Summary
Over the past five decades, thanatology has come to include the study of how individual cells in our bodies die appropriately and inappropriately in response to physiological and pathological stimuli. Morphological and biochemical criteria have been painstakingly established to create clarity around definitions of distinct types of cell death and mechanisms for their activation. Among these, ferroptosis has emerged as a unique, oxidative stress-induced cell death pathway with implications for diseases as diverse as traumatic brain injury, hemorrhagic stroke, Alzheimer’s disease, cancer, renal ischemia, and heat stress in plants. In this review, I will highlight some of the formative studies that fostered its recognition in the nervous system and describe how chemical biological tools have been essential in defining events necessary for its execution. Finally, I will discuss emerging opportunities for anti-ferroptotic agents as therapeutic agents in neurological diseases.
Graphical Abstract
eTOC
Iron has been causally linked to almost every neurological condition but CNS therapeutics targeted at iron have yet to be developed. In this review, Ratan highlights mechanistic understanding of ferroptosis-a unique, iron-dependent cell death pathway. This understanding provides clear opportunity to develop broad and effective neurological disease therapeutics.
Introduction
Ferroptosis is a non-caspase dependent form of programmed necrosis that is activated by oxidative stress induced by glutathione depletion, hemin, and/or reactive lipid species (Dixon et al., 2012; Murphy et al., 1989; Seiler et al., 2008; Zille et al., 2017). Oxidative stress, an imbalance between oxidants and antioxidants in favor of oxidants, has been implicated in almost every neurological condition and yet, established antioxidants such as vitamin E and its functional analogs have been disappointing in human trials (Antonic et al., 2018). These disappointments are understandable when one considers that our understanding of the molecular targets of antioxidants is poor; the important signaling role of oxidants in the normal central nervous system has been underestimated; and our unvalidated tendencies to treat oxidative stress as an undifferentiated whole rather than as a phenomenon whose consequences depend on the oxidant involved, its subcellular site of production, and known mechanisms of defense. As discussed below, ferroptotic inducers have provided a concrete path to address some of these extant challenges in understanding whether a particular subtype of oxidative stress is pathophysiologically relevant to the nervous system.
The system xc− (xCT) cystine/glutamate transporting agency, a novel target for toxicity mediated by the excitatory neurotransmitter glutamate and for understanding neuronal oxidative death
The term ferroptosis was coined in 2012 by Stockwell and colleagues with their seminal discovery that the anti-tumor drug erastin inhibits cystine transport via the system xc− cystine/glutamate transporting agency (xCT) to induce cell death in Kirsten rat sarcoma (KRAS) mutant and other types of tumor lines (Dixon et al., 2012). While these were groundbreaking observations, the table for this discovery was set decades earlier via studies by Shiro Bannai and colleagues in the late 1970s. They first showed that the sulfhydryl cysteine is rapidly oxidized to cystine—in which the sulfur groups are covalently linked—in the bathing medium of cultured cells, but that cells have the ability to transport cystine (in its minority, anionic form) and regenerate reducing agents that optimize the redox state of proteins intracellularly and extracellularly (Bannai and Kitamura, 1980; Makowske and Christensen, 1982; Takada and Bannai, 1984). The importance of these processes in cell survival was demonstrated by the findings that removal of cysteine from the bathing medium, or cystine transport blockade depleted cells of the versatile tripeptide antioxidant glutathione (γ-glutamyl-cysteinyl glycine) and caused an “antioxidant” reversible cell death (Bannai et al., 1977). Subsequently, the transporter for cystine was cloned (Sato et al., 1999). xCT is a chloride-dependent, non-electrogenic, and non-energy-requiring transporter that stoichiometrically exchanges anionic cystine for anionic glutamate. It is made up of two subunits. One subunit, SLC7A11, is 502 amino acids and has 12 membrane spanning domains; the other subunit, 4F2hc, is a cell surface glycoprotein with one transmembrane domain linked to SCL7A11 with a single disulfide bridge. The transporter exchanges intracellular glutamate (5-10 mM) for extracellular cystine (100 μM) with a molar ratio of 1:1 (Figure 1A (Koppula et al., 2018). Importantly, Bannai demonstrated that longer analogs of glutamate including quisqualate potently inhibited cystine transport while shorter analogs such as aspartate or N-methyl d-aspartate (NMDA) had no effect (Bannai, 1986). The ability of glutamate analogs to inhibit the xCT cystine antiporter correlated with toxicity in a number of cell types (Figure 1B).
Figure 1. Glutamate or its analogs induce ferroptosis in neurons and other cell types by competitively inhibiting the system xc− cystine/glutamate exchanger (xCT cystine antiporter) and blocking cystine uptake.
A) Structure of xCT cystine antiporter. The antiporter is composed of a 12 pass transmembrane subunit (SLC7A11) where anionic cystine moves into the cell via facilitated diffusion in exchange for anionic glutamate (which moves out of the cell via facilitated diffusion). The antiporter also has a 1 pass transmembrane regulatory subunit (SLC3A2) which is linked to the transporter via a disulfide bond.
B) Ability of glutamate analogs to induce ferroptosis in neurons is directly correlated with their ability to inhibit radiolabeled cystine uptake into the cell. Homocysteate nearly completely inhibits cystine uptake and it avidly induces ferroptosis. By contrast, N-methyl D-aspartate (NMDA) has no effect on radiolabeled cystine uptake and as expected, is not able to induce ferroptosis in neurons.
This line of the investigation caught the attention of Tim Murphy and Joe Coyle, neuroscientists based at Johns Hopkins University who were interested in novel mechanisms by which glutamate could induce cell death in the nervous system. Glutamate is a canonical excitatory neurotransmitter in the central nervous system which can “fire” neurons to death (excitotoxicity) via its ability to excessively activate synaptic and extrasynaptic ionotropic glutamate receptors (Choi et al., 1987). Vulnerability in this context accrues from the expression of specific subtypes of glutamate receptors and the level of exposure to glutamate. Indeed, excitotoxicity emerges from disruptions in glutamate homeostasis in a host of neurological conditions including stroke, traumatic brain injury, ALS, and Alzheimer’s disease (Rothstein, 1996). While antagonists to distinct glutamate receptors are effective in rodent models of neurological conditions, their success has not translated to human clinical trials. A likely reason for this is the important role that glutamate plays in the normal execution of diverse neurological functions such as blood pressure control, normal affect and cognition. Accordingly, doses of glutamate antagonists that were protective in rodents were unacceptably toxic in humans (Wu and Tymianski, 2018).
To explore distinct pathways of glutamate-induced death in the nervous system, Coyle, Murphy, and colleagues screened a host of cell lines and identified the N18-RE-105 (neuroblastoma-retinal hybrid cell line) as one that expressed high levels of the chloride-dependent, quisqualate-sensitive, glutamate-binding site on its plasma membrane that is identical to the xCT cystine antiporter identified by Bannai. As expected, N18-RE-105 neuroblastoma cells underwent cell death induced by glutamate or quisqualate (not NMDA) that could be inhibited by adding excess cystine to the medium or exacerbated by removing cystine from the medium (Murphy et al., 1989). Cell death by cystine deprivation or glutamate in neuroblastoma (N18-RE-105) cells or immature neurons was not blocked by antagonists to established glutamate receptors because N18-RE-105 cells do not express these receptors. The potency of cell killing in this cell line or primary neurons by glutamate analogs correlated with the potency of inhibition of radioactive cystine or radioactive glutamate uptake into the cells and was not correlated with any electrophysiological changes to glutamate (Figure 1B; (Murphy and Baraban, 1990; Murphy et al., 1990). Moreover, cell death was associated with depletion of the antioxidant glutathione and increases in fluorescence of a redox-sensitive reporter; it was also inhibitable by “antioxidants” (Figure 2). Of note, the antioxidants that most potently and durably inhibited non-receptor-mediated glutamate toxicity were those which targeted lipid peroxidation (Miyamoto et al., 1989).
Figure 2.). What are the consequences of inhibiting cystine uptake?
Cysteine added to the extracellular medium is rapidly oxidized to cystine (two cysteines covalently linked by a disulfide bond). Cystine is taken up into the cell via the xCT transporter in exchange for glutamate. Intracellularly cystine is reduced to cysteine via glutathione reductase. Cysteine (the rate limiting substrate for glutathione synthesis) is then covalently linked to glutamate via γglutamylcysteine synthetase. γglutamylcysteine is then covalently linked to glycine via glutathione synthetase to form γglutamylcysteineglycine (reduced glutathione). Inhibition of cystine uptake (by glutamate, erastin, homocysteate, quisqualate, or ibotentate) leads to depletion of cysteine and associated depletion of glutathione, a versatile antioxidant cofactor. Depletion of glutathione leads to ferroptotic death in embryonic cortical neurons, neuroblasts, oligodendrocytes and astrocytes although the primary neurons and immature oligodendrocytes are more sensitive to ferroptosis inducers. This likely relates to the higher antioxidant potential of astrocytes. Inhibitors of transcription (actinomycin-d) and translation (cycloheximide) abrogate ferroptosis in neurons-programmed necrosis or metabolic reprogramming? Two, non-exclusive models for mechanism of protection from ferroptosis by macromolecular synthesis inhibition: (left) metabolic reprogramming of cysteine from global protein synthesis into glutathione. (right) Inhibition of specific transcription factors (e.g. activating transcription factor 4 (ATF4) that drives expression of putative ferroptotic genes (Chop, Trb3, Chac1, REDD1, CARS, see text).
Glutathione is a tripeptide (gamma-glutamyl-cysteinyl-glycine) that is present in the brain and peripheral tissues in near millimolar concentrations (Aoyama et al., 2006; Dringen et al., 2015; Lee and Johnson, 2004). It is a versatile reservoir of electrons and can participate in a variety of redox reactions via its ability to interact with glutathione reductase, glutathione peroxidases, and glutathione transferases. The ability of glutamate to induce cell death by competitively inhibiting cystine uptake via the xCT transporter leading to glutathione depletion and not via selective binding to glutamate receptors supports this mechanism of glutamate toxicity (Figure 2).
From apoptosis to oxytosis to ferroptosis: how did we get here?
Accordingly, although glutamate or glutamate analog (HCA)-induced cell death is now referred to as ferroptosis, it was not recognized as such until Stockwell and Dixon defined it in 2012. Indeed, my own studies as a post-doctoral fellow in the early 1990s incorrectly labeled ferroptosis as apoptosis (Ratan et al., 1994a), but I ignored clues that this was a fundamentally different cell death pathway. First, glutamate-induced xCT cystine transport blockade (which is now known to trigger neuronal ferroptosis) was exacerbated by growth factors (brain-derived neurotrophic factor, BDNF) or serum; by contrast the canonical apoptosis of NGF deprivation in sympathetic neurons, a caspase dependent death, was blocked by the addition of NGF and other growth factors (Ratan et al., 1996). Second, the release of lactate dehydrogenase (LDH, a cytosolic enzyme) into the extracellular space, a sign of necrotic disruption of the plasma membrane, could be used 24 hours after exposure to glutamate to quantify cell death in what is now known as ferroptotic necrosis; by contrast, canonical apoptosis inducers such as staurosporine (a non-specific tyrosine kinase inhibitor) or Sindbis virus (used to model CNS encephalitides) showed no LDH release even after 48 hours (Esch et al., 1998).
A more appropriate formulation for non-receptor-mediated, oxidative glutamate toxicity that was thoughtfully advanced by Schubert and Maher was with the term oxytosis (Tan et al., 2001). Inherent in the term oxytosis was clear recognition that glutamate-induced xCT transport inhibition leads to death with features of apoptosis and necrosis whose mechanisms of death are, in aggregate, fundamentally different from apoptosis or necrosis. Schubert and Maher deserve enormous credit for their contributions to our understanding of oxytosis/ferroptosis in the nervous system, as will be outlined below.
However, it is undeniable that the concept of ferroptosis advanced by Stockwell has captured the imagination of many cell death researchers, for a host of reasons. First, it is notable because of its important biological contribution in showing that KRAS mutant tumors, which comprise 30-40% of all tumors, are particularly sensitive to erastin, a highly selective inhibitor of the xCT transporter (Stockwell et al., 2017). These studies codified differences between ferroptosis and other cell death pathways at a time when the richness of cell death diversity had been fully appreciated (Dixon et al., 2012). Notably, Stockwell and Dixon’s formulation went beyond the idea that ferroptotic death could only be used as an in vitro tool to understand oxidative death in the nervous system. Their formulation also expanded its relevance to other diseases and organs outside of the nervous system. Finally, and most importantly, studies of ferroptosis also led to the creation of a set of widely used and validated chemical biological tools (erastin, ferrostatin, RSL3, liproxstatin, adaptaquin, etc.) that have highlighted the important, relevant of work emerging from many labs. Ferroptosis is heretofore an underexamined concept relevant to cancer (Jiang et al., 2015), diseases in plants (Conrad et al., 2018; Distefano et al., 2017), and numerous diseases of many organs including the brain (Chen et al., 2015; Hambright et al., 2017). Additionally, Rosenberg, Volpe, and colleagues highlighted the important role that ferroptosis can play in non-neuronal CNS cells, especially immature oligodendrocytes, a topic beyond this review which will focus on neuronal vulnerability to ferroptotic stimuli (Back et al., 1998).
Glutamate-induced ferroptosis in neurons depends on macromolecular synthesis: Metabolic reprogramming or programmed cell death?
In the early 1990s, the prevailing notion was that oxidative stress triggered a thermonuclear-like chain reaction leading to cell death by the non-specific oxidation of proteins, lipids, and DNA. That paradigm shifted dramatically with the observation that glutamate-induced ferroptosis could be inhibited by inhibitors of transcription and translation (Figure 2A,B; Ratan et al., 1994a,b). Examination of the role of transcription in ferroptosis was stimulated by growing recognition that neuronal apoptosis during development or due to trophic factor deprivation could be abrogated by transcriptional and translational inhibitors (Martin et al., 1988). Two non-exclusive models emerged for the mechanism of action of macromolecular synthesis inhibitors. First, radioactive cyst(e)ine measurements showed that at concentrations sufficient to induce death, glutamate-induced cystine transport inhibition led to a 50% reduction of radioactive cystine incorporation into acid precipitable protein and an identical 50% reduction into acid soluble thiol (primarily glutathione) (Chatterjee et al., 2001; Esch et al., 1998; Ratan et al., 1994b). Inhibition of transcription or translation inhibited cyst(e)ine incorporation into de novo protein synthesis and shunted it into the formation of non-protein thiols including glutathione. Accordingly, glutathione levels rose above the minimum level required to maintain cell survival (Figure 2). Inhibition of incorporation of cysteine into glutathione with a selective inhibitor of gamma-glutamyl cysteine synthetase, buthionine sulfoxamine (Griffith and Meister, 1979), prevented the ability of macromolecular synthesis inhibitors to protect against ferroptosis (Ratan et al., 1994b). Indeed, agents that directly increased cysteine or glutathione including N-acetylcysteine (a cysteine prodrug), glutathione ethyl ester (a glutathione prodrug), and 5-oxothiazolidine carboxylate (a cysteine prodrug) all enhanced survival demonstrating that ferroptotic death can be rescued by cellular glutathione repletion (Karuppagounder et al., 2018; Ratan et al., 1994b).
Transcriptional inhibitors can also prevent neuronal ferroptosis by directly inhibiting de novo synthesis of proteins established to be pro-death in non-ferroptotic paradigms where cells die (Figure 2). Microarray studies of primary neurons showed that 6 to 8 hours following exposure to a ferroptotic stimulus (a glutamate analog), 119 genes were dysregulated (79 genes upregulated and 40 genes downregulated) (Lange et al., 2008). Among the genes induced, two established pro-death genes were notable-CCAT enhancer binding protein homologous protein (CHOP) and Tribbles homolog 3 (Trib3) (Karuppagounder et al., 2016; Lange et al., 2008). Analysis of the promoter regions for these two genes confirmed a previously identified, highly conserved response element called the CARE element (C/EBP-ATF response element, TGATGXAAX) that contains an ATF (Activating Transcription Factor Family) half site and a C/EPB (CCAT enhancer binding protein family) half site (Ma et al., 2002; Ohoka et al., 2005). A transcription factor known to bind with high affinity to this site and drive transcription of Trib3 and CHOP is ATF4 (Aime et al., 2020). Indeed, ferroptotic stimuli such as homocysteate (a glutamate analog that inhibits the xCT transporter, a Type 1 ferroptosis inducer [FIN]), erastin (a potent, selective inhibitor of the xCT transporter, and another Type 1 FIN), and RSL3 (an inhibitor of stability of the enzyme, GPX4 that neutralizes reactive lipids [Type II FIN]) all appear to induce the leucine zipper transcription factor, ATF4, at a protein level in primary neurons (Karuppagounder et al., 2016; Lange et al., 2008); Karuppagounder et al., in preparation; Figure 2).
While the specific upstream ferroptotic signals that activate ATF4 in neurons are unclear, amino acid (cysteine) depletion, iron redistribution, or reactive lipid species may separately or collectively contribute to ATF4-dependent gene activation. The mechanisms by which iron could activate ATF4-dependent transcription will be discussed later in this review. However, cellular cyst(e)ine depletion alone can lead to upregulation of ATF4 via the uncharging of cysteinyl tRNAs resulting in activation of a kinase called GCN2 (general control nonderepressible 2; (Hinnebusch, 1994). In turn, activated GCN2 can phosphorylate eif2α, a translational initiation factor whose affinity for eif2B (a guanine nucleotide exchange factor) increases when it is phosphorylated. The sequestration of eif2B leads to limiting levels of the ribosomal initiating ternary complex composed of eif2α-GTP, a methionine tRNA, and the 43S ribosomal subunit. Limiting amounts of this complex lead to the paradoxical translation of ATF4, and it then drives transcriptional activation of the integrated stress response which includes proferroptotic genes in primary neurons (Pathak et al., 2019). Consistent with this model, depletion of eif2α, which would inhibit the formation of eif2α-GTP, should dramatically reduce the amount of ternary translational initiating complex and prevent paradoxical translation of ATF4 which proved to be the case in glutamate-induced ferroptosis (Tan et al., 2001). Of note, the deprivation of certain amino acids, including methionine and possibly cysteine could lead to eif2α phosphorylation and ATF4 activation via glutathione depletion in the endoplasmic reticulum and GCN2-independent mechanisms (Mazor and Stipanuk, 2016; Wanders et al., 2016)
Whatever the stimulus for its activation, germline deletion of ATF4 can render primary immature neurons resistant to ferroptosis and cell death can be reinstated by forcing expression of wild type ATF4 but not ATF4 with its DNA binding domain mutated (Lange et al., 2008); Figure 2). Moreover, while 119 genes are dysregulated (79 significantly upregulated and 40 significantly downregulated) in cortical neurons exposed to the ferroptosis-inducer glutamate, only 3 genes change when glutamate is added to neurons with germline deletion of ATF4 (Figure 2). These results are congruent with the notion that transcription mediated via ATF4 drives transcription of upstream pro-ferroptotic death genes that facilitate subsequent ferroptotic stress (Figure 3). Of the genes induced by ATF4, Chac1 (Chen et al., 2017b), Trib3 (Ohoka et al., 2005), CHOP (Matsumoto et al., 1996), cysteinyl tRNA synthetase (CARS, (Hayano et al., 2016), and surprisingly the xCT cystine antiporter (Karuppagounder et al., 2016; Thorn et al., 2015) itself all present plausible schemes by which gene expression stimulated by ferroptotic stimuli could be integrated to promote cell death (Karuppagounder et al., 2016; Lange et al., 2008).
Figure 3.). The Cap N’ Collar transcription factor Nrf-2 has multiple modes of activation and can act non cell-autonomously (in glia) to protect adjacent neurons from ferroptosis.
A) Keap1 constitutively activates the proteasomal degradation of Nrf-2 under steady state conditions via the E3 ubiquitin ligase Cul3. In response to electrophiles capable of undergoing Michael Addition (sulphoraphane, dimethylfumarate), cysteine 151 on Keap1 is alkylated, thus altering the interaction between Keap1 and the degradation domain of Nrf2 (the Neh2 domain). Dissociation of this domain leads to Nrf-2 stabilization and consequent transcription of Nrf-2 dependent “antioxidant” genes.. (Adapted from Smirnova et al., 2011). Previous studies have proposed distinct mechanisms for distinct activators of Nrf-2. Specifcially, Nordihydroguarietic acid (NDGA) and fisetin either work by inhibiting a tyrosine kinase that leads to phosphorylation of Keap on Tyrosine 141 and destabilizes it. By contrast, gedunin is thought to directly disrupt interactions between Keap1 and Nrf-2, leading to Nrf-2 stabilization. TSA and geldanmycin appear to have similar kinetics and likely act as inhibitors of Hsp90. Canonical electrophiles such as TBHQ, sulforphane, pyrythiione and auranofin act as alkylators of Keap1 on cysteine 151, while inhibitors of Thioredoxin (cadmium) or specific 8 hydroxyquinolines appear to activate Nrf-2 by modifying the redox state of cysteines that coordinate Zn2+ (Cadmium) or remove Zn2+ from Keap1. In Smirnova et al. (2011), NDGA, gedunin, and fisetin all abrogate neuronal ferroptosis via activation of Nrf-2 in astrocytes.
B) Nrf-2 activation in astrocytes drives antioxidant and metabolic genes that support the ability of neurons to overcome proferroptotic signals. The model shows that glutathione (which is nearly a magnitude greater in astrocytes versus neurons) can be exported to the peri-neuronal space to bypass glutamate or glutamate analog induced xCT cystine antiporter blockade and augment cysteine and or glutathione levels in neurons (adapted from Shih et al., 2003). Glutathione release from astrocytes can bypass xCT inhibition via a host of potential intermediates that do not include cystine. Note the large gradient of intracellular and extracellular glutamate in neurons.
It should be noted that the role of ATF4 in mediating ferroptotic death may be context dependent. Indeed, agents that appear to drive ER stress (e.g. thapsigargin, tunicamycin, amino acid deprivation, cysteine aminoacyl tRNA synthetase deletion) and enhance ATF4 activation appear capable of preventing ferroptotic death (Hayano et al., 2016; Lange et al., 2008; Lewerenz and Maher, 2009). It may be that these ER stress inducers reduce sensitivity to ferroptotic death via their effects on global translation or by altering the heterodimeric transcriptional partners of ATF4.
Chac1 (Glutathione Specific Gamma Glutamyl Cyclotransferase) is a protein known to be regulated by ferroptotic stimuli in neurons as well as by ATF4 (Crawford et al., 2015) that catalyzes the cleavage of glutathione into 5-oxoproline and the cysteineglycine dipeptide. Accordingly, ferroptosis may lead to initial depletion of glutathione via inhibition of cystine transport that triggers feed-forward, morbid enhancement of glutathione depletion via activation of ATF4 and Chac1 expression. Consistent with this model, germline deletion of ATF4 in mouse cortical neurons has been shown to preserve glutathione levels in response to glutamate or the glutamate analog, HCA, by inhibiting degradation, not synthesis (Lange et al., 2008). As glutathione and glutathione peroxidase act together to neutralize reactive lipid species in ferroptosis, degradation of glutathione may reflect the commitment of neurons to irreversible cell death via the degradation of a prosurvival molecule (e.g. glutathione). In this context, the action of Chac1 could be analogous to the cleavage of bcl-2 or bcl-x by caspases to ensure apoptotic death once it is activated (Kirsch et al., 1999). As Chac1 generates 5-oxoproline as one of its products, future studies could look at 5-oxoproline generation during ferroptosis as a biomarker of Chac1 activity in vitro and in vivo.
In addition to Chac1, other putative, pro-death proteins are regulated downstream of ATF4. For example, Tribbles homolog 3 (Trib3) is induced in neurons in response to ferroptotic stimuli via an ATF4 dependent pathway (Figure 2). Trib3 was originally identified by Montminy and colleagues as a pseudokinase inhibitor of Akt which was activated downstream of insulin as a mechanism of feedback inhibition (Du et al., 2003). Inhibition of Akt has not been implicated in ferroptotic death. Rather it was shown that transient inhibition of Akt by pharmacological inhibitors forestalls ferroptosis so the role of Akt inhibition by Trib3 in ferroptosis remains unclear (Levinthal and DeFranco, 2004). However, there are other mechanisms by which Trib3 could induce ferroptosis. Specifically, Trib3 has also been postulated to induce death by reducing levels of the mitophagy regulator Parkin (Aime et al., 2020). The ability of ferroptotic stimuli to reduce Parkin could link ATF4 induction and diminished mitophagy with robust effects of bid translocation in mediating Drp-1 mitochondrial fragmentation, increased ROS, and loss of mitochondrial membrane integrity (Neitemeier et al., 2017; Tobaben et al., 2011).
Trib3’s toxic effects may also depend on its ability to dimerize with the prosurvival and pro-plasticity transcription factor CREB to reduce transcription of its target genes. As expected from these collective observations, the molecular deletion of Trib3 has been shown to reduce cell death in cell death models in vitro as well as in a global ischemia model in mice where ferroptosis has not been implicated (Wei et al., 2017).
CHOP was originally identified as a pro-death transcription factor whose basal expression is low but can be induced by a host of stresses including ER stress and DNA damage (Wang et al., 1996). ATF4 drives the upregulation of CHOP (a C/EBP family member) in ferroptosis, and like ATF4, CHOP possesses a leucine zipper dimerization domain and a basic DNA binding motif. In neuronal ferroptosis it appears to heterodimerize with ATF4 to bind CARE response element sites and drive proferroptotic genes. Among other genes with known pro-death activities downstream of ATF4 that could contribute to its integrated, negative effect on neuronal survival are the Regulated in development and DNA damage responses gene (REDD1, also known as RTP801 or DDIT4), a negative regulator of prosurvival mTOR activity (Malagelada et al., 2008) and cysteinyl tRNA synthetase (CARS) (Hayano et al., 2016) (Figure 2). Future studies will evaluate whether reduction of CHOP, Trib3, Chac1, REDD1, or CARS expression can separately regulate and inhibit neuronal ferroptosis, but we favor a model in which it is the integrated effect of these and other genes that ultimately determines whether a cell undergoes death (Figure 2).
Potential non-cell autonomous (glial) regulators of neuronal ferroptosis or excitotoxicity mediated by the transcriptional activators ATF4 or Nrf2
An important feature of nervous system function and dysfunction is the metabolic and trophic coupling between glial cells (astrocytes, oligodendrocytes, and microglia) and neurons that can lead to neuronal survival or neuronal death, specifically ferroptosis. Recent studies have highlighted the important role that astrocytes can play in modulating neuronal viability and how this can change at distinct stages of development (Liddelow et al., 2017). Studies of neuronal ferroptosis reflect a similarly complex but biologically meaningful relationship between glia and neurons. Microarray studies of the the two molecular components of the xCT cystine antiporter, SCL7A11 (the main 12 span transmembrane transporter) and SCL3A2 (4F2hc, a single pass membrane protein) showed that both are induced by ferroptotic stimuli in an ATF4-dependent manner (Lewerenz et al., 2013). The xCT cystine antiporter is highly expressed in glial cells (Shih et al., 2003) raising the possiblity that molecular reduction of ATF4 leads to complete resistance of neuronal cultures by reducing astroglial xCT(Lange et al., 2008). Consistent with this model, forced expression of ATF4 into glia or glial-based tumors was shown to enhance ferroptosis of cocultured neurons via xCT cystine antiporter mediated release of glutamate (Chen et al., 2017a). Deletion of the xCT cysteine antiporter can protect against a host of acute insults in vivo likely because of resultant diminished export of glutamate out of glia into extrasynaptic or synaptic sites would reduce excitotoxic cell death (Fogal et al., 2007). These findings are consistent with a distinct, non-cell-autonomous model by which ferroptotic stimuli in astrocytes might induce death in neurons. In diseases such as cerebral ischemia, ALS, Huntington’s disease, and Alzheimer’s disease, diverse pathobiologies may lead to accumulations of glutamate in extrasynaptic sites (Hardingham and Bading, 2010). Modest increases in glutamate would be expected to then inhibit cystine uptake via the xCT cystine antiporter in glial cells leading to cysteine and glutathione depletion. Depletion of cysteine or glutathione can activate GCN2 or PERK activation, eif2α phosphorylation, increases in ATF4 and maladaptive, feed-forward release of glutamate, ultimately leading to excitotoxicity (Lewerenz et al., 2013). Such a model provides a plausible scheme by which a ferroptotic signaling insufficient to trigger death in astrocytes could trigger excitotoxic death in adjacent neurons. Excitotoxic neuronal death appears to be independent of ferroptosis (Alim et al., 2019). However, ferroptotic inhibitors have been shown to inhibit excitotoxicity in vitro and in vivo; these agents might act by working on astrocytes or immune cells to inhibit neuronal death (Dixon et al., 2012; van Leyen et al., 2014).
In contrast to observations that show astrocytic xCT cystine antiporter activity can lead to the release of glutamate and excitotoxicity in mature neurons, immature neurons without glutamate receptors respond quite differently to xCT cystine antiporter activity on proximal glial cells. Specifically, meningeal cells and astrocytes, which possess high levels of plasma membrane xCT antiporter activity, are able to protect immature neurons from ferroptosis even when they are present at cellular ratios of 1:100 (Shih et al., 2006; Shih et al., 2003). This remarkable potency likely reflects the significantly higher antioxidant potential of astrocytes or meningeal cells as compared to immature neurons. Accordingly, excess glutathione released from astrocytes likely feeds pathways of cysteine delivery into neurons that can bypass xCT cystine transport inhibition and abrogate glutamate-induced ferroptosis (Dringen et al., 2000; Sun et al., 2005).
Nrf-2 is a Cap’n’Collar transcription factor which is constitutively degraded in the cytoplasm by an inhibitor protein called Keap1 (Kelch-like Ech-associated protein 1), which constitutively facilitates the interaction of Nrf-2 with a Cul3-Rbx1 E3 ubiquitin ligase that facilitates Nrf-2’s polyubiquitination and proteasomal degradation (Smirnova et al., 2011). An electrophilic (oxidative) attack of target cysteines on Keap1 disrupts its interactions with Nrf-2 at the Neh2 domain, thus leaving the complex stabilized (Figure 3A). Stabilized Nrf-2 can translocate into the nucleus to activate target genes including those involved in glutathione synthesis (e.g., the xCT transporter, gamma glutamylcysteine synthetase, and glutathione reductase) as well as electrophile counterattack (Glutathione S-transferases) (Kraft et al., 2004).
Electrophile activators of Nrf-2 include the FDA approved multiple sclerosis agent dimethyl fumarate whose trade name is Tecfidera (Brennan et al., 2016), the nutraceutical sulforaphane (Figure 3A; highly enriched in broccoli sprouts (Benedict et al., 2012), and the endogenous, cyclooxygenase-derived prostaglandin, 15-deoxy-Delta(12,14)-PGJ(2) (Haskew-Layton et al., 2013)). In the early 2000s the Murphy and Johnson laboratories teamed up to address the role of Nrf-2 in regulating glutamate-induced ferroptosis in primary neurons (Lee et al., 2003; Shih et al., 2003). Consistent with differences in the antioxidant potential of astrocytes compared to neurons, they showed that levels of an antioxidant response element reporter were activated at a much higher level in astrocytes than neurons in vitro and in vivo (Figure 3B) (Murphy et al., 2001). The forced expression of Nrf-2 in astrocytes protected against neuronal ferroptosis even when astrocytes were less than 2% of the total population of neurons in the dish (Shih et al., 2003). Molecular and biochemical studies demonstrated that this protection was due to the increased synthesis and release of glutathione from astrocytes to the culture medium (Figure 3B).
Glutathione can be degraded by cell surface ectopeptidases to its component parts or it can be taken up directly into neurons to bypass the xCT transport blockade and to protect neurons from ferroptosis (Dringen et al., 1999; Sagara et al., 1993a; Sagara et al., 1993b; Shih et al., 2003). It could be argued that forced expression of Nrf-2 in astrocytes (Figure 3B), which governs a large cassette of genes including the xCT transporter, is a superior way to overcome glutathione deficiency in neurons as compared to overexpression of the xCT cystine antiporter alone, as the rate-limiting enzymes in glutathione synthesis and oxidant detoxification are also induced by Nrf-2.
In addition to astrocytes, the forced expression of Nrf-2 in neurons also protect against ferroptosis, suggesting that putative inhibitors that act stoichiometrically govern the dramatically different Nrf-2 activity in neurons (Satoh et al., 2006; Shih et al., 2003). While Nrf-2 activation following injury occurs primarily in astrocytes, it remains unclear as to whether ferroptotic signals in astrocytes or neurons are sufficient to drive Nrf-2 expression. As expected, findings from some studies have failed to establish peroxide as a driver of Nrf-2 in astrocytes (Haskew-Layton et al., 2010). Nevertheless, the authors of recent studies have implicated BDNF or NGF as drivers of ceramide-mediated PkCζ activation upstream of Nrf-2 in astrocytes or AKT-mediated Nrf-2 activtion in PC12 cells (Ishii et al., 2019; Xia et al., 2015). Growing recognition of lipid electrophiles generated by ferroptotic stimuli, as discussed below, represent a whole new array of putative candidates for in vivo activators of the Nrf-2 pathway.
Ferroptosis in neurons: follow the lipids
One of the many attractive features of studying glutamate-induced xCT antiporter inhibition leading to glutathione depletion-induced ferroptosis in immature neurons is that oxidative stress-induced cell death is caused by physiological depletion of a versatile antioxidant rather than by the addition of non-physiological concentrations of oxidants. Given glutathione’s role as a cofactor for many enzymes, including glutathione reductases, glutathione transferases, and glutathione peroxidases, it was assumed that xCT cystine antiporter inhibition leading to glutathione insufficiency and cell death was caused by many oxidant species including superoxide, hydrogen peroxide, and lipid peroxides. Initial studies excluded hydrogen peroxide as a mediator of ferroptotic death in neurons, and instead reinforced the role that amino acid deprivation could play in modulating sensitivity to death (Esch et al., 1998).
A key set of observations came from the Schubert and Maher laboratories that focused attention on arachidonic acid metabolizing enzymes, specifically, 12/15 lipoxygenase, as essential mediators of ferroptosis in neurons (Li et al., 1997). Arachidonic acid is a 20-carbon polyunsaturated omega-6 fatty acid with unsaturated double bonds at carbons 5, 8,11, and 14 (Yang et al., 2019) (Figure 5A). It is esterified in phospholipids in the brain such as phosphatidyl inositol and phosphatidyl choline. Following its release from membranes by phospholipases, it can then be metabolized by lipoxygenases, cyclooxygenases, or epoxygenases to a host of reactive lipid species (Kagan et al., 2017). Using a chemical biological approach, the Schubert team showed that a non-metabolizable analog of arachidonic acid, eicosatetraynoic acid (ETYA), could completely abrogate glutamate-induced ferroptosis in immortalized hippocampal neuroblasts (Li et al., 1997). Selective inhibitors of each of the arachidonate-metabolizing enzymes were used to show that lipoxygenases mediate ferroptotic death but not cyclooxygenases and epoxygenases (Li et al., 1997). Specific lipoxygenases that incorporate oxygen into the 5, 12, and 15 positions of arachidonic acid are present in the nervous system, indicating the potential for arachidonic acid metabolism to contribute to ferroptotic death. Indeed, pharmacological or molecular deletion of 12/15-lipoxygenase abrogated ferroptosis induced by glutamate or its analogs (Khanna et al., 2003). Tocotrienol, a component of vitamin E, was shown to potently inhibit ferroptosis in neurons. Results from in silico modeling suggested that tocotrienol prevents arachidonic acid from entering the catalytic active site of 12/15-lipoxygenase (Khanna et al., 2003). Collectively, these were profound observations as they suggested that depleting a versatile and broad antioxidant such as glutathione in neurons leads to cell death via a specific, enzymatically generated lipid signaling pathway involving 12/15-lipoxygenase and not via multiple, distinct oxidant species (e.g., hydrogen peroxide, nitric oxide, etc.). They further suggested that lipid peroxidation occurs via an enzyme catalyzed reaction, not via random interactions between oxidants and lipids (Samhan-Arias et al., 2011), although other models have also been supported (Zilka et al., 2017).
Figure 5.). Persistent Erk signaling leads to ferroptosis via nuclear translocation of Erk and activation of gene expression.
Hyperactivation of Erk signaling can come from excessive growth factor stimulation, mutations in Ras or Raf, or via oxidative inactivation of Erk phosphatases. Phosphorylation of Erk by MEK leads to its nuclear translocation by binding to importin 7. Indeed, inhibiting Erk activation via the MEK inhibitor, UO126, completely inhibits neuronal ferroptosis. Of note, inhibiting nuclear translocation of activated Erk also blocks cell death suggesting that it is the nuclear translocation that is critical for death.
In the nucleus, Erk activation activates transglutaminase, an enzyme capable of polyaminating or monaminating glutamines on epigenetic proteins to modulate transcription. Inhibiting transglutaminase with cystamine or other structurally diverse TG inhibitors abrogates neuronal ferroptosis. Additionally, Erk can phosphorylate cMyc leading to its stabilization and activation. Mithramcyin, a DNA binding drug which reduces cMyc expression by binding to Sp1 sites in its promoter also abrogates neuronal ferroptosis. While gene expression downstream of Erk activation leading to ferroptosis has not been definitively shown, It is tempting to speculate that Erk activated cMyc triggers ferroptosis by driving expression of genes that increase iron uptake into the cell. Figure adapted from (Giltnane and Balko, 2014) with permission
These initial findings have been amplified and clarified by elegant lipidomic and protein biochemical studies from the Kagan and Bayir laboratories. They showed that in response to ferroptotic stimuli, 15-lipoxygenases utilize polyunsaturated phosphatidylethanolamines rather than free polyunsaturated fatty acids as substrates. This change in substrate specificity is mediated by a scaffold protein called phosphatidylethanolamine binding protein 1 (PEBP1) which can associate with both forms of 15-lipoxygenase to foster the generation of hydroperoxyl-eicosotetranoic acid phosphatidylethanolamines (HpETE-PE) (Wenzel et al., 2017) (Figure 4C). While the specific mechanisms by which HpETE-PE induce death remains unclear, they decompose easily to electrophiles capable of reacting avidly with nucleophilic protein targets (Anthonymuthu et al., 2018; Wenzel et al., 2017). These lipid electrophiles are selectively detoxified by one glutathione-dependent enzyme, glutathione peroxidase 4 (GPX4). Accordingly, Conrad and colleagues elegantly elucidated that the inducible deletion of GPX4 in neurons induces ferroptosis in vitro and in vivo (Figure 4B; a process that is dependent on 12/15- or 15-lipoxygenase activity (Seiler et al., 2008). In this context, GPX4 deficiency could be rescued by lipid peroxidation inhibitors such as a-tocopherol, consistent with an upstream role for oxidized lipids in cell death (Figure 4C). Glutathione and one of its associated enzymes, GPX4, thus qualifies as the Bcl-2 equivalent of ferroptotic death. Congruent with this view, the authors of recent studies have highlighted the role that inducers of GPX4 expression can play in abrogating ferroptotic death (Alim et al., 2019). Taken together, the results suggest that in the absence of glutathione, GPX4 levels are not adequate to compensate for increased lipid peroxidation. However, if GPX4 levels are augmented, GPX4 can still prevent death despite the depletion of glutathione. It may be that other thiols can provide electrons for GPX4 functioning in the absence of glutathione or that GPX4 inhibits the transcription of genes like Chac1 that enhance glutathione turnover. Additional studies have shown that the selenocysteine critical to GPX4’s anti-ferroptotic activity is required to prevent hyperoxidation of the enzyme by exogenously added oxidants (Ingold et al., 2018) (Figure 5A and B).
Figure 4.). Glutathione peroxidase 4 (GPX4), a selenoprotein, acts with reduced glutathione to neutralize reactive phospholipids that trigger ferroptosis.
A.) Selenocysteine (right) is the 21st amino acid and differs from cysteine (middle) in having a selenium molecule (Se) in the place of a sulfhydryl (SH). Selenocysteine has a lower pkA than cysteine and a lower reduction potential. It therefore is oxidized and reduced more rapidly than cysteine. The advantage of these properties is that proteins containing selenocysteine are much less like to be inactivated by oxidants than cysteine containing proteins, thus making them ideal antioxidant enzyme. Selenocysteine differs from serine (left) in possessing a Se rather than a hydroxyl group. A charged serine-tRNA is phosphorylated and then used to form the Selenocysteine-tRNA.
B.) GPX4 is found in multiple subcellular localizations including cytoplasm, nucleus, and mitochondria. It is unclear at this time which localization is required to interdict ferroptosis, but recent studies have shown that pharmacological selenium can drive expression of the nuclear and mitochondrial forms of GPX4 and interdict ferroptosis induced by erastin, HCA, or hemin in neurons (Alim et al.,2019).
C.) Under steady-state conditions cystine enters the neuron via the xCT cystine antiporter in exchange for glutamate. It is reduced intracellularly to cysteine by glutathione reductase (GR). Cysteine is then covalently linked to glutamate by gglutamylcysteine synthetase (GCS). Glycine is finally added to the dipeptide to form γglutamylcysteineglycine (glutathione) by glutathione synthetase (GS). Glutathione is a stable reservoir of electrons that can be utilized by GPX4 to neutralize reactive lipids to non-toxic alcohols. Following a ferroptotic stimulus 15 HpETE-PE is formed and ferroptosis ensues. Augmenting glutathione or GPX4 can convert toxic HpETE-PE to a non-toxic HETE-PE and the cell survives.
GPX4-regulation of cell survival may play an active role in maintaining brain homeostasis even in the absence of classic ferroptosis-inducing stimuli. GPX4 knock-in mice with the selenocysteine of GPX4 mutated to a cysteine alone (Figure 4A) are found to die early in life (depending on the genetic background). Remarkably, death from the substitution of cysteine for selenocysteine in GPX4 appears to occur due to selective loss of a class of parvalbumin-positive inhibitory neurons leading to uncontrolled epilepsy (Ingold et al., 2018). One could speculate that parvalbumin-positive interneurons have higher concentrations of polyunsaturated fatty acids, other oxidants, or diminished expression of GPX4, but the precise reason for their selective vulnerability is unknown. Of note, mice possessing selective deficiency of the selenocysteine tRNA which incorporates selenocysteine into proteins also show selective loss of parvalbumin-positive interneurons (Wirth et al., 2014). Additional insights as to why inhibitory interneurons are selectively vulnerable to the loss of selenocysteine in GPX4 will be of interest to those studying autism, brain development, epilepsy, and brain recovery fields where modulating the excitatory-to-inhibitory balance may be essential. Moreover, results from recent studies suggest that pharmacological levels of selenium can drive the expression of GPX4 (in multiple subcellular compartments; Figure 4B) and other selenoproteins to induce resilience to ferroptosis (Alim et al., 2019).
Ras-MEK-ERK signaling and neuronal ferroptosis
Another feature of neuronal ferroptosis that distinguished it in its infancy from the classical neuronal apoptosis in sympathetic neurons was that it was enhanced by, rather than inhibited by, growth factor stimulation (Ratan et al., 1996). Mechanistic clarity of this unexpected phenomenon emerged from studies by scientists in the DeFranco Laboratory who showed that glutamate-induced ferroptosis in hippocampal neuroblasts required the Ras-MEK-ERK pathway for its execution. Indeed, neuronal ferroptosis was abrogated by geldanamycin, an HSP90 inhibitor, even when the drug was delivered hours after exposure to glutamate (Xiao et al., 1999). Via a candidate-based approach, they showed that geldanamycin-induced protection of neuronal ferroptosis was associated with decreases in the levels of an oncogenic signaling protein called c-RAF1 that is an upstream modulator of the ERK1/2 serine/threonine kinases (Figure 5). Ras activation can be triggered by a number of growth factors or by products of 12/15-lipoxygenase (Roskoski, 2012; Stanciu et al., 2000). Activated Ras recruits Raf to the plasma membrane and together Ras-Raf activate MEK, which in turn phosphorylates tyrosine and threonine residues in the activation loop of ERK1/2 kinases in the cytoplasm (Kidger et al., 2018). Activated ERK1/2 kinases can phosphorylate many proteins involved in cell proliferation, axonal growth and tumorigenesis. DeFranco and colleagues expanded this biology to include ferroptosis. Indeed, ferroptotic stimuli in hippocampal neuroblasts (e.g. glutamate) lead to persistent activation (e.g. phosphorylation) of ERK1 (44 kD) and ERK2 (42 kD) near the time of cells to commit to death. Inhibition of MEK via a dominant negative MEK or the selective inhibitor (U0126) but not its inactive analog (U0124) abrogated ERK1/ERK2 phosphorylation (Figure 5) and prevented cell death (Luo and DeFranco, 2006; Stanciu et al., 2000). Of note, U0126 is known not to inhibit ERK, protein Kinase C, c-Raf1, Abl, JNK/SAPK, Cdk2, Cdk4, SEK, MKK-3, MKK-4, or MKK-6. Subsequent studies showed that once activated, ERK is phosphorylated by MEK1 on its TXY motif which allows it to dissociate from its scaffolding proteins including MEK, B-arrestin, and SEF1 (Berti and Seger, 2017). Detachment from scaffolding proteins exposes two serine residues (SPY motif) within the nuclear translocation signal of the ERK1/2 kinases. Phosphorylation in this domain allows ERK1/2 to bind to importin 7 which shepherds ERK1/2 into the nucleus where it can phosphorylate established transcriptional targets (e.g. cMyc, Elk1, and cfos; Figure 6). Under steady-state conditions, ERK1/2 is dephosphorylated by a dual specificity phosphatase, MKP3, that dephosphorylates tyrosines and threonines in the TXY domain of ERK1/2. This results in ERK1/2 translocation back to the cytoplasm. The persistence of nuclear ERK1/2 kinase activity results from reversible but persistent and directed redox-mediated inhibition of MKP3 phosphatases targeting ERK1/2 but not JNK or p38 Map Kinase phosphatases during glutamate-induced ferroptosis (Levinthal and DeFranco, 2004; Stanciu and DeFranco, 2002). This inhibition likely involves oxidation of a target cysteine within the MKP phosphatases by reactive lipid electrophiles produced from 15 hydroperoxyl-phosphatidylethanolamines, but this has not been formally tested (Figure 5; Levinthal and DeFranco, 2004). Cysteines within tyrosine and some serine/threonine phosphatases are particularly nucleophilic because their local environment reduces the pKa of the catalytic cysteine. As expected from this model, the forced expression of MKP3, a dual specificity phosphatase specific for ERK2, prevented cell death. Remarkably, a catalytically inactive MKP3 mutant which does not reverse ERK1/2 phosphorylation also prevented cell death (Levinthal and DeFranco, 2004) but its protection was correlated with inhibition of Erk nuclear transport and inhibition of Elk transcriptional activity (a known ERK nuclear target).
Figure 6.). Iron chelators inhibit iron dependent dioxygenases, the hypoxia inducible factor (HIF) prolyl hydroxylases to prevent proferroptotic gene expression and cell death.
A) Iron chelators can trigger adaptive transcriptional responses to hypoxia. Message levels for Hypoxia inducible factor-1α are constitutively expressed and translated into protein in normoxic cells. Furthermore HIF-1α is hydroxylated at Proline 403 and proline 564 in its oxygen dependent domain. Hydroxylation of HIF-1α enhances its affinity for an E3 ubiquitin ligase, the Von Hippel Lindau (VHL) protein. VHL tags HIF-1α with polyubiquitins targeting the protein for proteasomal degradation (Image on right). Under conditions of hypoxia or with small molecule Iron chelators, the HIF PHDs fail to function and HIF is neither hydroxylated nor is it proteasomally degraded. Instead it dimerizes with its constitutively expressed partner (HIF-1β), translocates to the nucleus where it activates more than 100 genes that act to alleviate the discrepancy in oxygen demand an supply. Structurally diverse HIF PHD inhibitors, including those that did not bind iron, also prevented ferroptosis, suggesting that HIF PHDs and not free iron are the target for iron chelator mediated protection of ferroptosis in neurons.
B) Structurally diverse HIF PHD inhibitors inhibit ferroptosis by inhibiting proferroptotic, ATF4-dependent transcription not by activating HIF-dependent transcription. Our initial model was that HIF PHD inhibitors were activating HIF dependent prosurvival genes to prevent ferroptosis. However, we found that molecular reduction of HIF-1α or HIF-2α did not abrogate the ability of iron chelators to inhibit ferroptosis (left model). A more selective HIF PHD inhibitor was identified via a screening campaign involving a cell based reporter.It was a branched oxyquinoline called adaptaquin which inhibited HIF PHDs selectively and did not effect other iron or zinc dependent enzymes linked to ferroptosis (e.g., 12/15 lipoxygenase). Unbiased transcriptomics correlated the protective effects of adaptaquin with reduction in ATF4-dependent genes (right model) not increases in HIF-dependent genes, consistent with prior results on HIF-independent protection by HIF PHD inhibitors.
C) Model for how ferroptotic stimuli lead to increases in iron and loading of iron-dependent metalloenzymes to trigger ferroptotic gene expression in neurons. Ferroptotic stimuli lead to glutathione depletion. In addition to its role as a cofactor in the glutathione dependent enzymes, glutathione has also been implicated in iron sulfur cluster biosynthesis. In this scheme, loss of glutathione releases iron from the iron sulfur clusters where it is released into the cytoplasm where it is taken up by PCPB1, a chaperone that loads iron onto iron free forms of HIF prolyl hydroxylase-1. In this way, the model predicts that the iron-free form of HIF PHD1 acts as an iron sensor and along with other stimuli, drives the activity of proferroptotic ATF4 gene expression. This model suggests that iron participates in ferroptosis as a signaling molecule not as a toxin.
Nuclear translocation of ERK1/2 appears critical for its cell death promoting functions in ferroptosis (Levinthal and DeFranco, 2004; Stanciu and DeFranco, 2002). Translocation of ERK1/2 to the nucleus was inhibited by prior heat shock or cycloheximide treatment, both of which protected neuronal cells from ferroptosis. Additionally, nuclear ERK1/2 can phosphorylate proto-oncogenic c-Myc (myelocytomatosis oncogene) on serine 62, leading to its stabilization and increased transcriptional activity (Figure 5). The DNA binding drug and c-Myc expression inhibitor mithramycin (Figure 5), which binds to the minor groove of GC-rich DNA, but not AT-rich DNA) potently inhibited neuronal ferroptosis (Chatterjee et al., 2001). Chemical biological approaches showed that mithramycin inhibited ferroptosis, whereas a mithramycin analog with no ability to bind DNA had no protective effect (Sleiman et al., 2011). The effects of protective mithramycin analogs (with varying affinities for DNA) correlated with their ability to reduce transcription of c-Myc and other genes in its network. Indeed, molecular reduction of c-Myc via a peptide inhibitor that disrupts Myc/Max interactions or via shRNA to c-Myc itself also prevented neuronal ferroptosis. While it is unclear which specific c-Myc-regulated genes control ferroptosis, it is well established that c-Myc can bind to its canonical response element (E-boxes, CANNTG) present in the promoters of Transferrin Receptor 1 (Tfr1), ferritin (heavy and light chains), IRP2 (iron regulatory protein 2), and Nramp1 (natural resistance associated macrophage protein 1) genes which coordinately regulate iron metabolism (Dang, 2012; Wu et al., 1999) (Figure 5). IRP2 was identified as one of the genes that could be reduced to prevent ferroptosis in cancer cells (Dixon et al., 2012). Moreover, given the ability of c-Myc to regulate mitochondrial biogenesis, c-Myc can also regulate genes such as citrate synthase, F1/F0 ATPase subunit 3, and acyl coA synthetase family member 2 whose molecular reduction also protects tumor cells from ferroptosis (Dixon et al., 2012).
In addition to transcription factors such as c-Myc, epigenetic modulators of ferroptosis may be activated downstream of ERK1/2. A candidate family of novel epigenetic modulators activated by ERK1/2 is the transglutaminases, which are enzymes that can add monoamines or polyamines to histones or transcription factors to promote transcription (Farrelly et al., 2019). Transglutaminase inhibitors abrogate glutamate-induced ferroptosis (Figure 5) even when added 14 hours after exposure to a ferroptotic stimulus (Basso and Ratan, 2013). ERK1/2 inhibitors block increases in TG1 and TG2 message and activity. By contrast, inhibitors of TGs have no effect on ERK activity. Converging lines of inquiry suggest that TGs can post-translationally modify histones or transcription factors to drive or repress transcription (Farrelly et al., 2019; McConoughey et al., 2010). However, whether they can modify transcription factors such as ATF4 or c-Myc that have been implicated in pro-death responses to neuronal ferroptosis in neurons is not established.
Iron in ferroptosis: More than a namesake
Dyshomeostasis of iron is observed in neuropathological specimens for almost every neurological condition including stroke, multiple sclerosis, Alzheimer’s disease, Parkinson’s disease, Friedreich’s ataxia, epilepsy, and restless legs syndrome (Ward et al., 2014). Because these changes are observed in human tissue, it is unclear whether changes in iron are causally related to neuronal dysfunction or death, or if changes in iron are simply the consequence of neurons or other cell types that are degenerating. Confidence that iron might be causally related to disease pathogenesis came from studies employing small molecule chelators of iron in animal models of neurological disease and in humans with Alzheimer’s disease (Barnham and Bush, 2014). In these studies, deferoxamine, a clinically approved iron chelating agent, showed benefit not only in animal models but also when delivered intramuscularly over two years in patients with Alzheimer’s disease. These observations created enormous opportunity for the therapeutic application of iron chelators but also highlighted challenges in their translation to humans.
Since iron is a cofactor in numerous enzymes in the nervous system including iron-sulfur cluster proteins in the mitochondria, heme-containing proteins such as hemoglobin or neuroglobin, and oxo-di-iron enzymes such as ribonucleotide reductase, it is important to understand the mechanism of action of iron chelation so one can understand how to nullify toxic levels of iron without interdicting physiological iron-dependent metalloenzymes. Initial hypotheses postulated that iron is toxic to neurons via its ability to act as a catalyst for the production of hydroxyl radicals from peroxide or peroxyl radicals from lipid peroxides also known as Fenton Chemistry, but the absence of good in vitro models of iron-mediated toxicity limited progress in understanding the mechanism of protection by iron chelators (Bergsland et al., 2019; Lee et al., 2019). Volpe and Rosenberg used a model of glutamate-induced ferroptosis in immature oligodendrocytes to show that iron chelators are protective (Yonezawa et al., 1996). They concluded that iron chelators protect from ferroptosis by inhibiting Fenton chemistry but did not explore any of the predictions of this model in detail (Winterbourn, 1995). Results from subsequent studies from our laboratory used immature primary neurons and confirmed that iron chelators could abrogate ferroptosis, but our data resulted in a model distinct from Fenton chemistry for how iron chelators abrogate ferroptosis (Zaman et al., 1999).
Studies of protective mechanisms of iron chelators in ferroptosis that are independent of classical Fenton chemistry focused attention on hypoxic adaptation. Under conditions of hypoxia (e.g. the condition where oxygen supply and demand are discrepant), a host of adaptive transcriptional responses occur (Johnson, 2016; Semenza, 2012). These include the transcriptional upregulation of erythropoietin to increase the oxygen-carrying capacity of blood in humans who are anemic or who reside at high altitudes (Wenger and Gassmann, 1997), transcriptional induction of vascular endothelial growth factor in the ischemic myocardium to enhance the synthesis of new vessels (Carmeliet and Baes, 2008), and transcriptional upregulation of glycolytic enzymes to allow for ATP generation in the absence of oxygen and oxidative phosphorylation (Semenza et al., 1994).
In a series of heroic experiments, Gregg Semenza and colleagues identified the DNA binding site (hypoxia response element) and the heterodimeric complex that binds to that response element to drive the expression of genes involved in adaptation to hypoxia (Wang and Semenza, 1993). The classical heterodimeric complex is composed of a 120 kD regulated subunit called hypoxia inducible factor (HIF)-1α and a 94 kD constitutively expressed subunit called HIF-1β. It is now known that other HIF isoforms exist and that HIF-1α and HIF-2α can regulate more than two hundred genes involved in hypoxic adaptation (Figure 6A). Over the past 25 years, we have learned a great deal about how HIF is regulated in normoxia and hypoxia. Seminal work emerged from the Kaelin and Ratcliffe labs (who shared the Nobel Prize in Physiology and Medicine in 2019 with Semenza) that revealed how iron chelators can drive the hypoxic transcriptional response to potentially protect from ferroptosis (Epstein et al., 2001; Ivan et al., 2001; Johnson, 2016). Under normoxic conditions, HIF-1α is hydroxylated on specific prolines (402 and 564). Hydroxylation of HIF-1α enhances its affinity for an E3 ubiquitin ligase, the Von Hippel Lindau (VHL) protein. Recruited VHL tags HIF-1α with ubiquitins which target the protein to the proteasome where it is degraded (Figure 6A).
Under conditions of hypoxia, oxygen-, iron-, and 2-oxoglutarate-dependent dioxygenases called the hypoxia inducible factor prolyl hydroxylases (HIF PHDs) cease to function because their Km for oxygen makes them sensitive to changes in oxygen tension. Accordingly, HIF-1α is not hydroxylated, ubiquitinated, or degraded. Hypoxia-stabilized HIF-1 dimerizes with its constitutively active partner and moves into the nucleus where it can drive a coordinated cassette of genes that act at the cellular, local, and systemic levels to alleviate a discrepancy in oxygen supply and demand (Epstein et al., 2001; Ivan et al., 2001). By binding iron, iron chelators can inhibit the HIF PHDs and “fool” the normoxic cell into thinking it is hypoxic and activate the adaptive HIF-dependent gene program (Zaman et al., 1999). Indeed, studies from our laboratory attributed protection from normoxic ferroptosis by iron chelators to selective inhibition of the HIF PHDs (Figure 6A). In this scheme, iron chelators can remove or prevent incorporation of iron into the HIF prolyl hydroxylases (Siddiq et al., 2005). As expected from this model, inhibitors of the HIF PHDs that do not bind iron (e.g., cobalt, dihydroxybenzoic acid, HIF PHD inhibitory peptides) also prevent ferroptosis. The structure-activity relationship of known chelators corresponds to their ability to inhibit the HIF prolyl hydroxylases and not to chelate-free iron (Siddiq et al., 2005; Zaman et al., 1999). Moreover, molecular reduction of the HIF PHDs in vitro or in vivo also protects neurons from ferroptosis (Karuppagounder et al., 2016; Siddiq et al., 2009).
In a screening campaign against a cellular HIF reporter used previously in vivo, our lab working with Irina Gazaryan’s laboratory identified adaptaquin, a branched oxyquinoline that could bind directly to the HIF PHDs and stabilize the HIF-1 (Figure 7B). Adaptaquin was named for its ability to drive the adaptive response to hypoxia (Karuppagounder et al., 2016; Smirnova et al., 2010). As expected, adaptaquin is a potent inhibitor of ferroptosis in neurons and cancer cells, and its effects cannot be attributed to inhibition of other metal-dependent enzymes linked to ferroptosis regulation including the lipoxygenases or histone deacetylases (Karuppagounder et al., 2016; Zille et al., 2019). Instead, we performed mechanistic studies in a classic model of neuronal ferroptosis and correlated adaptaquin’s protective effects with its ability to suppress ATF4-dependent transcription (Figure 6B), not with its ability to enhance HIF-dependent transcription (Aminova et al., 2005; Karuppagounder et al., 2016; Lange et al., 2008; Siddiq et al., 2009). Specifically, the dose of adaptaquin (1 μM) that provided 100 percent protection of neurons 14 hours after the onset of a ferroptotic stimulus did not activate HIF-dependent genes but suppressed a large cassette of ATF4-dependent genes (Karuppagounder et al., 2016).
Molecular and biochemical analyses showed that adaptaquin abrogated hydroxylation of ATF4 and this was correlated with reductions in ATF4-dependent gene expression (Figure 6B). The mutation of five conserved prolines on ATF4 showed that those modifications converted the wild-type transcription factor from inducing ferroptosis to having no effect on ferroptosis (Karuppagounder et al., 2016). Altogether, those studies favor a model in which iron and electrophilic lipids work in parallel rather than together to induce ferroptotic death. In this scheme, glutathione depletion induced by ferroptotic stimuli not only decreases GPX4 activity and leads to unopposed production of 15-hydroperoxy-phosphatidyethanolamines, but also releases iron from iron sulfur clusters formed via glutathione interacting with glutaredoxin 3 and/or glutaredoxin 4 (Figure 6C; Ratan, 2019). Iron released from the disruption of the glutathione-Grx3/Grx4 Fe/S cluster complex could then be taken up into ferritin or into iron chaperones (PCB1) for enzymes such as the HIF PHDs (Kumar et al., 2011; Nandal et al., 2011). Accordingly, iron could be toxic in ferroptosis not via its ability to foster the generation of reactive lipids, but rather via its ability to load iron-dependent enzymes (Figure 6C; HIF PHDs) present in the cytoplasm in their apo form that drive proferroptotic ATF4-dependent gene expression (Karuppagounder et al., 2016; Smirnova et al., 2010).
Ferroptosis in neurons and cancer cells: Is there a difference?
Cancer cells and post-mitotic neurons have a number of differences, including those related to metabolism, the make-up of the phospholipid bilayer, and the fact that one is mitotic while the other is post-mitotic (Heine et al., 2015). Since classical ferroptosis was identified in cancer cells with KRAS mutations, the question arises as to whether pathways identified in these cancer cells are similar to or different from those in neurons. Using erastin to inhibit the xCT antiporter and induce ferroptosis, Zille and colleagues show that ferroptosis in cancer cells and neurons (Table 1A) were identical (Zille et al., 2019). Moreover, almost all the agents identified in our laboratory as effective interventions for HCA (a glutamate analog)-induced death in primary neurons (e.g. mithramycin, transglutaminase inhibitors, adaptaquin) can interdict ferroptosis in cancer cells with one important exception: HDAC inhibitors (Table 1B). Simply, Histone deacetylase (HDAC) inhibitors are small molecules that enhance transcription by enhancing acetylation of lysine within the N-terminus of histone proteins. This change is associated with the unwinding of DNA and a greater propensity for gene transcription. Excitingly, HDAC inhibitors (which are in clinical trials in humans for cancer) appear to enhance death in cancer cells and prevent ferroptosis in neurons (Olson et al., 2015; Zille et al., 2019).
Table 1. Ferroptosis in cancer cells induced by erastin is similar to ferroptosis induced by erastin or HCA (glutamate analog) in primary neurons.
A.) Ferroptosis inhibitors abrogate ferroptosis in cancer cells (HT1080) and primary cortical neurons (PCNs). HT1080 cells were treated with 1mM erastin, PCNs with 5μM erastin or 5 mM glutamate analog HCA (all LD50) HCA and chemical inhibitors effective in ferroptosis were examined. Numbers show mean +/− SD at representative concentration in brackets. Gray scale coding indicates the continuum from no protection in the presence of erastin (black) to maximal cell viability (white). *P <.05 versus erastin or glutamate analog (HCA), #p <0.05 versus inactive analog U0124. (Table reproduced from Zille et al., 2019).
B.) Do known inhibitors of HCA (glutamate analog)-induced death in neurons prevent ferroptosis in cancer cells? HT1080 cells were treated with 1μM erastin (LD50) and chemical inhibitors effective in glutamate analog (HCA)-induced neuronal toxicity (values from published work and references indicated in last 2 columns) were examined. Numbers show mean +/−SD at representative concentration in brackets. Grayscale coding indicates the continuum from no protection in the presence of erastin (black) to maximal cell viability (white).*p<.05 versus erastin, #p<.05 versus Nullscript (negative control). (Table reproduced from Zille et al., 2019
| Cell Death by Ferroptosis | ||||||||
|---|---|---|---|---|---|---|---|---|
| Cell Death Inhibitor |
Target | Concentration Range |
HT1080 + 1 μM Erastin | PCN + 5 μM Erastin | PCN + 5 μM HCA | |||
| Representative Concentration |
Representative Concentration |
Representative Concentration |
||||||
| Actinomycin D | mRNA synthesis | 0.001 −1 μM | 62.15 ± 21.36 a | 10 nM | 73.44 ± 5.72 a | 1 nM | 45.88 ± 28.37 | 1 nM |
| Cycloheximide | Protein synthesis | 0.1- 50 μM | 49.77 ± 10.43 a | 1 μM | 76.55 ± 20.94 | 1 μM | 81.22 ± 8.62 a | 0.1 μM |
| Ferrostatin-1 | Canonical ferroptosis inhibitor,reactive lipid species (RLS) | 0.01 - 10 μM | 98.12 ± 12.20 a | 0.1 μM | 87.46 ± 11.04 a | 0.1 μM | 93.44 ± 6.08 a | 0.1 μM |
| Deferoxamine | Iron, hypoxia-inducible factor (HIF) prolyl hydroxylase domain-containing (PHD) inhibition | 10 – 100 μM | 82.25 ± 12.31 a | 10 μM | 85.60 ± 13.31 a | 50 μM | 81.00 ± 10.28 a | 50 μM |
| N-Acetylcysteine | Reactive oxygen species (ROS),RLS | 100 – 1000 μM | 110.26 ± 18.37 a | 500 μM | 112.10 ± 13.17 a | 100 μM | 102.50 ± 9.05 a | 100 μM |
| Trolox, vitamin E analog | RLS | 0.1 – 100 μM | 113.80 ± 9.41 a | 100 μM | 118.69 ± 7.02 a | 100 μM | 115.05 ± 10.84 a | 100 μM |
| U0126 | Mitogen-activated protein kinase kinase 1/2 (MEK 1/2) | 1 – 20 μM | 78.88 ± 4.76 a, b | 5 μM | 97.57 ± 5.56 a, b | 5 μM | 95.80 ± 8.21 a, b | 5 μM |
| U0124 | Inactive U0126 analog | 1 – 20 μM | 38.79 ± 16.11 | 5 μM | 34.39 ± 14.03 | 5 μM | 17.35 ± 9.23 | 5 μM |
| Vehicle | 30.93 ± 13.29 | 41.97 ± 15.77 | 37.10 ± 17.41 | |||||
| Cell Death Inhibitor |
Target | Concentration Range |
HT1080 + 1 μM Erastin | Published HCA or Glutamate in pCN |
References of published work |
||
|---|---|---|---|---|---|---|---|
| Representative Concentration |
Representative Concentration |
||||||
| Adaptaquin | Hypoxia-inducible factor (HIF) prolyl hydroxylase domain- containing (PHD) inhibition |
0.01 – 1 μM | 101.56 ± 28.55 a | 0.5 μM | 89.10 ± 7.56 a | 0.5 μM | Karuppagounder et al., 2016 |
| Mithramycin | Sp1 | 0.05 – 0.3 μM | 79.92 ± 17.35 a | 50 nm | 71.67 ± 7.57a | 200 nM |
Chatterjee et al., 2001 Ferrante et al., 2004 Sleiman et al., 2011 Sleiman et al., 2011 |
| Cystamine | Transglutaminase | 0.1 – 100 μM | 96.22 ± 17.75 a | 10 μM | 103.18 ± 0.86 a | 10 μM | Basso et al., 2012 |
| B003 | 25 – 200 μM | 87.79 ± 11.82 a | 200 μM | 98.83 ± 6.70 a | 200 μM | ||
| D004 | 12.5 – 100 μM | 45.82 ± 18.00 | 100 μM | 78.63 ± 29.17 a | 100 μM | ||
| Apicidin | Histone deacetylase |
0.01 −0.5 μM | 11.82 ± 4.35 a | 100 nM | 79.36 ± 10.17 a | 100 nM | Sleiman et al., 2014 Sleiman et al, 2011 |
| Butyrate | 0.1 – 10 μM | 16.71 ± 8.20 a | 5 mM | 76.84 ± 1.67 a | 5 mM | ||
| MS-275 | 0.5 – 5 μM | 22.32 ± 13.24 a | 1 μM | 64.17 ± 0.84 a | 100 nM | ||
| Scriptaid | 1 – 10 μM | 28.29 ± 6.75 | 5 μM | 70.00 ± 5.00a | 6.13 μM | Langley et al., 2008 | |
| Nullscript (Scriptaid negative control) | 1 – 10 μM | 24.29 ± 8.29 | 5 μM | 27.00 ± 0.96 | 6.13 μM | ||
| Vehicle | 34.23 ± 12.46 | 27.86 ± 6.12 | |||||
From these observations, several important conclusions can be drawn. First, mechanistic studies using inhibitors such as glutamate in neurons that occurred before the identification of ferroptosis as a distinct cell death pathway in 2012 provide equally important information about ferroptosis in multiple cell types. Second, as one of the goals of cancer chemotherapy is to kill cancer cells and protect neurons, HDAC inhibitors may be the perfect adjunctive therapies for pro-ferroptotic agents like erastin. Although it has not been formally tested, it is possible that if pro-ferroptotic agents penetrate into the CNS, they could cause toxicity in the form of chemo brain or peripheral neuropathy. Based on recent data (Zille et al., 2019), HDAC inhibitors could potentiate the pro-death action of erastin or other pro-ferroptotic agents in tumor cells while inhibiting their putative toxic side effects in central or peripheral neurons.
Relevance of ferroptosis inhibitors to neurological disease therapeutics
The applicability of chemical agents that protect against neuronal ferroptosis in vitro to disease models of neurological conditions in vivo has been extensive and robust. There are likely many reasons for this surprising observation. Among them the most direct reason may be the broad relevance of ferroptotic pathways to neuronal dysfunction and death in neurological conditions. As many of the pathways involving reactive lipids are also used by neurons to summon immune cells for damage consolidation and repair (Enyedi et al., 2016), it will be important to investigate whether the effects of ferroptosis inhibitors act cell-autonomously to inhibit death in neurons or act on other cell types to prevent neuronal ferroptosis.
A second feature that explains the robust translation to in vivo models is that ferroptosis modeling involves physiological depletion of an endogenous antioxidant rather than the non-physiological addition of an oxidant. For years, the oxidative stress field has been plagued by models of oxidative death that use high micromolar to millimolar concentrations of peroxide (or other oxidants) which cannot be achieved in an intact cell (Haskew-Layton et al., 2010). A third feature is that following exposure of erastin or glutamate to induce ferroptosis, there is a window of time when cells are operationally undergoing “oxidative stress”, an imbalance of oxidants and antioxidants in favor of oxidants, but when they can still be rescued. This property allows one to clearly delineate events that mediate death from those that are a consequence of death. Finally, ferroptotic death in cortical neuronal cultures from E15 mice embryos (2 DIV) is quite reproducible and concentrations of glutamate or erastin can be identified to define an LD50 or an LD80. Of note, almost all anti-ferroptotic agents defined to date in vitro that translate to neuroprotection in vivo show nearly complete protection at all concentrations of ferroptosis inducers. It is likely that the robustness of the neuroprotective effect that can be ascertained in vitro is an essential precursor to observe benefits in a more complex in vivo setting.
“Anti-ferroptotic” agents as neurological disease therapeutics
Among defined agents that have been translated from studies of ferroptosis in vitro to animals and in some cases humans, the first to be identified were activators of the transcriptional “antioxidant response” mediated by the transcription factor Nrf-2. Indeed, it was shown that electrophile inducers of quinone reductase (one of the genes activated by the Nrf-2 program) could condition neuroblastoma cells against a subsequent ferroptotic insult induced by glutamate (Murphy et al., 1991). Nrf-2 activators such as Tecfidera are now in clinical use to reduce exacerbations in multiple sclerosis and have shown promise in animal models involving other neurological conditions including ischemic or hemorrhagic stroke, Parkinson’s disease, and Huntington’s disease where ferroptotic death has been implicated (Sejbaek et al.,2019).
Agents that enhance delivery of cysteine via routes that bypass the xCT cystine antiporter can restore glutathione levels in response to ferroptotic stimuli such as erastin or glutamate and protect neurons (Dixon et al., 2012; Ratan et al., 1994b). N-acetylcysteine (NAC) protects against glutathione depletion and neuronal ferroptosis in vitro and has been shown to reduce cell death and behavioral deficits in a mouse model of brain hemorrhage where ferroptotic death has been observed (Karuppagounder et al., 2016). Ferroptosis has recently been described as a consequence of traumatic brain injury (Kenny et al., 2019). Accordingly, oral delivery of N-acetylcysteine to soldiers in Iraq less than 24 hours after a mild concussion from blast injury, resulted in symptom resolution within 7 days in 86% of those treated, whereas only 42% of those who received a placebo resolved during that same time period (Bhatti et al., 2018). Cystamine, which has been shown to increase glutathione levels in cortical neurons in vitro to inhibit ferroptosis (Karuppagounder et al., 2016), has also shown promise in animal models of Huntington’s disease (Karpuj et al., 2002), as well as hemorrhagic (Okauchi et al., 2009) and ischemic stroke (Li et al., 2015). Cystamine has the added potential benefit of enhancing Nrf-2 activation and inhibiting transglutaminase, both of which have been validated as targets for inhibiting ferroptotic death.
Among the inhibitors of reactive lipid species generation and toxicity, α-tocotrienol, a component of vitamin E, has shown promise in lisencephalic and gyrencephalic animal models of ischemic stroke and in vivo models of Parkinson’s disease (Khanna et al., 2003; Sen et al., 2007). Ferrostatin reduces cell death and improves functional recovery following hemorrhagic stroke in mice (Li et al., 2017). Future studies with newly developed, brain-penetrant forms of ferrostatin may allow its evaluation in other neurological disease models. It has been pointed out, however, that one of the challenges of tocotrienol or ferrostatin is the absence of a specific target. As such, 12/15-lipoxygenase, a documented mediator of ferroptosis in multiple cell types, may represent a specific druggable target in disease models where ferroptosis has been implicated in vivo. Germline deletion of 12/15-lipoxygenase can reduce infarct size following stroke (Conrad et al., 2018; Karatas et al., 2018). Increases in 12/15-lipoxygenase are observed in neurons and endothelial cells; selective inhibitors of the enzyme not only reduce infarct size but also reduce the hemorrhagic transformation that is associated with brain infarction (Zheng et al., 2019) . Whether cell death in neurons or endothelial cells following cerebral ischemia meets all the criteria of ferroptosis is as of yet uncertain, but these studies suggest further tests of this hypothesis are in order.
Of note, in establishing a screening campaign for 12/15-lipoxygenase inhibitors (e.g. LOXBlock1) for stroke, it was discovered that some inhibitors that are effective in mice do not work on the human enzyme (Armstrong et al., 2016; Yigitkanli et al., 2017). In contrast to ischemic stroke, teams of scientists from multiple laboratories have provided converging lines of evidence supporting the role of ferroptosis in hemorrhagic stroke (Alim et al., 2019; Li et al., 2017; Zhang et al., 2018; Zille et al., 2019). In this model, in vitro and in vivo studies suggest that it is 5-lipoxygenase localized at the nuclear envelope that executes ferroptosis (Karuppagounder et al., 2018). Therefore, the development of brain-penetrant inhibitors of 5-lipoxygenase that are active in mice and humans could be an effective strategy for treating parenchymal brain hemorrhage or Alzheimer’s disease (Vagnozzi et al., 2018).
Initial observations revealed that global inhibitors of transcription could abrogate ferroptosis in neurons, a finding that was recently validated (Ratan et al., 1994a; Zille et al., 2019). However, global inhibitors of transcription are problematic in that they not only nullify putative death genes, they also nullify genes involved in tissue homeostasis and repair. Because of this there was a therapeutic need to find small molecules that could modulate “death” gene expression without affecting global protein synthesis. Mithramycin, an aureolic acid antibiotic with DNA binding properties, was identified as a negative regulator of transcription of the ferroptosis mediator, lipoxygenase, and found to potently inhibit ferroptosis. Testing in animals revealed that mithramycin could prolong survival in the R6/2 model of Huntington’s disease (HD) (Ferrante et al., 2004; Sleiman et al., 2011; Voisine et al., 2007), which is a model in which cell death has features of ferroptosis (Lee et al., 2011; Paul et al., 2018; Paul et al., 2014). Structure-activity studies of mithramycin using analogs of the parent compound showed that in order to inhibit ferroptosis or to influence HD pathogenesis, mithramycin had to bind to DNA (Sleiman et al., 2011) and agents which bind DNA with higher affinity were more potent in inhibiting ferroptosis or treating animal models of HD (Sleiman et al., 2011). Mithramycin has also been shown to be effective in models of Alzheimer’s disease where ferroptosis has been similarly implicated (Wei et al., 2016). It has also been shown to prevent dopaminergic loss and functional deficits in a model of methamphetamine toxicity and cell loss in the CA1 region following transient global ischemia (Hagiwara et al., 2009; Osada et al., 2012). Studies involving DNA damage and endoplasmic reticulum stress show that mithramycin is protective in those models (Chatterjee et al., 2001; Kosuge et al., 2011), which raises questions about whether mithramycin targets ferroptosis specifically or whether it acts via convergent mechanisms in distinct models. Nevertheless, given its extensive experience in humans, it should be considered for diseases where ferroptosis is extant.
Epigenetic proteins sit above the genome and integrate external and internal signaling cues to modify chromatin structure and can thereby activate or repress gene expression (Ratan, 2009). Specific targets of epigenetic proteins included histones which form the protein structure for compaction of DNA into chromatin. Simply, electrostatic interactions between negatively charged DNA and positively charged histone tails can be reversibly modified via post-translational modifications such as acetylation. According to this scheme, histone acetyl transferases (HATs or readers) unwind DNA because they neutralize charge interactions between DNA and histones, while histone deacetylases (HDACs or erasers) do the opposite. HDACs are zinc hydrolases that can be inhibited by selective inhibitors with a cap linker and chelator moiety (Bradner et al., 2010). There are multiple classes of HDACs which are named based on molecular phylogenetic analysis of their primary structure. Results from previous studies have shown that chemicals such as scriptaid with inhibitory activity toward HDAC1, HDAC2, HDAC3, and HDAC6 inhibited glutamate- or HCA-induced ferroptosis in neurons, while nullscript (an analog of scriptaid with no HDAC inhibitory activity) had no effect (Langley et al., 2008; Ryu et al., 2003).
Subsequent chemical and molecular investigations confirmed a role for class I HDAC inhibitors in reducing damage in stroke (Fessler et al., 2013), traumatic brain injury (Wang et al., 2015), multiple sclerosis (Camelo et al., 2005), and ALS (Ryu et al., 2005) where ferroptotic pathways have been implicated. Moreover, HDAC6-selective inhibitors, whose activity might primarily reside in axons, have been developed and found not only to inhibit ferroptosis (Rivieccio et al., 2009) but also to improve axonal trafficking, reduce cell loss, and improve functional deficits in models of stroke, ALS, traumatic brain injury, and Charcot-Marie-Tooth disease (Dallavalle et al., 2012). As HDAC6 is primarily localized to axons, the protective effects of HDAC6 inhibitors against ferroptosis highlight how the cytoarchitectonics of neurons with dendrites, cell bodies, axons, and synapses may differentiate pathways to ferroptosis executed in the brain as compared to other organs.
As previously mentioned, non-selective iron chelators such as DFO have been shown to be propitious in the treatment of many neurological disorders. However, translation to humans has been hampered by the absence of a specific target. In a recent study, researchers examined the effect of DFO in intracerebral hemorrhage in humans (Selim et al., 2011). Cell death via ferroptosis has been shown to occur in vitro via exposure to Hgb/hemin and in vivo in rodents with experimental brain hemorrhage (Li et al., 2017; Zille et al., 2017). The human study by Selim and colleagues failed in its primary endpoints, likely because the concentration of DFO could not be optimized without a specific target (Selim et al., 2019).
Converging lines of inquiry outlined above suggest that iron chelators may offer protection from ferroptosis in the brain by targeting a family of iron-dependent dioxygenases, the HIF prolyl hydroxylases (HIF PHDs) (Davis et al., 2018). Adaptaquin was identified as a potent, novel, and selective inhibitor of HIF PHDs (Karuppagounder et al., 2016; Smirnova et al., 2010). It also has an 8-oxyquinoline backbone with known safety in humans. Moreover, as HIF transcription factors are stabilized by HIF prolyl hydroxylase inhibition, in vivo bioluminescence imaging of an HIF-luciferase reporter was used to validate concentrations of adaptaquin required to inhibit the HIF PHDs in the brain. These studies revealed that 30mg/kg was not only sufficient to inhibit the HIF PHDs, but that it also reduced cell death and improved functional recovery in mice and rats with experimental brain hemorrhage (Karuppagounder et al., 2016). Moreover, protection by adaptaquin was not associated with changes in iron levels or distribution in the CNS as measured by X-ray spectroscopy. As damage and dysfunction from brain hemorrhage can be reduced by a host of anti-ferroptotic agents including ferrostatin (Li et al., 2017), N-acetylcysteine (Karuppagounder et al., 2018), and GPX4 overexpression (Alim et al., 2019), the findings support the notion that iron participates in cell death and dysfunction not by preventing Fenton chemistry but rather by targeting a specific family of metalloenzymes, the HIF prolyl hydroxylases (Ratan, 2019).
Conclusion
The recognition of ferroptosis as a distinct cell death pathway by Stockwell and coworkers has expanded and clarified our understanding of the role of oxidative stress in nervous system diseases. The concept, as it applies to neurological conditions, leverages more than three decades of mechanistic studies on non-receptor-mediated glutamate toxicity in neurons initially revealed by Murphy and Coyle. Non-receptor-mediated glutamate (or HCA) toxicity has been shown to be identical to ferroptosis defined in cancer cells (Zille et al., 2019). The evidence that identified targets and chemicals that regulate ferroptosis in neurons in vitro are relevant to neurological disease models in vivo is already quite extensive. What is missing for many of these models and for studies of human neurons and human autopsy tissues is evidence of ferroptotic cell death. Exciting future studies will likely clarify the role of ferroptosis in an array of neurological conditions and set the table for therapeutic tests of its importance in neurological diseases where our ability to intervene remains frustratingly limited.
Acknowledgements:
RRR is funded with generous support from the Dr. Miriam and Sheldon G. Adelson Medical Research Foundation, the Burke Foundation, the Sperling Center for Hemorrhagic Stroke Recovery at the Burke Neurological Institute and the NIH (grant P01 NIA AG014930, project 1, to R.R.R.). I gratefully acknowledge Tim Murphy for introducing me to ferroptosis almost three decades ago, and to Jay Baraban, my post-doctoral mentor for giving me the freedom to work on this interesting model. I was fortunate to be able to leverage their insight and rigor. Space does not allow me to name them all them but I want to thank members of the Ratan Lab past and present (Hopkins (1994-1996), BIDMC/Harvard (1996-2004) and now BNI/Cornell (2004-present) for their diligence and creativity and their numerous contributions to this work. This piece is humbly dedicated to all of them. I also acknowledge careful review by Dr. Harley Kornblum, editorial assistance by Darlene White, and production of the figures by Laura Gilmartin.
Abbreviations:
- xCT
system Xc− cystine/glutamate antiporter
- KRAS
Kirsten rat sarcoma
- RSL3
ras-selective lethal compound 3
- CHOP
CCAT enhancer binding protein homologous protein
- ATF
Activating Transcription Factor Family
- FIN
ferroptosis inducer
- Chac1
glutathione specific gamma glutamyl cyclotransferase
- TRIB3
Tribbles homolog 3
- CREB
cAMP response element binding protein
- REDD1
regulated in development and dna responses
- CARS
cysteinyl tRNA synthetase
- ALS
amyotrophic lateral sclerosis
- Nrf-2
nuclear factor erythroid-2 related factor 2
- ETYA
eicosatetraynoic acid
- GPX4
glutathione peroxidase 4
- MEK
mitogen activated protein kinase kinase
- ERK
extracellular signal-regulated kinase
- RAF
rapidly accelerated fibrosarcoma
- cMyc
avian myelocytomatosis viral oncogene
- Tfr1
Transferrin receptor 1
- IRP2
iron regulatory protein 2
- NRAMP1
natural resistence associated macrophage protein 1
- HIF
hypoxia inducible factor
- HIF PHD
hypoxia inducible factor prolyl hydroxylases
- HCA
homocysteate
- HDAC
histone deacetylase
- NAC
N-acetylcysteine
- DFO
deferoxamine mesylate
Footnotes
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Conflict of Interest: RRR holds patents (owned by Cornell University or Burke Neurological Institute) related to inhibitors of ferroptosis and their therapeutic applications in the nervous system. He is also on the Scientific Advisory Board for Neuronasal, Inc. which has licensed patents from the Ratan Laboratory. RRR has an equity interest in Neuronasal, Inc. RRR also serves in an advisory capacity to Biogen.
References
- Aime P, Karuppagounder SS, Rao A, Chen Y, Burke RE, Ratan RR, and Greene LA (2020). The drug adaptaquin blocks ATF4/CHOP-dependent pro-death Trib3 induction and protects in cellular and mouse models of Parkinson's disease. Neurobiology of disease 136, 104725. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Alim I, Caulfield JT, Chen Y, Swarup V, Geschwind DH, Ivanova E, Seravalli J, Ai Y, Sansing LH, Ste Marie EJ et al. (2019). Selenium Drives a Transcriptional Adaptive Program to Block Ferroptosis and Treat Stroke. Cell 177, 1262–1279.e1225. [DOI] [PubMed] [Google Scholar]
- Aminova LR, Chavez JC, Lee J, Ryu H, Kung A, Lamanna JC, and Ratan RR (2005). Prosurvival and prodeath effects of hypoxia-inducible factor-1alpha stabilization in a murine hippocampal cell line. The Journal of biological chemistry 280, 3996–4003. [DOI] [PubMed] [Google Scholar]
- Anthonymuthu TS, Kenny EM, Shrivastava I, Tyurina YY, Hier ZE, Ting HC, Dar HH, Tyurin VA, Nesterova A, Amoscato AA, et al. (2018). Empowerment of 15-Lipoxygenase Catalytic Competence in Selective Oxidation of Membrane ETE-PE to Ferroptotic Death Signals, HpETE-PE. Journal of the American Chemical Society 140, 17835–17839. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Antonic A, Dottori M, Macleod MR, Donnan GA, and Howells DW (2018). NXY-059, a Failed Stroke Neuroprotectant, Offers No Protection to Stem Cell-Derived Human Neurons. Journal of stroke and cerebrovascular diseases : the official journal of National Stroke Association 27, 2158–2165. [DOI] [PubMed] [Google Scholar]
- Aoyama K, Suh SW, Hamby AM, Liu J, Chan WY, Chen Y, and Swanson RA (2006). Neuronal glutathione deficiency and age-dependent neurodegeneration in the EAAC1 deficient mouse. Nature neuroscience 9, 119–126. [DOI] [PubMed] [Google Scholar]
- Armstrong MM, Freedman CJ, Jung JE, Zheng Y, Kalyanaraman C, Jacobson MP, Simeonov A, Maloney DJ, van Leyen K, Jadhav A, et al. (2016). A potent and selective inhibitor targeting human and murine 12/15-LOX. Bioorganic & medicinal chemistry 24, 1183–1190. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Back SA, Gan X, Li Y, Rosenberg PA, and Volpe JJ (1998). Maturation-dependent vulnerability of oligodendrocytes to oxidative stress-induced death caused by glutathione depletion. The Journal of neuroscience : the official journal of the Society for Neuroscience 18, 6241–6253. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Bannai S, and Kitamura E (1980). Transport interaction of L-cystine and L-glutamate in human diploid fibroblasts in culture. The Journal of biological chemistry 255, 2372–2376. [PubMed] [Google Scholar]
- Bannai S, Tsukeda H, and Okumura H (1977). Effect of antioxidants on cultured human diploid fibroblasts exposed to cystine-free medium. Biochemical and biophysical research communications 74, 1582–1588. [DOI] [PubMed] [Google Scholar]
- Barnham KJ, and Bush AI (2014). Biological metals and metal-targeting compounds in major neurodegenerative diseases. Chemical Society reviews 43, 6727–6749. [DOI] [PubMed] [Google Scholar]
- Basso M, and Ratan RR (2013). Transglutaminase is a therapeutic target for oxidative stress, excitotoxicity and stroke: a new epigenetic kid on the CNS block. Journal of cerebral blood flow and metabolism : official journal of the International Society of Cerebral Blood Flow and Metabolism 33, 809–818. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Benedict AL, Mountney A, Hurtado A, Bryan KE, Schnaar RL, Dinkova-Kostova AT, and Talalay P (2012). Neuroprotective effects of sulforaphane after contusive spinal cord injury. Journal of neurotrauma 29, 2576–2586. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Bergsland N, Tavazzi E, Schweser F, Jakimovski D, Hagemeier J, Dwyer MG, and Zivadinov R (2019). Targeting Iron Dyshomeostasis for Treatment of Neurodegenerative Disorders. CNS drugs 33, 1073–1086. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Berti DA, and Seger R (2017). The Nuclear Translocation of ERK. Methods in molecular biology (Clifton, NJ) 1487, 175–194. [DOI] [PubMed] [Google Scholar]
- Bhatti JS, Vijayvergiya R, Singh B, and Bhatti GK (2018). Genetic susceptibility of glutathione S-transferase genes (GSTM1/T1 and P1) to coronary artery disease in Asian Indians. Annals of human genetics 82, 448–456. [DOI] [PubMed] [Google Scholar]
- Bradner JE, Mak R, Tanguturi SK, Mazitschek R, Haggarty SJ, Ross K, Chang CY, Bosco J, West N, Morse E et al. (2010). Chemical genetic strategy identifies histone deacetylase 1 (HDAC1) and HDAC2 as therapeutic targets in sickle cell disease. Proceedings of the National Academy of Sciences of the United States of America 107, 12617–12622. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Brennan MS, Patel H, Allaire N, Thai A, Cullen P, Ryan S, Lukashev M, Bista P, Huang R, Rhodes KJ et al. (2016). Pharmacodynamics of Dimethyl Fumarate Are Tissue Specific and Involve NRF2-Dependent and -Independent Mechanisms. Antioxidants & redox signaling 24, 1058–1071. [DOI] [PubMed] [Google Scholar]
- Camelo S, Iglesias AH, Hwang D, Due B, Ryu H, Smith K, Gray SG, Imitola J, Duran G, Assaf B et al. (2005). Transcriptional therapy with the histone deacetylase inhibitor trichostatin A ameliorates experimental autoimmune encephalomyelitis. Journal of neuroimmunology 164, 10–21. [DOI] [PubMed] [Google Scholar]
- Carmeliet P, and Baes M (2008). Metabolism and therapeutic angiogenesis. The New England journal of medicine 358, 2511–2512. [DOI] [PubMed] [Google Scholar]
- Chatterjee S, Zaman K, Ryu H, Conforto A, and Ratan RR (2001). Sequence-selective DNA binding drugs mithramycin A and chromomycin A3 are potent inhibitors of neuronal apoptosis induced by oxidative stress and DNA damage in cortical neurons. Annals of neurology 49, 345–354. [PubMed] [Google Scholar]
- Chen D, Fan Z, Rauh M, Buchfelder M, Eyupoglu IY, and Savaskan N (2017a). ATF4 promotes angiogenesis and neuronal cell death and confers ferroptosis in a xCT-dependent manner. Oncogene 36, 5593–5608. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Chen L, Hambright WS, Na R, and Ran Q (2015). Ablation of the Ferroptosis Inhibitor Glutathione Peroxidase 4 in Neurons Results in Rapid Motor Neuron Degeneration and Paralysis. The Journal of biological chemistry 290, 28097–28106. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Chen MS, Wang SF, Hsu CY, Yin PH, Yeh TS, Lee HC, and Tseng LM (2017b). CHAC1 degradation of glutathione enhances cystine-starvation-induced necroptosis and ferroptosis in human triple negative breast cancer cells via the GCN2-eIF2alpha-ATF4 pathway. Oncotarget 8, 114588–114602. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Choi DW, Maulucci-Gedde M, and Kriegstein AR (1987). Glutamate neurotoxicity in cortical cell culture. The Journal of neuroscience : the official journal of the Society for Neuroscience 7, 357–368. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Conrad M, Kagan VE, Bayir H, Pagnussat GC, Head B, Traber MG, and Stockwell BR (2018). Regulation of lipid peroxidation and ferroptosis in diverse species. Genes & development 32, 602–619. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Crawford RR, Prescott ET, Sylvester CF, Higdon AN, Shan J, Kilberg MS, and Mungrue IN (2015). Human CHAC1 Protein Degrades Glutathione, and mRNA Induction Is Regulated by the Transcription Factors ATF4 and ATF3 and a Bipartite ATF/CRE Regulatory Element. The Journal of biological chemistry 290, 15878–15891. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Dallavalle S, Pisano C, and Zunino F (2012). Development and therapeutic impact of HDAC6-selective inhibitors. Biochemical pharmacology 84, 756–765. [DOI] [PubMed] [Google Scholar]
- Dang CV (2012). MYC on the path to cancer. Cell 149, 22–35. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Davis CK, Jain SA, Bae ON, Majid A, and Rajanikant GK (2018). Hypoxia Mimetic Agents for Ischemic Stroke. Frontiers in cell and developmental biology 6, 175. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Distefano AM, Martin MV, Cordoba JP, Bellido AM, D'Ippolito S, Colman SL, Soto D, Roldan JA, Bartoli CG, Zabaleta EJ et al. (2017). Heat stress induces ferroptosis-like cell death in plants. The Journal of cell biology 216, 463–476. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Dixon SJ, Lemberg KM, Lamprecht MR, Skouta R, Zaitsev EM, Gleason CE, Patel DN, Bauer AJ, Cantley AM, Yang WS, et al. (2012). Ferroptosis: an iron-dependent form of nonapoptotic cell death. Cell 149, 1060–1072. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Dringen R, Brandmann M, Hohnholt MC, and Blumrich EM (2015). Glutathione-Dependent Detoxification Processes in Astrocytes. Neurochemical research 40, 2570–2582. [DOI] [PubMed] [Google Scholar]
- Dringen R, Gutterer JM, and Hirrlinger J (2000). Glutathione metabolism in brain metabolic interaction between astrocytes and neurons in the defense against reactive oxygen species. European journal of biochemistry 267, 4912–4916. [DOI] [PubMed] [Google Scholar]
- Dringen R, Pfeiffer B, and Hamprecht B (1999). Synthesis of the antioxidant glutathione in neurons: supply by astrocytes of CysGly as precursor for neuronal glutathione. The Journal of neuroscience : the official journal of the Society for Neuroscience 19, 562–569. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Du K, Herzig S, Kulkarni RN, and Montminy M (2003). TRB3: a tribbles homolog that inhibits Akt/PKB activation by insulin in liver. Science (New York, NY) 300, 1574–1577. [DOI] [PubMed] [Google Scholar]
- Enyedi B, Jelcic M, and Niethammer P (2016). The Cell Nucleus Serves as a Mechanotransducer of Tissue Damage-Induced Inflammation. Cell 165, 1160–1170. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Epstein AC, Gleadle JM, McNeill LA, Hewitson KS, O'Rourke J, Mole DR, Mukherji M, Metzen E, Wilson MI, Dhanda A, et al. (2001). C. elegans EGL-9 and mammalian homologs define a family of dioxygenases that regulate HIF by prolyl hydroxylation. Cell 107, 43–54. [DOI] [PubMed] [Google Scholar]
- Esch F, Lin KI, Hills A, Zaman K, Baraban JM, Chatterjee S, Rubin L, Ash DE, and Ratan RR (1998). Purification of a multipotent antideath activity from bovine liver and its identification as arginase: nitric oxide-independent inhibition of neuronal apoptosis. The Journal of neuroscience : the official journal of the Society for Neuroscience 18, 4083–4095. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Farrelly LA, Thompson RE, Zhao S, Lepack AE, Lyu Y, Bhanu NV, Zhang B, Loh YE, Ramakrishnan A, Vadodaria KC, et al. (2019). Histone serotonylation is a permissive modification that enhances TFIID binding to H3K4me3. Nature 567, 535–539. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Ferrante RJ, Ryu H, Kubilus JK, D'Mello S, Sugars KL, Lee J, Lu P, Smith K, Browne S, Beal MF, et al. (2004). Chemotherapy for the brain: the antitumor antibiotic mithramycin prolongs survival in a mouse model of Huntington's disease. The Journal of neuroscience : the official journal of the Society for Neuroscience 24, 10335–10342. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Fessler EB, Chibane FL, Wang Z, and Chuang DM (2013). Potential roles of HDAC inhibitors in mitigating ischemia-induced brain damage and facilitating endogenous regeneration and recovery. Current pharmaceutical design 19, 5105–5120. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Fogal B, Li J, Lobner D, McCullough LD, and Hewett SJ (2007). System x(c)- activity and astrocytes are necessary for interleukin-1 beta-mediated hypoxic neuronal injury. The Journal of neuroscience : the official journal of the Society for Neuroscience 27, 10094–10105. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Giltnane JM, and Balko JM (2014). Rationale for targeting the Ras/MAPK pathway in triple-negative breast cancer. Discovery medicine 17, 275–283. [PubMed] [Google Scholar]
- Hagiwara H, Iyo M, and Hashimoto K (2009). Mithramycin protects against dopaminergic neurotoxicity in the mouse brain after administration of methamphetamine. Brain research 1301, 189–196. [DOI] [PubMed] [Google Scholar]
- Hambright WS, Fonseca RS, Chen L, Na R, and Ran Q (2017). Ablation of ferroptosis regulator glutathione peroxidase 4 in forebrain neurons promotes cognitive impairment and neurodegeneration. Redox biology 12, 8–17. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Hardingham GE, and Bading H (2010). Synaptic versus extrasynaptic NMDA receptor signalling: implications for neurodegenerative disorders. Nature reviews Neuroscience 11, 682–696. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Haskew-Layton RE, Payappilly JB, Smirnova NA, Ma TC, Chan KK, Murphy TH, Guo H, Langley B, Sultana R, Butterfield DA, et al. (2010). Controlled enzymatic production of astrocytic hydrogen peroxide protects neurons from oxidative stress via an Nrf2-independent pathway. Proceedings of the National Academy of Sciences of the United States of America 107, 17385–17390. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Haskew-Layton RE, Payappilly JB, Xu H, Bennett SA, and Ratan RR (2013). 15-Deoxy-Delta12,14-prostaglandin J2 (15d-PGJ2) protects neurons from oxidative death via an Nrf2 astrocyte-specific mechanism independent of PPARgamma. Journal of neurochemistry 124, 536–547. [DOI] [PubMed] [Google Scholar]
- Hayano M, Yang WS, Corn CK, Pagano NC, and Stockwell BR (2016). Loss of cysteinyl-tRNA synthetase (CARS) induces the transsulfuration pathway and inhibits ferroptosis induced by cystine deprivation. Cell death and differentiation 23, 270–278. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Heine P, Ehrlicher A, and Kas J (2015). Neuronal and metastatic cancer cells: Unlike brothers. Biochimica et biophysica acta 1853, 3126–3131. [DOI] [PubMed] [Google Scholar]
- Hinnebusch AG (1994). The eIF-2 alpha kinases: regulators of protein synthesis in starvation and stress. Seminars in cell biology 5, 417–426. [DOI] [PubMed] [Google Scholar]
- Ingold I, Berndt C, Schmitt S, Doll S, Poschmann G, Buday K, Roveri A, Peng X, Porto Freitas F, Seibt T et al. (2018). Selenium Utilization by GPX4 Is Required to Prevent Hydroperoxide-Induced Ferroptosis. Cell 172, 409–422.e421. [DOI] [PubMed] [Google Scholar]
- Ishii T, Warabi E, and Mann GE (2019). Circadian control of BDNF-mediated Nrf2 activation in astrocytes protects dopaminergic neurons from ferroptosis. Free radical biology & medicine 133, 169–178. [DOI] [PubMed] [Google Scholar]
- Ivan M, Kondo K, Yang H, Kim W, Valiando J, Ohh M, Salic A, Asara JM, Lane WS, and Kaelin WG Jr. (2001). HIFalpha targeted for VHL-mediated destruction by proline hydroxylation: implications for O2 sensing. Science (New York, NY) 292, 464–468. [DOI] [PubMed] [Google Scholar]
- Jiang L, Kon N, Li T, Wang SJ, Su T, Hibshoosh H, Baer R, and Gu W (2015). Ferroptosis as a p53-mediated activity during tumour suppression. Nature 520, 57–62. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Johnson RS (2016). Profile of William Kaelin, Peter Ratcliffe, and Greg Semenza, 2016 Albert Lasker Basic Medical Research Awardees. Proceedings of the National Academy of Sciences of the United States of America 113, 13938–13940. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Kagan VE, Bayir H, Tyurina YY, Bolevich SB, Maguire JJ, Fadeel B, and Balasubramanian K (2017). Elimination of the unnecessary: Intra- and extracellular signaling by anionic phospholipids. Biochemical and biophysical research communications 482, 482–490. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Karatas H, Eun Jung J, Lo EH, and van Leyen K (2018). Inhibiting 12/15-lipoxygenase to treat acute stroke in permanent and tPA induced thrombolysis models. Brain research 1678, 123–128. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Karpuj MV, Becher MW, Springer JE, Chabas D, Youssef S, Pedotti R, Mitchell D, and Steinman L (2002). Prolonged survival and decreased abnormal movements in transgenic model of Huntington disease, with administration of the transglutaminase inhibitor cystamine. Nature medicine 8, 143–149. [DOI] [PubMed] [Google Scholar]
- Karuppagounder SS, Alim I, Khim SJ, Bourassa MW, Sleiman SF, John R, Thinnes CC, Yeh TL, Demetriades M, Neitemeier S et al. (2016). Therapeutic targeting of oxygen-sensing prolyl hydroxylases abrogates ATF4-dependent neuronal death and improves outcomes after brain hemorrhage in several rodent models. Science translational medicine 8, 328ra329. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Karuppagounder SS, Alin L, Chen Y, Brand D, Bourassa MW, Dietrich K, Wilkinson CM, Nadeau CA, Kumar A, Perry S et al. (2018). N-acetylcysteine targets 5 lipoxygenase-derived, toxic lipids and can synergize with prostaglandin E2 to inhibit ferroptosis and improve outcomes following hemorrhagic stroke in mice. Annals of neurology 84, 854–872. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Kenny EM, Fidan E, Yang Q, Anthonymuthu TS, New LA, Meyer EA, Wang H, Kochanek PM, Dixon CE, Kagan VE, et al. (2019). Ferroptosis Contributes to Neuronal Death and Functional Outcome After Traumatic Brain Injury. Critical care medicine 47, 410–418. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Khanna S, Roy S, Ryu H, Bahadduri P, Swaan PW, Ratan RR, and Sen CK (2003). Molecular basis of vitamin E action: tocotrienol modulates 12-lipoxygenase, a key mediator of glutamate-induced neurodegeneration. The Journal of biological chemistry 278, 43508–43515. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Kidger AM, Sipthorp J, and Cook SJ (2018). ERK1/2 inhibitors: New weapons to inhibit the RAS-regulated RAF-MEK1/2-ERK1/2 pathway. Pharmacology & therapeutics 187, 45–60. [DOI] [PubMed] [Google Scholar]
- Kirsch DG, Doseff A, Chau BN, Lim DS, de Souza-Pinto NC, Hansford R, Kastan MB, Lazebnik YA, and Hardwick JM (1999). Caspase-3-dependent cleavage of Bcl-2 promotes release of cytochrome c. The Journal of biological chemistry 274, 21155–21161. [DOI] [PubMed] [Google Scholar]
- Koppula P, Zhang Y, Zhuang L, and Gan B (2018). Amino acid transporter SLC7A11/xCT at the crossroads of regulating redox homeostasis and nutrient dependency of cancer. Cancer communications (London, England) 38, 12. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Kosuge Y, Taniguchi Y, Imai T, Ishige K, and Ito Y (2011). Neuroprotective effect of mithramycin against endoplasmic reticulum stress-induced neurotoxicity in organotypic hippocampal slice cultures. Neuropharmacology 61, 252–261. [DOI] [PubMed] [Google Scholar]
- Kraft AD, Johnson DA, and Johnson JA (2004). Nuclear factor E2-related factor 2-dependent antioxidant response element activation by tert-butylhydroquinone and sulforaphane occurring preferentially in astrocytes conditions neurons against oxidative insult. The Journal of neuroscience : the official journal of the Society for Neuroscience 24, 1101–1112. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Kumar C, Igbaria A, D'Autreaux B, Planson AG, Junot C, Godat E, Bachhawat AK, Delaunay-Moisan A, and Toledano MB (2011). Glutathione revisited: a vital function in iron metabolism and ancillary role in thiol-redox control. The EMBO journal 30, 2044–2056. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Lange PS, Chavez JC, Pinto JT, Coppola G, Sun CW, Townes TM, Geschwind DH, and Ratan RR (2008). ATF4 is an oxidative stress-inducible, prodeath transcription factor in neurons in vitro and in vivo. The Journal of experimental medicine 205, 1227–1242. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Langley B, D'Annibale MA, Suh K, Ayoub I, Tolhurst A, Bastan B, Yang L, Ko B, Fisher M, Cho S et al. (2008). Pulse inhibition of histone deacetylases induces complete resistance to oxidative death in cortical neurons without toxicity and reveals a role for cytoplasmic p21(waf1/cip1) in cell cycle-independent neuroprotection. The Journal of neuroscience : the official journal of the Society for Neuroscience 28, 163–176. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Lee J, Kosaras B, Del Signore SJ, Cormier K, McKee A, Ratan RR, Kowall NW, and Ryu H (2011). Modulation of lipid peroxidation and mitochondrial function improves neuropathology in Huntington's disease mice. Acta neuropathologica 121, 487–498. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Lee JM, and Johnson JA (2004). An important role of Nrf2-ARE pathway in the cellular defense mechanism. Journal of biochemistry and molecular biology 37, 139–143. [DOI] [PubMed] [Google Scholar]
- Lee JM, Shih AY, Murphy TH, and Johnson JA (2003). NF-E2-related factor-2 mediates neuroprotection against mitochondrial complex I inhibitors and increased concentrations of intracellular calcium in primary cortical neurons. The Journal of biological chemistry 278, 37948–37956. [DOI] [PubMed] [Google Scholar]
- Lee NJ, Ha SK, Sati P, Absinta M, Nair G, Luciano NJ, Leibovitch EC, Yen CC, Rouault TA, Silva AC, et al. (2019). Potential role of iron in repair of inflammatory demyelinating lesions. The Journal of clinical investigation 129, 4365–4376. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Levinthal DJ, and DeFranco DB (2004). Transient phosphatidylinositol 3-kinase inhibition protects immature primary cortical neurons from oxidative toxicity via suppression of extracellular signal-regulated kinase activation. The Journal of biological chemistry 279, 11206–11213. [DOI] [PubMed] [Google Scholar]
- Lewerenz J, Hewett SJ, Huang Y, Lambros M, Gout PW, Kalivas PW, Massie A, Smolders I, Methner A, Pergande M et al. (2013). The cystine/glutamate antiporter system x(c)(−) in health and disease: from molecular mechanisms to novel therapeutic opportunities. Antioxidants & redox signaling 18, 522–555. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Lewerenz J, and Maher P (2009). Basal levels of eIF2alpha phosphorylation determine cellular antioxidant status by regulating ATF4 and xCT expression. The Journal of biological chemistry 284, 1106–1115. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Li PC, Jiao Y, Ding J, Chen YC, Cui Y, Qian C, Yang XY, Ju SH, Yao HH, and Teng GJ (2015). Cystamine improves functional recovery via axon remodeling and neuroprotection after stroke in mice. CNS neuroscience & therapeutics 21, 231–240. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Li Q, Han X, Lan X, Gao Y, Wan J, Durham F, Cheng T, Yang J, Wang Z, Jiang C et al. (2017). Inhibition of neuronal ferroptosis protects hemorrhagic brain. JCI insight 2, e90777. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Li Y, Maher P, and Schubert D (1997). A role for 12-lipoxygenase in nerve cell death caused by glutathione depletion. Neuron 19, 453–463. [DOI] [PubMed] [Google Scholar]
- Liddelow SA, Guttenplan KA, Clarke LE, Bennett FC, Bohlen CJ, Schirmer L, Bennett ML, Munch AE, Chung WS, Peterson TC, et al. (2017). Neurotoxic reactive astrocytes are induced by activated microglia. Nature 541, 481–487. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Luo Y, and DeFranco DB (2006). Opposing roles for ERK1/2 in neuronal oxidative toxicity: distinct mechanisms of ERK1/2 action at early versus late phases of oxidative stress. The Journal of biological chemistry 281, 16436–16442. [DOI] [PubMed] [Google Scholar]
- Ma Y, Brewer JW, Diehl JA, and Hendershot LM (2002). Two distinct stress signaling pathways converge upon the CHOP promoter during the mammalian unfolded protein response. Journal of molecular biology 318, 1351–1365. [DOI] [PubMed] [Google Scholar]
- Makowske M, and Christensen HN (1982). Hepatic transport system interconverted by protonation from service for neutral to service for anionic amino acids. The Journal of biological chemistry 257, 14635–14638. [PubMed] [Google Scholar]
- Malagelada C, Jin ZH, and Greene LA (2008). RTP801 is induced in Parkinson's disease and mediates neuron death by inhibiting Akt phosphorylation/activation. The Journal of neuroscience : the official journal of the Society for Neuroscience 28, 14363–14371. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Martin DP, Schmidt RE, DiStefano PS, Lowry OH, Carter JG, and Johnson EM Jr. (1988). Inhibitors of protein synthesis and RNA synthesis prevent neuronal death caused by nerve growth factor deprivation. The Journal of cell biology 106, 829–844. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Matsumoto M, Minami M, Takeda K, Sakao Y, and Akira S (1996). Ectopic expression of CHOP (GADD153) induces apoptosis in M1 myeloblastic leukemia cells. FEBS letters 395, 143–147. [DOI] [PubMed] [Google Scholar]
- Mazor KM, and Stipanuk MH (2016). GCN2- and eIF2alpha-phosphorylation-independent, but ATF4-dependent, induction of CARE-containing genes in methionine-deficient cells. Amino acids 48, 2831–2842. [DOI] [PMC free article] [PubMed] [Google Scholar]
- McConoughey SJ, Basso M, Niatsetskaya ZV, Sleiman SF, Smirnova NA, Langley BC, Mahishi L, Cooper AJ, Antonyak MA, Cerione RA, et al. (2010). Inhibition of transglutaminase 2 mitigates transcriptional dysregulation in models of Huntington disease. EMBO molecular medicine 2, 349–370. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Miyamoto M, Murphy TH, Schnaar RL, and Coyle JT (1989). Antioxidants protect against glutamate-induced cytotoxicity in a neuronal cell line. The Journal of pharmacology and experimental therapeutics 250, 1132–1140. [PubMed] [Google Scholar]
- Murphy TH, and Baraban JM (1990). Glutamate toxicity in immature cortical neurons precedes development of glutamate receptor currents. Brain research Developmental brain research 57, 146–150. [DOI] [PubMed] [Google Scholar]
- Murphy TH, De Long MJ, and Coyle JT (1991). Enhanced NAD(P)H:quinone reductase activity prevents glutamate toxicity produced by oxidative stress. Journal of neurochemistry 56, 990–995. [DOI] [PubMed] [Google Scholar]
- Murphy TH, Miyamoto M, Sastre A, Schnaar RL, and Coyle JT (1989). Glutamate toxicity in a neuronal cell line involves inhibition of cystine transport leading to oxidative stress. Neuron 2, 1547–1558. [DOI] [PubMed] [Google Scholar]
- Murphy TH, Schnaar RL, and Coyle JT (1990). Immature cortical neurons are uniquely sensitive to glutamate toxicity by inhibition of cystine uptake. FASEB journal : official publication of the Federation of American Societies for Experimental Biology 4, 1624–1633. [PubMed] [Google Scholar]
- Nandal A, Ruiz JC, Subramanian P, Ghimire-Rijal S, Sinnamon RA, Stemmler TL, Bruick RK, and Philpott CC (2011). Activation of the HIF prolyl hydroxylase by the iron chaperones PCBP1 and PCBP2. Cell metabolism 14, 647–657. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Neitemeier S, Jelinek A, Laino V, Hoffmann L, Eisenbach I, Eying R, Ganjam GK, Dolga AM, Oppermann S, and Culmsee C (2017). BID links ferroptosis to mitochondrial cell death pathways. Redox biology 12, 558–570. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Ohoka N, Yoshii S, Hattori T, Onozaki K, and Hayashi H (2005). TRB3, a novel ER stress-inducible gene, is induced via ATF4-CHOP pathway and is involved in cell death. The EMBO journal 24, 1243–1255. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Okauchi M, Xi G, Keep RF, and Hua Y (2009). Tissue-type transglutaminase and the effects of cystamine on intracerebral hemorrhage-induced brain edema and neurological deficits. Brain research 1249, 229–236. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Olson DE, Sleiman SF, Bourassa MW, Wagner FF, Gale JP, Zhang YL, Ratan RR, and Holson EB (2015). Hydroxamate-based histone deacetylase inhibitors can protect neurons from oxidative stress via a histone deacetylase-independent catalase-like mechanism. Chemistry & biology 22, 439–445. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Osada N, Kosuge Y, Oguchi S, Miyagishi H, Ishige K, and Ito Y (2012). Protective action of mithramycin against neurodegeneration and impairment of synaptic plasticity in the hippocampal CA1 area after transient global ischemia. Neurochemistry international 60, 47–54. [DOI] [PubMed] [Google Scholar]
- Pathak SS, Liu D, Li T, de Zavalia N, Zhu L, Li J, Karthikeyan R, Alain T, Liu AC, Storch KF, et al. (2019). The eIF2alpha Kinase GCN2 Modulates Period and Rhythmicity of the Circadian Clock by Translational Control of Atf4. Neuron 104, 724–735.e726. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Paul BD, Sbodio JI, and Snyder SH (2018). Cysteine Metabolism in Neuronal Redox Homeostasis. Trends in pharmacological sciences 39, 513–524. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Paul BD, Sbodio JI, Xu R, Vandiver MS, Cha JY, Snowman AM, and Snyder SH (2014). Cystathionine gamma-lyase deficiency mediates neurodegeneration in Huntington's disease. Nature 509, 96–100. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Ratan RR (2009). Epigenetics and the nervous system: epiphenomenon or missing piece of the neurotherapeutic puzzle? The Lancet Neurology 8, 975–977. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Ratan RR (2019). Does iron loading of oxygen-sensing prolyl hydroxylases rather than random Fenton-driven radical formation drive programmed ferroptosis and degeneration in neurological diseases? Current Opinion in Physiology 7, 60–65. [Google Scholar]
- Ratan RR, Lee PJ, and Baraban JM (1996). Serum deprivation inhibits glutathione depletion-induced death in embryonic cortical neurons: evidence against oxidative stress as a final common mediator of neuronal apoptosis. Neurochemistry international 29, 153–157. [DOI] [PubMed] [Google Scholar]
- Ratan RR, Murphy TH, and Baraban JM (1994a). Macromolecular synthesis inhibitors prevent oxidative stress-induced apoptosis in embryonic cortical neurons by shunting cysteine from protein synthesis to glutathione. The Journal of neuroscience : the official journal of the Society for Neuroscience 14, 4385–4392. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Ratan RR, Murphy TH, and Baraban JM (1994b). Oxidative stress induces apoptosis in embryonic cortical neurons. Journal of neurochemistry 62, 376–379. [DOI] [PubMed] [Google Scholar]
- Rivieccio MA, Brochier C, Willis DE, Walker BA, D'Annibale MA, McLaughlin K, Siddiq A, Kozikowski AP, Jaffrey SR, Twiss JL, et al. (2009). HDAC6 is a target for protection and regeneration following injury in the nervous system. Proceedings of the National Academy of Sciences of the United States of America 106, 19599–19604. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Roskoski R Jr. (2012). ERK1/2 MAP kinases: structure, function, and regulation. Pharmacological research 66, 105–143. [DOI] [PubMed] [Google Scholar]
- Rothstein JD (1996). Therapeutic horizons for amyotrophic lateral sclerosis. Current opinion in neurobiology 6, 679–687. [DOI] [PubMed] [Google Scholar]
- Ryu H, Lee J, Olofsson BA, Mwidau A, Dedeoglu A, Escudero M, Flemington E, Azizkhan-Clifford J, Ferrante RJ, and Ratan RR (2003). Histone deacetylase inhibitors prevent oxidative neuronal death independent of expanded polyglutamine repeats via an Sp1-dependent pathway. Proceedings of the National Academy of Sciences of the United States of America 100, 4281–4286. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Ryu H, Smith K, Camelo SI, Carreras I, Lee J, Iglesias AH, Dangond F, Cormier KA, Cudkowicz ME, Brown RH Jr., et al. (2005). Sodium phenylbutyrate prolongs survival and regulates expression of anti-apoptotic genes in transgenic amyotrophic lateral sclerosis mice. Journal of neurochemistry 93, 1087–1098. [DOI] [PubMed] [Google Scholar]
- Sagara J, Miura K, and Bannai S (1993a). Cystine uptake and glutathione level in fetal brain cells in primary culture and in suspension. Journal of neurochemistry 61, 1667–1671. [DOI] [PubMed] [Google Scholar]
- Sagara JI, Miura K, and Bannai S (1993b). Maintenance of neuronal glutathione by glial cells. Journal of neurochemistry 61, 1672–1676. [DOI] [PubMed] [Google Scholar]
- Samhan-Arias AK, Tyurina YY, and Kagan VE (2011). Lipid antioxidants: free radical scavenging versus regulation of enzymatic lipid peroxidation. Journal of clinical biochemistry and nutrition 48, 91–95. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Sato H, Tamba M, Ishii T, and Bannai S (1999). Cloning and expression of a plasma membrane cystine/glutamate exchange transporter composed of two distinct proteins. The Journal of biological chemistry 274, 11455–11458. [DOI] [PubMed] [Google Scholar]
- Satoh T, Okamoto SI, Cui J, Watanabe Y, Furuta K, Suzuki M, Tohyama K, and Lipton SA (2006). Activation of the Keap1/Nrf2 pathway for neuroprotection by electrophilic [correction of electrophillic] phase II inducers. Proceedings of the National Academy of Sciences of the United States of America 103, 768–773. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Seiler A, Schneider M, Forster H, Roth S, Wirth EK, Culmsee C, Plesnila N, Kremmer E, Radmark O, Wurst W, et al. (2008). Glutathione peroxidase 4 senses and translates oxidative stress into 12/15-lipoxygenase dependent- and AIF-mediated cell death. Cell metabolism 8, 237–248. [DOI] [PubMed] [Google Scholar]
- Sejbaek T, Nielsen HH, Penner N, Plavina T, Mendoza JP, Martin NA, Elkjaer ML, Ravnborg MH, and Illes Z (2019). Dimethyl fumarate decreases neurofilament light chain in CSF and blood of treatment naive relapsing MS patients. Journal of neurology, neurosurgery, and psychiatry 90, 1324–1330. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Selim M, Foster LD, Moy CS, Xi G, Hill MD, Morgenstern LB, Greenberg SM, James ML, Singh V, Clark WM, et al. (2019). Deferoxamine mesylate in patients with intracerebral haemorrhage (i-DEF): a multicentre, randomised, placebo-controlled, double-blind phase 2 trial. The Lancet Neurology 18, 428–438. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Selim M, Yeatts S, Goldstein JN, Gomes J, Greenberg S, Morgenstern LB, Schlaug G, Torbey M, Waldman B, Xi G, et al. (2011). Safety and tolerability of deferoxamine mesylate in patients with acute intracerebral hemorrhage. Stroke 42, 3067–3074. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Semenza GL (2012). Hypoxia-inducible factors in physiology and medicine. Cell 148, 399–408. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Semenza GL, Roth PH, Fang HM, and Wang GL (1994). Transcriptional regulation of genes encoding glycolytic enzymes by hypoxia-inducible factor 1. The Journal of biological chemistry 269, 23757–23763. [PubMed] [Google Scholar]
- Sen CK, Khanna S, Rink C, and Roy S (2007). Tocotrienols: the emerging face of natural vitamin E. Vitamins and hormones 76, 203–261. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Shih AY, Erb H, Sun X, Toda S, Kalivas PW, and Murphy TH (2006). Cystine/glutamate exchange modulates glutathione supply for neuroprotection from oxidative stress and cell proliferation. The Journal of neuroscience : the official journal of the Society for Neuroscience 26, 10514–10523. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Shih AY, Johnson DA, Wong G, Kraft AD, Jiang L, Erb H, Johnson JA, and Murphy TH (2003). Coordinate regulation of glutathione biosynthesis and release by Nrf2-expressing glia potently protects neurons from oxidative stress. The Journal of neuroscience : the official journal of the Society for Neuroscience 23, 3394–3406. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Siddiq A, Aminova LR, Troy CM, Suh K, Messer Z, Semenza GL, and Ratan RR (2009). Selective inhibition of hypoxia-inducible factor (HIF) prolyl-hydroxylase 1 mediates neuroprotection against normoxic oxidative death via HIF- and CREB-independent pathways. The Journal of neuroscience : the official journal of the Society for Neuroscience 29, 8828–8838. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Siddiq A, Ayoub IA, Chavez JC, Aminova L, Shah S, LaManna JC, Patton SM, Connor JR, Cherny RA, Volitakis I et al. (2005). Hypoxia-inducible factor prolyl 4-hydroxylase inhibition. A target for neuroprotection in the central nervous system. The Journal of biological chemistry 280, 41732–41743. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Sleiman SF, Langley BC, Basso M, Berlin J, Xia L, Payappilly JB, Kharel MK, Guo H, Marsh JL, Thompson LM, et al. (2011). Mithramycin is a gene-selective Sp1 inhibitor that identifies a biological intersection between cancer and neurodegeneration. The Journal of neuroscience : the official journal of the Society for Neuroscience 31, 6858–6870. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Smirnova NA, Haskew-Layton RE, Basso M, Hushpulian DM, Payappilly JB, Speer RE, Ahn YH, Rakhman I, Cole PA, Pinto JT, et al. (2011). Development of Neh2-luciferase reporter and its application for high throughput screening and real-time monitoring of Nrf2 activators. Chemistry & biology 18, 752–765. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Smirnova NA, Rakhman I, Moroz N, Basso M, Payappilly J, Kazakov S, Hernandez-Guzman F, Gaisina IN, Kozikowski AP, Ratan RR, et al. (2010). Utilization of an in vivo reporter for high throughput identification of branched small molecule regulators of hypoxic adaptation. Chemistry & biology 17, 380–391. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Stanciu M, and DeFranco DB (2002). Prolonged nuclear retention of activated extracellular signal-regulated protein kinase promotes cell death generated by oxidative toxicity or proteasome inhibition in a neuronal cell line. The Journal of biological chemistry 277, 4010–4017. [DOI] [PubMed] [Google Scholar]
- Stanciu M, Wang Y, Kentor R, Burke N, Watkins S, Kress G, Reynolds I, Klann E, Angiolieri MR, Johnson JW, et al. (2000). Persistent activation of ERK contributes to glutamate-induced oxidative toxicity in a neuronal cell line and primary cortical neuron cultures. The Journal of biological chemistry 275, 12200–12206. [DOI] [PubMed] [Google Scholar]
- Stockwell BR, Friedmann Angeli JP, Bayir H, Bush AI, Conrad M, Dixon SJ, Fulda S, Gascon S, Hatzios SK, Kagan VE, et al. (2017). Ferroptosis: A Regulated Cell Death Nexus Linking Metabolism, Redox Biology, and Disease. Cell 171, 273–285. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Sun N, Zou X, Shi J, Liu X, Li L, and Zhao L (2005). Electroacupuncture regulates NMDA receptor NR1 subunit expression via PI3-K pathway in a rat model of cerebral ischemia-reperfusion. Brain research 1064, 98–107. [DOI] [PubMed] [Google Scholar]
- Takada A, and Bannai S (1984). Transport of cystine in isolated rat hepatocytes in primary culture. The Journal of biological chemistry 259, 2441–2445. [PubMed] [Google Scholar]
- Tan S, Schubert D, and Maher P (2001). Oxytosis: A novel form of programmed cell death. Current topics in medicinal chemistry 1, 497–506. [DOI] [PubMed] [Google Scholar]
- Thorn TL, He Y, Jackman NA, Lobner D, Hewett JA, and Hewett SJ (2015). A Cytotoxic, Co-operative Interaction Between Energy Deprivation and Glutamate Release From System xc-Mediates Aglycemic Neuronal Cell Death. ASN neuro 7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Tobaben S, Grohm J, Seiler A, Conrad M, Plesnila N, and Culmsee C (2011). Bid-mediated mitochondrial damage is a key mechanism in glutamate-induced oxidative stress and AIF-dependent cell death in immortalized HT-22 hippocampal neurons. Cell death and differentiation 18, 282–292. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Vagnozzi AN, Giannopoulos PF, and Pratico D (2018). Brain 5-lipoxygenase over-expression worsens memory, synaptic integrity, and tau pathology in the P301S mice. Aging cell 17. [DOI] [PMC free article] [PubMed] [Google Scholar]
- van Leyen K, Holman TR, and Maloney DJ (2014). The potential of 12/15-lipoxygenase inhibitors in stroke therapy. Future medicinal chemistry 6, 1853–1855. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Voisine C, Varma H, Walker N, Bates EA, Stockwell BR, and Hart AC (2007). Identification of potential therapeutic drugs for huntington's disease using Caenorhabditis elegans. PloS one 2, e504. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Wanders D, Stone KP, Forney LA, Cortez CC, Dille KN, Simon J, Xu M, Hotard EC, Nikonorova IA, Pettit AP, et al. (2016). Role of GCN2-Independent Signaling Through a Noncanonical PERK/NRF2 Pathway in the Physiological Responses to Dietary Methionine Restriction. Diabetes 65, 1499–1510. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Wang G, Shi Y, Jiang X, Leak RK, Hu X, Wu Y, Pu H, Li WW, Tang B, Wang Y et al. (2015). HDAC inhibition prevents white matter injury by modulating microglia/macrophage polarization through the GSK3beta/PTEN/Akt axis. Proceedings of the National Academy of Sciences of the United States of America 112, 2853–2858. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Wang GL, and Semenza GL (1993). General involvement of hypoxia-inducible factor 1 in transcriptional response to hypoxia. Proceedings of the National Academy of Sciences of the United States of America 90, 4304–4308. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Wang XZ, Lawson B, Brewer JW, Zinszner H, Sanjay A, Mi LJ, Boorstein R, Kreibich G, Hendershot LM, and Ron D (1996). Signals from the stressed endoplasmic reticulum induce C/EBP-homologous protein (CHOP/GADD153). Molecular and cellular biology 16, 4273–4280. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Ward RJ, Zucca FA, Duyn JH, Crichton RR, and Zecca L (2014). The role of iron in brain ageing and neurodegenerative disorders. The Lancet Neurology 13, 1045–1060. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Wei C, Zhang W, Zhou Q, Zhao C, Du Y, Yan Q, Li Z, and Miao J (2016). Mithramycin A Alleviates Cognitive Deficits and Reduces Neuropathology in a Transgenic Mouse Model of Alzheimer's Disease. Neurochemical research 41, 1924–1938. [DOI] [PubMed] [Google Scholar]
- Wei K, Wan L, Liu J, Zhang B, Li X, Zhang Y, Zhang C, and Yao W (2017). Downregulation of TRB3 protects neurons against apoptosis induced by global cerebral ischemia and reperfusion injury in rats. Neuroscience 360, 118–127. [DOI] [PubMed] [Google Scholar]
- Wenger RH, and Gassmann M (1997). Oxygen(es) and the hypoxia-inducible factor-1. Biological chemistry 378, 609–616. [PubMed] [Google Scholar]
- Wenzel SE, Tyurina YY, Zhao J, St Croix CM, Dar HH, Mao G, Tyurin VA, Anthonymuthu TS, Kapralov AA, Amoscato AA, et al. (2017). PEBP1 Wardens Ferroptosis by Enabling Lipoxygenase Generation of Lipid Death Signals. Cell 171, 628–641.e626. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Winterbourn CC (1995). Toxicity of iron and hydrogen peroxide: the Fenton reaction. Toxicology letters 82-83, 969–974. [DOI] [PubMed] [Google Scholar]
- Wirth EK, Bharathi BS, Hatfield D, Conrad M, Brielmeier M, and Schweizer U (2014). Cerebellar hypoplasia in mice lacking selenoprotein biosynthesis in neurons. Biological trace element research 158, 203–210. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Wu KJ, Polack A, and Dalla-Favera R (1999). Coordinated regulation of iron-controlling genes, H-ferritin and IRP2, by c-MYC. Science (New York, NY) 283, 676–679. [DOI] [PubMed] [Google Scholar]
- Wu QJ, and Tymianski M (2018). Targeting NMDA receptors in stroke: new hope in neuroprotection. Molecular brain 11, 15. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Xia B, Liu H, Xie J, Wu R, and Li Y (2015). Akt enhances nerve growth factor-induced axon growth via activating the Nrf2/ARE pathway. International journal of molecular medicine 36, 1426–1432. [DOI] [PubMed] [Google Scholar]
- Xiao N, Callaway CW, Lipinski CA, Hicks SD, and DeFranco DB (1999). Geldanamycin provides posttreatment protection against glutamate-induced oxidative toxicity in a mouse hippocampal cell line. Journal of neurochemistry 72, 95–101. [DOI] [PubMed] [Google Scholar]
- Yang B, Fritsche KL, Beversdorf DQ, Gu Z, Lee JC, Folk WR, Greenlief CM, and Sun GY (2019). Yin-Yang Mechanisms Regulating Lipid Peroxidation of Docosahexaenoic Acid and Arachidonic Acid in the Central Nervous System. Frontiers in neurology 10, 642. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Yigitkanli K, Zheng Y, Pekcec A, Lo EH, and van Leyen K (2017). Increased 12/15-Lipoxygenase Leads to Widespread Brain Injury Following Global Cerebral Ischemia. Translational stroke research 8, 194–202. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Yonezawa M, Back SA, Gan X, Rosenberg PA, and Volpe JJ (1996). Cystine deprivation induces oligodendroglial death: rescue by free radical scavengers and by a diffusible glial factor. Journal of neurochemistry 67, 566–573. [DOI] [PubMed] [Google Scholar]
- Zaman K, Ryu H, Hall D, O'Donovan K, Lin KI, Miller MP, Marquis JC, Baraban JM, Semenza GL, and Ratan RR (1999). Protection from oxidative stress-induced apoptosis in cortical neuronal cultures by iron chelators is associated with enhanced DNA binding of hypoxia-inducible factor-1 and ATF-1/CREB and increased expression of glycolytic enzymes, p21(waf1/cip1), and erythropoietin. The Journal of neuroscience : the official journal of the Society for Neuroscience 19, 9821–9830. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Zhang Z, Wu Y, Yuan S, Zhang P, Zhang J, Li H, Li X, Shen H, Wang Z, and Chen G (2018). Glutathione peroxidase 4 participates in secondary brain injury through mediating ferroptosis in a rat model of intracerebral hemorrhage. Brain research 1701, 112–125. [DOI] [PubMed] [Google Scholar]
- Zheng Y, Liu Y, Karatas H, Yigitkanli K, Holman TR, and van Leyen K (2019). Contributions of 12/15-Lipoxygenase to Bleeding in the Brain Following Ischemic Stroke. Advances in experimental medicine and biology 1161, 125–131. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Zilka O, Shah R, Li B, Friedmann Angeli JP, Griesser M, Conrad M, and Pratt DA (2017). On the Mechanism of Cytoprotection by Ferrostatin-1 and Liproxstatin-1 and the Role of Lipid Peroxidation in Ferroptotic Cell Death. ACS central science 3, 232–243. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Zille M, Karuppagounder SS, Chen Y, Gough PJ, Bertin J, Finger J, Milner TA, Jonas EA, and Ratan RR (2017). Neuronal Death After Hemorrhagic Stroke In Vitro and In Vivo Shares Features of Ferroptosis and Necroptosis. Stroke 48, 1033–1043. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Zille M, Kumar A, Kundu N, Bourassa MW, Wong VSC, Willis D, Karuppagounder SS, and Ratan RR (2019). Ferroptosis in Neurons and Cancer Cells Is Similar But Differentially Regulated by Histone Deacetylase Inhibitors. eNeuro 6. [DOI] [PMC free article] [PubMed] [Google Scholar]