Introduction

Alzheimer’s disease (AD) is a common neurodegenerative disease characterized by cognitive dysfunction that has become a major public health challenge, seriously impacting the quality of life and placing a considerable burden on the families of older individuals [1]. In 2018, Alzheimer’s Disease International estimated a dementia prevalence of approximately 50 million individuals worldwide, which is projected to triple by 2050, with two-thirds living in low-income and middle-income countries [2]. Diabetes-associated cognitive dysfunction (DACD) is a complex complication of diabetes occurring in the central nervous system [3]. Extensive epidemiological studies have revealed a close association between diabetes and the increased risk of cognitive dysfunction and AD. Individuals with type 2 diabetes mellitus (T2DM) have an approximately 60% higher risk of developing dementia than those without T2DM [4,5,6]. This risk of cognitive impairment and substantial dementia is further increased in individuals aged ≥ 60 years with diabetes [7]. Moreover, females with T2DM were found to have a higher risk of accelerated cognitive decline than males [7]. Both AD and DACD have been associated with cognitive impairment in the clinics, with some overlapping pathways identified in the underlying pathogenesis, such as insulin resistance, chronic inflammation, oxidative stress, loss of neurons and synapses, accumulation of β-amyloid (Aβ) plaques and hyperphosphorylation of Tau protein, and a metal ion imbalance; hence, AD has been considered “type 3 diabetes” [8, 9]. In the early stages of AD, a decrease in glucose utilization and energy metabolism in neurons is related to damage to the insulin signaling pathway [10, 11]. Recent studies have suggested that disorders in insulin signal transduction can cause neuronal damage. Insulin can regulate the metabolism of Aβ precursor protein and exerts a neuroprotective effect, while insulin receptor sensitizer can improve cognitive and learning functions [12, 13]. Therefore, diabetes-related cognitive impairment and AD exhibit several similarities, and there is considerable overlap and interaction in behavioral, morphological, and biochemical changes in the hippocampus, as well as pathological characteristics. However, the mechanisms underlying cognitive impairment related to AD and diabetes have not been fully elucidated. Ferroptosis is an erastin-triggered oxidative, non-apoptotic form of cell death that was discovered in 2012. Ferroptosis is morphologically, biochemically, and genetically distinct from apoptosis, various forms of necrosis, and autophagy and is controlled by multiple metabolic pathways [14]. Ferroptosis-inducing factors can directly or indirectly affect glutathione (GSH) peroxidase (GSH-Px) via different channels, resulting in a decline in cell antioxidant capacity and the accumulation of lipid reactive oxygen species (ROS), which eventually leads to the death of oxidized cells [15]. According to a growing body of evidence, there exists a link between ferroptosis and cognitive impairment associated with diabetes and AD [16, 17]. Interventions targeting the occurrence and development of cognitive impairment associated with diabetes and AD by regulating ferroptosis have gained momentum in etiological research and treatment.

The Regulatory Mechanism of Ferroptosis

Ferroptosis is primarily induced by lipid peroxidation, accompanied by the disruption of iron homeostasis, impaired antioxidant defense, and accumulation of ROS, leading to lipid peroxidation within the cell [14]. In recent years, there has been rapid progress in mechanistic studies of ferroptosis. Current research indicates that iron metabolism, oxidative stress, and lipid peroxidation are predominant features of ferroptosis (Fig. 1).

Fig. 1
figure 1

The regulatory mechanism of ferroptosis. The mechanisms involved in ferroptosis regulation mainly involve three aspects: iron metabolism, oxidative stress, and imbalance of lipid peroxidation defense systems. Firstly, the interruption of iron metabolism is an important driving factor of ferroptosis. Extracellular Fe3 + combines with TF to form TF-Fe3 + complex. This complex binds to TfR 1 and is transported into the cell. With the cell, Fe3 + is reduced to Fe2 + . The accumulation of Fe2 + promotes the Fenton reaction, produces hydroxyl radicals (-OH), contributes to the formation of LIP, and increases ROS production, thus inducing oxidative stress and accelerating lipid peroxidation. Secondly, the synthesis and utilization of GSH play a key role in inducing ferroptosis. Cysteine and glutamate are exchanged via System Xc- on the cell membrane to maintain GSH production. GPX4 catalyzes the conversion of GSH to its oxidized form (GSSG), preventing the accumulation of harmful lipid peroxidation products. Finally, PUFAs are transformed into PL-PUFA via the actions of ACSL4. Within the cell membrane, PL-PUFA undergoes oxidative reactions mediated by LOX-induced free radicals, ultimately resulting in the generation of PL-PUFA-OOH, thereby triggering ferroptosis. Abbreviations: ROS, reactive oxygen species; GSH, glutathione; GPX4, Glutathione Peroxidase 4; GSSG, oxidized glutathione; Xc, cystine/glutamate; TfR1, transferrin receptor 1; SLC3A2, solute carrier family 3 membrane 2; SLC7A11, solute carrier family 7 membrane 11; PUFA, polyunsaturated fatty acid

Full size image

Iron Metabolism

Disturbed iron metabolism is a crucial driver of ferroptosis and is one of its classic hallmarks [18]. Iron, an essential trace element in the human body, participates in numerous fundamental metabolic pathways and life processes, including oxygen transport, mitochondrial electron transport chain, energy metabolism, and nucleotide synthesis [19]. In the presence of an imbalance in the input, storage, and output of intracellular iron, the cell activates iron metabolic mechanisms to maintain the balance between iron uptake, transport, and utilization. The persistence of iron metabolic disorders may affect cell susceptibility to iron-induced cell death [20]. Iron imbalance can occur by disrupting mitochondrial electron transport chain activity, leading to an increase in ROS-mediated oxidative stress. Mitochondria, which are characterized by their iron-rich content and primary role in ROS generation, are considered crucial sites for ferroptosis. During ferroptosis, mitochondria typically exhibit contraction, reduction, or loss of cristae, along with increased membrane density [21]. Mitochondria not only serve as the primary sites for intracellular ROS generation but also provide specific lipid precursors for cellular ferroptosis through their fatty acid metabolism.

Proteins involved in iron metabolism, such as transferrin (TF), transferrin receptor (TfR), and divalent metal transporter 1 (DMT1), play crucial regulatory roles in ferroptosis. TF is a major protein responsible for iron transport, while the mobilization and upregulation of TfR1 have recently been reported as potential markers of ferroptosis [22]. DMT1 is the primary transmembrane transporter for Fe2+ entry into cells and plays a crucial role in Fe2+ release from endosomes during the TF cycle. When bound to TF, extracellular Fe3+ forms the TF-Fe3+ complex, which binds to TfR1 and is transported into the cells. Within the cell, metalloreductase STEAP3 reduces Fe3+ to Fe2+. Fe2+ is then transported into the cytoplasm via DMT1, forming the labile iron pool (LIP), an unstable pool of Fe2+. Fe2+ accumulation in LIP can specifically increase oxidative stress levels. The Fenton reaction leads to the generation of iron-mediated ROS [23]. ROS serve as the foundation for maintaining cellular homeostasis and contribute to immune responses, inflammation, synaptic plasticity, learning, and memory. However, excessive ROS can result in oxidative stress. ROS can react with polyunsaturated fatty acids (PUFAs) in the lipid membrane, inducing lipid peroxidation and ultimately resulting in ferroptosis [24].

Metabolism of System Xc-GSH-GSH-Px 4 (GPX4)

The synthesis and utilization of GSH constitute crucial pathways for ferroptosis induction. Inhibition of System Xc- to reduce GSH levels and prevent GPX4 inactivation are key driving factors in ferroptosis. In genetics, it has been demonstrated that GSH synthesis, System Xc- activity, and GPX4 collectively operate within cells to establish a comprehensive protective network [25]. This network serves to safeguard cells against death induced by various oxidative stress conditions [25]. System Xc- is an important antioxidant system that is widely distributed in phospholipid bilayers and comprises a heterodimer of the functional subunit SLC7A11 and the auxiliary subunit SLC3A2 [26]. Cysteine and glutamate undergo a 1:1 exchange at the cell membrane surface via System Xc. The absorbed cystine is then reduced to cysteine within the cells. Subsequently, the synthesis of GSH is catalyzed by the enzymes glutamate cysteine ligase (GCL) and glutathione synthetase (GSS). GPX4 is a lipid repair and lipid hydroperoxide detoxification enzyme that is essential for cell survival and is the core regulatory protein of ferroptosis [27]. GPX4 catalyzes the conversion of GSH to oxidized GSH (GSSG). Additionally, it transforms the peroxyl bond (L-OOH) of lipid peroxidation into a hydroxyl (L-OH) bond, thereby preventing the accumulation of harmful lipid peroxidation products and preserving the structure and function of the cell membrane. Consequently, activation of GPX4 hinders the build-up of lipid peroxides, consequently functioning as a defense against cellular ferroptosis [28].

Lipid Peroxidation

Unrestrained lipid peroxidation is a pivotal factor in ferroptosis induction [29]. PUFAs present in membrane phospholipids are susceptible to attack by ROS, leading to an excessive oxidative-reductive reaction, which is collectively referred to as lipid peroxidation [30]. Lipid hydroperoxides within lipid peroxides can disrupt the lipid bilayer of the membrane, thereby enhancing lipid oxidation and leading to ferroptosis [31]. Although the mechanisms triggering lipid peroxidation in ferroptosis are yet to be comprehensively elucidated, current research suggests that lipid peroxides can be generated during ferroptosis through two main pathways [32]: firstly, iron catalyzes lipid peroxidation through the Fenton reaction, producing ROS, and secondly, the esterification and oxidation of PUFA.

Enzymes involved in lipid metabolism, such as acyl-CoA synthetase long-chain family member 4 (ACSL4), lysophosphatidylcholine acyltransferase 3 (LPCAT3), and lipoxygenase (LOX), play crucial roles in the lipid peroxidation associated with ferroptosis, serving as notable factors contributing to ferroptosis. Among these enzymes, LOX is iron-dependent. As important constituents of phospholipids in neuronal membranes, PUFAs possess a variety of membrane functions, such as signal transduction, regulation of ion channels, and activity of membrane-bound proteins [33, 34]. PUFAs are converted into PUFA-containing phospholipids (PL-PUFA) via the action of ACSL4 and LPCAT3. Within the cell membrane, PL-PUFAs undergo oxidative reactions mediated by LOX-induced free radicals, ultimately generating PUFA-containing phospholipid hydroperoxides (PL-PUFA-OOH) and triggering ferroptosis [35]. Although it has been established that the regulation of GPX4 and GSH metabolism can prevent ferroptosis, the specific lipid peroxidation events that drive ferroptosis require further clarification.

Ferroptosis in AD

Brain Iron Metabolism Is Associated with Cognitive Impairment in AD

Excessive accumulation of iron ions increases the permeability of the blood–brain barrier, induces inflammation, affects the redistribution of iron ions in the brain, and consequently alters brain iron metabolism [36]. Iron metabolism in the brain is associated with cognitive impairments in patients with AD. Recent clinical studies utilizing magnetic resonance imaging (MRI) technology indicated a consistent correlation between cognitive impairment and iron deposition, primarily in the hippocampus, cortical regions, and basal ganglia [37]. In the preclinical stages of AD, iron levels were found to be substantially elevated in the hippocampus, cortex, and cerebellar neurons. In addition, many studies have observed substantial variations in iron levels across different brain regions, with lower levels of iron in white matter regions, such as the frontal lobe, whereas gray matter exhibits the highest iron content in the brain [38]. A clinical cohort study found that cortical iron may drive cognitive decline in patients with AD by inducing oxidative stress and ferroptosis [39]. Another cohort study utilizing the Alzheimer’s Disease Neuroimaging Initiative (ADNI) database suggested that cerebrospinal fluid ferritin levels, which reflect brain iron loading, are associated with preclinical cognitive decline in individuals carrying the APOE-ε4 allele [40]. Hepcidin and ferroportin (Fpn) are present in normal human and mouse brains. Nonetheless, concentrations of hepcidin and Fpn are substantially reduced in the brains of individuals with AD and the later developmental stages of the amyloid precursor protein (APP) transgenic mouse model [41]. Fpn1, a non-heme iron exporter produced during ferroptosis, is abnormally reduced in AD. Downregulation of Fpn1 leads to reduced iron export and elevated iron levels intracellularly, resulting in ferroptosis and memory impairment in APPswe/PS1dE9 mice. Furthermore, specific ferroptosis inhibitors, administered both in vivo and in vitro, were found to effectively reduce Aβ aggregation-induced neuronal death and memory deficits [42]. Drug therapy or electroacupuncture adjuvant therapy for AD was shown to effectively suppress ferroptosis and iron metabolism disorders, as well as substantially reduce iron metabolism-related proteins TfR1, ferritin heavy chain 1 (FTH1), FTL, and FPN1 [43,44,45,46].

Considering the aforementioned studies, oxidative stress and ROS generation induced by iron metabolism disorders may be associated with AD pathology. Interestingly, there is no reported evidence of an increased incidence of neurodegeneration or heightened brain iron concentration in peripheral iron-loading disorders, such as thalassemia and hemochromatosis, despite high iron accumulation in parenchymal tissues. Therefore, we believe that the brain represents a privileged body compartment in which peripheral iron does not interfere with iron in the brain, and under normal conditions, it does not respond to peripheral changes in iron status. The cognitive impairment observed in AD may arise from changes in brain iron levels from a state of rest to a neurotoxic state.

Interaction of Brain Iron Metabolism with Aβ and Tau Protein

Abnormal regulation of iron ions was shown to lead to neuronal oxidative stress, which is associated with the production of Aβ-associated misfolding by APP, tau protein aggregation, and hyperphosphorylation processes [47]. A clinical study utilizing iron-based MRI contrast agents discovered that increased levels of iron in the frontal cortex of individuals with AD were related to the quantity of Aβ plaques and tau protein lesions [48]. Additionally, ROS production during ferroptosis can induce Aβ aggregation by promoting APP misprocessing in both APPwt and APPsw cell models [49]. Likewise, in an APP transgenic mouse model of AD, astrocytes were capable of increasing APP and β-secretase expression under high iron conditions. Increased iron concentration is accompanied by Aβ aggregation, and iron deposition is colocalized with Aβ plaques [50]. Moreover, elevated levels of total brain iron reportedly coincide with early plaque formation [51]. Abnormal iron metabolism not only promotes Aβ production and aggregation but also Aβ toxicity. An experiment conducted in a fly model carrying UAS-Aβ transgene revealed that iron can alter the Aβ structure, which leads to the production of disordered protofibrils, promoting Aβ toxicity. However, iron chelation exerts a protective effect on the flies, mitigating the deleterious effects of Aβ. Accordingly, the study highlights the importance of iron metabolism as a cofactor that mediates Aβ toxicity [40]. It is noteworthy that excessive iron in the cytoplasm promotes Aβ deposition, and the aggregation of Aβ, in turn, can reduce Fe3+ to Fe2+, promoting iron overload within cells. This process generates hydroxyl radicals (OH), which promote lipid peroxidation and lead to ferroptosis, thereby creating a vicious cycle where iron accumulation and Aβ aggregation reinforce each other, contributing to the pathological progression in conditions like AD. In APP-transgenic mice, APP was found to exhibit iron oxidase activity mediated by a conserved H-ferritin-like active site, which promotes iron export and interacts with iron transport proteins to prevent iron accumulation and oxidative stress within the body [52]. ROS generated by iron overload may lead to the formation of oligomeric tau via Cys-Cys binding or kinase pathways. Iron can also generate -OH via the Fenton reaction, leading to aberrant tau phosphorylation in primary cultured neurons, which can lead to the accumulation of neurofibrillary tangles [53].

These findings suggest that the major pathological features of AD, including Aβ plaques, tau phosphorylation, and neurofibrillary tangles, are all associated with ferroptosis. However, whether iron and iron proteins in the process of ferroptosis act upstream by triggering Aβ aggregation through erroneous APP processing or directly enhance Aβ toxicity remains to be further elucidated.

Ferroptosis Mediates Oxidative Stress and Lipid Peroxidation in AD

Elevated oxidative stress and lipid peroxidation have been identified as driving factors of ferroptosis. Chen et al. reported increased lipid peroxidation in symptomatic 5xFAD mice, accompanied by elevated levels of the lysophospholipids lysophosphatidylcholine and lysophosphatidylethanolamine, using a lipidomic approach to determine the lipid profile of ferroptosis [54]. Recent studies have found that microglia in the white matter of the brain of patients with AD, after phagocytosis of damaged myelin debris, are in turn damaged by iron-rich myelin debris and undergo ferroptosis, characterized by lipid peroxidation damage and mitochondrial oxidative stress. This can adversely affect the integrity of the brain’s white matter and myelin sheath formation, ultimately resulting in cognitive impairment [55]. Ferroptosis is associated with excessive ROS generation. NADPH oxidase 4 (NOX4), a primary source of ROS, is substantially elevated in astrocytes of the cerebral cortex of patients with AD and APP/PS1 double-transgenic AD mouse models. Simultaneously, levels of the ferroptosis marker 4-hydroxynonenal (4-HNE) were found to be notably increased. Elevated levels of 4-HNE damage human astrocytes, whereas NOX4 promotes ferroptosis in astrocytes through lipid peroxidation, resulting in cognitive dysfunction in AD [56]. Furthermore, GSH is a crucial thiol antioxidant that eliminates lipid peroxides through the action of GSH-Px, thereby inhibiting the occurrence of ferroptosis. Reportedly, GSH, which plays an important role in ROS detoxification, is markedly reduced in the astrocytes and microglia of AD brains, leading to elevated oxidative stress and lipid peroxidation. These findings reinforce the crucial role of ferroptosis-mediated oxidative stress and lipid peroxidation in the pathogenesis and progression of AD.

Ferroptosis in DACD

Direct Evidence of Iron Deposition and Ferroptosis in DACD

Iron homeostasis plays a crucial physiological role in the central nervous system and is essential for brain health and development. Iron overload is a key trigger of ferroptosis. Recent research has provided direct evidence indicating the involvement of ferroptosis in the cognitive dysfunction associated with diabetes mellitus. According to a clinical cross-sectional study using Quantitative Susceptibility Mapping detection technology, patients with T2DM DACD (TDACD) exhibit substantially higher iron deposition in the hippocampus, bimaxillary caudate nucleus, left putamen, and right substantia nigra than normal individuals [57]. In a DACD model established in db/db mice, iron accumulation was detected in the brain, along with upregulated levels of TfR protein and downregulated expression of Fpn1 and FTH proteins [58]. Likewise, the hippocampus of a rat model of DACD exhibited a considerable increase in Fe2+ content. SLC40A1, which is responsible for transporting iron and is one of the genes associated with ferroptosis, was substantially downregulated. Concurrently, ferroptosis-related indicators, such as malondialdehyde (MDA), ROS, and lipid peroxidation, were also notably elevated [59]. More recently, ferroptosis in hippocampal neurons was identified as a key pathological mechanism and target for DACD prevention in a mouse model of T2DM induced by a high-fat diet combined with streptozotocin and a primary hippocampal neuron model treated with high glucose plus palmitic acid. Moreover, specific ferroptosis inhibitors effectively improved mitochondrial function, synaptic function, and cognitive deficits in mouse hippocampal neurons [60]. Conversely, T2DM is known to be strongly associated with an increased risk of cognitive dysfunction. High glucose levels can lead to ferroptosis by mediating mitochondrial oxidative stress [61]. Furthermore, excess iron catalyzes the Fenton reaction to produce excess ROS, which, in turn, leads to GSH depletion and lipid peroxidation, thereby exacerbating insulin resistance in diabetic mice [62]. Thus, the occurrence of ferroptosis may lead to pancreatic β-cell damage and dysfunction, contributing to the disruption of insulin signaling pathways and thereby establishing a vicious circle [53, 63]. The bidirectional “vicious circle” between ferroptosis and islet function may play an important role in the development of DACD. These studies reinforce the hypothesis that ferroptosis contributes to the development of DACD.

Ferroptosis Mediates Oxidative Stress and Lipid Peroxidation in DACD

ROS generated by mitochondrial ferroptosis induces cellular oxidative stress, leading to late-stage complications of T2DM through the inflammatory signaling pathway [64]. GPX4-mediated lipid peroxidation-induced ferroptosis is likely to be causally linked to DACD in a mouse model of TDACD and a high glucose-stimulated PC12 cell model. Additionally, erythropoietin was shown to mitigate neuronal ferroptosis by regulating ferroptosis-associated proteins, inhibiting lipid peroxidation, and reducing iron overload, thereby alleviating DACD symptoms [17].

Furthermore, studies suggest that ferroptosis in hippocampal neurons often manifests as iron deposition; imbalances in ferroptosis-related proteins including TF, GPX4, FTH1, and cystine/glutamate antiporter (xCT); and oxidative stress damage and disrupted lipid peroxidation. Suppression of lipid peroxidation, which is facilitated by hippocampal neuronal ferroptosis, was shown to successfully mitigate cognitive impairment [65]. In a mouse model of DACD, the levels of lipid peroxidation products (MDA) and ROS mediated by ferroptosis were substantially elevated in the hippocampus, whereas antioxidants such as superoxide dismutase (SOD) and GSH-Px and the antioxidant enzyme SOD2 were markedly reduced [66]. Liraglutide reportedly mitigates ferroptosis by reducing oxidative stress, lipid peroxidation, and iron overload. This reduces ferroptotic damage to hippocampal neurons and synaptic plasticity, delaying the occurrence and development of DACD [66]. These findings suggest that a better understanding of the emergence and reactions of ferroptosis-mediated oxidative stress and lipid peroxidation in DACD may aid in identifying novel therapeutic targets.

The Ferroptosis Pathway in AD and DACD

The Nuclear Factor Erythroid 2-Related Factor 2 (Nrf2)/GPX4 Pathway

The Nrf2/GPX4 axis is a crucial pathway that regulates oxidative stress, damage, and lipid peroxidation within the body. Nrf2, a transcription factor, is a key regulator of cellular antioxidant response. Several proteins and enzymes responsible for preventing lipid peroxidation are NRF2 target genes that prevent lipid peroxidation and free iron accumulation, thereby evading the initiation of ferroptosis [67]. GPX4 is a GSH-Px selenoprotein that specializes in ferroptosis and requires GSH as a substrate for lipid repair activities. GPX4 prevents ferroptosis by directly reducing phospholipids and cholesterol hydroperoxides into benign lipid alcohols [27]. Thus, reduced GPX4 activity can hinder the accumulation and peroxidation of lipid peroxides, resulting in ferroptosis [68]. In AD, the effect of ferroptosis on cognitive function via the Nrf2/GPX4 axis is mediated by modulating oxidative stress damage and lipid peroxidation in vivo. During ferroptosis in the brain, the loss of GPX4 not only upregulated protein levels of β-secretase but also substantially enhanced Aβ levels and amyloid plaque deposition [69]. Furthermore, treatment with forsythoside A (FA) reportedly suppressed p-Fyn expression and upregulated the expression of GPX2, Nrf1, and downstream proteins in the brain of male APP/PS3 double-transgenic AD mice, thereby activating the Nrf2/GPX4 axis to prevent erastin-induced ferroptosis and improve cognitive impairment in AD [70]. Similarly, the activation of the Nrf2/GPX4 axis was shown to inhibit ferroptosis and delay the progression of DACD. The occurrence of ferroptosis has been verified in the db/db mouse model of DACD, and dendrobine administration effectively suppressed ferroptosis, with a notable decrease in brain iron levels and an increase in the expression of Nrf2, GPX4, heme oxygenase-1 (HO-1), and quinone oxidoreductase 1 (NQO1) proteins [58].

The Nrf2/HO-1 Pathway

The Nrf2/HO-1 pathway plays an essential regulatory role in ferroptosis. Nrf2 triggers a series of events via distinct mechanisms, including protein stability, phosphorylation, and nuclear-cytoplasmic shuttling. Ultimately, Nrf2 protects against oxidative damage and plays an important role in GSH anabolism and the regulation of iron export and storage. Elevated HO-1 expression is thought to be an antioxidant protective mechanism that can lead to ferroptosis by inducing lipid peroxidation [71]. Under physiological conditions, HO-1 contains binding sites for Nrf2, forming a positive feedback loop that leads to self-enhancement and maintenance of antioxidant capacity [72]. The administration of eriodictyol inhibited ferroptosis in neuronal cells via vitamin D receptor-mediated activation of the Nrf2/HO-1 signaling pathway, suppressed Aβ aggregation and Tau phosphorylation in the brains of APPswe/PS1E9 transgenic mice, and alleviated cognitive impairment in AD [73]. Studies on DACD have revealed that ferroptosis mediates the Nrf2/HO-1 pathway to improve cognitive dysfunction. In vitro and in vivo experiments have reported that cycloheximide can reduce ferroptosis in hippocampal neurons by activating Nrf2/HO-1 signaling through increased epidermal growth factor (EGF) expression, exerting neuroprotective effects and alleviating TDACD rats [74]. Notably, HO-1 is an important source of intracellular iron. However, the integrated regulation of the iron content and antioxidant activity by HO-1 during ferroptosis remains controversial. Therefore, further research is needed to elucidate the role of the Nrf2/HO-1 pathway-mediated ferroptosis in AD and DACD.

The AMP-Activated Protein Kinase (AMPK) Pathway

AMPKs are a group of serine/threonine protein kinases that function as cellular energy sensors responsible for maintaining metabolic homeostasis. Previous studies found that abnormalities in the AMPK signaling pathway are closely associated with the development of AD, including tau protein hyperphosphorylation, inflammatory factor release, synapse loss, and apoptosis [75, 76]. AMPK can also regulate ferroptosis by mediating the phosphorylation of acetyl coenzyme A carboxylase (ACC) and biosynthesis of PUFAs [77]. A recent study reported that ferroptosis, mediated by the AMPK signaling pathway, was markedly enhanced in TDACD model mice. This discovery presents a new objective and strategy for the detection and management of DACD [17]. Ferroptosis reportedly occurs in hippocampal neurons but not in microglia or astrocytes in DACD. Moreover, AMPK activation improves DACD by hindering iron accumulation, enhancing TfR levels, and reducing ferritin, GPX4, and SLC7A11 levels in hippocampal neurons [78]. Additionally, the activation of inflammation via the AMPK pathway may be involved in ferroptosis [79]. High levels of ROS are implicated in the AMPK signaling pathway, thereby inducing ferroptosis [80]. These studies suggest that further exploration of the role of the ferroptosis-mediated AMPK signaling pathway in regulating DACD may provide novel targets for therapeutic interventions.

In summary, the Nrf2/GPX4, Nrf2/HO-1, and AMPK signaling pathways mediate ferroptosis in AD and DACD models (Table 1 and Fig. 2). The activation of these pathways reduces iron deposition and lipid peroxidation, thereby inhibiting ferroptosis and alleviating cognitive impairment in AD and DACD.

Table 1 Pathways of ferroptosis in AD and DACD
Full size table
Fig. 2
figure 2

Schematic illustration of potential action pathways of ferroptosis in AD and DACD

Full size image

The Use of Natural Products and Ferroptosis Inhibitors in the Regulation of Ferroptosis in AD and DACD

The Regulation of Ferroptosis by Natural Products in the Treatment of AD

Currently, no effective therapeutic agents are available to treat AD. Based on the role of ferroptosis in the development of AD, key molecules involved in the regulation of ferroptosis present potential new targets for treating AD. A growing number of drugs have been found to exert therapeutic effects by inhibiting ferroptosis, and certain natural products, such as FA, Rhodiola glycosides, and ginkgolide B, which suppress ferroptosis, reportedly alleviate cognitive deficits linked to AD (Table 2).

Table 2 The application of natural products in regulation of ferroptosis in AD and DACD
Full size table

FA

FA, an extract derived from the dried fruits of the Oleaceae family plant Forsythia suspensa, reportedly exhibits notable biological activities, including anti-inflammatory, antioxidant, and neuroprotective effects. FA has been shown to exert neuroprotective effects against Aβ-induced PC12 cell damage by reducing oxidative stress and neuroinflammation, altering the cholinergic system [82]. In APP/PS1 mice, forsythiaside B attenuated the activation of microglia and astrocytes in the cerebral cortex and hippocampus by decreasing Aβ deposition and phosphorylation of tau proteins, thereby ameliorating cognitive decline [83]. In addition, the most recent preclinical studies have reported that treatment with FA can substantially enhance the expression of GPX1 in the mesocortex of APP/PS4 mice; upregulate the expression of GPX2, Nrf1, and downstream proteins in the brains of APP/PS3 mice; effectively suppress the expression of TFRC and DMT1 in the brains of APP/PS1 mice; and upregulate the expression of FTH and FTL, thereby suggesting that FA can regulate ferroptosis by targeting the activation of the Nrf2/GPX4 axis to regulate ferroptosis-mediated neuroinflammation, thus exerting anti-AD properties [70].

Rhodiola Glycoside

Rhodiola glycosides are naturally occurring phenolic compounds extracted from Rhodiola rosea. It is well established that Rhodiola glycosides not only regulate metabolic disorders through AMPK signaling transduction but also exert effects by modulating multiple targets via various synergistic pathways [84]. These glycosides increase levels of catalase, GSH-Px, and SOD, thereby reducing the production of ROS and MDA, both oxidative stress products that mediate the occurrence and progression of ferroptosis, and alleviating cellular oxidative stress [85]. In an AD model, Rhodiola glycosides exerted an inhibitory effect on ferroptosis, thereby providing neuroprotection. Both in vitro and in vivo studies have revealed that Rhodiola glycosides can reduce lipid peroxidation and ROS levels, increase the expression of GPX4 and SLC7A11 proteins, and thus inhibit Aβ-induced cognitive impairment [86]. In an AD mouse model in SAMP8 mice, a 12-week Rhodiola glycoside treatment improved mitochondrial, iron, and lipid metabolism in the brain. Additionally, Rhodiola glycoside dose-dependently suppressed ferroptosis, reduced the accumulation of Aβ plaques, restored neuronal damage, and alleviated cognitive impairment [87]. These mechanisms are associated with the activation of the Nrf2/GPX4 signaling pathway. Rhodiola glycosides is a safe compound with no substantial adverse effects identified in preclinical and clinical trials, suggesting that it can be employed for targeting ferroptosis in AD therapy.

Ginkgolide B (GB)

GB is a bioactive terpene lactone that is extracted from the leaves and bark of Ginkgo biloba. GB exhibits various pharmacological activities, including antiplatelet aggregation, anti-inflammatory, antioxidant, anti-shock, and free radical-scavenging effects. GB was shown to exert neuroprotective effects in various in vivo and in vitro models of AD [88]. Moreover, recent studies have revealed that GB-induced neuroprotective effects are mediated via multiple mechanisms. GB was found to suppress the expression of TfR1, reducing levels of intracellular iron and ROS. Simultaneously, GB promoted the expression of the ferroptosis marker GPX4 and FTH1, a crucial iron storage protein [89]. Additionally, GB disrupted the interaction between nuclear receptor coactivator 4 (NCOA4) and FTH1, thereby inhibiting ferroptosis. This mechanism contributes to the alleviation of cognitive impairment caused by cerebral ischemia–reperfusion [90]. The most direct evidence has been obtained from studies undertaken in SAMP8 mice, indicating that GB can attenuate oxidative stress and inflammation and mitigate cognitive deficits mediated by GPX4-induced ferroptosis [91]. Overall, these findings suggest that GB exerts a neuroprotective role by inhibiting ferroptosis, attenuating oxidative stress, and suppressing neuroinflammation.

Ginsenosides

Pharmacological studies have indicated that ginseng and its extract, ginsenosides, exhibit favorable regulatory effects on iron metabolism, reducing cellular damage caused by iron overload. Ginsenosides reportedly activate the AMPK pathway to protect cells from iron-induced ROS production and mitochondrial damage [92]. Ginseng may also be involved in the regulation of iron metabolism in vivo by inducing estrogen-like effects [93]. Ginsenosides were found to not only reduce Aβ formation but also inhibit the neurotoxicity of Aβ [94, 95]. In addition, ginsenoside Rg2 can alleviate Aβ deposition-induced cognitive impairment in rats by activating the phosphatidylinositol 3-kinase/protein kinase B (PI3K/Akt) signaling pathway [96, 97]. More recently, ginsenoside RK1 was shown to regulate the phosphorylation of AMPK and its downstream target Nrf2, which optimizes mitochondrial membrane potential, reduces ROS levels, inhibits ferroptosis, and mitigates AD-like pathology, as demonstrated in the Aβ1-42 induced cellular injury model and the transgenic APP/PS1 mouse AD model [98]. Collectively, these findings suggest that ginsenosides are potential drug candidates for the treatment of AD via ferroptosis inhibition.

In conclusion, the identification of natural compounds with potential anti-ferroptotic properties as novel inhibitors of ferroptosis, along with the elucidation of their mechanisms of action, establishes a foundation for their future application as therapeutic agents for AD. Despite the optimistic research outcomes, the precise targets of these compounds warrant further validation and in-depth investigations.

The Regulation of Ferroptosis by Natural Products to Treat DACD

An important role of ferroptosis in the development and treatment of DACD is emerging. The natural products dendrobine, quercetin, and synephrine have been found to mediate ferroptosis, with some alleviating effects in DACD (Table 2).

Dendrobine

Dendrobine is a pyrrolizidine derivative alkaloid obtained through the extraction and isolation of Dendrobium stems from the Orchidaceae family. Dendrobine is known to elicit various health benefits in humans, including antioxidant, anti-inflammatory, antibacterial, anti-apoptotic, and anti-aging activities [99, 100]. Dendrobine not only improves mitochondrial function and reduces intracellular ROS production but also functions as a natural agonist of Nrf2, thereby increasing SOD, CAT, and GSH-Px levels and attenuating oxidative damage by inhibiting ferroptosis [101]. In a mouse model of AD, the administration of dendrobine restored iron metabolism in the brain tissue to normal levels. Furthermore, dendrobine substantially reduced iron levels by upregulating FTH proteins responsible for iron storage, Fpn1 proteins that regulate iron efflux, and downregulating TfR proteins that mediate iron uptake. Furthermore, dendrobine intervention effectively corrected ROS and mitochondrial membrane potential abnormalities induced by advanced glycation end products, thus improving mitochondrial morphology and dynamics in vivo [58]. Therefore, dendrobine inhibits ferroptosis and improves cognitive dysfunction associated with diabetes by influencing iron metabolism and lipid peroxidation processes.

Quercetin (QE)

QE is a flavonol belonging to the flavonoid class of compounds and is one of the most widely distributed polyphenols in fruits and vegetables. Various in vivo and in vitro studies have shown that QE not only possesses antioxidant, anti-inflammatory, antidiabetic, and antiviral properties but is also lipophilic, allowing it to easily cross the blood–brain barrier and alleviate oxidative stress, thereby improving cognitive function and reducing the risk of neurodegenerative diseases [102, 103]. QE has been shown to protect the mouse hippocampal cell line HT-22 from glutamate-induced oxidative toxicity and lipid peroxidation by blocking free radical production, increasing GSH levels and antioxidant enzyme function, and markedly attenuating Aβ1-42-induced cytotoxicity [104]. More recently, QE was shown to suppress ferroptosis in hippocampal neurons and attenuate DACD by reducing lipid peroxidation and iron deposition in the hippocampus by binding to KEAP1 and subsequently upregulating the Nrf2/HO-1 signaling pathway [81]. However, QE can also exert anti-inflammatory and neuroprotective effects against diabetes and diabetic complications by activating the AMPK pathway [105]. In summary, QE is an effective antioxidant that can prevent severe oxidative stress and lipid peroxidation, inhibit ferroptosis, and improve cognitive impairments associated with diabetes. Nevertheless, further research is required to elucidate the pathways through which QE suppresses ferroptosis in treating DACD to enhance its application in clinical practice.

Synephrine

Synephrine (SIN) is an active substance extracted from the traditional Chinese herbal medicine Qingfengteng, which reportedly exerts a wide range of pharmacological effects, including anti-inflammatory, immunosuppressive, and neuroprotective effects [106]. SIN contributes to the activation of the Nrf2 antioxidant system comprising Nrf2 and HO-1, thereby protecting neurons from cytotoxicity and oxidative stress [107]. At a dose of 100 mg/kg, SIN administration reduced MDA and ROS levels in rats and partially alleviated cognitive deficits after a trimethyltin attack [108]. In addition, in vivo results confirmed that SIN could reduce hippocampal neuronal ferroptosis via the EGF/Nrf2/HO-1 signaling axis and alleviate TDACD in rats [74]. Current research on the potential impact of SIN on DACD is relatively limited. However, these findings indicate that SIN may have therapeutic effects in improving cognitive function and combating oxidative stress. Further research will contribute to a deeper understanding of the mechanism of action of SIN in treating DACD and its application in clinical practice.

Collectively, preclinical studies on ferroptosis as a potential therapeutic target for DACD provide a pharmacological basis for developing optimal therapeutic agents. However, research on these natural plant products is in its infancy, and further validation across different in vitro and in vivo models needs to be undertaken in future investigations.

The Application of Ferroptosis Inhibitors in AD and DACD

Inhibiting ferroptosis in the central nervous system can effectively improve cognitive deficits, whereas promoting ferroptosis may exacerbate cognitive impairments. Specific ferroptosis inhibitors have been shown to effectively suppress neuronal ferroptosis and memory impairment induced by in vivo and in vitro Aβ aggregation. Certain ferroptosis inhibitors, developed for targeting ferroptosis specifically, have demonstrated therapeutic effects in preclinical studies on AD and DACD.

Ferrostatin-1 (Fer-1) was shown to suppress ferroptosis and thereby ameliorate neuronal death and memory deficits in AD brains by suppressing ROS levels and downregulating Nrf2 and GPX4 expression [42]. In addition, Fer-1 improved Aβ-induced spatial memory deficits [109]. Although in vivo and in vitro studies have demonstrated the notable effects of Fer-1 in ameliorating oxidative stress and preventing ferroptosis, no clinical trials are currently available. The efficacy of Fer-1 in AD, as well as associated side effects, need to be further explored. Moreover, ferroptosis can be prevented by directly targeting iron, and iron chelators can prevent ferroptosis by blocking lipid peroxidation driven by the iron-mediated Fenton reaction, thus reducing Aβ deposition in AD model mice [110, 111]. However, it is worth noting that iron chelators used clinically may induce side effects possibly related to the blockade of other iron-dependent physiological mechanisms, leading to off-target side effects and toxicity. In addition, hinokitiol, a novel ferroptosis inhibitor, reportedly possesses notable blood–brain barrier permeability and exerts neuroprotective effects by modulating the Nrf2 pathway and attenuating paclitaxel-induced ferroptosis-related neurotoxicity [112]. Likewise, synthesized hydroxylated chalcones, as a dual-function inhibitor of Aβ aggregation and ferroptosis, completely disrupted Aβ-induced lipid peroxidation, inhibited ferroptosis, and reduced neurotoxicity induced by oxidative stress [113]. However, the specific therapeutic targets of this novel agent warrant further investigations. Few studies have explored the potential of ferroptosis inhibitors in the treatment of DACD. Erythropoietin, which possesses neurotrophic effects, can exert effects similar to those of ferroptosis inhibitors and may improve TDACD by inhibiting ferroptosis, reducing iron overload, decreasing lipid peroxidation, and regulating the expression of ferroptosis-related proteins in a T2DM mouse model [17].

These studies suggest that novel potent ferroptosis inhibitors may serve as crucial disease-modifying therapies to delay the progression of AD and DACD (Table 3). Notably, the drug dosage and timing of intervention in the disease course are critical factors in the development of new ferroptosis inhibitors.

Table 3 The application of ferroptosis inhibitors in AD and DACD
Full size table

Conclusions and Future Perspectives

Taken together, ferroptosis is generally accepted to play a key pathogenic role in cognitive impairment associated with diabetes and AD. As summarized in this review, considerable compelling evidence suggests that targeting ferroptosis can provide promising new options to treat cognitive impairment associated with diabetes and AD. Despite the aforementioned findings, current studies on ferroptosis in the context of delaying the progression of cognitive impairment associated with diabetes and AD remain superficial. Although numerous targets for ferroptosis have been identified, the actual efficacy of these targets, specifically their applicability to patients and ease of modulation, is yet to be established. Moreover, the clinical application of ferroptosis-regulating drugs presents both opportunities and challenges, necessitating high specificity and stability in ferroptosis-targeted regulators. Therefore, it is crucial to further elucidate the molecular mechanisms of different target-regulating factors in the treatment of iron-dependent diseases. Furthermore, this review summarizes the available research on interventions with natural plant compounds to treat cognitive impairment associated with diabetes and AD via ferroptosis inhibition. However, most current studies are based on preclinical experiments and lack clinical trial research with multicenter, large cohorts, and long-term follow-ups, an area warranting future research endeavors. Finally, there are similarities between the modes of cell death, including ferroptosis, autophagy, and apoptosis. However, the specific mechanisms of these different types of cell death remain unclear. Whether these various cell death modes collaborate or antagonize each other, as well as their potential synthesis into a comprehensive regulatory network, requires further exploration.