Toxicological evaluation of mycotoxin mixtures: development of in vitro testing strategies and in vivo assessment of a binary mixture

Abstract

Mixtures assessment is a key challenge in the field of mycotoxins. The general aim of this work was to contribute to the current toxicological evaluation of mycotoxin mixtures with a special focus on genotoxicity. Firstly, the current evidence about the combined toxicity among aflatoxins and ochratoxin A was gathered. These mycotoxins were selected due to their individual toxicity and the probable human co-exposure. Inconclusive results were observed among the articles due to the diversity of experimental designs and methodologies employed for assessing the type of interaction. As there is a lack of tools for substance prioritization, we developed an in silico-in vitro strategy. The endpoint selected was genotoxicity as the main concern associated with mycotoxins is derived from the long-term exposure. The strategy allowed us to classify 12 mycotoxins considering their genotoxic potential, concretely, point mutations. The mycotoxins of the strategy were selected considering the toxicity potential and occurrence data. Two steps were applied. Firstly, two in silico tools were used. DEREK Nexus® is an expert knowledge tool commercially available commonly used in the pharmaceutical industry. VEGA QSAR© is a statistical based and publicly available tool that requires an expert knowledge on chemistry and modelling for its use. In phase 2, the in vitro SOS/umu test in absence and presence of liver S9 was used. It is a medium-throughput assay with high concordance with the Ames test; the gold standard test for gene point mutations. However, a high amount of time and test item is required. Moreover, this test also enables us to characterize the 12 mycotoxins using a kidney S9 fraction for in vitro bioactivation. We validated the test in these conditions with the negative and positive controls. Expected results were obtained with positive controls (4-nitroquinoline-N-oxide and 2-aminoanthracene) without metabolic activation or with liver S9, but a potent concentration-dependent effect with 4-nitroquinoline-N-oxide and no activity of 2-aminoanthracene with kidney S9 were observed. It showed to be a reproducible assay for clear negative and positive compounds. However, for some of the mycotoxins classified as equivocal a higher variability was observed with kidney S9. The SOS/umu test is normally performed only once. But if equivocal results are obtained is advisable to repeat it and to be used with complementary assays. Aflatoxin B1 and sterigmatocystin resulted in positive genotoxicity outcomes in silico and in vitro with liver S9. Moreover, both are part of the same biosynthetic route. Therefore, an in vivo assessment of the binary mixture was carried out in male Wistar rats. A single oral dose of AFB1 (0.25 mg/kg) and STER (20 mg/kg) alone or in combination did not cause histopathological alterations in kidney or liver. A slight DNA damage (DNA strand breaks) was detected in liver of AFB1 treated animals, as well as a slight increase in the Fpg-sensitive sites in liver of animals treated with the mixture. The exposure to the individual toxins or the mixture did not increase the % of micronuclei in the bone marrow. The absence of genotoxicity observed in this study might be due to the vehicle (corn oil) and the rat strain (Wistar) used, as the levels of mycotoxins reached in target organs were x1.5 times lower than in previous studies using the same AFB1 dose dissolved in NaHC03 and administered to F344 rats. The lack of genotoxic response avoided to discuss the interaction in more detail. However, the plasmatic and tissular levels in the mixture of both mycotoxins were higher than in the samples of animals treated with the mycotoxins alone. This could point a possible toxicokinetic interaction of both mycotoxins at level of metabolism.