Introduction

Alzheimer’s disease (AD) is a chronic and irreversible neurodegenerative disease associated with dementia in elderly people [1, 2]. In AD brain, neuropathological alterations are associated with the amyloid-beta (Aβ) aggregation which generates senile plaques (SPs) and primarily results in various consequences, for example, hyperphosphorylated aggregates of the microtubule-associated tau protein in neurofibrillary tangles (NFTs), impaired neuronal connectivity, and loss of neurons [3,4,5]. Numerous studies have extensively analyzed Aβ’s structure and its harmful effects in inducing oxidative stress (OS), autophagy, and neuroinflammation [6,7,8]. To treat AD, several drug candidates were developed to eliminate or decrease Aβ generation [9,10,11]. Nevertheless, most of them failed in clinical trials [12, 13]. In recent times, it has been revealed that Aβ aggregation is not regarded as the initial event of AD pathogenesis; rather, Aβ aggregation is a subsequent event of the disease [10, 14]. Thus, new research approaches are required to discover effective AD treatments. Several studies have demonstrated that homeostasis of essential biometals (for example calcium, magnesium, manganese, copper, zinc, and iron) is impaired in case of AD. Furthermore, these metals contribute significantly to the metabolism and aggregation of tau and Aβ. Depending on these findings, a metal hypothesis for AD has been proposed [12, 15], which is suggesting that targeting the interactions of metals with Aβ may prove more effective in AD prevention.

Several studies have demonstrated the pathophysiological impacts of metal imbalance in the brain [16, 17]. In a study, Akhtar et al. [18] revealed that chromium picolinate treatment attenuated streptozotocin-induced cognitive impairment. Furthermore, treatment with chromium picolinate reversed pathology of AD, as demonstrated by enhanced memory, decreased oxidative damage, mitochondrial dysfunction, neuroinflammation, and upregulated insulin signaling [18]. Nonetheless, still, there are arguments regarding impaired biometal activity as the causative factor for AD. The presence of the blood-brain barrier (BBB), makes it difficult to treat the brain diseases [19, 20]. As the BBB cannot be passively penetrated by the biometals, thus the mentioned metal imbalance in the AD brain cannot only be associated with the decreased or increased exposure to metals, but rather to a more initial distribution of intracellular ions in an unclear manner. Therefore, brain metal homeostasis is regulated by several metal exporters, importers, and metal sequestering proteins.

It has been demonstrated that heavy metal accumulation in the human body can be harmful for multiple organs. Particularly, heavy metals are well-known to exert toxic effects on the brain. Several studies have particularly focused on the neurological functions of cadmium, mercury, and lead in the brain [21]. In this article, we have focused on the biometals, heavy, and non-essential metals–induced molecular processes of AD.

Biometals and Alzheimer’s Disease

Iron

Iron (Fe, an essential trace metal) is involved with several vital neuronal functions in the brain, including synthesis of myelin, mitochondrial respiration, and transport of oxygen [22]. Furthermore, iron also plays a role as a cofactor for numerous metalloproteins associated with signal transduction and metabolism [23, 24]. Numerous studies also found increased levels of iron in AD brains [25, 26], particularly in putamen and globus pallidus areas [27, 28]. Nonetheless, as compared to healthy individuals, studies observed reduced or unchanged levels of iron in the serum of AD individuals [29, 30]. It has been revealed by a meta-analysis that there is a significantly decreased level of iron in the serum of AD individuals as compared to healthy controls. Nonetheless, iron level in CSF was not influenced by AD; however, further analyses are essential because of the relatively small number of CSF studies carried out till now [29]. Scientists have carefully studied iron content in 12 selective areas of the brain via separated meta-analyses by utilizing cross-referenced statistical methods [29]. They also observed that 8 certain areas of the brain contained an elevated level of iron that correlated with the clinical diagnosis of AD in a statistically validated manner. Indeed, these findings provided rigorous statistical support for the model that iron homeostasis was altered in individuals with AD, along with the finding of lower iron in their serum and evidence for iron overload in various specific areas of the brain [29]. Nevertheless, meta-regression analyses showed in several studies that differences in iron levels in serum might be owing to the different mean ages [30]. Unfortunately, clarification of this aforesaid imbalance is still not known. Studies also showed that in case of AD, excessive levels of iron can stimulate the production of hydroxyl radical by Fenton reaction, which can eventually lead to an elevated level of oxidative stress in AD.

Dynamic relationship between efflux and influx of iron is important to maintain homeostasis of intracellular iron, where various transporter proteins have significant contributions. In AD individuals, impairment of iron exporter ferroportin (FPN), along with iron importers, such as melanotransferrin (MTf), lactoferrin (Lf), divalent metal transporter 1 (DMT1), and transferrin (Tf), can significantly play role in the accumulation of iron in the affected areas of the brain (Fig. 1). Except for oligodendrocytes, DMT1 is expressed on microglia, astrocytes, and neurons. DMT1 is associated with the pathway involved in Fe2+ influx [31]. In the AD brain, DMT1+IRE (iron-responsive element) and DMT1-IRE are the 2 isoforms of DMT1 that were found to colocalize with Aβ in the plaques. In amyloid precursor protein (APP)/presenilin-1 (PS1) transgenic mouse model, these 2 isoforms of DMT1were also found to be elevated in the hippocampus and frontal cortex regions, along with a decreased level of FPN expression [32], which is further indicating that the dysregulated iron metabolism–associated protein DMT1 and FPN have significant contribution in the iron-facilitated AD neuropathogenesis. Hepcidin (a protein, involved in iron homeostasis) is co-located with FPN in astrocytes, and neurons exhibited a decreased level of FPN expression in AD brains. It has been observed that the downregulation of hepcidin can lead to the impaired iron export pathway which can eventually lead to retention of cellular iron [33].

Fig. 1
figure 1

The role of iron in the pathogenesis of Alzheimer’s disease. The ferrous form of iron (Fe2+) can directly enter into the cell via DMT1, whereas transferrin (Tf)-ferric iron (Fe3+) enters via the transferrin receptor (TfR)-facilitated endocytosis. Elevated Fe2+ levels stimulate the Fenton reaction to generate hydroxyl radical (•OH), which eventually results in oxidative stress and neurodegeneration. Furthermore, Fe2+ can elevate phosphorylation of tau via activation of glycogen synthase kinase 3β (GSK3β) and cyclin-dependent kinase 5 (CDK5) that lead to the formation of neurofibrillary tangles (NFTs). The iron chelators reduce the phosphorylation of tau by inhibiting the GSK3β and CDK5. In the cellular environment, Fe2+ binds with the iron-responsive element (IRE) in the 5′ UTR region of amyloid precursor protein (APP) mRNA, which results in the induction of APP translation that leads to the formation of amyloid beta (Aβ). This generated Aβ can interact with Fe2+ as well as increase Aβ aggregation. Fe2+ can only go out of the cell by using ferroportin (FPN) along with the activity of haptoglobin or ceruloplasmin. APP can also interact with FPN to oxidize Fe2+ into Fe3+ for Tf binding. Nevertheless, the binding of hepcidin (HP) with FPN leads to its internalization to avert the export of Fe2+

Full size image

APP can catalytically oxidize Fe2+ to Fe3+ (Fig. 1) through the activity of ferroxidase and subsequently interact with FPN to mediate the export of iron [34]; nevertheless, this mechanism is suppressed through extracellular zinc (Zn), which originates from Zn-Aβ complexes [35]. In the same study, researchers exhibited that loss of soluble tau can result in retention of iron via weakening APP-induced export of iron [36], and such inhibition can be achieved by the use of lithium [37] or an iron chelator [38]. In addition to this, sirtuin 2 controls homeostasis of cellular iron through deacetylation of nuclear factor erythroid-derived 2-related factor 2, which functions as a transcription factor to regulate the expression of FPN [39]. In a study, it was found that compounds derived from Chinese herbs can decrease the expression of DMT1 and can increase the expression of FPN, which is suggesting a new approach to reduce iron overload-mediated impairment in AD [32]. In BBB’s endothelial cells, Tf-transferrin receptor (TfR) complex was found to play a role in iron uptake. Iron transport across BBB can take place owing to receptor-facilitated endocytosis of Tf-bound iron [40]. A study showed that cerebrospinal fluid (CSF) levels of Tf were markedly different in case of familial AD (FAD) when compared between individuals who carried mutations and related non-carriers [41]. Lf’s structure is similar to the structure of Tf, where both of these iron importers have 2 lobes, in which each lobe contains a binding site for Fe3+ (Fig. 1) [42]. In AD patients, Lf expression is high in macrophages/monocytes and fibrillar-type SPs in the cerebral cortex region of individuals with AD [43]. Furthermore, the formation of SPs mediates the age-related deposition of Lf [44]. Interestingly, lipoprotein receptor–related protein (LRP, a cell surface receptor) is associated with the clearance of Aβ by an endocytic process. Furthermore, Lf can bind with LRP and can significantly increase soluble Aβ clearance instead of Aβ generation [45]. A liposomal system involving surface Lf was developed for the delivery of neuron growth factors through the BBB. Indeed, the aforesaid technique was useful to control AD progression [46, 47].

Copper

Copper (Cu, an essential trace metal) is also involved with various important cellular activities, such as it plays role as a structural component of enzymes that are essential for antioxidant defense and energy metabolism [48]. Association of copper with AD pathophysiology is complex. Increased copper levels have been identified in SPs [49]. In AD brain, a deficiency in the total copper levels in the AD brain has been reported by some studies [50]. Furthermore, another study revealed that even though the combined serum and plasma copper level was higher in AD individuals [51], however, total CSF levels of copper were not different when compared between AD individuals and healthy subjects [52, 53]. The reason for this heterogeneity is that a substantial amount of copper precipitates with SPs in AD-affected areas, which can further lead to a deficiency of copper in other areas. Indeed, copper can interact with both tau and Aβ and can exacerbate their pathological outcomes [54, 55].

Still, the processes associated with copper dislocation in the AD brain are not clear. Copper transporter 1 (CTR1) and the copper transporting P-type ATPases, such as ATPase copper transporting alpha and beta (ATP7A and B), are the main transporters associated with the cellular regulation of monovalent copper [56, 57] (Fig. 2). It was reported that DMT1 might contribute to divalent copper delivery into cells to synthesize copper-containing enzymes [58]. However, in terms of copper overload, dual-roles are played by ATP7A and ATP7B in the export of excess copper out of cells in an adenosine triphosphate (ATP) hydrolysis–dependent manner. Along with the transporters, a group of intracellular proteins, known as molecule chaperones, for example, copper chaperone for superoxide dismutase (SOD), cytochrome oxidase enzyme complex, and antioxidant protein-1, also play roles in copper delivery to certain targets [58, 59]. Interestingly, the genetic knockdown of copper transporter 1C (CTR1C) in a Drosophila model of AD markedly decreased the accumulation of copper in the brain [60]. Similar findings were also seen in flies when DmATP7 (a copper exporter) was increased, or when CTR1B (a copper importer) was suppressed in AD flies. Furthermore, these flies showed elevated Aβ generation but a decrease in Cu-Aβ complex-mediated OS, indicating that Aβ oligomers, or the elevated levels of Aβ aggregates, were less toxic in a decreased influx of copper-induced by CTR1 knockdown [60]. In an AD mouse model, it was observed that ATP7A can be increased in activated microglial cells where the amyloid plaques are gathered, which can further lead to a substantial change of copper trafficking in microglia. This finding suggests a neuromechanism, where inflammation-mediated copper dyshomeostasis in microglia is linked with AD [61]. On the other hand, genetic studies showed that a cohort analysis of a single nucleotide polymorphism in ATP7B is responsible for impairments in circulating non-ceruloplasmin-bound copper which can elevate the AD risk, which further indicates that alterations in copper homeostasis might speed up the neurodegeneration process which can lead to AD [53, 62].

Fig. 2
figure 2

The role of copper in the pathogenesis of Alzheimer’s disease. In the human body, copper is most commonly shifting between cuprous (Cu+) and cupric (Cu2+) forms. Cu+ enters into the brain cells via copper transporter 1 (CTR1), whereas Cu2+ uses divalent metal transporter 1 (DMT1) to enter into the brain cells. Cu+ accumulation is sequestered into specific locations in the cells via several chaperones of copper such as antioxidant protein-1 (ATOX1). ATOX1 plays a role in the Cu+ transfer to ATPase copper transporting alpha and beta (ATP7A and B), which also aids the Cu+ import into synaptic vesicles for release and/or directly facilitate copper efflux that interacts with amyloid beta (Aβ). Excessive levels of intracellular Cu2+ might cause activation of Fenton reaction to elevated oxidation of biomolecules. In addition, Cu2+ is associated with the activation of GSK3β that eventually leads to hyperphosphorylation of tau. Cu2+ also increases the phosphorylation of amyloid precursor protein (APP) and generate Aβ, which results in the formation of senile plaques

Full size image

Zinc

Zinc (Zn, an essential trace metal) is an important constituent of about 100 s of proteins and enzymes [63]. This trace metal is also extensively associated with cell signaling as compared to other metals, mainly because it can also play a role as a neurotransmitter [64]. Since a former study reported zinc redistribution into extracellular SPs, therefore functions of zinc have been widely studied in AD pathogenesis (Fig. 3) [65]. In AD individuals, inconsistent findings have been observed with the levels of zinc. However, numerous studies have reported increased zinc levels in the CSF and brains of AD individuals [66]. Other studies observed reduced or even no difference in the levels of zinc in the serum and brain of AD individuals as compared to controls [30, 67, 68]. Still, the exact cause of such diverse findings with the levels of zinc is not known. Henceforth, the development of zinc as an AD treatment is hindered. It is known that zinc binds with the histidine residues in the Aβ’s C-terminus and promotes the formation of aggregates. As compared to copper and iron, zinc has a greater affinity to bind with Aβ in a wide pH range [69].

Fig. 3
figure 3

The role of zinc in the pathogenesis of Alzheimer’s disease. In the plasma membrane, zinc importing protein (ZIP) mainly regulates the zinc (Zn2+) entry, while zinc transporter (ZnT) regulates the efflux of Zn2+. An elevated level of Zn2+ increases the accumulation of Aβ, tau modification as well as increase the formation of reactive oxygen species (ROS). Conversely, lower levels of Zn2+ reduce the bioavailability of Zn2+ and amyloid beta (Aβ) clearance as well as cause synaptic dysfunction

Full size image

Homeostasis of zinc in neurons is mainly regulated by 3 groups of transporters including metallothioneins (MTs), zinc-regulated transporter-like, and iron-regulated transporter-like proteins (ZIPs), and zinc transporters (ZnTs) (Fig. 3). It has been found that ZnTs mediate the efflux of zinc from cells or facilitate excessive zinc from the cytoplasm into intracellular vesicles and organelles [70]. Functions of ZIPs are nearly opposite to the roles of ZnTs. ZIPs mediate the import of zinc into cells or facilitate zinc movement from intracellular vesicles into the cytoplasm [70, 71]. MTs also play roles in the maintenance of zinc homeostasis and regulate the cellular levels of zinc and other related signaling mechanisms [72]. In the brains of AD individuals, immunofluorescence studies confirmed that various ZnTs (including ZnT1, 3, 4, 5, 6, and 7) are widely present in Aβ plaques of cortex region [73]. Among them, ZnT3 is mainly found on the synaptic vesicles of zinc-containing glutamatergic neurons [74]. ZnT3 levels reduce in elderly people with aging, especially in AD individuals [75].

Furthermore, an age-dependent impairment in cognitive functions has been observed in ZnT3 knockout mouse models [76]. However, in the ZnT3 deficit mouse models (while overexpressing APP), reduced levels of plaque burden and synaptic zinc were observed [77], which further indicating the role of synaptic zinc in amyloid plaque deposition in case of AD. In AD, ZnT3 knockout-mediated elevated levels of intra-neuronal zinc worsened the damages of neurons [49]. However, more studies are required on the mechanisms related to zinc reuptake into the presynapses. ZIPs play vital roles in facilitating the influx of zinc, and other processes may also indirectly contribute in this regard including presenilins (presenilin mutation is known as a causative factor for FAD) [78]. Moreover, MTs are the main zinc-buffering peptides that play roles in the maintenance of cytosolic zinc balance. MT-1 to MT-4 are the 4 major isoforms of MT that are found to be expressed in brains. In AD individuals, level of MT-1 and MT-2 increases, while the MT-3 level decreases [79]. Deficiency in MT-1/2 reduced the amyloid plaque burden in an animal model of AD and therefore recovered the APP-mediated alterations in mortality [80].

Interestingly, MT-3 is linked with the aggregation of Aβ via cysteine oxidation. The deficiency of MT-3 was found to partially rescue the APP-mediated mortality of females and resulted in alterations in APP-mediated behavioral phenotypes of mouse models [81]. Following the polymerization of actin, MT-3 controls the uptake of Aβ in astrocytes via its positive activity [82]. Studies involving spectroscopy and microscopy confirmed that MT3 averts Cu-Aβ-induced neurotoxic activities perhaps through a metal exchange in between aggregated Cu-Aβ (1–40) and Zn-MT-3, which can further result in inhibition of reactive oxygen species (ROS) generation [83]. A different study reported the protective functions of MTs (derived from astrocytes) on primary cortical neurons against Aβ toxicity via suppressing the generation of proinflammatory cytokines, increasing the B-cell lymphoma 2 levels, and decreasing contents of ROS [84].

Zinc has been found to induce tau aggregation under reducing conditions [85]. Interestingly, zinc suppressed the intramolecular disulfide bond formation but mediated intermolecular bonds between important cysteine residues. In addition to this, exposure to zinc elevated the phosphorylation of phosphoinositide 3 kinase and mitogen-activated protein kinase-dependent pathways which are vital for tau modifications [86]. For a better understanding of the AD-linked zinc dyshomeostasis, more studies are required to assess the functions of zinc transporters in AD pathogenesis.

Manganese

Manganese (Mn, an important trace element) exerts various vital physiological functions for intracellular homeostasis and growth [87]. Manganese plays an important role as a cofactor for important enzymes that are associated with normal cell function, for instance, glutamine synthetase and SOD. There is growing evidence that indicates that an overload of manganese is linked with neurodegenerative diseases and a slight increase in the level of manganese can trigger symptoms that are related to manganese poisoning [88]. Processes that are linked with manganese-induced cytotoxicity include reduced cellular antioxidant defense and autophagy, buildup of intracellular toxic metabolites, aberrant energy metabolism, mitochondrial dysfunction, and ROS overgeneration [89, 90].

Significantly increased level of manganese was found in the brain of AD individuals with dementia in comparison with the healthy subjects, while the maximum manganese level was found in the parietal cortex [91, 92]. This finding indicates that an overload of manganese might be associated with the cognitive deficit and AD pathology. Chronic exposure of manganese altered gene expression which dispersed Aβ plaques in non-human primates. Interestingly, p53 mostly targeted the altered genes; amyloid-beta precursor-like protein 1 (APLP1) was one such gene, which was found to be the major upregulated gene in the frontal cortex [93]. Exposure to manganese mainly affects the frontal cortex which can lead to incipient dementia [94]. Treatment with manganese elevated Aβ levels both in vitro and in vivo; the associated process is perhaps linked with the interruption of Aβ degradation [92]. In a study, it was revealed that manganese may weakly bind with the specific Aβ sites [95]. Nonetheless, more studies are required to demonstrate the effects of such manganese binding with Aβ in mediating Aβ aggregation. Indeed, manganese is a constituent of manganese (Mn) SOD (Mn-SOD), which is an antioxidant enzyme that significantly contributes to preserving the vitality of mitochondria. Elevated manganese levels can hinder oxidative respiration, which can increase ROS generation and eventually can result in mitochondrial dysfunction [96]. In a transgenic mouse model of AD, partial Mn-SOD deficiency elevated Aβ plaque deposition and phosphorylation of tau [97, 98]. However, Mn-SOD overexpression exerted various benefits against the pathology of AD via decreasing the burden linked with cortical plaques [99], which further confirmed the associations between the AD pathophysiology and mitochondrial oxidative stress. Moreover, manganese toxicity triggered cognitive impairment in humans, and it has been hypothesized that high level of manganese uptake can result in deficiency of iron in the Golgi apparatus, which is in line with the finding that iron and manganese compete with the same transport processes and binding sites, at least to some degree [100].

Manganese transport is facilitated by several importers, including dopamine transporter, ZIP14, ZIP8, Tf/TfR, DMT1, and also via multiple exporters including FPN, SLC30A10 (solute carrier family 30 member 10), park9/ATP13A2, and the secretory pathway Ca2+-ATPase 1 (SPCA1). Among them, DMT1 mediates iron influx and this transporter is the first mammalian transporter for cellular uptake of manganese. It has been found that DMT1 mediates manganese movement across the BBB, mainly under the conditions of iron deficiency [101]. Through a ligand-receptor endocytosis process, DMT1 transports divalent manganese, whereas Tf transports trivalent manganese into cells via a ligand-receptor endocytosis process [102]. ZIP14 and ZIP8 possess increased binding capacity with zinc; however, multiple studies revealed that these transporters are also associated with the manganese absorption from the lungs and liver [103,104,105]. The contribution of exporter proteins in the maintenance of manganese levels has also been studied. It has been identified by genome analysis that SLC30A10 might transport both manganese and zinc. Manganese accumulation has been detected in the carriers of mutations in SLC30A10 and individuals with Parkinson’s disease (PD) [106]. In the frontal cortex of APP/PS1 transgenic mouse models and AD individuals, the level of SLC30A10 was found to be markedly decreased, which further indicates that its dysregulation can play role in AD pathology [107]. In a pH-dependent manner, FPN (an iron exporter) can play a role as a cellular exporter of manganese to attenuate cytotoxicity and manganese accumulation [108, 109]. Moreover, ATP13A2 (ATPase cation transporting 13A2) plays a role as a cation transporter in the transportation of zinc and manganese. It has been revealed by studies that ATP13A2 overexpression decreases the intracellular level of manganese, which as a result can alleviate manganese-induced lethality; loss-of-function mutations in ATP13A2 are associated with the rise in both Aβ plaques and α-synuclein in Lewy body disease [110]. In addition to this, a SPCA1 homolog in yeast, plasma membrane ATPase-related 1, mediated the transport of manganese and Ca, and ectopic SPCA1 expression in yeast increased sensitivity to manganese toxicity [111]. Therefore, it has been indicated that SPCA1 functions as another regulator for cellular manganese homeostasis. However, further studies are required to investigate the affinity between manganese and SPCA1 and the functions of SPCA1 in the pathogenesis of AD.

Magnesium and Calcium

Magnesium (Mg, a major macro element) plays crucial roles in several enzymatic synthesis and cellular mechanisms such as synaptic plasticity, ion channels, and energy metabolism [112, 113]. Calcium (Ca, a major macro element) plays a role as a ubiquitous second messenger and its role in regulating cellular function has been extensively studied [114]. Levels of intracellular calcium are strictly controlled via several calcium-binding proteins, calcium channels, pumps, and are also controlled by other metal ions including magnesium. It has been found that magnesium can act as a calcium antagonist. Magnesium also plays a role in the maintenance of intracellular calcium concentrations and protecting neurons from excitatory responses mediated via calcium overload, under physiological conditions [115]. Nonetheless, disruptions in calcium and magnesium homeostasis change a series of processes that can result in various diseases including neurodegeneration [116].

Magnesium levels in serum and brain are considerably lower, while levels of calcium were found to be considerably increased in AD individuals as compared to age-matched healthy subjects [117,118,119]. Frequently elevated calcium concentrations can lead to the raised expression of ApoE and APP and can also mediate the generation of Aβ aggregation via a mechanism involving γ-secretase stabilization [120, 121], while Aβ aggregation can lead to altered membrane calcium permeability that can further worsen AD [122]. Some studies have also assessed the contribution of magnesium in AD pathogenesis. It has been revealed by an in vitro study that both magnesium and calcium can induce the mechanism of hyperphosphorylated tau aggregation [123]. Magnesium-l-threonate administration elevated concentrations of magnesium in the brain, which resulted in reduced β-secretase (BACE1) levels thus decreased the β c-terminal fragments and soluble APP levels, therefore alleviated the cognitive impairments and synaptic loss linked with Alzheimer’s symptoms [124]. Moreover, magnesium sulfate treatment decreased hyperphosphorylated tau levels via suppressing the phosphorylation of glycogen synthase kinase 3β (GSK3β) and elevating the activity of phosphatidylinositol 3 kinase (PI3K) and protein kinase B (Akt) [125, 126], therefore indicating that magnesium can play a role as a neuroprotective factor in AD development. The underlying mechanism associated with magnesium blocking the long-term activation of N-methyl- d-aspartate (NMDAR)-induced calcium influx and therefore decreasing calcium-mediated neuroinflammation. Indeed, NMDARs are cationic channels that are activated by glutamate, containing an increased permeability to calcium ions upon the synaptic activity, for instance, memory and learning. Aβ aggregation–mediated NMDAR over-activation might take place during the early AD stages [127]. Intracellular calcium concentrations can be increased by continuous calcium influx, which can further trigger numerous enzymatic processes that can lead to neuronal death, protein destruction, and peroxidation [128]. As an endogenous blocker, magnesium can bind with the NMDAR subtypes including NR1/2A and NR1/2B, which are components of NMDARs present in brain areas affected by AD, under normal conditions [129]. Blocking channels by the addition of magnesium decreased influx of calcium into post-synaptic neurons, to decrease excitotoxic cell death during dementia. In case of neurodegenerative diseases, activation of ATP-gated P2X purinergic receptors (P2XRs) linked with neuroinflammation has also been reported [130]. P2X7R can help in the formation of an oligomer to form membrane pores in microglia to facilitate the influx of calcium [131]. It has been revealed by using tissue culture that magnesium can decrease intracellular calcium levels via P2X7R, and that improved stimulation of purinergic receptor-activated neuroinflammation, which further indicates that elevated magnesium level can play role as an effective calcium entry inhibitor via cell surface channels [132].

Various factors including buffering proteins, exchangers, channels, and numerous transporters are associated with the maintenance of cellular calcium and magnesium homeostasis. Various channels can facilitate the influx of magnesium into cells, for instance, transient receptor potential melastatins 6 and 7 (TRPM6/TRPM7), magnesium transporter 1 (MagT1), and cyclin M (CNNM) transporter. Solute carrier family 41 member 1 (SLC41A1) and sodium-independent magnesium exchanger are essential to favor magnesium extrusion [133, 134]. In addition to this, intracellular calcium balance can be attained via several calcium transporters. NMDAR, store-operated calcium channels, voltage-gated calcium channels, and α-amino-3-hydroxy-5-methyl-4-isoxazole propionic acid receptors (AMPARs) are accountable for the increased level of calcium. On the other hand, calcium-binding buffering proteins including calbindin can mediate calcium storage in the endoplasmic reticulum (ER), while the activities of the calcium-ATPase pump and sodium-calcium exchanger facilitate calcium export out of cells. Indeed, mutant presenilins activate 2 types of calcium receptors and that plasma membrane calcium-permeable channels permit the leakage of calcium ions from the ER into the cytoplasm, therefore triggering a vital effect on ER-Ca dynamics in the AD brain [135, 136]. Furthermore, Aβ oligomers can either induce the formation of calcium-permeable channels or bind with NMDARs, therefore can mediate the entry of calcium through the plasma membrane [137, 138]. Nonetheless, the contribution of magnesium transporters in AD pathology is poorly understood. The physiological function of TRPM7 was found to be coordinated by presenilins, the mutation of which can result in familial AD [139]. TRPM2 removal in APP/PS1 mouse models ameliorated age-dependent memory impairments and decreased ER stress, while in vitro studies revealed that the knockdown of TRPM2 blocked the Aβ-induced rise in the magnitude of whole-cell current, therefore suggesting the significance of TRPM2 effect in Aβ neuronal toxicity [140]. It was also mentioned that TRPM2 alteration can result in calcium imbalance, even though its contribution in the regulation of magnesium linked with AD was overlooked [140]. Moreover, PD dementia and amyotrophic lateral sclerosis were found to be linked with lower levels of magnesium and calcium as compared to healthy subjects, which is further suggesting the dysfunction of TRPM2 and TRPM7 channels [141].

Pathogenic Mechanisms of Heavy Metal–Induced Alzheimer’s Pathology

Various studies have revealed the heavy metal-induced molecular mechanisms that are involved in AD pathogenesis. In this section, we have explained the molecular processes and analyzed the signaling mechanisms based on molecular networks linked with various heavy metals (cadmium, mercury, and lead) mediated AD pathogenesis. Furthermore, we have also discussed the connections of various molecular objects with the signaling pathways (Table 1).

Table 1 Signaling pathway–related molecular components involved in heavy metal (cadmium, mercury, and lead)-mediated Alzheimer’s disease
Full size table

Cadmium

Indeed, Aβ aggregation is an important hallmark of AD pathogenesis [172,173,174]. Interaction between cadmium (Cd) and Aβ is associated with elevated AD risk [172, 175, 176]. Certain metals are also involved with the NFTs formation [172, 177]. Cadmium also can interact with Aβ1–42 [172]. It is known that Aβ1–42 is an important component of SPs that contributes significantly in AD pathogenesis [178,179,180]. Cadmium is also associated with the Aβ aggregation (Fig. 4) [172, 181]. In line with this finding, tyrosine and histidine residues located at the N-terminal part of the peptide and the binding blocked ion channel of Aβ1–42 [172, 181]. AD risk can also be increased via decreased or elevated expression of certain proteins and these alterations are induced by the exposure to cadmium [182]. For instance, M1 receptor can be blocked by cadmium, which can result in AChE (acetylcholinesterase)-R downregulation and AChE-S overexpression [182]. In basal forebrain, cadmium was also found to activate the cell death of cholinergic neuronal cells [182]. This finding is similar to the mechanism involved in AD-related brain degeneration [182]. Interestingly, these symptoms are linked with elevated levels of Aβ, GSK3β, and formation of tau filament (Fig. 4) [182]. Indeed, GSK3β is an important constituent of tau paired helical filaments, which is located in the deposits of NFTs that disturb functions of neurons, and it is used as a marker for AD-related neurodegeneration [183].

Fig. 4
figure 4

The role of cadmium in the pathogenesis of Alzheimer’s disease. Cadmium (Cd) induced the aggregation of amyloid beta (Aβ) oligomers or fibrils by blocking the Aβ1–42 ion channel. Furthermore, Cd blocks M1 receptor that causes overexpression of AChE-S and downregulation of AChE-R and finally increase the expression of GSK3β that leads to hyperphosphorylation of tau

Full size image

As compared to healthy subjects, potential neurotoxic activity of cadmium has been detected owing to increased levels detected in plasma [184], liver [185], and brain tissues [67] of AD individuals [172]. Elevated cadmium levels were also identified in the hair and blood of AD individuals. Cadmium-exposed workers experienced neurobehavioral problems in memory, psychomotor speed, and attention [186]. Like humans, Li X et al. [187] confirmed that APP/PS1 transgenic mouse models exhibit symptoms of the ethological disorder including memory and learning following exposure to cadmium. Indeed, these observed symptoms are AD characteristics. It is considered that cadmium may play roles in AD and this heavy metal might be involved with the generation of NFTs and Aβ aggregation [172].

Mercury

The tubulin proteins polymerize into long chains or filaments that form microtubules [188, 189]. Mercury (Hg) suppresses the effects of tubulin, which results in neuronal damage and eventually AD [190]. This protein shows a very high affinity towards mercury and following binding this metal-ligand, the structural integrity of tubulin is affected, which leads to a suppression of polymerization of tubulin to micro-tubulin, which further leads to the generation of NFTs and SPs [190]. Indeed, these NFTs and SPs are characteristic features in the brains of AD individuals [190]. Activities of mercury were analyzed in studies involving animal neuronal cell experiments. In these studies, degeneration of axon and formation of NFTs were also reported [191]. A study on stem cells showed that mercury played role in neuronal apoptosis and this metal also inhibited the activities of tubulin [191]. Furthermore, mercury can cause tau hyperphosphorylation; as a result, stabilization of microtubules in the neuron can be affected [192, 193].

Mercury-induced OS can also affect the phosphorylation state of tau by elevating its level [193, 194]. Indeed, tau is closely linked with Aβ that can exacerbate both Aβ pathology and tau-Aβ interactions in AD [195]. Accumulation of Aβ can trigger tau hyperphosphorylation in AD [195], which further suggests that Aβ accumulation can stimulate the signal transduction pathways for the hyperphosphorylation of tau [195,196,197,198,199]. It was suggested that tau dysfunction might lead to AD pathogenesis [200]. Several studies exhibited that APP gene expression can be affected by mercury [200,201,202]. In a study, Song et al. [203] estimated whether mercury influences Aβ accumulation facilitated by an imbalance between Aβ generation and clearance. In that study, mercury and methylmercury elevated the level of APP, which is associated with Aβ generation. Levels of neprilysin were reduced in PC12 cells by the treatment of mercury and methylmercury. Collectively, these findings indicated that mercury stimulated Aβ accumulation via the overgeneration of APP and NEP reduction [203]. Astrogliosis caused by methylmercury was also observed in case of AD neuropathology, which eventually plays a role in APP expression via glial activation [200, 201, 204,205,206,207].

Increased levels of mercury were also identified in the blood of AD individuals and were also detected in brain tissue [208,209,210]. Increased levels of mercury were also found in hair [211]. Furthermore, as compared to the control group, increased mercury levels were identified in the hair of patients with degenerative brain disease [211]. In the nervous system, mercury can trigger dementia, attention deficits, and memory loss [212, 213]. In a study, Haut et al. [214] examined workers who were exposed to the vapor of mercury. Increased mercury levels were found in the blood of these workers and they also suffered from cognitive impairments.

Lead

Associations between AD and exposure to lead (Pb) have been extensively studied at the molecular level by generating oxidative DNA damage [195, 215]. In the brain, oxidative DNA damage has been identified during the aging process and this damage can also contribute to AD pathogenesis [195, 215,216,217,218]. Following exposure to lead, OS might play a role in the elevation of Aβ levels which can eventually result in oxidative damage of the nervous system [215, 219]. OS-mediated apoptosis is found to be involved with the accumulation of Aβ [220,221,222]. Lead toxicity–mediated oxidative DNA damage might be associated with the imbalance between 8-hydroxyguanosin (oxo8dG) accumulation and the effect of Ogg1 mediating oxidative DNA damage [215, 219]. In a study, Bolin et al. [219] stated that oxo8dG was briefly modified at the early stage of life (postnatal day 5); however, it later increased 20 months following cessation of lead exposure, but the activity of Ogg1 was found to be not altered. Besides, an age-dependent loss in the inverse correlation between accumulation of oxo8dG and Ogg1 activity was noticed. In old age, lead activity on oxo8dG levels did not take place if the animals were exposed to lead. Furthermore, these elevations in DNA damage took place in the absence of any lead-mediated alterations in manganese-SOD2, copper/zinc-SOD1, and reduced form glutathione (GSH). Collectively, these findings suggest that neurodegeneration and oxidative damage in the aging brain might be affected by the developmental disturbances [219]. The generation of oxo8dG owing to oxidative DNA damage is commonly known as a marker for oxidative DNA [215, 219]. Indeed, the imbalance may be associated with the processes of AD pathogenesis [215]. In a study, Wu et al. [223] found that AD-associated genes (BACE1, APP) were upregulated due to the lead exposure [223]. It was observed that snippets of APP caused the formation of Aβ aggregates in AD [224].

In a different study, Bolin et al. [219] revealed that lead exposure in an earlier stage of life resulted in gene alterations via hypomethylation of the APP gene, which is a gene responsive to lead [225]. Furthermore, this hypomethylation triggered APP gene overexpression and resulted in APP protein generation [219, 225]. Since the level of APP becomes high due to the lead exposure, thus the effect of Sp1 (a transcription factor) which controls AD-related proteins is increased [219, 224, 225]. Therefore, the aggregation of Aβ was induced and resulted in plaque formation in the brain [219, 224]. In a mouse hippocampal cell line, increased expression of APP was identified under chronic exposure to low-dose lead [226]. Besides, Wu et al. [223] also confirmed that lead might also induce AD pathogenesis in the monkey. Following lead exposure in the brain of monkeys, these researchers also observed intracellular staining of total Aβ and dense-core plaques via immunohistochemical analysis [223]. The obtained findings suggested that there was a buildup of immunoreactive Aβ aggregates inside neuronal cells and this further indicated the possibility of AD [223].

In the battery industry, occupation-related lead exposure is confirmed from the studies on workers [227, 228]. It has been found that workers of this industry are exposed to an increased level of lead in comparison with the average adults. These workers exhibited various symptoms of psychological dysfunctions including headaches, vertigo, forgetfulness, and paresthesia [229]. Moreover, an increased level of lead has also been detected in the blood [227,228,229]. In a study, Sharma et al. [227] confirmed that the workers of the industry suffered from schizophrenia-associated differential structural problems of the brain. Furthermore, workers in that study exhibited functional differences in brain activity in frontal lobes and hippocampus [227]. Indeed, these symptoms are usually seen in AD individuals. The researchers summarized that lead can trigger neurodevelopmental diseases showing neurocognitive impairments for example AD and schizophrenia [227].

Non-Biologically Relevant Metal

Aluminum

Indeed, the physiological roles of aluminum (Al) are not well-known, but this metal is toxic to organisms [230]. In daily life, we inevitably come into contact with aluminum because of its ubiquitous presence in the environment. Luckily, aluminum compounds that are present in our ingested food items are not dissolved at physiological pH; rather, it is then eliminated from the body. Indeed, aluminum toxicity can take place if increased levels of aluminum are inhaled or ingested. Aluminum accumulates primarily in the frontal cortex and hippocampal regions of the brain, therefore correlating to the impairment of other essential biometals. This can result in oxidative stress and can affect numerous signaling cascades, features which can further result in neuronal death and induction of neurodegenerative diseases [231,232,233].

Former studies stated the aluminum hypothesis, which indicated that exposure to aluminum is associated with AD etiology [234, 235]. Even though the validity of the aluminum hypothesis in AD has been argued for years and is progressively being regarded as only a fringe hypothesis as compared to various other theories in AD research, however, exposure to aluminum still remains important and endures to be the center of interest [236]. It has been reported by an in vivo study that chronic administration of aluminum increased Aβ generation in the cortical and hippocampal areas in rats [237]. In transgenic mouse models of AD, aluminum administration increased Aβ plaques accumulation and elevated amyloidogenesis, even though this activity might be removed through antioxidant treatment [238]. Collectively, these findings suggest that the neurotoxic activity of aluminum takes place in an elevated level of oxidative damage. In addition, these findings were observed in cultured neurons, where prolonged aluminum exposure caused aggregation of Aβ and fibrillar deposits on the surface of cells [239, 240]. In the body, increased aluminum levels influenced the activities of 3 crucial genes including presenilin-1, presenilin-2, and APP [241]. Aluminum also decreased the effects of some important enzymes associated with Aβ catabolism via triggering the activation of the amyloidogenic pathway [242, 243], which is suggesting a probable decrease of Aβ degradation. Moreover, aluminum increased hyperphosphorylated tau aggregation via suppressing the activity of protein phosphatase 2A [244]. Even though aluminum loading stimulates neurotoxic activities and causes behavioral alterations that partially model AD, but more studies are required to confirm whether toxic exposure of aluminum plays a role as a causal factor for AD.

In neurons, the molecular processes that are linked with aluminum transport are still unclear. Studies have revealed that uptake and transport of aluminum into cells is rather complicated because of other metal ions including iron, which is suggesting that aluminum competes with iron to bind with iron transporters (Tf/TfR or Lf/LfR), which are also involved with aluminum transport across the BBB [245]. A homolog of human DMT1, SMF-3, was involved with aluminum transport into neurons in Caenorhabditis elegans, which ultimately resulted in increased aluminum levels that decreased the levels of cellular ATP and mitochondrial membrane potential [246]. Besides, aluminum suppressed iron-mediated oxidation and the iron regulatory protein 2 (IRP2) degradation via the ubiquitin-proteasome pathway, which is indicating that aluminum stabilizes IRP2 to affect the balance of intracellular iron level [247]. Indeed, aluminum-mediated neurodegeneration seems to be linked with a different molecular pathway that is independent of tau- or Aβ-related toxicity and is mainly mediated by ROS generation and iron buildup in the brain [232].

Conclusion

The imbalances of intracellular biometal homeostasis and toxic metal exposure are linked to AD pathology. Various biometals have been reported to deposit in the brains of AD individuals, which further increased APP expression, Aβ aggregation, and hyperphosphorylation of tau. Toxic metal exposure might also trigger characteristics of AD pathology through various mechanisms including protein modification, neuroinflammation, and OS. Therefore, more studies are required to identify wider alterations of combined metal ion homeostasis in AD. Moreover, it is possible to identify the probable solutions for AD by recognizing the association between specific genes and particularly heavy metals. Indeed, major responsible genes for AD can also be targeted to treat this disease.