Alzheimer’s Disease: An Attempt of Total Recall

Abstract

This review is an attempt to compile existing hypotheses on the mechanisms underlying the initiation and progression of Alzheimer’s disease (AD), starting from sensory impairments observed in AD and concluding with molecular events that are typically associated with the disease. These events include spreading of amyloid plaques and tangles of hyperphosphorylated tau and formation of Hirano and Biondi bodies as well as the development of oxidative stress. We have detailed the degenerative changes that occur in several neuronal populations, including the cholinergic neurons in the nucleus basalis of Meynert, the histaminergic neurons in the tuberomammillary nucleus, the serotonergic neurons in the raphe nuclei, and the noradrenergic neurons in the locus coeruleus. Furthermore, we discuss the potential role of iron accumulation in the brains of subjects with AD in the disease progression which served as a basis for the idea that iron chelation in the brain may mitigate oxidative stress and decelerate disease development. We also draw attention to possible role of sympathetic system and, more specifically, noradrenergic neurons of the superior cervical ganglion in triggering of the disease. We also explore the alternative possibility of compensatory protective changes that may occur in these neurons to support cholinergic function in the forebrain of subjects with AD.

Keywords

Alzheimer’s disease amyloid astrocytes cholinergic neurons noradrenergic neurons neurodegeneration oxidative stress sympathetic system tauopathy

According to the International Classification of Diseases, Alzheimer’s disease (AD) is one of the most common forms of dementia. Dementias per se are characterized by the impairment of cognitive functions which is greater than would be expected at the individual’s age and overall premorbid level of cognitive functioning. Memory impairment is typical of most forms of dementia, but cognitive impairment is not limited to memory (i.e., there is impairment in other areas such as executive function, attention, language, social cognition and judgment, psychomotor speed, visuoperceptual or visuospatial abilities). In dementia, cognitive impairment is not related to normal aging and is severe enough to significantly affect a person’s independence in performing activities of daily living.

The current classification of dementias includes vascular dementia, frontotemporal dementia, dementia with Lewy bodies, dementias caused by psychoactive substances, dementias resulting from other diseases, and, of course, AD. AD, in turn, is also divided into several subtypes: 1) early-onset AD; 2) late-onset AD; 3) mixed type AD with cerebrovascular disease; 4) AD of mixed type with diseases of non-vascular etiology. In fact, only the first two types are used during diagnostics. The first type is when AD develops before age of 65 years and accounts for approximately 5% of all AD cases. This type of AD includes mainly hereditary forms of the disease, when the patient is a carrier of a certain mutation (for review, see 1 ). The second type is the AD form that develops after 65 years and accounts for 95% of all AD cases and, probably, has more complex genetic associations.

Since the disease was described in 1906, AD included changes in the histological structure of the brain of patients that we now know as amyloid plaques and neurofibrillary tangles (NFTs) of hyperphosphorylated tau protein. Later on, several other histological changes in the central nervous system were described. Despite the fact that the picture of pathogenetic changes in the brain and body in AD has expanded significantly, the main focus of research and approaches to the treatment of AD is substantially shifted towards processes associated with amyloid peptide. However, despite the “victorious” reports 2 on drugs based on anti-amyloid antibodies, there is still no understanding of what underlies the pathogenesis of AD and why these drugs only slow down the AD course but do not prevent its progression.

Here, we will only briefly dwell on the formation of amyloid deposits, as well as the NFT formation, since this topic is quite well presented in other reviews.3 –5 Instead, we will try to unite current knowledge on amyloid plaques and NFTs with data on other AD hallmarks and data obtained in experimental animals.

A BRIEF OVERVIEW OF AMYLOID-β

The main component of amyloid plaques found postmortem in the brains of patients with AD is amyloid-β42 (Aβ₄₂), a peptide consisting of 42 amino acids, although other shorter variant also were found in AD. Aβ is formed from amyloid-β protein precursor (AβPP), a transmembrane protein that can be proteolyzed by alpha-secretase or beta-secretase (BACE1), which cleave peptides of various lengths from the extracellular domain of AβPP, as well as by gamma-secretase (GS) cutting the intracellular domain. From the perspective of AD pathomorphology, the crucial pathway involves the cleavage of AβPP by BACE1 and GS to produce amyloid peptides consisting of 40 or 42 amino acids, which have the potential to aggregate.

It still remains unclear which factors determine path of AβPP cleavage and what exactly contributes to the intensification of AβPP processing along the path that promotes the formation of Aβ. According to modern concepts, the accumulation of amyloid deposits in the neocortex begins long before the onset of cognitive impairment. 3 The Aβ accumulation in the human brain structures follow a specific pattern, which is usually called Thal stages. 6 Aβ plaques appear in the neocortex and slowly spread to the subcortical structures. In the case of early forms of AD, it is assumed that mutations in the APP gene or the PSEN1 and PSEN2 genes (the protein products of the PSEN1 and PSEN2 genes are GS components) contribute to increased formation of Aβ₄₂, probably, by increasing the affinity of AβPP for BACE1, in case of mutations in APP, and by partial loss of function of GS, in case of mutations in PSEN1 and PSEN2 genes. 3

Amyloid plaques in the human brain are deposited at different rates in different parts of the brain, and it remains unclear which factors determine this unevenness. It is important to stress that RNAseq studies of human brain showed APP gene is expressed in all cell types of brain including (from the highest to lowest mRNA expression) neurons, oligodendrocytes, vascular cells, astrocytes, and microglia. 7 This extremely wide expression of AβPP in the brain cells raises question what serves as a seed for formation of Aβ aggregates in specific brain areas.

Several hypotheses have been proposed to explain plaque formation. One suggests that plaque formation may be triggered by disturbed cholesterol transport in cellular membranes. 8 Another idea is that the accumulation of hyperphosphorylated tau triggers intense Aβ production. 9 According to the third idea, isomerization of aspartates in Aβ (from L- to D-isoform) leads to an increase in the propensity of Aβ to form aggregates.10,11, 10,11 Most probably, all of these mechanisms may be involved in seeding and spreading of Aβ aggregates and the exact mechanism in given patient may be determined by some additional factors.

It is well described that in humans Aβ plaques are distributed not only diffusively throughout the brain but also exhibit strong accumulation around blood vessels. 4 The presence of amyloid deposits around blood vessels is very often associated with several other pathogenetic factors of AD. According to current understanding,12,13, 12,13 the drainage of interstitial fluid from brain tissue occurs along the perivascular space of the arteries facilitated by the wave-like movement of the blood vessel walls during the propagation of blood pressure waves. It was proposed that the accumulation of Aβ plaques around the vessels disturbs the flow of interstitial fluid and drainage of cellular metabolic products from the brain tissue. Whether Aβ itself is the agent that provokes this deterioration of drainage, or whether the deterioration of fluid flow by some other factors provokes the accumulation of Aβ aggregates in the tissue remains unknown. However, it should be noted that one of the diagnostic criteria for the development of AD is a decrease in the level of Aβ₄₂ in the cerebrospinal fluid (CSF). 14 It was hypothesized that this decrease is the result of Aβ aggregation in the brain tissue and a weakening of the brain tissue ability to washout Aβ.

The accumulation of Aβ aggregates along the vessels is one of the factors leading to disruption of the functioning of blood vessels and their constriction, 15 which inevitably impairs the blood supply to brain tissue and, as a result, can lead to the development of dementia due to a pathology similar to cerebrovascular illness.13,16, 13,16 In this regard, it is important to note that it still remains unclear to what extent the above-mentioned drugs aimed at Aβ removing from brain tissue lead to the clearing of the perivascular space from Aβ deposits, improving the state of blood vessels and restoring the outflow of metabolites from brain tissue.

It is well known that the small blood vessels in the brain are surrounded by astrocytic feet which form blood-brain barrier (BBB). Therefore, effect of Aβ on vascular characteristics may include also effects related to modulation of astrocytic function. Indeed, it was shown that astrocytes may be even better target for Aβ than neurons. The application of Aβ to neuronal cell cultures leads to an increase in intracellular calcium level, loss of mitochondrial membrane potential, and generation of reactive oxygen species (ROS) only in astrocytes but not in neurons.17,18, 17,18 This specific action of exogenous Aβ on astrocytes but not neurons seems to result from higher cholesterol content in astrocytic membranes where cholesterol promotes formation of pores by Aβ. 19 The ability of Aβ to induce calcium transients in astrocytes was also shown in acute brain slices. 20 In vivo microscopy in transgenic mice showed that astrocytes in mice with amyloidosis can generate spontaneous Ca^2 + waves spreading between cells and, importantly, these calcium waves did not depend on neuronal activity. 21 These findings raise several important questions. First of all, according to single cell RNAseq data, APP mRNA is expressed practically in all brain areas in neurons, oligodendrocytes, and vascular endothelium and, to a lesser extent in astrocytes and microglia, in both mice and human,7,22–24 , 7,22–24 i.e., practically any cell type in the brain may be a source of AβPP, and release of Aβ will be controlled by factors that modulate AβPP processing. Aβ is not toxic for astrocytes and does not lead to their degeneration, at least in culture,17,18, 17,18 suggesting that Aβ-induced Ca^2 + signals may have physiological role. Nevertheless, under some conditions these Aβ-induced changes in astrocytic function may lead to local induction of oxidative stress (see below) and, as a result, pathological changes in neurons, astrocytes, and BBB.

An important factor contributing to the formation of the focus of researchers specifically on AD is the fact that at least 3 mutations in APP leading to the development of early-onset AD have been described to date. Moreover, Down syndrome is associated with triplication of 21^st chromosome carrying APP gene, which leads to overproduction of Aβ. These observations served as a basis for the creation of readily available transgenic mice carrying human genes with mutations in APP and PSEN1/PSEN2 identical to those described in humans. The presence of an available experimental object with the corresponding mutations has led to a peculiar bias in AD research. However, many researchers pointed to the fact that mutations only reduce the age of onset of AD, but do not affect the rate of progression of the disease compared to sporadic forms of AD 25 (although it should be noted that some mutations accelerate time course of AD). This similarity in the rates is often interpreted as an indication that Aβ acts only as an agent instigating AD, and then the disease develops according to mechanisms that are less dependent on Aβ. 26 Moreover, it was also proposed that AD with early onset and sporadic AD may have strongly different pathogenesis during the onset phases and different factors may be critical for the disease induction in each case. 27 We will see below that the most possible variant of the development of sporadic AD is not related Aβ but rather results from spreading of tau pathology which induces Aβ accumulation.

HYPERPHOSPHORYLATED TAU AND CHANGES IN AXONS

Along with Aβ, deposits of NFTs of hyperphosphorylated tau protein (p-tau) are also one of the characteristic features of the pathogenesis of AD.3,28, 3,28 Tau is a microtubule-stabilizing protein and its hyperphosphorylation leads to dissociation from microtubules and their degradation. Moreover, tau hyperphosphorylation leads to the association of this protein into pretangle structures, which are considered as pathological feature, and, later on, into tangles. The predominant view is that tau hyperphosphorylation occurs as a result of the activity of CDK5 kinase29,30, 29,30 and GSK3β kinase31,32, 31,32 as well as some other kinases. 5

Analysis of the distribution of p-tau during AD progression showed that p-tau spreading follows rather standard stages currently referred to as Braak stages. First p-tau pretangles appear in the locus coeruleus (LC) at early stages and, then, spread to transentorhinal and entorhinal regions of the neocortex, hippocampus, prefrontal cortex, and secondary and, at final stages, primary visual, auditory, somatosensory and somatomotor cortices. 28 Importantly, the patterns of Aβ and p-tau spreading in the brain are different and p-tau pretangles in the LC occur years before Aβ deposits appear in the neocortex. Although Aβ is often considered as a factor provoking the formation of tau tangles, 26 it is more likely that Aβ-mediated onset of AD is typical of hereditary AD with early onset. It has been proposed that AD leads to the accumulation of tangles of hyperphosphorylated tau protein, which then spread throughout the brain through a prion-like mechanism 26 when the p-tau-containing neurons secrete NFTs in the extracellular space from their axonal terminals or release exosomes containing NFTs 33 and “infect” neurites of nearby cells with improperly folded p-tau. Accumulation of p-tau in these cells will inevitably lead to disruption of cytoskeleton and slowing or arrest of intracellular transport. It was hypothesized that this disruption of transport along cytoskeleton will impair transport of Aβ, which under conditions of normal cellular transport is harmless. In turn, accumulation of Aβ at the sites of damaged cytoskeleton and formation of intracellular Aβ aggregates will result in neuronal death and Aβ release and further spreading of pathology. 28

Currently, a lot of different therapeutic approaches are being developed to prevent accumulation or intensify removal of p-tau. These approaches range from inhibitors of kinases and agents that prevent tau aggregation to immunotherapy all of which are described in details in recent review. 34 Unfortunately, so far, no success was achieved in the overwhelming majority of these clinical trials despite the fact that the majority of drugs are well tolerated. Probably, this therapy may be combined with other types of therapy to slow down AD progression and achieve stable remission.

In addition to the widely described and intensively studied processes associated with p-tau, there is an even less discussed and studied phenomenon that is also associated with altered axonal structure in AD: Hirano bodies. Hirano bodies were found in axons of hippocampal neurons and their number increase in AD. 35 Their main components are actin, neurofilaments, tau, hippocampal cholinergic neurostimulating peptide, and C-terminal fragment of AβPP. 36 Hirano bodies remain poorly studied since it is unclear which factors lead to their formation. Nevertheless, a cell model of Hirano bodies formation was created on the basis of expression of a truncated version of the 34 kDa actin bundling protein. 37 These model Hirano bodies also contain tau and C-terminal fragment of AβPP 38 and do not induce death in the cell where they occurred. Moreover, it was shown that Hirano bodies may be a mechanism that protects cells against induction of cell death. 39 It is still unclear which role Hirano bodies play in AD pathology and why these intracellular protein aggregates are formed predominantly in the hippocampus.

AD-RELATED SENSORY IMPAIRMENTS

The first well-known hallmark of AD is reduction or loss of olfaction40 –43 which is associated with hypofunction of cortical areas processing smell information. 43 Recent study showed that olfactory decline appears approximately 5 years before mild cognitive impairment and the strength of olfactory loss correlated with postmortem p-tau and Aβ load. 44 In this respect, it is worth to note that amyloid pathology in 5xFAD mice carrying human APP and presenilins with 5 familial mutations (which never occur simultaneously in humans) is not associated with deficits in olfactory function even in elderly mice. 45 The latter suggests that diffuse amyloid pathology in these mice is not a critical factor that can induce changes in olfactory system and questions the pivot role of Aβ accumulation in AD in humans.

In addition to smell, taste function is also disturbed in patients with AD.46,47, 46,47 These negative changes in the ancient sensory systems (or strictly speaking, brain areas involved in the processing of this sensory information) are associated with disruption of eye movements, which are predominantly related to degenerative changes in the neocortex and associated cognitive decline and disruption of attention. 48 Moreover, pupillary responses seem to be also altered and these alterations are related to cholinergic dysfunction. 42

The list of functions disturbed in AD also includes vestibular impairment,49,50, 49,50 reduced ability to discriminate tactile stimuli,51 –53 hearing impairment, 42 and impaired pain and temperature processing. 54 The vestibular deficit seem to be a result of degeneration of vestibular cortex and other brain areas involved in spatial processing 55 whereas impaired tactile discrimination is attributed to cognitive decline. The changes described in this chapter seem to occur during AD phase associated with atrophy of various cortical areas, and, as we will discuss below, are highly likely to occur downstream of some event that triggered AD, probably, more than ten years ago. Nevertheless, impairments in various sensory systems described here are not due to changes in peripheral sensitivity but are usually related to cognitive decline and degeneration of certain cortical areas (i.e., information processing).

GLUCOSE METABOLISM AND INSULIN RESISTANCE

An important factor in the development of dementia in AD is a disruption of glucose metabolism in the brain, which is observed only in some areas of the brain, even in the late stages of the disease. PET analysis of glucose metabolism using fluorine-18 (¹⁸F)-labeled deoxyglucose showed that in AD there is a specific decrease in glucose metabolism in the frontal, temporal and parietal cortices. 56 These changes in glucose metabolism are not necessarily accompanied by an increase in blood glucose levels, however, in some areas of the brain there is an increase in glucose levels and a decrease in the activity of enzymes involved in glucose metabolism. 57 The question of why glucose hypometabolism develops only in certain areas of the neocortex still remains unanswered. However, a decrease in glucose metabolism is proposed to be considered as a marker of the later stages of AD, when the accumulation of Aβ and p-tau has already occurred. 58

Development of insulin resistance is another important aspect of AD. Insulin resistance is a condition that is characterized by the insensitivity of tissues and organs to the metabolic effects of insulin in combination with hyperinsulinemia. Insulin resistance is generally considered part of a broader condition called metabolic syndrome, 59 which often occurs in type 2 diabetes. Accordingly, the pathological phenomena that are usually observed in this condition (micro and macroangiopathy, thrombosis, metabolic changes in tissues) are not specific for the structures of the nervous system but are systemic.

Recent studies demonstrate that insulin resistance can be not only systemic but also local to the brain (cerebral insulin resistance, type 3 diabetes mellitus), which a number of researchers consider as an important component of the pathogenesis of neurodegenerative diseases, including AD. At the moment, there is no common understanding of the relationship between insulin resistance and neurodegenerative diseases but some relationships have been established.

One common view suggests that, when neurodegeneration is accompanied by type 2 diabetes mellitus, the descending signaling from the nervous system to the liver is changed resulting in altered lipid metabolism and the formation of toxic forms of ceramides and other toxic lipids. These lipids are released into the bloodstream after death of liver cells, can easily bypass the BBB, induce neuroinflammation, neurodegeneration, and aggravate cerebral insulin resistance.60 –62 However, this results only from systemic insulin resistance. In the case of AD, brain insulin resistance occurs, which is not identical to systemic insulin resistance and has some overlap. One of definitions is as follows: “brain insulin resistance can be defined as the failure of brain cells to respond to insulin”. 63 In contrast to systemic insulin resistance, brain insulin resistance has little relation to glucose uptake and metabolism in the brain because, in the brain, glucose uptake and metabolism are not dependent on insulin but are modulated by it.

The main source of insulin in the brain is insulin transported from blood whereas expression of insulin by brain cells is still debatable because different authors made different reports on the presence of insulin mRNA in the brain tissue in humans and rodents. 63 According to our RNAseq data, 64 insulin mRNA (both ins1 and ins2 genes in rat) are absent in the rat hippocampus. These findings are corroborated by single cell RNAseq data from mice which indicate that the neocortex and hippocampus24,65, 24,65 along with several other brain areas 23 do not express insulin mRNA. At the same time, these data suggest that cells in the choroid plexus can express ins2 mRNA pointing to the possibility that insulin may be synthesized in choroid plexus and then released into CSF.

In this context, it is worth to mention that AD is associated with pathological changes in the choroid plexus. These changes include alterations in the morphology of epithelial cells, accumulation Aβ deposits, lipofuscin granules, and Biondi bodies, which consist of tau, fibronectin, ubiquitin, and other proteins. 66 Biondi bodies were found only in primates and humans which complicates analysis of their role in ageing and AD but the majority of researchers believe that these protein aggregates reflects some pathological process. 67 Currently, it is not clear whether insulin expression in the choroid plexus is affected by these degenerative changes, however, it is possible to hypothesize that the decrease in the level of insulin and cytokines in the brain and CSF 68 may be related to the degenerative changes in the choroid plexus.

As we mentioned above, brain insulin resistance reflects impaired transmission of signals induced by activation of insulin receptors. One of the effects of insulin on the cell is the activation of Akt kinase. It was found that, in patients with AD and insulin resistance, there is a decrease in the level of proteins of the Akt cascade and an increase in the level of their phosphorylated forms69 –71 which reflects overactivation of Akt kinase leading to suppression of insulin signaling via feedback mechanism. 71 One of the functions of Akt kinase is to suppress the activity of the enzyme GSK3β, whose increased activity, as we mentioned above, is often associated with the development of tauopathy. Thus, a decrease in insulin level as well as suppression of activity of its downstream intracellular target Akt kinase may be the components of AD development. 72

Another target of insulin signaling is peroxisome proliferator activator receptor (PPAR) δ which modulates insulin-stimulated gene expression and has anti-inflammatory action. According to single cell RNAseq data, it is predominantly expressed in vascular endothelium and microglia and, to minor extent, in neurons in mice and humans.22,23, 22,23 It was shown that AD is associated with a decrease in PPAR-δ level and administration of its agonists to animal models of AD suppresses neuroinflammation and Aβ deposition. 73 The latter points to an additional role played by insulin as a modulator of the anti-inflammatory signaling pathway as well as to negative effects that may arise from a decrease in levels of both insulin and its downstream partners.

NEURODEGENERATION IN AD

Longitudinal studies using magnetic resonance imaging showed that AD is quite heterogeneous disease in terms of appearance of cognitive impairment and degenerative changes in the neocortex and hippocampus. Depending on various factors, cognitive decline in AD may occur in the absence of any degeneration in the neocortex and hippocampus or be associated with substantial cortical and hippocampal atrophy. The latter variant is more typical of people with higher intellectual abilities, presumably, because loss of neurons at initial stage of the disease is compensated by higher intellectual flexibility in these people. 74 Hippocampal and cortical atrophy occur as a secondary event after accumulation of NFTs in these brain areas. Currently, it is well established that this massive neuronal degeneration is preceded by smaller and more local degenerative changes in subcortical areas.

The cortical and hippocampal atrophy is preceded by degeneration of several brainstem and basal forebrain areas (Fig. 1). They include nucleus basalis of Meynert (NBM) and, sometimes, medial septum area (MSA) in the forebrain,75 –77 hypothalamic tuberomammillary nucleus (TMN),78 –80 LC, 81,82, 81,82 and raphe nuclei (RN) ( 83 and references therein) in the brainstem. These structures do not undergo complete degeneration but rather specific neuronal subpopulations fade away in them. In the basal forebrain, NBM loses cholinergic neurons resulting in acetylcholine deficit in the neocortex. If MSA is affected, this deficit may extend to the hippocampus and entorhinal cortex. TMN, LC, and RN are the only sources of histamine, noradrenaline, and serotonin in the brain, respectively, and degeneration of their neurons inevitably leads to deficit of histamine, noradrenaline, and serotonin in the forebrain. Therefore, attempts to compensate for these deficits were undertaken. However, among four mentioned areas providing cholinergic, histaminergic, noradrenergic, and serotonergic innervation, support of only cholinergic function gave some positive results, whereas the data on positive effects of modulators of the mentioned monoaminergic systems arecontradictory. 78

Fig. 1

Location of brain areas undergoing degeneration during progression of Alzheimer’s disease. The medial septum and nucleus basalis of Meynert are sources of cholinergic innervation of hippocampus and neocortex, respectively. The tuberomammaliary nucleus in the hypothalamus sends histaminergic projections throughout the entire brain. Locus coeruleus and raphe nuclei in the brainstem provide, respectively, noradrenergic and serotonergic innervation of the majority brain structures.

IMPAIRMENT OF THE CHOLINERGIC SYSTEM

It was shown that inhibition of acetylcholine esterase (AChE), an enzyme responsible for cleavage of acetylcholine in the extracellular space, can delay cognitive impairments in people with AD 75 and even reduce basal forebrain atrophy. 84 One of paradoxes related to positive effect of AChE inhibitors is that their positive effect may develop in all stages of AD, including cases where the majority of cholinergic axons is lost. The latter points to possibility that AChE inhibitors activate additional compensatory mechanisms not related to support of cholinergic function in the neocortex and hippocampus. It was proposed 75 that this compensatory mechanism may be enhanced involvement of striatal and/or thalamic cholinergic neurons, which remain intact in AD, 85 in supporting cognitive functions. There is also an alternative possibility, which is rarely discussed, because data are only available from rodent studies. It was shown that degeneration of cholinergic neurons either in NBM or MSA leads to a compensatory sprouting of noradrenergic sympathetic fibers from the superior cervical ganglion 86 (Fig. 2). Under normal conditions, these fibers innervate blood vessels in the neocortex or hippocampus; 87 however, loss of cholinergic innervation leads to sprouting of these fibers and shift in the released mediator – they become cholinergic.86,88, 86,88 Since degeneration of cholinergic fibers from the basal forebrain occurs slowly in AD, sprouting and transformation of sympathetic fibers may also slowly occur during AD and, at least, to some extent compensate loss of cholinergic forebrain innervation. However, the data on functioning and changes in the sympathetic ganglia and their axons during AD are lacking and it is not clear whether this sort of compensation may occur in human. In addition, it is also unclear whether function of noradrenergic sympathetic fibers is preserved in AD.

Fig. 2

Degeneration of the major sources of cholinergic projections, nucleus basalis of Meynert (nbM) and/or medial septum (MS), may induce transformation of cell phenotype in the superior cervical ganglion. Under normal conditions, sympathetic neurons of the superior cervical ganglion provide noradrenergic regulation of brain blood vessels but, after loss of cholinergic innervation of the neocortex and hippocampus, these neurons undergo transformation in the cholinergic cells which may partially compensate for the loss of cholinergic innervation.

There is also a highly speculative possibility about mechanism of positive effects of acetylcholine inhibitors. It is generally accepted that cholinergic innervation of the hippocampus and neocortex arises from the basal forebrain nuclei, and the neocortex and hippocampus do not possess intrinsic cholinergic interneurons, at least in humans. 89 However, the absence of cholinergic neurons in the neocortex and hippocampus is highly debated. The problem that appears here is that different research collectives using different antibodies either found or did not find neurons expressing choline acetyltransferase (ChAT) (enzyme responsible for synthesis of acetylcholine and obligatory marker of cholinergic neurons) in the mentioned brain areas (see introduction in 90 for discussion and references). In the mouse strains, where fluorescent protein (GFP) is expressed in cholinergic neurons, some hippocampal and cortical interneurons express GFP; however, some researchers suggest that it is an artifact of transgenesis 90 and ChAT-immunopositive neurons in the neocortex may not correspond to GFP-positive cells. 91 In our experience, the result of ChAT staining in neocortex depends on the antibodies used because different antibodies may give similar staining in the basal forebrain nuclei and striatum, however, give different staining in the neocortex. It is not clear what gives this difference in staining. In addition, it also worth to stress that several single-cell RNAseq studies showed that some neocortical/hippocampal interneurons may express ChAT mRNA but the level of ChAT transcripts in these neurons is 100 or more times lower than in the striatal neurons.23,24,65 , 23,24,65 From methodological point of view, this huge difference in the ChAT mRNA level in different neurons means that analysis of ChAT mRNA in the hippocampus or neocortex using in situ hybridization requires extremely sensitive approaches like single molecule in situ hybridization and previous approaches using standard in situ approach92,93, 92,93 may be not sensitive enough to detect ChAT-positive neurons in the hippocampus. The very low amount of ChAT mRNA in single cells raises concerns about the functional role of these transcripts as well as the amount of ChAT protein and acetylcholine produced by these ChAT-positive interneurons. It also worth to note that single cell RNA data of human neocortex in Allen brain atlas do not contain ChAT-expressing interneurons. However, it is not clear whether this absence results from limited sampling or reflects the real situation. Anyway, despite highly speculative possibility of functional compensation of acetylcholine deficit by these intrinsic hippocampal or cortical ChAT-positive interneurons, their existence and involvement in positive effects of AChE inhibitors remain absolutely unclear. Additionally, it has to be noted that the aforementioned positive effects of AChE inhibitors may be related to direct activation of NGF receptors. 94

An important aspect of cholinergic transmission is related to the fact that survival of the basal forebrain cholinergic neurons strongly depends on the availability of nerve growth factor (NGF) which activates p75 and TrkA receptors present in the adult forebrain only on cholinergic NBM and MSA neurons. Therefore, it was proposed that degeneration of cholinergic neurons may be prevented by local delivery of NGF to NBM (details on different approaches are presented in 75 ). It was shown that trophic support of NBM cholinergic neurons with NGF by implantation of cannula with cells secreting this growth factor helps to support cognitive functions of patients with AD.95 –97 It was hypothesized that loss of trophic support of NBM cholinergic neurons may result from disruption of NGF maturation in AD and therapeutic interventions addressing NGF maturation pathway may also have beneficial effects. 75

The basic principle underlying described approaches is supplementation of cholinergic neurons selectively in NBM and it is not clear to which extent they influence cholinergic terminals of NBM neurons in the neocortex. Anatomically, NBM is located distantly from the brain areas innervated by this nucleus and, under normal conditions, cholinergic NBM neurons receive NGF endocytosed by their terminals in these areas after axonal transport of NGF-receptor complex to soma. Hence, it is not clear whether trophic support of somas of cholinergic neurons is the most effective approach to support cholinergic function because currently the extent of preservation of cholinergic fibers in the neocortex after NGF therapy is not clear. The problem is that trophic support of NBM neurons in AD may be impaired by NFTs accumulated in their axons, which disrupts transport of NGF from terminals to soma and, finally, leads to disrupted trophic support of NBM neurons in AD and their degeneration. The positive effect of NGF supplementation to neuronal somas (if it is really related to maintenance of cholinergic transmission) suggests that it may maintain cholinergic function in distant areas of the neocortex which, probably, requires normalization of axonal transport and, probably, removal of NFTs in their axons.

Alternative modes of action of NGF are also possible. The axons coming to the hippocampus from the superior cervical ganglion, which can transform from noradrenergic phenotype to cholinergic one under conditions of cholinergic deficit in rodents, express NGF receptors. 98 Therefore, supplementation with NGF may lead to sprouting of sympathetic fibers in the NGF-enriched areas of the brain, acquisition of cholinergic phenotype by these fibers, and elevation of cholinergic tone. In addition, experiments with animals showed that overexpression of NGF in the hippocampus may induce positive effects without restoring cholinergic innervation.99,100, 99,100 Presumably, these effects of NGF may be mediated by non-classical SORCS3 receptors. 101

LOSS OF MONOAMINERGIC NEURONS

As we mentioned above, degeneration of histaminergic, serotonergic, and noradrenergic neurons is one of crucial steps in the AD development. 102 A single group of researchers also found degeneration of C1 neurons in the rostral ventrolateral medulla,103,104, 103,104 which express phenylethanolamine N-methyltransferase, an enzyme responsible for adrenaline synthesis. However, the events triggering degeneration of monoamine neurons are not clear.

Human data on histaminergic neurons in the hypothalamic tuberomammillary nucleus (TMN) and raphe serotonergic neurons in AD are very scarce and include predominantly data on aged patients with AD, in contrast to the studies on the locus coeruleus which include data on people of different ages. Therefore, it is really hard to make unambiguous conclusions about dynamics of degenerative changes in TMN and RN. It was shown that AD is associated with loss of histaminergic neurons and accumulation of NFTs in the survived neurons80,105, 80,105 but the age of manifestation of this pathological change is not clear. 79

For RN serotonergic neurons, it was hypothesized that degeneration starts with accumulation of NFTs in RN neurons at the early stages of AD.106,107, 106,107 However, AD progression is not associated with strong changes in the amount of p-tau inclusions in somas and dendrites of neurons in RN; 108 as for axons of these neurons, it remains obscure whether AD progression is associated with spreading of NFTs in serotonergic axons in the forebrain areas.

Several studies showed that degeneration in LC starts long before the onset of cortical atrophy 109 and before the onset of Aβ accumulation. 110 Current opinion is that accumulation of pretangle p-tau deposits in LC occurs more than 10 years before the onset of AD and may serve as a nucleus that disseminates tau tangles around cortical areas via axonal projections.28,110, 28,110 Moreover, it was proposed that Aβ accumulation in terminals of LC neurons in various brain areas may serve as seeds for Aβ spreading. 111

It was hypothesized that there may be several causes of susceptibility of LC noradrenergic neurons to pathological changes. First of all, the majority of LC axons are non-myelinated which reduces their resistance to tau propagation. 112 Second, the absence of myelination imposes large metabolic costs on the maintenance of non-myelinated axons which may reduce metabolic compensatory capacities of LC neurons which, taking into account specific features of noradrenaline metabolism, makes noradrenergic neurons very susceptible to various challenges. 113 Third, noradrenaline metabolism is mediated by monoamine oxidases (A and B), which are expressed in LC neurons. Activity of these enzymes leads to deamination of noradrenaline and formation of aldehyde 3,4-dihydroxyphenylglycolaldehyde (DOPEGAL) which is accompanied by H₂O₂ generation (Fig. 3). H₂O₂ under certain conditions may become a source of various ROS or lead to formation of highly reactive chemical products, and oxidative stress. Here, it is worth to mention that, in people with mild cognitive impairment and AD, the expression of SOD2 and GPX1 genes, whose protein products are responsible for cell protection against superoxide and H₂O₂, decreased in survived LC neurons 114 which provides high oxidative environment within aged LC neurons 115 and makes these neurons even more susceptible to ROS.

Fig. 3

Noradrenergic neurons in the locus coeruleus and superior cervical ganglion have similar metabolism when noradrenaline (norepinephrine) is metabolized by monoamine oxidases 1 and 2 (MAO) to 3,4-dihydroxyphenylglycolaldehydealdehyde (DOPEGAL) with formation of H₂O₂ which may result in the formation of radical form of DOPEGAL•leading to oxidative stress and enhanced formation of hyperphosphorylated tau (p-tau). The superior cervical ganglion innervates blood vessels in the brain, dura mater, and leptomeninges suggesting that p-tau accumulation and oxidative stress may occur not only in the brain parenchyma but also in the meninges.

As for DOPEGAL, it was shown that it is accumulated in LC neurons in AD 116 and, according to recent data, this product of noradrenaline metabolism may promote formation of tau tangles and tau hyperphosphorylation.117,118, 117,118 Importantly, H₂O₂ can react with DOPEGAL and form highly reactive radicals which can induce oxidative stress and damage mitochondria. 104 Accumulation of DOPEGAL per se in LC neurons and, possibly, their axons throughout the brain raises questions why the activity of aldehyde reductase, an enzyme responsible for reduction of DOPEGAL and its further metabolism, as well as enzymes responsible of H₂O₂ inactivation (described below) appears to be decreased. It was proposed that DOPEGAL accumulation results from impaired intracellular transport. 116 Presumably, local DOPEGAL accumulation leads to tau hyperphosphorylation, damage of axonal cytoskeleton and impairment of transport of protective enzymes, which leads to greater DOPEGAL accumulation, i.e., formation of vicious cycle leading to axon degeneration and, finally, neuronal death.

If we consider LC as one of points where AD starts, then detection of NFTs or, strictly speaking, tau pretangles in LC seems to be one of important steps in determination of age when AD development may emerge. However, studies in human suggest that tau pretangles in LC may be observed already at the age of 10 and by the age of 30 the majority of people have NFTs 119 or pretangles in LC.28,120, 28,120 These pathological changes were very rarely associated with Aβ accumulations in the cortical areas suggesting that NFT accumulation in LC may be one of primary morphological events that preclude development of sporadic AD. However, if NFT accumulation in LC neurons is a widespread phenomenon, it raises a question why AD do not develop in the majority of people and how brain withstands constant pressure of pretangle seeding by LC neurons in aged non-demented people.

Another important issue is related to the fact that the majority of postganglionic sympathetic neurons are noradrenergic. They are also non-myelinated and, in contrast to LC neurons, are not protected by BBB. Noradrenaline metabolism in these sympathetic neurons involves monoamine oxidase A 121 which inevitably results in DOPEGAL and H₂O₂ generation and development of neurotoxic consequences including NFT formation. As we mentioned above, superior cervical ganglion is a source of noradrenergic fibers that regulate tone of blood vessels in the brain and, if tau pretangles are formed in them, these axons may serve as another source of p-tau seeds (Fig. 4). Currently, it is not clear whether sympathetic fibers are prone to formation of NFTs and whether this mechanism of tau spreading via sympathetic projections is realized at least in some cases of AD. Occurrence of tau pretangles in LC is sometimes accounted for by strong vascularization of LC and, due to this stronger vascularization, higher probability of diffusion of some toxins to LC neurons from blood, however, if these blood vessels receive sympathetic innervation, p-tau may come from these noradrenergic sympathetic fibers whereas p-tau in these fibers may occur due to toxins that more easily passed from blood to the ganglia. It was proposed that various metals including iron, copper, zinc, lead, etc. may be slowly accumulated in neuromelanin of LC neurons 122 and potentially induce negative consequences. 123

Fig. 4

Scheme of hypothetical involvement of noradrenergic neurons from the superior cervical ganglion in the induction and progression of Alzheimer’s disease.

OXIDATIVE STRESS

The development of AD, like other neurodegenerative diseases, is associated with oxidative stress resulting in accumulation of oxidized metabolic products in the brain of patients with AD. 124 The oxidized products are accumulated due to overproduction of ROS and insufficient activity of antioxidant defense systems. The most commonly considered ROS are the superoxide anion radical ${\dot{O}}_{2}^{-}$ , the hydroxyl radical $O {\dot{H}}^{-}$ , and hydrogen peroxide. Although hydrogen peroxide is included in this list, it is a fairly stable compound which is formed after conversion of ${\dot{O}}_{2}^{-}$ by Cu/Zn-superoxide dismutase and Mn-superoxide dismutase.125,126, 125,126 These enzymes are expressed in the brain and play an important role in its antioxidant defense. As for H₂O₂, it may be either retained inside the cell or diffuse outside, and anywhere it can be a source of $O {\dot{H}}^{-}$ . The problem is that H₂O₂ can react with Fe² ⁺ /Fe^3 + and Cu⁺/Cu^2 + ions which leads to the generation of $O {\dot{H}}^{-}$ as a result of the Fenton reaction, 126 so peroxide removal is also an important task for cells. To degrade H₂O₂, brain cells use a number of enzymes, including catalase (expressed in the brain of both humans and mice), as well as glutathione peroxidases GPX1, GPX3 and GPX4, the latter being expressed most intensely in the brain (scRNAseq data from Allen brain atlas). In addition, the brain also expresses peroxiredoxins PRDX1-3, PRDX5, PRDX6, which are capable of inactivating peroxide, as well as auxiliary proteins thyroxins TXN1 and TXN2 and their reductase TXNRD1.

In addition to the Fenton reaction, ${\dot{O}}_{2}^{-}$ may be produced by a number of the following cellular enzymes (data on the expression are given according scRNAseq data from Allen brain atlas): xanthine oxidase (expressed only in brain vascular cells), lipoxygenases (most of them are practically not expressed by brain cells), cyclooxygenases (PTGS1 and PTGS2 are expressed by some neuronal subpopulations, as well as by brain microglia) and NADPH oxidases. The expression of genes encoding NADPH oxidase (NOX1, NOX3, NOX4, NOX5 (in humans), DUOX1, DUOX2) in the brain under normal conditions is very low and in some cases is practically undetectable.22–24,65 , 22–24,65 However, expression of the NADPH oxidase gene CYBB/NOX2, as well as an important component of its oxidase complex CYBA, was found in microglia in both mice and humans. 22 Despite the significant number of enzymes capable of generating ${\dot{O}}_{2}^{-}$ , in the majority of situations superoxide dismutases present in cells can effectively inactivate this oxygen radical, and the development of long-term oxidative stress in AD indicates some disruption in the mechanisms of defense against oxidative stress.

Recent studies point to curious twist in the story related to function of protein product of CYBB gene in AD. It was shown in previous studies that Aβ induces ROS generation in cultured astrocytes but not neurons and these ROS are generated by NADPH oxidase.17,18, 17,18 At that time, it was not clear which NADPH oxidase isoform was responsible for this effect. As we mentioned above, scRNAseq data suggest this can be only CYBB. This conclusion is supported by the data that CYBB expression and CYBB-dependent oxidative stress may be enhanced in astrocytes by Aβ application. 127 Moreover, it was shown that Aβ application leads to oxidative stress and glucose hypometabolism via activation of CYBB. 128 These findings suggest that one of possible pathways of disruption of astrocytic function (and, presumably, BBB functioning) in AD may be related to Aβ-induced oxidative stress and metabolic dysfunction selectively in astrocytes (Fig. 5). Therefore, support of astrocytic function by modulation of CYBB may be another possible approach to alleviation of oxidative stress and, possibly, metabolic disturbance in AD.

Fig. 5

Summary of events induced by Aβ in astrocytes. Aβ causes oxidative stress and mitochondrial dysfunction in astrocytes leading to disruption of functioning of the blood-brain barrier whereas accumulation of Aβ oligomers in perivascular spaces impedes interstitial fluid flow and cell metabolites. ROS, reactive oxygen species

As we mentioned above, free Fe² ⁺ /Fe^3 + ions may be a source of ROS generation and, therefore, regulation of level of free iron ions is one of possible ways to control ROS generation. Normally the level of iron ions in cells is regulated by buffer proteins. A significant portion of iron ions is bound by the proteins ferritin, transferrin, and ceruloplasmin where iron becomes non-reactive lacking the ability to participate in redox reactions. In AD, staining patterns for ferritin, transferrin, and iron are significantly altered in patients compared with staining observed in tissue samples collected from non-AD patients. An important aspect of these changes is that amyloid aggregates act as foci of these changes. Thus, transferrin was evenly distributed around senile plaques and most of the ferritin-positive microglia were located around senile plaques and vessels. Clusters of iron oxidation products were also detected near senile plaques and in cells associated with plaques. 129 It was also shown that aging is accompanied by the accumulation of iron in some parts of the brain, and in AD, iron ions become a component of senile plaques,130,131, 130,131 possibly, due to changes in the distribution of iron-binding proteins. Importantly, serum iron levels in AD remain unchanged or undergo a slight decrease132,133, 132,133 suggesting that the mentioned iron accumulation in some brain areas is not due to disruption of iron homeostasis at organismal level. Presumably, the mentioned redistribution of iron-binding proteins may be considered as an attempt of brain cells to bind free iron, which is intensively accumulated in the brain in AD. 129 The danger of free iron is related not only to H₂O₂ generation but also to the ability of free iron ions to bind Aβ 134 and induce its aggregation. It may be expected that this Aβ-mediated sequestration of iron ions may be have protective effect due to removal of toxic free iron. However, as discussed in the study, 130 the ability of iron ions in complex with Aβ to be involved in redox reactions and ROS generation depends on additional conditions and further studies are necessary to clarify the role of Fe-Aβ complexes in oxidative stress development in AD.

Similarly to Aβ, tau protein (both non- and hyperphosphorylated forms) can also interact with iron ions which leads to aggregation of tau or formation of tau-containing droplets135,136, 135,136 and iron chelation may reverse this process. 130

Since disruption of iron homeostasis and its excessive accumulation in the brain is one of the factors in AD, a number of authors attempted to treat the disease with metal chelating agents capable of forming stable compounds with a higher molecular weight, thereby converting metal ions into an inactive state. This approach is based on previous studies showing that iron and copper chelators, but not Ca^2 + and Mg^2 + (e.g., TPEN or α-lipoic acid), promote the dissolution of β-amyloid deposits from the brains of AD patients137,138, 137,138 and reverse aggregation of tau. 130

Deferoxamine has been proposed as a chelator capable of reducing iron accumulation. When deferoxamine was administered intranasally to mice with model AD, it led to memory restoration, and suppression of amyloidogenic processing of AβPP 139 and phosphorylation of the tau protein. 140 However, the bioavailability of deferoxamine is low, the size and hydrophilicity of the drug do not allow it to freely pass through the BBB 141 and it is currently proposed to be administered intranasally. 142 Despite the fact that the bioavailability of deferoxamine is quite low, clinical trials have shown its effectiveness in slowing the AD progression compared to placebo. 143 This result served as the basis for further clinical trials of iron chelators in AD, and to date, the second phase of clinical trials of another chelator deferiprone (ClinicalTrials.gov Identifier: NCT03234686) has ended but the results have not yet been published.

It was also proposed to use the protein lactoferrin as another iron chelator, the effectiveness of which was shown in mice with a model of AD. The lactoferrin molecule crosses the blood-brain barrier well, reducing Aβ deposition and improving cognitive function in mice with an AD model. 144 However, parallel studies in humans revealed an unexpected turn— a significant, although not very pronounced, increase in lactoferrin levels in the brain correlates with the development of AD and may be a predictor of the development of amyloidosis. 145 It was hypothesized that lactoferrin may also have protective function by sequestering excess extracellular iron (as suggested for the other iron chelators mentioned above) but also shift AβPP processing toward Aβ and activate the immune response. The latter is very important due to the fact that there is an assumption that AD is a consequence of microbial invasion 146 and Aβ is a toxin that prevents the spread of this invasion due to the presence of microbicidal properties, including the ability to generate ROS after binding to iron. 147

It was proposed that iron accumulation (especially Fe^2 + accumulation) in cells may be a result of Zn^2 + accumulation since extracellular zinc ions inhibit Fe² ⁺ oxidation to Fe^3 + by AβPP. 148 The role of zinc ions in AD is intensely studied and it was shown that zinc binds Aβ which results in formation of Aβ oligomers. 149 However, in contrast to iron, the data on zinc accumulation in the brain of AD patients is controversial (discussed in 149 ). Initially, it was proposed that administration of Zn^2 + chelator clioquinol may have beneficial effect in AD, 150 however, repeated analysis did not confirm initial optimistic interpretation. 151 Furthermore, the concept of zinc accumulation in the brain in AD is contrasted with the alternative idea that a deficiency in zinc could also be a contributing factor to the AD progression. Dietary zinc supplementation was proposed as a mean to ameliorate AD pathology, however, there is still no consensus on the efficacy of this approach. 149

OTHER CONSIDERATIONS

We mentioned above that slow decline of histaminergic, serotonergic, noradrenergic, and cholinergic neurons inevitably leads to loss of cognitive functions. It seems that slow degeneration of neurons of the mentioned subcortical signaling systems is somehow compensated during initial stages of AD and, currently, we do not know which mechanisms become activated during this degeneration. One of possibilities is that spreading of NFTs from LC or RN leads to seeding of NFTs in cortical and other areas and, presumably, impaired functioning of LC and RN axonal transport which results in axonal dysfunction and loss of axon collaterals. Axonal degeneration will inevitably lead to activation of microglia and inflammatory response. It is not clear whether this inflammation resulting from dying axons may serve as initiator of chronic neuroinflammation. In addition to inflammatory response, degeneration of the mentioned axons will be associated with loss of functions related to dying axons, i.e. deficit of histamine, serotonin, noradrenaline, and acetylcholine. The predominant focus on cholinergic system as a single system “important” for cognitive decline in AD and therapy implies that other signaling systems are “useless” or may be important only for correction of sleep or mood disorders. This point of view neglects complexity of AD and limits experimental studies addressing interaction of the mentioned systems. Clearly, the mentioned and other yet to be determined signaling systems interact under normal conditions and their interaction may be strongly distorted during degeneration of some of these systems. The latter means that compensatory changes that develop during loss of some signaling systems will result in overactivation of one set and hypoactivation of another set of signaling systems. This new homeostatic balance will be more sensitive to stimuli that can cause overload of already overactivated systems and promote their degeneration. Unfortunately, we still do not know how brain cells cope with loss of axons providing the mentioned mediators. Our studies in rats showed that loss of cholinergic innervation of the hippocampus after death of MS neurons does not induce significant transcriptomic changes. 152 One possible interpretation is that cholinergic innervation does not regulate expression of genes in hippocampal cells which looks quite paradoxical because cholinergic system is associated with memory formation which requires gene expression. An alternative and, probably, more real variant is that gene-regulating function of cholinergic innervation is executed by some system that underwent compensatory changes. Currently, it is not clear which systems are involved this “silent” compensation. From therapeutic viewpoint, this sort of compensation complicates development of approaches that would help to stop or reverse disease progression because therapeutic intervention should not only recover lost function but also return functioning of other signaling systems to normal state.

CONCLUDING REMARKS

AD is a slowly progressing disease which hypothetically occurs in subcortical areas and spreads to higher brain areas. The spreading of disease predominantly includes accumulation of Aβ plaques and NFTs in the neocortical areas and degeneration of some subcortical neuronal subpopulations.

AD is one of forms of dementia which is currently intensely studied and, despite huge efforts, very weak success is achieved in the development of therapy and cure of this disease. The critical factors for AD progression that were raised in our review are summarized in Fig. 6. The fundamental problem standing in front of scientific community is that there is no clear understanding of mechanisms that trigger AD and become activated in later stages. Currently, it looks like we have put some pieces of the puzzle together but we still do not know whether the assembled pieces are related to the same picture.

Fig. 6

Summary of factors involved in progression of Alzheimer’s disease. Solid lines indicate direct involvement of some factors in the disease progression. Dotted lines point to hypothetical influence of factors on each other or disease progression.

Studies of postmortem human brain gave rise to the strong hypothesis on the primary role of tauopathy spreading from noradrenergic LC neurons. However, this hypothesis raises many concerns about various aspects of p-tau spreading and triggers that can induce tau hyperphosphorylation. First of all, as we mentioned above, it is not clear whether noradrenergic sympathetic neurons and, more specifically, neurons of the superior cervical ganglion can be alternative source of p-tau seeding and, if not, which factors limit their involvement in this process. Second, p-tau seeding occurs around entire brain, however, spreading of NFTs follows a specific pattern known as Braak stages and include predominantly cortical areas whereas other brain areas resist NFT formation and the mechanism behind this resistance is unclear. Third, the role played by Aβ spreading is unclear because progression of cognitive impairments is associated with p-tau spreading and, to lesser extent with formation of Aβ-containing senile plaques. However, formation of plaques seems to be a secondary event after some trigger when Aβ, which is present in healthy brain, start to form oligomers and plaques. It seems that p-tau seeding may be one of the factors inducing Aβ pathology, however, some additional factor(s) is needed because p-tau seeding does not cause Aβ pathology immediately and neurons of various subcortical areas resist formation of senile plaques despite the presence of p-tau seeds. One of possible variants is that Aβ-containing plaques are formed in “traffic jams” in neurites of dying cells due to accumulation of NFTs. 28

Our review also shows that one part may be played Aβ-induced activation of NAPDH oxidase 2 followed by oxidative stress and impairment of glucose metabolism, presumably, in astrocytes. However, it is not clear which factors limit activation of this cascade under normal conditions because Aβ is present in the brain and, probably, can activate the mentioned cascade under normal conditions. Presumably, the major factor in this case can be intensity of Aβ formation and its concentration in specific area but it is not clear which factors provoke accumulation and/or arrest of metabolism/washout of Aβ.

AUTHOR CONTRIBUTIONS

Alexey P. Bolshakov (Conceptualization; Supervision; Writing – original draft; Writing – review & editing); Konstantin Gerasimov (Funding acquisition; Visualization; Writing – original draft); Yulia V. Dobryakova (Project administration; Writing – original draft; Writing – review & editing).

Footnotes

ACKNOWLEDGMENTS

Figures were made using Biorender software.

FUNDING

This research was funded by the Russian Science Foundation (RSF), grant number 23-75-30023

CONFLICT OF INTEREST

The authors have no conflict of interest to report.

References

Lambert

, Ramirez

, Grenier-Boley

, et al. Step by step: Towards a better understanding of the genetic architecture of Alzheimer’s disease. Mol Psychiatry 2023; 28: 2716–2727.

Jucker

and Walker

. Alzheimer’s disease: From immunotherapy to immunoprevention. Cell 2023; 186: 4260–4270.

Hampel

, Hardy

, Blennow

, et al. The amyloid-β pathway in Alzheimer’s disease. Mol Psychiatry 2021; 26: 5481–5503.

Walker

. Aβ plaques. Free Neuropathol 2020; 1: 1–31.

Basheer

, Smolek

, Hassan

, et al. Does modulation of tau hyperphosphorylation represent a reasonable therapeutic strategy for Alzheimer’s disease? From preclinical studies to the clinical trials. Mol Psychiatry 2023; 28: 2197–2214.

Thal

, Rüb

, Orantes

, et al. Phases of A beta-deposition in the human brain and its relevance for the development of AD. Neurology 2002; 58: 1791–1800.

Zhang

, Sloan

, Clarke

, et al. Purification and characterization of progenitor and mature human astrocytes reveals transcriptional and functional differences with mouse. Neuron 2016; 89: 37–53.

Area-Gomez

and Schon

. Towards a unitary hypothesis of Alzheimer’s disease pathogenesis. J Alzheimers Dis 2024; 98: 1243–1275.

Sadleir

, Kandalepas

, Buggia-Prévot

, et al. Presynaptic dystrophic neurites surrounding amyloid plaques are sites of microtubule disruption, BACE1 elevation, and increased Aβ generation in Alzheimer’s disease. Acta Neuropathol 2016; 132: 235–256.

10.

Roher

, Lowenson

, Clarke

, et al. β-amyloid-(1–42) is a major component of cerebrovascular amyloid deposits: Implications for the pathology of Alzheimer disease. Proc Natl Acad Sci U S A 1993; 90: 10836–10840.

11.

Kozin

, Kechko

, Adzhubei

, et al. Switching on/off amyloid plaque formation in transgenic animal models of Alzheimer’s disease. Int J Mol Sci 2024; 25: 72.

12.

Semyachkina-Glushkovskaya

, Postnov

, Khorovodov

, et al. Lymphatic drainage system of the brain: A new player in neuroscience. J Evol Biochem Physiol 2023; 59: 1–19.

13.

Greenberg

, Bacskai

, Hernandez-Guillamon

, et al. Cerebral amyloid angiopathy and Alzheimer disease - one peptide, two pathways. Nat Rev Neurol 2020; 16: 30–42.

14.

Palmqvist

, Zetterberg

, Mattsson

, et al. Detailed comparison of amyloid PET and CSF biomarkers for identifying early Alzheimer disease. Neurology 2015; 85: 1240.

15.

Nortley

, Korte

, Izquierdo

, et al. Amyloid b oligomers constrict human capillaries in Alzheimer’s disease via signaling to pericytes. Science 2019; 365: eaav9518.

16.

Chen

, Guo

, Zhu

, et al. Amyloid β oligomer induces cerebral vasculopathy via pericyte-mediated endothelial dysfunction. Alzheimers Res Ther 2024; 16: 56.

17.

Abramov

, Canevari

and Duchen

. Beta-amyloid peptides induce mitochondrial dysfunction and oxidative stress in astrocytes and death of neurons through activation of NADPH oxidase. J Neurosci 2004; 24: 565–575.

18.

Abramov

, Canevari

and Duchen

. Changes in intracellular calcium and glutathione in astrocytes as the primary mechanism of amyloid neurotoxicity. J Neurosci 2003; 23: 5088–5095.

19.

Abramov

, Ionov

, Pavlov

, et al. Membrane cholesterol content plays a key role in the neurotoxicity of β-amyloid: Implications for Alzheimer’s disease. Aging Cell 2011; 10: 595–603.

20.

Tyurikova

, Zheng

, Rings

, et al. Monitoring Ca2+elevations in individual astrocytes upon local release of amyloid beta in acute brain slices. Brain Res Bull 2018; 136: 85–90.

21.

Kuchibhotla

K V

, Lattarulo

, Hyman

et al. Synchronous hyperactivity and intercellular calcium waves in astrocytes in Alzheimer mice. Science 2009; 323: 1211.

22.

Zhang

, Chen

, Sloan

, et al. An RNA-sequencing transcriptome and splicing database of glia, neurons, and vascular cells of the cerebral cortex. J Neurosci 2014; 34: 11929–11947.

23.

Saunders

, Macosko

, Wysoker

, et al. Molecular diversity and specializations among the cells of the adult mouse brain. Cell 2018; 174: 1015–1030.e16.

24.

Zeisel

, Muñoz-Manchado

, Codeluppi

, et al. Cell types in the mouse cortex and hippocampus revealed by single-cell RNA-seq. Science 2015; 347: 1138–1142.

25.

Ryman

, Acosta-Baena

, Aisen

, et al. Symptom onset in autosomal dominant Alzheimer disease: A systematic review and meta-analysis. Neurology 2014; 83: 253.

26.

Busche

and Hyman

. Synergy between amyloid-β and tau in Alzheimer’s disease. Nat Neurosci 2020; 23: 1183–1193.

27.

Braak

and Del Tredici

. Reply: The early pathological process in sporadic Alzheimer’s disease. Acta Neuropathol 2013; 126: 615–618.

28.

Arnsten

AFT

, Datta

, Del Tredici

, et al. Hypothesis: Tau pathology is an initiating factor in sporadic Alzheimer’s disease. Alzheimers Dement 2021; 17: 115.

29.

González

, Calfío

, Churruca

, et al. Glucose metabolism and AD: Evidence for a potential diabetes type 3. Alzheimer’s Res Ther 2022; 14: 56.

30.

Popova

and Stepanichev

. Cell cycle induction, amyloid-beta, and free radicals in the mechanisms of neurodegenerative process progression in the brain. Neurochem J 2008; 2: 146–152.

31.

Sayas

and Ávila

. GSK-3 and tau: A key duet in Alzheimer’s disease. Cells 2021; 10: 721.

32.

Toral-Rios

, Pichardo-Rojas

, Alonso-Vanegas

, et al. GSK3β and tau protein in Alzheimer’s disease and epilepsy. Front Cell Neurosci 2020; 14: 485621.

33.

Jackson

, Guerrero-Muñoz

, Castillo-Carranza

. The prion-like transmission of tau oligomers via exosomes. Front Aging Neurosci 2022; 14: 974414.

34.

Congdon

, Ji

, Tetlow

, et al. Tau-targeting therapies for Alzheimer disease: Current status and future directions. Nat Rev Neurol 2023; 19: 715–736.

35.

Schmidt

, Lee

and Trojanowski

. Analysis of epitopes shared by Hirano bodies and neurofilament proteins in normal and Alzheimer’s disease hippocampus. Lab Invest 1989; 60: 513–522.

36.

Hirano

. Hirano bodies and related neuronal inclusions. Neuropathol Appl Neurobiol 1994; 20: 3–11.

37.

Maselli

, Davis

, Furukawa

, et al. Formation of Hirano bodies in Dictyostelium and mammalian cells induced by expression of a modified form of an actin-crosslinking protein. J Cell Sci 2002; 115: 1939–1949.

38.

Munoz

, Wang

, Greenberg

. Hirano bodies accumulate C-terminal sequences of β-amyloid precursor protein (β-APP) epitopes. J Neuropathol Exp Neurol 1993; 52: 14–21.

39.

Furgerson

, Fechheimer

and Furukawa

. Model Hirano bodies protect against tau-independent and tau-dependent cell death initiated by the amyloid precursor protein intracellular domain. PLoS One 2012; 7: e44996.

40.

Tian

, An

, Kitner-Triolo

, et al. Associations of olfaction with longitudinal trajectories of brain volumes and neuropsychological function in older adults. Neurology 2023; 100: E964–E974.

41.

Tian

, Bilgel

, Moghekar

, et al. Olfaction, cognitive impairment, and PET biomarkers in community-dwelling older adults. J Alzheimers Dis 2022; 86: 1275–1285.

42.

Murphy

. Olfactory and other sensory impairments in Alzheimer disease. Nat Rev Neurol 2019; 15: 11–24.

43.

Masurkar

and Devanand

. Olfactory dysfunction in the elderly: Basic circuitry and alterations with normal aging and Alzheimer’s disease. Curr Geriatr Rep 2014; 3: 91–100.

44.

Guo

, Dove

, Wang

, et al. Trajectories of olfactory identification preceding incident mild cognitive impairment and dementia: A longitudinal study. EBioMedicine 2023; 98: 104862.

45.

O’Leary

, Stover

, Mantolino

, et al. Intact olfactory memory in the 5xFAD mouse model of Alzheimer’s disease from 3 to 15 months of age. Behav Brain Res 2020; 393: 112731.

46.

Sakai

, Kazui

, Shigenobu

, et al. Gustatory dysfunction as an early symptom of semantic dementia. Dement Geriatr Cogn Dis Extra 2017; 7: 395.

47.

Kouzuki

, Ichikawa

, Shirasagi

, et al. Detection and recognition thresholds for five basic tastes in patients with mild cognitive impairment and Alzheimer’s disease dementia. BMC Neurol 2020; 20: 110.

48.

Molitor

, Ko

and Ally

. Eye movements in Alzheimer’s disease. J Alzheimers Dis 2015; 44: 1.

49.

Wei

, Oh

, Harun

, et al. Vestibular loss predicts poorer spatial cognition in patients with Alzheimer’s disease. J Alzheimers Dis 2018; 61: 995–1003.

50.

Huang

, Xu

, Zhang

, et al. Research progress on vestibular dysfunction and visual– spatial cognition in patients with Alzheimer’s disease. Front Aging Neurosci 2023; 15: 1153918.

51.

Zhang

, Chen

, Zhang

, et al. Tactile angle discrimination decreases due to subjective cognitive decline in Alzheimer’s disease. Curr Alzheimer Res 2020; 17: 168–176.

52.

Löffler

, Beier

, Bekrater-Bodmann

, et al. Reduced tactile sensitivity is associated with mild cognitive impairment. eBioMedicine 2024; 99: 104896.

53.

Freedman

and Oscar-Berman

. Tactile discrimination learning deficits in Alzheimer’s and Parkinson’s diseases. Arch Neurol 1987; 44: 394–398.

54.

Fletcher

, Downey

, Golden

, et al. Pain and temperature processing in dementia: A clinical and neuroanatomical analysis. Brain 2015; 138: 3360.

55.

Agrawal

, Smith

and Rosenberg

. Vestibular impairment, cognitive decline and Alzheimer’s disease: Balancing the evidence. Aging Ment Health 2020; 24: 705.

56.

Minoshima

, Cross

, Thientunyakit

, et al. 18F-FDG PET imaging in neurodegenerative dementing disorders: Insights into subtype classification, emerging disease categories, and mixed dementia with copathologies. J Nucl Med 2022; 63: 2S–12S.

57.

, Varma

, et al. Evidence for brain glucose dysregulation in Alzheimer’s disease. Alzheimers Dement 2018; 14: 318–329.

58.

Hammond

, Xing

, Wang

, et al. β-amyloid and tau drive early Alzheimer’s disease decline while glucose hypometabolism drives late decline. Commun Biol 2020; 3: 352.

59.

Huang

. A comprehensive definition for metabolic syndrome. Dis Model Mech 2009; 2: 231–237.

60.

de la Monte

. Insulin resistance and Alzheimer’s disease. BMB Rep 2009; 42: 475–481.

61.

Mei

, Liu

, Mei

, et al. Sphingolipid metabolism in brain insulin resistance and neurological diseases. Front Endocrinol (Lausanne) 2023; 14: 1243132.

62.

Terzo

, Amato

, Mulè

. From obesity to Alzheimer’s disease through insulin resistance. J Diabetes Complications 2021; 35: 108026.

63.

Arnold

, Arvanitakis

, Macauley-Rambach

, et al. Brain insulin resistance in type 2 diabetes and Alzheimer disease: Concepts and conundrums. Nat Rev Neurol 2018; 14: 168–181.

64.

Beletskiy

, Positselskaya

, Vinarskaya

, et al. Detailed analysis of dorsal-ventral gradients of gene expression in the hippocampus of adult rats. Int J Mol Sci 2022; 23: 9948.

65.

Zeisel

, Hochgerner

, Lönnerberg

, et al. Molecular architecture of the mouse nervous system. Cell 2018; 174: 999–1014.e22.

66.

Gião

, Teixeira

, Almeida

, et al. Choroid plexus in Alzheimer’s disease— the current state of knowledge. Biomedicines 2022; 10: 224.

67.

Liddelow

. Development of the choroid plexus and blood-CSF barrier. Front Neurosci 2015; 9: 123479.

68.

Lee

, Tong

, Hang

, et al. CSF and brain indices of insulin resistance, oxidative stress and neuro-inflammation in early versus late Alzheimer’s disease. J Alzheimers Dis Parkinsonism 2013; 3: 128.

69.

Liu

, Liu

, Grundke-Iqbal

, et al. Deficient brain insulin signalling pathway in Alzheimer’s disease and diabetes. J Pathol 2011; 225: 54–62.

70.

Tramutola

, Triplett

, Di Domenico

, et al. Alteration of mTOR signaling occurs early in the progression of Alzheimer disease (AD): Analysis of brain from subjects with pre-clinical AD, amnestic mild cognitive impairment and late-stage AD. J Neurochem 2015; 133: 739–749.

71.

Dierssen

, Fructuoso

, Martínez de Lagrán

, et al. Down syndrome is a metabolic disease: Altered insulin signaling mediates peripheral and brain dysfunctions. Front Neurosci 2020; 14: 640.

72.

Gabbouj

, Ryhänen

, Marttinen

, et al. Altered insulin signaling in Alzheimer’s disease brain - special emphasis on PI3K-Akt pathway. Front Neurosci 2019; 13: 629.

73.

de la Monte

. Insulin resistance and neurodegeneration: Progress towards the development of new therapeutics for Alzheimer’s disease. Drugs 2017; 77: 47–65.

74.

Poulakis

, Pereira

, Muehlboeck

, et al. Multi-cohort and longitudinal Bayesian clustering study of stage and subtype in Alzheimer’s disease. Nat Commun 2022; 13: 4566.

75.

Hampel

, Mesulam

, Cuello

, et al. Revisiting the cholinergic hypothesis in Alzheimer’s disease: Emerging evidence from translational and clinical research. J Prev Alzheimers Dis 2018; 6: 2–15.

76.

Shi

, Cui

, Sun

, et al. Exploring sexual dimorphism in basal forebrain volume changes during aging and neurodegenerative diseases. iScience 2024; 27: 109041.

77.

Kilimann

, Grothe

, Heinsen

, et al. Subregional basal forebrain atrophy in Alzheimer’s disease: A multicenter study. J Alzheimers Dis 2014; 40: 687–700.

78.

Trillo

, Das

, Hsieh

, et al. Ascending monoaminergic systems alterations in Alzheimer’s disease. translating basic science into clinical care. Neurosci Biobehav Rev 2013; 37: 1363–1379.

79.

Satpati

, Neylan

, Grinberg

, et al. Histaminergic neurotransmission in aging and Alzheimer’s disease: A review of therapeutic opportunities and gaps. Alzheimers Dement (N Y) 2023; 9: e12379.

80.

Nakamura

, Takemura

, Ohnishi

, et al. Loss of large neurons and occurrence of neurofibrillary tangles in the tuberomammillary nucleus of patients with Alzheimer’s disease. Neurosci Lett 1993; 151: 196–199.

81.

Zarow

, Lyness

, Mortimer

, et al. Neuronal loss is greater in the locus coeruleus than nucleus basalis and substantia nigra in Alzheimer and Parkinson diseases. Arch Neurol 2003; 60: 337–341.

82.

Tomlinson

, Irving

and Blessed

. Cell loss in the locus coeruleus in senile dementia of Alzheimer type. J Neurol Sci 1981; 49: 419–428.

83.

Aletrino

, Vogels

OJM

, Van Domburg

PHMF

, et al. Cell loss in the nucleus raphes dorsalis in Alzheimer’s disease. Neurobiol Aging 1992; 13: 461–468.

84.

Cavedo

, Grothe

, Colliot

, et al. Reduced basal forebrain atrophy progression in a randomized Donepezil trial in prodromal Alzheimer’s disease. Sci Rep 2017; 7: 11706.

85.

Kotagal

, Müller

MLTM

, Kaufer

, et al. Thalamic cholinergic innervation is spared in Alzheimer disease compared to Parkinsonian disorders. Neurosci Lett 2012; 514: 169.

86.

Scheiderer

, McCutchen

, Thacker

, et al. Sympathetic sprouting drives hippocampal cholinergic reinnervation that prevents loss of a muscarinic receptor-dependent long-term depression at CA3-CA1 synapses. J Neurosci 2006; 26: 3745–56.

87.

Crutcher

. Aging and neuronal plasticity: Lessons from a model. Auton Neurosci 2002; 96: 25–32.

88.

Nelson

, Kolasa

, McMahon

. Noradrenergic sympathetic sprouting and cholinergic reinnervation maintains non-amyloidogenic processing of AβPP. J Alzheimers Dis 2014; 38: 867–879.

89.

Mesulam

, Hersh

, Mash

, et al. Differential cholinergic innervation within functional subdivisions of the human cerebral cortex: A choline acetyltransferase study. J Comp Neurol 1992; 318: 316–328.

90.

Blusztajn

and Rinnofner

. Intrinsic cholinergic neurons in the hippocampus: Fact or artifact? Front Synaptic Neurosci 2016; 8: 6.

91.

Gamage

, Zaborszky

, Münch

, et al. Evaluation of eGFP expression in the ChAT-eGFP transgenic mouse brain. BMC Neurosci 2023; 24: 4.

92.

, Woolf

, Roghani

, et al. Cholinergic neurons in the rat central nervous system demonstrated by in situ hybridization of choline acetyltransferase mRNA. Neuroscience 1992; 47: 807–822.

93.

Lauterborn

, Isackson

, Montalvo

, et al. In situ hybridization localization of choline acetyltransferase mRNA in adult rat brain and spinal cord. Brain Res Mol Brain Res 1993; 17: 59–69.

94.

Autio

, Mätlik

, Rantamäki

, et al. Acetylcholinesterase inhibitors rapidly activate Trk neurotrophin receptors in the mouse hippocampus. Neuropharmacology 2011; 61: 1291.

95.

Eyjolfsdottir

, Koenig

, Karami

, et al. Fast alpha activity in EEG of patients with Alzheimer’s disease is paralleled by changes in cognition and cholinergic markers during encapsulated cell biodelivery of nerve growth factor. Front Aging Neurosci 2022; 14: 756687.

96.

Ding

, Li

, Geetha

, et al. Nerve growth factor in metabolic complications and Alzheimer’s disease: Physiology and therapeutic potential. Biochim Biophys Acta Mol Basis Dis 2020; 1866: 165858.

97.

Mitra

, Behbahani

, Eriksdotter

. Innovative therapy for Alzheimer’s disease-with focus on biodelivery of NGF. Front Neurosci 2019; 13: 38.

98.

Valle-Leija

, Cancino-Rodezno

, Sánchez-Tafolla

, et al. Presence of functional neurotrophin TrkB receptors in the rat superior cervical ganglion. Front Physiol 2017; 8: 474.

99.

Dobryakova

, Spivak

, Zaichenko

, et al. Intrahippocampal adeno-associated virus-mediated overexpression of nerve growth factor reverses 192IgG-saporin-induced impairments of hippocampal plasticity and behavior. Front Neurosci 2021; 15: 745050.

100.

Volobueva

, Dobryakova

, Bolshakov

, et al. Protective effect of nerve growth factor in the hippocampus against disturbances in synaptic plasticity caused by cholinergic deficit. Neurochem J 2022; 16: 147–154.

101.

Salasova

, Monti

, Andersen

, et al. Finding memo: Versatile interactions of the VPS10p-Domain receptors in Alzheimer’s disease. Mol Neurodegener 2022; 17: 74.

102.

Lyness

, Zarow

and Chui

. Neuron loss in key cholinergic and aminergic nuclei in Alzheimer disease: A meta-analysis. Neurobiol Aging 2003; 24: 1–23.

103.

Burke

, Galvin

, Chung

, et al. Degenerative changes in epinephrine tonic vasomotor neurons in Alzheimer’s disease. Brain Res 1994; 661: 35–42.

104.

Burke

, Li

, Chung

, et al. Neurotoxicity of MAO metabolites of catecholamine neurotransmitters: Role in neurodegenerative diseases. Neurotoxicology 2004; 25: 101–115.

105.

Shan

, Bossers

, Unmehopa

, et al. Alterations in the histaminergic system in Alzheimer’s disease: A postmortem study. Neurobiol Aging 2012; 33: 2585–2598.

106.

Grinberg

, Rüb

, Ferretti

REL

, et al. The dorsal raphe nucleus shows phospho-tau neurofibrillary changes before the transentorhinal region in Alzheimer’s disease. A precocious onset? Neuropathol Appl Neurobiol 2009; 35: 406–416.

107.

Rüb

, Del Tredici

, Schultz

, et al. The evolution of Alzheimer’s disease-related cytoskeletal pathology in the human raphe nuclei. Neuropathol Appl Neurobiol 2000; 26: 553–567.

108.

Ehrenberg

, Nguy

, Theofilas

, et al. Quantifying the accretion of hyperphosphorylated tau in the locus coeruleus and dorsal raphe nucleus: The pathological building blocks of early Alzheimer’s disease. Neuropathol Appl Neurobiol 2017; 43: 393–408.

109.

Theofilas

, Ehrenberg

, Dunlop

, et al. Locus coeruleus volume and cell population changes during Alzheimer’s disease progression: A stereological study in human postmortem brains with potential implication for early-stage biomarker discovery. Alzheimers Dement 2017; 13: 236.

110.

Matchett

, Grinberg

, Theofilas

, et al. The mechanistic link between selective vulnerability of the locus coeruleus and neurodegeneration in Alzheimer’s disease. Acta Neuropathol 2021; 141: 631.

111.

Muresan

and Muresan

. Seeding neuritic plaques from the distance: A possible role for brainstem neurons in the development of Alzheimer’s disease pathology. Neurodegener Dis 2008; 5: 250–253.

112.

Rubinski

, Franzmeier

, Dewenter

, et al. Higher levels of myelin are associated with higher resistance against tau pathology in Alzheimer’s disease. Alzheimers Res Ther 2022; 14: 139.

113.

Braak

, Del Tredici

, Schultz

, et al. Vulnerability of select neuronal types to Alzheimer’s disease. Ann N Y Acad Sci 2000; 924: 53–61.

114.

Kelly

, He

, Perez

, et al. Locus coeruleus cellular and molecular pathology during the progression of Alzheimer’s disease. Acta Neuropathol Commun 2017; 5: 8.

115.

Kelly

, Nelson

and Counts

. Pontine arteriolosclerosis and locus coeruleus oxidative stress differentiate resilience from mild cognitive impairment in a clinical pathologic cohort. J Neuropathol Exp Neurol 2021; 80: 325.

116.

Burke

, Li

, Schmitt

, et al. Accumulation of 3,4-dihydroxyphenylglycolaldehyde, the neurotoxic monoamine oxidase A metabolite of norepinephrine, in locus ceruleus cell bodies in Alzheimer’s disease: Mechanism of neuron death. Brain Res 1999; 816: 633–637.

117.

Kang

, Meng

, Zhang

, et al. Tau modification by the norepinephrine metabolite DOPEGAL stimulates its pathology and propagation. Nat Struct Mol Biol 2022; 29: 292.

118.

Kang

, Liu

, Ahn

, et al. Norepinephrine metabolite DOPEGAL activates AEP and pathological Tau aggregation in locus coeruleus. J Clin Invest 2019; 130: 422.

119.

Elobeid

, Soininen

and Alafuzoff

. Hyperphosphorylated tau in young and middle-aged subjects. Acta Neuropathol 2012; 123: 97–104.

120.

Braak

and Del Tredici

. The pathological process underlying Alzheimer’s disease in individuals under thirty. Acta Neuropathol 2011; 121: 171–181.

121.

Shigematsu

, Akiguchi

, Oka

, et al. Monoamine oxidase-containing nerve fibers in the major cerebral arteries of rats. Brain Res 1989; 497: 21–29.

122.

Bakulski

, Seo

, Hickman

, et al. Heavy metals exposure and Alzheimer’s disease and related dementias. J Alzheimers Dis 2020; 76: 1215.

123.

Capucciati

, Zucca

, Monzani

, et al. Interaction of neuromelanin with xenobiotics and consequences for neurodegeneration; promising experimental models. Antioxidants 2021; 10: 824.

124.

Smith

, Carney

, Starke-Reed

, et al. Excess brain protein oxidation and enzyme dysfunction in normal aging and in Alzheimer disease. Proc Natl Acad Sci U S A 1991; 88: 10540–10543.

125.

Phaniendra

, Jestadi

and Periyasamy

. Free radicals: Properties, sources, targets, and their implication in various diseases. Indian J Clin Biochem 2015; 30: 11.

126.

Andrés

CMC

, Pérez de la Lastra

, Juan

, et al. Chemistry of hydrogen peroxide formation and elimination in mammalian cells, and its role in various pathologies. Stress 2022; 2: 256–274.

127.

Wyssenbach

, Quintela

, Llavero

, et al. Amyloid β-induced astrogliosis is mediated by β1-integrin via NADPH oxidase 2 in Alzheimer’s disease. Aging Cell 2016; 15: 1140–1152.

128.

Malkov

, Popova

, Ivanov

, et al. Aβ initiates brain hypometabolism, network dysfunction and behavioral abnormalities via NOX2-induced oxidative stress in mice. Commun Biol 2021; 4: 1054.

129.

Connor

, Menzies

, St. Martin

, et al. A histochemical study of iron, transferrin, and ferritin in Alzheimer’s diseased brains. J Neurosci Res 1992; 31: 75–83.

130.

Liu

, Fan

, Yang

, et al. Iron and Alzheimer’s disease: From pathogenesis to therapeutic implications. Front Neurosci 2018; 12: 632.

131.

Smith

, Harris

PLR

, Sayre

, et al. Iron accumulation in Alzheimer disease is a source of redox-generated free radicals. Proc Natl Acad Sci U S A 1997; 94: 9866.

132.

Wang

, Tan

, Wang

, et al. Serum iron, zinc, and copper levels in patients with Alzheimer’s disease: A replication study and meta-analyses. J Alzheimers Dis 2015; 47: 565–581.

133.

, Zhang

, Wang

, et al. Serum copper, zinc, and iron levels in patients with Alzheimer’s disease: A meta-analysis of case-control studies. Front Aging Neurosci 2017; 9: 261619.

134.

Bousejra-Elgarah

, Bijani

, Coppel

, et al. Iron(II) binding to amyloid-β, the Alzheimer’s peptide. Inorg Chem 2011; 50: 9024–9030.

135.

Mantyh

, Ghilardi

, Rogers

, et al. Aluminum, iron, and zinc ions promote aggregation of physiological concentrations of beta-amyloid peptide. J Neurochem 1993; 61: 1171–1174.

136.

Mukherjee

and Panda

. Contrasting effects of ferric and ferrous ions on oligomerization and droplet formation of tau: Implications in tauopathies and neurodegeneration. ACS Chem Neurosci 2021; 12: 4393–4405.

137.

Cherny

, Legg

, McLean

, et al. Aqueous dissolution of Alzheimer’s disease Abeta amyloid deposits by biometal depletion. J Biol Chem 1999; 274: 23223–23228.

138.

Fonte

, Miklossy

, Atwood

, et al. The severity of cortical Alzheimer’s type changes is positively correlated with increased amyloid-beta Levels: Resolubilization of amyloid-beta with transition metal ion chelators. J Alzheimers Dis 2001; 3: 209–219.

139.

Guo

, Wang

, Zheng

, et al. Intranasal deferoxamine reverses iron-induced memory deficits and inhibits amyloidogenic APP processing in a transgenic mouse model of Alzheimer’s disease. Neurobiol Aging 2013; 34: 562–575.

140.

Fine

, Baillargeon

, Renner

, et al. Intranasal deferoxamine improves performance in radial arm water maze, stabilizes HIF-1α, and phosphorylates GSK3β in P301L tau transgenic mice. Exp Brain Res 2012; 219: 381–390.

141.

Crielaard

, Lammers

and Rivella

. Targeting iron metabolism in drug discovery and delivery. Nat Rev Drug Discov 2017; 16: 400–423.

142.

Kosyakovsky

, Fine

, Frey

, et al. Mechanisms of intranasal deferoxamine in neurodegenerative and neurovascular disease. Pharmaceuticals (Basel) 2021; 14: 95.

143.

Farr

and Xiong

. Challenges and opportunities of deferoxamine delivery for treatment of Alzheimer’s disease, Parkinson’s disease, and intracerebral hemorrhage. Mol Pharm 2021; 18: 593.

144.

Guo

, Yang

, Zhang

, et al. Intranasal lactoferrin enhances α-secretase-dependent amyloid precursor protein processing via the ERK1/2-CREB and HIF-1α pathways in an Alzheimer’s disease mouse model. Neuropsychopharmacology 2017; 42: 2504.

145.

Tsatsanis

, McCorkindale

, Wong

, et al. The acute phase protein lactoferrin is a key feature of Alzheimer’s disease and predictor of Aβ burden through induction of APP amyloidogenic processing. Mol Psychiatry 2021; 26: 5516–5531.

146.

Kim

H soo

, Kim

SMS

, Shin

, et al. Gram-negative bacteria and their lipopolysaccharides in Alzheimer’s disease: Pathologic roles and therapeutic implications. Transl Neurodegener 2021; 10: 1–23.

147.

Huang

, Atwood

, Hartshorn

, et al. The A beta peptide of Alzheimer’s disease directly produces hydrogen peroxide through metal ion reduction. Biochemistry 1999; 38: 7609–7616.

148.

Duce

, Tsatsanis

, Cater

, et al. Iron-export ferroxidase activity of β-amyloid precursor protein is inhibited by zinc in Alzheimer’s disease. Cell 2010; 142: 857–867.

149.

Shippy

, Oliai

and Ulland

. Zinc utilization by microglia in Alzheimer’s disease. J Biol Chem 2024; 300: 107306.

150.

Ritchie

, Bush

, Mackinnon

, et al. Metal-protein attenuation with iodochlorhydroxyquin (clioquinol) targeting Abeta amyloid deposition and toxicity in Alzheimer disease: A pilot phase 2 clinical trial. Arch Neurol 2003; 60: 1685–1691.

151.

Sampson

, Jenagaratnam

and Mcshane

. Metal protein attenuating compounds for the treatment of Alzheimer’s dementia. Cochrane Database Syst Rev 2014: CD005380.

152.

Dobryakova

, Gerasimov

, Spivak

, et al. The induction of long-term potentiation by medial septum activation under urethane anesthesia can alter gene expression in the hippocampus. Int J Mol Sci 2023; 24: 12970.