A number of models exist to study DNT endpoints, addressing cellular key events and neurodevelopmental outcomes. However, the ability of these models to detect ED-induced DNT is largely unknown. In ENDpoiNTs, we have assessed the responsiveness of well-established DNT key events to endocrine interference. Potentially relevant ED pathways were selected using in silico modelling to predict targets of EDCs associated with neurodevelopmental impairments in humans. Using receptor agonists and antagonists, their link to cellular key events was established in a number of in vitro models. This screening effort showed that all investigated key events are dependent on one or several of the selected hormonal receptors. Thereby, novel roles for hormonal receptors in neurodevelopmental key events were uncovered. For the most sensitive and robust assays, positive and negative controls as well as Standard Operating Procedures (SOPs) were established, and the response to EDCs associated with DNT outcomes in humans have been tested. Results show that some of these EDCs indeed affect cellular endpoints in low doses. To establish human relevance of the test methods, doses producing an adverse effect in some of the models were compared with human exposure data for both single EDCs and for their mixture. This comparison showed that the newly developed in vitro models are more sensitive as they would predict risk for neurodevelopmental adversities at levels found in the general population. For six in vitro assays addressing ED-induced DNT, inter-laboratory validation has been initiated. Additionally, brain organoids have been established for ED-DNT testing and a strategy developed to use such 3D models to unravel the contribution of genetic variability in response to ED-induced DNT.
To link the molecular and cellular key events addressed in the in vitro models to adverse outcomes in whole organisms, several in vivo models were employed. Rats were developmentally exposed to six EDCs. Behavioural analyses in the adults showed sex-specific effects on learning and memory for some of these EDCs. Molecular analyses, including transcriptomics, epigenomics, metabolomics, lipidomics, and steroidomics identified a number of gene networks, steroids, neurotransmitters and lipids dysregulated by the six EDCs with compound and sex-specific effects. Correlation analyses revealed sets of genes correlated with the behavioural outcomes and transcriptional data sets from in vitro models, which provided a link between the cellular and the organismal outcomes. Also in Xenopus laevis, a systematic correlation between early transcriptional dysregulation and adult adverse outcomes using the same EDCs as well as hormone agonists and antagonists has been conducted. Additionally, a zebrafish testing approach was developed, from molecular (steroid hormones and neurotransmitters) to morphological (thyroid hormone follicle activity) and behavioral level (swimming distance), to confirm ED linked DNT effects of EDCs. In humans, several epigenetic changes partly mediating associations between EDC exposure and neurodevelopmental impairment were identified, which could be developed into biomarkers for ED-DNT.
For the management and handling of the experimental data, a standardised database platform was developed. It includes data capture, curation procedures, and statistical analysis approaches. All ENDpoiNTs data is contained in the database and will be made available as an invaluable resource for the community. Furthermore, 12 first-tier screening high confidence Quantitative Structure Activity Relationships (QSARs) were developed for predicting the agonistic and antagonistic modes of action of the receptors captured by the developed in vitro models. This is the basis for in silico models identifying chemicals that can induce DNT via these interactions. To extrapolate in vitro to in vivo concentrations of EDCs, a physiology-based toxicokinetic (PBTK) approach has been developed, predicting fetal brain concentrations. Furthermore, data produced in ENDpoiNTs have been integrated into existing Adverse Outcome Pathways (AOPs) and AOP networks. 5 AOPs have been submitted to AOP Wiki.
To ensure the uptake of the findings and developed tools into the regulatory context, ENDpoiNTs was actively engaging with key stakeholders and communication channels with relevant working groups of the OECD have been strengthened to enable continuous discussion of the readiness of the developed assays for validation and regulatory implementation. Finally, to disseminate the project’s results to different target groups, ENDpoiNTs has produced project flyers and 3 animated short films, and will maintain its webpage (
https://endpoints.eu/(s’ouvre dans une nouvelle fenêtre)) as well as continuously publishing its results in scientific journals with open access beyond the project’s end.
