The ability of GPX and other selenoproteins to be selectivel

The ability of GPX4 and other selenoproteins to be selectively induced by ferroptotic stimuli suggests that the stress-induced increase in the transcription of a number of putative, “antioxidant” selenoproteins is an adaptive homeostatic, but insufficient response to prevent cell death in the face of ferroptotic insults. As Se supplementation has been previously shown to load selenocysteine proteins co-translationally (Ingold et al., 2018, Rafferty et al., 1998, Rusolo et al., 2013), we examined the effect of Se supplementation on this response. As expected, Se supplementation led to protection in the face of ferroptotic insults. What was unexpected was that Se supplementation could drive transcription of a host of selenoproteins. Detailed promoter analysis of the anti-ferroptotic gene, GPX4, identified TFAP2c and Sp1 as driving transcription in response to ferroptotic stimuli or Se supplementation. A model emerges that TFAP2c and Sp1 sense pharmacological changes in cellular “selenium” or selenocysteine and act sequentially to activate GPX4 transcription and a cassette of other genes that form a selenome. It is formally possible that increases in selenoprotein levels represent a way for adiponectin receptor to compensate for increased free and potentially toxic selenium, or selenocysteine, by increasing its incorporation into selenoproteins, where its redox reactivity can be homeostatically controlled. It also suggests that the set point for GPX4 levels in any cell could be set via transcriptional regulation by selenium availability, implying that the sensitivity to ferroptotic stimuli in any tissue, including cancer cells (Dixon et al., 2012, Viswanathan et al., 2017), could be determined by the amount of free selenium (or selenocysteine) that is delivered into the cell. The ability of Se or TFAP2c/Sp1 to drive a cassette of genes raises the possibility that these genes are activated as part of a coordinated stress response to compensate for ferroptotic or other stresses (e.g., ER stress and excitotoxicity). Se-induced significant changes in the expression of 238 genes including the mitochondrial and nuclear forms of GPX4 (Figures 3 and S4). Expression of mitochondrial and nuclear GPX4 isoforms are driven by distinct promoters and could have distinct roles in regulating signaling in distinct death paradigms as previously suggested (Hauser et al., 2013, Savaskan et al., 2007a, Savaskan et al., 2007b). A recent study from our lab demonstrated that nuclear ALOX5 mediates cell death following ICH (Karuppagounder et al., 2018) raising the possibility that nuclear GPX4 is the necessary isoform induced by Se to counteract lipid peroxidation and prevent cell death seen in ICH. Of course, other lipoxygenase (ALOX12 and 15) with distinct subcellular localizations have been implicated in ischemic stroke (Dixon et al., 2012, Yigitkanli et al., 2013) suggesting that depending on the clinical condition that GPX4 may mediate protection from distinct subcellular sites. Nutritional selenium is preferentially targeted to the brain and testes (in men) under conditions where selenium is rate limiting in the body (Brown and Burk, 1973, Schomburg and Schweizer, 2009). Indeed, removal of the testes in male mice that are genetically deficient in selenoprotein enzymes abrogates neurodevelopmental and neurodegenerative effects of selenium deficiency (Pitts et al., 2015). To overcome the nutritional targeting of selenium in the whole body to achieve pharmacological levels of Se in the brain, we injected Se directly into the ventricular system and found elevated GPX4 message specifically in this organ and reduced ferroptotic death and improved functional recovery following ICH in mice (Figure 5). However, we found that Se supplementation in vitro and in vivo has a parabolic dose response curve, raising concerns not only about the mode of delivery, but the ultimate safety of this approach. To overcome this challenge, we developed a peptide (Tat SelPep), which possesses a Tat transduction domain linked to six amino acids in the C-terminal domain of SelP. This peptide not only possesses a wide therapeutic window in vitro and in vivo, but also can induce GPX4 expression in a number of organs that were tested including the brain, heart, and liver (Figures 6 and S7B). In vivo studies showed that the salutary effects of Tat SelPep are inhibited by forced expression of a dominant-negative Sp1, consistent with the notion that Tat SelPep, like Se, drives transcription of GPX4 via a Sp1-mediated pathway (Figure 6). The ability to drive expression of the GPX4 and other genes of the selenome in any organ of the body offers a host of potential therapeutic opportunities for conditions where GPX4 deficiency is associated with cell death or dysfunction (Stockwell et al., 2017). Moreover, it offers a way to supplement Se in the whole body with less concern for toxicity (Burk and Hill, 2015, Jones et al., 2017). Our studies also suggest that agents that disarm transcription of GPX4 and other genes of the selenome might be a mechanism to potentiate tumor kill in response to ferroptosis inducers such as erastin (Dixon et al., 2012, Yang et al., 2014, Yu et al., 2015). Overall, these studies describe a novel transcriptional mode of regulation for Se in controlling cell death in vitro and in vivo with potentially broad nutritional (Jones et al., 2017) and therapeutic implications.