The ER is known as an organelle involved in protein maturation and folding (Read and Schröder, 2021). Toxic protein aggregates in neurodegenerative diseases, pathogen-associated molecular patterns (PAMPs), danger-associated molecular patterns (DAMPs), ROS, and reactive nitrogen species (RNS), can disrupt protein folding processes in the ER lumen, leading to ER stress (Zhang and Kaufman, 2008). In the condition of ER stress, the aggregation of unfolded or misfolded proteins within the ER lumen of neurons and neuroglia leads to failure of ER in maintaining protein homeostasis through UPR and ERAD. For instance, accumulation of tau protein in AD can affect essential components of ERAD and block this pathway, leading to the accumulation of more misfolded proteins in the ER lumen (Hetz and Saxena, 2017; Ghemrawi and Khair, 2020). Subsequently, UPR-dependent inflammation and apoptotic pathways are induced, resulting in neuronal cell death (Sprenkle et al., 2017; Ghemrawi and Khair, 2020). The ER stress can also be induced by ER Ca²⁺ dysregulation, impairments in vesicular trafficking, or any defects in UPR components (Cooper et al., 2006; Sprenkle et al., 2017). PKR-like ER kinase (PERK), inositol-requiring transmembrane kinase/endoribonuclease 1 α (IRE1α), and activating transcription factor 6 (ATF6) are three vital sensor proteins that are involved in UPR regulation (Ghemrawi and Khair, 2020). Under normal conditions, these proteins are inactive due to association with ER chaperone proteins such as Immunoglobulin binding protein (BiP) or 78 kDa glucose-regulated protein (GRP78), which are members of heat shock protein families (Halperin et al., 2014).

Under ER stress conditions, the misfolded proteins interact with the substrate binding domain of BiP. Consequently, BiP is released and leads to dimerization and auto-phosphorylation of PERK, as well as intramembrane proteolysis of ATF6 and phosphorylation of IRE1α. Subsequently, the UPR cascade activates to maintain protein homeostasis (Ghemrawi and Khair, 2020). To elaborate, the phosphorylation of the alpha subunit of eukaryotic translation initiation factor (eIF2α) followed by activation of PERK occurs through BiP dissociation. This process inhibits protein synthesis to prevent overload of proteins in the ER lumen (Hetz and Saxena, 2017; Almeida et al., 2022), therefore attempting to restore protein homeostasis (Da Silva et al., 2020). Besides, under prolonged ER stress conditions and failure in the UPR mechanism, p-eIF2α promotes activating transcription factor 4 (ATF4) translation, which enhances up-regulation of pro-apoptotic factors, including CHOP (also known as GADD153; Ghemrawi and Khair, 2020). Eventually, down-regulation of anti-apoptotic Bcl-2 family makes neurons more susceptible to death (Doyle et al., 2011; Hetz and Saxena, 2017; Da Silva et al., 2020; Figure 1A). Moreover, it has been claimed that TRB3 genes, GADD34, death receptor 5 (DR5), ER oxidase 1 (ERO1), and other apoptotic molecules can potentially receive apoptosis signals from CHOP and induce cell death (Taalab et al., 2018; Da Silva et al., 2020). ATF4 also induces transcription of the p53-upregulated modulator of apoptosis (PUMA), which results in ER-stress-induced neuronal apoptosis (Galehdar et al., 2010). Interestingly, experiments have indicated that CHOP could not induce apoptosis in PUMA-deficient neurons, demonstrating the key role of PUMA in CHOP-induced neuronal apoptosis (Galehdar et al., 2010). Moreover, in IRE1-α signaling pathway, the second arm of UPR, after the release of BiP by aggregated proteins, IRE1-α undergoes oligomerization and auto-phosphorylation. p-IRE1α facilitates neuronal death by activation of the apoptotic-signaling kinase-1 (ASK1) and other apoptotic factors as a result of c-Jun N-terminal kinase (JNK) and p38 mitogen-activated protein kinase (p38 MAPK) pathways activation (Ghemrawi and Khair, 2020). This pathway may induce p53 activation and lead to the upregulation of Bcl-2 associated X (BAX) protein, which triggers the release of cytochrome C (Cyt C) from the mitochondria to the cytosol and cause apoptotic neuronal cell death (Stefani et al., 2012). Furthermore, the RNase activity of IRE1α plays a critical role in splicing the mRNA coding for X-box binding protein 1 (XBP1) and increases the expression of genes involved in ER machinery, such as BiP (Lee et al., 2003; Hirota et al., 2006; Chen et al., 2022). Besides outlined functions, IRE1α participates in the degradation of some mRNAs and microRNAs, known as “regulated IRE1α-dependent decay” (RIDD; Hetz and Saxena, 2017; Figure 1B). Mutations in Presenilin 1 and 2 (PS1 and PS2), which are frequently involved in AD, can inhibit IRE1 and impair UPR, leading to AD pathology and neuronal cell death (Doyle et al., 2011). ATF6 is the third sensor protein of UPR cascades which is embedded in the ER membrane. By interaction of aggregated proteins with ATF6 in the ER lumen and release of BiP, ATF6 translocates to the Golgi apparatus and undergoes proteolysis. Subsequently, cleaved ATF6 induces transcription of ER chaperones and XBP1 in the nucleus and participates in protein homeostasis (Da Silva et al., 2020; Figure 1C).

In AD, Aβ can trigger ER stress, mitochondrial fragmentation, and neuronal death through ER Ca²⁺ release by ryanodine receptors (RyRs) and inositol triphosphate receptors (IP3R; Chami and Checler, 2020). Based on studies, overexpression of RyRs contributed to Ca²⁺ dysregulation in AD mouse models and cell lines. In addition, increase in IP3 receptor-mediated Ca²⁺ signaling was indicated in AD patients’ fibroblast cells (Callens et al., 2021). Aβ oligomer-dependent ER stress responses can subsequently activate different kinases which phosphorylate specific epitopes on tau leading to the development of neurofibrillary tangles (NFTs) and propagating AD pathology (Sprenkle et al., 2017). Aβ peptides can activate ASK1 and JNK pathways, which can subsequently mediate ER stress-induced apoptosis (Ghemrawi and Khair, 2020). Both ASK1 and JNK were reported to be upregulated in transgenic mouse brains and post-mortem AD samples, respectively (Galvan et al., 2007; Sbodio et al., 2019). It has been revealed that CHOP activation plays a crucial role in the triggering and progression of pathological hallmarks of AD. In agreement, CHOP and its downstream effectors, including caspase-12 and GADD34, are markedly upregulated in the brains of AD patients (Ghemrawi and Khair, 2020). In addition, phosphorylated forms of PERK and eIF2α were significantly increased in the hippocampal pyramidal cells and frontal cortex of AD patients (Stutzbach et al., 2013). The evidence also shows that ER chaperones, including BiP, are also upregulated in the cerebrospinal fluid (CSF) and AD brains (Ghemrawi and Khair, 2020).

Mutations in PARK7, a gene involved in familial PD, might activate ASK1-induced neuronal death in PD. This can be due to the dysfunction in protecting against the Daxx-ASK1 cell death axis, which plays a key role in the completion of signaling pathways from cell surface death receptors (Chang et al., 1998; Homma et al., 2009). In addition, upregulation of ER stress markers, such as GRP78, p-PERK, and p-eIF2α in dopaminergic (DA) neurons of post-mortem PD samples (Shi et al., 2022), demonstrate their function in initiating apoptosis pathways, which could cause serious clinical implications. According to evidence, upregulation in ER stress markers, including BiP and CHOP in post-mortem HD brains, may be associated with neuronal death in HD (Shi et al., 2022). Mutation in genes such as SOD1, a gene encoding Superoxide dismutase 1 (SOD1), can also induce ER stress in neurons in ALS and cause neuronal damages (Sprenkle et al., 2017). ALS-associated mutations in vesicle-associated membrane protein-associated protein B (VAPB) can physically interact with ATF6 and disturb its natural function (Hetz and Saxena, 2017). Patients with ALS-associated VAPB mutations indicated malfunctions in Ca²⁺ signaling and storage, excessive ER stress, and neuronal death as a result of inhibition of ATF6 (Ghemrawi and Khair, 2020). Furthermore, upregulation of PERK, IRE1α, and ATF6 was found in the ALS mouse models (Ghemrawi and Khair, 2020; Zhao et al., 2022).

According to the critical role of ER stress in the occurrence of neuronal cell death in neurodegenerative diseases, targeting associated pathways seem to have hopeful effects on protecting neurons from death (Figure 1). Among three arms of UPR in ER stress conditions, the PERK pathway is the most well-studied in the neuroprotective effects of inhibition of ER stress. In parallel with this, Salubrinal, an anti-ER stress compound, has been well investigated in neurodegenerative disease pathology and treatment (Gupta S. et al., 2021; Ajoolabady et al., 2022). Salubrinal is an activator of UPR, which raises ER chaperone levels, including BiP. It inhibits eIF2α dephosphorylation which can attenuate neuronal death by interfering with death-related signaling pathways, including ATF4 or ASK1 (Figure 1, Table 2; Niso-Santano et al., 2011; Wu et al., 2014; Sprenkle et al., 2017). Accumulating evidence indicates that Salubrinal reduced ER accumulation of α-synuclein and significantly protected against α-synuclein-mediated dopaminergic (DA) neuronal death in transgenic mouse models (Colla et al., 2012). Also, Salubrinal reduced the accumulation of mutant huntingtin (mHTT) by upregulation of BiP and p-eIF2α, and prevent neuronal cell death (Maity et al., 2022). In addition, the drug Adaptaquin blocks Tribbles pseudokinase 3 (Trib3) induction by inhibiting ATF4 and CHOP activity probably through an eIF2α-independent mechanism, leading to neuronal protection in mouse models of PD (Figure 1, Table 2). More investigation is required for the neuroprotective effects of Adaptaquin in ER stress-induced neuronal cell death (Karuppagounder et al., 2016; Aime et al., 2020). Moreover, the PKR inhibitor “C16” can reduce transcriptional induction of pro-apoptotic target genes of ATF4, such as CHOP, Trib3, and PUMA (Figure 1, Table 2). This could significantly reduce MPP+ and 6-OHDA neurotoxin-induced neuronal cell death in PD models (Demmings et al., 2021). According to the experimental study, PUMA expression can be downregulated by directly targeting CHOP to decrease ER stress-induced neuronal apoptosis (Galehdar et al., 2010). Notably, pharmacological inhibition of ATF4, using imidazole-oxindole PKR inhibitor, indicated neuroprotection against neurotoxin-induced cell death in PD models (Demmings et al., 2021). Comparing motor neuron death in ATF4-ablated transgenic ALS mouse models with those expressing normal levels of ATF4 demonstrated the possible role of ATF4 ablation in neuroprotection against ALS by reducing apoptosis components, including CHOP (Matus et al., 2013). Likewise, another study revealed an increase in neuronal death in PD rat models by overexpression of ATF4 using recombinant Adeno-Associated Virus (rAAV; Gully et al., 2016). Halliday et al. (2017) revealed the inhibition of UPR-induced p-eIF2α signaling and neuronal survival by two chemical compounds termed “Trazodone” and “dibenzoylmethane” (DBM) in prion-infected mice (Figure 1, Table 2), presumably by reversing translational attenuation and lowering levels of ATF4 and CHOP which needs to be more inquired in other neurodegenerative diseases including AD and PD. Interestingly, it has been demonstrated that PERK inhibitor GSK2606414 (Figure 1, Table 2), despite its pancreatic toxicity (Halliday et al., 2015), inhibits and reduces PERK expression, which has a neuroprotective effect on DA neurons in Substantia Nigra pars compacta (SNpc) of PD mouse models, and improves the motor performance and neuronal excitability of PD mice (Mercado et al., 2018). In addition, inhibition of PERK signaling with IRSIB has been investigated in ALS rodent models, and a reduction in ATF4 and CHOP levels has been indicated (Figure 1, Table 2), which results in significant neuronal survival. In the same study, a reduction in IRE1-dependent signaling has also been indicated (Halliday et al., 2015, 2017).

Inhibition of the IRE1 pathway is also a possible way to attenuate neuronal cell death. For instance, Kinase-Inhibiting RNase Attenuator 6 (KIRA6) inhibits apoptosis by breaking IRE1 oligomers and inhibiting RNase activity of IRE1α (Figure 1, Table 2; Ghosh et al., 2014). Given the vital role of ASK1 in IRE1-mediated UPR and inducing apoptosis, targeting and deletion of ASK1 in mutant SOD1-transgenic mice have been indicated to mitigate motor neuronal death (Homma et al., 2009). Additionally, evidence shows that overexpression of XBP1 protects DA neurons against neurotoxin-induced ER Stress-associated cell death (Valdes et al., 2014; Shi et al., 2022). Furthermore, upregulation of autophagy by targeting XBP1 in ALS and HD models is known to be another way of protection from neuronal cell death (Remondelli and Renna, 2017). The experiments have been demonstrated that ablation of ATF6 facilitates DA neuronal death caused by neurotoxins, including 6-hydroxydopamine (6-OHDA) and 1-methyl-4-phenyl-pyridinium (MPP+; Shi et al., 2022). This indicates the plausible role of the third arm of UPR pathways in inducing neuronal death. Kaempferol (Table 2), a plant-derived ER stress-induced cell death inhibitor, has also reduced the expression of ATF6, PERK, IRE1α, as well as CHOP in Brefeldin A (BFA)-induced ER stress in IMR32 cell lines. More investigations is needed to determine whether it is effective in animal and human neurodegenerative models (Abdullah and Ravanan, 2018). It is also claimed that 3β-Hydroxysteroid-Δ24 reductase (DHCR24) can protect neuronal cells by reducing BiP and CHOP levels and attenuating ER stress-specific apoptotic signaling pathways (Table 2; Lu X. et al., 2014). Targeting other indirect factors involved in ER stress, such as IP3 receptors and Ryanodine receptors, has also been examined. Remarkably, the first research confirming blocking Inositole triphosphate receptors (IP3Rs) and ryanodine receptors (RyRs) to decrease ER stress-induced Ca²⁺ dyshomeostasis in DA neurons revealed that a RyRs blocker (RY) markedly reduced 6-OHDA-induced cytosolic Ca²⁺ increases. In contrast, an IP3Rs blocker (Xes) had no considerable effect on cytosolic Ca²⁺ levels and neuronal cell death (Table 2). Moreover, pre-treatment with an ER stress inhibitor 4-phenyl butyric acid (4-PBA) had a neuroprotective effect on DA neurons from 6-OHDA-induced apoptosis (Table 2; Huang et al., 2017).

Healthy mitochondria produce ROS as a byproduct of oxidative phosphorylation mainly as signaling messengers (Hajam et al., 2022; Trushina et al., 2022), while defective mitochondria generate aberrant amounts of ROS and cause oxidative stress and suspend cellular homeostasis due to the disruption of the balance between ROS generation and antioxidant function (Figure 2; Höhn et al., 2020; Holubiec et al., 2022). Neurons are susceptible to produce free radicals due to being metabolically very active. Evidently, any pathological situation or dysfunction in neurons can generate excess ROS leading to oxidative stress (Bhat et al., 2015). Given that the metabolism rate of neurons is very high, the brain has a high oxygen consumption rate (20%–25% of the total body oxygen consumption). Furthermore, the high content of easily peroxidizable unsaturated fatty acids (PUFA) and the relative paucity of antioxidant enzymes compared with other organs makes the brain vulnerable to free radical damage (Nunomura et al., 2007; Rocha et al., 2018). Therefore, the excessive production of ROS and RNS resulting from various factors, including calcium influx and mitochondrial dysfunction, can compromise cell fidelity and exacerbate disease progression. Hydrogen peroxide (H₂O₂), superoxide anion (O₂⁻), and highly reactive hydroxyl radical (HO^•) are the ROS involved in neurodegeneration. The RNS, such as nitric oxide (NO), are also found to have a deleterious effect on neurons (Singh et al., 2019; Korovesis et al., 2023).

ROS adversely affects the oxidation or peroxidation of specific macromolecules such as lipid peroxidation to malondialdehyde (MDA), protein carbonylation, and oxidation of specific nucleic acids (Singh et al., 2019). It has been claimed that ROS allows Cyt C and AIF to be released from the inner mitochondrial membrane (IMM) and initiate an apoptotic cascade (Figure 2; Bhat et al., 2015). The neural brain cells are enriched in PUFA, such as docosahexaenoic acid, arachidonic acid, and cardiolipin, which makes cells susceptible to lipid peroxidation and subsequent outcomes (Höhn et al., 2020; Falabella et al., 2021). For example, Cardiolipin (CL), a specific phospholipid of IMM, located in the sites of ROS production in the mitochondrial electron transport chain, can potentially be a target for ROS due to its high composition of unsaturated acyl chains. After peroxidation by ROS, CL is supposed to be involved in the conformational changes in IMM and the release of pro-apoptotic proteins, including Cyt C (Bhat et al., 2015; Falabella et al., 2021). However, a serine protease called HTRA2 takes part in the inhibition of pro-apoptotic protein release from mitochondria, but its function may not be sufficient, or it may be disturbed (Bhat et al., 2015). Moreover, the brain is also enriched in redox-active metals (copper and iron) that involve in generating free radicals and peroxidation of lipids (Sbodio et al., 2019; Falabella et al., 2021). These metals could be reduced by proteins such as Aβ, which leads to the formation of H₂O₂ and pro-apoptotic lipid peroxidation (LPO) products, such as 4-hydroxy-2-nonenal (HNE; Opazo et al., 2002; Jiang et al., 2009). In addition to the role of HNE in decreasing antioxidant levels by reacting with sulfhydryl groups (Taso et al., 2019), HNE forms stable adducts with amine or thiol groups in proteins and may eventually anomalously activate caspases and triggers neuronal cell death (Figure 2; Gaschler and Stockwell, 2017; Barrera et al., 2018). It has been demonstrated that Cadmium (Cd) could easily penetrate the blood-brain barrier and contribute to neurotoxicity. Cd can induce mitochondrial ROS production in neurons as well as downregulation of x-linked inhibitor of apoptosis (XIAP), leading to an increase in mouse double minute 2 (MDM2). Consequently, decrease in p53 facilitates neuronal apoptosis cell death (Figure 2; Zhao et al., 2020).

ROS can influence protein oxidation and contribute to the formation of insoluble protein aggregates (Figure 2), including Aβ peptides and NFTs, α-synuclein, and mSOD1. Oxidative stress can enhance expression of gamma-secretase and beta-secretase (BACE1) through activation of MAPK pathway and involves in Aβ production in neurons and amyloidogenic processing of amyloid precursor protein (APP; Tamagno et al., 2005; Lin and Beal, 2006; Höhn et al., 2020). Oxidative stress also increases tau phosphorylation by activation of glycogen synthase kinase 3 (GSK3; Lin and Beal, 2006). ROS also mediate JNK/stress-activated protein kinase pathways, which subsequently contributes to hyper-phosphorylation of tau proteins, formation of intracellular NFTs, and Aβ-induced neuronal death (Liu et al., 2017). Indeed, hydrogen peroxide and deficiency of mitochondrial antioxidant enzymes has been tested in animal models, which led to increase in Aβ levels (Gerakis and Hetz, 2019) and neuronal cell death (Lin and Beal, 2006). Interestingly, it has been revealed that neurons close to Aβ-plaques seem to be more at risk of cell death due to more severe toxicity of the microenvironment caused by oxidative stress in AD (Xie et al., 2013). Aβ-mediated oxidative stress can enhance the activity of a serine/threonine phosphatase, known as calcineurin, and promotes neuronal death by associating with caspases and/or triggering pro-apoptotic Bcl-2 proteins (Figure 2; Awasthi et al., 2005; Akanji et al., 2021). Moreover, ROS can have noxious effects by affecting Ca²⁺ cation channels on the ER and plasma membrane. Impaired Ca²⁺ channels can, in turn, lead to Ca²⁺ influx to the cytosol, as well as impairment in pumping intracellular Ca²⁺ out of the cell to maintain homeostasis (Brini et al., 2014). Toxic levels of calcium can trigger cell death through activation of apoptotic factors, including calcium-dependent proteases calpain and caspases (Figure 2; Fairless et al., 2014). Importantly, ROS has also a deleterious effect on nuclear factor erythroid 2-related factor 2 (Nrf2) regulation. Nrf2 is a transcription factor that has an essential role in regulating cellular redox homeostasis (Kovac et al., 2015). Reduced levels of Nrf2 can subsequently result in mitochondrial dysfunction and apoptosis. Therefore, upregulation of Nrf2 reduces oxidative stress by promoting the expression of antioxidant enzymes (Figure 2; Li Y. C. et al., 2021). Besides, the Repressor element 1-silencing transcription factor (REST) regulates cell death-associated genes, including BAX, BH3 interacting domain death agonist (BID), and also PUMA, and maintains resistance to stress conditions. REST-depleted neurons are more susceptible to oxidative stress and anomalously express apoptosis-inducing genes which facilitates neuronal death in AD (Lu T. et al., 2014). Remarkably, upregulation of the transient receptor potential melastatin-2 (TRPM2) in the SNpc of human PD brains agrees with the role of TRPM2 in ROS-induced cell death in PD pathogenesis (Malko et al., 2021). The function of Parkin can be affected by mitochondrial dysfunction and oxidative stress. This will promote Cyt C release and caspase-9 activation, which leads to neuronal cell death and facilitating PD pathogenesis (Figure 2; Lin and Beal, 2006). Defect in Complex I of the mitochondrial electron transport chain by aggregation of α-synuclein and PTEN-induced putative kinase 1 (PINK1) mutations can be also involved in PD pathogenesis by inducing neuronal apoptosis and failure in maintaining mitochondrial membrane potential, respectively (Liu et al., 2017; Morales-Martínez et al., 2022).

Besides the oxidation of macromolecules is elevated in the brain of patients, decreased levels of antioxidants, including uric acid, vitamin C and E, superoxide dismutase (SOD), catalase, and especially the antioxidant glutathione (GSH), lead to decreased detoxification of ROS in the brain cells which has been discovered in various AD, PD, and other neurodegenerative disease patients (Singh et al., 2019). Oxidative damage occurs before the onset of significant plaque pathology in the AD by triggering glycogen synthase kinase 3 (Lin and Beal, 2006; Wu et al., 2019). Overproduction of ROS and RNS in AD patients has been detected. Additionally, 8-hydroxydeoxyguanosine (8-OHdG), a biomarker of oxidative damage, was elevated in AD ventricular CSF (Niedzielska et al., 2016). However, alterations in plasma levels of antioxidants in AD patients are paradoxical in experimental reports (Niedzielska et al., 2016). Aβ plaques can cause Ca²⁺ dyshomeostasis in ER leading to Ca²⁺ influx in the cytosol. Consequently, endogenous GSH levels are reduced, and ROS can cause neurotoxic effects (Liu et al., 2017). The alterations of transition metals, including Cu²⁺, Zn²⁺, Fe³⁺, have been assessed in AD samples. The results indicated that transition metals seem to be imbalanced in AD brains which contributed to oxidative damage and subsequent neuronal death (Bhat et al., 2015). Evidence also shows reduction in nuclear REST levels in neurons of degenerated regions in AD brains, such as prefrontal cortical and hippocampal neurons, which transcriptional dysregulation in apoptotic genes and vulnerability to oxidative stress has made them susceptible to apoptosis cell death (Lu T. et al., 2014). Elevated levels of activated JNK have been reported in post-mortem AD samples, which is probably associated with Aβ formation. JNK implicates in the upregulation of BACE1 and promotes the formation of Aβ, leading to oxidative stress and neuronal apoptosis cell death (Yao et al., 2005; Guglielmotto et al., 2011; Sbodio et al., 2019).

HNE levels were significantly high in the CSF of AD and PD patients, which can be considered as an important reason for neuronal demise and behavioral symptoms in neurodegenerative diseases (Taso et al., 2019). Decreased level of GSH in the Substantia Nigra (SN) of PD patients is one of the earliest biochemical alterations that facilitate the neurotoxic effects of ROS (Niedzielska et al., 2016). Moreover, overexpression of α-synuclein in transgenic mice results in mitochondrial dysfunction and increased oxidative stress (Song et al., 2004). It has been observed that aberrant activity of mutant human SOD1 in ALS patients leads to increase in the level of free radicals in CSF, serum, and urine samples of ALS patients, which exacerbates neuronal damage (Liu and Wang, 2017). Indeed, mSOD1 accumulation in outer mitochondrial membrane (OMM) can result in mitochondrial dysfunction and promotes aberrant ROS production. In an experimental study, mice expressing mSOD1 showed more oxidative damage to mitochondrial lipids and molecules (Mattiazzi et al., 2002; Liu et al., 2004). P53 can regulate genes involved in oxidative stress and mitochondrial function. Environmental toxicants such as bisphenol A (BPA) play a role in inducing neurotoxicity by significantly increased oxidative stress. BPA subsequently leads to upregulation in apoptotic inducing factors, including p53, PUMA, and Drp-1 (Ishtiaq et al., 2021). Additionally, in Huntington’s disease pathology, mHTT can interact with p53 and increase p53 levels, and eventually upregulates apoptotic factors BAX and PUMA, which leads to apoptosis (Bae et al., 2005; Lin and Beal, 2006). mHTT also interacts with mitochondrial membranes, causing mitochondrial abnormalities and an increase in ROS generation, which potentially leads to neuronal degeneration and cell death (Ross and Tabrizi, 2011; Liu et al., 2017). Elevated levels of lipid peroxidation and decreased levels of GSH content have been indicated in the plasma of HD patients (Klepac et al., 2007). In addition, increased levels of 8-OHdG have been observed in the serum of HD patients and post-mortem HD samples (Sbodio et al., 2019).

Undoubtedly, oxidative stress has neurotoxic effects in the pathogenesis of neurodegenerative diseases. However, the exact molecular pathways remain unclear and need to be more inquired about finding a promising therapeutic strategy to decrease neuronal death in neurodegenerative diseases and extend the lifespan of patients. In this regard, the antioxidant properties of many candidate compounds have been reported. In addition, many other molecules that mediate oxidative stress-induced apoptosis have been targeted to prevent neuronal cell death and disease progression (Figure 2). A cytoprotective polypeptide called Humanin (HN), which is encoded by mtDNA, has neuroprotective activity against cellular stress conditions, such as oxidative stress. HN regulates mitochondrial function by targeting apoptotic factors and inhibits apoptosis by upregulation of Bcl-2 and downregulation of Bid and Bax (Figure 1, Table 2; Hazafa et al., 2021). Also, some amino acid derivatives have shown anti-apoptotic effects in ALS models in vitro (Sirianni et al., 2015). N-acetyl-L-tryptophan (L-NAT) and N-acetyl-DL-tryptophan (DL-NAT) have inhibited neuronal cell death in H₂O₂-induced NSC-34 motor neurons (Figure 1, Table 2). L-NAT inhibits the release of Cyt C/Smac/AIF from mitochondria, as well as inhibition of caspase activity, thereby preventing neuronal apoptosis cell death (Sirianni et al., 2015). Another study by Yoo et al. (2017) showed that N-acetyl serotonin (NAS) has anti-apoptotic properties by activating neurotrophic signaling TrkB/CREB/BDNF pathways. NAS induces and activates the expression of antioxidant enzymes to reduce the level of ROS (Figure 1, Table 2). It also regulates anti- and pro-apoptotic factors and restores mitochondrial membrane potential to prevent neuronal cell death in neurodegenerative disease models (Yoo et al., 2017). According to evidence, mitochondrial permeability transition pore (mPTP) can increase mitochondrial calcium retention and cause cell death. CoQ10 is considered as an inhibitor of mitochondrial permeability transition pore and protects neurons from oxidative stress and apoptosis. The exact protective mechanism of CoQ10 is still indistinct and needs more experiments. CoQ10 may decrease apoptosis by maintaining the integrity of the mitochondrial membrane and inhibiting Cyt C release. CoQ10 may also decrease the Bcl-2 protein level and prevent caspase activation (Figure 1, Table 2; Akanji et al., 2021). As previously mentioned, protein aggregates can potentially induce oxidative stress in neuronal cells. For instance, Aβ can interact and bind to Aβ-binding alcohol dehydrogenase (ABAD), a mitochondrial-matrix protein, and induce apoptosis and free-radical generation. Blocking the interaction of Aβ and ABAD with a “decoy peptide” suppress oxidative stress and neuronal death. In contrast overexpression of ABAD in mouse models contribute to exaggerating cellular stress and further complications (Lustbader et al., 2004). Moreover, NADPH oxidase (NOX) catalyzes the formation of O₂⁻ and participates in elevating neurotoxicity and increasing cell death in HD. Treatment of HD models with NOX inhibitors, including diphenyleneiodonium, apocynin, and VAS2870, prevented neurons from cell death (Figure 1, Table 2; Sbodio et al., 2019).

In several studies, it has been reported that plant iridoids have therapeutic applications in several neurodegenerative diseases by regulating apoptotic factors and neuroprotective proteins (Dinda et al., 2019). Aucubin (AU) is an iridoid glycoside with neuroprotective properties, which significantly increases cell viability in neurons via oxidative stress reduction. AU enhances the antioxidant capacity of cells through the Nrf2 signaling pathway and decreases ROS-induced neuronal apoptosis by regulating mitochondrial membrane potential and reducing ROS generation (Figure 1, Table 2; Li Y. C. et al., 2021). The pharmacological effects of other herbal compounds have also been investigated. Salidroside (Sald), is a Chinese plant-derivative compound that could detoxify neurons by suppressing the elevation of the intracellular ROS level and induction of antioxidant enzymes. Sald also participates in the downregulation of pro-apoptotic protein Bax and upregulation of anti-apoptotic protein Bcl-xl, and prevents neuronal cell death (Figure 1, Table 2; Zhang et al., 2010). Some plant-derived organic oils, including a bicyclic monoterpene termed “Borneol”, indicated neuroprotective effects against H₂O₂-induced apoptosis in vitro. Borneol alleviates neuronal apoptosis by inhibiting Cyt C and AIF release through increase in the expression of anti-apoptotic protein Bcl-2 and decrease in expression of pro-apoptotic protein Bax (Figure 1, Table 2; Hur et al., 2013). -gingerol also attenuates Aβ-induced oxidative stress. Studies revealed that -gingerol scavenges free radicals and decreases phospholipid peroxidation, as well as improves cellular redox balance (Figure 1, Table 2; Lee et al., 2011). Increase in the activity of glycogen synthase kinase-3β (GSK-3β) under oxidative stress condition leads to Nrf2 dysregulation (Kumar et al., 2012). Thus, GSK-3β inhibitors, including an anti-oxidative phytochemical known as Isoorientin, can have neuroprotection against oxidative damage by regulating Nrf2 antioxidant activity (Figure 1, Table 2; Lim et al., 2007; Gianferrara et al., 2022).

Research findings have indicated that several neurodegenerative diseases are associated with inflammation (Kwon and Koh, 2020). Activated microglia and T lymphocytes have been detected in the SN of PD patients (Dias et al., 2013). In parallel with this, high expression levels of chemokines, interleukins, interferons, and tumor necrosis factor-α (TNF-α) have been discovered in the striatum and substantia nigra of PD post-mortem brain samples and CSF of AD patients (Hirsch and Hunot, 2009; Llano et al., 2012; Gelders et al., 2018). A neurotoxic microenvironment can be promoted by the continuous secretion of inflammatory mediators from microglia and astrocytes, thus facilitating neural degeneration, and glial cell death (Pardillo-Díaz et al., 2022; Song et al., 2022). In addition to apoptosis cell death, pyroptosis, a non-apoptotic programed cell death, can also occur in the CNS, which is mainly mediated by inflammatory processes. Pyroptosis is characterized by cell swelling, formation of pores in the plasma membrane carried out by cleaved Gasdermin D, and the release of pro-inflammatory cytosolic contents into the extracellular space (Figure 3; Walle and Lamkanfi, 2016; Man et al., 2017; Wang et al., 2019). Besides, some specific caspases, including caspase-1, 4, 5, 11, are called “inflammatory caspases”, can mediate pyroptosis (Taylor et al., 2008; Gaidt and Hornung, 2016). Some factors are associated with initiation of inflammatory cascades and promotion of disease pathology. For instance, ER stress can cause inflammation in neurodegenerative diseases. In other words, inducing ER stress in neurons mostly initiates apoptosis, whereas intense ER stress in glial cells can potentially trigger inflammation in neurodegenerative diseases (Sprenkle et al., 2017). The UPR can increase the production and the release of inflammatory factors, such as transcription factor “nuclear factor kappa-light-chain-enhancer of activated B cells” (NF-κB), interleukin 1 (IL-1), IL-6, IL-8, and TNF-α (Feng et al., 2017). Furthermore, in ER stress condition, p-IRE1 can bind to the TRAF-2 protein and forms the TRAF2-IRE1 complex. The complex may bind to ASK-1 and activate the JNK signaling pathway that enhances inflammation (Vukic et al., 2009; Mohammed-Ali et al., 2015). In addition, p-PERK also facilitates neuroinflammation by inducing the JAK1/STAT3 signaling pathway in glial cells (Meares et al., 2014). Moreover, neuroinflammation is a cause and a consequence of chronic oxidative stress. Studies indicate that the production of free radicals (such as ROS) are elevated in neurodegenerative diseases, which can be due to neuroinflammation (Dias et al., 2013; González-reyes et al., 2017). On the other hand, increased levels of ROS can contribute to pro-inflammatory gene transcription and release of cytokines, including IL-1, IL-6, and TNF-α (Sochocka et al., 2013; Teleanu et al., 2022). Debris of dead neurons may trigger glia-mediated neuroinflammation and initiate a pro-inflammatory cascade that can exacerbate disease progression (Wang et al., 2015; Joshi et al., 2019). Recently, it was revealed that microglial pro-inflammatory cytokines are associated with increased α-synuclein aggregation (Guo et al., 2020). Thereby, protein aggregates can be involved in inducing neuroinflammation in the CNS parenchyma.

It has been investigated that the failure to clear apoptotic cells and activation of glial cells by DAMPs and PAMPs can enhance inflammation response and subsequent neuronal damage and loss (Imbeault et al., 2014; Salter and Stevens, 2017; Voet et al., 2019; Cheng Y. et al., 2021). Activation of complex signaling cascades such as the NLR family pyrin domain containing 3 (NLRP3) inflammasome can be triggered by a wide range of factors including cellular stress, infection (Bader and Winklhofer, 2020; Mahboubi Mehrabani et al., 2022), protein aggregates, and activated microglia (Nichols et al., 2019; Bader and Winklhofer, 2020; Tansey et al., 2022). This phenomenon contributes to producing more neurotoxic cytokines and chemokines such as IL-1β, IL-6, TNF-α, and CCL2 (also known as monocyte chemo-attractant protein-1 or MCP-1), that cause enhancement in neurotoxicity and cell death (Sprenkle et al., 2017; Rocha et al., 2018; Joshi et al., 2019; Nichols et al., 2019). The exact mechanism of neuronal death through activation of NLRP3 inflammasome in microglia has not been perfectly discovered yet (Lee et al., 2019). In fact, NLRP3 inflammasome induces heteromer formation or aggregation of apoptosis-associated speck-like protein containing a caspase recruitment domain (ASC), which subsequently activates caspase-1. It results in the maturation of IL-1β and induction of pyroptosis using cleaved Gasdermin D and membrane pores (Figure 3; Stancu et al., 2019; Wang et al., 2019; Bader and Winklhofer, 2020; Feng et al., 2020; Onyango et al., 2021).

Specific receptors, including TLRs and NLRs expressing on neurons or neuroglia, can recognize extracellular neurotoxic protein aggregates (Nichols et al., 2019; Leng and Edison, 2021; Heidari et al., 2022). This may facilitate neuronal cell death (Li Y. et al., 2021) by involvement in caspase activation and secretion of pro-inflammatory factors through NF-κB activation (Figure 3; Rocha et al., 2018; Leng and Edison, 2021). The expression of the receptors for cytokines was indicated in DA neurons, making neurons more susceptible to damage and death (Hirsch and Hunot, 2009). Complement receptors and Fc receptors on microglia can also mediate pro-inflammatory responses independent from extracellular protein aggregates (Leng and Edison, 2021). Activation of microglia can potentially contribute to activating astrocytes, which rapidly upregulates inflammatory signaling molecules (Sims et al., 2022), and can potentially initiate or enhance nitrosative stress due to producing NO (Rocha et al., 2018). In support of this claim, the presence of activated astrocytes is confirmed in post-mortem brain samples of various neurodegenerative disease patients (Hashioka et al., 2021). DAMPs released by dying neurons may also activate microglia through the ionotropic P2X and metabotropic P2Y purinergic receptors and initiate an inflammatory response. For example, under pathological conditions, P2X7 receptors can be overexpressed in the CNS. ATP acts as a DAMP and activates P2X7 receptors, and promotes chronic inflammatory neurological disorders (Thawkar and Kaur, 2019). Post-mortem brain samples of AD patients showed overexpression of P2X7 receptors, which can be associated with disease pathology and progression. It is also suggested that P2X4 receptor overstimulation may result in neuronal cell death (Thawkar and Kaur, 2019). Interestingly, upregulation of P2X and P2Y receptors in ALS patients can subsequently lead to overproduction of TNF-α and cyclooxygenase-2 (COX2), which facilitates neurotoxicity (Liu and Wang, 2017).

Even though the neuroinflammatory cascades exacerbate neurodegenerative disease progression and participate in neuronal death, the exact mechanism of this involvement and therapeutic strategies regarding targeting neuroinflammation is not well studied. According to some investigations, targeting neuroinflammation can be a promising therapeutic approach to decrease inflammation and its further complications in neurodegenerative diseases (Figure 3). Therefore, it can improve patients’ neuronal function in mental and physical activities. Incipiently, some specific microRNAs, such as miR-155-5p, can be key regulators of inflammatory cascades in neurodegenerative diseases. Overexpression of miR-155-5p has been reported in the CSF of AD and MS patients (Lv et al., 2020). It has been recently found that Rosmarinic acid (RA) can inhibit neuroinflammation in neurodegenerative disease samples by regulating miR-155-5p, leading to attenuation in inflammation-associated neuronal damage and loss (Figure 3, Table 2; Lv et al., 2020). Calpains as non-caspase proteases participate in the execution of neuronal cell death and cooperate with key factors of neuronal cell death. Thus, targeting these proteases may result in neuroprotection. Studies have demonstrated that Alpha1-antitrypsin (A1AT) can attenuate microglial neuroinflammation as well as inhibition of calpain activity (Figure 3, Table 2; Feng et al., 2020). Other novel strategies have also been recruited with a focus on neuroinflammation. A recent study by Cheng C.-Y. et al. (2021) targeted inflammation in the substantia nigra of lipopolysaccharide (LPS)-treated rats by liposomes carrying Epigallocatechin-3-gallate (EGCG), a natural antioxidant in green tea. It demonstrated neuroprotection by inhibiting neuroinflammation (Figure 3, Table 2; Cheng C.-Y. et al., 2021). Aucubin, which showed neuroprotective effects in oxidative stress-induced neurotoxicity, can also reduce phosphorylation levels of NF-κB, JNK, p38, and ERK, leading to a decrease in the level of inflammatory factors (Figure 3, Table 2; Li Y. C. et al., 2021). Additionally, a polyphenol named Resveratrol indicated a similar effect by down-regulation of the transcription factor NF-κB in vitro (Figure 3, Table 2; Zhong et al., 2012; Zhang et al., 2017). Targeting TLR4/NF-κB signaling pathway by Hesperetin, a Citrus flavonoid, protected neurons from neuroinflammation and apoptosis (Table 2; Muhammad et al., 2019). Many other similar signaling pathways can be inhibited, aiming to alleviate neuroinflammation in the CNS (Hou et al., 2021). Furthermore, targeting inflammation-associated receptors expressed on neurons can also be an approach. For example, 15d-PGJ2 is a peroxisome proliferator-activated receptor-gamma (PPAR-γ) agonist that inhibits the production of some interleukins and suppresses inflammation in microglial cells in vitro (Figure 3, Table 2; Xu et al., 2008). Suppressing neuroinflammation by targeting P2X7R has been studied, but blood-brain barrier (BBB) permeability limits candidate drugs, so more studies are needed in this case (Thawkar and Kaur, 2019). Intriguingly, Anakinra, an IL-1 receptor antagonist (Mahboubi Mehrabani et al., 2022), reaches CNS easily and inhibits the activity of IL-1β by binding to its receptor and mitigate pyroptosis in neurons (Figure 3, Table 2; Wang et al., 2019).

As mentioned in the former section, since the NLRP3 inflammasome plays a significant role in the enhancement of neurotoxicity, inhibition of NLRP3 and its subsequent pathways might be an effective method to decrease neuroinflammation-induced cell death. MCC950 is a small-molecule NLRP3 inhibitor that has inhibited inflammasome activation in rodent PD models leading to substantial neuroprotection, mitigation in motor deficits, and accumulation of α-synuclein aggregates (Figure 3, Table 2; Gordon et al., 2018). Noteworthy, the neurotransmitter dopamine can bind to the dopamine D1 receptor, which results in ubiquitination and degradation of NLRP3 via the binding of cAMP with NLRP3, leading to the restriction of NLRP3 activation (Figure 3, Table 2; Yan et al., 2015). Some experiments revealed attenuation of NLRP3-mediated neuroinflammation in PD mouse models using peroxisome proliferator-activated receptor beta/delta (PPAR-β/δ) agonist GW501516 (Figure 3, Table 2). There were some limitations with this drug, such as the resistance of the BBB to pass the drug to reach the brain parenchyma. Hence, this compound cannot be considered a candidate for PD treatment until the problem is not solved (Chen et al., 2019). A flavonoid derived from the roots of Scutellaria baicalensis Georgi, termed “Baicalein”, indicated anti-inflammatory and anti-pyroptosis properties in animal models of PD. Experiments suggest that Baicalein may play a role in preventing the loss of DA neurons by reducing the production of various pro-inflammatory cytokines. It can also inhibit NLRP3 and caspase-1 activation, and simultaneously suppress pyroptosis by targeting Gasdermin D in 1-methyl-4phenyl-1,2,3,6-tetrahydropyridine (MPTP) Induced Mice Model of PD (Figure 3, Table 2; Rui et al., 2020). Benzyl isothiocyanate (BITC) and dihydromyricetin (DHM) are other plant-derived compounds with anti-inflammation properties (Lee et al., 2016; Feng et al., 2018). BITC seems to have the neuroprotective effects by inhibition of IL-1β release and NLRP3 inflammasome inhibition in the BV2 microglial cells (Lee et al., 2016). Treatment of APP/PS1 transgenic mice with DHM improved neuroinflammation and memory function, as a result of decreased NLRP3 inflammasome activation (Figure 3, Table 2; Feng et al., 2018). Taken together, the detrimental effects of neuroinflammation in induced neuronal cell death must not be underestimated, and more research is required to provide a better understanding of mechanisms and underlying therapeutic strategies.