Histone methylation plays a key role in regulating chromatin structure, transcriptional activity and the DNA damage response. The importance of histone methylation in the DNA damage response is highlighted by the observation that inactivation of H3K9me3 methyltransferases leads to genomic instability and an inability to correctly repair DSBs caused by IR. Several KDMs, including KDM4B and KDM3A, contain Hypoxia Response Elementsand are transcriptionally activated by Hif1a under hypoxic conditions, suggesting that histone methylation may decrease under hypoxic conditions. Accordingly, we examined how the stabilization of Hif1a by DMOG impacts methylation of H3K9me3. In figure 4A, exposure of cells to DMOG rapidly increased H3K9me3 in cells. This was somewhat unexpected, since Hif1a transcriptionally upregulates 2 H3K9me3 specific KDMs, KDM4B and KDM3A, and should therefore decrease H3K9me3 levels. However, KDMs and the Hif1a prolylhydroxylases belong to the larger family of dioxygenases. Both the Hif1a prolylhydroxylases and KDMs utilize an Feand 2-oxoglutarate-dependent dioxygenase mechanism to either hydroxylate proline on Hif1aor remove methyl groups from methylated lysines on histone tails. DMOG, which is a 2-oxoglutarate analog, is a competitive inhibitor of the Hif1aPH and is therefore likely to inhibit KDMs as well. In figure 4B, the H3K9me3 specific demethylase KDM4A was transiently expressed in cells. Cells expressing vector showed increased H3K9me3 methylation when cells were exposed to DMOG, consistent with inhibition of endogenous KDMs. Evofosfamide 918633-87-1 Overexpression of KDM4A resulted in almost complete loss of endogenous H3K9me3 in the cells; however, addition of DMOG to cells expressing KDM4A restored H3K9me3 levels to near MK-2206 normal. Figure 4B therefore clearly demonstrates that DMOG, in addition to increasing Hif1a protein levels, can also inhibit endogenous KDMs and therefore increase levels of H3K9me3 in the cells. To further analyze how DMOG may regulate H3K9me3 levels, we also determined if Hif1a could increase expression of Suv39h1, a key H3K9me3 methyltransferase. Surprisingly, figure 4C and 4D demonstrate that both DMOG and, to a lesser extent, CoCl2, increased expression of the Suv39h1 methyltransferase. Further, the increase in Suv39h1 protein closely followed the dosedependentand time-dependentincrease in Hif1a levels caused by DMOG. Importantly, both basal and DMOG dependent induction of Suv39h1 were abolished in cells lacking Hif1a, indicating that Hif1a directly regulates the levels of Suv39h1 in cells. Figure 4 therefore demonstrates that DMOG can influence H3K9me3 levels by directly inactivating H3K9 demethylasesas well as upregulating H3K9 methyltransferases through a Hif1a dependent mechanism 
Figure 4C�CE). The previous results indicate that DMOG may influence radiosensitivity through 2 distinct pathways. First, DMOG can directly inhibit KDMs, increasing H3K9me3 levels by a mechanism which is independent of Hif1a. Second, DMOG can directly inhibit Hif1a-prolylhydroxylases, leading to stabilization and accumulation of transcriptionally active Hif1a. To determine the relative contributions of these 2 pathways to the ability of DMOG to function as a radioprotector, we examined how silencing of Hif1a with shRNA impacted radioprotection. Figure 5A demonstrates that DMOG increased radioresistance in MEFs expressing a non-specific shRNA, but this effect was lost when Hif1a was silenced. The ability of DMOG to function as a radioprotector therefore requires Hif1a, indicating that DMOG exerts its primary effect on radiosensitivity through inhibition of the Hif1a-prolylhydroxylase and accumulation of Hif1a.