Predictive mechanistic in silico simulations of mixture of azole fungicides acting on cranio-facial morphogenesis
Résumé
The huge number of combinations of chemical, the paucity of efficient test strategies for the risk assessment of mixtures, and the increasing societal need to reduce animal testing make the study of mixtures a very complex issue. Single azole fungicides, can affect cranio-facial morphogenesis during the early development. Data previously obtained in post-implantation rat whole embryo cultured in vitro described specific teratogenic effects at the branchial structures, while the coexposure resulted in mixture effects, accounting for a common mode of action (MoA) for the azole fungicides. The proposed MoA for azoles is the inhibition of CYP26 enzymes, involved in the retinoic acid (RA) catabolism, with a consequent indirect increase in the local endogenous concentration of RA at the branchial apparatus level. With the aim of investigating this hypothetical mechanism, the specific teratogenic effect of some azole fungicides, alone or in mixture, has been tested in vitro. Then, data were interpreted developing an innovative in silico tool, combining in vitro experiments and mathematical and probabilistic modelling of RA pathway enzymes. After training with single azoles dose–response data, predictive simulations for the mixtures were performed. The model appears to be reasonably predictive for the mixture effects (experimental data and model predictions are in promising agreement) confirming the accuracy of the hypothesised pathogenic pathway. In addition, the formation of physiological RA levels in the rat embryo hindbrain can be reasonably estimated after azole exposures. This research will provide a better understanding of the toxicity mechanism of single and combined chemicals affecting RA (not just azole fungicides), and a predictive tool for human risk assessment for all mixtures acting on craniofacial morphogenesis. Finally, predictive mechanistic simulations of mixture effects, just using dose–response data for single chemicals, will significantly reduce the use of animals in developmental toxicity studies.