A catalogue of approximately 50 molecules was established. Considering the good potential of those molecules, preliminary toxicological investigation was undergone using
in silico tools.
In silico analysis showed that the majority of test compounds was predicted to be non-mutagenic. No firm conclusion could be drawn regarding carcinogenicity due to unreliable predictions (i.e. predictions outside the applicability domain, or predictions not relevant to humans in the case of the barbituric derivatives). In regards to endocrine toxicity, compounds were not predicted to exert effects via estrogen, androgen or thyroid receptors. Predicted oral lethal dose (LD) 50 in rats indicated only very moderate acute toxicity.
In silico investigations were also done on environmental aspects. No compounds were predicted to be bioaccumulative (log (BCF) < 3.3), and the majority of compounds was predicted to be readily biodegradable. Computational analysis on persistence in different matrices (sediment, soil, and water) did not retrieve particular alerts.
In vitro experiments were further conducted with two lead compounds : sinapoyl malate (SM) and and di-geranyl sinapoyl malate (DGSM). The bacterial reverse mutation assay (also known as Ames test) was performed in compliance with OECD Test guidelines No. 471. Five different strains (TA98, TA100, TA1535, TA1537, and
E.colishort forEscherichia coli WP2) were incubated with SM or DGSM (concentration range from 10
µgshort formicrogram/plate to 5000
µgshort formicrogram/plate) in the presence or absence of rat S9 using the plate incorporation method. Neither SM nor DGSM showed any mutagenic potential under the tested conditions. To complete the data on mutagenicity, the micronucleus test was performed. V79 cells were incubated with SM or DGSM in the presence or absence of rat S9. Neither SM nor DGSM showed a clastogenic/aneugic potential under the conditions tested. In addition, experiments were conducted in the target organs lung, skin, brain, small intestine and liver. Possible induction of γH2AX, an unspecific marker for genotoxicity, was investigated by means of immortalized cell lines. A549 cells were used as surrogate for human alveolar epithelial cells, while HaCaT and Neuro-2a cells were used as surrogate for human keratinocytes and rat neuronal cells, respectively. Caco-2 cells were used as surrogate for human enterocytes while HepaRG cells were used as surrogate for human hepatocytes. Concentration range were first established using cytotoxicity assays (NRU and WST-1). Results showed that SM had no or very slight effect on the NRU assay irrespective of the tested cell line. However, SM showed concentration-dependent cytoxicity in HepaRG cells with estimated IC50 value of 2.60 mM in the WST-1 assay. DGSM showed concentration-dependent cytoxicity in Neuro-2a cells with estimated IC50 values of 1.63 mM and 1.13 mM, for NRU and WST-1 respectively. Data showed no induction of γH2AX following treatment with SM irrespective of the tested cell lines while DGSM slightly increased γH2AX (1.6-fold) in A549 and HepaRG cells at high concentrations. IC50 values determined for SM and DGSM are very high and do not indicate strong
in vitro cytotoxic potential. Furthermore, induction of γH2AX responses occured at very high concentrations and may be the result of unspecific apoptotic events. It is noteworthy that the used concentrations would not reflect realistic human .