The Validated Embryonic Stem Cell Test (EST) and its Applicability Domain

The validated embryonic stem cell test (EST) is a system for testing developmental toxicity ofchemicals in vitro. In a stand alone approach, it comprises three assays: i) the disturbance ofmouse embryonic stem cell line D3 differentiation towards beating cardiomyocytes, ii) thechemical-mediated cytotoxicity to D3 cells, and iii) to the mouse embryonic fibroblast cell line3T3. The half-maximal effective concentrations determined in the three assays are then usedin a prediction model that computes a predicted embryotoxicity class for the individual testsubstance. Since its validation by the European Centre for the Validation of Alternative Methods(ECVAM), the original EST has been used in a number of studies that both underscoredits potential, but also demonstrated several limitations. In the validation study, the ESTshowed an overall accuracy of 78% with 20 substances. Moreover, by using seven congeners ofvalproic acid, a teratogenic anticonvulsant, the EST nicely differentiated the structurallyrelated molecules and ranked them in the same order and in full accordance with dataobtained in the NMRI exencephaly in vivo mouse model. In addition, during the ReProTectProject, funded within the 6th Framework Programme of the European Union, two sets ofchemicals were tested. One set comprised reproductive toxicants, and all substances labelledas reprotoxicant according to the European Union Classification, Labelling and Packaging(CLP) Regulation were detected as developmentally adverse by the EST. By contrast, only15% of substances were correctly identified in another set comprising a broad array of chemicals.Reviewing the misclassified substances of all sets unveiled several limitations of theEST, many of which also affect other in vitro and ex vivo alternative approaches to animaltesting. The major limitation of the EST is its representation of early organogenesis only. Inparticular, substances affecting specifically neural or osseous tissues have to be detected byseparate, more specialised assays. Secondly, the use of aqueous media restricts the testing ofhighly lipophilic substances, a shortcoming of most in vitro assays. In addition, some componentsof the media are necessarily supplemented at artificially high concentrations that counteractsome of the tested substancesf effects. A specific shortcoming of the validated EST is itsreadout of beating cell clusters, which is inhibited by muscle relaxants. Conversely,approaches where this physiological endpoint has been replaced by molecular markers,including our FACS-EST, remain unaffected by this class of substances. Moreover, some of thetested substancesf effects, or lack thereof, depends on maternal tissues and their pharmacokineticbehaviour, influencing factors that cannot be provided by the EST. Finally, the in vivoclassification of the substances is an important issue, requiring the consistent use of criteriaand recognition that species specificity also applies to in vitro systems. In conclusion, many ofthe limitations revealed can be overcome by modifications and additions to the EST, makingit a promising basis for a regulatory acceptable in vitro assay system for reproductive toxicity.