New computational approaches for repeated dose toxicity prediction in view of the safety assessment of cosmetic ingredients - Ineris - Institut national de l'environnement industriel et des risques Accéder directement au contenu
Communication Dans Un Congrès Année : 2014

New computational approaches for repeated dose toxicity prediction in view of the safety assessment of cosmetic ingredients

Andrea-Nicole Richarz
  • Fonction : Auteur
M.N. Berthold
  • Fonction : Auteur
Elena Fioravanzo
  • Fonction : Auteur
Daniel Neagu
  • Fonction : Auteur
Chihae Yang
  • Fonction : Auteur
Mark Cronin
  • Fonction : Auteur

Résumé

Since the full EU ban of marketing cosmetics containing ingredients tested on animals entered into force on 11 March 2013, alternatives have become necessary to evaluate the consumer safety of cosmetics. The EU COSMOS Project within the SEURAT-1 Research Initiative is developing computational modelling approaches, focusing on repeated dose toxicity, to support the safety assessment of cosmetics-related chemicals, without relying on animal testing. Efficient storage of relevant information in flexibly searchable databases is needed to allow for development of structure-based models for toxicity prediction. A comprehensive database has been developed and made publicly available (COSMOS DB ver1.0; http://cosmosdb. cosmostox. eu). It includes an inventory of over 5,500 cosmetics-related substances with high quality structures linked to data compiled from over 12,000 toxicological studies for over 1,660 compounds with data across 27 endpoints, including in vitro and in vivo genetic toxicity and oral repeated dose toxicity. Based on a new oral repeated-dose toxicity database within COSMOS DB, a new COSMOS non-cancer TTC database has been established enriched by cosmetic ingredients. This has been developed to evaluate the potential of applying the threshold of toxicological concern (TTC) concept to risk assessment of cosmetics, with particular consideration given to the dermal route of exposure. Moreover, grouping approaches for toxicity prediction have been developed and applied, e.g. for hair dyes. Data mining of the oRepeatTox DB has identified structural fragments capable of inducing hepatotoxicity (steatosis/steatohepatitis/fibrosis). This knowledge has been captured in the form of “chemotypes” relevant for liver toxicity of cosmetics-related chemicals. The studies were supported by molecular modelling to predict binding to two nuclear receptors (LXR and PPARy) considered to be involved in steatosis. Elucidating the mechanisms underlying toxic events can be used to inform the development of Adverse Outcome Pathways. In order to support oral-to-dermal extrapolation a new database of skin permeability values has been developed along with QSAR models for dermal absorption, skin and liver metabolism and hepatic clearance. Furthermore, to support in vitro to in vivo extrapolations, physiologically-based toxicokinetic (PBTK) models have been developed and calibrated. Combined with cell based assays, incorporating aspects of chemical fate, cell growth, toxicity and feedback, they enable realistic estimates of in vivo concentration at the organ level to be extrapolated from in vitro data. COSMOS models are being coded using open access KNIME workflow software. Workflows can be executed from locally installed KNIME software or via a web browser using the KNIME WebPortal (http://knimewebportal.cosmostox.eu).

Domaines

Toxicologie
Fichier non déposé

Dates et versions

ineris-01855536 , version 1 (08-08-2018)

Identifiants

  • HAL Id : ineris-01855536 , version 1

Citer

Andrea-Nicole Richarz, M.N. Berthold, Elena Fioravanzo, Daniel Neagu, Alexandre R.R. Pery, et al.. New computational approaches for repeated dose toxicity prediction in view of the safety assessment of cosmetic ingredients. 16. International Workshop on Quantitative Structure-Activity Relationship in Environmental and Health Sciences (QSAR 2014), Jun 2014, Milan, Italy. pp.68. ⟨ineris-01855536⟩

Collections

INERIS
42 Consultations
0 Téléchargements

Partager

Gmail Facebook X LinkedIn More