If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Department of Neurology and State Key Laboratory of Biotherapy, West China Hospital, Sichuan University, and Collaborative Innovation Center for Biotherapy, Chengdu, P.R. ChinaMelbourne Dementia Research Centre, Florey Institute of Neuroscience and Mental Health, The University of Melbourne, Victoria, Australia
Treatments for Alzheimer’s disease (AD) directed against the prominent amyloid plaque neuropathology are yet to be proved effective despite many phase 3 clinical trials. There are several other neurochemical abnormalities that occur in the AD brain that warrant renewed emphasis as potential therapeutic targets for this disease. Among those are the elementomic signatures of iron, copper, zinc, and selenium. Here, we review these essential elements of AD for their broad potential to contribute to Alzheimer’s pathophysiology, and we also highlight more recent attempts to translate these findings into therapeutics. A reinspection of large bodies of discovery in the AD field, such as this, may inspire new thinking about pathogenesis and therapeutic targets.
Alzheimer’s disease (AD), the most common form of dementia, is increasingly prevalent and a worsening healthcare burden. The cognitive deterioration of AD has remained frustratingly recalcitrant to candidate disease-modifying therapeutics despite massive efforts over the last 35 years. Most research into therapeutics has been philosophically guided by the connection of the hallmark histopathology of AD, cortical amyloid plaques, and neurofibrillary tangles, with familial dementia-causing mutations associated with their most insoluble component proteins, the amyloid-β peptide (Aβ) (
). Both proteins are normal and soluble components of tissue that become denatured by events that are not simply related to overproduction.
Alois Alzheimer himself came to the conclusion 5 years after his description of plaques and tangles that despite their dramatic appearance, these histopathologies were not the cause of neurodegeneration in AD but, rather, a signature epiphenomenon (reviewed [
]). Yet, dogged efforts have been made in the modern era to causatively link the aggregation of these proteins to the brain atrophy, synaptic disintegration, and neuronal loss that characterize AD, through death mechanisms that remain unproven after decades of research (e.g., the Amyloid Cascade Hypothesis [
]). The discovery of familial AD (FAD) causative mutations of the amyloid protein precursor (APP) and the presenilins (that cleave the carboxyl terminus of Aβ from APP) as well as mutations of tau that cause fronto-temporal dementia have been interpreted simplistically through the prism of the toxic proteinopathy theory. Efforts to investigate the neurodegeneration mechanisms of the genetic lesions of AD outside of the formation of putatively toxic aggregates have received, in our opinion, insufficient attention, for example, that pathogenic presenilin mutations cause neurodegeneration without proteinopathy through a loss of trophic function (
Billions of dollars are being spent by big pharmaceutical companies on lowering Aβ or tau as therapeutic strategies. This approach was justified on the premise of the earliest data from murine knockouts of APP and tau, which have minimal phenotypes in youth, leading to the conclusion that these proteins therefore must be functionless rogues that humans can live without. But, the safe redundancy of tau and APP is unlikely because the adverse phenotypes relevant for neurodegeneration, particularly those related to brain metal dyshomeostasis, are, like AD itself, a product of aging and do not emerge until the postreproductive epoch (
). Nonetheless, clinical trials of Aβ-lowering agents proceed despite more than 30 phase 3 trials failing to demonstrate conspicuous cognitive benefit to AD patients or sometimes being harmful, even upon successful clearance of amyloid plaques (
). One of these, aducanumab, has been presented for registration to the Food and Drug Administration on the basis of debatable benefits that were seen in one but not both of its two phase 3 trials and could be explained as a placebo effect caused by the unblinding when treatment is temporarily suspended upon activation of the amyloid related imaging artefact protocols, which is much more common in the active arm (
). In no instance has an amyloid-lowering treatment shown a reliable and indisputable benefit.
With amyloid being challenged as a therapeutic target, other neurochemical changes in AD have attracted growing interest. Hence, the subject of this monograph. Biometals such as zinc, copper, and iron, which have essential roles in normal physiology, have been implicated in AD pathogenesis for more than 25 years, while commanding a tiny fraction of the research and clinical trial resources committed to proteinopathy research. These are physiologically essential metal ions, but their nutritional (or genetic) dysregulation causes neurotoxicity and neurological damage. These metal ions are stringently regulated by multiple handling systems because excess can also be neurotoxic. These should not be called “trace” metal ions because their concentrations in the brain are within the same order of magnitude as magnesium. Also, the epithet “heavy metal” should not be applied to these versatile and essential metal ions but should be reserved for metals such as lead, mercury, and cadmium that are conspicuously neurotoxic and serve no biochemical purpose. Although aluminum has been investigated as a neurotoxicant that may influence AD, we place it outside of this review of essential elements because it is a nonessential “light” metal with no biochemical function but is very abundant in the environment (present at low micromolar concentrations in plasma as an environmental contaminant) and only potentially toxic at high concentrations (
). Over this time, evidence has accumulated to indicate that these biological elements impact Aβ and tau production, posttranslational modification, aggregation, and toxicity. Sensitive multielement assay technology, e.g., inductively coupled plasma mass spectrometry, has enabled metallomics (“elementomics”, actually, because the array of elements measured simultaneously frequently includes nonmetals, such as selenium [Se]) to be adapted to examining biological samples. Furthermore, biological metal dyshomeostasis alone has been shown to cause neuronal and cognitive dysfunction. Here, we review the associations of the brain’s three most abundant physiological transition metals, iron, zinc, and copper, with the pathophysiology and neuropathology of AD. Because the iron-dependent regulated cell death pathway, ferroptosis, is so closely involved with the selenoenzyme glutathione peroxidase 4 (GPx4) (
), we also discuss the role of the essential trace metal Se.
Zinc is essential for brain function, and it participates in protein structure stabilizing and catalytic reactions in living organisms. It is concentrated in the gray matter of the brain, where 20 to 30% of brain zinc is located in glutamatergic vesicles (
). The high flux of zinc in the synapse contributes to synaptic plasticity, and long-term potentiation (LTP) in the hippocampal CA3 region is modulated by zinc at presynaptic and postsynaptic targets (
), highlighting the potential for zinc dysregulation to contribute to cognitive impairment in AD. Zinc homeostasis is mostly regulated by the SLC39 family (zinc regulated transporter-like iron regulated transporter-like proteins, ZIPs), which has 14 members that transport Zn2+ into the cytoplasm (from organelles and cellular uptake), and the SLC30 family (zinc transporters, ZnTs), which has 10 members that transport Zn2+ out of the cytoplasm (extracellularly and into organelles) (
). These two families of transporters are believed to be relative selective for Zn2+, but a few ZIPs and ZnTs transport other metals, such and Fe, Mn, and Cd. The expression of various members of these families is tissue-specific. ZnT3 expression is selectively expressed in cortical tissue, accounting for 20% of total brain zinc content, and, by loading Zn2+ into glutamatergic synaptic vesicles, is responsible for the high concentrations of extracellular Zn2+ released during neurotransmission (
), promoting the stability of Aβ aggregates (Fig. 1). Importantly, the rat/mouse homolog of Aβ is exceptional among mammalian sequences for having a His13Arg substitution that attenuates Zn2+ binding and Zn2+-induced precipitation (
) unless made transgenic to overexpress the human Aβ sequence. These substitutions also impair the binding of Cu2+ and Fe3+ at an overlapping binding site (vide infra). Curiously, APP possesses an ectodomain high-affinity zinc-binding site remote from the Aβ sequence that promotes the affinity of APP for heparin (
Zn2+ can induce different forms of Aβ aggregates depending on the ratio between Aβ and zinc: stoichiometric concentrations of zinc induce nonfibrillary aggregates enriched in the reversible α-helical structure, whereas fibrillar, β-sheet–enriched aggregates are formed with substoichiometric concentrations of zinc as a consequence of seeding (
). This is a major differentiation between the fibrillar pathway of amyloid formation that occurs with a micromolar concentration of peptide in vitro at a slow rate through hydrophobic β-sheet forces, taking days, compared with the millisecond reversible precipitation of Aβ by Zn2+, mediated by an ionic histidine bridge (
The significance of amyloid plaques themselves in the etiopathogenesis of AD is uncertain. It is understood from both postmortem and PET ligand studies that amyloid deposition commences 1 to 2 decades before the onset of functional impairments in the natural history of AD. However, 30 to 40% of people in the age of risk for AD have conspicuous amyloid pathology without cognitive impairment. Indeed, there is no association of amyloid burden with the rate of premortem cognitive decline (
). With the failure of more than 30 phase 3 clinical trials that lower brain Aβ, it is possible that amyloid plaque pathology might be a biomarker of another process, such as zinc dyshomeostasis. Recent evidence has brought to light a mechanism that may explain amyloid deposition caused by the slow turnover of synaptic Zn2+ released during glutamatergic synaptic transmission (Fig. 1). This pool of Zn2+ is normally rapidly cleared by regional mechanisms that are still uncertain. Extracellular Zn2+ clearance from stimulated rat hippocampal slices is impaired by the advanced age of the donor and female sex, two prominent risk factors for extracellular amyloid pathology, which increase the average extracellular Zn2+ concentration over time and promote the aggregation of Aβ (
), but the drug was supplanted for development by PBT2 (5,7-dichloro-2-[(dimethylamino)methyl]quinolin-8-ol), which was more easily synthesized. Like clioquinol, PBT2 rescued the amyloid burden, lowered phosphorylated tau, and rapidly improved memory performance in the APP/PS1 transgenic model of AD (
In a small phase 2a double-blind, placebo-controlled, randomized controlled trial (RCT) of PBT2 for AD (n = 29 placebo versus n = 29,250 mg/day), PBT2 caused significant executive function improvement in only 12 weeks of treatment (
PBT2 EURO Study Group Safety, efficacy, and biomarker findings of PBT2 in targeting Abeta as a modifying therapy for Alzheimer's disease: a phase IIa, double-blind, randomised, placebo-controlled trial.
). In other words, PBT2 did not just arrest boosted performance. How could a nootropic benefit from a purportedly disease-modifying drug emerge after only 12 weeks? While both clioquinol and PBT2 were developed to dissipate amyloid pathology through releasing Zn2+-bridged Aβ oligomers, this was on the presumption that Aβ aggregates were neurotoxic. A second, smaller, phase 2 RCT of PBT2 used changes in amyloid burden by PiB PET imaging as its primary readout. This exploratory study showed only a trend to decreasing amyloid burden after treatment with PBT2 (250 mg/d, n = 25) compared with placebo (n = 15) for 12 months (
), with no differences in cognitive performance. The study was underpowered for a cognitive readout, and the placebo group cognitive performance did not measurably deteriorate throughout the study (a confound of smaller studies of AD). Thus, the possible nootropic boost of PBT2 in 12 weeks at the first RCT could not be caused by a reduction of amyloid burden. Indeed, the PiB ligand detects fibrillar forms of Aβ, which were never the target of PBT2 (
). As the clinical trials were underway, the mechanism of action of both clioquinol and PBT2 was further investigated, and it became appreciated that these compounds are not high-affinity chelators that lower brain metals, but rather they are copper/zinc ionophores that foster the uptake of Zn2+ and Cu2+ into cells with notable impact on multiple relevant neurochemical pathways (
). Thus, it became apparent that these ionophores might be therapeutic not by clearing amyloid but by normalizing the bioavailability of these essential metal ions otherwise trapped in Aβ aggregates (Fig. 2). The Zn2+ released during glutamatergic neurotransmission must be recycled to maintain intracellular stores for various physiological events, including maintaining the expression of the N-methyl-D-aspartate (NMDA) receptor submits. The trapping of Zn2+ by extracellular Aβ aggregates can impair neuronal function by causing a deficiency of intracellular Zn2+ (
). Drug candidates such as clioquinol and PBT2 may act, therefore, to normalize the distribution of Zn2+ by facilitating its reuptake and distribution to intracellular targets, as demonstrated in models of autism (
), rather than acting as chelators or reversing Aβ aggregation.
While ZnT3 is responsible for supplying the extracellular Zn2+ that promotes extracellular Aβ aggregates, its expression is essential for maintaining cognition with aging, as demonstrated with the accelerated decline in cognition in aging ZnT3 knockout mice (
). Higher levels of ZnT3 were associated with slower antecedent cognitive decline in an unbiased large-scale proteomic analysis of postmortem brain from two tissue banks, even when adjusted for AD pathology (
). Critically, treatment of ZnT3 knockout mice (with no amyloid) with the zinc/copper ionophore, clioquinol, for 6 weeks corrects the early onset cognitive deficits and normalizes changes in synaptic proteins (
). Similarly, treatment with PBT2 of normal old (22 months) C57Bl6 mice (also without amyloid) restores age-dependent cognitive deficits within 12 days, while rejuvenating synaptic architecture and markers, decreasing phosphorylated tau and significantly increasing zinc (but not copper) in the CA1 hippocampal region but not in any other cortical region (
). The regional selectivity of the zinc elevation in PBT2-treated old mice probably reflects the greater dynamic zinc release and uptake physiology in this region, where zinc turnover is impaired with age (
) are mediated by restoring cortical zinc homeostasis and that the amyloid aggregates are a proxy for perturbed zinc trafficking that may exaggerate the problem by trapping more zinc. Therefore, the significant cognitive improvement in a strikingly rapid time frame, 12 weeks, observed in trials (
PBT2 EURO Study Group Safety, efficacy, and biomarker findings of PBT2 in targeting Abeta as a modifying therapy for Alzheimer's disease: a phase IIa, double-blind, randomised, placebo-controlled trial.
) is consistent with the time frame in cognitive improvement in each report of these animal models treated with zinc ionophores and therefore most likely reflects the treatment benefits of correcting cortical zinc homeostasis.
Disruption of cortical zinc homeostasis in AD has not been reflected in reports of bulk zinc levels from postmortem tissue (
). However, factors including the accuracy of diagnosis, statistical power, methods of sample preparation, and detection limits may have made changes inconsistent between studies. Also, the total tissue zinc levels might not rise until the plaque burden is severe (
The blood–brain barrier prevents passive fluctuations of plasma zinc from being transduced into the brain. Nevertheless, some reports have explored the impact of dietary zinc on Aβ transgenic models, with inconsistent results reported. Prenatal and postnatal zinc-enriched diets in Tg2576 and TgCRND8 were described to induce accelerated cognitive impairment in these mice (
), although it is difficult to know whether these changes are, like ZnT3, potentially upstream in the pathological process or whether they represent homeostatic responses. The message RNA levels of several ZnT family proteins such as ZnT1, ZnT4, and ZnT6 are increased in AD tissue and correlate with Braak pathological staging (
). The protein level of ZnT6 has been reported to be elevated in the hippocampus/parahippocampal gyrus region of pathologically confirmed AD cases, but the level of ZnT1 was significantly decreased in the same region (
), which is the main zinc storage protein in neurons.
Interestingly, zinc may regulate the production of Aβ via affecting the secretases that are responsible for its production. The activity of β-secretase responsible for APP cleavage into the nonamylogenic pathway, a disintegrin and metalloprotease 10, is a zinc metalloproteinase, and mutation of its zinc-binding site abolishes its activity (
). Free Zn2+ is reported to promote tau phosphorylation and aggregation (Fig. 1). Several kinases and phosphorylases were suggested as mediators of zinc-induced tau hyperphosphorylation in cell culture and mice, including glycogen synthase kinase 3β, cyclin-dependent kinase 5, extracellular signal-regulated kinase, c-Jun N-terminal kinase, and protein phosphatase 2A (PP2A) (
The slowing of zinc synaptic turnover with normal aging could lie upstream of the amyloid pathology of AD as well as some facets of cognitive impairment. This warrants more in-depth research, especially because the nootropic benefits of correcting in aging mice without amyloid are provocative and could be the basis of drug interventions for which there is already some clinical trial evidence. The mechanisms for clearing the synapse of zinc released during neurotransmission need to be elaborated urgently. While zinc dyshomeostasis may hamper neuronal function, its connection to neurodegeneration is still unclear, as is the mechanism for synaptic loss in ZnT3 knockout mice that is corrected by zinc ionophore treatment (
Copper is a redox-active metal that is involved in multiple metabolic activities in the brain, and it serves as the active site for a range of cuproenzymes such as ceruloplasmin (Cp), superoxide dismutase 1 (SOD1), tyrosinase, cytochrome oxidase, etc (
). Several studies have reported that copper supplementation in cultured neurons inhibits the activation of receptors for NMDA, α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid, and gamma aminobutyric acid (
). Despite the decrease in total copper in AD-affected tissue, the proportion of “labile” or loosely bound exchangeable copper ions was increased, indicating a disruption of the average coordination environment of cellular copper ions in the tissue (
). Furthermore, there is evidence that copper concentrates with other metals in amyloid plaques (vide infra). Thus, there is a change in the distribution of copper in the AD brain tissue where it is deficient in the cells but trapped in the extracellular plaques. This complex picture is consistent with experimental results, reviewed here, that intracellular copper deficiency promotes Aβ production, whereas extracellular Cu2+ pooling can promote Aβ precipitation (under acidic conditions) and oxidative cross-linking and modification. Therefore, neither copper chelation nor copper supplementation are likely to have unopposed benefits, and the theoretical ideal agent would mobilize extracellular Cu2+ to be taken back into the cell. Here, we review the evidence for this.
Lowering cellular copper has been shown to increase Aβ production (
) binding sites remote from the Aβ sequence, the copper/zinc binding site in Aβ is overlapping and only emerges once the carboxyl terminus of Aβ is cleaved from full-length APP through the activity of the presenilins. Copper-Aβ interaction was first described in 1994, where Cu2+ was observed to strikingly induce soluble dimer formation of Aβ1–40 at neutral pH (
) with highest apparent affinities of Cu2+ for the peptide aggregates being measured as ≈50 pM for Aβ1-40 and ≈6 aM for Aβ1-42, with the aggregates binding up to three equivalents of Cu per Aβ peptide (
). The very high apparent affinity of Aβ1-42 for Cu2+ may be a product of the perturbed equilibrium of the peptide–metal complex coming out of solution, but nonetheless the peptide aggregation is reversible with chelation, evidencing proof of principle of pharmacological targeting of the metal center for reversing amyloid formation, which was recapitulated by the solubilization of Aβ from the insoluble fraction of AD-affected brain tissue by copper chelation (
Copper binding to the amyloid-beta (Abeta) peptide associated with Alzheimer's disease: folding, coordination geometry, pH dependence, stoichiometry, and affinity of Abeta-(1-28): insights from a range of complementary spectroscopic techniques.
) and Zn2+ binding (vide supra). This is important because mice and rats are exceptional for lacking brain amyloid deposition with age. Zinc (and copper under low pH) induces Aβ oligomerization that favored by greater α-helix content in the peptide. In contrast to metal-free aggregation that proceeds by β-sheet–mediated hydrophobic attraction, zinc-induced aggregation is reversible by dissociating the metal ion (
). Even the trace contaminant metal concentrations (nM) found in neutral buffers is sufficient to promote the seeding and profibrillar aggregation of Aβ peptide solutions and is important to consider in experimental studies (
). The structural basis for this reaction and the pathophysiological significance of this dramatic difference in response to these metal ions has not yet been resolved. Mildly acidic conditions where Cu2+ could precipitate Aβ are thought to be present in the synapse, but this view has been challenged (
Importantly, copper–Aβ interaction can form a catalytic redox-cycling complex that embeds in lipid membrane and recruits substrates like cholesterol to produce hydrogen peroxide and promote oxidative stress that causes neurotoxicity in cell culture (
). This redox activity is abrogated in the rat/mouse Aβ, which is not only less able to promote the catalytic cycling of Cu2+ (and Fe3+) but also lacks the Tyr at position 10 (which becomes Phe) to permit dityrosine modification (
). The metal-centered catalytic cycling of human-sequence Aβ in an oxygenated environment not only generates products such as hydrogen peroxide and 4-cholesten-3-one but also oxidizes the side-chains of the peptide, creating dityrosine cross-linked (highly resistant to catabolism) (
Dityrosine cross-linked Abeta peptides: fibrillar beta-structure in Abeta(1-40) is conducive to formation of dityrosine cross-links but a dityrosine cross-link in Abeta(8-14) does not induce beta-structure.