*3.1. Ferroptosis in Alzheimer's Disease*

Alzheimer's disease (AD) is considered a neurodegenerative disease associated with multiple brain complications. It was initially described by the German Alois Alzheimer in 1907 [134,135]. AD is characterized by progressive disorder in the cortical and hippocampal neuronal areas which leads to both loss of neuronal function and cell death, and is the most common type of dementia (Figure 4). The hallmark of AD is the histopathological presence of an extracellular β-amyloid (Aβ) deposition in senile plaques (SPs) and intracellular neurofibrillary tangles (NFTs) formed from the hyperphosphorylation of the tau protein. Neurocognitive decline is associated with synapse decrease and neurotransmitter oxidation. These changes are due to the increase in oxidative stress, mainly an increase in ROS and intra- and extracellular hydrogen peroxides [136–138]. Moreover, genetic changes include alterations in amyloid precursor protein (APP), apolipoprotein E (APOE), presenilin 1 (PSEN1) and presenilin 2 (PSEN 2) genes [139].

**Figure 4.** Alzheimer's disease. The development and progression of Alzheimer's disease (AD) lead to atrophy, loss and dysfunction of both neurons and glial cells. AD begins in the dorsal raphe nucleus with subsequent progression to the cortex, which is the center of information processing and memory storage. The factors that promote the development of AD are still unknown. However, it seems that the intracellular accumulation in neurons of the phosphorylated Tau protein (neurofibrillary tangle) and the formation of amyloid-B plaque (senile plaque) in the extracellular environment and brain tissue both lead to neuron loss and dysfunction. In addition, the formation of neurofibrillary tangle and senile plaque alters the functions of glial cells, such as oligodendrocytes (responsible for the myelination of neurons), microglia cells (phagocytic cells) and astrocytes (responsible for the absorption and exchange of nutrients between neurons and blood vessels). Dysregulation of cholesterol transport and iron metabolism in the central nervous system contributes to poor prognosis of Alzheimer's disease. All these associated factors lead to an increase in neuroinflammation and oxidative stress associated with mitochondrial dysfunction, compromising the production of ATP, altering the concentration of neurotransmitters in the synaptic cleft, finally promoting cell death.

Evidence of the association between the accumulation of iron in the cerebral cortex and the development of Alzheimer's disease emerged in the early 1960s [140,141]. Since then, several studies have demonstrated the direct association between free iron, oxidative stress, lipid peroxidation and cell death of neurons, usually associated with apoptosis and/or necrosis, due to increased neuroinflammation. The iron dyshomeostasis is associated with ROS production and neurodegeneration in AD [138]. Furthermore, aging and changes in iron metabolism are associated with the development of Aβ plaques and NFTs. Svobodová et al. [142] demonstrated in an APP/PS1 transgenic mice model that free iron and ferritin accumulation follows amyloid plaque formation in the cerebral cortex area. In fact, iron deposition has been involved in the misfolding process of the Aβ plaques and NFTs [143].

Iron is related to the development of Tau protein and, consequently, NFTs. In fact, iron is present through the induction and regulation of tau phosphorylation [143,144]. The association of NFT with neurodegenerative dysfunctions is termed tauopathy [145–147]. The oxidation process slows down or excludes the regular action of the Aβ and tau protein [148]. In animal models of tauopathies, increased iron associated with aging and neurodegeneration has been observed [149]. Indeed, animals with tauopathies treated with the iron chelator deferiprone showed a trend toward improved cognitive function associated with the decrease in brain iron levels and sarkosyl-insoluble tau [150].

APP is a type 1 transmembrane protein and its function in heathy individuals appears to be associated with the development of synaptic activity [145]. Proteolytic cleavage of the β-amyloid precursor protein (APP) to form the β-amyloid peptide (Aβ) is related to the pathogenesis of AD because APP mutations that influence this process induce familial AD or decrease the risk of AD [145]. The amyloid cascade hypothesis states that the agglomeration and production of Aβ plaques in the brain would occur, resulting in cell death. Presenilins 1 (PSEN1) and presenilins 2 (PSEN2) precisely cleave the APP and other proteins as they are part of the catalytic protease compounds [151]. Acetylcholinesterase participates in the aggregation of Aβ plaques [152]. Moreover, the Aβ plaques in the presence of free iron participate efficiently in the generation of ROS resulting in increased lipid peroxidation, protein oxidation and DNA damage [153]. Deferiprone derivatives act as acetylcholinesterase inhibitors and in iron chelation [154].

The proteolytic cleavage in APP occurs by enzymatic complexes involving α-secretase or β-secretase and γ-secretase. The proteolytic cleavage in APP by β-secretase produces a neurotoxic 40 to 42 amino acid amyloid [155]. Tsatsanis et al. [156] showed that APP promotes neuronal iron efflux by stabilizing the cell-surface presentation of ferroportin, and that β-cleveage of APP depletes surface ferroportin, leading to intracellular iron retention independently on the generation of Aβ. Furthermore, these findings indicate how β-secretase's processing of APP might indirectly promote ferroptosis. Iron overload alters the neuronal sAPPα distribution and directly inhibits β-secretase activity [157]. Cortical iron has been strongly associated with the rate of cognitive decline [158]. Iron in the brain increases lipid peroxidation, oxidative stress, and neuroinflammation due to the depletion of neuronal antioxidant systems—mainly the glutathione system [143]. In addition, increased hepcidin expression in APP/PS1 mice astrocytes improves cognitive decline and partially decreases the formation of Aβ plaques in the cortex and hippocampus. Indeed, decreased iron levels in neurons led to a reduction in oxidative stress (induced by iron accumulation), decrease in neuroinflammation and decreased neuronal cell death in the cortex and the hippocampus. [159]. As mentioned before, the hepcidin peptide binds ferroportin, which is followed by cell internalization and further degradation [160].

In order to investigate whether neurons of the cerebral cortex and hippocampus severely affected in patients with AD may be vulnerable to ferroptosis, Hambright et al. [161] have shown in GPx4BIKO mice (a mice model with a conditional deletion in neurons of the forebrain of GPx4) that tamoxifen led to the deletion of GPx4 mainly in neurons of the forebrain. GPx4BIKO mice exhibited significant deficits in spatial learning and memory function, as well as hippocampal neurodegeneration. These results were associated with ferroptosis markers, such as increased lipid peroxidation, ERK activation and neuroinflammation. In addition, GPx4BIKO mice fed a vitamin E-deficient diet had an accelerated rate of hippocampal neurodegeneration and behavioral dysfunction. On the other hand, treatment with Liproxstatin-1, a ferroptosis inhibitor, improved neurodegeneration in these mice. Moreover, in an in vitro model, iron increased nerve cell death in conditions where GSH levels were reduced, by decreasing the activity of glutamate cysteine ligase [162].

TheHT22 cellline has high concentrations of extracellular glutamate, whichinhibit the glutamate-cystine antiport, leading to the depletion of intracellular GSH and resulting in excessive ROS production. In a study with these cells, Hirata et al. [163] found that an oxindole compound, GIF-0726-r, prevented cell death induced by oxidative stress, including oxytosis induced by glutamate and ferroptosis induced by erastin. Moreover, an excess of extracellular glutamate associated with high levels of extracellular iron cause the overactivation of glutamate receptors. As a consequence, there was an increase in iron uptake in neurons and astrocytes, increasing the production of membrane peroxides. Glutamate-induced neuronal death can be mitigated by iron chelating compounds or free radical scavenging molecules. Ferroptosis is induced by reactive oxygen species in the excitotoxicity of glutamate [110,164,165]. In addition, the sterubin compound maintained GSH levels in HT22 cell lines treated with erastin and RSL3, suggesting protection against ferroptosis [166]. 7-*O*-esters of taxifolin 1 and 2 were described as having neuroprotective action against ferroptosis induced by RSL3 in HT22 cells [167].

Chalcones 14a–c were shown to inhibit β-amyloid aggregation, and in addition, protect neural cells against toxicity induced by Aβ aggregation and from erastin and RSL3-induced ferroptosis in human neuroblastoma SH-SY5Y cells [160]. The authors suggested that the inhibition of toxicity induced by Aβ plaques' aggregation and of ferroptosis occurs due to the presence of hydroxyl groups in the chalcone derivatives. Chalcone 14a-c can react with lipid peroxyl radicals by transferring the hydrogen (H) atom, thus inhibiting lipid peroxidation [168].

After treatment with high dietary iron (HDI), WT (wild type) mice and the APP/PS1 double Tg mouse model of ADon (HDI) showed upregulation of divalent metal transporter 1 (DMT1) and ferroportin expression, and downregulation of TFR1 expression, with fewer NeuN-positive neurons in both animal models. Moreover, the iron-induced neuron loss may involve increased ROS production and oxidative mitochondria dysfunction, decreased DNA repair, and exacerbated apoptosis and autophagy [169]. Using X-ray spectromicroscopy and electron microscopy it was found that the coaggregation of Aβ and ferritin resulted in the conversion of the ferritin inert ferric core into more reactive low oxidation states [170].

Ates et al. [171] showed in an animal model that inhibition of fatty acid synthase (FASN) by CMS121 decreased lipid peroxidation. CMS121 treatment reduced the levels of 15LOX2 in the hippocampus compared to those of untreated WT mice. Relative levels of endocannabinoids, fatty acids, and PUFAs were significantly higher in untreated AD mice as compared to CMS121-treated AD mice, suggesting that other enzymes may be involved in the process of ferroptosis in Alzheimer's disease.

It is important to highlight the heterogeneity of Alzheimer's disease and the involvement of multiple metabolic pathways which contribute to the poor prognosis of this disease (Figure 4). In fact, multiple patterns of cell death are involved in the neurodegeneration process, such as apoptosis, necrosis, and autophagy associated with disturbed BBB (brain blood barrier) permeability. In vitro experiments are the main evidence of ferroptosis in human neurodegenerative processes. The identification of ferroptosis in in vivo models of Alzheimer's disease is difficult since specific markers for ferroptotic cells, such as specific antibodies, are not available. In addition, other metal ions, such as copper, can also regulate ferroptosis and lipid peroxidation [172,173]. Taking all under consideration, we still do not know whether ferroptosis is the cause or consequence of neurodegeneration processes such as Alzheimer's disease.

#### *3.2. Ferroptosis in Parkinson's Disease*

Parkinson's disease (PD) is one of the most common and best-known diseases of the nervous system, affecting roughly 0.1–0.2% of the general population and 1% of the population above 60 years [174]. PD is characterized as a slowly progressing neurodegenerative ailment with motor and non-motor clinical manifestations, due to an intense decrease in dopamine production [175]. Classic hallmarks in PD are still related to the motor manifestation such as bradykinesia, resting tremor and rigidity [176]. However, non-motor symptoms associated with PD have recently gained more attention due to their relevance and impact on the patient's quality of life. Non-motor symptoms of PD include anosmia, constipation, pain, anxiety, depression, psychosis and cognitive disorders that can progress to dementia [177–179].

The pathophysiological characteristics of PD include the slow and progressive degeneration of dopaminergic neurons in the pars compacta of the substantia nigra (SNpc), which is associated with a systematic and progressive accumulation of iron, leading to striatum dopamine depletion, disappearance of neuromelanin and the appearance of intracellular Lewy bodies having aggregated α-synuclein as the main component [180,181]. During PD progression there is an increase in oxidative stress, lipid peroxidation, and mitochondrial dysfunction associated with the depletion of antioxidant enzymes in the glutathione systems. All of these associated factors lead to neuronal death and the functional disability of the organism (Figure 5). Currently, the pharmacological treatment of PD aims to increase dopamine levels in the synaptic cleft. Levodopa is the drug of choice, being associated with dopamine agonists, dopamine metabolism inhibitors and decarboxylase inhibitors. Treatment is stable for a period of 5–6 years. Then, however, the disease progresses with marked neurodegeneration and development of dementia [182–184].

**Figure 5.** Parkinson's disease. Parkinson's disease (PD) occurs due to the decrease in and/or oxidation of dopamine in the substantia nigra, involving the motor system. The incorrect folding of α-synuclein leads to the accumulation of protein (Lewy body) in nervous tissue. The formation of Lewy bodies may be due to a highly pro-oxidative environment, due to dysfunction in the transport of lipids, iron, inflammation and mitochondrial changes. The increase in Lewy bodies is the trigger for the development of dementia, neurotoxicity and neuronal death.

The association between iron and PD is long-standing [185,186]. Daily exposure to elevated iron levels is a risk factor for the development of PD [187]. In addition, an increase in the iron content in substantia nigra and globus pallidus of PD patients was observed by magnetic resonance imaging (MRI). This increase was associated with time of disease, neurodegeneration and severity of motor impairment [188,189]. In mice, treatment with deferiprone (DFP) was shown to significantly reduce labile iron and biological damage in oxidation-stressed cells, improving motor functions while raising striatal dopamine levels. Furthermore, in patients with PD, a decrease in iron overload has been described in the substantia nigra after 6 months of deferiprone treatment, as has an increase in ceruloplasmin activity in cerebrospinal fluid (CSF) [190–193].

Ferroptosis is also proving to be a mechanism of immeasurable importance for the pathogenesis of PD [194]. In fact, since the early 2000s it has been known that elevated levels of iron could be found in the brain of patients with PD, although an iron-dependent cell death mechanism had not yet been proposed at that time. Additionally, several genes and proteins related to iron metabolism of brain cells have been found to be mutated in PD patients, strengthening the correlation between iron metabolism and Parkinson's disease [195,196].

Some previous works on Parkinson's disease described the presence of PUFA peroxidation, a decrease in GPx4 activity and exhaustion of the glutathione system, associated with increased oxidative stress. The first evidence of ferroptosis in Parkinson's disease was described by Do Van et al. [197]. PD models, both in vitro and in vivo, have shown that the characteristic features of ferroptosis were present in differentiated Lund human mesencephalic (LUHMES) cells intoxicated with erastin. The characteristics of ferroptosis in LUHMES cells were different from those reported for other cell lines. Moreover, the calcium chelator 1.2-bis(o-aminophenoxy)ethane-N,N,N- ,N- -tetraacetic acid (BAPTA) and protein kinase C (PKC) inhibitors (the bisindolylmaleimide analog Bis-III, and small interfering RNA (siRNA)) were very effective in counteracting erastin-induced cell death. In LUHMES cells, ferroptosis requires activated mitogen-activated protein kinase (MEK) signaling but is independent of Ras activation. Moreover, ferroptosis involvement in dopaminergic cell death was observed in a

mouse model in which toxicity was inhibited by the specific ferroptosis inhibitor ferrostatin 1. Lastly, the regulation of dopaminergic cell death by ferroptosis and its inhibition by PKC were also confirmed ex vivo by studying organotypic slice cultures (OSCs). It is important to note that the ferroptosis activation pathways were initially described in cancer models, in which there was elevated metabolic activity and cellular proliferation due to uncontrolled cellular repair pathways. These pathways are not activated in neurodegeneration models. Therefore, ferroptosis can be triggered by different mechanisms in different cells and different tissues.

A plethora of new evidence is clarifying the molecular mechanisms involved in the interaction of PD and ferroptosis cell death. α-Synuclein, a protein abundantly expressed in the nervous system and a main component of Lewy bodies, has been widely studied in PD as its pathogenic effects are strongly correlated with PD's pathophysiology [198]. Additionally, it has been recently shown that α-synuclein aggregation (a common feature in PD) is responsible for the production of ROS followed by lipid peroxidation in an iron-dependent manner, resulting in increased calcium influx and consequent cell death [199]. In this way, the use of ferroptosis inhibitors such as ferrostatin or iron chelators [200] has been sufficient to suppress cell death, supporting the hypothesis that ferroptosis is a major player in this process and may harbor therapeutic potential. Several studies suggest the modulation in ferroptosis as a therapeutic target in neurodegenerative diseases [201,202].

Another molecule that may link PD to ferroptosis is the transcription factor Nrf2. As Nrf2 is involved in the regulation of processes such as the metabolism of iron, lipids and glutathione, several works have focused on understanding how the modulation of this transcription factor can intervene in the ferroptosis pathway [203]. For instance, it was shown that Nrf2 overexpression in brain tumor cells was an indication of poor survival outcomes since Nrf2 bestowed these tumor cells with resistance to cell death mechanisms such as ferroptosis [204]. In addition, the activation of the Nrf2 pathway (p62-Keap1-Nrf2) has also been shown to prevent 6-hydroxydopamine (6-OHDA)-induced ferroptosis in a human dopaminergic cell line (SH-SY5Y) [205]. Taken together, these results demonstrate that the modulation of Nrf2 expression can provide new therapeutic approaches for neurodegenerative diseases such as PD [206]. In addition, studies performed on a monkey model of PD has shown that clioquinol (CQ), a drug primarily used as an antiparasitic agent that also presents iron chelation properties [207], was able to improve both, motor and non-motor manifestations in treated monkeys. Shi et al. [208] observed that CQ not only led to a decrease in iron levels in the substantia nigra but also managed to suppress the well-known oxidative stress present in PD. At a molecular level, CQ was shown to be able to reduce p53-mediated cell apoptosis and to diminish the iron content and oxidative stress by activating the AKT/mTOR pathway, which was found downregulated in the PD monkey model [208]. These results, taken together, point out once again to the involvement of ferroptosis in PD and demonstrate how pharmacological interventions could be useful to revert this outcome. Unfortunately, available therapies for PD patients are only capable of mitigating their symptoms and cannot reverse the loss of dopaminergic cells [209]. Therefore, seeking new therapeutic options that may intervene in this primary process of cell death could potentially change PD treatment.

#### *3.3. Ferroptosis in Huntington's Disease*

Huntington's disease (HD) is an autosomal dominant late-onset neurodegenerative disorder (age of onset: 30–50 years). HD is caused by a polymorphic sequence of three CAG nucleotides in exon 1 of the IT15 gene (Huntingtin (HTT)), which is located at 4p16.3. HD was described by George Huntington in 1872, after he observed in Long Island a rare disease present in some families in the region. He called this disease "hereditary chorea" (Huntington, 1872). The main observed clinical signs are motor disorders (such as involuntary movements), cognitive, emotional, and psychiatric disorders (such as personality change and dementia). Carriers of this disease may also have dysphagia, which leads to weight loss. In patients affected with juvenile HD under the age of 20, the most observed disorders are behavioral disorders, learning difficulties, and often seizures [210–214].

The mutation occurs in exon 1 of the HTT gene; this region is polymorphic and encodes a polyglutamine segment (polyQ)—in this fragment, expansion and generation of mutant proteins that can lead to the development of HD may occur [215]. The huntingtin protein has a molecular weight of ~348 kDa, and the expression level is different according to the cell types in which it is found: neuronal cell bodies, dendritic cells, and axons. Inside the cell, the huntingtin protein is located in the cytoplasm and partially in the nucleus and can move between these compartments [216–218].

Such evidence of expression and location suggests that this protein plays an important role in the nervous system, suggesting that changes in its conformation can lead to an imbalance in the performance of its functions, which in turn can result in the development of HD. The pathophysiological mechanism associated with HD includes loss of glial cells (astrocytes and oligodendrocytes), neuronal death, and atrophy of brain tissues, which may start in the striatum, followed by the cerebral cortex (Figure 6). Although mutated huntingtin protein (mHTT) is found in the brain, its expression has been evidenced in cells other than the ones in the central nervous system [216].

**Figure 6.** Huntington's disease. Huntington's disease is caused by the repetition of autosomal dominant CAG trinucleotide in the Huntingtin gene (HTT gene) on chromosome 4, giving rise to the mutant huntingtin protein. The mutated protein translocates to the nucleus and remains in the cytoplasm. In the nucleus, association, oligomerization and aggregation with other proteins occurs, leading to the formation of inclusions. Protein inclusions disrupt the transcriptional process in nerve tissue cells. In the cytoplasm, the oligomerization, aggregation and precipitation of the huntingtin protein occurs. This process alters the metabolism and both intra- and extra-cellular signaling pathways. The increase in oxidative stress, lipid peroxidation and iron dyshomeostasis contribute to the aggregation and oligomerization of huntingtin protein with other cytoplasmic proteins. Aberrant protein aggregation increases the excitotoxicity of glutamate. Mitochondrial dysfunction changes autophagy mechanisms, and transport in the neuronal axon, leading to nerve cell degeneration.

One of the cellular processes involved in the development of HD is ferroptosis. A study in mouse models with HD showed an accumulation of toxic iron in neurons compared to the wild model, which suggests that iron accumulation may contribute to the neurodegenerative process [219]. Another study with mice that had GPx4 ablation showed the importance of inhibiting ferroptosis to prevent spinal motor neuron degeneration since mice with this characteristic showed motor disorders. The tests

carried out to confirm ferroptosis included the absence of apoptotic markers (caspase-3 and TUNEL) and activation of the ERK pathway [220]. Magnetic resonance images showed an accumulation of iron in the brain of patients with HD [221]. However, the pathway that induces ferroptosis in the brain has not been fully elucidated [222]. The mutated huntingtin protein (mHTT) is cleaved at different points since it has different cleavage sites than those present in the normal protein, which result in fragments with different sizes inside neurons (small oligomers or monomers) [223]. It has been shown that mHTT and the wild version are associated with the outer mitochondrial membrane [224].

Inside the cells, there must be a balance between the state of mitochondrial fission and fusion for the proper functioning of this organelle [225]. When an imbalance occurs, cellular respiration is affected and, consequently, there is an increase in ROS inside cells [225]. Mitochondria can generate significant amounts of ROS, resulting from normal organelle metabolism and the electron transport chain that contributes to oxidative stress. Additionally, an mHTT leads to increased oxidative stress, which consequently increases ROS levels in the cell [226]. Under normal conditions, GSH regulates the activity of GPx4, and its function is to inhibit ferroptosis and eliminate the excess of lipid peroxides [225]. However, an increase in ROS levels and, consequently, an increase in lipid peroxides, which leads to a depletion of GSH, decreases GPx4 [225]. There is a series of intracellular signals that culminate in an imbalance in cell homeostasis and leads to the process of ferroptosis. In patients with HD, there is a deregulation of GSH which interferes with their functions and enzymes dependent on its action [117,225].

In patients with HD and asymptomatic carriers, high lipid peroxidation and low GSH plasma levels have been found, showing that oxidative stress may be linked to the pathophysiological mechanism of HD [227]. Iron chelators could be an alternative for treatment [228].

Nfr2 is a transcriptional regulator of genes involved with ferroptosis, such as GPx4, which can be found in cytosol and modulates mitochondrial function [206]. These are essential components for the process of ferroptosis and demonstrate the importance of Nfr2 in protecting against ferroptosis. Therefore, Nfr2 can be an alternative therapy for reducing ferroptosis [206]. Currently there is no specific treatment available for HD, only palliative care [212]. In general, the multidisciplinary treatment available at the moment is focused on the palliative treatment of symptoms and neuroprotection of patients [218,229]. Although there is no specific treatment for HD, several studies are being developed focusing on silencing the DNA or mRNA of the mutated allele whose gene varies in the number of copies [215,229]. Another promising alternative is the creation of mouse models with HD and the involvement of GSH activity as sources of possible target treatments [117].
