ORIGINAL ABSTRACT

The increased use of agrochemicals raises concerns about environmental, animal, and mainly human toxicology. The development of New Approach Methodologies (NAMs) for toxicological risk assessment including new in vitro tests and in silico protocols is encouraged. Although agrochemical mutagenicity testing is well established, a complementary alternative approach may contribute to increasing reliability, with the consequent reduction of false-positive results that lead to unnecessary use of animals in follow-up in vivo testing. Additionally, it is unreasonable to underestimate the phototoxic effects of an accidental dermal exposure to agrochemicals during agricultural work or domestic application in the absence of adequate personal protection equipment, especially in terms of photomutagenicity. In this scenario, we addressed the integration of in vitro and in silico techniques as NAMs to assess the mutagenic and phototoxic potential of agrochemicals. In the present study we used the yno1 S. cerevisiae strain as a biomodel for in vitro assessment of agrochemical mutagenicity, both in the absence and in the presence of simulated sunlight. In parallel, in silico predictions were performed using a combination of expert rule-based and statistical-based models to assess gene mutations and phototoxicity. None of the tested agrochemicals showed mutagenic potential in the two proposed approaches. The Gly and 2,4D herbicides were photomutagenic in the in vitro yeast test despite the negative in silico prediction of phototoxicity. Herein, we demonstrated a novel experimental approach combining both in silico and in vitro experiments to address the complementary investigation of the phototoxicity and (photo)mutagenicity of agrochemicals. These findings shed light on the importance of investigating and reconsidering the photosafety assessment of these products, using not only photocytotoxicity assays but also photomutagenicity assays, which should be encouraged.