Tau is a microtubule-associated protein primarily responsible for stabilizing microtubule structures in neurons and is involved in intracellular transport (Pidara and Yamada, 2022). Microtubules are a crucial component of the cytoskeleton. Tau in their normally phosphorylated state ensure the stability and function of microtubules by regulating their binding capacity to microtubules (Barbier et al., 2019). Long axons and dendrites of neurons require robust scaffolds to maintain their morphology and function, and microtubule stabilization is particularly important for preserving neuronal structure and function (Kelliher et al., 2019). Furthermore, Tau are highly expressed in axons, and their phosphorylation status plays a regulatory role in the transport of substances within axons (Iwata et al., 2019). Axonal transport involves the directed movement of various molecules and organelles from the cell body to the synaptic terminal and is essential for normal neuronal function and long-range signaling (Guedes-Dias and Holzbaur, 2019). Normally phosphorylated Tau ensure smooth and efficient transport, thereby supporting neuronal growth, maintenance, and synaptic plasticity. By regulating the dynamic balance of microtubules and axonal transport, Tau indirectly affect the efficiency of synaptic transmission and the functional integration of neural networks, which are critical for higher functions such as learning and memory (Wang and Mandelkow, 2016). Conversely, in response to cellular stress and injury, the phosphorylation state of tau changes, and these changes may contribute to cellular adaptation and survival (Guo T. et al., 2017). The roles of phosphorylated tau in stress responses include regulating cytoskeletal reorganization, modulating cellular signaling pathways, and protecting cells from excessive damage. When neurons experience trauma or stress, dynamic phosphorylation of tau can facilitate rapid cellular responses and recovery, maintaining nervous system stability (Jenkins et al., 2000). Thus, by stabilizing microtubules, regulating axonal transport, maintaining neuronal signaling, and participating in cellular stress responses, phosphorylated tau ensure the health and function of the nervous system. The synergy of these activities is crucial for proper nervous system function and the realization of higher neural functions (Figure 2A).

In AD patients, tau phosphorylation primarily occurs at serine and threonine residues. Tau phosphorylation in AD has been reported to involve multiple sites, including Ser202/Thr205, Thr217, Thr212/Ser214, Thr181, Ser396, Ser422, Ser199, Thr231, and Ser262 (Mai et al., 2022; Hart de Ruyter et al., 2023; Hines et al., 2023). Hyperphosphorylation at these sites causes detachment from microtubules and leads to the formation of insoluble NFTs. This abnormal phosphorylation is a key pathological feature of AD. Mechanisms of tau phosphorylation include imbalances in kinases and phosphatases, oxidative stress, iron accumulation, and inflammatory responses. Among these, iron accumulation plays a crucial role in the phosphorylation and aggregation of tau (Rao and Adlard, 2018). Iron, the most abundant metal ion in the brain, can directly bind to the iron-binding motif in tau. Specifically, Fe^{3 +} binding to Tau induces the hyperphosphorylation of Tau, which promotes its aggregation into paired helical filaments (PHF Tau), ultimately contributing to the formation of NFTs (Yamamoto et al., 2002). The formation of NFTs can hinder the extracellular iron efflux pathway regulated by Tau, further exacerbating the accumulation of iron. However, this effect only occurs when iron is in its oxidized state (Fe^{3 +}) and not when it is in its reduced state (Fe^{2 +}). The redox state transition of iron plays a crucial role in the aggregation of PHF Tau. Studies indicate that Fe^{3 +} can promote the aggregation of Tau, while Fe^{2 +} affects the aggregation state by inducing reversible conformational changes in Tau and influences the formation of NFTs through interaction with Thr residues under the influence of Tau phosphorylation (Rao and Adlard, 2018). Further research has shown (Wan et al., 2019) that in primary neuronal cultures from sub-iron-treated SD rats and in C57BL/6 mice fed a high-iron diet, the phosphorylation levels of Tau at the Ser202/Thr205, Thr181, and Ser396 sites increased significantly, while total Tau expression remained unchanged, further explaining the pathological mechanism between iron and PHF tau.

Amyloid precursor protein is a transmembrane glycoprotein widely found in mammalian cell membranes, with particularly high expression in neurons. There are three main isoforms of APP, which contain 695, 751, and 770 amino acids, respectively. These isoforms are present in the brain and are generated by alternative transcriptional splicing (Hefter et al., 2020). APP can be processed through two main pathways: the non-amyloid pathway and the amyloid pathway. In the non-amyloid pathway, APP is cleaved by α-secretase (ADAM10) at amino acid position 17 of the Aβ sequence, generating sAPPα and C83 fragments. Subsequently, γ-secretase cleaves the C83 fragment to generate the P3 peptide and the APP intracellular domain, yielding a fragment that is protective for neuronal cells (Cho et al., 2022). In AD, however, phosphorylation of APP alters its metabolic pathway, making it more susceptible to cleavage by β- and γ-secretase, leading to the production of Aβ. In this process, phosphorylation occurs mainly in the intracellular segment of APP, such as at the threonine (Thr688/181) site, which has been shown to increase APP’s affinity for β- and γ-secretase, facilitating the production of Aβ and the formation of insoluble amyloid plaques (Lewczuk et al., 2020; Figure 2B).

Iron accumulation has been reported to affect iron storage and distribution by disrupting the interaction between iron-regulatory proteins IRP1, IRP2, and the iron-responsive element (IRE) (Khan, 2020). Further studies showed that the 5’ untranslated region of APP mRNA contains an IRE that binds iron-regulatory proteins. Iron can bind directly to IRP, leading to its dissociation from IRE, thereby promoting APP mRNA translation and Aβ production. Binding of IRP1 and IRP2 directly enhances APP translation (Zhou and Tan, 2017). Therefore, the degree of iron accumulation directly affects APP expression and Aβ production. More importantly, APP on the cell surface not only contributes to amyloid plaque formation but also maintains iron efflux by stabilizing the ferroportin (FPN), thus playing a key role in neuronal iron homeostasis. It has been shown (Tsatsanis et al., 2019) that phosphorylation of the extracellular region of APP (especially at site Ser206) is required for APP translocation to the cell surface, and that N-glycosylation and phosphorylation affect its localization and binding to FPN. Deletion of the phosphorylation site Ser206A resulted in a significant reduction in cell surface APP levels and a decrease in total soluble APP, suggesting that phosphorylation plays an important role in APP secretion. These mutations also led to a significant increase in intracellular iron levels, suggesting an important role of APP phosphorylation in iron efflux and intracellular iron regulation. Furthermore, using a neuroblastoma cell line stably expressing APP, it was found (Tsatsanis et al., 2019) that mutations in the phosphorylation and N-glycosylation sites not only reduced the presence of APP and FPN on the cell surface but also caused intracellular iron retention. This further revealed that these mutants exhibited significant iron accumulation in a high-iron environment, suggesting that APP plays an important role in regulating intracellular iron homeostasis. In addition, tau was found to further promote the transport of APP to the cell surface, aiding in iron export (Li et al., 2015). By injecting Aβ oligomers into the mouse hippocampus, it was found that eliminating tau prevented Aβ-induced cognitive impairment, neuronal loss, and iron accumulation. Increased APP levels, tau phosphorylation, and iron accumulation persisted even after Aβ was removed, suggesting that these factors continue to contribute to AD pathology. Additionally, transgenic mice expressing APP and progerin 1 (APP/PS1) treated with high doses of iron can serve as a model for AD. High-dose iron increases APP expression and phosphorylation, enhances amyloid APP cleavage and Aβ deposition, and impairs spatial learning and memory (Kupershmidt et al., 2012; Guo et al., 2013b). In contrast, the iron chelators deferoxamine and M30 can significantly improve memory in mice by inhibiting iron-induced amyloid APP processing, shifting APP processing to the non-amyloid pathway, reducing APP protein expression and phosphorylation at the Thr668 site, and attenuating Aβ deposition (Amit et al., 2017; Stergiou et al., 2023).

Oxidative stress is a process in which excessive levels of ROS exceed the capacity of the body’s antioxidant defense mechanisms, leading to cellular damage. Although the pathological features of AD are dominated by amyloid Aβ plaques and hyperphosphorylation of tau, oxidative stress is an early event, occurring even before the formation of Aβ plaques and tau phosphorylation (Mandal et al., 2023). Excessive iron accumulation during AD pathology generates ROS via the Fenton reaction, producing highly reactive hydroxyl radicals that exacerbate oxidative stress and damage neurons and cellular structures. In an oxidative stress environment, the production of ROS exceeds the cell’s antioxidant capacity, making tau more prone to phosphorylation and aggregation. Previous studies have shown (Xie et al., 2012) that levels of the antioxidant glutathione in the hippocampus and plasma are significantly decreased in patients with AD, and that this reduction further exacerbates iron-mediated oxidative stress by lowering the cell’s antioxidant capacity. Treatment with the antioxidant ebselen was able to reverse metallotransporter protein 1-mediated iron transport and increase glutathione level (Xie et al., 2012). In addition, the initiation of signals such as cystathionine depletion and lipid peroxidation, associated with ferroptosis (iron-dependent cell death), leads to increased iron levels, making neurons more susceptible to ferroptosis, resulting in neuronal death (Lane et al., 2021). Related studies have shown (Li et al., 2022) that increased phosphorylation of nuclear factor erythroid 2-related factor 2 (Nrf2) at Ser40 promotes the release of Nrf2 from Kelch-like ECH-associated protein 1 and its translocation into the nucleus, which activates heme oxygenase-1 (HO-1) expression. This effectively reduced ferroptosis in the brains of APP/PS1 mice, thereby inhibiting Aβ-induced tau hyperphosphorylation and neurotoxic 1-42 oligomerization in HT-22 cells. However, another study found (Hui et al., 2011) that long-term overexpression of HO-1 significantly increased the release of iron from the catabolism of hemoglobin, enhanced the local iron load in the brain, and induced phosphorylation of tau at the Ser199/202/396 site, promoting the aggregation of tau in HO-1 overexpressing transgenic mice. The use of HO-1 inhibitors can reduce iron release and local iron accumulation, thereby reducing oxidative stress and inflammatory responses, and protecting neuronal function. Additionally, knockdown or pharmacological inhibition of HO-1 reduces iron accumulation and neuronal damage, and improves cognitive function in AD mouse models (Li et al., 2022). Thus, short-term or moderate activation of HO-1, an inducible enzyme that catabolizes heme to produce iron, carbon monoxide, and biliverdin/bilirubin, may be protective by reducing ferroptosis and lowering oxidative stress, whereas long-term overactivation increases iron release and accumulation, promoting neurotoxicity and tau phosphorylation.

In addition to iron-mediated pro-phosphorylation, dephosphorylation also plays an important role in the pathogenesis of AD. BCL-xL/BCL-2-associated death promoter (BAD) is a pro-apoptotic protein in the Bcl-2 family, whose activity is regulated by its phosphorylation status. Unphosphorylated BAD can bind to anti-apoptotic proteins and block their inhibitory effects on pro-apoptotic factors, leading to cell death (Deng, 2017). In contrast, when BAD is phosphorylated by specific kinases (e.g., AKT/ Protein Kinase A), it binds to 14-3-3 proteins, which are then released from the mitochondria, losing their inhibitory effect on Bcl-2 family members (Fu et al., 2000). In AD, oxidative stress and cellular stress responses lead to BAD dephosphorylation, restoring its pro-apoptotic function and promoting neuronal apoptosis. Dephosphorylated BAD binds to Bcl-2 and Bcl-xL, inhibiting their anti-apoptotic functions, leading to loss of mitochondrial membrane potential and cytochrome c release, initiating the apoptotic cascade (Zhao et al., 2021). Studies have shown (Kuperstein and Yavin, 2003) that the combined action of Aβ and Fe^{2 +} significantly reduces protein kinase C (PKC) isoforms, decreases Akt kinase activity, and reduces BAD phosphorylation, while enhancing p38 MAPK and caspase-9 and -3 activation, exhibiting pro-apoptotic features. Iron chelators can effectively block these pro-apoptotic activities. Notably, although Aβ alone enhanced PKC isoforms, Akt activity, and BAD phosphorylation, and Fe^{2 +} could transiently enhance p38 MAPK and caspase activity, neither was sufficient to cause cell death alone. However, direct inhibition of the PKC or Akt pathway enhances Aβ/Fe^{2 +} toxicity, whereas inhibition of p38 MAPK prevents cell injury and apoptosis (Kuperstein and Yavin, 2003).

In addition, iron plays an important role in the pathogenesis of AD by regulating the phosphorylation of the Ser637 site of Dynamin-related protein 1 (Drp1), which consists of several structural domains, including the GTPase domain, the intermediate domain, and the GTPase Effector Domain (GED) (Park et al., 2015; Kandimalla et al., 2016). The GTPase domain is responsible for hydrolyzing GTP to provide energy, while the intermediate and GED domains are involved in protein interactions and multimerization (Hill et al., 2017). Drp1 exists as a monomer in the cytoplasm and aggregates at the outer mitochondrial membrane during mitochondrial division, forming a loop or helix by interacting with receptor proteins (e.g., Fis1, Mff, Mid49, Mid51) at the outer mitochondrial membrane. By interacting with these receptor proteins, Drp1 forms a ring or helical multimeric structure, splitting the mitochondria using energy generated from GTP hydrolysis (Osellame et al., 2016). In addition, by regulating mitochondrial fission and fusion, Drp1 plays a key role in maintaining mitochondrial quality and function, helping cells remove damaged or malfunctioning mitochondria (Zerihun et al., 2023). The pathogenesis of AD is often accompanied by cellular stress and injury, during which Drp1 is activated and translocated to the mitochondria to promote mitochondrial fission. This fission is often accompanied by a loss of mitochondrial membrane potential and increased permeability of the outer mitochondrial membrane, leading to the release of apoptotic factors such as cytochrome c, initiating a cascade of apoptotic responses (Kim et al., 2016). Studies have shown (Xie et al., 2012; Zhang et al., 2020) that excessive iron accumulation leads to dephosphorylation of Drp1 at Ser637, increasing the activity of dephosphorylated Drp1 and inducing excessive mitochondrial fragmentation and dysfunction. In contrast, inhibiting calmodulin phosphatase activation prevents iron-induced, Drp1-dependent mitochondrial fragmentation and apoptosis, maintains the phosphorylated state of Drp1 at Ser637, and protects neurons from iron overload (Park et al., 2015; Yu et al., 2022).

The NMDAR is a key glutamate receptor and ion channel protein in the central nervous system. It is composed of three subunits: GluN1, GluN2, and GluN3, and plays a crucial role in synaptic transmission, synaptic plasticity, learning, memory, and other processes related to neuronal excitability (Yu et al., 2021). Under normal conditions, NMDARs influence calcium ion signaling by regulating synaptic strengthening and weakening mechanisms, gene expression, and neuronal survival and apoptosis (Zhang et al., 2007; Chen et al., 2014). Additionally, during neurodevelopment, NMDAR activation promotes the stabilization and refinement of dendritic spines, thus facilitating the formation of mature neural networks (Kehoe et al., 2014). It is noteworthy that NMDAR function is influenced by several factors, including magnesium blockade, phosphorylation, internalization and exocytosis, and regulatory proteins (Liu et al., 2023). Phosphorylation typically enhances receptor function, increases ion permeability, and regulates its localization on the cell membrane (Wang et al., 2014). In AD, abnormally phosphorylated tau can bind to the GK domain of PSD-95 through multivalent interactions, causing the dynamic arrest of PSD condensates and clusters, thereby affecting the dynamics and phosphorylation status of NMDA receptors. Specifically, this interaction can disrupt the tyrosine phosphorylation of the GluN2B subunit, leading to altered receptor trafficking and reduced synaptic stability. These changes further trigger calcium overload and neuroexcitotoxicity, even at normal glutamate concentrations. Moreover, persistent NMDAR overactivation, particularly involving GluN2B-containing extrasynaptic receptors, exacerbates chronic excitotoxicity, resulting in neuronal loss, dysfunction, and a decline in neural network function (Xuan et al., 2013; Miyamoto et al., 2017; Thakral et al., 2023). Additionally, NMDAR activation accelerates divalent metal transporter 1 (DMT1) mediated iron influx, with Fe^{2 +} and Ca^{2 +} entering the NMDAR channel competitively, promoting iron release from lysosomes, increasing intracellular iron levels in primary neurons, and exacerbating neuronal damage (Xu et al., 2017). Notably, a stable Fe^{2 +} level in neurons can generate ROS, activating NMDAR, increasing NMDAR-mediated Ca^{2 +}, and further promoting Ca^{2 +} generation and persistence through Ryanodine receptor-mediated calcium release, maintaining normal intracellular redox homeostasis (Sanmartín et al., 2014). Previous studies have shown (Muñoz et al., 2011) that iron deficiency or the use of iron chelating agents can significantly reduce NMDAR-mediated calcium signaling and weaken intracellular calcium signaling. This reduction in calcium signaling and NMDAR-dependent extracellular signal-regulated kinases 1 and 2 (ERK1/2) phosphorylation leads to reduced synaptic transmission in the CA1 region of hippocampal neurons, impairing long-term potentiation (LTP) and, consequently, affecting learning and memory functions.

The gamma-aminobutyric acid type A receptor (GABA_AR) is one of the most important inhibitory neurotransmitter receptors in the central nervous system and is responsible for mediating fast inhibitory synaptic transmission in the brain (Ghit et al., 2021). The GABA_AR is a ligand-gated chloride channel. When GABA binds to the receptor, the receptor channel opens, allowing chloride ions (Cl^–) to enter the cell. The proper functioning of the receptor directly affects the hyperpolarization of the neuronal membrane and is essential for maintaining the excitatory-inhibitory balance of neurons (Sallard et al., 2021). Phosphorylation is an important mechanism for regulating GABA_AR function in this process. By altering the structure, activity, and localization of the receptor, phosphorylation can rapidly and dynamically regulate inhibitory signal transmission (Saliba et al., 2012). Phosphorylation of GABA_AR typically occurs in the intracellular region and primarily involves specific serine (Ser359, Ser343, Ser383, Ser327) and threonine (Thr375) residues (Kellenberger et al., 1992; Meier and Grantyn, 2004; Houston et al., 2007; Mukherjee et al., 2011; Nakamura et al., 2020). In physiological processes, the phosphorylation of GABA_AR receptors is a key mechanism for regulating synaptic plasticity. The phosphorylation state of GABA_AR can regulate the response of neurons to inhibitory signals, affecting learning and memory processes (Nakamura et al., 2015). Under stress conditions, the phosphorylation state of GABA_AR changes, affecting neuronal excitability and the overall inhibitory tone of the brain, thereby regulating behavior and emotional responses (Maguire, 2014).

Today, benzodiazepines (BZDs) are widely used to treat anxiety, insomnia, seizures, and as anticonvulsants by binding with high affinity to the GABAAR and potentiating the effects of GABA (Furukawa et al., 2017). However, clinical reports suggest (Ettcheto et al., 2019) that long-term use of BZDs may increase the risk of AD. Additionally, chronic BZD intervention induces the overexpression and phosphorylation of the GABA_AR subunit α5 in the mouse brain and significantly upregulates the expression of lipocalin 2 (Lcn2), resulting in intracellular iron overload, increased oxidative stress levels, and cell damage or death (Furukawa et al., 2017). Additionally, knocking down Lcn2 expression in hippocampal neurons from AD mice has been shown to alleviate glial activation, memory impairment, and synaptic abnormalities in mice (Zhou et al., 2023). Diazepam administration has also been found to increase Lcn2 expression in the cerebral cortex, hippocampus, and amygdala, a process that is not accompanied by the upregulation of pro-inflammatory genes and can be reversed by the iron chelator deferoxamine mesylate (DFO) (Furukawa et al., 2017). This suggests that the phosphorylation of GABA_AR and the role of Lcn2 in the nervous system are closely related to the regulation of iron homeostasis. However, it is important to note that Lcn2, as an iron-related protein, plays a dual role in regulating intracellular iron homeostasis by binding to iron carriers. On one hand, Lcn2 binds to iron chelates, reducing the availability of free iron in the cell and thus preventing oxidative stress caused by iron overload (Devireddy et al., 2005); On the other hand, Lcn2 ensures that cells can obtain sufficient iron when needed by promoting the storage and utilization of iron ions (Xu et al., 2012). Therefore, changes in Lcn2 expression may be important in the regulation of iron homeostasis. Overall, although targeting Lcn2 expression by regulating the expression and phosphorylation state of GABA_AR may offer a means to regulate iron metabolism, the potential benefits need to be carefully weighed against the risk of neurotoxicity. In the central nervous system, the complex regulatory relationship between GABA_AR and Lcn2 requires careful consideration to ensure the safety and efficacy of long-term treatment.

In AD, abnormal phosphorylation of AKT and GSK3β leads to increased activity, which further promotes the overphosphorylation of Tau, forming NFTs, ultimately leading to the destruction of the structure and function of neurons (Zhao et al., 2024). In this process, hyperphosphorylated Tau can cause iron deposition, activate recombinant calcineurin (CaN) to dephosphorylate the Ser9 site of GSK3β, relieving the inhibition of GSK3β and leading to an increase in GSK3β activity, which in turn mediates the hyperphosphorylation of tau. At the same time, CaN further dephosphorylates the cyclic adenosine monophosphate response element binding protein, causing synaptic defects and memory impairment, creating a vicious cycle (Zuo et al., 2024). Previous studies have shown that in erastin (iron death inducer)-treated N2a cells (a cell line of mouse neuroblastoma), iron death inhibitors can significantly restore erastin-induced GSH depletion and GPX activity reduction, reduce ROS and iron levels, decrease GSK-3β activation (restore Ser9 phosphorylation levels), and ultimately reduce Tau hyperphosphorylation and aggregation (Wang et al., 2022). Zhang et al. (2015) found in an animal study that oral deferoxamine intervention can reduce iron accumulation induced by intraperitoneal injection of lipopolysaccharide (LPS) in the mouse brain by chelating iron ions, thereby reducing neuroinflammation and apoptosis. In addition, DFO further reduced LPS-induced neuroinflammation and cognitive impairment by increasing the phosphorylation of GSK3β at Ser9 and inhibiting its activity (Zhang et al., 2015). In addition to targeting iron accumulation, activation of vitamin D receptor agonists can also significantly reduce iron accumulation in the cerebral cortex and hippocampus of APP/PS1 mice by downregulating IRP2 and TfR, and reduce the phosphorylation of Tau at Ser396 and Thr181 sites by inhibiting the phosphorylation of GSK3β (Tyr216) (Wu et al., 2022).

In addition, GSK3β is active in the basal state and does not require activation by external signals. It can phosphorylate multiple substrates, including glycogen synthase, transcription factors, and other proteins. When growth factors (such as insulin, Insulin-like growth factor 1, and epidermal growth factor) bind to their corresponding receptor tyrosine kinases, the receptors self-phosphorylate and activate phosphoinositide 3-kinase (PI3K). PI3K phosphorylates the membrane phospholipid phosphatidylinositol 4,5-bisphosphate (PIP2) to generate phosphatidylinositol (3,4,5)-trisphosphate (PIP3) (Maqbool et al., 2016; Hermida et al., 2017). PIP3 acts as a second messenger to recruit AKT with a pleckstrin homology domain to the cell membrane. AKT is fully activated by phosphorylation at Thr308 by 3-phosphoinositide-dependent protein kinase 1 and at Ser473 by mechanistic target of rapamycin complex 2. The activated AKT inactivates GSK-3β by directly phosphorylating its Ser9 site. This phosphorylation alters the conformation of GSK-3β, preventing it from binding to substrates and reducing its kinase activity (Truebestein et al., 2021). In AD, iron accumulation and ROS produced by lipid peroxidation can oxidize the cysteine residues in the AKT molecule, forming disulfide bonds or sulfhydryl oxides, leading to a decrease in the phosphorylation level of AKT (Rochette et al., 2022). Studies have shown (Uranga et al., 2013) that exposing the mouse hippocampal neuronal cell line HT22 to 25–200 μM Fe2⁺ for 24 h activates the PI3K/AKT pathway, resulting in increased phosphorylation of AKT and GSK3β. It is worth noting that PI3K inhibitors and AKT inhibitors can increase iron-induced oxidative levels in HT22 cells, but GSK3β inhibitors have no significant effect, which further indicates the key role of AKT activity in inhibiting GSK3β and the production of oxidants (Uranga et al., 2013). Similarly, intervention with the iron chelator M30 in rat primary cortical cells directly enhanced the levels of phosphorylated AKT (Ser473) and phosphorylated GSK-3β (Ser9) and reduced the phosphorylation of Tau, protecting neurons from Aβ toxicity (Avramovich-Tirosh et al., 2010).

In addition, both iron accumulation and a high-fat diet can lead to cognitive deficits, and their combination can cause more severe cognitive damage. This combination also induces insulin resistance and alters insulin signaling pathways in the hippocampus, including reduced phosphorylation of AKT (do Nascimento et al., 2024). Normally, insulin binds and self-phosphorylates, leading to the activation of insulin receptor substrate 1 (IRS-1). Phosphorylated IRS-1 recruits and activates PI3K. The regulatory subunit of PI3K, PI3K p85α, binds to phosphorylated IRS-1 and activates PI3K, converting PIP2 to PIP3, which subsequently recruits and activates AKT (Kearney et al., 2021). Studies have shown (Wan et al., 2019) that ferrous chloride can reduce tyrosine phosphorylation levels of insulin receptor β (IRβ), IRS-1, and PI3K p85α in primary neurons, thereby disrupting the insulin signaling pathway and impairing learning and memory functions. Exogenous insulin supplementation can reverse the abnormal phosphorylation of tau induced by iron, restore tyrosine phosphorylation of IRβ, IRS-1, and PI3K p85α, and activate insulin signaling, thereby improving iron accumulation-induced neuronal pathology (Wan et al., 2019).

The role of the MAPK/P38/HIF-1 regulatory network is crucial in the cellular response to external stress and hypoxic environments (Guan et al., 2020). Among these, hypoxia-inducible factor-1α (HIF-1α) is a transcription factor stabilized under hypoxic conditions, and its activity is regulated by phosphorylation (Daly, 2020). Under normal physiological conditions, cell surface receptors (e.g., Receptor Tyrosine Kinases, G Protein-Coupled Receptors) autophosphorylation and activate downstream rat sarcoma proteins after binding to their corresponding ligands (e.g., growth factors, cytokines). This activates rapidly accelerated fibrosarcoma kinase, which phosphorylates and activates MAPK/ERK kinase (MEK), which in turn phosphorylates and activates ERK (extracellular signal-regulated kinase) (Zhao and Luo, 2022). In AD, oxidative stress and the inflammatory response induced by iron accumulation activate upstream mitogen-activated protein kinase kinase 3/6 (MKK3/6), which phosphorylates and activates p38 MAPK. The activated p38 MAPK translocates to the nucleus, promoting the transcription and translation of HIF-1α (Fang et al., 2017). This phosphorylation stabilizes HIF-1α and promotes the expression of hypoxia-related genes. Meanwhile, under stress conditions, the p38 MAPK and HIF-1 pathways are cross-regulated to jointly respond to hypoxia and oxidative stress.

Studies have shown that aberrantly activated HIF-1α upregulates the expression of various downstream genes, including pro-inflammatory factors and vascular endothelial growth factor (VEGF), the overexpression of which exacerbates neuroinflammation, cerebrovascular lesions, and neuronal damage (Fehsel, 2023). Additionally, HIF-1α phosphorylation plays a key role in the pathogenesis of AD through multiple mechanisms, including promoting neuroinflammation, increasing Aβ accumulation, disrupting mitochondrial function, and accelerating neuronal death (Guo C. et al., 2017). The neuroprotective effects of iron chelators in AD have been shown to involve mechanisms such as reducing ROS production and stabilizing HIF-1α expression (Fine et al., 2012). In vivo studies have shown (Guo et al., 2015) that desferrioxamine reduces Aβ deposition and synaptic loss by upregulating HIF-1α mRNA and protein levels and inducing proteins encoded by HIF-1α adaptor genes, including TfR, DMT1, and BDNF. In vitro studies have shown (Guo et al., 2015) that DFO enhances the stability and phosphorylation levels of HIF-1α through activation of the P38/HIF-1α pathway, which further reduces iron in the CA3 region of the hippocampus and has no significant effect on the cortical and CA1 regions. Similarly, M30, a multifunctional iron chelator, was shown to increase HIF-1α protein levels in various brain regions and the spinal cord of adult mice, induce the expression of its dependent target genes such as VEGF, erythropoietin, TfR, HO-1, and other proteins, and further enhance the transcriptional levels of BDNF and growth-related protein-43 (Kupershmidt et al., 2011). These findings suggest that iron chelators activate the P38 MAPK signaling pathway to upregulate HIF-1α, increase the expression of HIF-1α-dependent genes, and participate in the pathological process of AD, with the mechanism potentially promoting the redistribution of iron rather than its removal for neuroprotective effects.

As mentioned earlier, IRP1 is a key protein responsible for regulating iron metabolism in mammalian cells. IRP1 regulates the stability and translation of iron metabolism-related mRNAs (such as APP) by binding to IREs, thereby influencing intracellular iron homeostasis. In general, IRP1 exhibits high IRE-binding activity, and this enhanced activity enables IRP1 to more effectively regulate mRNAs containing IREs, thereby modulating iron uptake and storage (Calvo-López et al., 2023). In this process, PKC, a serine/threonine kinase widely involved in cell signaling and regulation of various cellular functions, affects the phosphorylation state of IRP1. Specifically, PKC can phosphorylate IRP1 at the Ser-138 site, increasing its IRE-binding activity (Brown et al., 1998; Johnson et al., 2017). Additionally, the phosphorylation state of PKC affects the stability of the iron-sulfur (Fe-S) cluster of IRP1, making it more susceptible to oxidative stress. This, in turn, alters its dissociation from cytosolic aconitase to an IRE-binding protein, indirectly affecting the expression of TfR and ferritin (Brown et al., 1998). Several studies have shown (Alkon et al., 2007; Lu et al., 2022) that PKC phosphorylation levels are closely related to AD, and that abnormal Aβ accumulation and stress can regulate PKC activity. In addition, PKC overactivation can affect APP cleavage and rebalance the APP production pathway. Furthermore, the levels of PKC isoenzymes in nerve cells treated with Aβ and Fe^{2 +} were significantly increased. Both Ca^{2 +}-dependent and non-dependent isoenzymes showed a significant increase within 4 h, while PKC levels significantly decreased after 12 h (Kuperstein and Yavin, 2003). This may indicate that prolonged exposure to Aβ and Fe^{2 +} leads to PKC depletion or negative feedback regulation. Further studies have shown that treating cortical primary cells from adult rats with the iron chelator M30 enhances PKC phosphorylation and upregulates neuroprotective adaptive mechanisms in the brain.