Introduction
Alzheimer’s disease (AD) is a degenerative neurological condition that
gradually diminishes memory and cognitive functions. It accounts for a significant
portion of dementia cases, ranging from 50% to 75%. In the United
States, roughly 6.5 million individuals aged 65 and above have been diagnosed with
clinical AD in 2021 [1]. Unless medical
advances are made to prevent, delay, or cure AD, this number is projected to rise to
13.8 million by 2060. The total medical expenses for AD and other dementia patients
will escalate from $ 321 billion in 2022 to nearly $1 trillion by
2050. Furthermore, individuals with AD or related dementia conditions are more prone
to developing other chronic ailments, resulting in escalating healthcare costs.
Consequently, it has become a major global health concern [2].
The therapeutic approaches for AD are primarily focused on two signature
abnormalities: extracellular plaque formation of β-amyloid (Aβ) and
intraneuronal neurofibrillary tangles (NFTs) created by hyperphosphorylated tau
[3]. The Aβ peptide, which ranges
from a total of 39 to 43 amino acids in length, is a result of the cleavage of the
amyloid precursor protein (APP). Despite being widely recognized as the primary
culprit for AD genesis, the Aβ peptide also plays a vital role in numerous
physiological processes such as memory consolidation, promotion of neuronal growth,
safeguarding against toxins and pathogens, etc. [4]. In individuals with normal cognitive function, the concentration of
Aβ40 in cerebrospinal fluid (CSF) is approximately 2–3
ng/ml, while that of Aβ42 is around 0.7–0.8 ng/ml.
Elevated production or reduced clearance of Aβ peptides, or a combination of
both, can lead to the accumulation of inflammatory SPs, which can impair cell
signaling pathways, resulting in subsequent synaptic degeneration, neuronal loss,
and cognitive function decline. Oligomeric Aβ has high neurotoxicity, with
Aβ40 and Aβ42 isoforms being the primary culprits that aggregate and
form Aβ plaques in the brain. However, autopsies of healthy individuals who
had no history of cognitive impairment before death also showed the presence of
Aβ plaques in their brains. Therefore, the existence of Aβ plaque
buildup does not necessarily correlate with abnormal clinical findings [5]
[6]
[7]
[8].
Tau, a significant protein involved in Alzheimer’s disease (AD), is a
microtubule-associated protein (MAP) that is soluble in nature. In normal
physiological conditions, tau plays a pivotal role in maintaining the stability of
neuronal microtubules, thereby preserving neuronal morphology and physiology.
Typically, tau is found in axons and is composed of a family of six isomeric forms
ranging from 352 to 441 amino acids in length with sizes ranging from 37 to 46 KD,
which arise from the MAPT gene’s exons 2, 3, and 10 alternative splicing as
shown in [Fig. 1]. Tau is a protein that
lacks a defined structure, which can stabilize microtubules, control their assembly,
and influence the growth and morphology of neurons. When there are pathological
conditions such as AD, PSP, FTD, or other neurodegenerative diseases, tau’s
abnormal assembly gives rise to insoluble aggregates that accompany a range of
neurodegenerative diseases such as synaptic dysfunction and nerve cell death [9]
[10].
During the progression of AD, tau aggregates abnormally to form NFTs that exist in a
characteristic distribution pattern, which allows differentiation into six stages,
Braak grades, including transentorhinal stages I–II (clinically silent
cases), limbic stage III–IV (incipient AD), and neocortical stages
V–VI (fully developed AD). Based on the pathologic features of tau,
prion-like aggregation and transmission of tau have been suggested. Pathologic tau
acts as a template to induce conformational changes in normal tau protein, making it
more likely to aggregate and causing more tau pathology in peripheral regions,
resulting in the spread of tau pathology to wider brain regions. Research has
demonstrated that AD patients have a substantial accumulation of tau protein in
brain tissue compared to healthy control groups [10]
[11]
[12]
[13]
[14]
[15].
Fig. 1 Illustrates a schematic representation of the protein
structures of tau, which consists of six isoforms of 352–441 amino
acids, namely 2N4R, 2N3R, 1N4R, 1N3R, 0N4R, and 0N3R, resulting from
alternative splicing of exon 2 (E2), E3, and E10. Tau has four regions,
namely N-terminus, proline-rich domain, microtubule-binding domain, and
C-terminus.
The phosphorylated isomeric forms of tau help stabilize microtubules, with the 4R tau
isoform promoting microtubule assembly more than the 3R isoform. However, when tau
becomes hyperphosphorylated, it disturbs microtubule assembly and sequesters MAPs
into paired helical filaments (PHFs), which are insoluble entities that damage
cytoplasmic scaffolding and impede axonal transport, ultimately leading to
neurodegeneration. Furthermore, aggregated tau protein forms neurofibrillary tangles
(NFTs), which are a fundamental pathological finding in AD. Autopsy studies of
AD-affected neocortical and CA1 regions reveal a direct correlation between the
location of tau pathology and the severity and progression of the disease [16]
[17]
[18].
Nevertheless, growing evidence suggests that there is a two-way interaction between
Aβ and tau. Elevated CSF tau levels were also observed in APP/PS1
transgenic mice that overexpress Aβ, whereas tau deficiency mitigated
Aβ deposition. The combination of Aβ and tau synergistically impairs
glucose metabolism, induces brain atrophy in AD brain, and results in the
deterioration of cognitive function. Alzheimer’s disease (AD) is a
neurodegenerative illness that is linked to age and classified into two subtypes
based on the onset of symptoms: early-onset AD (EOAD), which occurs before the age
of 65, and late-onset AD (LOAD), which usually occurs after the age of 65. LOAD is
the most prevalent form of AD and is a multifactorial disease caused by genetics,
aging, environmental factors, lifestyle, and chronic illnesses like obesity. EOAD
accounts for only 10% of all AD cases, and a mere 5% of EOAD
patients have a pathogenic variant in the AD genes (APP, PSEN1, and PSEN2) or the
apolipoprotein E (APOE) ε4 allele. The pathogenesis of EOAD remains largely
unknown in most patients [19]
[20]
[21]
[22]
[23]
[24]
[25].
Currently, the etiological theories regarding Alzheimer’s disease mainly
comprise the amyloid cascade theory, tau theory, inflammatory theory, cholinergic
theory, among others. Among them, the amyloid cascade theory is the most widely
accepted. However, the majority of anti-Aβ drugs have not demonstrated
satisfactory therapeutic efficacy over the past twenty years. It is a promising
development that aducanumab, a humanized recombinant monoclonal antibody that
targets Aβ, has become the first disease-modifying therapy (DMT) drug for AD
in 2021, approved by an expedited pathway of the US Food and Drug Administration
(FDA). Nonetheless, the approval, which was based on reduced amyloid markers and its
clinical efficacy, has sparked controversy. In January 2023, another monoclonal
antibody against Aβ called lecanemab was approved by the FDA for the
treatment of early AD, also through its accelerated approval pathway [26]
[27]
[28].
Lecanemab exhibited significant reduction in brain amyloid accumulation in
early-stage Alzheimer’s disease (AD) and moderate cognitive decline compared
to placebo over an 18-month period, albeit with the occurrence of adverse effects
[29]. Further verification is essential to
establish the effectiveness and long-term safety of these drugs. Additionally, it is
imperative to reassess the amyloid cascade hypothesis and explore alternative
targets such as anti-tau drug development for AD, given mounting evidence that
cognitive dysfunction and disease severity are more closely linked to tau pathology
[30]. Inflammatory responses, particularly
neuroinflammation, are implicated in the entire AD pathogenesis process and also
associated with cognitive dysfunction. Hence, this review concentrates on tau
pathology’s pathological changes in AD progression and recent research on
the interdependent regulation and influence of neuroinflammation and tau pathology.
The current clinical drug development centered on the tau and inflammation
hypotheses is also analyzed. This review offers novel insights into AD pathogenesis
and drug treatment strategies [31]
[32].
Pathological state of tau
In 1975, the discovery of Tau marked a significant milestone. This
microtubule-associated protein is highly expressed and soluble in neurons across
the central nervous system (CNS). It is predominantly present in the axons of
neurons and plays a pivotal role in binding to microtubules, thereby enhancing
assembly and regulating stability. This function is critical in neurite
outgrowth, cell shape and polarity, and intracellular cargo transport, including
neurotransmitters. The MAPT gene on chromosome 17q21 encodes human tau,
comprising 16 exons [33]
[34]
[35]. Alternative splicing of exons 2 and 3 of the tau gene in the
human brain results in three isoforms with 0, 1, or 2 N-terminal repeats (0N,
1N, 2N). The absence or presence of exon 10 produces tau species with either
three (3R) or four (4R) carboxyl-terminal microtubule-binding domain. Thus,
within the human brain, six tau isoforms of varying lengths ranging from 352 to
441 amino acids are expressed, with their expression being developmentally
regulated. In the adult brain, all six isoforms are present with roughly
equivalent 4R and 3R isoforms, while the fetal brain expresses solely 0N3R tau.
The 4R tau isoforms demonstrate greater affinity when binding to microtubules
compared to the 3R isoforms. Furthermore, research has demonstrated that certain
mutations within the tau gene affect the alternative splicing of exon 10,
leading to an altered 4R:3R ratio, a critical element of primary tauopathies.
These disorders belong to a subgroup of FTLD illnesses characterized by neuronal
and glial tau inclusions, with predominant atrophy in the frontal and temporal
lobes [36]
[37]
[38].
Pathological tau in AD undergoes severe post-translational modifications (PTMs),
which initiate prior to the formation of neurofibrillary tangles (NFTs), even
decades before the manifestation of symptoms. It is crucial to note that not all
PTMs lead to pathological consequences, as the location, type, and extent of
modification play a decisive role. Hyperphosphorylation, the most extensively
studied PTM of tau, was one of the earliest modifications identified. Previous
investigations have demonstrated that aggregated tau, isolated from AD brains,
has an overall phosphorylation level that is 3–4 times higher than
healthy controls (2–3 mol per protein) [39]. Hyperphosphorylated tau leads to reduced binding ability with
microtubules, as well as decreased stability of microtubules. The dissociation
of tau leads to self-aggregation, resulting in the formation of oligomers and
NFTs. This process disrupts the normal structure of cells and inhibits
intracellular material exchange, leading to aggregation and ultimately,
neurodegeneration. Furthermore, acetylation and N-glycosylation have been found
to increase in AD, with acetylation inhibiting tau protein degradation and
promoting pathological tau aggregation and propagation [40]
[41]
[42]
[43]
[44]
[45]. N-Glycosylation, on
the other hand, stimulates tau polymerization by promoting phosphorylation and
conformational changes. Additionally, nitration of the Tyr29 site of tau, found
only in AD patients, significantly affects tau polymerization and facilitates
tau aggregation as shown in [Fig. 2].
Fig. 2 Illustrates the therapeutic mechanism of targeting tau
protein.
The cleavage modification of caspase in AD patient brain also plays a key role.
Caspases 2 and 3 cleave the C-terminus at Asp314 and Asp421, respectively, while
caspase 6 cleaves the N-terminus. In addition to affecting microtubule binding
and aggregation, cleaving the C-terminal of tau has been linked to mitochondrial
and synaptic damage [46]
[47]
[48]
[49]
[50]. However, it is important to note that
not all post-translational modifications (PTMs) contribute to the acceleration
of tau pathology and Alzheimer’s disease (AD) progression. In fact,
ubiquitination has been found to aid in the degradation of tau through autophagy
and proteasome, while also inhibiting tau aggregation and maintaining
microtubule stability. Additionally, lysine methylation is a normal PTM observed
in tau in the human brain and can help to protect against pathological tau
aggregation to some extent as shown in [Table
1]. Another tau PTM, dityrosine (DiY) cross-links, is induced by
oxidative stress and has been observed on human AD-derived tau oligomers and
PHFs. Interestingly, DiY cross-links can facilitate the formation of non-toxic,
soluble tau oligomers, inhibit the formation of β-sheets and further
extension of prefibrils, and increase the insolubility and stability of tau
fibrils in AD [51]
[52]
[53]
[54].
Table 1 The investigation of the modalities and
advancements of drugs that are aimed at tau protein therapy is being
conducted [55].
Entry
|
Drug
|
Targeting Site
|
Action
|
1
|
LMTM
|
Alzheimer’s Disease (phase 3)
|
Aggregation inhibitors
|
2
|
Lithium
|
MCI (phase 4), expected to be completed by 2023
|
GSK-3β inhibitor
|
3
|
LY3372689
|
Alzheimer’s Disease (phase 2)
|
O-GlcNAcase inhibitor
|
4
|
Tideglusib
|
Alzheimer’s Disease (discontinued)
|
GSK-3β inhibitor
|
5
|
Curcumin
|
Alzheimer’s Disease (phase 2)
|
Aggregation inhibitors
|
6
|
Epothilone
|
Alzheimer’s Disease (discontinued)
|
Microtubule stabilizers
|
7
|
Gosuranemab
|
Alzheimer’s Disease (discontinued)
|
A humanized IgG4 monoclonal antibody targeting the N-terminal
region of tau protein
|
8
|
Bepranemab
|
Alzheimer’s Disease (phase2), will run until November
2025.
|
A humanized IgG4 monoclonal antibody targeting tau
235–250 near the MTBR of tau
|
9
|
Tilavonema
|
Alzheimer’s Disease (discontinued)
|
A humanized IgG4 antibody recognizes 25–30 aa of tau
and targets extracellular tau
|
10
|
Sodium selenate
|
Alzheimer’s Disease (phase 2)
|
PP2A activators
|
Phosphorylation state inhibitor
The tau protein comprises multiple phosphorylation sites that are integral to its
ability to bind to microtubules. The degree of phosphorylation determines the
protein’s affinity for microtubules, with hyperphosphorylated tau
exhibiting decreased affinity and even a loss of the ability to bind altogether
[55]
[56]. This results in a compromised cytoskeleton, which could lead to
neuronal death if hyperphosphorylated tau accumulates in neurons to form PHFs.
Maintaining normal neuronal physiological functions necessitates inhibiting tau
hyperphosphorylation. The phosphorylation level of tau is governed by a balance
of protein kinases and phosphatases, and any disruption to this balance is
linked to abnormal tau phosphorylation observed in AD, particularly the
hyperactivation of glycogen synthase kinase 3β (GSK-3β) and
cyclin-dependent kinase 5 (CDK5) and the inhibition of protein phosphatase-2A
(PP2A). Hampering tau pathology and AD therapy may be achieved by inhibiting
protein kinase activity and increasing phosphatase activity [57].
GSK-3β inhibitors
Tideglusib (also known as NP031112, Nypta, Zentylor, NP12) is a GSK-3β
irreversible inhibitor that is widely examined as a tau kinase inhibitor. Its
inhibition effectively counteracts tau hyperphosphorylation. Mouse entorhinal
cortex and hippocampus preclinical studies have indicated that tideglusib can
reduce a range of disease outcomes including tau phosphorylation, Aβ
deposition, neuron loss, and gliosis, and reverse spatial memory impairment in
transgenic mice. Furthermore, tideglusib’s neuroprotective,
anti-inflammatory, and neurogenesis-inducing effects have also been validated in
animal models [58]. In a phase 2 clinical
trial with mild to moderate AD patients, tideglusib demonstrated an acceptable
safety profile for a short-term (26 weeks) treatment, with the exception of a
transient increase in serum transaminase levels and diarrhea
(14–18% in active, 11% placebo). However, it was not
successful in slowing cognitive dysfunction, and only a small number of patients
showed significant reductions of β-secretase in CSF. At present, the
study of tideglusib in AD has been terminated. Lithium has been widely used as a
medication for bipolar disorder and has been examined for its ability to inhibit
GSK-3β in the treatment of AD [59].
While clinical trials have not shown significant improvements in cognitive scores
or CSF-based biomarker concentrations after 10 weeks of lithium treatment for
mild or mild-to-moderate AD, a 12-month double-blind trial for patients with
amnestic mild cognitive impairment (MCI) has demonstrated promising results. The
group treated with lithium showed decreased levels of phosphorylated tau in CSF
and improved cognitive ability compared to the control group receiving a
placebo. Furthermore, the use of lithium was deemed safe and well-tolerated at a
serum concentration range of 0.25–0.5 mmol/l. In patients
diagnosed with AD, the effectiveness of microdose lithium (300
μg/day) in preventing cognitive loss was demonstrated during a
15-month study. Previous studies had failed to discover a significant impact of
lithium on cognition or AD-related biomarkers, potentially due to the high
incidence of side effects such as renal and neurologic dysfunction, endocrine
abnormalities, and high withdrawal rates. Furthermore, shorter follow-up periods
and the inclusion of patients with cognitive degradation related to the later
stage may have contributed to these findings. Overall, clinical evidence
suggests that long-term microdose lithium therapy is necessary for a positive
outcome, and the use of microdose lithium could potentially solve the toxicity
issues that have hindered medication development. Currently, there is a phase 4
clinical trial investigating the use of lithium in patients with MCI, which is
set to conclude in 2023 [60].
CDK5 inhibitors
CDK5 is a protein kinase that is activated through interaction with its
activators, namely p35 and p39, as well as their cleavage products, p25 and p29,
respectively. This activation governs several cellular processes in neurons. The
dysregulation of CDK5 is primarily attributed to the formation and accumulation
of neurotoxic p25 and p29. The resultant hyperactivation and mislocalization of
CDK5 lead to the hyperphosphorylation of its substrates. CDK5 phosphorylates
Tau, which is a crucial substrate of CDK5, and causes it to disassociate from
microtubules, self-aggregate, and eventually form PHFs and insoluble NFTs in
neurons[61]
[62]
[63]
[64]
[65].
Studies have demonstrated that knocking down CDK5 in triple transgenic (3xTg) AD
mice, as well as inhibiting CDK5 with CDK5 inhibitory peptide in vitro, can
reduce tau hyperphosphorylation and NFT formation. Minocycline has been found to
alleviate Alzheimer’s disease (AD)-like symptoms and improve cognitive
impairment by inhibiting the CDK5/p25 signaling pathway. During the
advanced stages of the disease, CDK5 primes phosphorylation sites on tau for
GSK-3β, which then synergistically promotes the GSK-3β-mediated
tau hyperphosphorylation. However, a previous study has reported that the
administration of CP681301, a CDK5 inhibitor, actually enhanced tau
phosphorylation in p25-overexpressing transgenic mice, contradicting the
conventional understanding of the synergistic relationship between CDK5 and
GSK-3β. CDK5 can indirectly phosphorylate GSK-3β at S9 and
inhibit its activity, suggesting that CDK5 inhibition may enhance tau
phosphorylation through the activation of GSK3β. Despite contradictory
reports regarding the effect of CDK5 on tau phosphorylation, an equally
significant concern is the lack of tau-specificity exhibited by CDK5 inhibitors.
Given that CDK5 can phosphorylate molecules other than tau, greater attention
should be devoted to the development of tau-targeting agents with CDK5
inhibitory activity. As of yet, no successful small molecule candidates have
been identified as selective inhibitors of CDK5 kinase activity [66]
[67]
[68].
Phosphatase activator
The major constituents of phosphatases in the human brain are PP2A, PP5, PP1, and
PP2B, comprising 71%, 11%, 10%, and 7% of the
total tau dephosphorylation activity, respectively. In individuals with
Alzheimer’s disease (AD), there is a noticeable decrease in the total
phosphatase activity, PP2A activity, and PP5 activity towards tau, however, an
increase in PP2B activity is also observed. The downregulation of PP2A activity
may lead to abnormal hyperphosphorylation of tau, which can result in a
50% reduction in its activity in AD brain, consistent with the level of
tau phosphorylation. Conversely, PP1 and PP5 activity is much less
downregulated. In vitro, co-incubation of tau aggregates with PP2A has been
found to restore the binding of tau to microtubules to a level similar to the
control group (approximately 80%). Therefore, PP2A is the primary tau
phosphatase, and its activators have promising potential for AD treatment [69]
[70].