Phenolic antioxidants reduce the effect of oxidative stress within cells. They are found in a various fruits, vegetables and as food additives to reduce spoilage. Consumption of antioxidants by humans has been linked with increased lifespan and reduced incidence of cancer and cardiovascular disorders (Cabrera et al. 2006; Kuriyama 2008). In cultured mammalian cells however, some of these phenolic antioxidants have been reported to generate reactive oxygen species (ROS), leading to chromosomal breakage (Long et al. 2007; Long & Halliwell 2001). It is clear then, that amongst this group of compounds, in vitro toxicological study is not a reliable prediction of human hazard. It is for this reason that the work described in this thesis was undertaken: the principal aim was to gain a better understanding of the reasons underlying this contradiction. It has been suggested that excessive ROS generated in vitro might be a result of the higher levels of oxygen (~20%) compared to (1-7%) in vivo: (Yusa et al. 1984; Turrens et al. 1982). With clearer understanding, new experimental approaches might be taken to highlight or reduce positive in vitro genotoxicity test results that might be considered misleading. A diverse set of test compounds was first chosen. It included polyphenolic (PPA), monophenolic (MPA) and non-phenolic antioxidants (NPA), in addition to mechanistically characterised oxidants, genotoxins and cytotoxic, non-genotoxins as controls. Genotoxicity was assessed in vitro using the GADD45a, GFP reporter assay and in silico using Derek Nexus . Amongst the 19 antioxidants assessed, the 11 of 12 of PPAs, 0 of 4 MPAs and 1 of 3 NPAs (ethoxyquin) produced positive results in vitro and 8 of 12 PPAs generated alerts of at least plausible genotoxicity in silico. To discover whether these results were the result of cellular hyperoxia-promoted generation of physiologically irrelevant ROS in cells, genotoxicity was reassessed in the presence of 1 and 5% oxygen. This reduced oxygen exposure had no effect upon the qualitative result for any of the assessed compounds and a negligible effect upon the dose at which any positive result was produced. An assessment of the ability of antioxidants to generate potentially genotoxic ROS within cells was carried out using the intracellular fluorescent dye, dichlorofluorescin diacetate (DCFH-DA). 10 of 12 PPAs, 0 of 4 MPAs and 1 of 3 NPAs (ethoxyquin) were shown to increase the level of ROS within TK6 human lymphoblastoid cells within 4 hours of compound exposure. Within this same timeframe, the mitochondrial membranes in cells treated with 10 of 12 PPAs, 2 of 4 MPAs and 1 of 3 NPAs (ethoxyquin) were shown to become depolarised using JC-1 dye. It was unclear however, whether mitochondrial membrane depolarisation was a cause or a consequence of ROS generation within the cells. In order to assess whether the increase in intracellular ROS led to an increase in oxidised DNA within treated cells, 8-oxoguanine (8-OG) was quantified using a FITC conjugated anti8-OG antibody. This assessment revealed that levels of the oxidised base were only increased in cells exposed to two of the 12 PPAs (quercetin and resorcinol). The level of 8OG detected was lower than the vehicle control for cells treated with 10 of the 15 antioxidants. One interpretation of this is that these agents induce the repair pathway for oxidative damage, which leads to a lower level of oxidised DNA bases in the genome. The results showed that while a large proportion of PPAs produce genotoxic results in vitro and lead to increased levels of ROS, the amount of oxidised DNA is not higher in treated cells. This would suggest the presence of a different mechanism for the observed genotoxicity.
|Date of Award
|31 Dec 2015
- The University of Manchester
|Richard Walmsley (Supervisor) & Christopher Grant (Supervisor)
- Genetic toxicology