activity to predict risk for aquatic species and ecosystems.In this context, our objective was to determine 1) whether estrogenic activity measured in vitro led to an in vivo response in a fish biological model, and 2) on the basis of such an in vitro / in vivo comparison, whether there exists an in vitro threshold value above which an in vivo effect is detected/evidenced. This study reports the use of the zebrafish-based EASZY assay to assess in vivo estrogenic activity of 33 surface (SW) and waste water (WW) samples collected across Europe and that were previously well-characterized for estrogen hormones and in vitro estrogenic activity (Könemann et al., 2018). We showed that 18 out of the 33 SW and WW samples induced estrogenic responses in transgenic zebrafish embryos leading to significant and concentration-dependant dependent up-regulation of the ER-regulated cyp19a1b gene expression in the developing brain. The in vivo 17?-estradiol-equivalents (EEQs) were highly correlated with both the chemical analytical risk quotient (RQ) based on steroidal estrogen concentrations and EEQs reported from 5 different in vitro assays. Based on in vitro vs. in vivo comparison, regression analyses allowed to determine for each in vitro assay an optimal cut-off value which was defined as an effect-based trigger values (EBTs) above which in vivo responses were observed. The definition of these in vitro assay-specific EBTs values allowed to improve the sensitivity and specificity of the individual in vitro assays for predicting a risk associated with substances acting through the same mode of action within environmental mixtures. Altogether, this study provides experimental demonstration of the toxicological relevance of in vitro-based assessment of estrogenic activity and recommends the use of such in vitro/in vivo comparative approach to refine and validate EBTs for mechanism-based bioassays.