DNA Methyltransferases Modulate the Bystander Effect in Mouse Embryonic Stem Cells.

Cells exposed to radiation or other genotoxic agents can induce DNA damage and other stress responses in non-irradiated cells that are either cultured with the irradiated cells or have been exposed to culture medium from irradiated cells. This is called the bystander effect. In a previous study we found that the descendents of bystander cells exposed to Mitomycin C (MMC) are themselves capable of inducing homologous recombination in un-exposed cells. This suggests that MMC induces persistent and transmissible changes in expression in bystander cells. Bystander effects are likely caused by epigenetic mechanisms rather than “classic” mutations, i.e. changes in DNA sequence. One of the epigenetic mechanisms cells employ for changing expression is DNA methylation in which DNA methyltransferases (DNMTs) add a methyl group to the 5 carbon of cytosine. In this study we asked if ionizing radiation can induce transmissible DNA damage in bystander cells by examining if bystander cells exposed to irradiated cells were themselves able to induce damage in naive cells. Furthermore, we asked if this was dependent on DNMT activity in the irradiated cells. We irradiated wild-type (WT) and DNMT triple knockout (DNMT TKO) mouse embryonic stem cells (ESCs) and after two weeks of continuous culture, we collected conditioned medium (CM). CM was then added to cultures of naive WT ESCs (primary bystanders). Three weeks later, CM was collected from the primary bystanders and added to naïve WT cells (secondary bystanders). We assessed DNA damage by evaluating strand breaks using the alkaline Comet assay and sister chromatid exchange (SCE) analysis. As expected, we found that medium from cells irradiated with 5 Gy induced modest damage in bystander cells. The median Olive tail moment was 2.8 in bystander cells exposed to conditioned medium from irradiated cells compared to 1.0 in control bystander cells (p < 0.0001). Homologous recombination was 0.15 chromatid exchanges per chromosome compared to 0.092 in control bystanders (p < 0.0001). We also observed an increase in strand breaks in secondary bystanders of a similar magnitude to that found in primary bystanders, indicating that radiation-induced bystanders are themselves able to induce damage. In contrast to WT cells, the irradiated DNMT TKO cells did not induce strand breaks in bystander cells, as measured by the Comet assay, but did induce HR. Surprisingly, we also observed that un-irradiated DNMT TKO cells induce DNA damage in bystanders, and furthermore that the magnitude of the effect is similar to that induced by irradiated WT cells. These data suggest that methyltransferases have a complex role in bystander effects. Bystander effects may be mediated by free radicals. To see if the DNMT TKO cells had changes in antioxidant levels, glutathione (GSH) and glutathione peroxidase (GPX) activity were determined. There was no significant change in GSH levels between WT and DNMT TKO cells. However, DNMT TKO cells had significantly higher levels of GPX activity (275.4 + 19.8 mU/mg protein) compared to control cells (122.0 + 16.4 mU/mg, p= 0.0001). Taken together, these results show that radiation-induced bystander cells can themselves induce damage in un-irradiated cells and suggest that cells lacking DNA methylation activity can induce bystander effects.