Application of the TGx-28.65 genomic biomarker to assess the genotoxicity of disperse orange and 1,2,4-benzenetriol in human cells [DO]

In vitro gene expression signatures to predict toxicological responses can provide mechanistic context for human health risk assessment purposes. We previously developed the TGx-28.65 genomic biomarker from a database of gene expression profiles in human TK6 cells exposed to 28 well-known compounds, and it comprises 65 genes that can classify chemicals as DNA damaging or non-DNA damaging. In this study, we applied the TGx-28.65 genomic biomarker in parallel with the in vitro micronucleus (MN) assay to determine if two chemicals of regulatory interest at Health Canada, disperse orange (DO: the orange azo dye 3-[[4-[(4-Nitrophenyl)azo]phenyl]benzylamino]propanenitrile) and 1,2,4-benzenetriol (BT: a metabolite of benzene) are genotoxic or non-genotoxic. Both chemicals caused dose-dependent declines in relative survival (RS) and increases in apoptosis. A strong significant increase in micronucleus induction was observed for all concentrations of BT; the top two concentrations of DO also caused a statistically significant increase in MN, but these increases were less than 2-fold above controls. TGx-28.65 analysis classified BT as genotoxic at all three concentrations and DO as genotoxic at the mid and high concentrations. Thus, although DO only induces a small increase in MN, this response is sufficient to induce a cellular DNA damage response that is likely relevant for risk assessment. Benchmark dose modeling revealed that BT is much more potent than DO; the BT benchmark dose for MN induction was similar to that of benzo[a]pyrene, (BaP: a genotoxic carcinogen), which was used as a positive control. The results strongly suggest that follow-up work is required to assess whether DO and BT are also genotoxic in vivo. This is particularly important for DO, which may require metabolic activation by bacterial gut flora to fully induce its genotoxic potential. Our previously published data and this proof of concept study suggest that the TGx-28.65 genomic biomarker has the potential to add significant value to existing approaches used to assess genotoxicity. This experiment examined the whole genome transcriptional response of human TK6 cells exposed to 3 concentrations of disperse orange (50, 500 and 750 µg/ml). TK6 cells were exposed to the test compound for 4 hours followed by a recovery time of 3-4h in fresh media at 3 different concentrations, including a low, a medium, and a high concentration, in addition to a negative vehicle control (DMSO (+S9)) and a positive control (benzo[a]pyrene at 10 µg/ml). Each concentration and time point had 3 biological replicates. There were a total 13 samples (26 arrays) included in the final analysis using a two-colour dye swap design.