Pharmacological Inhibition of Ferroptosis as a Therapeutic Target for Neurodegenerative Diseases and Strokes

Abstract Emerging evidence suggests that ferroptosis, a unique regulated cell death modality that is morphologically and mechanistically different from other forms of cell death, plays a vital role in the pathophysiological process of neurodegenerative diseases, and strokes. Accumulating evidence supports ferroptosis as a critical factor of neurodegenerative diseases and strokes, and pharmacological inhibition of ferroptosis as a therapeutic target for these diseases. In this review article, the core mechanisms of ferroptosis are overviewed and the roles of ferroptosis in neurodegenerative diseases and strokes are described. Finally, the emerging findings in treating neurodegenerative diseases and strokes through pharmacological inhibition of ferroptosis are described. This review demonstrates that pharmacological inhibition of ferroptosis by bioactive small‐molecule compounds (ferroptosis inhibitors) could be effective for treatments of these diseases, and highlights a potential promising therapeutic avenue that could be used to prevent neurodegenerative diseases and strokes. This review article will shed light on developing novel therapeutic regimens by pharmacological inhibition of ferroptosis to slow down the progression of these diseases in the future.


Introduction
Ferroptosis, a nonapoptotic mode of regulated cell death (RCD), was first described by Dixon in 2012. It is characterized by the iron-dependent oxidative modification of phospholipid membranes. [1] Ferroptosis is triggered by the toxic build-up of lipid peroxides on cellular membranes through inhibition of the antioxidant defense system and accumulation of iron-dependent ferroptosis to treat and prevent organ damage, has been convincingly established in NDs and strokes, thus opening new opportunities to treat these diseases using pharmacological inhibition of ferroptosis.
This review summarizes recent knowledge about the underlying regulatory mechanism of ferroptosis. Additionally, we discuss the potential pathophysiological roles of ferroptosis in the contribution to NDs and strokes. We review the emerging data from studies in treating NDs and strokes through pharmacological inhibition of ferroptosis. This review article suggests that pharmacological inhibition of ferroptosis by bioactive compounds (ferroptosis inhibitors) as a therapeutic target for NDs and strokes, and highlights a potential promising therapeutic avenue that could be used to negate and counteract NDs and strokes. This article provides a rationale for the development of new neuroprotective bioactive compounds, based on their ability to inhibit ferroptosis to slow down the progression of these diseases.

Ferroptosis: A Historical Perspective
The concept of ferroptosis was introduced by Stockwell lab in 2012 based on three major research areas which provide the foundational understanding of ferroptosis, i.e., the mechanisms of lipid and amino acid metabolism, [13][14][15] the control of ROS, [16,17] and the regulation of iron [18] (Figure 1). The pioneering work by Eagle have shown that deprivation of cysteine (an amino acid) can lead to cell death, [13] while endogenous synthesis of cysteine causes cell resistance to this type of cell death. [19,20] Ferroptosis was discovered in an attempt to find molecules selectively inducing death in cancer cells carrying an oncogenic form of rat sarcoma virus (RAS) mutant subtypes through highthroughput screening. During 2001-2003, Stockwell and coworkers discovered a new compound named as eradicator of RAS-transformed cells (erastin), which can induce nonapoptotic cell death through a high-throughput screen in cancer cells carrying an oncogenic form of RAS. [21] This pattern of cell death www.advancedsciencenews.com www.advancedscience.com was dependent on the accumulation of cellular iron and oxidative stress and inhibited by iron-chelating agents. [3] Following this discovery, Stockwell and co-workers performed a larger screen and identified two additional compounds, RAS-selective-lethal-3 small molecule (RSL3) and RSL5, works as synthetic compounds that selectively killed BJeLR cells in a similar iron-dependent oxidative non-apoptotic cell death in 2008. [22] Almost at the same time, Conrad and co-workers revealed that genetic ablation of GPX4 resulted in a LPO-induced non-apoptotic cell death that could be inhibited by alpha-tocopherol [23] and that enforcing overexpressed the light chain of cystine/glutamate antiporter (system Xc − ) protected cells from this type of cell death. [24] In 2012, Stockwell and co-workers found that erastin caused an iron-dependent, non-apoptotic mode of cell death through inhibiting cystine uptake by system Xc − accompanying with the accumulation of lipid ROS in human fibrosarcoma HT-1080 cells, involving an unique constellation of biochemical, morphological, and genetic features. [1] According to those pioneering works, Stockwell and co-workers proposed the concept of ferroptosis that was inhibited by a specific inhibitor of ferroptosis, such as ferrostatin-1(Fer-1). [1]

Molecular Mechanisms of Ferroptosis
Ferroptosis reflects a delicate antagonistic balance between inducers of ferroptosis and ferroptosis defense systems. When ferroptosis-promoting factors that include iron-dependent ROS and LPO significantly override the antioxidant defense systems, a lethal accumulation of lipid peroxides on cellular membranes results in membrane rupture, leading to ferroptotic cell death [2] (Figure 2). Ferroptosis defense systems include the SLC7A11reduced glutathione (GSH)-glutathione peroxidase 4 (GPX4) system, the dihydroorotate dehydrogenase-dihydroubiquione (DHODH-CoQH2) system, the ferroptosis suppressor protein 1-ubiquinol (FSP1-CoQH2) system, and the GTP cyclohydrolase 1-tetrahydrobiopterin (GCH1-BH4) system. [2] Ferroptosis is mainly driven by iron accumulation and LPO, and subsequent plasma membrane rupture. [25] The induction of ferroptosis needs two key signals, namely the accumulation of free iron and the inhibition of antioxidant SLC7A11-GSH-GPX4 system. [26] There are many connections between lipid reactive oxygen species and each metabolic process (Figure 3).

Iron Homeostasis
Iron, an essential redox-active metal for normal physiological functions, is important for neuronal activity and plays an important role in energy metabolism, neurotransmitter biosynthesis, and myelination. [27] In the human body, the cellular iron metabolism is regulated by transport, absorption and recycling of iron. At the transcriptional level, iron was regulated by iron response elements (IRE) and iron regulatory proteins (IRP), [28] avoiding the accumulation of iron leading to iron overload, which is highly harmful to tissues. [8,29] Iron homeostasis is highly regulated and is crucial to maintain its normal function in the brain. [28] The dysregulation of iron within the brain causes oxidative stress-dependent injury to neurons, leading to neurological diseases. Iron-dependent ROS and LPO are of great significance in the regulation of ferroptosis. Iron is involved in free radical formation and propagation of LPO.
The transferrin receptor 1 (TfR1), which is expressed on the luminal side of the brain microvascular endothelial cells (BMECs), is responsible for regulating the content of iron transported into the brain through the blood-brain barrier (BBB) to maintain iron homeostasis for proper brain function. [30] The iron (Fe 3+ ) binds to circulating transferrin (Tf) to form a complex holo transferrin (holo-Tf), [31] which binds to TfR1 and then enters into the BMECs through endocytosis mediated by clathrin. [32] The Fe 3+ -Tf-TfR complex is then transported into cells to the endosome where Fe 3+ detaches from Tf, and is reduced to Fe 2+ by metalloreductases six-transmembrane epithelial antigen of prostrate 3 (STEAP3) [33] or duodenal cytochrome b (DCYTB). [34] Fe 2+ then enters the cytosol by divalent metal transporter 1 (DMT-1). [35] The unbound redox-active iron (Fe 2+ ) in the cytosol constitutes a labile iron pool (LIP), which acts as an intermediate between stored, utilized, or imported iron, feeds iron-catalyzed ROS production. [36] The utilized unbound Fe 2+ can be transported to the mitochondrion for the biosynthesis of heme, formation of iron-sulfur clusters, and as co-factors for mitochondrial enzymes. Also, the unbound Fe 2+ can be stored in a non-toxic form as Fe 3+ by ferritinin, which consists of two isoforms, i.e. heavy (FTH) and light (FTL). FTH has ferroxidase activity, involving in rapid iron uptake and reutilization. FTL plays a role in long-term storage of iron through nucleation. [27,28,37,38] Ferritin can be degraded through ferritinophagy, an autophagy-like process by which releases labile iron and facilitates LPO driving ferroptosis. [39] The unbound excess Fe 2+ can be exported from the cells via ferroportin (FPN) with the help of ceruloplasmin or hephaestin to oxidize Fe 2+ to form Fe 3+ . [40,41] Cellular iron homeostasis is maintained by iron regulatory proteins (IRP1 and IRP2), [42,43] which bind to the iron response element (IRE) of iron metabolizing genes and regulate their expression, leading to changes in cellular iron metabolism. [44,45] When intracellular iron is low, Fe-S is released from the active site of the IRP, permitting IRP binds to IRE of DMT1 and TfR gene transcripts to activate their translation, while IRPs bind to FPN gene transcripts to inhibit its translation, resulting in promoting absorption of free iron to increase intracellular dose and reducing cellular iron excretion. [8,29,46] The accumulation of free iron is involved in the accumulation of lethal lipid peroxides and the initiation of ferroptosis. [26] Iron catalyzes the Fenton reaction or acts as an essential cofactor for cytochrome P450 oxidoreductase (POR) or arachidonate lipoxygenases (ALOXs), which promote LPO. [2] The Fenton reaction induced the generation of free radicals that attack membranebound PUFAs in a nonenzymatic pathway, while nonheme ironcontaining lipoxygenases (LOXs) initiate the dioxygenation of membrane phospholipids containing PUFAs, i.e. linoleic acid (LA, 18:2 n-6) to arachidonic acid (AA, 20:4 n-6) in the enzymatic LPO pathway. [47] Due to its ability of oxidation-reduction, the iron (Fe 2+ ) converts hydrogen peroxide (H 2 O 2 ) to toxic hydroxyl radicals (HO • ), an important ROS acting on LPO, resulting in oxidative stress and cell death via the Fenton and Haber-Weiss reaction, known as: Fe 2+ + H 2 O 2 → Fe 3+ + OH • + OH − , first described in 1894. [48] The free radicals generated near membranes react with PUFAs, i.e., initiate peroxidation of In nonenzymatic LPO, PUFAs are ligated with CoA by ACSL4 to produce acyl-CoA, which can be re-esterified in phospholipids through LPCATs to produce PL. PUFAs can be scavenged from the environment and dietary sources and can be synthesized from the basic building block acetyl CoA, through the action of acetyl CoA carboxylase (ACC). [39] Once polyunsaturated acyl tails (PL-PUFAs) are incorporated into membrane environments, ALOXs and cytochrome P450 oxidoreductase (POR), and labile iron use molecular oxygen (O 2 ) to do a peroxidation reaction, leading to the generation of peroxidated PL-PUFAs (PL-PUFA-OOH). [39,52] This process requires hydrogen peroxide (H 2 O 2 ) derived from an iron-dependent Fenton reaction, or POR and NADPH oxidase (NOX) activation, or mitochondria electron transport chain pathways. The last step of ferroptosis, LPO or its secondary products such as 4-HNE and MDA cause pore formation in plasma or organelle membranes, eventually triggering . Metabolic processes impinging on the susceptibility of cell toward ferroptosis. Insufficiently controlled intracellular iron storage and PUFAenriched phospholipids are prerequisite for the execution of cell death by ferroptosis. Accordingly, MUFAs that compete with PUFAs for incorporation into phospholipids exert antiferroptotic effects. Although mitochondria are dispensable for the execution of ferroptosis, under certain conditions TCA cycle fueled by glutaminolysis may play a major role in ferroptosis induction. Ferroptosis is governed by at least four defense systems with different subcellular localizations to detoxify lipid peroxides and thus protect cells against ferroptosis, i.e., cytosolic GPX4 (GPX4 cyto ) cooperates with FSP1 on the plasma membrane (and other non-mitochondrial membranes) and mitochondrial GPX4 (GPX4 mito ) cooperates with DHODH in the mitochondria to neutralize lipid peroxides. The SLC7A11/GSH/GPX4 axis, the GCH1/BH4 axis, and the FSP1/CoQ10 axis are fueled by NADPH. Furthermore, the SLC7A11/GSH/GPX4 axis depends on the cystine glutamate antiporter system Xc − and the transsulfuration pathway providing cysteine, while the FSP1/CoQ10 axis relies on the mevalonate pathway that generates CoQ10. cell death. In recent years, ferroptosis has gained substantial attention and sparked great interest in the research community of neurological diseases, and targeting ferroptosis through pharmacological inhibition might provide new therapeutic opportunities to treat neurological diseases. [53,54]

SLC7A11-GSH-GPX4 Axis
The cellular antioxidant systems that directly neutralize lipid peroxides is one of the ferroptosis defense systems. There exist four anti-ferroptosis defense systems with distinctive subcellular localizations. The SLC7A11-GSH-GPX4, which is related to amino acid metabolism, is well-defined and believed to constitute the major cellular system to protect against ferroptosis. [2,8] Solute carrier family 7 member 11 (SLC7A11; also known as xCT) and solute carrier family 3 member 2 (SLC3A2) consists of the system Xc − . [55,56] xCT, the transporter subunit, is a core component of system Xc − . xCT import extracellular cystine and export intracellular glutamate to mediate the antiporter activity of system Xc − . [55,57] The extracellular cystine taken up by SLC7A11 is rapidly reduced to cysteine through a nicotinamide adenine dinucleotide phosphate (NADPH)-consuming reduction reaction in the cytosol. Then cysteine serves as the rate-limiting precursor for the biosynthesis of GSH, which is a principle cofactor for GPX4-mediated LPO detoxification. [56] Blocking transporter activity of SLC7A11 or depleting cystine in culture media can induce ferroptosis in various cancer cells. [56] GPX4 belongs to the GPX protein family with activity of lipid repair enzyme, [58,59] and was identified as a key inhibitor in ferroptosis. [1,60,61] GPX4 can convert PL hydroperoxides (L-OOH) to non-toxic lipids PL alcohols and simultaneously oxidize two GSH to oxidized glutathione (GSSG). [16,62] Previous studies have established the critical role of GPX4 in ferroptosis suppression through genetic or pharmacological manipulation. [63,64] GPX4 is regulated by epigenetic, transcriptional, and post-translational modifications (PTMs) such as phosphorylation, ubiquitination, succination, and glycosylation. [65]

The FSP1-CoQH 2 System
Coenzyme Q 10 (CoQ 10 ) is found in diverse membranes throughout cells, including mitochondria, and serves as a second endogenous mechanism for protecting membranes against lipid peroxidation that drives ferroptosis. The ferroptosis suppressor protein 1 (FSP1; also known as AIFM2 [66] ) was firstly identified to operate independently of GPX4 to defend against ferroptosis. [67,68] FSP1 localizes on the plasma membrane (as well as on other subcellular compartments) and its plasma membrane localization appears to be both necessary and sufficient for the function of FSP1 in suppressing ferroptosis. [67,68] FSP1 functions as an NAD(P)H-dependent oxidoreductase to reduce ubiquinone (also known as coenzyme Q 10 ) to regenerate the reduced form of CoQ 10 ubiquinol (CoQ 10 -H 2 ), which can trap lipid peroxyl radicals, thereby suppressing lipid peroxidation and ferroptosis. [50,51] In some cases, FSP1 inhibits ferroptosis by activating ESCRT-III-dependent membrane repair instead of its oxidoreductase function. [69] The ESCRT III complex repairs damage to the plasma membrane and slows cell death by ferroptosis. [70]

The GCH1-BH 4 System
Recent studies have shown that GTP cyclohydrolase 1 (GCH1)tetrahydrobiopterin (BH 4 ) system functions as another critical GPX4-independent suppressor of ferroptosis through suppressing lipid peroxidation. [71,72] BH 4 is a cofactor of aromatic amino acid hydroxylases and other enzymes, and GCH1 mediates the rate-limiting reaction in the BH 4 biosynthesis pathway. GCH1, which generates the endogenous metabolite BH 4 , was discovered in a CRISPR activation screen as a suppressor of ferroptosis, [71] as well as an enhancer of ferroptosis in a CRISPR loss-of-function screen. [72] GCH1 suppresses ferroptosis through the generation of BH 4 as a radical-trapping antioxidant as well as via GCH1mediated production of CoQH 2 and PL-PUFAs.

The DHODH-CoQH 2 System
The DHODH-CoQH 2 system is a newly identified GPX4independent mitochondria-localized ferroptosis defense that can compensate for GPX4 loss to detoxify mitochondrial lipid peroxidation. [73] DHODH, an enzyme involved in pyrimidine synthesis that can reduce CoQ10 to CoQH 2 in the inner mitochondrial membrane, functions by reducing mitochondrial CoQ10, analogous to the function of FSP1 in extra mitochondrial membranes. [73] Following the inactivation of GPX4, the flux through DHODH is significantly increased, leading to promoting CoQH 2 generation, which neutralizes lipid peroxidation and inhibits ferroptosis in mitochondria. [73] Inactivation of both GPX4 and DHODH in mitochondria triggers robust ferroptosis through unleashing potent mitochondrial lipid peroxidation. Cells with a low expression of GCH1 or DHODH confer more sensitivity to ferroptosis, and those with high expression confer resistance to ferroptosis.

Parkinson's Disease
Parkinson's disease (PD) is a neurodegenerative disease characterized by loss of dopaminergic neurons in the substantia nigra and the classical motor features of parkinsonism are associated with Lewy bodies. [74] The pathogenesis of PD remains largely unknown, but seems to develop as a result of complex interaction among multiple predisposing genes and environmental factors. [75] The oxidative stress hypothesis of PD holds that oxidative stress leads to the neurodegeneration of dopaminergic neurons and results in the pathogenesis of PD, which was proposed in 1992, based on the observation that increased oxidative stress by PD-related neurotoxins, increased LPO and decreased GSH, increased iron and reduced ferritin and decreased activities of glutathione peroxidase and catalase of in the SN of patients with PD. [76] The PD pathological hallmarks include iron dyshomeostasis, LPO, GSH depletion, and oxidative stress are all related to ferroptosis. After coined the term ferroptosis, which is characterized by iron dysregulation, LPO, and GSH depletion, scientists begin to revisit the contributions of aforementioned signatures of PD to the disease development, and a ferroptosis hypothesis of PD is emerging, which claims that ferroptosis is involved in the pathogenesis of PD [47,[77][78][79][80] (Figure 4). The direct relationship between ferroptosis and PD was confirmed in vitro in organotypic slice cultures and in vivo in the MPTP mouse model. [81,82] Role of Iron and Ferroptosis in PD: Iron dys-homeostasis and accumulation contribute to progression of neuropathology through ROS production and are a common feature of PD. [47] Ample evidence has shown that increased iron in glia and dopaminergic neurons in the substantia nigra pars compacta (SNpc) in PD [83][84][85][86][87][88] is associated with disease severity. [89,90] The microglial iron and ferritin are elevated in the SN in the animal models of PD induced by 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP) and 6-hydroxydopamine (6-OHDA). [91] The iron accumulation in PD probably results from the dysregulation of IRP triggered iron control homeostatic mechanisms, resulting in increased intracellular iron import or decreased export, which constitute vulnerability factors for PD development. [81,[92][93][94] PD patients with mutations in iron-bound proteins are more prone to develop PD. [92][93][94][95] Elevated DMT1 levels in the SNpc of PD patients and PD mouse models are likely related to an increased cellular iron import. [96,97] It has been shown that the levels of iron storage protein ferritin decreased in the SN from post-mortem brains of patients with PD and PD models, [98] and ferroportin (iron efflux protein) is decreased in MPTP or 6-OHDA-induced PD models. [99][100][101] Ferritin, the main iron storage . Potential molecular mechanisms of ferroptosis in PD development. The activated microglia and astrocytes release the inflammatory cytokines (IL-1 , IL-6, and TNF-), which promote iron accumulation in neurons by downregulating FPN1 and upregulating DMT1. IL-6 promotes the release of hepcidin from astrocytes and thus prevents iron release from neurons mediated by FPN1. Activated microglia release ROS to promote neuronal oxidative stress. Upregulation of Nrf2 and the release of MTs in astrocytes protect neuron against oxidative stress. Two types of glia regulate glutamate-mediated toxicity and influence ferroptosis. Astrocytes effectively control the level of glutamate in the synaptic cleft by regulating GSH synthesis and inhibiting neuronal xCT. Activated microglia release TNF-, promoting neuronal glutamine release. The gene mutations of LRRK2 and PARKIN contribute to excess iron accumulation in PD. The accumulation of iron potentiates -synuclein-mediated neurotoxicity. The -synuclein induces oxidative stress and lipid peroxidation.
protein important in maintaining iron homeostasis, confers protection against iron-mediated ferroptosis through chelating iron. [102] Astrocytes increased ferritin release to respond to iron overload, which might inhibit iron-mediated oxidative damage and ferroptosis of dopamine (DA) neurons in PD. [102] SEC24Bregulated ferroptosis in microglia contributes to neurodegeneration in PD. [103] The microglia grown in a human induced pluripotent stem cell (iPSC)-derived triculture system that contains microglia, neurons, and astrocytes are highly responsive to iron and susceptible to ferroptosis. Removal of microglia from the triculture system substantially delayed iron-induced neurotoxicity, suggesting a critical role for microglia iron overload and ferrop-tosis in neurodegeneration in PD patients. [103] Taken together, there exists a dysregulation of iron influx and efflux mechanisms in PD, which contribute to the accumulation of intracellular iron that causes susceptibility of iron-mediated free radical formation and ferroptosis.
LPO and Ferroptosis in PD: The oxidative stress and iron dependent increased LPO leading to nigral cell death was first reported by Dexter in 1986. [104] Postmortem analyses of brain found that PUFAs, but not MUFAs, decreased in the SN of PD patients. [104] The levels of MDA were elevated in SN [105] and levels of lipid hydroperoxide (LOOH) in plasma is increased. [106] A previous study showed that 4-HNE was associated with Lewy www.advancedsciencenews.com www.advancedscience.com bodies in the SN of PD patients. [107] The elevated 4-HNE in the cerebrospinal fluid (CSF) of PD patients is positively correlated with an increase of iron in the SN. [107,108] Recent studies have demonstrated that LPO is increased in PD in vitro and in vivo models induced by rotenone, [109] 6-OHDA, [110][111][112][113] MPP + , [114] and MPTP. [115] Another study showed a direct association between endogenous -synuclein ( -syn) levels and the sensitivity of dopaminergic neurons to lipid peroxidation and ferroptosis via modulation of plasma membrane ether-linked phospholipids. [116] This study uncovered that -syn functions as a positive modulator of ferroptosis, supporting ferroptosis as a key mechanism involved in the pathology of PD and providing potential ferroptosis-based therapeutic opportunities in PD. [116] This observation was supported by other studies, which demonstrated that -syn overexpression sensitizes neuronal cells to ferroptosis induction through suppressing Nrf2 protein. [117] Arachidonic acid (AA) cotreatment with iron, increases AA-containing phospholipid and synergistically increased the susceptibility of dopaminergic neurons to ferroptosis through promoting high lipid peroxidation, which can be rescued by Ferrostatin-1. Inhibiting ACSL4 and lipoxygenases 15/15B attenuate dopaminergic neuronal cell death. [118] The mutations of calcium-independent phospholipase A 2 (iPLA 2 , PLA2G6, or PNPLA9 gene) are likely involved in PD pathogenesis. iPLA 2 plays a critical part in ferroptosis by regulating the metabolism of PUFAs. iPLA 2 can act as an antiferroptotic guardian by eliminating the proferroptotic signal. Deficiency in iPLA 2 , caused by genetic factors or chemical or pharmacological poisoning, may be associated with increased sensitivity to ferroptotic cell death in PD. [119] The Role of Inhibition of GSH/GPX4 Axis in PD: Decreased GSH metabolism and a decreased GSH/GSSG ratio are associated with oxidative stress in PD. [120,121] GPX4 is decreased in the SN of PD patients and increased in the surviving dopaminergic neurons, indicating that upregulation of GPX4 may act as a compensatory mechanism to protect surviving neurons against ferroptosis-induced injury during disease progression. [122] Meanwhile, the expression of SLC7A11 is downregulated in the SN of PD patients, which may contribute to a reduced GSH levels in the SN of PD. [123] Paraquat (PQ), a neurotoxicant which is linked to increased PD risk and PD-like neuropathology, can promote neurotoxicity of dopaminergic neurons through downregulating SLC7A11/GPX4, upregulating Cox2 expression, and causing iron accumulation through ferritinophagy pathway induced by NCOA4. [124] Recent studies have revealed that PD-related neurotoxins downregulate the expression of SLC7A11 by rotenone [109] and MPP + , [114] and downregulate the expression of GPX4 by rotenone, [109] 6-OHDA, [110,112] and MPP + [125] in PD in vitro and in vivo models.
The Role of FSP1-CoQH 2 System in PD: An iron-free form of ferritin, apoferritin improves motor deficits in MPTP treated mice by inhibiting iron aggregation through down-regulating iron importer divalent metal transporter 1 (DMT1) and ferroptosis, which was resulting from effectively down-regulating ACSL4 and the up-regulating FSP1. [126] The Role of GCH1-BH 4 System in PD: GCH1 catalyzes the rate-limiting step in tetrahydrobiopterin (BH4) synthesis, an essential cofactor in the synthesis of monoaminergic neurotransmitters, including dopamine (DA). It was found that the level of BH4 and activity of GCH1 were markedly decreased in the sub-stantia nigra and striatum of PD patients. [127,128] Loss-of-function mutations in GCH1 lead to striatal DA depletion and nigrostriatal cell loss. [129] Mounting genetic association studies have also identified that GCH1 variants increased the risk for PD. [129][130][131][132] The putative damaging variants of GCH1 contributed to the collective risk for early-onset Parkinson's disease (EOPD) in ethnic Chinese. [133] An early study reported a lower BH4 concentration in the CSF of PD patients. [134,135] These studies suggest that there exist an impairment of GCH1-BH4 system in PD. Most recent study revealed a mechanism that GCH1 deficiency may contribute to PD. GCH1 −/− zebrafish develop a marked deficiency of monoaminergic neurotransmitter. Tyrosine hydroxylase (Th) protein levels were markedly reduced. L-DOPA treatment improved survival without ameliorating the motor phenotype in GCH1 −/− larvae. Microglial activation was found in GCH1 −/− larvae. These results suggest that GCH1 deficiency may unmask early, subclinical parkinsonism and only indirectly contribute to neuronal cell death via immune-mediated mechanisms. [136] The Role of PD-Related Regulator in PD: Specific regulators play a role in modulating ferroptosis and PD. 6-OHDA upregulates super-enhancer-driven sorting nexin 5 in PD rats. Superenhancer-driven sorting nexin 5 promotes dopaminergic neuronal ferroptosis in 6-OHDA-lesioned SNc of PD rats. [137] Silencing SNX5 significantly inhibit ferroptosis in 6-OHDA-induced PC12 cells, suggesting the correlation between the SNX5, ferroptosis, and PD. Overall, SNX5 functions as an inducer of ferroptosis in PD. [137] This observation was corroborated by study from the same group, which reported SNX5 promotes ferroptosis in the PD model. [138] Mutations and increased leucine-rich repeat kinase 2 (LRRK2) kinase activity are associated with both familial and idiopathic PD pathology. The wild-type cells were more resistant to ferroptosis than the LRRK2 knockout RAW 264.7 murine macrophages. Inhibition of the LRRK2 kinase increased the sensitivity of cell to erastin. These results suggest that LRRK2 protects against erastin-induced ferroptosis in RAW macrophages. [139]

Alzheimer's Disease
Alzheimer's disease (AD) is the most common neurodegenerative disease characterized by cerebral atrophy, accumulation of amyloid beta-peptide (A ) in senile plaques (SPs), and intraneuronal accumulation of hyperphosphorylated microtubuleassociated protein tau in neurofibrillary tangles (NFTs), resulting in a progressive decline in cognitive function. [140] The iron dyshomeostasis, LPO, GSH depletion, and oxidative stress are implicated in the pathogenesis of AD. [141][142][143] After the introduction of the term ferroptosis, which is characterized by iron dysregulation, LPO, and GSH depletion, the science community begins to revisit the contributions of aforementioned signatures of AD to the disease development, and proposed"Ferroptosis hypothesis in Alzheimer's disease," which claims that ferroptosis is involved in the pathogenesis of AD. [144][145][146][147][148] The involvement of ferroptosis in the pathogenesis of AD result from iron metabolism dysregulation, LPO, and the inhibition of GSH-GPX4 axis ( Figure 5).
The Role of Iron Metabolism Dysregulation in AD: Elevated iron in the brains of patients with AD was first revealed in 1953. [141] This observation was supported by Aging, inflammation, and oxidative stress could dysregulate the iron transport proteins and cause iron retention. Increased irons could be exported by FPN1/Cp or FPN1/Heph with the help of APP, which is transported to stabilize FPN1 by soluble tau protein. The overload Fe 2+ could upregulate the expressions of ferritin, FPN1, and APP by IRP-IRE interactions, while suppress the function of furin, thereby enhancing -secretase activity by reducing furin protein expression and thus accelerating A production and deposition. Aberrant phosphorylation of the tau protein can lead to increased APP and A 40 aggregation. A directly down-regulate FPN and decrease levels of GPX4 and elevate levels of ferritin. Excessive iron in neurons can lead to tau hyperphosphorylation and NFT formation. When xCT or GSH decreases in neurons, the decreased GPX4 cannot exert the function of anti-lipid peroxidation. After Fenton reaction or ALOX-catalyzed process, PUFA-OOH can accumulate to a lethal level to trigger ferroptosis, which could be responsible for the tau hyperphosphorylation, A formation, and neuronal loss.
subsequent post-mortem [84,[149][150][151][152] and in vivo [145,149,152 ] studies. Ample evidence has shown that aberrant levels of iron regulatory molecules, such as elevated CSF ferritin, [153] hepcidin, [154] Tf (iron transport), [155,156] ferritin (iron storage), and decreased ferroportin (Fpn, iron export), [157,158] are associated with the pathogenesis of AD. Elevated iron occurs even at the early stage as mild cognitive impairment of AD [159,160] and result in cognitive decline prior to the disease. [161,162] Excess free iron are found within SPs, which upregulates ferritin expression, and impairs spatial memory in AD mouse models. [163,164] A is generated from amyloid precursor protein (APP), a single-pass transmembrane protein that helps protect cells from iron-mediated oxidative stress by loading Fe 3+ onto transferrin and stabilizing FPN. [152] APP knockout mice exhibited decreased FPN and increased iron accumulation with concurrent oxidative stress in cortical neurons. [152] High level of iron upregulates APP transla-tion, which is modulated by an IRE. Iron binds to and promotes toxicity of A and tau. [165][166][167][168][169][170] Iron overload injures microglia through increasing ROS production. Meanwhile, some ferroptosis inhibitors, such as vitamin E and desferrioxamine (DFE) have shown clinical benefits for the patients with AD. [171,172] Loss of Fpn induces memory impairment by promoting ferroptosis in AD. [173] Fpn was downregulated in the brains of APPswe/PS1dE9 mice as an AD mouse model and AD patients. Deletion of Fpn in mice led to AD-like hippocampal atrophy and memory deficits. Inhibitors of ferroptosis effectively reduced the neuronal death and memory impairments induced by A aggregation in vitro and in vivo. Restoring Fpn ameliorated ferroptosis and memory impairment in APPswe/PS1dE9 mice. [173] Decreased FTH1 and SAT1 resulted in a decrease in the accumulation of lipid ROS and eventually result in ferroptosis of astrocytes in AD, suggesting that activated ferroptosis in astrocytes may contribute to the pathophysiological process in the entorhinal cortex in AD. [174] Recent study showed that A 1-40 caused blood-brain barrier (BBB) disruption by upregulating CD36 expression in pericytes via inducing pericyte mitophagy-dependent ferroptosis through the CD36/PINK1 /Parkin pathway, uncovering a molecular mechanism by which pericytes of the BBB disrupted in AD. [175] A caused spatial learning and memory impairment, along with intracellular A deposits, which were reversed by treatment with Fer-1. A decreased the expression of GPX4 and SLC7A11 and increased the level of TfR, suggesting ferroptosis are involved in A neurotoxicity. [176] ORMDL3 gene, a susceptibility gene closely related to the occurrence of childhood asthma, promote AD through inducing ferroptosis by PERK/ATF4/HSPA5 signaling pathway. [177] The Role of LPO in AD: The mammalian brain is inadequately equipped with antioxidant defense systems and is prone to oxidative damage due to its high O 2 consumption, and neuronal membrane lipids rich in high PUFAs side-chains, especially DHA (C22:6) residues, [178] Iron is found throughout the brain, [28,179,180] neuronal mitochondria generate O 2 ·− , and brain metabolism generates a lot of H 2 O 2 . [178] LPO can generate some aldehyde byproducts, including 4-HNE, malondialdehyde (MDA), and acrolein. [181] Previous studies have shown increased LPO and decreased PUFAs in the brain of patients with AD, and increased 4-HNE in ventricular fluid, among other markers of oxidative stress. [143,[182][183][184] Mounting evidence indicates elevated levels of 4-HNE [185][186][187] and acrolein [186,187] in AD brains. The increased 4-HNE and acrolein also occur in MCI and early stage of AD, suggesting that LPO is an early event in the pathogenesis of AD. [186,187] 4-HNE exert cytotoxicity to neurons through increasing Ca 2+ levels, damaging neurofilament proteins, and inactivating glutamate transporters. [188,189] Co-localization of A plaques and LPO products were observed in the brain of AD patients. [190] A peptides can induce LPO, which increases APP processing, resulting in a vicious circle between A and LPO in AD pathology. [191] Previous studies indicate that some enzymes responsible for LPO, including LOXs, [192][193][194] COXs, [195,196] and NOXs, [197,198] are increased in AD [199,200] and play vital roles in the pathogenesis of AD. Iron-chelating agents such as desferrioxamine can largely inhibit LPO of isolated brain tissues. [201] Pharmacologically inhibiting 12/15-LOX can reverse AD-like phenotype in a mouse model. [199] A recent study has shown that NADPH oxidase 4 (NOX4), a major source of ROS, was significantly elevated in impaired astrocytes of the cerebral cortex from patients with AD and the APP/PS1 double-transgenic mouse model of AD. Overexpression of NOX4 promotes cytotoxicity by inducing ferroptosis through the activation of lipid peroxidation in human astrocytes. [202] In general, dysregulation of LPO and perturbation of lipid metabolism are involved in the pathophysiology of AD.
The Role of Inhibition of GSH/GPX4 Axis in AD: GSH is the most abundant nonprotein thiol to exert antioxidant defense role and maintain redox homeostasis in neurons. The depletion of GSH in the brain is a common finding and is linked to loss of neurons, resulting in neurological diseases such as PD, stroke, and AD. [203,204] GSH maintains redox homeostasis through binding to Fe 2+ in the LIP to inhibit the irondependent oxidization [205] and functions as the substrate of GPX4-mediate lipid detoxification. [63,64] Decreased levels of GSH in the brain were found in both animal models [206,207] and autopsy specimens [208,209] of AD. A treatment induced depletion of GSH in cultured neurons. [210] Recent evidence indicates that depletion of GSH destroyed redox homeostasis and is associated with the occurrence of ferroptosis in AD. [211,212] Moreover, a reduction of brain GPX4 level was observed in AD mice. [213] Inhibition of GPX4 decreases the sensitivity of brain to ferroptosis in AD pathogenesis. [63] Overexpression of GPX4 protects cortical neurons against A -induced cytotoxicity through suppressing LPO. [214] Ablation of GPX4 can cause a neuronal loss in the hippocampal regions in neonatal [23] and adult mice, [215] accompanied by elevated activation of astrocytes, [215] suggesting an suppressive function of GPX4 in neurodegenerative disorders. The conditional deletion in forebrain neurons of GPX4 in mice leads to cognitive impairment and hippocampal neurodegeneration. [212] Moreover, ferroptosis inhibitors can reverse GPX4 ablation-induced elevation of LPO, activation of ERK and augmentation of neuroinflammation, which are markers associated with ferroptosis, thereby ameliorating neurodegeneration. [212] Overexpression of GPX4 inhibited neuronal loss and the production of lipid ROS in frontal cortex, accompanied by a reduced amyloid plaque formation in frontal cortex tissues and improved learning and memory abilities in 5×FAD (5×FAD/GPX4) mice. [216] 5×FAD/GPX4 mice exhibited attenuated markers of ferroptosis, evidenced by decreased 4-HNE, supporting the notion that ferroptosis is a key contributor to AD pathogenesis. [216] A 1-42 induced PC12 cell death via inducing VDAC1-dependent ferroptosis, which was reversed by inhibition of VDAC1, thereby activating the AMPK/mTOR and Wnt/catenin pathways. [217] These findings strongly suggested that inhibition of the GSH/GPX4 axis is associated with AD pathology.
The Role of FSP1-CoQH 2 System in AD: Recent study has shown that the upregulation of FSP1 is involved in the inhibition of a ketogenic diet for chronic sleep deprivation-induced AD. [218] Ketogenic diet prevents the chronic sleep deprivationinduced cognitive deficiency, amyloid deposition, and hyperphosphorylated tau protein. Ketogenic diet inhibits iron dyshomeostasis by down-regulating the expression of TfR1 and DMT1 and upregulating the expression of FTH1 and FPN1. Ketogenic diet promoted the elevation of xCT/GPX4 axis, FSP1, and reduced MDA. Meanwhile, ketogenic diet activated Sirt1/Nrf2 signaling pathway in the hippocampus in SD-exposed mice. These results suggested that KD could prevent chronic SD-induced AD through inhibiting ferroptosis via Sirt1/Nrf2 signaling pathway, thereby upregulating GPX4 and FSP1. [218] The Role of AD-Related Regulator in AD: Mutations in presenilin 1 and 2 (PS1 and PS2) or the amyloid-(A ) precursor protein (APP), all of them are involved in the generation of A , cause autosomal dominant familial AD (FAD). A recent study demonstrates that the potential for presenilin mutations to promote neurodegeneration irrespective of A is through sensitizing multiple cell types to ferroptosis. [219] Mutant PS1(mPS1) sensitizes cells to ferroptosis. mPS1 promotes the expression of GPX4 by quenching the membrane propagation of lethal hydroperoxyl radicals. Presenilin -secretase activity cleaves Notch-1 to signal LRP8 expression, which then controls GPX4 expression by regulating the supply of selenium into the cell since LRP8 is the uptake receptor for selenoprotein P. mPS1 disrupts selenium uptake, thus suppressing GPX4 expression. Presenilin mutations may promote neurodegeneration by www.advancedsciencenews.com www.advancedscience.com promoting ferroptosis, which highlights a disease-modifying concept for therapeutics. [219] Allelic variation of APOE gene confers the greatest genetic risk for sporadic AD (SAD). Recent study reported that apoE activates the PI3K/AKT pathway to inhibit ferritinophagy, thus inhibiting iron-dependent LPO. [220] Postmortem study has revealed that there is a strong association of iron with pathologically confirmed clinical AD, particularly in those with the adverse APOE-4 allele. APOE 4 carriers increased susceptibility to ferroptosis in AD. APOE 4 carriers express high levels of oxidation-sensitive PUFAs that promote ferroptosis. The decreased level of apoE boost ferritinophagy and iron release from ferritin, which further increases LPO and ferroptosis. [220]

Amyotrophic Lateral Sclerosis
Amyotrophic lateral sclerosis (ALS), also known as motor neuron disease (MND), is a devastating, progressive, and late-onset neurodegenerative disease affecting the upper motor neurons in the motor cortex and lower motor neurons in the brainstem and spinal cord, leading to progressive muscle weakness and death from respiratory failure within 2-5 years of symptom onset. [221][222][223] The neuropathological hallmark of ALS is the aggregation and accumulation of ubiquitinated proteinaceous inclusions in the motor neurons.90% of ALS cases are considered sporadic ALS (sALS) without correlation with family history, while about 10% of people with ALS show some form of family history, classifying their condition as familial ALS (fALS). [223] More than 30 genes have been identified which are causative in or confer an increased risk of the development of ALS. The most common mutations occur in Cu/Zn superoxide dismutase 1(SOD1), TAR DNA-binding protein 43 (TDP-43), fused in sarcoma (FUS), and Chromosome 9 open reading frame 72(C9orf72), accounting for the disease in up to 70% of patients with fALS. [221] Role of Iron and Ferroptosis in ALS: Brain iron metabolic deregulation and ferroptosis were associated with neurodegeneration in ALS in the studies in the last two decades. Increased levels or deposition of iron were found in the spinal cord of transgenic SOD1 G93A and SOD1 G37R mice [224][225][226] and in motor cortex and substantia nigra of ALS patients. [227,228] Increased levels of ferritin were detected in the serum and in CSF of ALS patients, [229][230][231][232][233][234] which were associated with a poor clinical outcome. [229,230,[232][233][234][235] An abnormal iron accumulation was observed in the tissues of SOD1 G93A mice already at a presymptomatic stage of the disease. [224,236] Increased levels of iron were correlated with an increase of markers of LPO throughout the disease progression, [236] suggesting that the progression of ALS symptoms was correlated with abbrant iron accumulation and increased LPO.
LPO and Ferroptosis in ALS: ROS production and lipid peroxidation have been shown to be a hallmark of pathology in ALS, where excessive levels of 4-HNE have been found in the serum and cerebrospinal fluid in sALS patients. [237] 4-HNE levels were positively correlated with the disease stage.
The Role of Inhibition of GSH/GPX4 Axis in ALS: Another parameter that associates ALS to ferroptosis is GPX4. In neuronal inducible knockout of GPX4 mouse model, the condi-tional ablation of GPX4 led to a rapid development and progression of a motor phenotype, i.e., paralysis, severe muscle atrophy and, consequently, death. [238] The dramatic motor neuron degeneration observed in the spinal cord of the mouse model was correlated with a significant decrease of GPX4 level and with an increase of 4-HNE, the markers of ferroptosis. [238] Chen et al. reported a compromised anti-ferroptosis defense in SOD1 G93A mice and ALS patient samples, which revealed a significant decreased GPX4 activity and GSH levels in the spinal cord of SOD1 G93A animal models and in samples from sALS patients, that were correlated with an increase in lipid peroxidation products. [239] It has been shown that the overexpression of GPX4 rescued motor neuron integrity and reduced the levels of lipid peroxidation in the spinal cord of the models. [239] Furthermore, a delay in the disease onset, a slower progression of the motor symptoms, and an extension of the overall survival were observed. [239] A most recent study showed that ferroptosis mediates selective motor neuron death in ALS. [240] Depletion of GPX4 occurred in post-mortem spinal cords of both sALS and fALS patients. GPX4 depletion was also an early and universal characteristic of spinal cords and brains of SOD1 G93A , TDP-43 and C9orf72 mouse models of ALS. GPX4 depletion resulted from an impaired Nrf2 pathway and GSH synthesis in mutant SOD1 mice. [240] GPX4 overexpression improves lifespan, motor function and delays disease onset in SOD1 G93A mice, which was attributed to attenuated lipid peroxidation and motor neuron preservation. [240] The Role of Inhibition of FSP1-CoQH 2 System Axis in ALS: The myeloperoxidase (MPO), a heme-containing enzyme, converts hydrogen peroxide (H 2 O 2 ) and Cl-to H 2 O to form hypochlorous acid (HOCl), a powerful oxidant of the reactive oxygen species (ROS) family. [241] MPO/HOCl pathway was activated by hSOD1 G93A mutation in SOD1 G93A ALS NSC-34 motor neuron models. Activation of the MPO/HOCl pathway occurred differently in motor neurons of the motor cortices, brain stems, and spinal cords in male and female SOD1 G93A transgenic mice. Inhibition of MPO improved the motor performance of SOD1 G93A transgenic mice. [242] The activation of MPO/HOCl pathways facilitated ferroptosis through inhibiting the expressions of GPX4 and NQO1 and thus leading to irreversible lipid peroxidation. [242] These results suggested that aggregation of mutant hSOD1 proteins led to activation of the MPO/HOCl pathway, triggering ferroptosis in ALS. [242] Overexpressed FSP1 could ameliorate ferroptosis in SOD1 G93A ALS NSC-34 motor neurons models, evidenced by suppressed the MDA levels. FSP1 overexpression increased cell viability in hSOD1 G93A cells. NQO1 expression was increased by FSP1 overexpression, which compensated for the inability of GPX4, by potential cooperating with NQO1 in hSOD1 G93A cells. [242] The Role of Inhibition of GCH1-BH 4 System in ALS: Compared with WT, an obvious increased lipid peroxidation, content of LIP, and decreased GSH/GSSG was observed in hSOD1 G93A cells. Meanwhile, upregulated ALOX15 and GDF15 (respectively a driver and a marker of ferroptosis), down-regulated GCH1 and GPX4, and unaltered FSP1 was observed in hSOD1 G93A cells. Mutant SOD1 promotes ferroptosis via generating TfR1-imported excess free iron, decreasing GSH, upregulating ALOX15, and inactivating GCH1 and GPX4 in ALS. [243] The study also showed that a highly conserved "cyclin-like" protein, speedy/RINGO cell Figure 6. The molecular mechanisms of ferroptosis during cerebral ischemia. Following ischemic stroke, the stimulation of the NMDA receptor increases Ca 2+ , which activates cPLA2 to provide substrates for lipid peroxidation. NMDAR overactivation boosts neuronal iron uptake and produces ferroptosis. The dyshomeostasis of glutamatergic neurotransmission elevates extracellular glutamate levels, which inhibits cystine uptake and limits the biosynthesis of GSH. The reduction of soluble tau protein after cerebral ischemia prevents the dissociation of immature APP from the ER, abolishing the trafficking of APP to the neuronal surface, where APP interacts with FPN, allowing iron export from neurons. The absence of this interaction prevents iron from exiting neurons, leading to a toxic intracellular accumulation of iron and, ultimately, neuronal ferroptotic damage after ischemic stroke.

Ischemic Stroke
The Role of Iron Metabolism Dysregulation in Ischemic Stroke: The dysregulation of iron metabolism and iron accumulation occurs in the brain after ischemic stroke [244] (Figure 6). The acidic environment in brain tissue following cerebral ischemia can inhibit the binding of iron to transferrin, leading to iron disassociation from transferrin. [245] The neurons easily take up this unbound iron, causing intracellular iron elevation. [246] In the middle cerebral artery occlusion (MCAO) ischemic stroke model, in-creased accumulation of iron in cells along the lesion border is detected at 72 h after stroke.

Spontaneous Intracerebral Hemorrhage
Spontaneous intracerebral hemorrhage (ICH) is an acute subtype of acute cerebral stroke and accounts for 80% of hemorrhagic stroke and 10-15% of all types of strokes. [289] ICH causes two types of injury to the brain. The first is the primary brain injury, which is caused by the hematoma compressing the surrounding brain tissues. The second injury is the secondary brain injury after ICH (SBI-ICH), which is caused by blood components, such as hemoglobin (Hb), iron, and other neurotoxic substances released by the hematoma, all of which contribute to oxidative stress and neuroinflammation. [290,291] Both the primary injury and SBI-ICH lead to significant loss of neurological functions. Previous studies have shown that ferroptosis occurs in the ICH model in mice, and contributes to ICH-induced neuronal death. [292] A growing experimental evidence implicates that ferroptosis is involved in the pathogenesis of ICH [290,291] (Figure 7).
The Role of Iron Metabolism Dysregulation in ICH: Iron accumulation occurs and causes brain damage after ICH. [293] DMT1 and Fpn increase in ICH rats, both of which are positively corre-lated with Fe 2+ . [294] Certain lines of evidence suggest that dysregulation of proteins involving in iron influx and efflux, and mobilization of iron stored in ferritin (i.e., ferritin, TF, and TfR levels and HO-1), were significantly increased in the brain after ICH, leading to increased intracellular redox-active iron. [295] Following ICH, lysed red blood cells released Hb, which can be engulfed by activated microglia and macrophages in the perihematomal zone and release ferrous/ferric iron, thereby inducing the formation of lethal ROS and LPO. [296][297][298][299] Subsequently, excessive ferrous iron accumulates in neurons is transported out of microglia via the transferrin (Tf)-Tf receptor system and reacts with H 2 O 2 to form hydroxyl radicals (•OH) with a highly toxic effect via the Fenton reaction. [300] Heme oxygenase (HO) is the initial and ratelimiting enzyme of heme catabolism that catalyzes the oxidation of heme to Fe 2+ , and the expression of Heme oxygenase-1 (HO-1) is rapidly induced following ICH. [301,302] HO-1 knockout mice show smaller infarct volumes after ICH, [303] indicating a harmful role played by HO-1 in ICH. It was revealed that overexpression of IRP2 mRNA after ICH, [292] and the ablation of IRP2 in neurons can possess neuroprotection against the toxicity of Hb after ICH. [304] The attenuation of hemoglobin-induced toxicity was also observed in cortical cell cultures from IRP2 knockout mice. [305] The molecular mechanism underlying the neuroprotection in IRP2 knockout mice is likely related to sequester the excess hemorrhage-induced iron by increasing the expression of ferritin. Recent studies using in vitro and in vivo models have revealed that iron elevation and deposition were increased after ICH. [306][307][308][309][310][311][312] In hemin-treated N2A and SK-N-SH neuronal cells, hemin induced ferroptosis accompanied by an increment of global level of trimethylation in histone 3 lysine 9 (H3K9me3) and its methyltransferase Suv39h1. H3K9me3 was enriched at the promoter region and transferrin receptor gene 1 (Tfr1) and repressed its expression upon hemin stimulation. Inhibition of H3K9me3 or knockdown of Suv39h1 aggravated ferroptosis by upregulating Tfr1 expression. Furthermore, Suv39h1-H3K9me3 mediated repression of Tfr1 contributes to the progression of ICH in mice. These data suggest that H3K9 trimethylation dictates neuronal ferroptosis through repressing Tfr1 post ICH. [313] The Role of LPO in ICH: ICH-associated ROS can cause cell damage through ferroptosis-like LPO. [314] The ICH promotes heme incorporation into the plasma membrane, thereby increasing the sensitivity to the exogenous H 2 O 2 to facilitate LPO. [300,315] It was shown that heme significantly increased the contents of MDA and 4-HNE and significantly depleted glutathione in primary hippocampal neurons. [316] The content of 4-HNE and MDA were increased in brain tissue around hematoma 7 days after ICH in ICH model rats. [316] The content of MDA was significantly increased after ICH in mice. [311] This observation was corroborated by other studies. [306][307][308][309]312,317,318] Previous studies have demonstrated that specific regulators of ferroptosis, including CDGSH iron sulfur domain 2 (CISD2), [319] NOX4, [320] SRY-box transcription factor 10 (SOX10), [321] and forkhead box O3 (FOXO3) [322] play crucial roles in regulating ferroptosis in secondary brain injury after ICH. Increased CISD2 alleviates brain injury by inhibiting lipid peroxidation and ferroptosis via AKT/mTOR in mice. [319] The transcription factor SOX10 inhibits ferroptosis of hippocampal neurons after ICH through increasing miR-29a-3p expression, resulting in suppression of ACSL4 transcription. [321] Silencing FOXO3 ameliorates Figure 7. Mechanisms governing ferroptosis in ICH. Following a hemorrhagic stroke, iron derived from hemoglobin and/or heme enters neurons and produces massive lipid peroxidation. The increased permeability of the blood-brain barrier (BBB) causes a variety of components rich in Fe 3+ in the bloodstream to infiltrate into the brain parenchyma. Fe 3+ -TF then binds to TfR1 on the surface of brain cell membranes, enters cells through endocytosis. Fe 3+ is reduced to Fe 2+ , and transported to the cytoplasm through DMT1. Fe 2+ then initiates the Fenton reaction to form reactive oxygen species (ROS) and affects the catalytic activity of lipoxygenase (LOX). Elevated extracellular glutamate leads to an inhibition of system XC − , depletion of GSH, and diminution in GPX4 activity. Meanwhile, glutamate binds to and activates NMDAR, which further exacerbates iron uptake. Regarding lipid metabolism, ACSL4 is upregulated and facilitated the production of PUFA-CoA and PL-PUFAs which is then catalyzed by LOX to PL-PUFA-OOH. These components together lead to ferroptosis, account for the secondary brain injury after ICH.
post-ICH brain damage through inhibiting neuronal ferroptosis via down-regulating NOX4 transcription levels. [322] The role of Inhibition of GSH/GPX4 Axis in ICH: The depletion of GSH and the reduction of GPX4 induced ferroptosis after ICH. GSH was decreased following ICH [307,308,316,323] and caused brain edema and neural injury, which were reversed by GSH supplement. [324] The downregulation of glutathione reductase may account for the depletion of GSH after ICH. [309] previous investigations have shown that GPX4 levels were markedly reduced in neurons after ICH, [263,307,[310][311][312][316][317][318]325,326] the inhibition of GPX4 could exacerbate SBI-ICH, while the overexpression of GPX4 ameliorates ferroptosis-mediated hemorrhagic brain injury in rats. [327] The expression of SLC7A11 was markedly reduced in neurons after ICH. [311,318,325] The FOXO3 has been reported to play crucial roles in regulating ferroptosis in secondary brain injury after ICH through repressing GPX4. [322] Excitatory amino acid transporter 3 (EAAT3) that mediates mature neurons take up astrocyte-derived cysteine and plays an important role in neuronal resistance to ferroptosis through maintaining the activity of the GSH-GPX4 antioxidant pathway and scavenging lipid peroxides. [328] NOX4-mediated peroxidation and Tf/TfR-mediated iron overload exacerbate neuronal ferroptosis in a rat model of ICH. [320] In addition, astrocyte-derived glutamine synthesis was increased in ICH rats, while neuronal cysteine Figure 8. Mechanisms governing ferroptosis in SAH. Following SAH, the iron homeostasis is disrupted, intracellular LPO elevated and expression of GPX4 decreased, which induces ferroptosis, thereby leading to early brain injury after SAH. The iron metabolism is dysregulated, increased LPO and the inhibition of GSH/GPX4 were observed in SAH. ACSL4 was upregulated and FSP1 was downregulated in brain tissues after SAH. Elevated iron causes brain damage after SAH, which maybe be related to dysregulation of hepcidin. uptake was diminished, evidenced by downregulation of EAAT3 and GPX4 expression. Inhibiting NOX4 and iron chelation partially restored neuronal levels of EAAT3 and GPX4 expression and inhibited neuronal ferroptosis, suggesting that inhibition of these pathological signals can protect the hemorrhagic brain. [320] Increased methyltransferase-like 3, a N 6 -methyladenosine (m 6 A) methyltransferases ("writers") promoted secondary brain injury after ICH by inhibiting GPX4 expression in an m 6 A-dependent manner. [329] These results implied that m 6 A-dependent epigenetic modification of GPX4 is involved in the genesis of secondary brain injury after ICH.

Subarachnoid Hemorrhage
Subarachnoid hemorrhage (SAH), an acute hemorrhagic stroke commonly caused by ruptured intracranial aneurysms. SAH is characterized by the complexity of pathophysiological responses following extravasation of blood from cerebral circulation, and carries high mortality and morbidity. [330] Early brain injury after SAH (EBI-SAH) plays an important role in the poor prognosis of SAH. [331] Although the exact molecular mechanisms of EBI-SAH are being elucidated, it is widely accepted that increased oxidative stress, blood-brain barrier disruption, neuroinflammation, and cell death within the brain contribute to brain injury. [331,332] A previous study has shown that ferroptosis occurs in the EBI-SAH, and contributes to neuronal death after SAH, [333] accumulating evidence show ferroptosis is involved in the pathogenesis of EBI-SAH (Figure 8). Using both in vivo and in vitro models, it was shown that iron metabolism dysregulation, increased LPO and the inhibition of GSH/GPX4 were observed in in vivo and in vitro models with SAH. [333] The Role of Iron Metabolism Dysregulation in SAH: Elevated iron causes brain damage after SAH, [334,335] which is related to dysregulation of hepcidin. Pharmacological upregulation of hepcidin increased the expression of DMT1, decreased the expression of FPN1, and aggravated ferroptosis and EBI-SAH, indicating that hepcidin played a role in regulating iron metabolism and contribute to ferroptosis via activation of DMT1 signaling in rats with SAH. [336] A recent study showed that ferritinophagy is involved in the pathogenesis of EBI-SAH through inducing ferroptosis. [337] As a type of autophagy, ferritinophagy mediated by nuclear receptor activator 4 (NCOA4) plays a role in inducing ferroptosis by regulating iron homeostasis and producing ROS in cells. [338][339][340] NCOA4 acts as a selective autophagy receptor and binds to FTH1 of ferritin to mediate the transport of intracellular ferritin to autophagy lysosomes and finally releases free iron, which increases the content of available iron in cells. [341] SAH can disrupt iron homeostasis, elevated intracellular LPO, and decreased expression of GPX4 and FTH1. The autophagy inhibition by ATG5 gene knockout can reduce the intracellular iron level and LPO, increase the expression of GPX4, thereby alleviating SAH-induced cell death, suggesting that SAH triggers neuronal ferritinophagy-dependent ferroptosis and regulating iron homeostasis through ferritinophagy, which provides clues for the prevention of EBI-SAH. [337] Transferrin infiltration was increased in the brain parenchyma 24 h after SAH, which was positively correlated with neuronal ferroptosis. Overexpression of aquaporin 4 in the mouse brain can effectively improve post-SAH neuronal ferroptosis and brain injury through inhibiting transferrin infiltration into the brain parenchyma. [342] The Role of LPO in SAH: ACSL4 level in brain tissue increased significantly after SAH. Inhibiting ACSL4 alleviated EBI-SAH, evidenced by decreased inflammation, blood-brain barrier (BBB) impairment, oxidative stress, brain edema, and behavioral and cognitive deficits, and increased the number of surviving neurons, indicating that ACSL4 exacerbated SAH-induced EBI by mediating ferroptosis. [343] The Role of Inhibition of GSH/GPX4 Axis in SAH: The depletion of GSH [335,344,345] and the reduction of GPX4 [335,[345][346][347][348] induced ferroptosis after SAH, leading to EBI-SAH. The overexpression of GPX4 ameliorates ferroptosis-mediated EBI-SAH, evidenced by reduced LPO and cell death in both in vitro and in vivo experimental SAH models, and decreased brain edema and neurological deficits after SAH. [349] However, the role of ferroptosis in EBI-SAH remains an open conundrum for future investigation.

Pharmacological Inhibition of Ferroptosis to Treat NDs and Strokes
It was shown that Fer-1 prevents glutamate induced ferroptosis cell death in postnatal rat brain, suggesting that the inhibition of ferroptosis could be exploited to preserve neuronal cells and protect organisms from specific oxidative, iron-dependent neurodegeneration, such as AD, PD, and strokes. [1] The existing evidence suggests a strong correlation between ferroptosis and neurodegenerative diseases and stroke through a shared mechanism involving dysregulation of iron metabolism, induction of LPO, and inhibition of GSH/GPX4 axis. Thus, inhibition of ferroptosis could be a promising target for treatment of neurodegenerative diseases and stroke. Since the discovery of the involvement of ferroptosis in pathogenesis of neurological diseases, scientists have claimed an approach of targeting anti-ferroptosis to treat these diseases. A growing number of drugs have been uncovered to exert their therapeutic efficacy by inhibiting ferroptosis. During the past decade, many experimental compounds and clinical drugs have been shown to inhibit ferroptosis by preclinical and clinical studies to achieve therapeutic purposes. Mounting evidence indicates that pharmacological inhibition of ferroptosis exert neu-roprotection in in vitro or in vivo disease models in PD (Table 1), AD ( Table 2), ALS (Table 3), IS (Table 4), and ICH (Table 5), and SAH ( Table 6).

Conclusions and Perspectives
In conclusion, this review article summarizes the recent progress of understanding the pathological pathways and regulatory mechanisms of ferroptosis in neurological diseases including PD, AD, ALS, IS, ICH, and SAH. We discuss the application of ferroptosis inhibitors in mitigating neurodegenerative diseases and provide a new target for future treatment and prevention of these diseases through targeting ferroptosis. Ferroptosis is involved in the neurodegeneration of neurological diseases including PD, AD, ALS, IS, ICH, and SAH. Ferroptosis is involved in the pathogenesis of neurodegeneration, while stroke leads to ferroptosis, resulting in secondary brain injury. This review highlights the promising potential of pharmacological inhibition of ferroptosis for the treatment of neurological diseases. However, the current research on the role of ferroptosis in neurological diseases is still poorly understood. Despite many advances, relatively little is known about how ferroptosis orchestrates diverse cellular events. Firstly, the regulatory mechanism underlying ferroptosis in neurological diseases needs to be uncovered. Secondly, we should note that ferroptosis plays important roles in multiple diseases besides neurological diseases, and inhibition of ferroptosis in neurological diseases as a therapeutic might result in drug resistance in cancer. Therefore targeting ferroptosis maybe a double-edged sword, and lead to unexpected toxicity and injury. Third, the specific biomarker for ferroptosis is urgent needed for accurate predicting the efficiency of inhibition of ferroptosis. Furthermore, most data obtained from experimental studies are far from reaching impact on clinical application and more clinical approaches are necessary. Nevertheless, targeting ferroptosis will bring new directions for the treatment of neurological diseases.