Iron is the most abundant transition metal element in the Earth’s crust and plays a critical role in oxygen transport, cellular energy metabolism, and many enzymatic reactions [29, 30]. Under normal conditions, iron is absorbed from food in the duodenum and mucosa of the upper jejunum [31]. The ferrous ions released from the ferric transporter on the outer surface of the capillary endothelial cell are oxidized by ceruloplasmin into ferric ions, which are then captured by transferrin (Tf) in the intercellular fluid and cerebrospinal fluid. Tf binds the ferric ions and forms a Tf-TFR1 complex with transferrin receptor 1 (TFR1), which is then endocytosed via clathrin-coated pits at the cell surface [32, 33]. In the acidic endosomes, ferric ions are reduced into ferrous ions and transported by divalent metal transporter 1 (DMT1) to labile iron pools in the cytoplasm. Here, the ferrous ions are stored in ferritin or imported into mitochondria via mitotic ferritin. Non-transferrin-bound iron at the cell surface is directly introduced into the labile iron pool by DMT1 [34, 35, 36]. The accumulation of intracellular ferrous ions promotes lipid peroxidation, leading to the production of ROS and ultimately to ferroptosis (Fig. 1) [37].

Fig. 1.

Mechanism of ferroptosis. (a) Tf binds ferric ions and forms a Tf-TFR1 complex with TFR1. The following endocytosis into acidic endosomes, ferric ions are reduced into ferrous ions and transported by DMT1 to labile iron pools in the cytoplasm. The accumulation of intracellular ferrous ions promotes lipid peroxidation, leading to the production of ROS and ultimately to ferroptosis. (b) System xc-exchanges intracellular glutamate for extracellular cystine in a 1:1 ratio, which is then synthesized into GSH by GCL and GSS. GSH is a reducing cofactor for glutathione peroxidase (GPX4), a membrane lipid repair enzyme that eliminates lipid hydroperoxides to protect against ferroptosis. (c) Dihydroorotate Dehydrogenase (DHODH) in the mitochondrial membrane resists ferroptosis by oxidizing dihydroorotate (DHO) to orotic acid (OA) and reducing ubiquinone (CoQ10) to Dihydroubiquinone (CoQH2). (d) Ferroptosis suppressor protein 1 (FSP1) converts ubiquinone into ubiquinol and generates a lipophilic, radical-trapping antioxidant (RTA) that prevents lipid peroxide formation and inhibits ferroptosis. (e) Polyunsaturated fatty acids (PUFA) are substrates for lipid peroxidation and react with acyl-CoA synthetase long-chain family member-4 (ACSL4) to generate Malondialdehyde (MDA) and 4-hydroxynonenal (4-HNE), eventually leading to ferroptosis.

Lipid peroxidation is the hallmark of ferroptosis. In the initial stage, lipid peroxidation can occur through both enzymatic and non-enzymatic reaction processes [38, 39]. However, which of these two is essential for ferroptosis remains controversial. Several studies have suggested that non-enzymatic lipid peroxidation is responsible for ferroptosis [40]. This is driven by the Fenton reaction, which uses iron and oxygen to catalyze a chain reaction that leads to the production of phospholipid hydroperoxides (PLOOHs) [40]. Polyunsaturated fatty acid (PUFA) phospholipids are substrates for lipid peroxidation. Enzymes such as acyl-CoA synthetase long-chain family member-4 (ACSL4) inhibit ferroptosis by reducing PUFA phospholipid synthesis. ACSL4 is therefore a marker of sensitivity to ferroptosis [41]. Mechanistically, ACSL4 enriches cellular membranes with long polyunsaturated $\omega$6 fatty acids. The peroxidation of polyunsaturated fatty acids (PUFAs) and unlimited lipid peroxidation generates products such as malondialdehyde (MDA) and the formation of reactive aldehydes, especially 4-hydroxynonenal (4-HNE). Eventually, this leads to the loss of cell membrane integrity and to ferroptosis (Fig. 1) [42, 43].

System xc${}^{-}$ is an important intracellular antioxidant system consisting of two subunits: SLC7A11 and SLC3A2 [44]. SLC7A11 is mainly responsible for transport activity and is highly specific for cystine and glutamate, while SLC3A2 acts as a chaperone protein [45]. System xc${}^{-}$ exchanges intracellular glutamate for extracellular cystine (Cys2) in a 1:1 ratio. The cystine is then synthesized into glutathione (GSH) by glutamate-cysteine ligase (GCL) and glutathione synthetase (GSS) [46, 47, 48]. GSH is a reducing cofactor for glutathione peroxidase (GPX4), which is a membrane lipid repair enzyme [16]. Inhibition of system xc${}^{-}$ leads to the depletion of intracellular cystine which is the rate-limiting substrate for the synthesis of the antioxidant glutathione (GSH) [49]. Due to the antioxidant properties of GSH, its depletion causes an imbalance between the cellular oxidation and antioxidant systems, as well as the inhibition of glutathione-dependent enzymes such as GPX4. GPX4 is a member of the glutathione peroxidase family [40]. These can reduce hydroperoxides produced from phospholipids and cholesterol to their corresponding alcohols by using two electrons from the GPX4 catalytic cysteine residues [16]. Genetic studies have established that GPX4 is a key regulator of ferroptosis [50]. In the normal state, PUFAs are frequently oxidized by lipoxygenases such as ACSL4 but are then immediately reduced by GPX4 and its GSH cofactor [51]. Hence, the loss of GPX4 leads to peroxidation chain reactions that cause lipid peroxidation-dependent cell death by blocking hydroperoxide conversion into non-toxic lipid alcohols [11, 52, 53]. In conclusion, disruption of the System xc${}^{-}$-GSH-GPX4 axis results in the inability to eliminate PLOOH, the accumulation of lipid peroxides, and finally to ferroptosis (Fig. 1).

Previous studies on ferroptosis defense mechanisms have focused on GPX4 [54]. Recently, two articles published in 2019 reported on a novel ferroptosis suppressor, FSP1, that was independent of GSH [55, 56]. FSP1 was originally named AIFM2 and has homology to apoptosis-inducing factor (AIF) [55]. However, it lacks the N-terminal mitochondrial targeting sequence present in AIF and does not promote apoptosis. AIFM2 was therefore renamed FSP1 [55]. It is located in lipid droplets (LDs) and the plasma membrane [57]. It can convert ubiquinone on cell membranes into its reduced form, panthenol, which inhibits peroxidation and prevents ferroptosis [55]. By employing a synthetic lethal CRISPR/Cas9 screen, Doll et al. [55] found that FSP1 was an oxidoreductase that can reduce coenzyme Q10 (CoQ) to generate a lipophilic, radical trapping antioxidant (RTA) that prevents lipid peroxide formation (Fig. 1). Venkatesh et al. [58] reported that inhibition of MDM2 and MDMX resulted in increased levels of FSP1 protein, which in turn led to increased levels of the endogenous lipophilic antioxidant CoQ, thereby inhibiting ferroptosis. Song et al. [59] showed that miR-4443 regulates the expression of FSP1 through METLL3 in an m6A manner, thus regulating ferroptosis.

DHODH is a flavin-dependent enzyme located in the mitochondrial membrane [60]. By selectively inhibiting dihydroorotate dehydrogenase, thereby inhibiting pyrimidine synthesis [61]. This novel, GSH-independent ferroptosis defense system was first reported by Boyi’s team in 2021 [61]. They demonstrated that DHODH inhibits ferroptosis by oxidizing dihydroorotate (DHO) to orotic acid (OA) and reducing ubiquinone (CoQ10) to Dihydroubiquinone (CoQH2) in the mitochondrial membrane (Fig. 1). To our knowledge, there have been no further advances regarding the DHODH and ferroptosis mechanism. The molecular mechanisms that link ferroptosis with various diseases also warrant further exploration.

Depression is a major global health issue that affects about 300 million people worldwide [62]. It is characterized by persistent mood depression and restlessness and includes psychomotor disorders and cognitive impairments [63]. It also has the characteristics of being universal, chronic, and recurrent [64]. Unfortunately, completely effective treatment for depression has yet to be discovered. Currently, treatment for depression mainly involves drug therapy and psychotherapy. However, their efficacy is quite limited and drug therapy requires long-term medication, while psychotherapy is slow [64, 65, 66]. Therefore, it is very important to explore the pathophysiological mechanism of depression. Studies have shown that hippocampal volume is significantly reduced in both adults and adolescents with depression [67]. Patients who have recovered from depression also show a decreased hippocampal volume [67]. The destruction and atrophy of neurons during depression may be the cause of this reduced volume [68, 69]. Previous studies on depression and the death of hippocampal neurons have focused mainly on apoptosis [70]. For example, Zhao et al. [71] found that repetitive transcranial magnetic stimulation (RTMS) reduces depression-like behavior by inhibiting the apoptosis of hippocampal neurons. Dexmedetomidine can also reduce the apoptosis of hippocampal neurons in Sprague Dawley rats and decrease depressive behavior by inhibiting the p38 MAPK/c-Myc /CLIC4 signaling pathway [72]. As described above, ferroptosis is a newly discovered form of programmed cell death that is caused mainly by the accumulation of iron and lipid peroxides [73]. Ferroptosis offers a new treatment strategy for depression. Cao et al. [74] confirmed that mice with chronic unpredictable mild stress (CUMS) could be used as an animal model for depression. Jiao et al. [75] found that GPX4 expression was decreased in this model whereas that of FTH1, ACSL4, and COX2 was increased, indicating that ferroptosis was activated. Wang et al. [76] found that sodium hydride (NaHS) can reduce iron deposition and oxidative stress and increase the protein level of GPX4 and SLC7A11 in the prefrontal cortex of type 1 diabetic mice, thereby reducing ferroptosis and alleviating anxiety-depressive behavior (Table 2, Ref. [76, 77, 78, 79, 80, 81, 82, 83, 84, 85]). Using the chronic social defeat stress (CSDS) mouse model, Dang et al. [77] found that edravone (EDA) could upregulate GPX4 expression to alleviate depressive behavior, but that EX527 and ML385 can reduce Nrf2, HO-1, and Gpx4 protein levels and thereby attenuate the antidepressant effects of EDA (Table 2). In conclusion, these results confirm that ferroptosis occurs in mouse models of depression. Although the underlying molecular mechanism remains unknown, the upregulation of GPX4 can inhibit ferroptosis and thereby reduce depressive behavior in mice.

In conclusion, ferroptosis is a potential target for the treatment of depression, but there is still no clinical evidence to support the use of ferroptosis inhibitors in humans.

Neurodegenerative diseases are caused by the accumulation of oligomers and the aggregation of misfolded proteins [86, 87]. Due to the aging of society, neurodegenerative diseases have a considerable socio-economic impact. They include Alzheimer’s disease (AD), Parkinson’s disease (PD), and Huntington’s disease (HD) [88, 89], and are characterized by a loss of function of specific neuronal groups and of their connections [88]. Emerging evidence has shown that ferroptosis can disrupt neuronal viability and is involved in the pathological mechanism of neurodegenerative diseases [78, 90].

Stroke is an acute disorder of blood circulation in the brain tissue caused by arteriovenous stenosis, occlusion, or rupture due to various inducing factors [119, 120]. It is also one of the most important causes of disability [121, 122]. The clinical symptoms of a stroke are language disorder and cognitive impairment [123]. Stroke can be divided into ischemic stroke and cerebral hemorrhage (ICH) stroke [124]. It has been reported that ferroptosis plays an important role in the pathophysiological processes of various cancer types [125, 126]. In recent years there has also been substantial evidence that ferroptosis is associated with the progression of stroke [127].

TBI is a series of pathological changes to normal brain function caused by external physical forces [134]. It often leads to death and disability and can be divided into primary and secondary impairments [135]. Several types of programmed cell death occur during TBI pathophysiology, including necrosis, apoptosis, necroptosis, pyroptosis, and ferroptosis [136, 137]. Research has shown that TBI is followed by several characteristics of ferroptosis during the secondary impairment stage, including iron accumulation, dysfunction of iron metabolism, and up-regulation of ferroptosis-related genes [135, 82]. GPX4 activity decreases significantly at 6 h, 1 d, and 3 d after TBI, while the MDA concentration and lipid ROS levels increase significantly at 6 h, peak at 3 d, and return to baseline levels at 7 d [134, 138]. Mitochondrial shrinking was observed by transmission electron microscopy three days after TBI. Ferrostatin-1 (Fer-1) injection through the ventricle significantly reduced iron deposition and neurodegeneration, while also alleviating injury damage and improving long-term motor and cognitive performance [82] (Table 2). This strongly suggests the involvement of ferroptosis after TBI, but the precise mechanism underlying ferroptosis in TBI remains unknown [139]. Reduction of iron accumulation and inhibition of lipid peroxidation may therefore be potential treatments for TBI.

Sepsis-associated encephalopathy (SAE) can be defined as a diffuse brain dysfunction caused by an inflammatory response but without an accompanying infection of the central nervous system (CNS) [140]. Research suggests that ferroptosis is induced during sepsis and is involved in the pathogenesis of SAE [141]. In support of this, the iron concentration and GPX4 expression level are increased in the hippocampal neurons of SAE mice. Furthermore, iron chelating agents can alleviate SAE-associated cognitive dysfunction [83] (Table 2). Wei et al. [84] found that inhibition of miR-9-5p led to increased expression of transferrin receptor (TFRC) and exacerbation of SAE (Table 2). More recent research has also shown that inhibiting ferroptosis after SAE can reduce glutamatergic excitability [142]. Hence, ferroptosis may offer a new strategy for the treatment of SAE. In summary, an imbalance of the antioxidant system results in a large accumulation of lipid peroxidation. By reducing the iron concentration and the expression level of GPX4, ferroptosis can be inhibited in SAE mice, thereby improving cognitive dysfunction.